Atoll 3.3.0 Technical Reference Guide.pdf

Version 3.3.0 Technical Reference Guide for Radio Networks AT330_TRR_E1 AT330_TRR_E1 Atoll 3.3.0 Technical Reference Guide for Radio Networks Release: AT330_TRR_E1 (March 2015) © Copyright 1997-2015 Forsk. All Rights Reserved. Published by: Forsk 7 rue des Briquetiers 31700 Blagnac, France Tel: +33 562 747 210 Fax: +33 562 747 211 The software described in this document is provided under a licence agreement. The software may only be used or copied under the terms and conditions of the licence agreement. No part of the contents of this document may be reproduced or transmitted in any form or by any means without written permission from the publisher. The product or brand names mentioned in this document are trademarks or registered trademarks of their respective registering parties. The publisher has taken care in the preparation of this document, but makes no expressed or implied warranty of any kind and assumes no responsibility for errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of the use of the information contained herein. Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter : Introduction AT330_TRR_E1 Introduction This Technical Reference Guide is aimed at radio network engineers with an advanced knowledge of Atoll and radio network planning. It provides detailed information about the inner workings and formulas that are implemented by Atoll. About Atoll Atoll is a 64-bit multi-technology wireless network design and optimisation platform. Atoll is open, scalable, flexible, and supports wireless operators throughout the network life cycle, from initial design to densification and optimisation. Atoll’s integration and automation features help operators smoothly automate planning and optimisation processes through flexible scripting and SOA-based mechanisms. Atoll supports a wide range of implementation scenarios, from standalone to enterprise-wide server-based configurations using distributed and multi-threaded computing. Atoll Microwave is a complete backhaul and microwave link planning solution based on the leading Atoll platform, which includes a high performance GIS and advanced data and user management features. Atoll Microwave can share its site database with Atoll radio planning and optimisation modules, thus allowing easy data consistency management across the operator organisation. If you are interested in learning more about Atoll, please contact your Forsk representative to inquire about our training solutions. About Forsk Forsk is an independent company providing radio planning and optimisation software solutions to the wireless industry since 1987. In 1997, Forsk released the first version of Atoll, its flagship radio planning software. Since then, Atoll has evolved to become a comprehensive radio planning and optimisation platform and, with more than 5000 installed licenses worldwide, has reached the leading position on the global market. Atoll combines engineering and automation functions that enable operators to smoothly and gradually implement SON processes within their organisation. Today, Forsk is a global supplier with over 300 customers in 100 countries and strategic partnerships with major players in the industry. Forsk distributes and supports Atoll directly from offices and technical support centres in France, USA, and China as well as through a worldwide network of distributors and partners. Since the first release of Atoll, Forsk has been known for its capability to deliver tailored and turn-key radio planning and optimisation environments based on Atoll. To help operators streamline their radio planning and optimisation processes, Forsk provides a complete range of implementation services, including integration with existing IT infrastructure, automation, as well as data migration, installation, and training services. Getting Help The online help system that is installed with Atoll is designed to give you quick access to the information you need to use the product effectively. It contains the same material as the Atoll 3.3.0 User Manual. You can browse the online help from the Contents view, the Index view, or you can use the built-in Search feature. You can also download manuals from the Forsk web site. Printing Help Topics You can print individual topics or chapters from the online help. To print help topics or chapters: 1. In Atoll, click Help > Help Topics. 2. In the Contents tab, expand the table of contents. 3. Right-click the section or topic that you want to print and click Print. The Print Topics dialog box appears. 4. In the Print Topics dialog box, select what you want to print: • • If you want to print a single topic, select Print the selected topic. If you want to print an entire section, including all topics and sections in that section, select Print the selected heading and all subtopics. 5. Click OK. 3 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter : Introduction © 2015 Forsk About Atoll Documentation The following PDF manuals are available for Atoll and Atoll Microwave and can be downloaded from the Forsk web site at: http://www.forsk.com/support. • • • • • • Atoll 3.3.0 User Manual Atoll 3.3.0 Administrator Manual Atoll 3.3.0 Data Structure Reference Guide Atoll 3.3.0 Technical Reference Guide Atoll 3.3.0 Task Automation Guide Atoll 3.3.0 Model Calibration Guide To read PDF manuals, you can download Adobe Reader from the Adobe web site at: http://get.adobe.com/reader/ Hardcopy manuals are also available. For more information, contact to your Forsk representative. Contacting Technical Support Forsk provides global technical support for its products and services. To contact the Forsk support team, visit the Forsk Support web site at: http://www.forsk.com/support. Alternatively, depending on your geographic location, contact one of the following support teams: • Forsk Head Office For regions other than North and Central America and China, contact the Forsk Head Office support team: • • • Tel.: +33 562 747 225 Fax: +33 562 747 211 Email: [email protected] Opening Hours: Monday to Friday 9.00 am to 6.00 pm (GMT +1:00) • Forsk US For North and Central America, contact the Forsk US support team: • • • Tel.: 1-888-GO-ATOLL (1-888-462-8655) Fax: 1-312-674-4822 Email: [email protected] Opening Hours: Monday to Friday 8.00 am to 8.00 pm (Eastern Standard Time) • Forsk China For China, contact the Forsk China support team: • • • Tel: +86 20 8557 0016 Fax: +86 20 8553 8285 Email: [email protected] Opening Hours: Monday to Friday 9.00am to 5.30pm (GMT+08:00) Beijing, Chongqing, Hong Kong, Urumqi. 4 AT330_TRR_E1 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents Table of Contents Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 1 Antennas and Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 1.1 1.1.1 1.1.2 1.1.3 Antenna Attenuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Calculation of Azimuth and Tilt Angles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Antenna Pattern 3D Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Additional Electrical Downtilt Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 1.2 Antenna Pattern Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 1.3 Power Received From Secondary Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 1.4 1.4.1 1.4.2 Transmitter Radio Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 GSM Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 1.5 1.5.1 1.5.1.1 1.5.1.2 1.5.1.3 1.5.1.4 1.5.2 1.5.2.1 1.5.2.2 1.5.3 1.5.3.1 1.5.3.2 Repeaters and Remote Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Total Gain Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Repeater Noise Figure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Appendix: Carrier Power and Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 GSM Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 EIRP Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Donor-side Parameter Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Azimuth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Mechanical Downtilt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 1.6 1.6.1 1.6.1.1 1.6.1.2 1.6.1.3 1.6.1.4 1.6.2 1.6.3 1.6.4 Beamforming Smart Antenna Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Downlink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Uplink Beamforming. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Uplink Beamforming and Interference Cancellation (MMSE). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Downlink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Uplink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Uplink Beamforming and Interference Cancellation (MMSE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 1.7 Grid-of-Beams Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 1.8 Adaptive Beam Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 1.9 Statistical Smart Antenna Gain Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 2 Radio Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57 2.1 2.1.1 2.1.2 2.1.2.1 2.1.2.2 2.1.3 2.1.4 Path Loss Calculation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Ground Altitude Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Clutter Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Clutter Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Clutter Heights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Geographic Profile Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Resolution of the Extracted Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 2.2 List of Default Propagation Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 2.3 2.3.1 2.3.2 Okumura-Hata and Cost-Hata Propagation Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Hata Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Corrections to the Hata Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 5 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 6 © Forsk 2015 2.3.3 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 2.4 2.4.1 2.4.2 2.4.3 ITU 529-3 Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 ITU 529-3 Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 Corrections to the ITU 529-3 Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65 2.5 2.5.1 2.5.2 2.5.2.1 2.5.2.2 2.5.2.3 2.5.2.4 2.5.2.5 2.5.2.6 2.5.2.7 2.5.3 2.5.3.1 2.5.3.2 2.5.4 Standard Propagation Model (SPM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65 SPM Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66 Visibility and Distance Between Transmitter and Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66 Effective Transmitter Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66 Effective Receiver Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69 Correction for Hilly Regions in Case of LOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70 Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70 Losses due to Clutter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70 Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Automatic Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 General Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 Sample Values for SPM Path Loss Formula Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73 Unmasked Path Loss Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74 2.6 2.6.1 2.6.2 WLL Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 WLL Path Loss Formula. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 2.7 2.7.1 2.7.2 ITU-R P.526-5 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 ITU 526-5 Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 2.8 2.8.1 2.8.2 ITU-R P.370-7 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 ITU 370-7 Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 2.9 2.9.1 2.9.2 2.9.3 Erceg-Greenstein (SUI) Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78 SUI Terrain Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78 Erceg-Greenstein (SUI) Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79 2.10 2.10.1 2.10.1.1 2.10.1.2 2.10.1.3 2.10.1.4 2.10.1.5 2.10.1.6 ITU-R P.1546-2 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80 Step 1: Determination of Graphs to be Used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80 Step 2: Calculation of Maximum Field Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81 Step 3: Determination of Transmitter Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81 Step 4: Interpolation/Extrapolation of Field Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81 Step 5: Calculation of Correction Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Step 6: Calculation of Path Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84 2.11 Sakagami Extended Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84 2.12 Free Space Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86 2.13 2.13.1 2.13.2 2.13.3 2.13.4 2.13.5 Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86 Knife-edge Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86 3 Knife-edge Deygout Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87 Epstein-Peterson Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 Deygout Method with Correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 Millington Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 2.14 2.14.1 2.14.1.1 2.14.1.2 2.14.2 2.14.2.1 2.14.2.1.1 2.14.2.1.2 2.14.2.2 2.14.2.2.1 2.14.2.2.2 Shadow Fading Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 Shadowing Margin Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95 Shadowing Margin Calculation in Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95 Shadowing Margin Calculation in Monte-Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 Macro-Diversity Gains Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 Uplink Macro-Diversity Gain Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 Shadowing Error PDF (n Signals). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 Uplink Macro-Diversity Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Downlink Macro-Diversity Gain Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Shadowing Error PDF (n Signals). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Downlink Macro-Diversity Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 AT330_TRR_E1 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 2.15 2.15.1 2.15.2 2.15.3 2.15.3.1 2.15.3.2 Path Loss Matrices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Calculation Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Validity of Path Loss Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Path Loss Tuning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Transmitter Path Loss Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Repeater Path Loss Tuning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 2.16 2.16.1 2.16.1.1 2.16.1.2 2.16.1.3 2.16.2 2.16.2.1 2.16.2.2 2.16.2.3 2.16.3 2.16.3.1 2.16.3.2 2.16.3.3 2.16.3.4 2.16.3.5 2.16.3.6 2.16.4 2.16.4.1 2.16.4.2 2.16.5 2.16.5.1 2.16.5.2 2.16.5.3 2.16.5.4 File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Path Loss Matrix File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Pathloss.dbf File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Pathloss.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 LOS File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Path Loss Tuning File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Pathloss.dbf File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Pathloss.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 PTS File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Interference Matrix File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 CLC Format (One Value per Line). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 CLC File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 DCT File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 IM0 Format (One Histogram per Line) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 IM1 Format (One Value per Line, TX Name Repeated) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 IM2 Format (Co- and Adjacent-channel Probabilities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 "Per Transmitter" Prediction File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 <per_transmitter_prediction>.dbf File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 <per_transmitter_prediction>.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Coverage Prediction Export and Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Filtering Coverage Predictions at Export. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Smoothing Coverage Predictions at Export . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Examples of Prediction Export Filtering and Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Coverage Prediction Reports Over Focus/Computation Zones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 3 GSM GPRS EDGE Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 3.1 3.1.1 3.1.2 3.1.3 3.1.3.1 3.1.3.2 3.1.4 3.1.4.1 3.1.4.1.1 3.1.4.1.2 3.1.4.1.3 3.1.4.1.4 3.1.4.1.5 3.1.4.1.6 3.1.4.1.7 3.1.4.1.8 3.1.4.2 3.1.4.2.1 3.1.4.2.2 Signal Level Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 DL Signal Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 UL Signal Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Signal Level-based DL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 DL Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Best Signal Level and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Second Best Signal Level per HCS Layer and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 HCS Servers and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Highest Priority HCS Server and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Best Idle Mode Reselection Criterion (C2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Coverage Display. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 3.2 3.2.1 3.2.2 3.2.3 3.2.3.1 3.2.3.2 3.2.3.2.1 3.2.3.2.2 3.2.3.3 3.2.3.3.1 3.2.3.3.2 Interference-based DL Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 DL Carrier-to-Interference Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Interference-based DL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Service Area Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Coverage Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Interference Condition Satisfied by At Least One TRX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Interference Condition Satisfied by The Worst TRX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Coverage Display. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 3.3 3.3.1 GPRS/EDGE Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 Coding Scheme Selection and Throughput Calculation Without Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 7 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 8 © Forsk 2015 3.3.1.1 3.3.1.2 3.3.1.3 3.3.2 3.3.2.1 3.3.2.2 3.3.2.3 3.3.3 3.3.4 3.3.5 3.3.5.1 3.3.5.1.1 3.3.5.1.2 3.3.5.1.3 3.3.5.1.4 3.3.5.1.5 3.3.5.1.6 3.3.5.1.7 3.3.5.1.8 3.3.5.2 3.3.5.2.1 3.3.5.2.2 Calculations Based on C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C/(I+N) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coding Scheme Selection and Throughput Calculation With Ideal Link Adaptation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C/(I+N) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Application Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . BLER Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GPRS/EDGE Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Signal Level per HCS Layer and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Second Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HCS Servers and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Highest Priority HCS Server and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Idle Mode Reselection Criterion (C2). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 137 137 138 138 138 139 139 140 140 140 140 140 141 141 141 141 142 142 142 142 143 3.4 3.4.1 3.4.2 3.4.2.1 3.4.2.2 3.4.3 3.4.3.1 3.4.3.2 3.4.4 3.4.4.1 3.4.4.1.1 3.4.4.1.2 3.4.4.1.3 3.4.4.1.4 3.4.4.1.5 3.4.4.1.6 3.4.4.1.7 3.4.4.2 3.4.4.2.1 3.4.4.2.2 Codec Mode Selection and CQI Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Circuit Quality Indicator Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CQI Calculation Without Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C/N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C/(I+N) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CQI Calculation With Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C/N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations Based on C/(I+N) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Circuit Quality Indicators Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Signal Level per HCS Layer and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Second Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HCS Servers and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Highest Priority HCS Server and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 147 147 147 148 148 148 148 149 149 149 149 150 150 150 150 151 151 151 151 3.5 3.5.1 3.5.1.1 3.5.1.2 3.5.1.3 3.5.1.4 3.5.1.5 3.5.1.6 3.5.1.7 3.5.1.8 3.5.2 3.5.2.1 3.5.2.2 3.5.2.2.1 3.5.2.2.2 3.5.2.2.3 3.5.2.2.4 3.5.3 3.5.3.1 3.5.3.2 3.5.3.3 3.5.3.4 3.5.3.4.1 UL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DL Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . All Servers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Second Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HCS Servers and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Highest Priority HCS Server and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Idle Mode Reselection Criterion (C2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage by UL Signal Level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UL Signal Level (in dBm, dBµV, dBµV/m) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best UL Signal Level (in dBm, dBµV, dBµV/m). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UL Total Losses (dB). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Minimum UL Total Losses (dB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage by UL C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UL C/I Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 152 152 152 153 153 153 154 154 155 155 155 155 155 155 156 156 156 156 156 156 156 156 AT330_TRR_E1 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 3.5.3.4.2 3.5.3.4.3 3.5.4 3.5.4.1 3.5.4.2 3.5.4.3 3.5.5 3.5.5.1 3.5.5.2 Max C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Min C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Coverage by UL Coding Schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Service Area Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Coding Scheme Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Coverage by UL Codec Modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Service Area Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Codec Mode Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 3.6 3.6.1 3.6.1.1 3.6.1.1.1 3.6.1.1.2 3.6.1.2 3.6.1.2.1 3.6.1.2.2 3.6.1.3 3.6.1.3.1 3.6.1.3.2 3.6.2 3.6.2.1 3.6.2.1.1 3.6.2.1.2 3.6.2.1.3 3.6.2.2 3.6.2.2.1 3.6.2.2.2 3.6.2.2.3 Traffic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Traffic Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Circuit Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Packet Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Concentric Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Circuit Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Packet Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Circuit Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Packet Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Calculation of the Traffic Demand per Subcell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Concentric Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Concentric Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 3.7 3.7.1 3.7.1.1 3.7.1.2 3.7.1.2.1 3.7.1.2.2 3.7.1.2.3 3.7.2 3.7.2.1 3.7.2.1.1 3.7.2.1.2 3.7.2.2 3.7.2.2.1 3.7.2.2.2 3.7.2.2.3 3.7.2.2.4 3.7.2.2.5 3.7.2.2.6 Network Dimensioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Dimensioning Models and Quality Graphs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Circuit Switched Traffic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Blocking Probability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Network Dimensioning Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Network Dimensioning Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Network Dimensioning Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Step 1: Timeslots Required for CS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Step 2: TRXs Required for CS Traffic and Dedicated PS Timeslots . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Step 3: Effective CS Blocking, Effective CS Traffic Overflow and Served CS Traffic . . . . . . . . . . . . . . . 178 Step 4: TRXs to Add for PS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Step 5: Served PS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Step 6: Total Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 3.8 3.8.1 3.8.1.1 3.8.1.2 3.8.1.3 3.8.2 3.8.2.1 3.8.2.1.1 3.8.2.1.2 3.8.2.1.3 3.8.2.1.4 3.8.2.1.5 3.8.2.1.6 3.8.2.2 3.8.2.2.1 3.8.2.2.2 3.8.2.2.3 3.8.2.2.4 Key Performance Indicators Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Circuit Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Erlang B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Erlang C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Served Circuit Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Case 1: Total Traffic Demand > Dedicated + Shared Timeslots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Packet Switched Traffic Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Throughput Reduction Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Blocking Probability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Served Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Case 2: Total Traffic Demand < Dedicated + Shared Timeslots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Packet Switched Traffic Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Throughput Reduction Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 9 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 3.8.2.2.5 3.8.2.2.6 © Forsk 2015 Blocking Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Served Packet Switched Traffic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 3.9 3.9.1 3.9.1.1 3.9.1.2 3.9.1.3 3.9.1.4 3.9.1.5 3.9.1.6 3.9.1.7 3.9.1.8 3.9.1.9 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radio Resource Management in GSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GSM Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Servers Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Codec Mode Assignment and DL Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coding Scheme Assignment, Throughput Evaluation and DL Power Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subcell Traffic Loads Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Half-Rate Traffic Ratio Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DL Power Control Gain Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DTX DL Gain Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GSM Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 184 184 187 187 188 189 189 189 190 190 3.10 3.10.1 3.10.2 3.10.3 3.10.4 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neighbour Allocation for a Group of Transmitters or One Transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neighbour Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix: Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 191 194 194 195 3.11 3.11.1 3.11.1.1 3.11.1.2 3.11.1.2.1 3.11.1.2.2 3.11.1.2.3 3.11.2 3.11.2.1 3.11.2.2 3.11.2.3 3.11.3 3.11.3.1 3.11.3.2 3.11.3.3 3.11.3.4 3.11.3.4.1 3.11.3.4.2 3.11.3.4.3 3.11.3.5 AFP Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The AFP Cost Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cost Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cost Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Separation Violation Cost Component. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interference Cost Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I_DIV, F_DIV and Other Advanced Cost Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The AFP Blocked Traffic Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation of New Traffic Loads Including Blocked Traffic Loads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Recalculation of CS and PS From Traffic Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Testing the Blocked Cost Using Traffic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Using Interferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cumulative Density Function of C/I Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Precise Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Precise Interference Distribution Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Direct Availability of Precise Interference Distribution to the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . Efficient Calculation and Storage of Interference Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Robustness of the IM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic Load and Interference Information Discrimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 196 197 198 198 200 201 202 203 204 205 205 205 205 206 206 206 206 206 207 4 UMTS HSPA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 4.1 4.1.1 4.1.2 4.1.2.1 4.1.2.2 4.1.3 4.1.3.1 4.1.3.1.1 4.1.3.1.2 4.1.3.1.3 4.1.3.2 4.1.3.2.1 4.1.3.2.2 General Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plot Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 211 211 211 212 212 212 212 212 213 213 213 213 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ec/I0 Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DL Eb/Nt Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UL Eb/Nt Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 214 215 222 223 224 4.3 4.3.1 4.3.1.1 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Generating a Realistic User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Simulations Based on User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 10 AT330_TRR_E1 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 4.3.1.1.1 4.3.1.1.2 4.3.1.2 4.3.1.2.1 4.3.1.2.2 4.3.1.2.3 4.3.2 4.3.2.1 4.3.2.2 4.3.2.3 4.3.2.3.1 4.3.2.3.2 4.3.2.3.3 4.3.2.3.4 4.3.2.3.5 4.3.2.3.6 4.3.2.3.7 4.3.2.4 4.3.2.4.1 4.3.2.4.2 4.3.2.4.3 4.3.2.4.4 4.3.2.5 4.3.3 4.3.3.1 4.3.3.2 4.3.3.2.1 4.3.3.2.2 4.3.3.2.3 4.3.3.2.4 4.3.4 4.3.4.1 4.3.4.2 4.3.4.2.1 4.3.4.2.2 4.3.4.2.3 4.3.4.3 4.3.4.3.1 4.3.4.3.2 4.3.4.4 4.3.4.5 4.3.4.6 Circuit Switched Service (i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 Packet Switched Service (j). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226 Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Throughputs in Uplink and Downlink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Total Number of Users (All Activity Statuses) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 Number of Users per Activity Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 Power Control Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 R99 Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 HSDPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238 HSDPA Power Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238 Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users . . . . . . . . . . . . . . . . . 238 HSDPA Bearer Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Fast Link Adaptation Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 MIMO Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252 Scheduling Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253 Dual-Cell HSDPA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255 HSUPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 Admission Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 HSUPA Bearer Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 Noise Rise Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 Radio Resource Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 Convergence Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 R99 Related Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 HSPA Related Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Statistics Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Mobiles Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270 Cells Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 Sites Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 Admission Control in the R99 Part . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 Resources Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 OVSF Codes Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 Channel Elements Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278 Iub Backhaul Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278 Downlink Load Factor Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 Downlink Load Factor per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 Downlink Load Factor per Mobile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 Uplink Load Factor Due to One User . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 Inter-carrier Power Sharing Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 Best Serving Cell Determination in Monte Carlo Simulations - Old Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 4.4 4.4.1 4.4.2 4.4.2.1 4.4.2.2 4.4.2.3 4.4.2.4 4.4.2.5 4.4.3 4.4.3.1 4.4.3.1.1 4.4.3.1.2 4.4.3.2 4.4.3.2.1 4.4.3.2.2 4.4.3.3 4.4.3.3.1 4.4.3.3.2 4.4.3.4 4.4.3.4.1 4.4.3.4.2 4.4.3.5 4.4.3.5.1 4.4.3.5.2 UMTS HSPA Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284 Best Serving Cell and Active Set Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284 Point Analysis - AS Analysis Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Bar Graph and Pilot Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Downlink R99 Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288 Uplink R99 Sub-Menu. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 HSDPA Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294 HSUPA Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Pilot Quality Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Prediction Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Study Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Downlink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 Prediction Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 Study Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 Uplink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 Prediction Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 Study Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 Downlink Total Noise Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303 Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 HSDPA Prediction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 Prediction Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Study Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 11 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 4.4.3.6 4.4.3.6.1 4.4.3.6.2 4.4.3.6.3 © Forsk 2015 HSUPA Prediction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Prediction Study Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 309 310 310 4.5 4.5.1 4.5.2 4.5.3 4.5.3.1 4.5.3.2 4.5.4 4.5.4.1 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neighbour Allocation for a Group of Transmitters or One Transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Importance of Intra-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Importance of Inter-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 312 316 316 316 318 319 319 4.6 4.6.1 4.6.1.1 4.6.1.2 4.6.1.2.1 4.6.1.2.2 4.6.1.3 4.6.1.3.1 4.6.1.3.2 4.6.1.3.3 4.6.2 4.6.2.1 4.6.2.1.1 4.6.2.1.2 4.6.2.1.3 4.6.2.1.4 4.6.2.2 Primary Scrambling Code Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Options and Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multi-Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transmitter Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Site Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allocation Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allocation Strategies and Use a Maximum of Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy: Clustered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy: Distributed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy: ‘One Cluster per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy: ‘Distributed per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allocate Carriers Identically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 320 320 321 321 322 323 323 325 325 326 326 326 327 328 328 329 4.7 4.7.1 4.7.2 4.7.2.1 4.7.2.2 4.7.2.3 4.7.2.3.1 Automatic GSM-UMTS Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Algorithm Based on Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 329 330 330 331 333 333 5 CDMA2000 Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 5.1 5.1.1 5.1.2 5.1.2.1 5.1.2.2 5.1.3 5.1.3.1 5.1.3.1.1 5.1.3.1.2 5.1.3.1.3 5.1.3.2 5.1.3.2.1 5.1.3.2.2 General Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plot Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 337 338 338 338 338 338 338 338 339 339 339 339 5.2 5.2.1 5.2.1.1 5.2.1.2 5.2.1.3 5.2.1.4 5.2.1.5 5.2.2 5.2.2.1 5.2.2.2 5.2.2.3 5.2.2.4 Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameters Used for CDMA2000 1xRTT Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ec/I0 Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameters Used for CDMA2000 1xEV-DO Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ec/I0 and Ec/Nt Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340 340 340 345 346 347 348 350 350 354 355 356 12 AT330_TRR_E1 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 5.3 Active Set Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358 5.4 5.4.1 5.4.1.1 5.4.1.1.1 5.4.1.1.2 5.4.1.2 5.4.1.3 5.4.2 5.4.2.1 5.4.2.1.1 5.4.2.1.2 5.4.2.1.3 5.4.2.2 5.4.2.2.1 5.4.2.2.2 5.4.2.2.3 5.4.3 5.4.3.1 5.4.3.2 5.4.3.2.1 5.4.3.2.2 5.4.3.3 5.4.3.3.1 5.4.3.3.2 5.4.3.4 5.4.3.5 Simulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358 Generating a Realistic User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359 Number of Users, User Activity Status and User Throughput. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359 Simulations Based on User Profile Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359 Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362 Transition Flags for 1xEV-DO Rev.0 User Throughputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 User Geographical Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 Network Regulation Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 CDMA2000 1xRTT Power Control Simulation Algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 Presentation of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 Convergence Criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376 CDMA2000 1xEV-DO Power/Data Rate Control Simulation Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 Presentation of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 Convergence Criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386 Admission Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386 Resources Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386 Walsh Code Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386 Channel Element Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 Downlink Load Factor Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 Downlink Load Factor per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 Downlink Load Factor per Mobile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Best Server Determination in Monte Carlo Simulations - Old Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Radio Bearer Allocation Algorithm for Multi-carrier EVDO Rev.B - Old Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 5.5 5.5.1 5.5.1.1 5.5.1.2 5.5.1.2.1 5.5.1.2.2 5.5.1.3 5.5.1.3.1 5.5.1.3.2 5.5.2 5.5.2.1 5.5.2.2 5.5.2.2.1 5.5.2.2.2 5.5.2.3 5.5.2.3.1 5.5.2.3.2 5.5.2.4 5.5.2.4.1 5.5.2.4.2 CDMA2000 Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Point Analysis: The AS Analysis Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Bar Graph and Pilot Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Downlink Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400 Uplink Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402 CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402 CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406 Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 Pilot Quality Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 Downlink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413 CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413 CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415 Uplink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417 CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417 CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419 Downlink Total Noise Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422 Analysis on the Best Carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422 Analysis on a Specific Carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423 5.6 5.6.1 5.6.2 5.6.3 5.6.3.1 5.6.3.2 5.6.4 5.6.4.1 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423 Neighbour Allocation for all Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424 Neighbour Allocation for a Group of Transmitters or One Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427 Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428 Importance of Intra-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428 Importance of Inter-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 5.7 5.7.1 5.7.1.1 5.7.1.2 5.7.1.2.1 5.7.1.2.2 5.7.1.2.3 5.7.1.3 5.7.1.3.1 PN Offset Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431 Options and Constraints. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431 Allocation Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432 Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432 Multi-Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433 Difference between Adjacent and Distributed PN-Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433 Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433 Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433 13 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 5.7.1.3.2 5.7.1.3.3 5.7.2 5.7.2.1 5.7.2.2 5.7.2.3 5.8 5.8.1 5.8.2 5.8.2.1 5.8.2.2 5.8.2.3 6 © Forsk 2015 Transmitter Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Site Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allocation Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy: PN Offset per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy: Adjacent PN-Clusters Per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy: ‘Distributed PN-Clusters Per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435 436 436 436 437 437 Automatic GSM-CDMA Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Algorithm Based on Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438 438 438 438 439 441 LTE Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445 6.1 Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445 6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.2.5 6.2.6 6.2.7 6.2.8 6.2.9 6.2.10 6.2.11 6.2.12 6.2.13 6.2.14 6.2.15 6.2.16 6.2.17 6.2.18 6.2.19 6.2.20 Calculation Quick Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink Transmission Powers Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Co- and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interference Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Noise Rise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation of Downlink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation of Uplink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation of Downlink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation of Uplink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation . . . . . . . Scheduling and Radio Resource Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450 450 453 453 455 456 460 460 462 462 462 463 463 463 464 465 465 465 466 467 469 6.3 6.3.1 6.3.1.1 6.3.1.2 6.3.1.3 6.3.1.4 6.3.2 6.3.2.1 6.3.2.2 6.3.2.3 6.3.2.4 6.3.3 6.3.4 6.3.4.1 6.3.4.1.1 6.3.4.1.2 6.3.4.2 Available Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Profile View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reception View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interference View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Details View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink Signal Level Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Effective Signal Analysis Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/(I+N)-based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cell Identifier Collision Zones Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations on Subscriber Lists. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simulations Based on User Profile Traffic Maps and Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . . Simulations Based on Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470 470 470 470 470 470 471 471 472 473 475 476 476 476 476 478 479 6.4 6.4.1 6.4.2 6.4.2.1 6.4.2.2 6.4.2.3 6.4.2.4 6.4.3 6.4.3.1 6.4.3.2 Calculation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink Transmission Power Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Co- and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conversion From Channel Numbers to Start and End Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Co-Channel Overlap Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adjacent Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total Overlap Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subframe Pattern Collision Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subframe Pattern Normalisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determination of Effective Subframe Patterns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485 485 493 494 495 496 496 497 497 498 14 AT330_TRR_E1 6.4.3.3 6.4.4 6.4.4.1 6.4.4.2 6.4.4.3 6.4.4.4 6.4.4.5 6.4.4.6 6.4.4.7 6.4.4.8 6.4.4.8.1 6.4.4.8.2 6.4.4.9 6.4.4.10 6.4.5 6.4.6 6.4.6.1 6.4.6.1.1 6.4.6.1.2 6.4.6.2 6.4.6.2.1 6.4.6.2.2 6.4.6.3 6.4.7 6.4.7.1 6.4.7.2 6.5 6.5.1 6.5.2 6.5.3 6.5.3.1 6.5.3.2 6.5.3.3 6.5.4 6.5.4.1 6.5.4.2 6.5.4.3 6.5.5 6.5.5.1 6.5.5.2 6.5.5.3 6.5.6 6.5.6.1 6.5.6.2 7 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents Calculation of Subframe Collision Probabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499 Signal Level and Signal Quality Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501 Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501 Noise Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505 Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505 C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516 C/(I+N) and Bearer Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518 Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523 Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525 Interference Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525 Interfering Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526 Noise Rise Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528 C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529 C/(I+N) and Bearer Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532 Best Server Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535 Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 Calculation of Downlink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 Calculation of Uplink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543 Calculation UE Capacities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545 Calculation of Downlink UE Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545 Calculation of Uplink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation . . . . . . 547 Scheduling and Radio Resource Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552 Scheduling and Radio Resource Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552 User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561 Automatic Planning Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563 Automatic Neighbour Planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563 Automatic Inter-technology Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567 Automatic Frequency Planning Using the AFP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570 Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570 Cost Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571 Automatic Physical Cell ID Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572 Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572 Cost Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576 Automatic PRACH RSI Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576 Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576 Cost Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578 Interference Matrix Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578 Distance Importance Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579 3GPP Multi-RAT Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .583 7.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583 7.2 7.2.1 7.2.2 Multi-RAT Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583 User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583 Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585 7.3 Multi-RAT Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585 8 3GPP2 Multi-RAT Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .589 8.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589 8.2 8.2.1 8.2.2 Multi-RAT Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589 User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589 Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590 8.3 Multi-RAT Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591 9 TD-SCDMA Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .595 15 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents © Forsk 2015 9.1 9.1.1 9.1.2 9.1.3 9.1.4 9.1.5 9.1.6 9.1.7 Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P-CCPCH Eb/Nt and C/I Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DwPCH C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DL TCH Eb/Nt and C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UL TCH Eb/Nt and C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA Dynamic Power Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595 595 600 600 601 601 602 602 9.2 9.2.1 9.2.1.1 9.2.1.2 9.2.2 9.2.2.1 9.2.2.2 9.2.2.2.1 9.2.2.2.2 9.2.2.3 9.2.2.4 9.2.2.5 9.2.2.5.1 9.2.2.5.2 9.2.2.6 9.2.2.6.1 9.2.2.6.2 9.2.2.7 9.2.2.7.1 9.2.2.7.2 9.2.2.8 Signal Level Based Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RSCP Based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P-CCPCH RSCP Coverage Prediction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Best Server P-CCPCH Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P-CCPCH Pollution Analysis Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DwPCH RSCP Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UpPCH RSCP Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Baton Handover Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scrambling Code Interference Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602 602 603 603 603 603 604 604 604 605 605 605 605 606 606 606 606 607 607 607 607 9.3 9.3.1 9.3.1.1 9.3.1.1.1 9.3.1.1.2 9.3.1.2 9.3.1.2.1 9.3.1.2.2 9.3.1.2.3 9.3.2 9.3.2.1 9.3.2.2 9.3.2.2.1 9.3.2.2.2 9.3.2.2.3 9.3.2.2.4 9.3.2.2.5 9.3.2.2.6 9.3.2.2.7 9.3.2.3 9.3.2.3.1 9.3.2.3.2 9.3.2.3.3 9.3.2.3.4 9.3.2.3.5 9.3.2.4 Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Generating a Realistic User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simulations Based on User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Circuit Switched Service (i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packet Switched Service (j) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Throughputs in Uplink and Downlink. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total Number of Users (All Activity Statuses) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Number of Users per Activity Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Power Control Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Algorithm Initialisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . R99 Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determination of Mi’s Best Server (SBS(Mi)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dynamic Channel Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Uplink Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Uplink Signals Update . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink Signals Update. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Control of Radio Resource Limits (Downlink Traffic Power and Uplink Load) . . . . . . . . . . . . . . . . . . HSDPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA Power Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Connection Status and Number of HSDPA Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA Admission Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HSDPA Dynamic Channel Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ressource Unit Saturation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Convergence Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608 608 608 609 609 612 612 613 613 613 614 614 614 615 617 619 621 621 621 622 622 624 624 625 625 625 9.4 9.4.1 9.4.2 9.4.3 9.4.4 9.4.5 9.4.6 9.4.7 9.4.8 9.4.9 TD-SCDMA Prediction Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P-CCPCH Reception Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DwPCH Reception Analysis (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink TCH RSCP Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Uplink TCH RSCP Coverage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink Total Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Downlink Service Area Analysis (Eb/Nt) or (C/I). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Uplink Service Area Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Effective Service Area Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cell to Cell Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626 626 628 629 630 631 631 633 635 636 16 AT330_TRR_E1 9.4.10 9.4.11 Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents UpPCH Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637 HSDPA Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637 9.5 9.5.1 9.5.1.1 9.5.1.2 9.5.1.3 9.5.1.4 9.5.1.5 9.5.2 9.5.3 9.5.4 9.5.4.1 9.5.4.1.1 9.5.4.1.2 9.5.4.2 9.5.4.2.1 9.5.4.2.2 9.5.4.2.3 Smart Antenna Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 638 Modelling in Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639 Grid of Beams Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639 Adaptive Beam Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 640 Statistical Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641 Beamforming Smart Antenna Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641 3rd Party Smart Antenna Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642 Construction of the Geographic Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642 Modelling in Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643 HSDPA Quality and Throughput Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644 Fast Link Adaptation Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644 CQI Based on P-CCPCH Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644 CQI Based on HS-PDSCH Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648 Coverage Prediction Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 Colour per CQI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 Colour per Peak Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650 Colour per HS-PDSCH Ec/Nt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650 9.6 9.6.1 N-Frequency Mode and Carrier Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650 Automatic Carrier Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650 9.7 9.7.1 9.7.2 9.7.3 9.7.4 Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Neighbour Allocation for a Group of Transmitters or One Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655 Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655 Appendix: Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656 9.8 9.8.1 9.8.1.1 9.8.1.2 9.8.1.3 9.8.1.3.1 9.8.1.3.2 9.8.1.4 9.8.1.4.1 9.8.1.4.2 9.8.1.4.3 9.8.2 9.8.2.1 9.8.2.1.1 9.8.2.1.2 9.8.2.1.3 9.8.2.1.4 9.8.2.2 Scrambling Code Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656 Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657 Allocation Constraints and Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657 Allocation Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658 Allocation Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658 Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658 Multi-Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659 Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660 Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660 Transmitter Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662 Site Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663 Scrambling Code Allocation Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663 Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663 Strategy: Clustered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664 Strategy: Distributed per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664 Strategy: One SYNC_DL Code per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665 Strategy: Distributed per Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665 Multi Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665 9.9 9.9.1 9.9.1.1 9.9.1.2 9.9.1.3 9.9.1.3.1 Automatic GSM/TD-SCDMA Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666 Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666 Algorithm Based on Distance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667 Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669 Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669 10 WiMAX BWA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .673 10.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673 10.2 10.2.1 10.2.2 10.2.3 10.2.4 10.2.5 10.2.6 10.2.7 Calculation Quick Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 678 Co- and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 678 Preamble Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 Preamble Noise Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 Preamble Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 Preamble C/N Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 Preamble C/(I+N) Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680 Traffic and Pilot Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680 17 . . . . . . Traffic and Pilot C/N Calculation (DL) . . . . . . . . Traffic and Pilot Interference Signal Levels Calculation (DL) . . . .3. . . . . . . . . . . . . . . .4.8. . . . . . . . . . . . . . . .2 10. . . . . . . . . . . .6. . . . . . . . . . . .12 10. . .6 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702 702 703 704 705 705 706 707 707 708 710 712 712 713 714 714 715 715 717 718 718 722 726 727 729 730 731 731 733 734 737 740 740 740 741 18 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . .2 10. . . . . . . . . . . . . . . . . . . . . . . .4.9 10. . . . . . . . . . . . . . . . . . . . . . . . . Traffic and Pilot Signal Level and Quality Calculations .4. . Effective Signal Analysis Coverage Predictions . . . . . . . . . . . . . . . .3 10. . . . . . . . . . . . . . . . . . . . . Traffic Interference Signal Levels Calculation (UL) . . . . . . . . . . . 690 690 690 690 690 690 691 691 692 693 695 695 696 696 696 698 699 10. . . . . . . . . . . . .2. . . . .5 10. . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . .14 10. . . . . . .3. . . . . . . . . . .4. . . . . . . . . . . . . . Traffic Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic and Pilot C/(I+N) and Bearer Calculation (DL) . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7. . .1 10. . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . Reception View . . . . Preamble Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . .6 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 10. . . . . and Per-user Throughput Calculation . . . . . .6. . . . . . . . . . . . . . . . .4. . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . Calculations on Subscriber Lists. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic and Pilot Noise Calculation (DL) . Traffic Interference Calculation (UL) . . . . .4 10. . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . Profile View . . . . . . . .1 10. . . . . . . . . . . . Traffic and Pilot Signal Level Calculation (DL) . . . . . . . . Calculation of Total Cell Resources . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . .4. . . . .1 10. . . .2. . . . . . . . . . . . . . . . . .3 10. . . . . . . . . . .4. . .4. . . . . . . . .2. . . User Distribution . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . Preamble C/N Calculation. . . . . . . . . . . . . . . . . . . . .17 10.3 10. .1. . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . .2 10.0 Technical Reference Guide for Radio Networks Table of Contents 10. . . . . . . . . . . . . . . . . . . . . .4 10. .1 10. . . . . . . .7 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . .2. Traffic C/(I+N) Calculation (UL) . . . . . . . .4. . . . . . . . . . Preamble Signal Level Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . User Throughput Calculation . . . . . . . . . . . . . . . . . . .5 10. . . . . . . . . . . . . . . . .2 Calculation Details . . . . .4 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 10. . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 10. . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interference View . . . . . . . . . . . . . . .2. . . . . . .3. . . . . .3 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Co. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 10. . .7 10. . . . . . . . . . . . . . . . . . . . . . Traffic Noise Calculation (UL) . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/(I+N)-based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 10. . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . .1. . . . . . . . . . . . .1. . . . . .4. . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . .2. .13 10. . . . . . . . . . . . . . . . . Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . .2 Available Calculations . . . . . . . . . . . . . . . . .and Adjacent Channel Overlaps Calculation . . . . .4. . .1. . .1 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 10. . . . . . . . . . . . . . . . . . . . . . . . Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic and Pilot C/(I+N) Calculation (DL) . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3.1 10. . . . . . . . . Best Server Determination. . . .1. . . . . . . . . . . .1 10. . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10 10. . . . . . . . . . . . . . .4. . . . . . . Coverage Predictions . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . .1 10.2. . . . . . . .3 10. . . . . .1. . . . . . . . . . . . . . . . Details View . . . . . . . . . . . . . . . . .1. . 680 681 682 682 683 683 683 684 684 684 685 687 689 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic Signal Level Calculation (UL) . . . . . .4. . . . . . . . . . . . Preamble Signal Level and Quality Calculations. . . . . . . . . . . . . . . . Traffic Interference Calculation (UL) .4. . . . . . .1 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . Traffic Noise Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 10. . . . . . . . . . . .16 10. . . . . . . . . . . Calculation of Sampling Frequency . . . . . . .2. . . . . .9 10. . . . . . . . . . . . . . . . .2. . . . . . . . . . .6. . . . . . .3. . . . .1. . . . .3. . . . . Traffic C/(I+N) and Bearer Calculation (UL) . . .4. Permutation Zone Selection . . . . . . . . . . . . . . . . . . . . . . . . . .10 10. . . . . . .4. . . . . . . . . . . . Co-Channel Overlap Calculation. . .1 10. . . . .4 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 10. . . . . . . . . . . . . . .3. . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 10. . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allocated Bandwidth Throughput. . . . . . . . . . . . . . . . . . . . . . . .4. . . . . .3 10. . . . . . . . . . . . .2 10. . . . . . . . . . . . . . . . Simulations Based on Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . .4 10. . . . . . . . . . . . Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . .4. . .19 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cell Identifier Collision Zones Coverage Prediction . . . . . . . . . . . .5 10. . . .2 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic C/N Calculation (UL) . . . . . . . .Atoll 3. .4. . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . Preamble Noise Calculation . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . Noise Rise Calculation (UL) . . . . . . .1 10. . Calculation of Symbol Duration . . . . . . . . . . . . . . . . . . . . . . . .5 10. . . . . . . . . . . . . . . . . . . . . . . . . . . .2 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic C/N Calculation (UL) . . . . . . . . . . . . . . . . . .6. . FDD – TDD Overlap Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . Cell Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total Overlap Ratio Calculation . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . Adjacent Channel Overlap Calculation . .2 10. . . . . . . . . . . . . Effective Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . Service Area Calculation. . . . . . . . . Simulation Process . . . . . . . . . . . . . . . . . . . . Conversion From Channel Numbers to Start and End Frequencies . Scheduling and Radio Resource Management. . . . . . . . . . . . . . . . Channel Throughput.3 10. . . . . . . . . . . . . . . . . . . Preamble C/(I+N) Calculation . .3. . . . . . . . . . . . . . . Simulations Based on User Profile Traffic Maps and Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . .3. . . .3. . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . .2. . . . . . . . . . . . . . . . . .2 10. . . . . . . .7. . . . . . . . . . . . . . . . .20 © Forsk 2015 Traffic and Pilot Noise Calculation (DL). . . . . . . Preamble Interference Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic and Pilot C/N Calculation (DL) . . . . . . .7.1. . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . .1 11. . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767 Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . . .1 10. 783 C/N Calculation (DL). . . . . . . . . . . . . . . . . . . .1. 742 Channel Throughput. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . .4. . . . . . . . . . . . . . . . . . . 774 Wi-Fi Networks . . . . . . . . 769 AFP Algorithm . . . . . . . . . . . . . . . . . . . . 772 AFP Algorithm .7 11. . . . .3. . . . . 783 Interference Calculation (DL) .4. . . . . . . . . . 789 Point Analysis .1 10. .2. 786 Scheduling and Radio Resource Management . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . 800 Co-Channel Overlap Calculation . . . . . . . . . . .2 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789 Signal Level Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 10. . . . . . . . .2 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . 788 11. . 789 Interference View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . 801 19 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . .2 Available Calculations . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772 Cost Calculation. . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . 766 Cost Calculation. . . . . . . 784 Signal Level Calculation (UL) . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771 Automatic Zone PermBase Planning Using the AFP . . . . . . . . . . .1 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771 Constraint and Relationship Weights . . . . . . . . . .7. . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . .3 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791 Calculations on Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . 743 Scheduling and Radio Resource Management . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . .TDD Networks . .1 10. . . . . . . . . . . . . . . . . . . . . . . 767 Automatic Preamble Index Planning Using the AFP .5 11. . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 11. .3 10.9 11. . . . . . . . . . . . . . . . . 757 Automatic Planning Algorithms . . . . . . .5. . . . . . . . . . . . . . . . . . 793 User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 11 Atoll 3. . . . .4. . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . 789 Effective Signal Analysis Coverage Predictions. . . . . . . . . . 784 Noise Calculation (UL) . . . . .4 11.2. . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 11. . . .5. . . . . . . . . . 784 C/N Calculation (UL). . . . 782 Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . .3.6 10. . . . . . . . . . . . . . . . 766 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . 799 Co. . . . . 748 User Throughput Calculation . . . .1. . .2. and Per-User Throughput Calculation . . . . . . 784 C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 11. . . . . .3. . . . . . . . .7.2. . . . . . . . . . . . . . . . . .4 10. . . . . . . . . . . . . . . . .2 10. . 748 Scheduling and Radio Resource Allocation. . . . . . . . . . . . . . . . . . . Allocated Bandwidth Throughput. .4. . . . . . . . . . . . . . . . . . . . . . . . . . . 793 Simulations Based on User Profile Traffic Maps and Subscriber Lists . . . . . . .3. . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789 Reception View . . . . . . . . . 794 Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 11. . . . . . . . .4 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10 11. . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . 785 Calculation of Total Cell Resources. . . . . . . . . . . . . . . . . . . . . . . 800 Adjacent Channel Overlap Calculation . . . . 779 11.5. . . . . . . . . . . .5 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . .2 11. . . . . . .4. . . . . . . . . . . . . . 799 Conversion From Channel Numbers to Start and End Frequencies . . . . . . . . . . . . . . . .8 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . .14 11. . . . .13 11. . . . . . . . . . . . . . . . . . . . .4. . .2 11. . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . and Per-user Throughput Calculation . . . . . . . . . . . . . .1 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 11. . . . . . . . . . . . . . . . . . . . . . . . . . . Cell Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . .4 11. . . . . . . . . . . . .4 10. . . . . . . . . 793 Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . .779 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 11. . . . . .1 11. . . . . . . 785 Channel Throughput. . . . . . . . . . . . . . . . . . . . . . . . . 774 Interference Matrix Calculation. . . . . . . . . . . . . .8. . . . . .2 11. . . . . . . . . . . . . . . . . . . .3 10. . . . . . . . . . . . . . . . . . . . 782 Co. . . . . . . . . . . . . . . .0 Technical Reference Guide for Radio Networks Table of Contents Calculation of Total Cell Resources . . . . . . . . . . .2. . . . . . . .2. . . .8. . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795 Simulation Process . . 765 Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7. . .5. . . . . . . . . .1 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 11. . . . . . . . . . . . . .8 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . .1. . . . . . . . . . . . . 789 Profile View . . . . . . . . . . . . . . .2 10. . . . . . . . . . .5. . . . . 759 Automatic Inter-technology Neighbour Planning . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . .2. . . . . . 784 Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 11. . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. .4. . . .AT330_TRR_E1 10. . . . . . . . . . 791 C/(I+N)-based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 11. . . . . . . . . .5 10. . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . .3 Calculation Details . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789 Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783 Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 10. . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . .3. . . . . . . . . . . . . . . . . . . .3 10. . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .FDD Networks . . . . . . 774 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763 Automatic Frequency Planning Using the AFP. . . . . .4. . . . . . . . . . 741 Calculation of Total Cell Resources . . . . . . . . . . . . 758 Automatic Neighbour Planning. . . . . . .5. . . .1 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774 Distance Importance Calculation. . . .5. . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . .2 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . .2.3. . . . . . . . . . . . .and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . . . . .3 11. .1 11. . . . . . . . . . . . . . . .3 11. . .3 10. . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1.4. . . .5. . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785 C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 11. . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cell Capacity. . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768 Cost Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 Calculation Quick Reference. . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . .4 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . .1. . . . . . . . . . . . . . . . . . .3 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signal Level and Quality Calculations . . . . . . . .1 12. .1 11. . . . . . . . . .2. . . . . .2 12. . . . . . . . . . . . . . . . . . . . . Configuring an Optimisation Setup . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . .6 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CDMA2000 Quality Indicators . . . . . . . . . . . . . . .4 11. . . . . Antenna Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Quality Predictions and the Antenna Masking Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Remote Antennas. .4. . . . . . . . Additional Electrical Tilt (AEDT) . . . . . . . .8 11. . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . WiMAX Quality Indicators. . . . . . . . . . . . . . . . . . . . . . .4. . .2. . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . Relative Electrical Tilt Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Automatic Frequency Planning Using the AFP . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . .6. . Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . LTE Quality Indicators . . . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Atoll 3. . . . . . . . . . .5 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839 839 839 840 841 841 841 841 841 842 842 842 843 843 843 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . GSM Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . .1. . . . . . . . . . . . . . . . .2 11. . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . .5 11. . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interference Matrix Calculation . . . . Interference Calculation (UL) . . . Interference Signal Levels Calculation (UL) . . . . . . . . . . . . .5 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 11. . . . . . . . . . . . Basic Method . . . .5. . . . . . . Appendices . . . . . . .2. . . . . . Definition and Evaluation . . . . . . . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .0 Technical Reference Guide for Radio Networks Table of Contents 11.4. . . . . . . .3. . . . . . . . . . Noise Calculation (UL) .4. . . . . . . . . . . . . . . . .1 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802 803 803 804 804 806 807 809 810 810 811 811 812 813 815 815 816 816 817 820 820 823 Automatic Planning Algorithms . . . . . . . . . . .6 11. . . . . . . . . . . . . . . . . . . . . . Antenna Masking Modes for Non-Native Propagation Models . . . . . . . . . . . . Advanced Objective Configuration . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . Signal Level Calculation (UL) . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7 11. . . . . . . . . . .1. Distance Importance Calculation . . .1 12. . . . . . . . . . . . . . . . . .4. . . . . . . . . . . Channel Throughput. . . . . . . . . . . . .4. . . . . . . . . . . . . . . . .1. . . and Secondary Antennas . . .4. . . . . . . . . . . . . . . . . . . .4. . .3 12. . . . . . . . . . . . . . . . . . . .2 11. . . . . . . . . . . . . . . . . . . . . . . .1. .2. . .4. . . . . . . .1 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 11. . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . .3 11. . . . . . . . . . . . . . . . . . . . . . . . . . .2. .2 11. . . .2 11. . . . . . . . . . . . . . . . Antenna Masking and Repeaters. . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Throughput Calculation . . . Cost Objective . . . . . . . . . . . . . . . . . . . .3 12. . .2. . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . .1. . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . 844 844 844 844 845 845 845 846 846 12. . . . . . . . . . . . . . .4 11. . . . . . . . . . . . . . . . . . . . . . . . Scheduling and Radio Resource Allocation . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 11. . . . . . . . . . . . . . . . . .3 12. . .4. . . . . . . . . . . . C/(I+N) and Bearer Calculation (UL). . . . . . . . . .4. . . .2. . . . .4. . . . . . . . . . . . . . . . . . . . . . . . .4 12. . . . . . . . . . . Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Improved Method . . . . Quality Objective . . . . . . . .2. . . . . . . UMTS Quality Indicators . . . . . . . . . . .1. . . . .2 12 © Forsk 2015 Total Overlap Ratio Calculation . . . . . . . . . . . . . . . .2. . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . Target Filtering. . . .2. . .2 11. . . . . . . . . . . . . . . . . . . . . Optimised Method . . . . . . . . . . . . . . . . . . . . . . . . . Cell Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . 839 12. . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 11. and Per-user Throughput Calculation . . . . . . . . . . . . . . . . . . . . Constraint and Relationship Weights . . . . . . . . . . . .10 11.2. . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . .1 11. . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8. . . C/(I+N) and Bearer Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8 12. . . . . . . . . Scheduling and Radio Resource Management. . . . . . . . . . . . . . . . . . . . . . . .3 Objectives . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . .8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User Throughput Calculation . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . Cost Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846 846 846 847 847 20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . .2. . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . Antenna Correction Method . . .4. . . . . . . .5. Signal Level Calculation (DL) . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Automatic Neighbour Planning . .2. .4 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C/N Calculation (DL) . . . Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . .3. . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . Full Path Loss Method. . . . . .5. . . . . . . . . .1. . . . . . . . AFP Algorithm . . . . . . . . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . .1. . . . . . . . . . . . . . .3 11. . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . .2 12. .2. . . . . . . . . . . . . . . . . . . .1. . . . .1 11. . . . . . . . . . . . . . . . . . . Quality Indicators in the ACP . . . . . . . . . . . . . . .2 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 12. . 824 825 829 831 832 832 833 833 833 834 ACP Module . . . . . . . . . . . . .3 12. . . . . .1 12. . . . . . . . . . . . . Automatic Inter-technology Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 11. . .2. . . . . . . . . . . . . . . . . . . . . .1 11. . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . .3. . . Atoll and ACP Prediction Matching . . . . .1. . . . . . . . . . . . . . . . . .1. . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . .9 11. . . . . . . . . . . . . . . . . . . Best Server Determination. . . .3 Configuration . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . .4 11. . . .2 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Noise Rise Calculation (UL) . . . . . . CrossWave Propagation Model . . . . . . . . . . . Service Area Calculation. . . . . . . . . . .3. . . . . . Quality Indicator Parameters and Reference Maps . Progressive Thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . 863 Worst-case Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . 867 12. . . . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . .7. . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5.1 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852 Controlling the Optimisation. . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . 855 Traffic Capture for Load Balancing .5 12. . . . . . . . . . . . .6. . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865 Results . . . . . . . . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . . . . 850 Sorting Algorithm . . . . . . . . . . . . . . . . . . .9. . . . . . . . 863 EMF Exposure Calculation . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862 The Contribution of Transmitter Power to EMF Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9. . . . . . . . .9. . . . . . . . . . . . . . . . . . . . . . . . 851 Tuning Algorithm . . . . .9 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . .4. . . . . . . .3 12. . . . . . . . . . . . . . . . .7. . . . 848 Multi-RAT and Co-planning Modes. . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . .8 Shadowing Margin and Indoor Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . .7 12. . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . 867 Distribution of Evaluation Points . 858 Load Balance . . . .3 12. . . . . . . . . . . . . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859 Load Balancing Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 12. . . . . . . . 862 Terrain Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7. . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865 Pixel Weighting. . . . . . . . . . . . 861 12. . . . . . . . . . . . .5. . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863 12. . . . . .4. . . . . . . . . . . . . . . . . . . .9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . .4 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848 Technology Layer Definition . . . . . . . . . . . . . . . . . . .6. . 857 Captured Traffic Ratio. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859 Optimisation Results . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854 Memory Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9. . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . .1 12. . . . . .3. . .6. . . . . . . . . . . . . . . . . . . . . . .9.2. . . . . . . . . . . . . . . .4 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . 853 Memory Usage and Optimisation Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . 859 Average Load. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862 Propagation Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867 The Contribution of Transmitter Power to EMF Exposure . .3. . . . . . . . . . . . . . . . . . . . . . . . 851 Search Algorithm. . . . .1 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . .9. . . . . . . . . . .3 12. . . . . . . 860 Impact on the Global Score Function . 867 General Workflow . . . . . . . . . 865 Notes . . . . . 863 General Workflow . . . . . . . . . . . . . . . . . . . .5 12. . . . . . . . . . . . . . . . . . . .5. . . . . . . . . .6 12. . . . . . . . . . . . . . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 12. . . . . . . . . . . . . . 857 Introduction of Load Balancing as a Quality Indicator. . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . .0 Technical Reference Guide for Radio Networks Table of Contents 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866 Terrain Profile . . . . . . . . . . . . . . .3 12. . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . .6.1 12. . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . .9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858 Quality Figures Used for Graphs and Statistics Results. . . . . .4 12. . . . . .2 Optimisation Methodology. .10 ACP Software Data Flow . . . . . . . . . . . . . . . . . . .7. . . . . . . . . . . . . . . . . . . . . . . . . . 865 Path Loss Calculation and Data Caching.2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . . . .6 12. . . . . . . . . . . . 852 Implementation Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867 Worst-case Mode .1. .4 12. . . . . . . . . . . . . . . . . . . . . . . . . . . .6 Multi-Storey Optimisation . . . . . . . . . . . . . . . .6 12. . . . . . . . . . . . . . . . . . . . 854 Internal Data Management and Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . 855 Optimisation Principle . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . 868 21 . . . . 854 Disk Space Usage . .AT330_TRR_E1 Atoll 3. . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . .9. . . .7 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . .1. . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 12. . . . . . . . . . . . . . . . . . . . . . . . .5 12. . . . . . . . . . .5 12. . . . . .1. . . . . . . 849 12. . . . . . . . . . .4. . . . . . . . . . . 855 Principle Used in ACP . . . . . . . . .7. . . . . . . . . . . .2 12. . . . . . . . . . . . 859 Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 12. . . . . . .2 Multi-RAT and Co-planning Support . . . . .5. . . . . . 851 Global Score Function . . . . . . . . . . . . . . . . .5 12. . . . . . . . . . . . . . . . . . . . . .1. . . 866 Concepts of ACP EMF Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856 Load Quality Index . . . 849 Tuning Algorithm . . . . . . . . . .1 12. . . . . . 865 12. . . . . . . . . . . . . . .7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866 Propagation Classes . . . . . . . .3 12. . . . . . . . . . . . . . . . . . . . . . .1 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . .7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856 Load Balancing Score Function . . . . 855 Cell Capacity Load Calculation . . . . . . . . . . . . . . . . .2 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9. . . . . . . . . . . . . . . . . . . . . . . . 861 Concepts of ACP EMF Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9. . . . . . . . . . .3 EMF Exposure . . . . . . . . . . . . . .9 12. . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . 849 Search Algorithm . . . 852 Weighting . . . . . . . . . . . . . . 862 Distribution of Evaluation Points . . . . . . . . . . .4 12. . . .8 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 Load Balancing Objective . . .7. . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 12. . . . . . Atoll 3.3.0 Technical Reference Guide for Radio Networks Table of Contents 22 © Forsk 2015 . Chapter 1 Antennas and Equipment This chapter covers the following topics: • "Antenna Attenuation" on page 25 • "Antenna Pattern Smoothing" on page 27 • "Power Received From Secondary Antennas" on page 29 • "Transmitter Radio Equipment" on page 30 • "Repeaters and Remote Antennas" on page 32 • "Beamforming Smart Antenna Models" on page 43 • "Grid-of-Beams Smart Antenna Model" on page 51 • "Adaptive Beam Smart Antenna Model" on page 52 • "Statistical Smart Antenna Gain Model" on page 53 . Atoll 3.0 Technical Reference Guidefor Radio Networks © Forsk 2015 24 .3. the receiver coordinates are: x Rx cos  e Rx   sin  a Rx   d y Rx = cos  e Rx   cos  a Rx   d z Rx (1) – sin  e Rx   d Let az and el respectively be the azimuth and tilt of the receiver in the transmitter antenna coordinate system S Tx  x'' y'' z''  . 1. Atoll calculates the accurate azimuth and tilt angles and performs 3D interpolation of the horizontal and vertical patterns. aRx and eRx are respectively the azimuth and tilt of the receiver (Rx) in the coordinate system S 0  x y z  .1. These angles describe the direction of the transmitter-receiver path in the transmitter antenna coordinate system. aTx and eTx are respectively the transmitter (Tx) antenna azimuth and tilt in the coordinate system S 0  x y z  .Atoll 3. we have the following relations: x' y' = z' cos  a Tx  – sin  a Tx  0 x  sin  a Tx  cos  a Tx  0 y z 0 0 1 (3) and 25 . Therefore.3.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 1 Antennas and Equipment 1. Atoll determines the receiver position relative to the direction of the transmitter antenna (i. Figure 1. the direction of the transmitter-receiver path in the transmitter antenna coordinate system). the receiver coordinates in S Tx  x'' y'' z''  are: x'' Rx y'' Rx = z'' Rx cos  el   sin  az   d cos  el   cos  az   d – sin  el   d (2) According to the figure above.e. d is the distance between the transmitter (Tx) and the receiver (Rx).1: Azimuth and Tilt Computation In the coordinate system S 0  x y z  .1 Antenna Attenuation To determine the transmitter antenna attenuation.1 Calculation of Azimuth and Tilt Angles From the direction of the transmitter antenna and the receiver position relative to the transmitter. 0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment 1 0 0 x'' x' =  cos   – sin  e  0 e y'' y' Tx Tx 0 sin  e Tx  cos  e Tx  z'' z' ©Forsk 2015 (4) Therefore. the description of the antenna pattern must satisfy the following conditions: H(a0)=V(e0) 26 and H(180+a0)=V(180-e0) .  H  a 0  – V  el   + ------------------. In this case. (1) and the receiver coordinates in the system STx from Eq.+ ---------------------------------------------tan  a Rx – a Tx  sin  a Rx – a Tx  and cos  e Tx   tan  e Rx    – sin  e Tx  el = atan sin  az    ---------------------------------. (6) leads to a system where two solutions are possible: 1st solution: If a Rx = a Tx . Atoll considers these values to determine transmitter antenna attenuations in horizontal and vertical patterns. the vertical pattern is a plane section with a rotation a0 degrees from the vertical plane. the description of the antenna pattern must satisfy the following: H(0)=V(0) and H()=V() If the electrical tilt is e0.1.+ ---------------------------------------------- sin  a Rx – a Tx    tan  a Rx – a Tx  If sin  az   sin  a Rx – a Tx   0 . the relation between the system S 0  x y z  and the transmitter antenna system S Tx  x'' y'' z''  is: 1 0 0 cos  a Tx  – sin  a Tx  0 x'' x =   0 e cos   – sin  e  y'' sin  a Tx  cos  a Tx  0 y Tx Tx z'' z 0 sin  e Tx  cos  e Tx  0 0 1 (5) We get. If the electrical azimuth is a0. then 1 az = atan ---------------------------------------------------------------------------------------cos  e Tx  sin  e Tx   tan  e Rx  ----------------------------------. It reads the following: • • H(az) H(a0) V(el) the attenuation in the horizontal pattern for the calculated azimuth angle az the attenuation in the horizontal pattern for the electrical azimuth angle a0 the attenuation V(el) in the vertical pattern for the calculated tilt angle el Then it calculates the antenna total attenuation. then az = 0 and el = eRx – e Tx 2nd solution: If a Rx  a Tx .  H  180 + a 0  – V  180 – el   if |el| ≠ 90° L antTx  az el  = H  az  – ---------------------------------180 180 Else: L antTx  az el  = V(el) Atoll assumes that the horizontal and vertical patterns are cross-sections of a 3D pattern.3.2 Antenna Pattern 3D Interpolation The direction of transmitter-receiver path in the transmitter antenna coordinate system is given by angle values az and el. x'' y'' = z'' cos  a Tx  – sin  a Tx  0 x cos  e Tx   sin  a Tx  cos  e Tx   cos  a Tx  – sin  e Tx   y z sin  e Tx   sin  a Tx  sin  e Tx   cos  a Tx  cos  e Tx  (6) Then.Atoll 3. (2) in Eq. substituting the receiver coordinates in the system S0 from Eq. the horizontal pattern is a conical section with an elevation of e0 degrees off the horizontal plane. In other words. L antTx  az el  : 180 – az – a 0 az – a 0 . then az = az + 180 1. 3 Additional Electrical Downtilt Modelling The additional electrical downtilt. this implies that: • • Interpolated horizontal and vertical patterns respectively fit in with the entered horizontal and vertical patterns. 1. Angle values in formulas are stated in degrees. • • • This interpolation is performed in dBs.Atoll 3. Then. This interpolation is not used with 3D antenna patterns. the vertical pattern is transformed as follows: V  x  = V  x – AEDT  when x  [– 90.90] V  x  = V  x + AEDT  when x  [90. ASmoothing = 90 degrees. also referred to as remote electrical downtilt or REDT. Let DPeak-to-Null be 10 dB.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 If the constraints listed above are satisfied.2 Antenna Pattern Smoothing Empirical propagation models. You can smooth vertical as well as horizontal antenna patterns in Atoll. AEDT. Otherwise. L antTx  az el  . 27 .2: Vertical Pattern Transformation due to Electrical Downtilt 1. Finally. In order to take it into account. It determines the antenna attenuation in the transformed vertical pattern for the calculated tilt angle (V(el)) and applies the 3D interpolation formula in order to calculate the antenna total attenuation. introduces a conical transformation of the 3D antenna pattern in the vertical axis. Then. Peaks (P) are the lowest attenuation angles and nulls (N) are the highest attenuation angles in the pattern. The vertical pattern transformation is represented below.3 on page 28. Signal level predictions can be improved by smoothing the highattenuation points of the vertical pattern. and The contribution of both front and back parts of the vertical pattern are taken into account. it determines the nulls to be smoothed (NSmoothing) and their corresponding angles according to the defined Peak-to-Null Deviation (DPeak-to-Null). and FSmoothing = 0. Let’s take an example of an antenna pattern to be smoothed. as shown in Figure 1. The left picture shows the initial vertical pattern when there is no electrical downtilt and the right one shows the vertical pattern transformation due to an electrical downtilt of 10°.5. DPeak-to-Null is the minimum difference of attenuation in dBs between two peaks and a null between them. like the Standard Propagation Model (SPM). only the second point is guaranteed. Figure 1. The antenna pattern smoothing algorithm in Atoll first determines the peaks and nulls in the pattern within the smoothing angle (ASmoothing) defined by the user. Atoll smoothes the pattern between 0 and the smoothing angle (ASmoothing) by applying the smoothing factor (FSmoothing) defined by the user.3. require antenna pattern smoothing in the vertical plane to simulate the effects of reflections and diffractions.1. even in case of electrical tilt. Atoll proceeds as explained in the previous section. the angle values are in degrees.270] Where. Atoll verifies whether the difference of attenuation at a given angle is DPeak-to-Null less than the before and after it.3: Vertical Antenna Pattern Atoll first determines the peaks and nulls in the part of the pattern to be smoothed by verifying the slopes of the pattern curve at each angle.2 . Nulls to be smoothed (NSmoothing) 28 Angle (°) Attenuation (dB) 15 33.6 Then.5 21 13.6 49 32.6 38 16. Figure 1.2 30 37.5 30 37.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 Figure 1.Atoll 3.1 15 33.9 49 32.3.4: Peaks and Nulls in the Antenna Pattern Peaks (P) and Nulls (N) Angle (°) Attenuation (dB) 1 0. This comparison determines the nulls to be smoothed (NSmoothing).2 67 15. and PDLRS in LTE). • azi : the difference between the receiver antenna azimuth and azimuth of the transmitter secondary antenna. Atoll applies the smoothing algorithm to all the attenuation values at all the angles between the first peak.3 Power Received From Secondary Antennas When secondary antennas are installed on a transmitter. PTx is the transmitter power (Ppilot in UMTS HSPA and CDMA2000.(not in dB1) L model  P rec  Where. 1.Atoll 3. Formula cannot be directly calculated from components stated in dB and must be converted in linear values. • • 1. az and el. G ant – i Tx is the gain of the secondary antenna. the signal level received from it is calculated as follows:     G ant – m Tx  G ant – i  X i  ---------------------Tx  P Tx   1 – P Tx  X i  -------------------- L Tx    L Tx i  -----------------------------------------------------------------.3. G ant – m Tx is the gain of the main antenna installed on the transmitter. i. PPreamble in WiMAX. i.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 Once the nulls are known. AAngle is the attenuation at any given angle which can be i. i. Smoothing Algorithm For all nulls n  N Smoothing surrounded by two peaks P1 and P2 at angles  1 and  2 . i. i is the angle in degrees from  1 to  2 incremented by 1 degree.  1 or  2 . installed on the transmitter. Tx The definition of angles. Method 1 (must be indicated in an Atoll. Lmodel is the path loss calculated by the propagation model. i is the secondary antenna index. • elm: the difference between the receiver antenna tilt and tilt of the transmitter main antenna. xi is the percentage of power dedicated to the secondary antenna. i. • eli : the difference between the receiver antenna tilt and tilt of the transmitter secondary antenna. L ant – m  az m el m  is the attenuation due to main antenna pattern. Tx L ant – i  az i el i  is the attenuation due to pattern of the secondary antenna. Method 2 (default): • azm : the receiver azimuth in the coordinate system of the transmitter main antenna. i.+ ---------------------------------------- L ant – m  az m el m  L ant – i  az i el i    Tx Tx i   = -------------------------------------------------------------------------------------------------------------------------------. LTx are transmitter losses (LTx=Ltotal-DL). • azi : the receiver azimuth in the coordinate system of the transmitter secondary antenna.ini file): • azm: the difference between the receiver antenna azimuth and azimuth of the transmitter main antenna. • eli : the receiver tilt in the coordinate system of the transmitter secondary antenna.  A 2 – A  1   -   i –  1   A i Smoothed = A i – F Smoothing A i –  A  +  ---------------------1  2 – 1    Where. i. the null. 29 . and the last peak. • elm : the receiver tilt in the coordinate system of the transmitter main antenna. PP-CCPCH in TD-SCDMA. depends on the used calculation method. and FSmoothing is the smoothing factor defined by the user. In Atoll. the transmitter noise figure corresponds to the BTS noise figure.Atoll 3. are taken into account to evaluate: • Total UL and DL losses of transmitter ( L total – UL L total – DL ) and transmitter noise figure  NF Tx  in UMTS HSPA. • 30 UL L Misc are the miscellaneous reception losses (Transmitter property). Uplink Total Losses Atoll calculates total UL losses as follows: UL UL UL UL L Total – UL = L Misc + L Feeder + L BTS – Conf + NR Repeaters – G Ant – div – G TMA Where. and LTE documents. WiMAX.1 GSM Documents Atoll calculates DL total losses as follows: DL DL DL DL L Total – DL = L TMA + L Feeder + L Misc + L BTS – Conf Where. we have NF TX = NF BTS .3. The entry to the BTS is considered the reference point which is the location of the transmission/reception parameters. DL • L TMA is the TMA transmission loss. TD-SCDMA. Novosad T. CDMA2000. . the transmission feeder length in metre and the connector transmission loss). DL 1. the transmitter-equipment pair is modelled a single entity.Location of the Transmission/Reception parameters • According to the book “Radio network planning and optimisation for UMTS” by Laiho J. TD-SCDMA. the noise figure corresponds to the loss in case of passive components.4. UL NF Feeder = L Feeder • Loss and gain inputs specified in ATL documents must be positive values. Therefore.4. Wacker A. • L BTS – Conf are the losses due to BTS configuration (BTS property). • CDMA2000 1xRTT 1xEV-DO. WiMAX... I Feeder and L Connector DL DL DL DL DL DL are respectively the feeder loss per metre. DL • L Misc are the miscellaneous transmission losses.2 UMTS. 1. Therefore. Transmitter total losses  L Total  in GSM GPRS EGPRS documents. and LTE Documents As the reference point is the BTS entry.5: Reference Point . • L Feeder is the feeder transmission loss ( L Feeder = L Feeder  I Feeder + L Connector . Figure 1. feeder noise figure is equal to the cable uplink losses.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 1. Where NF BTS is the BTS noise figure. where L Feeder ..4 Transmitter Radio Equipment Radio equipment such as TMA. feeder and BTS. The noise rise at transmitter due to repeaters is calculated as follows:  NR Repeaters = 10  Log  1 +   1 ------------------  NIM Rp  r r For each active repeater ( k ). This gain does not exist in WiMAX and LTE documents. which is calculated as follows: WithoutTMA G TMA = NF Composite WithTMA WithTMA – NF Composite WithoutTMA Where NF Composite and NF Composite are the composite noise figures with and without TMA respectively.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 • UL UL UL UL UL L Feeder are the feeder reception losses ( L Feeder = L Feeder  I Feeder + L Connector . • L Feeder is the feeder reception loss ( L Feeder = L Feeder  I Feeder + L Connector .Atoll 3. • L • For each active repeater ( k ). • G TMA is the TMA reception gain. Friis' equation is used to calculate the composite noise figure when there is a TMA. where L Feeder . • NR Repeaters is the noise rise at transmitter due to repeaters. This is the difference between the k donor transmitter noise figure ( NF TX ) and the repeater noise figure received at the donor. • NF TMA is the TMA noise figure. where L Feeder . UL • L BTS – Conf are the losses due to BTS configuration (BTS property). • G Feeder is the feeder UL gain G Feeder = – L Feeder . the reception feeder length in metre and the connector reception loss). the reception feeder length in metre (Transmitter property) and the connector reception losses. I Feeder and UL L Connector are respectively the feeder loss per metre (Feeder property). NF Composite = NF BTS + NF Feeder Where. WithTMA NF Composite NF Feeder NF BTS  NF  -----------------------------------------TMA  ------------------ 10 10 10 10 – 1 10 – 1 + ---------------------------------. I Feeder and UL UL UL UL UL UL UL UL UL UL L Connector are respectively the feeder loss per metre. Atoll calculates a noise injection margin ( NIM Rp ). 31 . This parameter is taken into account only if the UL transmitter has active repeater(s). Rp k NIM Rp = NF TX –  NF Rp + G amp – L  r k TX – Rp k   Where. k to calculate the noise rise at the donor transmitter due to active repeaters ( NR Repeaters ).+ ----------------------------------------------- = 10  Log  10   UL UL UL G TMA G TMA G Feeder  ----------------------------------------------------  10 10 10  10 10  10 WithoutTMA And. Then. • NF Feeder is the feeder noise figure. • NF BTS is the BTS noise figure. Atoll converts the noise injection margin ( NIM Rp ) to Watt. it uses the values TX – R p k are the losses between the donor transmitter and the repeater (repeater property). • G Ant – div is the antenna diversity gain (Transmitter property). Downlink Total Losses Atoll calculates total DL losses as follows.3. k Rp k • G amp is the repeater amplification gain (repeater property). • G TMA is the gain due to TMA. • NF Rp is the repeater noise figure. The donor side receives the signal from a donor (transmitter. DL • L TMA is the TMA transmission loss. WiMAX. and channels in WiMAX and LTE documents). • L Feeder is the feeder transmission loss ( L Feeder = L Feeder  I Feeder + L Connector .1 Signal Level Calculation The received signal level (dBm) on a carrier ic from a donor D at a pixel/mobile Mi via a repeater or remote antenna R (see Figure 1. WiMAX. and the related downlink and uplink gains and EIRP. Repeaters increase the coverage area of their donors by retransmitting all the frequencies (TRXs in GSM. A remote antenna is connected to the base station with an optic fibre. remote antennas should be connected to base stations that do not have any antennas. Atoll calculates the signal level received from a repeater or a remote antenna by determining the total downlink and uplink gains (described in "UMTS. the transmission feeder length in metre and the connector transmission losses). and the server side amplifies and re-transmits the received signal. as opposed to a repeater.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment DL DL DL ©Forsk 2015 DL L Total – DL = L TMA + L Feeder + L Misc + L BTS – Conf Where. WiMAX.1 UMTS.6 on page 34) is calculated as follows: Mi Mi R R – Mi C UL = P UL + G Total – L Path – M Shadowing – L Indoor + G Mi –L Mi Mi Mi Here: 32 D • P DL  ic  is the downlink transmission power of a donor D on carrier ic.5 Repeaters and Remote Antennas A repeater receives. and LTE Documents" on page 32). In UMTS. In GSM documents. DL • L Misc are the miscellaneous transmission losses. TD-SCDMA. the total signal D R strength is the sum of the two signals: C DL  ic  + C DL  ic  The received signal level (dBm) from a pixel/mobile Mi at a donor D via a repeater or remote antenna R (see Figure 1. TD-SCDMA. DL 1. CDMA2000. carriers in UMTS. It has a donor side and a server side. A remote antenna. and re-transmits the radiated or conducted RF carrier both in downlink and uplink. repeater. CDMA2000. TD-SCDMA. and LTE documents.1. at locations that would normally require long runs of feeder cable. does not have any equipment and therefore generates neither amplification gain nor noise. The following sections describe how received signal levels. Mi R – L Ant – L Body – L Misc – UL . • P UL is the uplink transmission power of a pixel/mobile Mi. the received signal level from a repeater or a remote antenna is calculated by determining the EIRP transmitted by the repeater or remote antenna (described in "GSM Documents" on page 39). Remote antennas allow you to ensure radio coverage in an area without a new base station.3. 1. where L Feeder . Donors and repeaters may be linked through: • • • Air: Microwave Links: Optical Fibre Links: User-defined or calculated propagation losses User-defined link losses User-defined link losses Remote antennas are antennas located far from the transmitters.5. amplifies. or remote antenna). CDMA2000.6 on page 34) is calculated as follows: R D R – Mi R C DL  ic  = P DL  ic  + G Total – L Path – M Shadowing – L Indoor + G Mi –L Mi Mi Mi R – L Ant – L Body – L Misc – DL If a pixel/mobile Mi receives signals from the donor D and its repeater R. are calculated from a repeater or remote antenna R with a donor D. and LTE Documents 1. I Feeder and L Connector DL DL DL DL DL DL are respectively the feeder loss per metre. CDMA2000 and TD-SCDMA. In Atoll.5. • L BTS – Conf are the losses due to BTS configuration (BTS property).Atoll 3. L Ant . For Mi calculating the interfering signal level from any interferer. • L Path is the path loss (dB) calculated as follows: R–M R–M i i R L Path = L Model + L Ant . R 33 . • Mi L Body is the body loss defined for the service used by the pixel/mobile Mi. with: • L Model is the path loss calculated using a propagation model. see the technology-specific chapters.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 R • G Total is the total gain. L Mi . • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the repeater or remote R • antenna R. Mi L Ant is the terminal antenna attenuation (from antenna patterns) calculated for the pixel/mobile Mi (available in WiMAX and LTE only). Mi For calculating the useful signal level from the best serving cell.3. R • L Misc – DL is the miscellaneous transmission losses defined for the repeater or remote antenna R. L Ant is determined in the direction (H.Atoll 3. is the terminal loss for the pixel/mobile Mi. For more information. user-defined or calculated as explained in "Total Gain Calculation" on page 34.0) from the antenna patterns of the antenna used by Mi. M Shadowing is the shadowing margin. L Ant is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi.G Mi Mi Mi . and L Body are not used in all the calculations. • G • • L Mi Mi is the terminal antenna gain for the pixel/mobile Mi.V) = (0. while the antenna is pointed towards Mi’s best serving cell. • L Misc – UL is the miscellaneous reception losses defined for the repeater or remote antenna R. • L Indoor is the indoor loss. 0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 Figure 1. and LTE: Signal Level Calculation 1.7: Downlink Total Gain: Over-the-Air Repeaters 34 . TD-SCDMA. WiMAX.3.5. CDMA2000.Atoll 3.2 Total Gain Calculation The total gain is calculated from the donor transmitter reference point ( ) to the repeater or remote antenna reference point ( ) as follows: Over-the-Air Repeaters D D D–R R R R G Total = – L Total – DL + G Ant – L Model + G Donor – Ant – LDonor RX – Feeder R R + G Amp – LCov TX – Feeder R + G Cov – Ant Figure 1.6: UMTS.1. this is 0. • LDonor R are the donor-side reception feeder losses for the repeater or remote antenna R. Optical Fibre Link Repeaters and Remote Antennas D–R R R R G Total = – L Fibre + G Amp – LCov TX – Feeder R + G Cov – Ant 35 . • L Model is the path loss between the donor D and the repeater or remote antenna R. • G Ant is the gain of the antenna used at the donor D.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 Here: D • L Total – DL are the total downlink losses of the donor D.8: Downlink Total Gain: Microwave Link Repeaters Here: D–R • L MW are the user-defined microwave link losses between the donor D and the repeater or remote antenna R. the propagation losses between the donor and the repeater or remote antenna are calculated using the ITU 526-5 propagation model. • LCov R are the coverage-side transmission feeder losses for the repeater or remote antenna R. • LCov R R are the coverage-side transmission feeder losses for the repeater or remote antenna R. TX – Feeder • R G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R. For remote antennas. • G Amp is the amplifier gain of the repeater R. This can be user-defined or D D–R calculated using the selected propagation model. RX – Feeder R • G Amp is the amplifier gain of the repeater R. R • G Donor – Ant is the gain of the donor-side antenna used at the repeater or remote antenna R. Microwave Link Repeaters D–R R R R R G Total = – L MW + G Amp – LCov TX – Feeder + G Cov – Ant Figure 1. For remote antennas. If you do not select a propagation model. this is 0. TX – Feeder • R G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R. Secondary antennas are fully supported in the evaluation of the repeater gains.3. when the mobile receiver is located in the vicinity of the repeater. • LCov R R are the coverage-side transmission feeder losses for the repeater or remote antenna R. D R R R R P DL  ic  + G Total  P Max + G Cov – Ant – LCov TX – Feeder Here: • D P DL  ic  is the downlink transmission power of a donor D on carrier ic.3.5. 36 . Atoll considers the highest power.3 Repeater Noise Figure You can define and assign a repeater equipment to each repeater. This noise figure has an impact on the donor total reception losses. see "Transmitter Radio Equipment" on page 30. 1.4 Appendix: Carrier Power and Interference Calculation This section explains how Atoll calculates the received carrier power and interference when a transmitter has a connected repeater. when a mobile receiver is in the vicinity of the donor transmitter. these equipment contain a noise figure which is applied to the repeater they are assigned to.1. Similarly. Similarly.5. the signal from the mobile is received at the donor transmitter as well as its repeater. the signal to/from the donor transmitter would be very weak due to the same reason. 1. In addition to the allowed ranges of gains and powers allowed to each repeater.1. TX – Feeder • R G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R. For remote antennas. TX – Feeder • R G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R. R • G Total is the total downlink gain. When the donor has more than one cell.9: Downlink Total Gain: Optical Fibre Link Repeaters or Remote Antennas Here: D–R • L Fibre are the user-defined optical fibre link losses between the donor D and the repeater or remote antenna R. For information. Repeater Downlink Power Limitation Atoll verifies that the downlink power after amplification is consistent with the repeater equipment limitation. In practice. user-defined or calculated as explained in "Total Gain Calculation" on page 34. this is 0. A mobile receiver receives signal from the donor transmitter as well as its repeater. Atoll does not differentiate between the mobile receiver being in the transmitter coverage area or being in its repeater coverage area. Atoll adds the signals received from the donor transmitter and its repeater to generate a combined pathloss matrix that is associated with the donor transmitter and includes the effect of its repeater.Atoll 3. • G Amp is the amplifier gain of the repeater R. • LCov R R are the coverage-side transmission feeder losses for the repeater or remote antenna R. the signal to/from the repeater would be very weak due to high pathloss between the repeater and the mobile receiver.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 Figure 1. • P Max is the maximum downlink power allowed by the equipment. D G Ant L Total = --------------------------------------------------------------------------------------------------D R  G Total G Ant D . D L Total – DL is the transmission feeder loss of the donor transmitter. As a mobile in the donor transmitter/repeater coverage area is likely to be far from the repeater/donor transmitter coverage area. R – Mi L Path is the path loss between the repeater and the mobile receiver So. is calculated by computing a downlink budget. L Total . Similarly.+ ------------– Mi R – Mi   L DTotal – DL  L DPath  L Path  Since. Txd on a carrier ic. the total carrier power received at the mobile receiver is: D R  G Ant G Total D–R R R D P Rec  ic  = P Rec  ic  + P Rec  ic  = P Pilot  ic    --------------------------------------------- . without considering any shadowing margin or indoor loss. D D P Pilot  ic   G Ant L Total = ----------------------------------------------------------------------------------------------------------------------------D R  G Ant G Total D D L Total – DL  P Pilot  ic    ---------------------------------------------. If we take the case of a CDMA project.+ ------------L Total – DL   --------------------------------------------- – Mi R – Mi   L DTotal – DL  L DPath  L Path  This total path loss depends on the location of the mobile receiver in realistic network scenarios. D – Mi L Path is the path loss between the donor transmitter and the mobile receiver. This implies that we can study the two cases separately without influencing the results much. R P Rec  ic  is the carrier power received at the mobile receiver from the repeater on a carrier ic (in W) D P Pilot  ic  is the pilot power of the donor transmitter on the carrier ic (in W) R G Total is the total gain of repeater linked to a donor transmitter with an air link.3.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 Calculation of Total Path Loss The total pathloss.+ -------------- – Mi R – Mi   L DTotal – DL  L DPath  L Path  Hence. D P Rec  ic  is the carrier power received at the receiver from the donor transmitter on a carrier ic (in W) D P Pilot  ic  is the pilot power of the donor transmitter on the carrier ic (in W) D G Ant is the donor transmitter antenna gain. the respective pathloss value will be very large. the power received at the mobile receiver from the repeater R is: D R  P Pilot  ic   G Total  R P Rec  ic  = -------------------------------------------R – Mi L Path Where. • Case 1: Receiver in Donor Transmitter Coverage Area 37 .Atoll 3. at the mobile receiver can be stated as (for a link over the air): D D  P Pilot  ic   G Ant  D P Rec  ic  = ------------------------------------------D D – Mi  L Total –DL  L Path  Where. D D P Pilot  ic   G Ant L Total = -----------------------------------------------D–R D L Total – DL  P Rec  ic  Therefore. the power received from the donor transmitter. Atoll 3. ---------------------P Rec – UL = ---------------------------------------------.= ---------------------------------------------P Rec – UL  ic  = ---------------------------------------------D D D – Mi  L Total – UL  L Total   L Total – UL  L Path  Where. the quality level at the transmitter on a traffic channel is: E b C W  ---= --. so the term ------------R – Mi L Path D – Mi L Total = L Path Considering this total pathloss value.3.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 R G Total R – Mi . Mi P Output  ic  is the transmitted power from the mobile terminal (in W) D L Total – UL is the reception feeder loss of the transmitter Calculation of Eb/Nt Uplink In the uplink.= ---------------------------------------R R R G Total  L Total – DL  D  G Total  -----------------------------------------   L Total – DL   ------------R – Mi – Mi   L RPath  L Path D D D R  P Pilot  ic   G Ant   P Pilot  ic   G Total  D . This implies that: L Path is likely to be very high. Mi P Output  ic  is the transmitted power from the mobile terminal on the carrier ic (in W) D L Total – UL is the reception feeder loss of the transmitter • Case 2: Receiver in Repeater Coverage Area D G Ant D – Mi . This implies that: L Path is likely to be very high.= -------------------------------------------P Rec – DL  ic  = ------------------------------------------D R – Mi  L Total – DL  L Total   L Path  Mi R D Mi D  P Output  ic   G Total  L Total – DL  P Output  ic   G Ant  D . ---- N t UL I R Where. so the term --------------------------------------------D D – Mi  L Total – DL  L Path  D D G Ant G Ant L Total = --------------------------------------------------. Mi D P Output  G Ant C = P Total – UL = --------------------------------------D L Total – UL  L Total I = I Total + N 0 38 . the total received power in the uplink and in the downlink can be stated as: D D D D  P Pilot  ic   G Ant   P Pilot  ic   G Ant  D P Rec – DL  ic  = ------------------------------------------.= --------------------------------------------D D D – Mi  L Total – DL  L Total   L Total – DL  L Path  Mi D Mi D  P Output  ic   G Ant   P Output  ic   G Ant  D .can be ignored. C is the carrier power received from the mobile terminal (in W) I is the total interference (in W) W is the spreading bandwidth (Hz) R is the effective service throughput in the uplink (bits/s) (W/R is the service processing gain in the uplink) C and I are both evaluated at the same reference point. which is the entry of BTS using the following formulas.can be ignored.= ------------------------------------------------R – Mi D D  L Path  L Total – UL  L Total – UL  L Total  Where. the total signal D R strength is the sum of the two signals: C DL  tt  + C DL  tt  Here: R • EIRP DL  tt  is the effective isotropic radiated power of the repeater or remote antenna R on the TRX type tt. D N 0 = NF  K  T  W Where.2 GSM Documents 1.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 Where. • L Indoor is the indoor loss. D NF is the noise figure of the transmitter equipment at the reference point. is the miscellaneous transmission losses defined for the repeater or remote antenna R. • G Mi Mi is the terminal antenna gain for the pixel/mobile Mi. • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the repeater or remote R • antenna R.5. the entry of the BTS K is Boltzman constant T is the ambient temperature (in K) Hence N 0 = NF BTS KTW 1. P  tt  is the power offset defined for the TRX type tt. for each mobile terminal Mi. • L Path is the path loss (dB) calculated as follows: R – Mi R – Mi R L Path = L Model + L Ant . the downlink transmission power of a donor D on carrier ic.e.Atoll 3.  I Total = Mi Mi D  P Output  G Ant   --------------------------------------  L DTotal – UL  L Mi Total And.2. • L • R L Misc – DL is the terminal loss for the pixel/mobile Mi. It can be • • user-defined or calculated as explained in "EIRP Calculation" on page 40. M Shadowing is the shadowing margin. 39 . i.10 on page 40) is calculated as follows: R R – Mi R C DL  tt  = EIRP DL  tt  – P  tt  – L Path – M Shadowing – L Indoor + G Mi –L Mi R – L Misc – DL If a pixel/mobile Mi receives signals from the donor D and its repeater R.5. I Total is the sum of the signals received from mobile terminals inside the same cell and those outside (in W) N 0 is the transmitter equipment thermal noise (in W) Therefore.1 Signal Level Calculation The received signal level (dBm) on a TRX type tt from a donor D at a pixel/mobile Mi via a repeater or remote antenna R (see Figure 1.3. with: • L Model is the path loss calculated using a propagation model. • L Total – DL are the total downlink losses of the donor D.10: GSM: Signal Level Calculation 1.2 EIRP Calculation D The EIRP of a repeater or remote antenna R is calculated at the repeater or remote antenna reference point ( ) w. r.2.3.11: EIRP: Over-the-Air Repeaters Here: 40 D • P DL is the downlink transmission power of the donor D.5.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 s Figure 1.Atoll 3. P DL at the donor reference point ( ) as follows: Over-the-Air Repeaters R D D D D–R R R EIRP DL  tt  = P DL – L Total – DL + G Ant – L Model + G Donor – Ant – LDonor RX – Feeder R Figure 1. D R + G Amp – LCov TX – Feeder R + G Cov – Ant . t. If you do not select a propagation model.3. • LCov R are the coverage-side transmission feeder losses for the repeater or remote antenna R. • L MW are the user-defined microwave link losses between the donor D and the repeater or remote antenna R. R • G Donor – Ant is the gain of the donor-side antenna used at the repeater or remote antenna R. TX – Feeder • R G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R. • G Amp is the amplifier gain of the repeater R.12: Downlink Total Gain: Microwave Link Repeaters Here: D • P DL is the downlink transmission power of the donor D. Secondary antennas are fully supported in the evaluation of the repeater gains. This can be user-defined or D–R calculated using the selected propagation model. this is 0. • L Model is the path loss between the donor D and the repeater or remote antenna R. For remote antennas. the propagation losses between the donor and the repeater or remote antenna are calculated using the ITU 526-5 propagation model.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 D • G Ant is the gain of the antenna used at the donor D. Microwave Link Repeaters D D–R R R R EIRP DL  tt  = P DL – L MW + G Amp – LCov TX – Feeder R + G Cov – Ant Figure 1. • LDonor R are the donor-side reception feeder losses for the repeater or remote antenna R. For remote antennas. • LCov D–R R R are the coverage-side transmission feeder losses for the repeater or remote antenna R. this is 0. RX – Feeder R • G Amp is the amplifier gain of the repeater R.Atoll 3. TX – Feeder • R G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R. Optical Fibre Link Repeaters and Remote Antennas D D–R R R R EIRP DL  tt  = P DL – L Fibre + G Amp – LCov TX – Feeder R + G Cov – Ant 41 . It is the absolute horizontal angle at which the donor-side antenna of the repeater should be pointed in order to be aligned with the donor antenna.1 Azimuth This is the angle at which the donor antenna is situated with respect to the North at the repeater or remote antenna. • G Amp is the amplifier gain of the repeater R. TX – Feeder • R G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R.5.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 Figure 1.5.14: Angle from North (Azimuth) 1. 1. Repeater Downlink Power Limitation Atoll verifies that the EIRP after amplification is consistent with the repeater equipment limitation. 42 .3 Donor-side Parameter Calculations 1. For remote antennas. this is 0. downtilt angles are considered positive and uptilt angles negative.5. R R R R EIRP DL  tt   P Max + G Cov – Ant – LCov TX – Feeder Here: R • EIRP DL  tt  is the effective isotropic radiated power of the repeater R on the TRX type tt.2 Mechanical Downtilt This is the tilt angle for the repeater’s donor-side antenna. This angle is measured clock-wise as shown in the figure below. TX – Feeder R • G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R.13: Downlink Total Gain: Optical Fibre Link Repeaters or Remote Antennas Here: D • P DL is the downlink transmission power of the donor D. • LCov D–R R R are the coverage-side transmission feeder losses for the repeater or remote antenna R. • L Fibre are the user-defined optical fibre link losses between the donor D and the repeater or remote antenna R.3. • P Max is the maximum downlink power allowed by the equipment. • LCov R R are the coverage-side transmission feeder losses for the repeater or remote antenna R. which ensures that it points towards the donor antenna in the vertical plane. As a general rule. Figure 1.3.3.Atoll 3. • H Ant is the height of the antenna of the donor D. The signal processor dynamically applies weights to each element of the adaptive antenna system to create array patterns in real-time. • Conventional Beamformer: The Conventional Beamformer smart antenna model performs dynamic beamforming in downlink and uplink as explained in "Downlink Beamforming" on page 46 and "Uplink Beamforming" on page 48.15: Positive/Negative Mechanical Downtilt Since this parameter depends on the difference of heights/altitudes between the donor transmitter and the repeater.. Smart antenna results are later on used in coverage prediction calculations. These algorithms are based on optimization methods such as the minimum mean square error method.17 on page 44. and beamforming and interference cancellation in uplink using the minimum mean square error algorithm as explained in "Uplink Beamforming and Interference Cancellation (MMSE)" on page 49. along with smart signal processing.3. i. • D D D–R is the distance between the antenna of the donor D and the antenna of the repeater or remote antenna R.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 Figure 1. Here: R • H Donor – Ant is the height of the donor-side antenna of the repeater or remote antenna R. Smart antenna results are later on used in coverage prediction calculations. the corresponding tilt angle can be found out and applied using the Calculate button. and maximize useful signal reception. • Optimum Beamformer: The Optimum Beamformer smart antenna model performs dynamic beamforming in downlink as explained in "Downlink Beamforming" on page 46. 43 . to locate and track various types of signals. Adaptive algorithms can also be used in order to minimize the interference received by the cells. the useful signal. 1. Example Figure 1.16: Tilt Angle Computation The tilt angle repeater’s donor-side antenna in the above figure would be: R D  H Donor – Ant – H Ant R - T Donor – Ant = atan  ------------------------------------------D–R   D As obvious.6 Beamforming Smart Antenna Models Adaptive antenna systems use more than one antenna elements. Beamforming smart antennas dynamically create antenna patterns with a main beam pointed in the direction of the user being served. such as the one shown in Figure 1.e. These smart antenna models support linear adaptive array systems. If the height/altitude of the antenna is modified. to dynamically minimize interference. this angle will be negative for uptilts and positive for downtilts of the antenna. The following beamforming smart antenna models are available in Atoll.Atoll 3. it can be automatically calculated in the repeater’s Donor side properties. respectively. 1 Definitions Name Value Unit Description E SA Smart antenna model parameter None Number of smart antenna elements  Calculation parameter Degrees Angle of arrival for the useful signal  Calculation parameter Degrees Angle at which the smart antenna effect is calculated d  --. e e wn e 44 2 j  -----.3.Atoll 3. 1.1..6.1.6. where  is the wavelength of the signal 2 m Distance between two adjacent antenna elements 1.17: Linear Adaptive Antenna Array In the following explanations.  E SA – 1 d  sin   2 – j  -----.. 2d  sin   T 2 j  -----. 1.1 Definitions and Formulas The tables in the following subsections list the parameters and formulas used in beamforming smart antenna models. d is the distance between two adjacent antenna elements. we assume: • There are a total of E SA elements in the adaptive antenna system.2 Downlink Beamforming Name Value Unit Description gn    Smart antenna model parameter None Gain of a single element None Steering vector for the direction of  None Complex smart antenna weight None Array correlation matrix for a given user direction  None Smart antenna gain in any direction  S 1 e 2 j  -----.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 Figure 1.. nd  sin   – j    n  sin  with d =  --2 H R S  S G SA    g n     S   R   S  = g n     S   S   S   S  = g n     E SA H H H 2 . • •  is the angle at which we want to calculate the smart antenna gain. d  sin   e  . •  is the angle of arrival for the useful signal.6. 1.1.3 Uplink Beamforming Name Value Unit Description w S ----------E SA None Vector of ESA complex weights for the conventional beamformer None Total noise correlation matrix None Thermal noise correlation matrix None Interference correlation matrix J RN Rn + RI = 2 n I+  pj  Sj  Sj H j=1 2 n  I Rn J  pj  Sj  Sj RI H j=1 H PN w  RN  w W Total uplink noise power P p   w  S   S   w = p   E SA H W Total power received from the served user CINR UL P p   E SA -----.3.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 Name Value Unit Description None Downlink array correlation matrix for iteration k None Average downlink array correlation matrix over a simulation (K iterations) J  pj  Rj Rk j=1 K --1.= -------------------------H PN w  RN  w None C/(I+N) in the uplink (WiMAX) Q UL P p   E SA -----.4 Uplink Beamforming and Interference Cancellation (MMSE) Name Value Unit Description ˆ w   RN  S None Vector of ESA complex weights for the optimum beamformer  E SA ----------------------------H –1 S  RN  S None MMSE optimization constant –1 45 .6. K R Avg  Rk k=1 1.Atoll 3. K Avg  RN k k=1 I UL    H w  RN 2 Avg  w – n 2 NR UL    I UL    +  n --------------------------2 n 1.6.= -------------------------H PN w  RN  w None Signal quality in the uplink (TD-SCDMA) G SA E SA None Uplink smart antenna beamforming gain in the direction of the served user W Average noise correlation matrix W Uplink interference None Angular distribution of uplink noise rise SA H K RN --1. = p   S   R N  S  ˆP PN N None Signal quality in the uplink (TD-SCDMA) G SA S   I  S  = E SA None Uplink smart antenna beamforming gain in the direction of the served user W Average inverse noise correlation matrix W Uplink interference None Angular distribution of uplink noise rise SA H 2 –1 H 2 –1 H 2 K –1 RN Avg --1. w n . to each antenna element in order to form a beam towards the served user.2 Downlink Beamforming Figure 1.= p   S   R N  S  PN Pˆ N None C/(I+N) in the uplink (WiMAX) Q UL P H Pˆ –1 -----. K  RN –1 k k=1 I UL    E SA 2 -----------------------------------. The smart antenna processor applies complex weights.= -----.18: Downlink Beamforming Beamforming dynamically creates a beam towards the served user.6.3. The magnitude of these complex weights is 46 .Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 Name Value Unit Description None Total noise correlation matrix None Thermal noise correlation matrix None Interference correlation matrix W Total uplink noise power (optimum beamformer) W Total power received from the served user (optimum beamformer) J RN 2 n Rn + RI = I+  pj  Sj  Sj H j=1 2 n  I Rn J RI  pj  Sj  Sj H j=1 Pˆ N   S  RN  S Pˆ  p     S  RN  S  CINR UL P H Pˆ –1 -----.– n H –1 S  RN  S Avg 2 NR UL    I UL    +  n --------------------------2 n 1.= -----. the powers fed to each antenna element are combined for transmission. J is the number of served mobiles during the iteration. Since each element transmits the same input power. p j is the EIRP transmitted towards the mobile j... The smart antenna is able to form the beam only in the horizontal plane. g n    is the gain of the nth antenna element in the direction  . The smart antenna gain in dB will be G SA    = 10  Log  G SA     . Atoll calculates a moving average of the array correlation matrices calculated in each iteration. w n = e .. given by: H R = S  S For the direction of the served user. therefore.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 set to 1.e. which is the complex conjugate transpose of a matrix. Therefore.. the complex weight at any nth antenna element can be given by: wn = e 2 – j  -----. i. at the angle of arrival  is given by: S  = 1 e 2 j  -----. the vertical pattern is assumed to remain the same. 2d  sin    . and R j is the array correlation matrix for the mobile j. representing the complex weights for forming a beam towards the served user.. At the end of a simulation with K iterations. The sum of these array correlation matrices for all the users served in one iteration k is calculated as follows: J Rk =  pj  Rj j=1 Where R k for any cell is the downlink array correlation matrix for iteration k. e T 2 j  -----. Power Combining Gain Cell transmission power is fed to each antenna element of the smart antenna system. i. S  . this results in a gain due to power combination. therefore.  E – 1 d  sin  SA  Where T represents the transpose of a matrix. nd  sin   – j    n  sin  In Atoll. Atoll calculates the smart antenna gains (array correlation matrix R  ) for each served mobile in a cell’s coverage area in each iteration. the average downlink array correlation matrix for any cell is given by: K 1 R Avg = --..3. the smart antenna gain is calculated as follows: H H H 2 G SA    = g n     S   R   S  = g n     S   S   S   S  = g n     E SA The smart antenna gain includes the gain of the beamforming as well as the gain of power combination.e.Atoll 3. i. and R  is the array correlation matrix for a given user direction  . 2 The smart antenna gain in any direction  can be given by: H G SA    = g n     S   R   S  Where H represents the Hilbert transform. The beamforming is performed using only the phase of the complex weights. The steering vector.  . d =  --. K  Rk k=1 47 . Additional Processing in Monte Carlo Simulations During Monte Carlo simulations.e. d  sin   e 2 j  -----.  . p j is the power received by one element of the smart antenna from the jth interfering mobile.= -------------------------H PN w  RN  w In WiMAX.Atoll 3.3 Uplink Beamforming Figure 1. The total noise received in the uplink. R N .3. the total power received from the served user is given by: H H P  = p   w  S   S   w = p   E SA Where p  is the power received by one element of the smart antenna from the served user. I is the identity matrix. In TD-SCDMA. The total noise correlation matrix is the sum of the thermal noise correlation matrix R n . including thermal noise and interference from all uplink interferers. is stored in a total noise correlation matrix. given by: J RN = Rn + RI = 2 n I+  pj  Sj  Sj H j=1 J 2 Where R n =  n  I and R I =  pj  Sj  Sj H j=1 2  n is the thermal noise power. S j is the steering vector in the direction of the jth interfering mobile. i. J is the total number of interfering mobiles.6. The total noise power. the uplink signal quality is calculated by: p   E SA P SA Q UL = -----. interference and thermal noise.= -------------------------H PN w  RN  w 48 . w is given by: S w = ----------E SA Where S  is the steering vector in the direction of the served user. and the interference correlation matrix R I . the C/(I+N) in the uplink is then calculated by: P p   E SA CINR UL = -----.  . received by a cell is given by: H PN = w  RN  w And..0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 1.e.19: Uplink Beamforming Let w represent the vector of ESA complex weights for the beamformer. G SA = E SA .0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 From the above equation. and  n is the thermal noise power. Additional Processing in Monte Carlo Simulations The noise correlation matrix R N for each iteration k includes the effect of the matrix calculated for the previous iteration. In TD-SCDMA.e. These weights do not try to fully cancel E SA – 1 interference signals. we can determine the uplink smart antenna beamforming gain in the direction of the served user. The average of the noise correlation matrices is calculated as follows: K RN Avg 1 = --. The interference can be isolated from the thermal noise and can be calculated for any direction using the formula. and R N k is the noise correlation matrix of the kth iteration. S  is the steering 2 vector in the direction  . the angular distribution of the uplink noise rise is given by: 2 I UL    +  n NRUL    = --------------------------2 n 1. A simple null steering beamformer can cancel the interference from the most interfering E SA – 1 interfering mobiles.4 Uplink Beamforming and Interference Cancellation (MMSE) The optimum beamformer uses the Minimum Mean Square Error algorithm in the uplink in order to cancel interference. The Minimum Mean Square Error algorithm optimizes the useful signal as well as maximizes the signal quality.3. w ˆ is given by: Let w 49 . H I UL    = w  R N 2 Avg  w – n Where I UL    is the interfering signal in the direction  . but rather try to reduce the overall received interference as much as possible. Figure 1. the uplink load is calculated from the average noise correlation matrix. ESA is the number of smart antenna elements. It calculates the optimum smart antenna weights using the knowledge of directions and power levels of interference. This angular distribution of the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX) can be stored in the Cells table. which equals the number of smart antenna elements. The optimum beamforming method used in Atoll overcomes this limitation. K  RN k k=1 Where R N Avg is the average of the noise correlation matrices of all the iterations from k = 1 to K. In WiMAX. which is calculated from the noise correlation matrix obtained at the end of the last iteration of a Monte Carlo simulation. i.6. The result is the angular distribution of the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX).Atoll 3.20: Uplink Adaptive Algorithm ˆ represent the vector of ESA complex weights for the beamformer.. The result is the angular distribution of the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX).   . J is the total number of interfering mobiles. including thermal noise and interference from all uplink interferers.e. i. which is calculated from the inverse of the noise correlation matrix obtained at the end of the last iteration of a Monte Carlo simulation. given by: J RN = Rn + RI = 2 n I+  pj  Sj  Sj H j=1 J Where R n = 2 n  pj  Sj  Sj  I and R I = H j=1 2  n is the thermal noise power.Atoll 3. RSCP TCH – UL = p   S   I  S  = p   E SA H In WiMAX. received by a cell is given by: H 2 –1 Pˆ N =    S   R N  S  And. Additional Processing in Monte Carlo Simulations –1 The inverse noise correlation matrix R N for each iteration k includes the effect of the matrix calculated for the previous iteration. which is a constant value for a given useful signal that optimizes the beamformer weights.3. The total noise power.  . This angular distribution of the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX) can be stored in the Cells table. The average of the inverse noise correlation matrices is calculated as follows: 50 . RSCP TCH – UL (TD-SCDMA) or C UL (WiMAX) can be calculated from the above equation by considering the interference and –1 noise to be null. p j is the power received by one element of the smart antenna from the jth interfering mobile. Hence. the uplink signal quality is calculated by: H Pˆ SA –1 Q UL = -----. This gives: H In TD-SCDMA. the C/(I+N) in the uplink is then calculated by: H Pˆ –1 CINR UL = -----. i.. the total power received from the served user is given by: 2 H 2 –1 Pˆ  = p       S   R N  S   Where p  is the power received by one element of the smart antenna from the served user. R N = I . G SA = E SA . C UL = p   S   I  S  = p   E SA From the above equation.  .0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 ˆ =    R –N1  S  w Where S  is the steering vector in the direction of the served user.= p   S   R N  S  ˆP N From the above equation. the uplink smart antenna beamforming gain equals the number of smart antenna elements. S j is the steering vector in the direction of the jth interfering mobile. and the interference correlation matrix R I . we can determine the uplink smart antenna beamforming gain in the direction of the served user. Atoll is able to calculate an average of the smart antenna interference-cancellation effect. It is given by the equation: E SA   = ----------------------------H –1 S  RN  S –1 R N is the inverse of the total noise correlation matrix.e. The total noise correlation matrix is the sum of the thermal noise correlation matrix R n . I is the identity matrix..= p   S   R N  S  Pˆ N In WiMAX. In TD-SCDMA. horizontal. ESA is the number of smart antenna elements. and vertical attenuations of the beams of the GOB. and horizontal and vertical SA SA SA SA attenuations. consists of more than one directional antenna pattern (beam) in different directions. L Beam . The following example shows how Atoll calculates the GOB gains and losses. During the simulations. the angular distribution of the uplink noise rise is given by: 2 I UL    +  n NRUL    = --------------------------2 n 1. G UL . In words. and L Beam are the gains.Atoll 3. E SA 2 I UL    = -----------------------------------. Atoll determines the most suitable beam from the GOB for each user served by the smart antenna. Each beam of a GOB has a different azimuth so that the GOB as a whole covers an entire sector. and whose horizontal patterns are pointed towards different directions as shown in the figure below: 51 . K  RN –1 k k=1 –1 Where R N –1 Avg is the average of the inverse noise correlation matrices of all the iterations from k = 1 to K. The most suitable beam (best beam) is the one which provides the highest gain towards the served user: BeamBest = Beam H H V Max  G Beam – L Beam – L Beam V Where G Beam . In WiMAX. S  is the steering 2 vector in the direction  . and  n is the thermal noise power. called GOB. the uplink load is calculated from the average inverse noise correlation matrix. and L UL ) are determined from the selected best beam.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 K –1 RN Avg 1 = --. Example: Let us assume a GOB with 5 beams that have the same vertical patterns.– n H –1 S  RN  S Avg Where I UL    is the interfering signal in the direction  . The interference can be isolated from the thermal noise and can be calculated for any direction using the formula. The gains and losses of the GOB ( G DL . L DL .3.7 Grid-of-Beams Smart Antenna Model A grid-of-beams smart antenna. the best beam is the one among all the beams of a GOB that has the highest difference between gain. and R N k is the inverse noise correlation matrix of the kth iteration. In TD-SCDMA. 60 .15 0.21: Grid Of Beams Modelling Let us assume that all the beams and the main antenna have the same 18 dBi gain. Atoll determines the best beam. the total gain of the beam at 60° is the highest.22: GOB Modelling .21 15 18 . SA SA H V G UL = 18 dB and L UL = L Beam + L Beam = 17.21 dB If this beam has been selected in the uplink.2. and the vertical attenuation at the user location is 15 dB. which has the highest gain towards  . If this beam has been selected in the downlink.15 -57 60° 18 2.Determination of the Best Beam In our example.15 -57 -60° 18 60 15 18 .15 -57 30° 18 60 15 18 . which is also the same for all the beams because we assume that the vertical patterns are the same.60 . Therefore this beam is selected as the best beam.15 -57  Transmitter Centre of the pixel where the served user is located  Angle between the user and the transmitter azimuth Figure 1. If the user is located at  = 70 azimuth.21 dB 52 .Atoll 3. as shown in the figure below.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 Figure 1.79 -30° 18 60 15 18 . SA SA H V G DL = 18 dB and L DL = L Beam + L Beam = 17.3. as follows: Beam Gain (dBi) Horizontal Vertical Attenuation (dB) Attenuation (dB) G Beam – L Beam – L Beam Total Gain (dB) H V 0° 18 60 15 18 .60 .21.60 . Let us assume that the adaptive beam and the main antenna have the same 18 dBi gain. For each smart antenna equipment based on statistical modelling. and reads the smart antenna C/ I gain defined for the Probability = 1 – TProb SA corresponding to the spreading angle.  Spread . Urban and dense urban clutter types introduce more multipath and spread the signal at a wider angle than an open or rural clutte type. Once you have assigned the spreading angles to clutter classes. The following example shows how Atoll calculates the adaptive beam gains and losses. it reads the probability threshold from the smart antenna properties. Different clutter types have different spreading effects on the propagation of radio waves. and the vertical attenuation at the user location is 15 dB. the gain and losses of the adaptive beam at  are: SA SA H V G UL = 18 dB and L UL = L Beam + L Beam = 15 dB H In fact. During the simulations. 1. this adaptive beam is oriented in the direction of each served user in order to model the effect of the smart antenna.3. called a beam because of its highly directional shape. as shown in the figure below:  Transmitter Centre of the pixel where the served user is located  Angle between the user and the transmitter azimuth Figure 1.23: Adaptive Beam Modelling . In Atoll. Example: Let us assume an adaptive beam smart antenna selected for a transmitter along with a main antenna. as well as for calculating the uplink interfering signals received at transmitter when decoding signal received from the served user.Determination of the Best Beam If the adaptive beam smart antenna is selected in the downlink.9 Statistical Smart Antenna Gain Model A statistical modelling approach is also available in Atoll which can be used to model the effect of smart antennas through C/ I gains. SA SA SA The adaptive beam gains ( G DL and G UL ) are the antenna gains defined for the beam. it reads the spreading angle from the clutter class properties. the gain and losses of the adaptive beam at  are: SA SA H V G DL = 18 dB and L DL = L Beam + L Beam = 15 dB If the adaptive beam smart antenna is selected in the uplink. in the smart antenna equipment based on the statistical model. These values are used in interference calculation to determine the downlink interfering signal due to transmission towards the served user. Atoll determines the clutter class of the served user. as the ideal beam steering algorithm steers the beam towards the served user. and the adaptive beam losses ( L DL and SA H V L UL ) are the horizontal and vertical pattern attenuations L Beam + L Beam towards the user direction. TProb SA . If the user is located at  = 60 azimuth. Atoll reads the clutter class in which the served user is located to determine the spreading angle. this is modelled using a single antenna pattern. L Beam = 0 . You can assign a spreading angle to each clutter class in your document. you can enter the C/I gains and their cumulative probabilities for each spreading angle. To find the smart antenna gain.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment AT330_TRR_E1 1. 53 .8 Adaptive Beam Smart Antenna Model An adaptive beam smart antenna is capable of steering a given antenna pattern towards the direction of the served signal. You can create smart antenna equipment in Atoll based on the statistical approach by providing C/I gains and their cumulative probabilities for different spreading angles. you can set a probability threshold.Atoll 3. 7196 dB .Atoll 3. 54 . If a gain for the exact probability value of 20% is not defined.6298 dB and G SA Prob = 20. Atoll linearly interpolates the gain value from the two surrounding values. Their are no losses for this type of smart antenna equipment. If G SA Prob = 19% = 4.0 Technical Reference Guide for Radio Networks Chapter 1: Antennas and Equipment ©Forsk 2015 The following example shows how Atoll calculates the statistical C/I gains and losses.3.6941 dB The smart antenna gains are the same for uplink and downlink. The smart antenna equipment SA SA has TProb = 80 % . Example: Let us assume that the served user is located at a an urban clutter class with  Spread = 10 . then G SA Prob = 20% = 4. Negative values of C/I gains are considered as losses. Atoll will read the smart antenna C/I gain G for Prob = 20 % .4% = 4. 526-5 Propagation Model" on page 76 • "ITU-R P.1546-2 Propagation Model" on page 80 • "Sakagami Extended Propagation Model" on page 84 • "Free Space Loss" on page 86 • "Diffraction" on page 86 • "Shadow Fading Model" on page 90 • "Path Loss Matrices" on page 103 • "File Formats" on page 107 .370-7 Propagation Model" on page 76 • "Erceg-Greenstein (SUI) Propagation Model" on page 78 • "ITU-R P.Chapter 2 Radio Propagation This chapter covers the following topics: • "Path Loss Calculation Prerequisites" on page 57 • "List of Default Propagation Models" on page 62 • "Okumura-Hata and Cost-Hata Propagation Models" on page 63 • "ITU 529-3 Propagation Model" on page 64 • "Standard Propagation Model (SPM)" on page 65 • "WLL Propagation Model" on page 75 • "ITU-R P. 3.Atoll 3.0 Technical Reference Guidefor Radio Networks © Forsk 2015 56 . 1: Digital Terrain Model Four points (hence. Propagation models are mathematical representations of the average loss in signal strength over distance. Diffraction loss and shadow fading margins are added to this average loss in order to get more precise path loss values. Monte Carlo simulations. Path loss matrices are calculated for each transmitter and their results used in other calculations (coverage predictions. Figure 2. Space between points is defined by pixel size P (in metre). This chapter describes the various propagation models available in Atoll.3. point analysis. 57 . The method used to evaluate site altitude is based on a bilinear interpolation. Pixel size must be the same in both directions. Abscissa and ordinate axes are respectively oriented in right and downwards directions.1 Ground Altitude Determination Atoll determines reception and transmission site altitude from Digital Terrain Model map. With the Standard Propagation Model.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 2 Radio Propagation Path loss calculations are carried out between a transmitter and a receiver using propagation models and other calculations related to radio wave propagation such as diffraction and shadow fading. etc. you can choose between radial or systematic. 2.1.Atoll 3. and other radio wave propagation phenomena such as diffraction and shadow fading.1 Path Loss Calculation Prerequisites 2. The method of calculation may differ depending on the analysis being performed: Analysis type Receiver position Calculation Profile extraction Result Coverage predictions Centre of each bin inside the calculation area Based on path loss matrices Radiala One value for the bin’s surface area Point analysis (Profile) Anywhere Real-time Systematic Different values inside a calculation bin Point analysis (other) Anywhere inside the calculation areas Based on path loss matrices Radiala One value for the bin’s surface area Monte Carlo simulations Mobile coordinates Based on path loss matrices Radiala One value at the mobile location Subscriber lists Subscriber coordinates Real-time Radiala One value at the subscriber location a. four altitude values) are necessary to describe a “bin”. Therefore. DTM files provide altitude value z (in metre) on evenly spaced points.). The first point given in the file corresponds to the centre of the top-left pixel of the map (northwest point georeferenced by Atoll). these points are bin vertices. a DTM file that contains N x N bins requires N2 points (altitude values). DEM (Digital Elevation Model) is the same as Digital Terrain Model (DTM). Figure 2. By definition. This line respectively intersects (S’1.S’’1) and (S’2. Atoll knows the altitudes of four bin vertices. Let us suppose a site S located inside a bin.4: Ground Altitude Determination . In literature. S’’1. Atoll performs a second linear interpolation to evaluate the S altitude.3 3. S’2 and S’’2.3.Atoll 3. DEM refers to the altitude above sea level including ground and clutter. from the DTM file (centre of each DTM pixel).2 2.5: Ground Altitude Determination . S’1.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Figure 2. Atoll determines the S1 and S2 altitudes using a linear interpolation method. DEM and DTM do not always have the same meaning. Figure 2.2: Schematic view of a DTM file • In Atoll. Figure 2. Atoll draws a vertical line through S. while DTM refers to the ground altitude above sea level alone. 58 .3: Ground Altitude Determination .1 1. S’’2) lines at S1 and S2. S’’1.1.Atoll 3. 59 . Atoll reads clutter heights of four points around the site.3.8: Clutter Height If you do not have any clutter height file. clutter height is an average height related to a clutter class.1 Clutter Classes Atoll uses clutter classes file to determine the clutter class. therefore Atoll takes the S”2 clutter height as clutter height of S. S’2 and S’’2. Figure 2.2. In the clutter heights file. S’1.4 2. clutter height of a site is the height of the nearest point in the file.7: Clutter Classes Atoll supports a maximum of 255 clutter classes (8 bits/pixel).2 Clutter Determination Some propagation models need clutter class and clutter height as information at receiver or along a transmitter-receiver profile. Atoll takes clutter height information in clutter classes file.2 Clutter Heights To evaluate the clutter height. Pixel size must be the same in both directions. Here.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Figure 2. The first point given in the file corresponds to the centre of the top-left pixel of the map (northwest point geo-referenced by Atoll.1. Atoll uses clutter heights file if available in the . It is possible to specify an average height for each clutter class in Atoll. Clutter class files provide a clutter code per bin. Bin size is defined by pixel size P (in metre). Abscissa and ordinate axes are respectively oriented in right and downwards directions. In this case.1. the nearest point to S is S”2.2. A clutter classes file file that contains N x N bins requires N2 code values. 2.6: Ground Altitude Determination . 2. The clutter classes map is a grid representing the ground with each bin assigned a clutter class code corresponding to its clutter type. Figure 2. Example: Let us suppose a site S.atl document. Profiles can be based on DTM only or on DTM and clutter both.3.9: Radial calculation method Transmitter location Radials (Atoll extracts a geographic profile for each radial) Centres of bins located on the calculation border Receiver location Figure 2. depending on the selected propagation model. Method 1: Radial Extraction Atoll draws radials from the site (where transmitter is located) to each calculation bin located along the transmitter calculation area border. In other words.1. the receiver may be located at the centre of a calculation bin (coverage predictions) or anywhere within a calculation bin. Figure 2. 60 .10: Site-bin centre profile Depending on the calculation being carried out.3 Geographic Profile Extraction Geographic profile extraction is needed in order to calculate diffraction losses. Atoll determines a geographic profile between site and each bin centre.Atoll 3. Atoll uses the profile nearest to the receiver for calculations (the receiver is assumed to be located on the profile).0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 2. 11: Radial calculation method Transmitter location Geographic profile Receiver location 2.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Method 2: Systematic Extraction Atoll extracts a precise geographic profile between the site and the receiver. Example 1 (Using the Standard Propagation Model) A DTM map with a 40 m resolution and a clutter heights map with a 20 m resolution are available. Atoll uses the bilinear interpolation method described in "Ground Altitude Determination" on page 57. ground altitude and clutter height. Atoll uses the clutter classes map to determine clutter height. It means that Atoll will extract geographic information. The profile resolution will be 20 m. Every 20 m. To get ground altitude every 20 m. Atoll takes the clutter height of the nearest point every 20 m. No clutter height file has been imported in the document. Example 2 (Using the Standard Propagation Model) A DTM map with a 40 m resolution and a clutter classes map with a 20 m resolution are available. ground altitude and clutter height. It means that Atoll will extract geographic information. • If the propagation model uses both DTM and clutter heights along the profile. every 20 m. To get ground altitude every 20 m. The profile resolution will be 20 m. the profile resolution will be the highest of the two. Example (Using the Cost-Hata Propagation Model) 61 .4 Resolution of the Extracted Profiles Geographic profile resolution depends on resolution of geographic data used by the propagation model (DTM and/or clutter). the profile resolution will be the highest resolution among the DTM files. • If the propagation model uses only DTM along the profile. it determines clutter class and takes associated average height. every 20 m.3.Atoll 3. Clutter heights are read from the clutter heights map. The selected profile resolution does not depend on the geographic layer order. Figure 2. Atoll uses the bilinear interpolation method described in "Ground Altitude Determination" on page 57.1. broadcast Fixed receivers WLL Fixed receivers WLL. Microwave links. clutter) L(d. every 25 m. HRx) (per environment) Diffraction loss L(d. HRxeff.2 List of Default Propagation Models Propagation models available in Atoll are listed in the table below along with their main characteristics. Atoll uses the bilinear interpolation method described in "Ground Altitude Determination" on page 57. f. It means that Atoll will extract geographic information. The profile resolution does not depend on the geographic layer order in the Geo tab of the Explorer window. the geographic layer order has influence on the usage of the data. Diff loss. For example. HTxeff. f. HRx) (per environment) Diffraction loss Diffraction calculation method Deygout (3 obstacles) Epstein-Peterson (3 obstacles) Deygout corrected (3 obstacles) Millington (1 obstacle) Deygout (1 obstacle) Deygout (1 obstacle) Deygout (1 obstacle) 62 . Atoll will use DTM 1 for extracting the profile where DTM a is available and it will use DTM 2 elsewhere. The profile resolution will be 25 m. Propagation model ITU 370-7 (Vienna 93) ITU 1546 ITU 526-5 WLL Frequency band 100-400 MHz 30-3000 MHz 30-10000 MHz 30-10000 MHz Physical phenomena Free space loss Corrected standard loss Free space loss + corrections Free space loss Diffraction loss Free space loss Diffraction loss Diffraction calculation method - - Deygout (3 obstacles) Deygout corrected (3 obstacles) Deygout (3 obstacles) Profile based on - - DTM DTM Clutter Profile extraction mode - - Radial Radial Cell size Macro cell Macro cell Macro cell - Receiver location Rooftop Rooftop Street Street Rooftop Receiver Fixed Mobile Fixed Fixed Use d > 10 km Low frequencies Broadcast 1 < d < 1000 km Land and maritime mobile. WiMAX Propagation model Standard Propagation Model Erceg-Greenstein (SUI) ITU 529-3 COST-Hata Okumura-Hata Frequency band 150-3500 MHz 1900-6000 MHz 300-1500 MHz 150-2000 MHz Physical phenomena L(d.3. Geo Tab of the Explorer Window > DTM > DTM 1 (25m) > DTM 2 (40m) > Clutter > Clutter (20m) 2. when DTM 1 is on the top of DTM 2.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 DTM maps with 40 m and 25 m resolutions and a clutter map with a 20 m resolution are available.Atoll 3. However. HTx. only the ground altitude. To get ground altitude every 25 m. f. HRx) (per environment) Diffraction loss L(d. d is the distance between the transmitter and the receiver (km). A3.33 log  f  – 40.82 -13.94 2 2 a(hRx) is a correction for a receiver antenna height different from 1.75h Rx   – 4. urban environments and 1.97 2 63 . LTE 2.90 A3 -13.5 metre mobile antenna height. For other environments and mobile antenna heights. • For rural/small cities: a  h Rx  =  1.55 49.4 28 • For quasi-open rural areas: L model1 = Lu – a  h Rx  – 4.5m. Path loss (Lu) is calculated (in dB) as follows: Lu = A 1 + A 2 log  f  + A 3 log  h Tx  +  B 1 + B 2 log  h Tx  + B 3 h Tx  log d f is the frequency (MHz). CDMA2000.78  log  f   + 18.78  log  f   + 18.55 -6.16 33.3 Okumura-Hata and Cost-Hata Propagation Models 2. This formula is valid for flat. • For urban areas: L model1 = Lu – a  h Rx  • f 2 For suburban areas: L model1 = Lu – a  h Rx  – 2  log  ------  – 5.90 44. The parameters A1. and B3 can be user-defined. Default values are proposed in the table below: Parameters Okumura-Hata f 1500 MHz Cost-Hata f > 1500 MHz A1 69.33 log  f  – 35. B2. corrective formulas must be applied. hTx is the transmitter antenna height above ground (m) (Hb notation is also used in Atoll).56 log  f  – 0. CDMA2000.7 h Rx –  1.90 B2 -6. receiver-transmitter distance and antenna heights for an urban environment. the Hata formula is valid for urban environment and a receiver antenna height of 1. LTE Urban and suburban areas 100 m < d < 8 km Fixed WiMAX 1 < d < 100 km GSM.3.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Propagation model Standard Propagation Model Erceg-Greenstein (SUI) ITU 529-3 COST-Hata Okumura-Hata Profile based on DTM Clutter DTM DTM DTM Profile extraction mode Radial Systematic Radial Radial Radial Cell size Macro cell Mini cell Macro cell Mini cell Macro cell Mini cell Macro cell Mini cell Receiver location Street Rooftop Street Street Street Receiver Mobile and Fixed Fixed Mobile Mobile Use 1 < d < 20 km GSM. B1.3. UMTS.5m.55 B3 0 0 2.1 log  f  – 0.8  • For large cities: a  h Rx  = 3.2  log  11.3. WiMAX.2 Corrections to the Hata Path Loss Formula As described above.Atoll 3.94 • For open rural areas: L model1 = Lu – a  h Rx  – 4. LTE GSM.82 B1 44.1 Hata Path Loss Formula Hata formula empirically describes the path loss as a function of frequency.30 A2 26. A2. UMTS. CDMA2000. Atoll stops calculations.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 When receiver antenna height equals 1. L model = L model1 + L model2 2.Atoll 3. the Hata formula is valid for urban environment.82 log h Tx –  44. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation mode.9 – 6.55 + 26.4. 2nd step: This step depends on whether the ‘Add diffraction loss’ option is checked.37 + 20 log f – E which gives the following path loss formula for the ITU 529-3 model: Lu = 69. a conversion has to be made. is given by: E = 69.3. 1st step: For each calculation bin.16 log f + 13. Atoll proceeds as follows: a. • If the ‘Add diffraction loss’ option is unchecked. 64 .82 – 6. This formula is valid for flat.4 ITU 529-3 Propagation Model 2. h Tx is the transmitter antenna height above ground (m) (Hb notation is also used in Atoll) h Rx is the receiver antenna height above ground (m) d is the distance between the transmitter and the receiver (km) b is the distance correction The domain of validity of such is formula is: • • • • Frequency range: 300-1500 MHz Base Station height: 30-200 m Mobile height: 1-10 m Distance range: 1-100 km Since Atoll needs the path loss (Lu) formula.1 ITU 529-3 Path Loss Formula The ITU 529. receiver-transmitter distance and antenna heights for a urban environment.55 log h Tx   log d  b where: E is the field strength for 1 kW ERP f is the frequency (MHz). urban environments and 1. its formula empirically describes the path loss as a function of frequency. It determines the largest obstacle along the profile in accordance with the Deygout method and evaluates losses due to diffraction L model2 . This clutter bin corresponds to a clutter class. One can find the following conversion formula: Lu = 139.3 Calculations in Atoll Hata models take into account topo map (DTM) between transmitter and receiver and morpho map (clutter) at the receiver. for a receiver located on a urban environment. corrective formulas must be applied.9 – 6. The standard ITU 529-3 formula. L model = L model1 • If the ‘Add diffraction loss’ option is selected.3.2 Corrections to the ITU 529-3 Path Loss Formula Environment Correction As described above.4.5m. Then. it uses the Hata formula assigned to this clutter class to evaluate L model1 .5 metre mobile antenna height.82 log h Tx +  44. a(hRx) is close to 0 dB regardless of frequency. For other environments and mobile antenna heights. 2. For this reason. Atoll determines the clutter bin on which the receiver is located. b.3 model is a Hata-based model.55 log hTx   log d  b 2.16 log f – 13. HTxeff: effective height of the transmitter antenna (m). b.Atoll 3. This clutter bin corresponds to a clutter class. Atoll stops calculations. 2nd step: This step depends on whether the ‘Add diffraction loss’ option is checked.8 d>20 km: b = 1 +  0.78  log f  + 18. • For rural/small cities: a  h Rx  =  1.8  • For large cities: a  h Rx  = 3.4.5. Then. Atoll proceeds as follows: a. Atoll determines the clutter bin on which the receiver is located.14 + 1.7 h Rx –  1. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation mode. • d<20 km: b = 1 • h Tx –4 –3 d 0. L model = L model1 • If the ‘Add diffraction loss’ option is selected. 65 .87  10 f + 1. It determines the largest obstacle along the profile in accordance with the Deygout method and evaluates losses due to diffraction  L model2  . it uses the ITU 529-3 formula assigned to this clutter class to evaluate L model1 . K1: constant offset (dB). d: distance between the receiver and the transmitter (m). 1st step: For each calculation bin.75h Rx   – 4.5 Standard Propagation Model (SPM) 2. K2: multiplying factor for log(d). a  h Rx  is the environment correction and is defined according to the area size. K4 has to be a positive number. L model = L model1 + L model2 2.1 SPM Path Loss Formula SPM is based on the following formula: L model = K 1 + K 2 log  d  + K 3 log  H Txeff  + K 4  DiffractionLoss + K 5 log  d   log  H Txeff  + K 6  H Rxeff  + K 7 log  H Rxeff  + K clutter f  clutter  with.07  10 h' Tx    log ------ with h' Tx = ---------------------------------------- 20 –6 2 1 + 7  10 h Tx 2.1 log  f  – 0. K3: multiplying factor for log(HTxeff).56 log  f  – 0.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 L model1 = Lu – a  h Rx  for large city and urban environments f 2 L model1 = Lu – a  h Rx  – 2  log  ------  – 5.97 2 Distance Correction The distance correction refers to the term b above. K4: multiplying factor for diffraction calculation.33 log f – 40.4 for suburban area   28  2 L model1 = Lu – a  h Rx  – 4.2  log  11. • If the ‘Add diffraction loss’ option is unchecked.3 Calculations in Atoll Hata-based models take into account topo map (DTM) between transmitter and receiver and morpho map (clutter) at the receiver.3.94 for rural area Area Size Correction In the formulas above. 2.K2)NLOS. H 0 is the average ground height above sea level along the profile (m). If the receiver is in the transmitter line of sight.1 Visibility and Distance Between Transmitter and Receiver For each calculation bin. H Rxeff : effective mobile antenna height (m). f(clutter): average of weighted losses due to clutter. Height Above Ground The transmitter antenna height is above the ground (HTx in m). HTxeff equals HTx only.K2)LOS. • Whether the receiver is in the transmitter line of sight or not. H Txeff =  H Tx + H 0Tx  – H 0Rx + K  d where. Slope at Receiver Between 0 and Minimum Distance The transmitter antenna height is calculated using the ground slope at receiver. If the receiver is not in the transmitter line of sight. 66 .3. Distance min and Distance max are minimum and maximum distances from the transmitter respectively.2 Effective Transmitter Antenna Height Effective transmitter antenna height (HTxeff) may be calculated with six different methods. H 0Tx is the ground height (ground elevation) above sea level at transmitter (m). Atoll will use the set of values marked “Near transmitter”. Atoll determines: • The distance between the transmitter and the receiver.2 Calculations in Atoll 2. receiver is considered far from transmitter. K5: multiplying factor for log  d   log  H Txeff  K6: multiplying factor for H Rxeff K7: multiplying factor for log  H Rxeff  . the receiver is considered to be near the transmitter. Atoll will use the set of values (K1. If the profile is not located between the transmitter and the receiver.5. 2. Atoll will use the set of values “Far from transmitter”. Kclutter: multiplying factor for f(clutter). If the distance Tx-Rx is greater than the maximum distance.5. H Txeff = H Tx +  H 0Tx – H 0  where.5. The profile length depends on distance min and distance max values and is limited by the transmitter and receiver locations. The LOS is defined by no obstruction along the direct ray between the transmitter and the receiver. HTxeff = HTx Height Above Average Profile The transmitter antenna height is determined relative to an average ground height calculated along the profile between a transmitter and a receiver. If the distance Tx-Rx is less than the maximum user-defined distance (break distance).Atoll 3.2.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Diffraction loss: loss due to diffraction over an obstructed path (dB). 2. Atoll will take into account the set of values (K1. The LOS line equation is:   H 0Tx + H Tx  –  H 0Rx + H Rx   . • If H Txeff  20m then. Figure 2.Res  i  Los  i  =  H 0Tx + H Tx  – ----------------------------------------------------------------------d where. i is the point index. 67 .12: Enhanced Slope at Receiver Let x-axis and y-axis respectively represent positions and heights.3. Res is the profile resolution (distance between two points).0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 H 0Rx is the ground height (ground elevation) above sea level at receiver (m). Enhanced Slope at Receiver Atoll offers a new method called “Enhanced slope at receiver” to evaluate the effective transmitter antenna height. Atoll extracts the transmitter-receiver terrain profile. In this case. H Rx is the receiver antenna height above the ground (m). H Txeff = H Tx Absolute Spot Ht H Txeff = H Tx + H 0Tx – H 0Rx Distance min and distance max are set to 3000 and 15000 m according to ITU recommendations (low frequency broadcast f < 500 Mhz) and to 0 and 15000 m according Okumura recommendations (high frequency mobile telephony). Distance min is a distance from receiver. Atoll takes 200m. This calculation is achieved in several steps: 1. H Txeff = H Tx +  H 0Tx – H 0Rx  If H 0Tx  H 0Rx then. Spot Ht If H 0Tx  H 0Rx then. K is the ground slope calculated over a user-defined distance (Distance min).Atoll 3. Atoll uses 20m in calculations. These values are only used in the two last methods and have different meanings according to the method. 2. We assume that x-axis is oriented from the transmitter (origin) towards the receiver. • If H Txeff  200m then. Atoll determines line of sight between transmitter and receiver. 05. 0. H filt – Rx  i  = H filt – Rx  i + 1  +  H orig  i  – H orig  i + 1   H orig  i  – H orig  i + 1  ii. Original terrain height is determined from extracted ground profile. H filt – Tx  i  = H orig  i  • Filter starting from receiver Let us assume that H filt  Rx  = H orig  Rx  For each point. H filt – Rx  i  = H filt – Rx  i + 1  If H filt  i   H orig  i  additionally Then. It corresponds to the distance from receiver at which the original terrain profile plus 30 metres intersects the LOS line for the first time (when beginning from transmitter). H filt – Tx  i  = H filt – Tx  i – 1  iii. 0. H filt  i  = max  H filt – Tx  i  H filt – Rx  i   4. H orig is the original height. Profile points are evenly spaced on the basis of profile resolution. H orig  i  – H orig  i – 1  .05 ii. If H orig  i   H orig  i + 1  and ---------------------------------------------------. 0. we have three different cases: i. 3rd case: If H orig  i   H orig  i + 1  Then. for every point of profile. H orig  i  – H orig  i + 1  .0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 3. Atoll compares the two filtered heights and chooses the higher one. in order for them not to unfavourably influence the regression line calculation. Hills and mountains are already taken into account in diffraction calculations. one established from the transmitter and another from the receiver. Therefore. R. To determine filtered terrain height at a point. H filt is the filtered height. we have three different cases: i. • Filter starting from transmitter Let us assume that H filt – Tx  Tx  = H orig  Tx  For each point.05 . Atoll evaluates ground slope between two points and compares it with a threshold set to 0. H filt – Rx  i  = H filt – Rx  i + 1  iii. The influence area must satisfy additional conditions: • 68 R  3000m . Some notations defined hereafter are used in next part. H filt – Rx  i  = H orig  i  Then. H filt – Tx  i  = H filt – Tx  i – 1  +  H orig  i  – H orig  i – 1   H orig  i  – H orig  i – 1  . If H orig  i   H orig  i – 1  and --------------------------------------------------Res Then.05 . Atoll calculates two filtered terrain profiles. Atoll filters the terrain profile. If H orig  i   H orig  i – 1  and --------------------------------------------------Res Then. Atoll determines the influence area. If H orig  i   H orig  i – 1  Then. where three cases are possible. It determines filtered height of every profile point. 0.Atoll 3. H filt – Tx  i  = H filt – Tx  i – 1  If H filt  i   H orig  i  additionally Then. If H orig  i   H orig  i + 1  and --------------------------------------------------Res Then.3.05 Res Then. 63 + ----------10  1000  1000 2. dm = d – R --2 d(i) is the distance between i and the transmitter (m). 1000m will be used in calculations.3 Effective Receiver Antenna Height H Rxeff =  H Rx + H 0Rx  – H 0Tx where. Only points within R are taken into account. Therefore.3   H where. Atoll calculates effective transmitter antenna height. 69 . H Rx is the receiver antenna height above the ground (m).2. Atoll performs a linear regression on the filtered profile within R in order to determine a regression line. R = d. H 0Tx + H Tx – b H Txeff = --------------------------------2 1+a If HTxeff is less than 20m. Then. 5. Atoll extends the regression line to the transmitter location. If d < 3000m. The regression line equation is: y = ax + b   d  i  – dm   Hfilt  i  – Hm  i and b = H m – ad m a = --------------------------------------------------------------------2  d  i  – dm   i where. • • When several influence areas are possible. Therefore.3. Atoll evaluates path loss using H Txeff = 20m and applies a correction factor. Then. 7. If H Txeff is still less than 20m (even negative). K lowant = ------Txeff – 20   – -------------------------------------------------------------------------5 d -   6.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 • • R  0. an additional correction is taken into account (7th step).5. Atoll chooses the highest one. • If H Txeff is still less than 20m. its equation is: regr  i  = a   i  Res  + b 6.93 + ----------d -  9.–  0. which begins at transmitter.Atoll 3.01  d R must contain at least three bins. 1 H m = --n  Hfilt  i  i i is the point index. if H Txeff  20m . • In case H Txeff  1000m . L model = L model   H Txeff = 20m  d f  + K lowant 20   1 –  H Txeff – 20   d . Atoll recalculates it with a new influence area. H Txeff (m). H 0Rx is the ground height (ground elevation) above sea level at the receiver (m). Therefore. the path loss formula is: L model = K 1 NLOS + K 2 NLOS log  d  + K 3 log  H Txeff  + K 4  Diffraction + K 5 log  H Txeff  log  d  + K 6  H Rx + K clutter f  clutter  K hill LOS is determined in three steps.2.atl document.Atoll 3. w: weight determined through the weighting function.3.75 log  h  – 11. K h = 0 2 Else K h = 7. it considers average clutter height specified for each clutter class in the clutter classes file description. we have: L model = K 1 LOS + K 2 LOS log  d  + K 3 log  H Txeff  + K 5 log  H Txeff  log  d  + K 6  H Rx + K clutter f  clutter  + K hill LOS When the transmitter and the receiver are not in line of sight.4 Correction for Hilly Regions in Case of LOS An optional corrective term enables Atoll to correct path loss for hilly regions when the transmitter and the receiver are in Line-of-sight.2.1924   H 0Rx + H Rx – regr  i Rx   H 0Rx + H Rx – regr  i Rx  2 Else K hf = – 2   – 1. • Or only ground altitude. one which is exceeded by 10% of the profile points and the other one by 90%. it sorts points according to the deviation and draws two lines (parallel to the regression line). Along the transmitter-receiver profile. 2nd step: Atoll evaluates the terrain roughness. and regression line are supposed available.5 Diffraction Four methods are available to calculate diffraction loss over the transmitter-receiver profile.73  log  h   – 15. 2. K hf = – 2  0. Four weighting functions are available: 70 . Otherwise. 3rd step: Atoll calculates K hill LOS . We have K hill LOS = K h + K hf If 0  h  20m .2.29 log  h  + 6. 2. Atoll uses clutter height information from clutter heights file if available in the .6 Losses due to Clutter n Atoll calculates f(clutter) over a maximum distance from receiver: f  clutter  =  Li wi i=1 where. R. Then.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 H 0Tx is the ground height (ground elevation) above sea level at the transmitter (m). In this case.21   ----------------------------------------------------h iRx is the point index at receiver. it is the distance between the two lines. The calculation of effective antenna heights ( H Rxeff and H Txeff ) is based on extracted DTM profiles.746 If 0  h  10m . n: number of points taken into account over the profile.5. h. if the receiver is in the transmitter line of sight and the Hilly terrain correction option is active. Influence area. Points are evenly spaced depending on the profile resolution. 2. 1st step: For every profile point within influence area.616  log  h   + 14. you may consider: • Either ground altitude and clutter height (Consider heights in diffraction option). Atoll calculates height deviation between the original terrain profile and regression line. They are not properly performed if you have not imported heights (DTM file) beforehand.5. L: loss due to clutter defined in the Clutter tab by the user (in dB).5. specify receiver clearance (m) per clutter class. where d’i is the distance between the receiver and the ith point and D is the maximum distance defined. f(clutter)=0.even better . This approach is recommended if the clutter height information is either semi-deterministic (clutter roughly defined. 71 . Both ground altitude and clutter height are considered along the whole transmitter-receiver profile except over a specific distance around the receiver (clearance).0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 • Uniform weighting function: w i = --1n • di Triangular weighting function: w i = -----------n  dj j=1 • d i = D – d' i .via a clutter height file). To avoid this. The clearance information is used to model streets. In case of semi-deterministic clutter information.5. we advise: 1. This approach is recommended if the clutter height information is statistical (clutter roughly defined. altitude defined with an average height per clutter class or .3. • d log  ----i + 1 D  Logarithmic weighting function: w i = -----------------------------------n d log  ----j + 1 D  j=1 di ---D • e –1 Exponential weighting function: w i = -----------------------n e dj ---D –1 j=1 The chart below shows the weight variation with the distance for each weighting function.7 Recommendations Beware that the clutter influence may be taken into account in two terms. Losses due to clutter are only taken into account in the computed Diffraction loss term. Not to consider clutter heights to evaluate diffraction loss over the transmitter-receiver profile if you specify losses per clutter class. no altitude). In this case.13: Losses due to Clutter 2. Diffraction loss and f(clutter) at the same time. altitude defined with an average height per clutter class) or deterministic (clutter sharply defined.Atoll 3.2. Not to define any loss per clutter class if you take clutter heights into account in the diffraction loss. Figure 2. where Atoll proceeds as if there was only the DTM map. Or 2. 3. Even with no clearance. Here. m is the number of measurement points. Atoll calculates the path loss by considering potentially some diffraction loss at reception. Atoll does not consider any difraction (and clearance) from the building but takes into account the indoor loss as an additional penetration loss. A calibrated model must restore the behaviour of CW measurements depending on their configuration on a large scale.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Figure 2. the ground altitude and clutter height (in this case. given a real m x n matrix A. without having been calibrated on these new CW measurements. Calibration is based on imported CW measurement data.5. 2. Clearance definition is not necessary in case of deterministic clutter height information. K3. For a complete description of the calibration procedure (including the very important prerequisite filtering work on the CW measurement points).1 General Algorithm Propagation model calibration is a special case of the more general Least-Square problems. find a real n-vector x0 that minimises the Euclidean length of Ax . the clutter height (extracted either from clutter class or clutter height folders) is never considered at the last profile point. and a real m-vector b. please refer to the User Manual and the SPM Calibration Guide. 72 . The following SPM formula parameters can be estimated: • • • • K1.3 Automatic Calibration The goal of this tool is to calibrate parameters and methods of the SPM formula in a simple and reproducible way. K2. average height specified for each clutter class in the clutter classes map description) are taken into account along the profile.14: Tx-Rx profile In the above figure. K5. K4. 2. You must keep in mind that the model calibration and its result (standard deviation and root mean square) strongly depend on the CW measurement samples you use. i. K6 and K7 Losses per clutter class (Kclutter must be user-defined) Effective antenna height method Diffraction method Automatic model calibration provides a mathematical solution. The relevance of this mathematical solution with a physical and realistic solution must be determined before committing these results. The calibrated model has to give correct results for every new CW measurement point in the same geographical zone.5. It is the process of limiting the difference between predicted and measured values.e. Two cases can be considered: 1. • • To consider indoor losses in building only when using a deterministic clutter map (clutter height map). If the receiver is supposed to be inside a building (clutter height higher than receiver height). and not just totally coincide with a few number of CW measurements.b. the 'Indoor Coverage' box must not be checked in predictions unless this loss will be counted twice inside buildings (on the entire reception clutter class and not only inside the building). 2. Clutter height information is accurate enough to be used directly without additional information such as clearance.3. If the receiver is in the street (clutter height lower than receiver height). n is the number of parameters to calibrate. Lawson & Hanson [2] proposed a theoretical solution of the least-square problem with general linear inequality constraints on the vector x0. SIAM.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 A is the values of parameter associated variables (log(d). Here are some sample values: Project type Frequency (MHz) K1 GSM 900 935 12. The theoretical mathematical solution of this problem was found by Gauss (around 1830). typical losses (in dB) per clutter class are: Dense urban From 4 to 5 Woodland From 2 to 3 Urban 0 73 . Further enhancements to the original method were proposed in the 60's in order to solve the numerical instability problem. log(heff).3.5 0. Since Kclutter is a constant.6 2500 26. “Numerical Methods for Least Square Problems”.8 K5 -10 -6. The vector x0 is the set of parameters found at the end of the calibration.8 2700 27. which is explained in detail in [1]. 1974. It is the average of the downlink and uplink centre frequencies of the band. its value is strongly dependant on the values given to the losses per clutter classes. From experience. K6 is a multiplicative coefficient to a value in dB.3. 2045 MHz = (2140 + 1950)/2. “Solving Least Squares Problems”. Atoll implementation is based on this method.55 0 K6 -1 0 0 K7 -10 0 0 It is recommended to set K6 to 0.9 70 K3 -20 5. In 1974.L. It is highly recommended to calibrate the SPM using measurement data collected on the field for WiMAX networks before using the SPM for predictions. [2] Lawson C..) at each measurement point. The above K1 values for WiMAX are extrapolated estimates for different frequency ranges. K1 depends on the frequency and the technology.5 GSM 1800 1805 22 GSM 1900 1930 23 UMTS 2045a 23.2 Sample Values for SPM Path Loss Formula Parameters The following tables list some sample orders of magnitudes for the different parameters composing the Standard Propagation Model formula. All K paramaters can be defined by the automatic calibration wizard.Atoll 3.7 WiMAX a. etc. 2. 1996. Hanson R. Minimum Typical Maximum K1 Variable Variable Variable K2 20 44.J.5.8 1xRTT 1900 23 2300 25.83 20 K4 0 0. which means that slight variations in K6 have considerable impact on the path loss. and use K7 instead of K6.9 3500 31. References: [1] Björck A.9 3300 30. SIAM. and b is the vector of measurement values. So. and change its signature and type as shown above. you will be able to keep different versions of the SPM.1 However. The above values are normalized for urban clutter types (0 dB for urban clutter class).. Positive values correspond to more dense clutter classes and negative values to less dense clutter classes. backup it’s signature. 5.e. 4. 2.3.StdPropagModelUnmasked. To make the SPM calculate path losses excluding the antenna pattern attenuation. it is possible to create different instances of the SPM. has the following characteristics: • • Signature: Type: {D5701837-B081-11D4-931D-00C04FA05664} Atoll. Such path losses are useful when using path loss matrices calculated by Atoll with automatic optimisation tools. valid for urban environments. i. 6. Atoll automatically creates different signatures for different instances of the same propagation model.Atoll 3.4 Unmasked Path Loss Calculation You can use the SPM to calculate unmasked path losses. Reset the path loss storage directory to the original one. 3. you have to change the signature of the model as well. Backup the storage directory of path loss matrices. Select the SPM used. and to invalidate the path loss matrices calculated with this model. Therefore. with different parameter settings. Perform optimisation using the path loss matrices calculated by the unmasked version of the SPM. Set a different storage directory for calculating and storing unmasked path loss matrices.StdPropagModel. those that calculate path losses with antenna pattern attenuation. and create unmasked versions of these instances. The Standard Propagation Model is derived from the Hata formulae. because the signature of the propagation model is still the same.1 You can access these parameters in the Propagation Models table by double-clicking the Propagation Models folder in the Modules tab. you have to change the type of the SPM to: • Type: Atoll. changing the type only does not invalidate the already calculated path loss matrices. 74 . If you want Atoll to recognize that the SPM has changed. Unmasked path losses are calculated by not taking into account the transmitter antenna patterns. The usual process flow of an ACP working on an Atoll document through the API would be to: 1. the attenuation due to the transmitter antenna pattern is not included. The instance of the SPM available by default. 2. You can change the signature and type of the original instance of the SPM. but it is recommended to make a copy of the SPM in order not to lose the original SPM parameters. and others that calculate path losses without it. under the Propagation Models folder in the Modules tab.5.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Suburban From -5 to -3 Industrial From -5 to -3 Open in urban From -6 to -4 Open From -12 to -10 Water From -14 to -12 These values have to be entered only when considering statistical clutter class maps only. Restore the type and the signature of the SPM. The default signature for the SPM that calculates unmasked path loss matrices is: • Signature: {EEE060E5-255C-4C1F-B36C-A80D3D972583} The above signature is a default signature. 2.ini options) or the other one (new identifier & signature) but not to combine both. • Receiver Clearance Define receiver clearance (m) per clutter class when clutter height information is either statistical or semideterministic. do not define any receiver clearance (m) per clutter class. Otherwise. 75 . 2. If the clutter is deterministic. When selecting the option ‘Line of sight only’. see "Free Space Loss" on page 86. Using the SPM. Atoll takes clutter height information from the clutter heights file if available in the . Otherwise. However. the "unmasked" version does not take Atoll. where Atoll proceeds as if there was only the DTM map (see SPM part). Atoll considers that Lmodel tends to infinity. It still takes into account Atoll. it considers average clutter height specified for each clutter class in the clutter classes file description. Atoll checks for each calculation bin if the Diffraction loss (as defined in the Diffraction loss: Deygout part) calculated along profile equals 0. Both ground altitude and clutter height are considered along the whole profile except over a specific distance around the receiver (clearance).2 Calculations in Atoll Free Space Loss For free space loss calculation. This method is described under "Diffraction" on page 86. and F Diff is the diffraction multiplying factor defined in the model properties. Atoll uses the clearance information to model streets.6. losses due to clutter are taken into account in diffraction losses.ini options. In this case. The final diffraction losses are determined by multiplying the diffraction losses calculated using the Deygout method by the Diffraction multiplying factor defined in the model properties.ini options into account.1 Signature: {659F0B9E-2810-4e59-9F0D-DA9E78E1E64B} The "masked" version of the algorithm has not been changed.atl document. receiver is considered in ‘line of sight’ and Atoll computes Lmodel on each calculation bin using the formula defined above. • Receiver Height Entering receiver height per clutter class enables Atoll to consider the fact that receivers are fixed and located on the roofs. • Visibility If the option ‘Line of sight only’ is not selected.StdPropagModelIncidence.1 WLL Path Loss Formula L model = L FS + F Diff  L Diff Where L FS is the free space loss calculated using the formula entered in the model properties.6 WLL Propagation Model 2.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 • • • • It is not possible to calibrate the unmasked version of the SPM using measurement data. Therefore.6. L Diff is the diffraction loss calculated using the 3-obstacle Deygout method. The Deygout construction (considering 3 obstacles) is used. clutter height information is accurate enough to be used directly without additional information such as clearance (Atoll can locate streets).3. Diffraction Atoll calculates diffraction loss along the transmitter-receiver profile built from DTM and clutter maps. Atoll computes Lmodel on each calculation bin using the formula defined above. • • In this case. you can also calculate the angles of incidence by creating a new instance of the SPM with the following characteristics: Type: Atoll.Atoll 3. It’s highly recommended to use one method (Atoll. 2. Cn is the field strength received in dBV/m. Diffraction Atoll calculates diffraction loss along the transmitter-receiver profile is built from the DTM map.8 ITU-R P.HTxeff) is the field received in dBV/m) is read from charts for a distance.7 ITU-R P. see "Free Space Loss" on page 86. Atoll evaluates the effective transmitter antenna height.7. HTxeff (in m).8.2 Calculations in Atoll Free Space Loss For free space loss calculation. let us assume that Cn=En(d. see "Free Space Loss" on page 86. L model = 1000 If 1<d<1000 km.HTxeff) (where En(d. • Cn Calculation The Cn value is determined from charts Cn=f(d. CorrectedS tan dardLoss = 60 – C n – A H Rxeff – A cl – 108. In the following part. f is the frequency in MHz. HTxeff).0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 2. These methods are described under "Diffraction" on page 86. and an effective transmitter antenna height. H Txeff = H 0Tx + H Tx – H 0Rx 76 .2 Calculations in Atoll Free Space Loss For free space loss calculation.3. d (in km).54 – 20 log f where.Atoll 3. First of all.75 + 31. The Deygout construction (considering 3 obstacles).370-7 Propagation Model 2. L model = max  L FS CorrectedS tan dardLoss  d is the distance between the transmitter and the receiver (km). AH Rxeff is a correction factor for effective receiver antenna height (dB). Acl is the correction for terrain clearance angle (dB). Corrected Standard Loss This formula is given for a 60 dBm (1kW) transmitter power. as follows: If 0  d  3km . is used.1 ITU 370-7 Path Loss Formula If d<1 km.526-5 Propagation Model 2.7. H Txeff . with or without correction.1 ITU 526-5 Path Loss Formula L model = L FS + L Diff Where L FS is the free space loss calculated using the formula entered in the model properties and L Diff is the diffraction loss calculated using the 3-obstacle Deygout method. 2. L model = L FS If d>1000 km.8. 2. 1200) • AHRxeff Calculation AH Rxeff H Rx c = --.  is the clearance angle (in radians) determined according to the recommendation 370-7 (figure 19).Atoll 3. f is the frequency stated in MHz.5) + En(25.1  + 1  +   – 0. we have C n = E n  d + 142 – d horizon 1200  Else Cn=En(d. 20  log  ------10  6 where.15  is the average ground height (m) above sea level for the profile between a point 3 km and another 15 km from transmitter. H 0  3 . A cl = 8.9 + 20 log     – 0. Then.9 + 20 log     – 0.5) – En(dhorizon.5  Else Cn=En(d.1    2 Otherwise. If HTxeff < 37.5 HTxeff 1200.1 –  6. c is the height gain factor. 37.1  + 1  +   – 0.d  is the average ground height (m) above sea level for the profile between a point 3 km from transmitter and the receiver (located at d km from transmitter).1  H Txeff (d is stated in km) Therefore. 37.15  where. c values are provided in the recommendation 370-7. Cn= En(d. 77 .HTxeff) Otherwise. Atoll considers d horizon = 4.5 If d  d horizon . If 37. 1200) – En(dhorizon. we have C n = E n  d + 25 – d horizon 37.1    f With  = –   4000  --------300 where.d  If 15  d . H Txeff = H 0Tx + H Tx – H 0  3 . HRx is the user-defined receiver height.5) If HTxeff > 1200 If d  d horizon .9 –  6. for example.3. • Acl Calculation If f  2 300 MHz. 37. H 0  3 . H 0Tx is the ground height (ground elevation) above sea level at the transmitter (m). A cl = 14. c=4 in a rural case. H Tx is the transmitter antenna height above the ground (m). Atoll determines Cn using bilinear interpolation as follows. H Txeff = H 0Tx + H Tx – H 0  3 .0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 If 3  d  15km . depending on d and HTxeff. 1200) + En(142. . where d depends  2 on the terrain type. al.S. 17.0075 0. The basic path loss equation of the Erceg-Greenstein propagation model is: d PL = A + 10  a  H BS   Log 10  -----  d 0 4d 0 Where A = 20  Log 10  ------------ . d is the distance  between the base station antenna and the receiver terminal and d0 is a fixed reference distance (100 m).G.77 Vol.3. and c are correction coefficients which depend on the SUI terrain type.Atoll 3. 2. References: • • [1] V. namely A. pp. Select Areas Commun. 2005. The basic path loss equation with correction factors is presented below: d PL = A + 10  a  H BS   Log 10  ----- + a  f  – a  H R  d0 f Where a(f) is the correction factor for the operating frequency. The Erceg-Greenstein propagation model is further developed through the correction factors introduced by the Stanford University Interim model. Erceg et. V.9 Erceg-Greenstein (SUI) Propagation Model Erceg-Greenstein propagation model is a statistical path loss model derived from experimental data collected at 1. "Comparison of empirical propagation path loss models for fixed wireless access systems. The constants used for a. 1205-1211. Type C is associated with minimum path loss and applies to flat terrain with light tree densities. B and C.. a(HBS) is the correction factor for base station antenna heights. I. • • • Type A is associated with maximum path loss and is appropriate for hilly terrain with moderate to heavy tree densities. D..0065 0. The path loss model applies to base antenna heights from 10 to 80 m. This propagation model is well suited for distances and base station antenna heights that are not well-covered by other models. a(HR) is the correction factor for the receiver antenna height. M. July 1999.005 The word ‘terrain’ is used in the original definition of the model rather than ‘environment’. 1 2. with f being the operating 2000 HR frequency in MHz. a  H R  = X  Log 10  ------ . M. b. • • a(HR) = 0 for HR = 2 m. 78 Model Parameter Terrain A Terrain B Terrain C a 4. vol. no. 30 May-1 June 2005 Page(s):73 .16 include channel models developed by Stanford University. This is a fixed quantity which depends upon the frequency of operation.6 b (m-1) 0. Wassell." Vehicular Technology Conference.9. [2] Abhayawardhana.1 to 8 km.P. “An empirically based path loss model for wireless channels in suburban environments. HBS: ca  H BS  = a – b  H BS + ------H BS Where 10 m  HBS  80 m . b. Sellars. Type B is characterised with either mostly flat terrains with moderate to heavy tree densities or hilly terrains with light tree densities. base-to-terminal distances from 0. IEEE 61st Volume 1.1 SUI Terrain Types The SUI models are divided into three types of terrains2. and c are given in the table below.. and it distinguishes between different terrain categories called the Stanford University Interim Terrain Models.6 4. The model is for suburban areas. 7.0 3.. a  f  = 6  Log 10  ------------ .. The standards proposed by the IEEE working group 802. Crosby.9 GHz in 95 macrocells.” IEEE J.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 2. . and three distinct terrain categories. and a. Hence it is used interchangeably with ‘environment’ in this description.J. Brown. 366 + 26  Log 10  f  + 10  a  H BS    1 + Log 10  d   – a  H R  (2) Where. the path loss has been restricted to the free space path loss with correction factors for operating frequency and receiver height: 4d 4d PL = 20  Log 10  ------------------ + a  f  – a  H R  instead of PL = 20  Log 10  ------------------   Where a(f) and a(Hr) have the same definition as given above. You can get the same resulting equation by setting a(hBS) = 2. 79 . Atoll determines the clutter bin on which the receiver is located. The final path loss is the sum of the path loss determined in 1st step and L Diffraction . If the ‘Add diffraction loss’ option is selected. we get.3.366 + 26  Log 10  f  + 10  a  H BS    1 + Log 10  d   The above path loss formulas are valid for d > d0. i.e.634 + 26  Log 10  f  + 20  Log 10  d  The above equation is not user-modifiable in Atoll except for the coefficient of Log 10  f  . see "3 Knife-edge Deygout Method" on page 87. • • If the ‘Add diffraction loss’ option is not selected. For more information on the Deygout method. It determines the largest obstacle along the profile in accordance with the Deygout method and evaluates losses due to diffraction L Diffraction .6 17. PL = 12.e. Simplifying the above equation.9. • • • • f is the operating frequency in MHz d is the distance from the transmitter to the received in m in equation (1) and in km in equation (2) HBS is the transmitter height in m HR is the receiver height in m The above equation is divided into two parts in Atoll: PL = Lu – a  H R  Where.1 20 X 10. Atoll proceeds as follows: a.8 10. 1st step: For each pixel in the calculation radius. Atoll uses the Erceg-Greenstein (SUI) path loss formula assigned to this clutter class to evaluate path loss. b. i. and it can also take clutter into account at the receiver location. For d < 100 m.9.2 Erceg-Greenstein (SUI) Path Loss Formula The Erceg-Greenstein (SUI) propagation model formula can be simplified from the following equation: 4d d PL = 20  Log 10  -----------0- + 10  a  H BS   Log 10  ----- + a  f  – a  H R      d 0 (1) to the equation below: PL = – 7. Lu = – 7.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Model Parameter Terrain A Terrain B Terrain C c (m) 12.634 + 26  Log 10  f  + 20  Log 10  d  – a  H R  . 2.8 20 2. 26. Atoll uses the same coefficient as the one you enter for Log 10  f  in Atoll for the case d > d0.3 Calculations in Atoll The Erceg-Greenstein (SUI) propagation model takes DTM into account between the transmitter and the receiver. or Lu = 12. It extracts a geographic profile between the transmitter and the receiver using the radial calculation method.Atoll 3. 2nd step: This step depends on whether the ‘Add diffraction loss’ option is selected or not. 1st step gives the final path loss result. This clutter bin corresponds to a clutter class. d > 100 m. • Receiver antenna height. the graphs represent field strength values for a receiver antenna height of 10 m. Similarly.1546-2 Propagation Model This propagation model is based on the P. Once f n1 and f n1 are known. Moreover. 10 and 50 % of time. and the clutter at the receiver location. and 2000 MHz) between which the transmission frequency is located. For other values of receiver antenna height.. and the graphs for 1000 MHz are applicable to frequencies from 1000 to 3000 MHz. 80 . f is the transmission frequency.1. the distance between the transmitter and the receiver. and t is the percentage of time for which the path loss has to be calculated. f n : 100. land or sea). 20. The minimum value of the representative height of clutter is 10 m. the sets of graphs which will be used for the calculation are also known.1 Step 1: Determination of Graphs to be Used First of all. the type of environment (i. It is possible for the value of h 1 to be negative. These recommendations are not valid for transmitter-receiver distances less than 1 km or greater than 1000 km. a correction is applied according to the environment of the receiver. i. the upper and lower nominal frequencies are determined for any given transmission frequency. In the following calculations.3707 recommendations. given in db  V  m  .10. t : 1. those for 600 MHz are applicable to frequencies from 300 to 1000 MHz. For h 1 below 10 m. equal to the representative height of the clutter around the receiver. The upper and lower nominal frequencies are the nominal frequencies (100. Therefore in Atoll. The path loss is calculated by this propagation model with the help of graphs available in the recommendations. 2. and are suited for operating frequencies from 30 to 3000 MHz. these recommendations are not valid for transmitter antenna heights less than the average clutter height surrounding the transmitter. and 50 % The propagation curves represent the field strength values exceeded for 1. along with the information about the percentage of time t and the type of path (land or sea). § 4. f n1  f  f n2 . 75.. § 6). The mixture of land and sea paths is not supported by Atoll. the path loss between a transmitter and a receiver over more than 1000 km is the same as the path loss over 1000 km.e.10. the extrapolation to be applied is given in Annex 5. h 1 : 10. • Transmitter antenna heights. an interpolation or extrapolation from the appropriate two curves is used. The method for interpolation is described in the recommendations (Annex 5.1). The method for calculating this correction is given in Annex 5. § 4. The following calculations are performed in Atoll to calculate the path loss using this propagation model. • • The cold sea graphs are used for calculations over warm and cold sea both. • Time variability. § 4. h 2 : 10 m For land paths. The graphs provided in the recommendations represent field (or signal) strength. 600.5. transmitter and receiver heights.3. For sea paths. as a function of distance for: • Nominal frequencies.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Shadow fading is computed in Atoll independent of the propagation model.Atoll 3. § 9. These recommendations extend the P. 300. d is the transmitter-receiver distance. 600. 37.1 Calculations in Atoll The input to the propagation model are the transmission frequency. and 1000 MHz The graphs provided for 100 MHz are applicable to frequencies from 30 to 300 MHz. the path loss between a transmitter and a receiver over less than 1 km is the same as the path loss over 1 km. 150. 10.10 ITU-R P. see "Shadow Fading Model" on page 90. 2. as described in the recommendations (Annex 5. 600. For more information on the shadow fading calculation. the precentage of time the field strength values are exceeded.3. in which case the method is given in Annex 5.2. 2.1546-2 recommendations of the ITU-R. the graphs represent field strength values for a receiver antenna height above ground.e. and 1200 m For any values of h 1 from 10 to 3000 m. i.. the required field strength is obtained directly from the corresponding graph. h Up = 1200 m if h 1  1200 m . E .Atoll 3. 2.. the thrid and final interpolation/extrapolation is performed over the interpolated/extrapolated values of E d to determine E f . namely 10. E SE is an enhancement for sea graphs.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 2. E Low is the field strength value for h Low at the required distance.e. h a is the antenna height above ground and h eff is the effective height of the transmitter antenna. i. 81 . E Max . 37.1  h 1 .4 Step 4: Interpolation/Extrapolation of Field Strength The interpolations are performed in series in the same order as described below. • Land paths h 1 = h eff • Sea paths h 1 = Max  1 h a  Here.e.1  h 1 . 75.3 Step 3: Determination of Transmitter Antenna Height The transmitter antenna height to be used in the calculation depends on the type and length of the path. if d  4.94    Log  50  t  for sea paths. or E h1 = E 10  12.9 km  + E 10  d  – E 10  d H  h 1   because d H  10  = 12. otherwise h Up is the nearest nominal effective height above h 1 . and 1200 m.3.9 – 20  Log  d  for land paths. and h a ) are in expressed in m. E h1 = E 10  d H  10   + E 10  d  – E 10  d H  h 1   .9 – 20  Log  d  + 2. Where E FS is the free space field strength for 1 kW ERP.9 km • For land path if the transmitter-receiver distance is greater than or equal to the smooth-Earth horizon distance d H  h 1  = 4.9 km + d – d H  h 1   because d H  10  = 12.2  d and d km from the transmitter in the direction of the receiver.38  1 – exp  – d  8. and E Max = E FS + E SE = 106. all antenna heights (i. which is given by: E Max = E FS = 106. 150. • If 0 m  h 1  10 m • For land path if the transmitter-receiver distance is less than the smooth-Earth horizon distance d H  h 1  = 4.. which is its height over the average level of the ground between distances of 0. 2.5. 300. And. The second interpolation/extrapolation is performed over the interpolated/extrapolated values of E h1 to determine E d . Step 4. if d  4.1. h 1 .1  h 1 .1. and E Up is the field strength value for h Up at the required distance. 600. The first interpolation/extrapolation is performed over the field strength values.10. otherwise h Low is the nearest nominal effective height below h 1 .1  h 1 .1. E h1 = E 10  d H  10  + d – d H  h 1   . h eff .10. Otherwise: • If 10 m  h 1  3000 m The field strength is interpolated or extrapolated from field strengths obtained from two curves using the following equation: Log  h 1  h Low  E h1 = E Low +  E Up – E Low   ------------------------------------Log  h Up  h Low  Where h Low = 600 m if h 1  1200 m . 20.1: Interpolation/Extrapolation of Field Strength for Transmitter Antenna Height If the value of h 1 coincides with one of the eight heights for which the field strength graphs are provided. from the graphs for transmitter antenna height to determine E h1 . or E h1 = E 10  12.e.9 km Where E x  y  is the field strength value read for the transmitter-receiver distance of y from the graph available for the transmitter antenna height of x.2 Step 2: Calculation of Maximum Field Strength A field strength must not exceed a maximum value.10. 6 = Max  0. –h1  . h 1 should not be less than 1 m.. 300. This distance is given by: D h1 = D 0. 925. d H  10  + d – d H  h 1   1000 km even though d  1000 km . 19. 6.. 2. • For sea path. 160. 800.1  + 1 +  – 0.35 for 100 MHz. 3. 975. 14.6 Fresnel clearance distance for the sea path where the transmitter antenna height is 20 m is also calculated as: D 20 = D 0. 90. 450. 575. and Log  20  10  Log  20  10  E'' is the field strength calculated as described for land paths. 15. If d  D h1 the 0. 60.a = 6370 km . 400. 8. 180. 525. E D ED 20 h1 is E Max for d = D h1 . 775. and F S =  d – D 20   d . This correction is the maximum of the diffraction correction. then interpolation of field strength is not required and the field strength can be directly read from the graphs. This calculation requires the distance at which the path has 0.3. 1000. 5. 825. 600.6 of the first Fresnel zone just unobstructed by the sea surface.0000389  f  h 1  h 2 (frequency-dependent term). 650. 675. 11. 425. (radius of the Earth).with  e = -------------------C h1t = 30  Log  ---------------------. the field strength for the required distance is given by: E h1  E Max   Log  d  D h1  =  E D +  E D – E D   ---------------------------------h1 20 h1 Log  D 20  D h1     E'   1 – F S  + E''  F S for d  D h1 for D h1  d  D 20 for d  D 20 Where E Max is the maximum field strength at the required distance as calculated in "Step 2: Calculation of Maximum Field Strength" on page 81. E 10  y  and E 20  y  are field strengths interpolated for distance y and h 1 = 10 m and 20 m . 875. 82 . 750. and k = 4  3 is the   e +  eff2 ak effective Earth radius factor for mean refractivity conditions. 25. 30. 350. 130. 950. respectively. 4. 95. 725. 12.00 for 2000 MHz. E' = E 10  d  +  E 20  d  – E 10  d    ------------------------------. 625. 150. 375. the field strength is plotted against distance from 1 km to 1000 km. E h1 . calculated in the above description in order to take into account the diffraction and tropospheric scattering. 900. 200. Log  h1  10  Log  h1  10  = E 10  D 20  +  E 20  D 20  – E 10  D 20    ------------------------------. 75. 85. 225.6  f h 1  h 2 = 10 m   (km) Df  Dh Where D 0.1   h 1 + h 2  (asymptotic term defined by the horizon distance). 17.31 for 600 MHz. 850. 275. 18. If the transmitter-receiver distance is a value from this list. The distance values for which field strengths are tabulated are 1. 55.Atoll 3. 500. 80. 475. 9. and K  is 1.2: Interpolation/Extrapolation of Field Strength for Transmitter-Receiver Distance In the field strength graphs in the recommendations.9 + 20  Log    – 0. 3. 550. the field strength is determined from linear extrapolation for Log (distance) of the graph given by: Log  d  D Low  E h1 = E Low +  E Up – E Low   -------------------------------------Log  D Up  D Low  Where D Low is penultimate tabulation distance (km). 325. D Up is the final tabulation distance (km). 140.03 – J    with 2 J    =  6. E Low is the field strength value for D Low . and  D f + D h D h = 4. 40. 250. 13.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 If in the above equation.1   and  = K    eff2 . .001 ----------------- (km) with D f = 0. and tropospheric scattering correction. 20. 6... 35. 10. 50. 70. 16. • If h 1  0 m A correction is applied to the field strength. 190. 170. Step 4.6  f  h 1 = 20 m   h 2 = 10 m   (km) Once D h1 and D 20 are known. 100. 7. 120. C h1 = Max  C h1d C h1t  Where C h1d = 6. 65. and E Up is the field strength value for D Up . 45. 110.  eff2 = arc tan  ----------9000 e  180  d. 700. Atoll 3.2  Log  f    Log  -----2-  10 • For sea path d 10 and d h2 are determined as distances at which at which the path has 0. the field strength values are extrapolated from the two nearer nominal frequency values. 2000 MHz otherwise).2 + 6.6 of the first Fresnel zone just unobstructed by the sea surface with h 2 = 10 m and variable h 2 .3: Interpolation/Extrapolation of Field Strength for Transmission Frequency The field strength at the transmission frequency is interpolated from the graphs available for the upper and lower nominal frequencies as follows: Log  f  f Low  E f = E Low +  E Up – E Low   ----------------------------------Log  f Up  f Low  Where f Low is the lower nominal frequency (100 MHz if f < 600 MHz.5  d + R . The field strength values given by the graphs for land paths are for a reference receiver antenna at a height. f Up is the higher nominal frequency (600 MHz if f < 600 MHz. Examples of reference heights are 20 m for an urban area. These distances are given by 83 . 30 m for a dense urban area.  R'  • For land path other zones h C Receiver =  3. the elevation angle of the arriving ray is taken into account by calculating a modified representative clutter  1000  d  R – 15  h 1  height R' .5 Step 5: Calculation of Correction Factors Step 5. given by R' = Max  1 ----------------------------------------------------------.0108  f   R' – h 2   arc tan  --------------- . For sea paths the notional value of R is 10 m. E Low is the field strength value for d Low .3.03 – J    for h 2  R'  C Receiver =  h   3. representative of the height of the clutter surrounding the receiver.1: Correction for Receiver Antenna Height The receiver antenna height correction depends on the type of path and clutter in which the receiver is located.9 + 20  Log    – 0. d Up is the higher value of the nearest tabulated distance to d . and E Up is the field strength value for f Up . For land paths. R'  R .2  Log  f    Log  ------ . 600 MHz otherwise). subject to a minimum height value of 10 m. Step 4.1.2 + 6. C Receiver is reduced by  3.2  Log  f    Log  ----2- for h 2  R'  R'   R' – h 2 2 With J    =  6. The above equation is used for all land paths and sea paths. The different correction factors are calculated as follows: • For land path in urban and suburban zones  6.1   and  = 0. respectively.10.1  + 1 +  – 0. and 10 m for a suburban area.2 + 6.   1000  d – 15 Note that for h 1  6. In the case of transmission frequencies below 100 MHz or above 2000 MHz. the field strength are linearly interpolated or extrapolated for the logarithm of the distance using the following equation: Log  d  d Low  E d = E Low +  E Up – E Low   ------------------------------------Log  d Up  d Low  Where d Low is the lower value of the nearest tabulated distance to d . R (m). E Low is the field strength value for f Low . 2. and E Up is the field strength value for d Up .  27  10 If R'  10 m .0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 If the transmitter-receiver distance does not coincide with the list of distances for which the field strengths are accurately available from the graphs. E f .2 + 6.9 + 20  Log    – 0.11 Sakagami Extended Propagation Model The Sakagami extended propagation model is based on the simplification of the extended Sakagami-Kuboi propagation model. • h2 If h 2  10 m .1  + 1 +  – 0.1.85  Log  d    1 – 0. It is added to the field strength and is given by: C Building = – 3.46  Log  1 + h a – R   Where h a is the antenna height above the ground. in urban and suburban zones.. Therefore. The correction is added to the field strength and is given by: C Clearance = J  '  – J    2 Where J    =  6. Step 5. C Receiver =  3. This correction gives more precise field strength prediction over small reception areas. The Sakagami extended propagation model is valid for frequencies above 3 GHz. The calculated field strength is given by: E Calc = E f + C Receiver + C Building + C Clearance The resulting field strength is given by E = Min  E Calc E Max  . over a transmitter-receiver distance less than 15 km.3. it is only available in WiMAX documents by default.065   Clearance  f  Clearance is the clearance angle in degrees determined from: • •  : The elevation angle of the line from the receiver which just clears all terrain obstructions in the direction of the transmitter over a distance of up to 16 km but not going beyond the transmitter. L B ) is calculated as follows: L B = 139 – E + 20  Log  f  2.2  Log  f    Log  ------ 10 • h2 If h 2  10 m and d  d 10 . and over a transmitter-receiver distance less than 16 km. respectively. This correction takes into account the presence of buildings in these zones.Atoll 3. C Receiver =  3.3: Correction for Receiver Clearance Angle This correction is only applied when the path loss is to be calculated over land paths.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 D f  D h . from which the path loss (basic transmission loss. This correction is only applied when d  15 km and h 1 – R  150 m .2  Log  f    Log  ------   ----------------------------------  10  Log  d 10  d h2  Step 5.6 Step 6: Calculation of Path Loss First.3  Log  f    1 – 0. 2. C Receiver = 0 • Log  d  d h2  h2 If h 2  10 m and d  d 10 and d  d h2 .2 + 6. the final field strength is calculated from the interpolated/extrapolated field strength.036  f .  Ref : The reference angle.2: Correction for Short Urban/Suburban Paths This correction is only applied when the path loss is to be calculated over land paths.001 --------------- D f + D h explained earlier.  Ref = arc tan  ------------------- 1000  d Where h 1S and h 2S are the heights of the transmitter and the receiver above sea level. h 1S – h 2S .2 + 6. ' = 0.as d 10 = D 0. The buildings are assumed to be of uniform height. C Receiver =  3.6  f h 1 h 2  (km). Here D 0. 84 . and R is the clutter height of the clutter class where the receiver is located.2  Log  f    Log  ------ 10 • If h 2  10 m and d  d 10 and d  d h2 .6  f h 1  h 2 = 10 m   and d h2 = D 0. respectively.1   . and  = 0. by applying the corrections calculated earlier. The correction represents a reduction in the field strength due to building clutter.6 = Max  0.10. The path loss formula for the extended Sakagami-Kuboi propagation model is: L Model = 54 + 40  Log  d  – 30  Log  h b  + 21  Log  f  + a Where a is a corrective factor with three components: hm  H W a = a  H 0  + a  W  + a  h m  = 11  Log  -----0- – 7. heights of the buildings close to the receiver.1  Log  ------ – 5  Log  -----20 20 1. • • • • • • • • • W is the width (in meters) of the streets where the receiver is located  is the angle (in degrees) formed by the street axes and the direction of the incident wave hs is the height (in meters) of the buildings close to the receiver H1 is the average height (in meters) of the buildings close to the receiver hb is the height (in meters) of the transmitter antenna with respect to the observer hb0 is the height (in meters) of the transmitter antenna with respect to the ground level H is the average height (in meters) of the buildings close to the base station d is the separation (in kilometres) between the transmitter and the receiver f is the frequency (in MHz) The Sakagami-Kuboi propagation model is valid for: 5m <W< 50 m 0° < < 90° 5m < hs < 80 m 5m < H1 < 50 m 20 m < hb < 100 m 0.4  Log  h s  + 6.37 – 3.1 km <d< 3 km 0.5 m < hm < 5m 85 . such as widths of the streets where the receiver is located.23  Where.1  Log  h b    Log  d  + 20  Log  f  + e 13   Log  f  – 3.Atoll 3. which also allows a simplification in terms of the input required by the model.5 • • • • • • • W is the width (in meters) of the streets where the receiver is located H0 (= hs = H1) is the height (in meters) of the buildings close to the receiver hb (= hb0) is the height (in meters) of the transmitter antenna with respect to the ground hm is the height (in meters) of the receiver antenna H is the average height (in meters) of the buildings close to the base station d is the separation (in metres) between the transmitter and the receiver f is the frequency (in GHz) The extended Sakagami-Kuboi propagation model is valid for: 5m <W< 50 m 10 m < H0 < 30 m 10 m < hb < 100 m 0.1  Log  H 1  – 24.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 The Sakagami-Kuboi propagation model requires detailed information about the environment. the angles formed by the street axes and the directions of the incident waves. The path loss formula for the Sakagami-Kuboi propagation model is [1]: H 2 L Model = 100 – 7.1  Log  W  + 0.7   --------  Log  h b  + h b0  43.5 km <d< 10 km 450 MHz <f< 2200 MHz h b0  H Studies [2] have shown that the Sakagami-Kuboi propagation model can be extended to frequencies higher than 3 GHz.2 – 3.023   + 1. etc.8 GHz <f< 8 GHz 1.3. which corresponds to the ratio of the obstruction height (h) and the radius of the Fresnel zone (R). d is the Tx-Rx distance in km.4 (calculated for 3. depends on the obstruction parameter ().5 GHz) K2 40 K3 -30 K4 0 K5 0 K6 0 K7 -5 For more information on the Standard Propagation Model. Jesus Perez-Arriaga. 2. see "Standard Propagation Model (SPM)" on page 65. Catedra. L FS = 32. "Path Loss Prediction Formula for Urban and Suburban Areas for 4G Systems.3. 1999. Free space loss is stated in dB. References: • • [1] Manuel F.13. In Atoll.12 Free Space Loss The calculation of free space loss is based on ITU 525 recommendations. "Cell Planning for Wireless Communications. [2] Koshiro Kitao.1 km <d< 3 km 3 GHz <f< 8 GHz 1.4 + 20 log  f  + 20 log  d  where. 86 ." IEEE. f is the frequency in MHz. the path loss predicted by the extended model is almost independant of the input parameters such as street widths and angles. Calculations take the earth curvature into account through the effective Earth radius concept (K factor=1. The diffraction loss. 2. All of them are based on this same physical principle presented hereafter.13 Diffraction The calculation of diffraction is based on ITU 526-5 recommendations. Shinichi Ichitsubo. the extended Sakagami-Kuboi propagation model can be simplified to the extended Sakagami propagation model: L Model = 54 + 40  Log  d  – 30  Log  h b  + 21  Log  f  – 5  Log  h m  The extended Sakagami propagation model is valid for: 10 m < hb < 0.5 m < hm < 5m 100 m The path loss calculation formula of the Sakagami extended propagation model resembles the formula of the Standard Propagation Model. but differ in the way they consider one or several obstacles.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Studies also show that above 3 GHz. Therefore. J(). Four construction modes are implemented in Atoll. General method for one or more obstacles (knife-edge diffraction) is used to evaluate diffraction losses (Diffraction loss in dB).Atoll 3.1 Knife-edge Diffraction The procedure checks whether a knife-edge obstructs the first Fresnel zone constructed between the transmitter and the receiver." Artech House Publishers. 2006. 2. this model is in fact a copy of the Standard Propagation Model with the following values assigned to the K coefficients: K1 65.333). Deygout method is based on a hierarchical knifeedge sorting used to distinguish the main edges.2 3 Knife-edge Deygout Method The Deygout construction.Atoll 3.99792 x108 ms-1).1   Else. J    = 0 In case of multiple-knife edge method. if   – 0.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Figure 2. d2 is the distance from obstacle to receiver in m. 2. Hence.1  + 1 +   – 0. and secondary edges. The edge hierarchy depends on the obstruction parameter () value. which induce the largest losses. We have:  = h --r where. This method is used to evaluate path loss incurred by multiple knife-edges. R r = ------2 h is the obstruction height (height from the obstacle top to the Tx-Rx axis). J    = 6.15: Knife-Edge Diffraction R = c0  n  d1  d2 -------------------------------f   d1 + d2  where. n is the Fresnel zone index. limited to a maximum of three edges.78. f is the frequency in Hz d1 is the distance from the transmitter to obstacle in m. 2 For 1 knife-edge method.3.7 .13. c0 is the speed of light (2. which have a lesser effect. the minimum  required to estimate diffraction loss is -0.9 + 20  log    – 0. 87 . is applied to the entire profile from transmitter to receiver. DiffractionLoss = J   P  88 . if  P  0 we have DiffractionLoss = J   P  + J   t  + J   r  Otherwise.3.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 1 Obstacle Figure 2. Therefore. Therefore. Figure 2. di. are called ‘secondary edges’. p. Therefore.Atoll 3. The point with the highest  value is termed the principal edge. another half. (points t and r). J(t) and J(r). the total loss is evaluated by adding all the intermediary losses obtained. is also recorded. Losses induced by the secondary edges. i are evaluated from these data. between the transmitter and the knife-edge section.17: Deygout Construction – 3 Obstacles The same procedure is repeated on each half profile to determine the edge with the higher . constituted by the knife-edgereceiver section.7 . the main edge (point p) is considered as a secondary transmitter or receiver. The obstruction position. and the corresponding loss is J(p). If  P  – 0. The two obstacles found. Once the edge hierarchy is determined. are then calculated.16: Deygout Construction – 1 Obstacle A straight line between transmitter and receiver is drawn and the height of the obstacle above the Tx-Rx axis. hi. is calculated. we have DiffractionLoss = J   P  3 Obstacles Then. the profile is divided in two parts: one half profile. hp. Diffraction loss term is determined as follows: • If  P  – 0.13. we have DiffractionLoss = J   P  +  J   t  + J   r    t • Otherwise DiffractionLoss = 0 J  P   - 1 Here.78 . The main edge position dp is recorded and p and J(p) are evaluated from these data.4 Deygout Method with Correction The Deygout method with correction (ITU 526-5) is based on the Deygout construction (3 obstacles) plus an empirical correction.13. we have DiffractionLoss = J   P  + t   J   t  + J   r  + C  • Otherwise DiffractionLoss = 0 J  P   - 1 and C = 8.Atoll 3.0 + 0. is applied over the entire profile. Diffraction loss term is determined as follows: • If  P  – 0.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 In case of ITU 526-5 and WLL propagation models. A straight line between the transmitter and the receiver is defined and the height of the 89 . t = min  ---------- 6 2. t = min  ---------- 6 the transmitter and the receiver. is recalculated according to the Epstein-Peterson construction. First. 2.04d with d = distance stated in km between Here.13. Deygout construction is applied to determine the three main edges over the whole profile as described above. Then. we have DiffractionLoss = J   P  + J   t  + J   r  2. the main edge height. hp is the height above a straight line connecting t and r points. C. limited to a single edge. Figure 2.5 Millington Method The Millington construction. we have DiffractionLoss = J   P  + J   t  + J   r  + C Otherwise DiffractionLoss = J   P  + C In case of ITU 526-5 propagation model. Two horizon lines are drawn at the transmitter and at the receiver. Therefore.78 .18: Epstein-Peterson Construction Therefore. If  P  0 .3.3 Epstein-Peterson Method The Epstein-Peterson construction is limited to a maximum of three edges. As a mobile passes under a shadow. and distance. is calculated. which is due to the mobile receiving multipath copies of a signal. Therefore. h and J(h) are evaluated using the same previous formulas. The shadowing effect is modelled by a log-normal (Gaussian) distribution. However.14 Shadow Fading Model Propagation models predict the mean path loss as a function of transmission and reception parameters such as frequency. etc.) make large shadows that cause variations in the path loss over long distances. the path loss to the mobile keeps varying from point to point.3. from these values. Therefore. as shown in Figure 2. hills. we have DiffractionLoss = J   h  Figure 2.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 intersection point between the two horizon lines above the Tx-Rx axis.Atoll 3. the predicted path loss between a transmitter and a receiver is constant. It is crucial to account for the shadow fading in order to predict the reliability of coverage provided by any mobile cellular system. Some paths undergo more loss while others are less obstructed and may have higher received signal strength. etc. hh. Shadow fading varies as the mobile moves. depending on the type of environment is called shadow fading (shadowing) or slow fading. The position dh is recorded and then. whose standard deviation  depends on the type of clutter. in reality different types of clutter may exist in the transmitter-receiver path. antenna heights. The variation of path loss with respect to the mean path loss values predicted by the propagation models.20: Log-normal Probability Density Function 90 . "Slow" fading implies that the variations in the path loss due to shadow fading occur comparatively slower than the fast fading effect (Rayleigh fading). Figure 2.19: Millington Construction 2. Therefore. The location of the receiver in different types of clutter causes variations with respect to the mean path loss values given by the path loss models. in a given environment and for a given distance. the path losses for the same distance could be different along paths that pass throught different types of environments.20 on page 90. Different types of clutter (buildings. while fast fading can vary even if the mobile remains at the same location or moves over very small distances. Therefore.. 309-310 [5] Laiho J.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Different clutter types have different shadowing effects. Leonard M. 180-198 [2] Holma H. “Systèmes de radiocommunications avec les mobiles” pp. How up-to-date the digital map is. Application Throughput/Timeslot. The definition (bin size) of the digital map.Atoll 3. The accuracy of this model depends upon: • • • • • The suitability of the range of standard deviation used for each clutter class. The number of clutter classes.3. a coverage prediction with a cell edge coverage probability of x % means that the signal level predicted on each pixel is reliable x % of the time.G. Wacker A. Coverage by GPRS/EDGE Coding Scheme. Ec/I0. Shadowing margins are calculated for a given cell edge coverage probability. 80-81 GSM GPRS EGPRS Documents The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 95. Novosad T.. “Antennas and propagation for Wireless Communication Systems” pp. Interfered Zones. The cell edge coverage probability is the probability of coverage at a pixel located at the cell edge. Toskala A. In these calculations. The accuracy of assignment of clutter classes... The shadowing margin is calculated for a given cell edge coverage probability. References: • • • • • [1] Saunders S. RLC Throughout/Timeslot. and depends on the model standard deviation (  model in dB) associated to the clutter class where the receiver is located. the shape of the Gaussian distribution curve remains similar. and do not depend on interference. Coverage by Signal Level. This shadowing margin is calculated for a given cell edge coverage probability and depends on the C/I standard deviation (  C  I in dB) associated to the clutter class where the receiver is located. and the overall predicted coverage area is reliable at least x % of the time. 309-315. and whether it is applied to predictions or simulations.20 on page 90.. each clutter type in Atoll can have a different standard deviation representing its shadowing characteristics. or Eb/Nt as explained below. The following sections explain how shadowing margins are calculated and applied to different technology documents. ( C  I calculations). “Radio network planning and optimisation for UMTS” pp. a shadowing margin ( M Shadowing – model ) is applied to the received signal level calculated for each pixel. Circuit Quality Indicator Analysis) and calculations in point analysis window’s Interference tab that require calculation of the received signal level and interference received from other base stations. and applied to signal level or C/I as explained below. the shadowing margin ( M Shadowing – C  I ) is applied to the ratio of the carrier power (C) and the interfering signal levels (I) received from the interfering base stations. and corresponds to the reliability of coverage that you are planning to achieve at the cell edge. • Signal Level-Based Predictions 91 . 10511053” [4] Remy J.. • Signal Level-Based Predictions Signal level-based predictions include coverage predictions (Coverage by Transmitter. For different standard deviations. “CDMA systems engineering handbook” pp. Therefore. In these calculations (signal level calculations). • Interference-Based Predictions Interference-based predictions include coverage predictions (Coverage by C/I Level. Siben C. Shadowing is applied to the predicted path loss differently depending on the technology. as shown in Figure 2. Cueugnet J. and Overlapping Zones) and calculations in point analysis tabs (Profile and Reception) that require calculation of the received signal level only. UMTS HSPA and CDMA2000 1xRTT 1xEV-DO Documents The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 95 and "Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97. “WCDMA for UMTS” [3] Jhong S. a cell edge coverage probability of 75 % means that the users located at the cell edge will receive adequate signal level during 75 % of the time. and applied to signal level. For example. • Signal Level-Based Predictions Signal level-based predictions include coverage predictions (Best Server and RSCP P-CCPCH Coverages.) that require calculation of the received signal level and interference received from other base stations. • Macro-Diversity Gains UL DL Atoll calculates the uplink and downlink macro-diversity gains ( G macro – diversity and G macro – diversity ) depending on the receiver handover status. Handoff Status. The shadowing margin is calculated for a given cell edge coverage probability. or ) are applied to Eb/Nt.   Eb  Nt  DL .Atoll 3. or   Eb  Nt  UL . Baton Handover Coverage.3.   Eb  Nt  DL . in dB) associated to the clutter class where the receiver is located. and do not depend on interference. These gains are respectively taken into account to evaluate the uplink Eb/Nt in case of soft handover and the downlink Ec/Io from best server. In these calculations (signal level calculations). TD-SCDMA Documents The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 95 and "Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97. A shadowing margin for each transmitter-receiver link in each simulation is obtained by taking a random value from the probability density distribution for the appropriate clutter class. Cell to Cell Interference. and do not depend on interference. a shadowing margin ( M Shadowing – model ) is applied to the received signal level calculated for each pixel. etc. or M Shadowing –  Eb  Nt  UL ) are applied to Ec/I0 or Eb/Nt. please refer to "Macro-Diversity Gains Calculation" on page 97. In these calculations. M Shadowing –  Eb  Nt  DL .) and point analysis (AS Analysis tab) that require calculation of the received signal level and interference and noise received from other base stations. and Scrambling Code Interference) and calculations in point analysis tabs (Profile and Reception) that require calculation of the received signal level only. The shadowing margin is calculated for a given cell edge coverage probability. and depends on the model standard deviation (  model in dB) associated to the clutter class where the receiver is located. a shadowing margin ( M Shadowing – model ) is applied to the received signal level calculated for each pixel. • Monte-Carlo Simulations Random values for shadowing margins are calculated for each transmitter-receiver link and applied to the predicted signal level. A shadowing margin for each transmitter-receiver link in each simulation is obtained by taking a random value from the probability density distribution for the appropriate clutter class. The probability distribution is a lognormal distribution as explained above. in dB) associated to the clutter class where the receiver is located. DwPCH and UpPCH Coverages. the shadowing margins ( M Shadowing – Ec  Io . In these calculations. • Interference+noise-Based Predictions Interference+noise-based predictions include coverage predictions (Pilot Quality Analysis. Service Area Analyses. The probability distribution is a lognormal distribution as explained above. and Overlapping Zones) and calculations in point analysis tabs (Profile and Reception) that require calculation of the received signal level only. M Shadowing –  Eb  Nt  DL . In these calculations (signal level calculations). These shadowing margins are calculated for a given cell edge coverage probability and depend on the Ec/I0 or Eb/Nt standard deviations (  Ec  Io . and applied to signal level or interference+noise predictions as explained below. the shadowing margins ( M Shadowing –  Eb  Nt  M Shadowing –  Eb  Nt  UL P – CCPCH . • Interference+noise-Based Predictions Interference+noise-based predictions include coverage predictions (P-CCPCH Eb/Nt and C/I Coverages. These shadowing margins are calculated for a given cell edge coverage probability and depend on the Eb/Nt standard deviations (   Eb  Nt  P – CCPCH . and depends on the model standard deviation (  model in dB) associated to the clutter class where the receiver is located. Downlink Total Noise. Service Area Analsyses for downlink and uplink Eb/Nt and C/I. or   Eb  Nt  UL . For detailed description of the calculation of macro-diversity gains. etc. 92 . • Monte-Carlo Simulations Random values for shadowing margins are calculated for each transmitter-receiver link and applied to the predicted signal level. Coverage by Signal Level. P-CCPCG Pollution Analysis.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Signal level-based predictions include coverage predictions (Coverage by Transmitter. Coverage by Throughput. • Interference+noise-Based Predictions Interference-based predictions include coverage predictions (Coverage by C/(I+N) Level. • I P : The predicted received interference power without any shadowing margin. and depends on the model standard deviation (  model in dB) associated to the clutter class where the receiver is located.Atoll 3. • Signal Level-Based Predictions Signal level-based predictions include coverage predictions (Coverage by Transmitter. etc. M Shadowing – C  I is calculated for a given cell edge coverage probability and depends on the C/I standard deviation (  C  I in dB) associated to the clutter class where the receiver is located. • m C : Shadowing margin based on the model standard deviation ( 10 • m C  I : Shadowing margin based on the C/I standard deviation ( 10 • N : Thermal noise M Shadowing – model ---------------------------------------------------------10 M Shadowing – C  I -------------------------------------------------10 ) ) Calculations The effective received carrier power is given by: C = mC  CP The effective C/I is given by: CP C --. the effective C/(I+N) is given by: mC  CP C . In these calculations (signal level calculations). the ratio M Shadowing – model – M Shadowing – C  I is applied to the interfering signal levels (I). The reason why the ratio M Shadowing – model – M Shadowing – C  I is used can be understood from the following derivation (linear. The shadowing margin is calculated for a given cell edge coverage probability. not it dB): Inputs • C P : The predicted received carrier power without any shadowing margin.) that require calculation of the received signal level and interference.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 WiMAX Documents The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 95 and "Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97 . mC  I Therefore. Coverage by Signal Level. and applied to signal level or C/(I+N) as explained below. (C/(I+N) calculations).= ---------I = ---------------------mC  I P CP CP m C  I  ----m C  I  ----IP IP mC Where ----------.= -------------------------------------------------I + N m C  --------- -I +N  mC  I P  • Monte-Carlo Simulations 93 .3. and do not depend on interference. In these calculations. and Overlapping Zones) and calculations in point analysis tabs (Profile and Reception) that require calculation of the received signal level only.= m C  I  ----IP I The above equations lead to: mC mC  CP C . a shadowing margin ( M Shadowing – model ) is applied to the received signal level calculated for each pixel.corresponds to M Shadowing – model – M Shadowing – C  I in dB. in addition to the shadowing margin ( M Shadowing – model ) applied to the received signal level calculated for each pixel. Coverage by Bearer.= ----------------------I . Coverage by Throughput. a shadowing margin ( M Shadowing – model ) is applied to the signal level calculated for each pixel. M Shadowing – C  I is calculated for a given cell edge coverage probability and depends on the C/I standard deviation (  C  I in dB) associated to the clutter class where the receiver is located.) that require calculation of the received signal level and received interference. not it dB): Inputs • C P : The predicted received carrier power without any shadowing margin. etc. A shadowing margin for each transmitter-receiver link in each simulation is obtained by taking a random value from the probability density distribution for the appropriate clutter class. The reason why the ratio M Shadowing – model – M Shadowing – C  I is used can be understood from the following derivation (linear. • m C : Shadowing margin based on the model standard deviation ( 10 • m C  I : Shadowing margin based on the C/I standard deviation ( 10 • N : Thermal noise M Shadowing – model ---------------------------------------------------------10 M Shadowing – C  I -------------------------------------------------10 Calculations The effective received carrier power is given by: C = mC  CP The effective C/I is given by: C C --.Atoll 3.= ---------I = ---------------------mC  I P CP CP m C  I  ----m C  I  ----IP IP mC Where ----------. and Overlapping Zones) and calculations in point analysis tabs (Profile and Reception) that require calculation of the received signal level only. • Interference+noise-Based Predictions Interference-based predictions include coverage predictions (Coverage by C/(I+N) Level. (C/(I+N) calculations). mC  I Therefore.corresponds to M Shadowing – model – M Shadowing – C  I in dB. and do not depend on interference. Coverage by Bearer.= ----------------------I . the effective C/(I+N) is given by: 94 ) ) . • Signal Level-Based Predictions Signal level-based predictions include coverage predictions (Coverage by Transmitter.3. The shadowing margin is calculated for a given cell edge coverage probability. and applied to signal level or C/(I+N) as explained below.= m C  I  ----PIP I The above equations lead to: mC mC  CP C . In these calculations (signal level calculations). In these calculations. and depends on the model standard deviation (  model in dB) associated to the clutter class where the receiver is located. The probability distribution is a lognormal distribution as explained above.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Random values for shadowing margins are calculated for each transmitter-receiver link and applied to the predicted signal level. Coverage by Signal Level. • I P : The predicted received interference power without any shadowing margin. in addition to the shadowing margin ( M Shadowing – model ) applied to the signal level calculated for each pixel. LTE Documents The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 95 and "Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97 . the ratio M Shadowing – model – M Shadowing – C  I is applied to the interfering signal levels (I). • Lpath is the predicted path loss. and applied to the path loss. 2.14.1 Shadowing Margin Calculation in Predictions Shadowing margins.14. and required cell edge coverage probability. are calculated from standard deviation values defined for the clutter class where the pixel (probe mobile) is located. Shadowing Error PDF (1 Signal) The measured path loss in dB can be expressed as a Gaussian random variable: L = Lpath +  dB  G  0 1  where. I P + N  mC  I  • Monte-Carlo Simulations Random values for shadowing margins are calculated for each transmitter-receiver link and applied to the predicted signal level.3. 95 . The probability distribution is a lognormal distribution as explained above.1. Lpath.Atoll 3.= -------------------------------------------------m I + N C  ---------. The following shadowing margins are calculated using the method described below: Network Type Standard Deviation MShadowing Applied to  model M Shadowing – model C C  I M Shadowing – C  I C/I  model M Shadowing – model C  Ec  Io M Shadowing – Ec  Io Ec/I0 GSM GPRS EGPRS UMTS HSPA   Eb  Nt    Eb  Nt  CDMA2000 DL UL UL Eb/Nt (UL) C  Ec  Io M Shadowing – Ec  Io Ec/I0 DL UL  model   Eb  Nt  P – CCPCH   Eb  Nt    Eb  Nt  LTE Eb/Nt (DL) M Shadowing – model   Eb  Nt  WiMAX M Shadowing –  Eb  Nt  DL  model   Eb  Nt  TD-SCDMA M Shadowing –  Eb  Nt  DL UL M Shadowing –  Eb  Nt  M Shadowing –  Eb  Nt  DL Eb/Nt (DL) UL Eb/Nt (UL) M Shadowing – model M Shadowing –  Eb  Nt  P – CCPCH M Shadowing –  Eb  Nt  M Shadowing –  Eb  Nt  C Eb/Nt P-CCPCH DL Eb/Nt (DL) UL Eb/Nt (UL)  model M Shadowing – model C and C/(I+N) C  I M Shadowing – C  I C/(I+N)  model M Shadowing – model C and C/(I+N) C  I M Shadowing – C  I C/(I+N) 2. MShadowing. A shadowing margin for each transmitter-receiver link in each simulation is obtained by taking a random value from the probability density distribution for the appropriate clutter class.1 Shadowing Margin Calculation The following sections describe the calculation method used for determining different shadowin margins.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 mC  CP C . M Shadowing . L Rx are receiver losses. 2  z --------- dB e 2 x – ----2 z dx = Q  --------   dB where Q is the complementary cumulative function. Confidence in the prediction can be expressed as: C d = P' Tx – L  P rec  L  P' Tx – P rec  G  0 1    dB  M Shadowing where.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation • • ©Forsk 2015 dB is the user-defined standard deviation of the error. • G antRx is the receiver antenna gain. 96 . Random variations also exist in the interfering signals. the shadowing margin is only applied to the signal from the interfered transmitter (C).Atoll 3.21: Normalised Margin M arg in = ------------------------- dB In interference-based predictions. e  dB 2 2 x – -------------2 2 dB The probability that the shadowing error exceeds z dB is 2 x  – ------------2 2 dB  PL  x  z  =  pL  x  dx z 1 = -------------------. The shadowing margin is calculated such that: M Shadowing P  C d  P rec  = R L  M Shadowing  = 1 – P L  x – M Shadowing  0  = 1 – Q  -------------------------   dB A lookup table is used for mapping the values of Q vs. • P rec is the signal level predicted at the receiver. M Shadowing Figure 2. P rec = P' Tx – L path – M Shadowing • P' Tx = EIRP + G antRx – L Rx • • EIRP is the effective isotropic radiated power of the transmitter. R L . Therefore. where signal to noise ratio is calculated.1) is a zero-mean unit-variance Gaussian random variable. for the predicted value. [3] explains how a certain level of interference is maintained by congestion control in CDMA-based networks. is added to the link budget. We consider that the interference value is not altered by the shadowing margin. e  dB 2  dx z Normalising x by dividing it bydB:  1 P L  x  z  = ---------.3. a set of cell edge coverage probabilities. a shadowing margin. G(0. To ensure a given cell edge coverage probability. the probability density function (pdf) for the random (shadowing) part of path loss is: 1 p L  x  = -------------------. but taking only the average interference gives accurate results. and required cell edge coverage probability.2 Macro-Diversity Gains Calculation The following sections explain how uplink and downlink macro-diversity gains are calculated in UMTS HSPA and CDMA2000 1xRTT 1xEV-DO documents for predictions and AS Analysis tab of the point analysis tool. the method influences the calculated macro-diversity gain in case the mobile is in soft handover. 97 .2 Shadowing Margin Calculation in Monte-Carlo Simulations Shadowing margins. a standard deviation  -------------- 2  So. as shadowing errors on the transmitter-receiver links are uncorrelated. 2. Lpath. MShadowing. On the other hand. Here.1. an activity status. are calculated from standard deviation values defined for the clutter class where the pixel (probe mobile) is located. and remains the same for all links between the receiver and the base stations from which it is receiving signals.  P models the error related to the path between the transmitter and the receiver. a mobility type. which gives:  model  model  L = --------------. we have: 2 2 2  model =  L +  P 2 L  = --------------2  model Therefore. we have: 1  1 =  L +  P for link 1 2  2 =  L +  P for link 2 i Standard deviations of  L   L  and  P   P  can be calculated from  i .  is a zero mean gaussian random variable G  0  dB  representing variation due to shadowing. These values also have a zero-mean gaussian distribution with  model . and the correlation coefficient    between  1 and  2 . path loss (L) can be broken down to L = L path +  . Hence.. we have: Receiver Path E Shadowing – model = E Shadowing – model + E Shadowing – model Random shadowing error has its mean value at zero. a geographic position and a random shadowing value. It can be expressed as the sum of two uncorrelated zero mean gaussian random variables.Atoll 3.  L and  P .  L models the error related to the receiver’s location (surrounding environment). values are randomly generated for each receiver. Each user is assigned a service. Assuming all  P have the same standard deviations. and added to the path loss. Atoll generates another random value for each transmitter-receiver pair.5 in Atoll. to model shadowing error common to all the signals received at a receiver ( E Shadowing – model ).14. in case of two links. the model standard deviation   model  . where   model  is the model standard deviation associated with the receiver’s clutter class.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 2..14. This values represents the shadowing error Path not related to the location of the receiver ( E Shadowing – model ).and  P = --------------2 2 Receiver Therefore. this shadowing modelling method has no impact on the simulated network load. These values have a zero-mean gaussian distribution with a standard deviation of model  --------------.3. 2 2 2 2  P =  model   1 –    L =  model    is set to 0. Random values are generated during Monte-Carlo simulation. Therefore.  2  Next. For each link. Ni is the noise level for link i.1 Shadowing Error PDF (n Signals) For each link. It can be expressed as the sum of two uncorrelated zero mean gaussian random variables. As for one link.2. 2.  L models error related to the receiver local environment. For more information. In this case.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 The calculation and use of macro-diversity gains can be disabled through the Atoll. mobiles may be in soft handoff (mobile connected to cells located on different sites). 2 2  P =   Eb  Nt  2 UL 2  L =   Eb  Nt  UL  1 –   2 Signals Without Recombination In technologies supporting soft handoff (UMTS and CDMA2000). cell is interference limited. Therefore. it is the same whichever the link.14. path loss (L) can be broken down as: L = L path +   is a zero mean gaussian random variable G  0  dB  representing variation due to shadowing. we add to each link budget a shadowing margin.ini file. the uplink Eb/Nt standard deviation    Eb  Nt   and the correlation coefficient  between  1 and  2 .2.3.  L and  P .Atoll 3. 2.14. Prediction reliability in order to have Eb/Nt higher or equal to Eb/Nt from the best server can be expressed as: Cd 1 1 -------1 = P' Tx1 – L 1 – N 1  CI pred   1  P' Tx1 – L path – N 1 – CI pred 1 N1 or Cd 1 1 -------2 = P' Tx2 – L 2 – N 2  CI pred   2  P' Tx2 – L path – N 2 – CI pred 2 N2 where i CI pred is the quality level (signal to noise ratio) predicted at the receiver for link i.  P models error related to the path between transmitter and receiver. in case of two links.1 Uplink Macro-Diversity Gain Evaluation In UMTS HSPA and CDMA2000 1xRTT 1xEV-DO.1. see the Administrator Manual. to ensure a required cell edge coverage probability R L for the prediction. we can consider the shadowing error pdf described below. we have: 1  1 =  L +  P for the link 1 2  2 =  L +  P for the link 2 Knowing  i . 2signals M Shadowing –  Eb  Nt  UL . We have: 2   Eb  Nt  2 UL 2 = L + P 2 L  = ----------------------2   Eb  Nt  UL Therefore. We note: 98 . we can UL calculate standard deviations of  L   L  and  P   P  (assuming all  P have the same standard deviations). Therefore.3.e 2  –   2 –xL --------2 2signals 2signals 2  M Shadowing  M Shadowing – x L   Eb  Nt   – x L   Eb  Nt   –  1  –  Eb  Nt  –  Eb  Nt  UL UL UL UL     Q  -----------------------------------------------------------------------------------------------  Q  ------------------------------------------------------------------------------------------------------------ dx L 1– 1–   Eb  Nt    Eb  Nt       UL UL 99 .Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 2signals M Shadowing –  Eb  Nt  i UL = P' Txi – L path – N i – CI pred i and 2 1 2  1 = CI pred – CI pred 2  1 is the minimum needed margin on each link.  P and  P P  2signals   1  2  1 UL = P  L   P L P 2signals M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt   1 1 2  P  P  P 2signals  M Shadowing –  Eb  Nt  2signals   1  2  1 2signals 2signals UL 1 L 2signals P noMRC UL 2 UL 2 –  1  L =  L –  L  P  M Shadowing –  Eb  Nt  M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt  = P    L   P    P  M Shadowing –  Eb  Nt  RL 2 UL UL 2 UL – 1 – L  2 –  1 L =  L  2signals –  L  P    P  M Shadowing –  Eb  Nt  P 2 UL – 1 –  L  2signals  M Shadowing –  Eb  Nt   UL    1 2signals 2 2signals 2 =  1 – P    L   P    P  M Shadowing –  Eb  Nt  –  L   P    P  M Shadowing –  Eb  Nt  –  1 –  L  d L   L P UL P UL   –  i 2signals P    P  M Shadowing –  Eb  Nt  P    1  =  ----------------  P 2  2signals M   Shadowing –  Eb  Nt  UL – L   e UL –  L  2 –x --------2 2 P  2signals  M Shadowing –  Eb  Nt  UL –  L  dx = Q  ---------------------------------------------------------------- P     Then. the probability of having a quality at least equal to the best predicted one is: noMRC RL noMRC RL Cd  Cd  2signals 1 1  M Shadowing –  Eb  Nt   = 1 – P L1 L2  -------1  CI pred -------2  CI pred UL N N  1  2 2signals  M Shadowing –  Eb  Nt   = 1 – P  UL 1 1  2 2signals 2signals   1  M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt  UL 2 UL – 1  2 We can express it using  L . we have: noMRC RL 2signals  M Shadowing –  Eb  Nt   UL  2signals  M Shadowing –  Eb  Nt UL 2signals 2   –  L  M Shadowing –  Eb  Nt  UL –  1 –  L =  1 – P    L   Q  ----------------------------------------------------------------  Q  ---------------------------------------------------------------------------- d L   L P P       –  If we introduce user defined standard deviation    Eb  Nt   and correlation coefficient    . and consider that P  is a UL L Gaussian pdf: noMRC RL 2signals  M Shadowing –  Eb  Nt   UL    1 =  1 – ---------. Atoll 3. If angle is 0 and lengths are the same.Atoll may consider up to 3 signals): noMRC RL nsignals  M Shadowing –  Eb  Nt   UL    1 =  1 – ---------. In a normal handover status. To model this function.  L and  P .  P models the error related to the path between the transmitter and the receiver.  is close to  (+/. in case of two links.2. Correlation Coefficient Determination There is currently no agreed model for predicting correlation coefficient    between  1 and  2 .  L models the error related to the receiver local environment.when  T     D2  T is a function of the mean size of obstacles near the receiver and  is also linked to the receiver environment. assuming a hexagonal design for sites. in case of soft handoff. A simple model has been found [1]: T  =  -----   D1 ------.1.14. The Softer/soft case is equivalent to the two signals case.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 n Signals Without Recombination We can generalize the previous expression to n signals (n is the number of available signals . 2.  = 0. in the shadowing margin calculation. we will use combined SNR to calculate the availability of the link. mobiles are able to switch from one cell to another if the best pilot drastically fades. Therefore. we have: UL nsignals G macro – diversity = M Shadowing –  Eb  Nt  UL – M Shadowing –  Eb  Nt  UL Where n is the number of cell-mobile signals. Let us consider the shadowing error pdf described below. correlation is high. path loss (L) can be broken down as: L = L path +   is a zero mean gaussian random variable G  0  dB  representing variation due to shadowing. 2.2.2. we have: 100 .5 when  = 0. which is the same for all links.14. Correlation is different from zero when path lengths differ. both for the best signal and for all the other available signals. 10 In Atoll. The relative values of the two signal lengths. Two key variables influence correlation: • • The angle between the two signals.14.  In [1.5.2 Downlink Macro-Diversity Gain Evaluation In UMTS HSPA and CDMA2000 1xRTT 1xEV-DO. If this angle is small.2 Uplink Macro-Diversity Gain UL Atoll determines the uplink macro-diversity gain ( G macro – diversity ) from the shadowing margins calculated in case of one signal and n signals. 2.3 and  T = -----.5].  is set to 0. It can be expressed as the sum of two uncorrelated zero mean gaussian random variables.1 Shadowing Error PDF (n Signals) For each link.. correlation is zero./3) and D1/D2 is close to 1. we have to consider the probability of fading over the shadowing margin.e 2  –   2 –xL --------2 2  M nsignals  M nsignals – x L   Eb  Nt   – x L   Eb  Nt   –  1  Shadowing –  Eb  Nt  Shadowing –  Eb  Nt  UL UL UL UL     Q  -----------------------------------------------------------------------------------------------  Q  ------------------------------------------------------------------------------------------------------------ dx L 1– 1–   Eb  Nt    Eb  Nt       UL UL The case where softer handoff occurs (two signals from co-site cells) is equivalent to the one signal case.2. Therefore. For the path associated with the softer recombination.3.   ----- Io Io pred Io Io pred 2signals 2signals 2signals 2  M Shadowing – Ec  Io  = 1 – P 1 2   1  M Shadowing – Ec  Io  2  M Shadowing – Ec  Io –  1  1 2 We can express it by using  L .  ----- for the best server can be expressed as: Io Io pred Ec Ec 1 Ec 1 -------1.  P and  P 2signals 2signals 2 P 1 2   1  M Shadowing – Ec  Io  2  M Shadowing – Ec  Io –  1  L =  L  = P  L   P L 1 1 2  P  P  P 2signals 2 2signals 2  M Shadowing – Ec  Io –  L P  M Shadowing – Ec  Io –  1–  L  2signals 2signals 2 P 1 2   1  M Shadowing – Ec  Io  2  M Shadowing – Ec  Io –  1  L =  L 1 2signals 2 2signals 2 = P    L   P    P  M Shadowing – Ec  Io –  L   P    P  M Shadowing – Ec  Io –  1 –  L  L P P 101 .  -----  -------.= P pilot – L 1 – Io   -----   1  P pilot – L m – Io –  -----  Io  pred  Io  pred 1 1 1 Io Or Ec Ec 1 Ec 1 -------2.Atoll 3. to each link budget. Ec Ec Prediction reliability to have ----. As for one link.3.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 1  1 =  L +  P for the link 1 2  2 =  L +  P for the link 2 Knowing  i . Therefore. to ensure a 2signals required cell edge coverage probability R L for the prediction. we can calculate standard deviations of  L   L  and  P   P  (assuming all  P have the same standard deviations). We have: 2 2 2  Ec  I o =  L +  P 2 L  = ------------2  Ec  I o Therefore. we add a shadowing margin.= P pilot – L 2 – Io   -----   2  P pilot – L m – Io –  -----  Io  pred  Io  pred 2 2 2 Io We note: 1 Ec 2signals M Shadowing – Ec  Io = P pilot – L m – Io –  -----  Io  pred i i Ec 1 Ec 2 2  1 =  ----- –  ----- Io pred Io pred 2  1 is the minimum needed margin on each link. probability of having a quality at least equal to the best predicted one is: noMRC RL noMRC RL Ec 2 Ec 1 Ec 1 Ec 1  2signals  M Shadowing – Ec  Io  = 1 – P L1 L2  -------. 2 2 2 2  P =  Ec  I o   1 –    L =  Ec  I o   2 Available Signals In technologies supporting soft handoff (UMTS and CDMA2000) cells are interference limited. the Ec/Io standard deviation   Ec  I o  and the correlation coefficient  between  1 and  2 . M Shadowing – Ec  Io . 22: Margin .0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation noMRC RL ©Forsk 2015 2signals  M Shadowing – Ec  Io   = 1–  PL  L   PP  P  MShadowing – Ec  Io – L   PP  P  MShadowing – Ec  Io – 1 – L  dL 1 2signals 2 2signals 2 –  1 i 2signals P    P  M Shadowing – Ec  Io –  L  = ----------------P  P 2   SHO –  L e 2 –x --------2 2 P 2signals  M Shadowing – Ec  Io –  L - dx = Q  ----------------------------------------------------P   Then. we have:  noMRC 2signals RL  M Shadowing – Ec  Io  = 1– 2signals  – 2signals 2  M Shadowing – Ec  Io –  L  M Shadowing – Ec  Io –  1 –  L P    L   Q  ------------------------------------------------------  Q  ------------------------------------------------------------------ d L L  P  P    If we introduce a user defined Ec/Io standard deviation    and a correlation coefficient    and consider that P  is a L Gaussian pdf: noMRC RL 2signals  M Shadowing – Ec  Io   1 = 1 – ---------2  e 2 –xL --------2 – 2signals 2signals 2  M Shadowing – Ec  Io – x L  Ec  I o   M Shadowing – Ec  Io –  1 – x L  Ec  I o   Q  ----------------------------------------------------------------------------  Q  ---------------------------------------------------------------------------------------- dx L  Ec  I o 1 –   Ec  I o 1 –      n Available Signals We can generalize the previous expression for n signals (n is the number of available signals .3.Atoll may consider up to 3 signals): noMRC RL nsignals  M Shadowing – Ec  Io   1 = 1 – ---------2  – e 2 –xL --------2 nsignals  M Shadowing – Ec  Io – x L  Ec  I o  - x  Q  -------------------------------------------------------------------------- Ec  I o 1 –    n  i=2 nsignals i  M Shadowing – Ec  Io –  1 – x L  Ec  I o  Q  ---------------------------------------------------------------------------------------- dx L  Ec  I o 1 –    2  1 =1 dB 2  1 =5 dB 2  1 =10 dB Figure 2.Atoll 3.Probability (Case of 2 Signals) 102 . we have: DL nsignals G macro – diversity = M Shadowing – Ec  Io – M Shadowing – Ec  Io Where n is the number of available signals.2 Downlink Macro-Diversity Gain DL Atoll determines the downlink macro-diversity gain ( G macro – diversity ) from the shadowing margins calculated in case of one signal and n signals. 103 . a transmitter fulfilling the conditions detailed above will be called TBC transmitter. and a second matrix over a larger radius computed with a low resolution and another propagation model (extended matrix). delta1 = 1dB) 2 signals 3  1 =5 dB 3  1 =10 dB Figure 2.Atoll 3.Probability (Case of 3 Signals with sigma = 8dB.3. Atoll determines a path loss value ( L path ) on each calculation bin (calculation bin is defined by the resolution) of the calculation area of the TBC transmitter.2. 2. In the rest of the document. and It must have a calculation area.2. To be considered for calculations.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 2 signals 3  1 =5 dB 3  1 =10 dB Figure 2. delta1 = 2dB) Correlation Coefficient Determination For further information about determination of the correlation coefficient.14.Probability (Case of 3 Signals with sigma = 8dB.15 Path Loss Matrices Atoll is able to calculate two path loss matrices per transmitter. a first matrix over a smaller radius computed with a high resolution and a propagation model (main matrix). You may have one or two path loss matrices per TBC transmitter. 2. It must satisfy filter criteria defined in the Transmitters folder. please see "Correlation Coefficient Determination" on page 103. a transmitter must fulfil the following conditions: • • • It must be active.23: Margin .24: Margin . Therefore. The path loss matrix size of a TBC transmitter depends on its calculation area. 3.26: Example 2: Multiple Calculation Areas Computation zone Rectangle containing the computation zone(s) Calculation area defined (square) Transmitter Actual calculation area on which Atoll calculates path losses 2.15. Dx and Dy) Invalid Path loss matrices Sufficient Not necessary Antennaa height Invalid Path loss matrices Sufficient Not necessary . One side of the square equals twice the entered calculation radius. Figure 2. transmitter calculation area corresponds to the intersection area between its calculation square and the rectangle containing the computation zone area(s). This table lists these modifications and also changes that have an impact only on coverage predictions.2 Validity of Path Loss Matrices Most geographic data modifications and some radio data changes make path loss matrices invalid. Since the computation zone can be made of one or several polygons.15.1 Calculation Area Determination Transmitter calculation area is made of a rectangle or a square depending on transmitter calculation radius and the computation zone.25: Example 1: Single Calculation Area Figure 2. Calculation radius enables Atoll to define a square around the transmitter. 104 Modification Matrix validity Impact on Calculate Force calculation Frequency Invalid Path loss matrices Sufficient Not necessary Antenna coordinates (site coordinate: X and Y.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 2.Atoll 3. 0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 a. Except if this action has an impact on the site positions/altitudes. 2.3.Atoll 3. All these tuning paths are stored in a catalogue. This catalogue is stored under a . Since a tuning file can contain several measurement paths.3 Path Loss Tuning Atoll can tune path loss matrices obtained from propagation results by the use of real measurements (CW Measurements or Test Mobile Data). Modification Matrix validity Impact on Calculate Force calculation Antennaa pattern Invalid Path loss matrices Sufficient Not necessary Downtilta Invalid Path loss matrices Sufficient Not necessary Azimutha Invalid Path loss matrices Sufficient Not necessary % Power (secondary antennas) Invalid Path loss matrices Sufficient Not necessary Site position/altitude Invalid Path loss matrices Sufficient Not necessary Grid resolution (main or extended) Invalid Path loss matrices Sufficient Not necessary Propagation model (main or extended) Invalid Path loss matrices Sufficient Not necessary Propagation model parameters Invalid Path loss matrices Sufficient Not necessary Calculation areas (Calculation areas gets smaller) Valid Coverage predictions Sufficient Not necessary Calculation areas (Calculation areas gets larger) Invalid Path loss matrices Sufficient Not necessary Receiver height Invalid Path loss matrices Sufficient Not necessary Receiver losses Valid Coverage predictions Sufficient Not necessary Receiver gain Valid Coverage predictions Sufficient Not necessary Receiver antenna Valid Coverage predictions Sufficient Not necessary Geographic layer order Invalid Path loss matrices Insufficientb Necessary Geographic file resolution Invalid Path loss matrices Insufficientb Necessary New DTM map Invalid Path loss matrices Insufficientb Necessary Clutter class edition Invalid Path loss matrices Insufficientb Necessary Coverage prediction resolution Valid Coverage predictions Sufficient Not necessary Cell edge coverage probability Valid Coverage predictions Sufficient Not necessary Coverage prediction conditions Valid Coverage predictions Sufficient Not necessary Coverage prediction display options Valid Coverage predictions Sufficient Not necessary Modification of any parameter related to main or other antennas makes matrix invalid.15.dbf file and one . Atoll tries to merge measurements and predictions on the same points and to smooth the surrounding points of the path loss matrices for homogeneity reasons. 105 . all these measurements are added to the tuning file. A transmitter path loss matrix can be tuned several times by the use of several measurement paths.tuning folder containing a . For each measured transmitter.pts file per tuned transmitter. b. Mean error is calculated for each pathloss matrix (main and extended) of each transmitter.The ellipse is user-defined by two parameters: • • The radius of the axis parallel to the Profile (A) The radius of the axis perpendicular to the Profile (B) Let’s take M a measurement value and P i the path loss value at point i. Path Losses tuning will be done using two steps.Atoll 3. Local correction for each measured value For each measured value. Then. 2. a second tuning ( R i ) is applied where: Ri =  1 – Di    M – g – Pi new  so R i =  1 – D i    M – g –  P i old + E Where g is (measurement gain . The main axis of the ellipse is oriented to the transmitter. M is limited by the minimum measurement threshold defined in the interface. the final tuned path loss is: Pi 106 tuned = Pi new + R i so P i tuned = Pi old + E + Ri . the path loss value is tuned using E: Pi new = Pi old +E Finally.1 Transmitter Path Loss Tuning The same algorithm is used for CW Measurement and Test Mobile Data. see the Administrator Manual.3.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 For more information on the tuning files.+ ----------------------2 2 A B Where: X i and X M are the X-coordinates of i and M respectively Y i and Y M are the Y-coordinates of i and M respectively The mean error for the first tuning is given by: 1 E =  ---   n  ei i Where e i is the error between measurement and prediction at point i E is limited by the maximum total correction defined in the interface. So.losses). Total matrix correction A mean error is calculated between each measured value and the corresponding bin in the pathloss matrix. 2.3. before any tuning. an ellipse is used to define the pathloss area which has to be tuned. The squared elliptic distance between i and M is given by: 2 2  Xi – XM   Yi – YM  D i = ----------------------.15. This tuning is done to smooth the local corrections (step 2) of measured values and not the tuned bins. It is also the same for main and extended matrices. 1. R i is limited by the maximum local correction defined in the interface. This mean error is then applied to all the matrix bins. 2 Repeater Path Loss Tuning In the case of repeaters. The formats of the pathloss.16. the donor transmitter (resp.1 Path Loss Matrix File Format When path loss matrices are stored externally.15. Atoll provides only a composite measured value per pixel which is a combination of the contribution of both a transmitter and one or several repeaters.dbf and LOS files are described here. The file can be opened in Microsoft Access. one can define the contribution of each element as follows: Cd Cr and M r = M  ---------------. D4 = dBaseIV.dbf file has a standard DBF (dBase III) format.dbf File Format The pathloss. D3 = dBaseIII+. Fp = FoxPro. M = Md + Mr where : M d represents the contribution of the donor transmitter in the measured value. but it should not be modified without consulting the Forsk customer support. In order to tune the path loss matrices of donor transmitters and repaters. If C d and C r represent respectively the filtered signal level from the donor transmitter and the repeater on a pixel. D5 = dBaseV. CL = Clipper 107 .16.3. For general information.e. M r represents the contribution of the repeater in the measured value. outside the ATL file. the path loss matrices folder contains a ‘pathloss.16 File Formats 2.dbf file and one LOS (path loss results) file per transmitter that has an extended path loss matrix. M r ) values.3.dbf’ file containing the calculation parameters of the transmitters and one LOS (path loss results) file per calculated transmitter.Atoll 3. i. Fb = FoxBase. the repeater) is then tuned using M d (resp. Let’s take M the measured value. the final tuned path loss is given by:     1 – d j P j  tuned   j = ------------------------------------------------  d j n –    Pi tuned  j Where n is the number of overlapping ellipses 2. M d = M  ---------------Cd + Cr Cd + Cr Following the path loss tuning process described in "Transmitter Path Loss Tuning" on page 106.1. the format of DBF files in any Xbase language is as follows: Notations used in the following tables: FS = FlagShip. All the values are used in Watts. The path loss matrices folder also contains a LowRes folder with another pathloss.. its is mandatory to split the contribution of each element in the measured value as starting point.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 When several ellipses overlap a pathloss bin. 2. 2.1 Pathloss. .Atoll 3. CL 0x04 plain . D5 0x00 normal visible All 16 12 0 (1) multi-user environment use D4.n. byte10) ….fmp memo Fp 01 3 YYMMDD Last update digits All 04 4 ulong Number of records in file All 08 2 ushort Header size in bytes All 10 2 ushort Record size in bytes All 12 2 0. Fp. D5 (FS) 0x05 plain . D5 0x00 index upon demand All 29 1 n language driver ID D4.dbf is not empty DBF Header The DBF header size is variable and depends on the field count.n Fld address in memory D3 .0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 DBF Structure Byte Description 0.dbv memo var size FS 0xB3 with . Fb.D5 28 1 0x01 production index exists Fp. Fb.dbt memo in D4 format D4.. D4.n DBF header (see next part for size. D3. last record last optional: 0x1a (eof byte) If . CL 0.n. CL 0x8B with .dbt memo FS 0x83 with . Fb.dbf FS. D5 0x00 ignored FS. D5 0xF5 with . D4. D4. D5 0x00 End Transaction D4. D3. Byte Size Contents Description Applies to 00 1 0x03 plain . D3. Fp (FS) 0x43 with . D5.3. D5. byte 8) Remarks n+1 1st record of fixed length (see next parts).0 Reserved All 14 1 0x01 Begin transaction D4. D5 0x8E with SQL table D4. Fb.dbv and .0 reserved All Field Descriptor (see next paragraph) all Header Record Terminator all 0x0D Field descriptor array in the DBF header (32 bytes for each field): 108 Byte Size Contents Description Applies to 0 11 ASCI field name. D5 0x01 codepage437 DOS USA Fp 0x02 codepage850 DOS Multi ling Fp 30 2 32 n*32 +1 1 0x03 codepage1251 Windows ANSI Fp 0xC8 codepage1250 Windows EE Fp 0x00 ignored FS. D3. CL 15 1 0x01 Encrypted D4.dbf D4. Fp. 2nd record (see next part for size.dbt memo FS. Fp. 0x00 termin all 11 1 ASCI field type (see next paragraph) all 12 4 n. Fp.dbf D5. Fb..point all n = 1.3. Fp.0. D5 21 2 n. Fp 22 short int binary int max +/.0 reserved 31 1 0x01 Field is in . bin (see next paragraph) all \ FS. of float.. Fp.0 ignored FS.. bin all / both used for fld lng 18 2 0. D5 0x00 ignored FS..64kb (using deci count) FS n = 1. CL n = 1.. the start block posit. Fb. CL n = 1. Fp.20 FS.32kb (using deci count) Fp.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Byte Size Contents Description Applies to n. D5 (FS) Memo 10 digits repres.. CL 16 1 byte Field length.n Char ASCII (OEM code page chars) rest= space. Fb Logical ASCII chars (YyNnTtFf space) FS.. or 10 spaces if no entry in memo all V 10 Variable Variable.18 D3.0.dbv FS P 10 Picture binary data in . D4.. field descriptor (1 byte): Size Type Description/Storage Applies to C 1.0123456789) fix posit/no float.0. Fp N 1. Fp. CL 23 1 0x01 Set Fields D3.2147483647 FS 88 double binary signed double IEEE FS L1 M 10 Each DBF record (fixed length): 109 ..CL: for C field type 17 1 byte decimal count. D5. D4. in . Fb. D5 0x00 unused FS.point n = 1. Fp...n 0....9) in the YYYYMMDD format all F 1.0 ignored FS. D4.dbt file.20 FS. CL ASCII chars (YyNnTtFf?) D4...dbt structure like M D5 G 10 General OLE objects structure like M D5.254 all D8 Date 8 ASCII digits (0.dbv 4bytes bin= start pos in memo 4bytes bin= block size 1byte = subtype 1byte = reserved (0x1a) 10 spaces if no entry in .Atoll 3. Fb.. D5 0x00 ignored FS..n Numeric ASCII digits (-.n Numeric ASCII digits (-. bin/asc data in . D4. D5.0123456789) variable pos. D4. D5. CL Field type and size in the DBF header. not \0 term.n.0 offset from record begin Fp 0.ftp structure like M Fp B 10 Binary binary data in .32767 FS 44 long int binary int max +/.0 reserved all 20 1 byte Work area ID D4.. Fb.. D3. CL multi-user dBase D3. Fp.. CL all 24 7 0. D3. D3.mdx index D4.. Fb. all n = 1. This information is used. the associated model ID changes.16. 0: antenna losses not taken into account 1: antenna losses included INC_ANT 110 .2 Pathloss.los file MODEL_NAME Text Name of propagation model used to calculate path loss MODEL_SIG Text Signature (identity number) of model used in calculations. when no calculation radius is set. ULXMAP Float X-coordinate of the top-left corner of the path loss matrix upper-left pixel ULYMAP Float Y-coordinate of the top-left corner of the path loss matrix upper-left pixel RESOLUTION Float Resolution of path loss matrix in metre NROWS Float Number of rows in path loss matrix NCOLS Float Number of columns in path loss matrix FREQUENCY Float Frequency band TILT Float Transmitter antenna mechanical tilt AZIMUTH Float Transmitter antenna azimuth TX_HEIGHT Float Transmitter height in metre TX_POSX Float X-coordinate of the transmitter TX_POSY Float Y-coordinate of the transmitter ALTITUDE Float Ground height above sea level at the transmitter in metre RX_HEIGHT Float Receiver height in metre ANTENNA_SI Float Logical number referring to antenna pattern. The Model_SIG is used for the purpose of validity. Boolean Atoll indicates if losses due to the antenna pattern are taken into account in the path loss matrix. MAX_LOS Float Maximum path loss stated in 1/16 dB. You can check it in the propagation model properties (General tab). path) of . In the same way.dbf File Contents The DBF file provides information that is needed to check validity of each path loss matrix. This enables Atoll to detect path loss matrix invalidity. Field Type Description TX_NAME Text Name of the transmitter FILE_NAME Text Name (and optionally. two identical propagation models in different projects do not have the same model IDa. see (2) byte 10…11 All 2. A unique Model_SIG is assigned to each propagation model. unterminated. to check the matrix validity.1.Atoll 3. When model parameters are modified. For n. fixed length. Antennas with the same pattern will have the same number.3.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Byte Size Description Applies to 0 1 deleted flag "*" or not deleted " " All 1…n 1… x-times contents of fields. CAREA_XMIN Float Lowest x-coordinate of centre pixel located on the calculation radiusb CAREA_XMAX Float Highest x-coordinate of centre pixel located on the calculation radius CAREA_YMIN Float Lowest y-coordinate of centre pixel located on the calculation radius CAREA_YMAX Float Highest y-coordinate of centre pixel located on the calculation radius WAREA_XMIN Float Lowest x-coordinate of centre pixel located in the computation zonec WAREA_XMAX Float Highest x-coordinate of centre pixel located in the computation zone WAREA_YMIN Float Lowest y-coordinate of centre pixel located in the computation zone WAREA_YMAX Float Highest y-coordinate of centre pixel located in the computation zone LOCKED Boolean Locking status 0: path loss matrix is not locked 1: path loss matrix is locked. These coordinates enable Atoll to determine the rectangle including the computation zone. b. "No data" values are represented by +32767. 2. Header: • • • • • • • • • • • • • 4 bytes: version 4 bytes: flag (can be used to manage flags like active flag) 50 bytes: GUID 4 bytes: number of points 255 bytes: original measurement name (with prefix "Num" for drive test data and "CW" for CW measurements) 256 bytes: comments 4 bytes: X_RADIUS 4 bytes: Y_RADIUS 4 bytes: gain = measurement gain .pts file AREA_XMIN Float Not used AREA_XMAX Float Not used AREA_YMIN Float Not used AREA_YMAX Float Not used 2. Path losses are tuned by merging measurement data with propagation results on pixels corresponding to the measurement points and the pixels in the vicinity.1 Pathloss. For more information on the path loss tuning algorithm.16.dbf file and one PTS (path loss tuning) file per transmitter.dbf File Contents The DBF file provides information about the measured transmitters involved in the tuning.2 Path Loss Tuning File Format Atoll can tune path losses calculated by propagation models using CW measurements or drive test Data.16. users must retrieve the propagation models from the same central database.2 Pathloss. The file contains 16-bit signed integer values in the range [-32768.3. A tuning file can contain several measurement paths. Path losses surrounding the measurement points are smoothed for homogeneity. 2. c. Field Type Description TX_NAME Text Name of the transmitter FILE_NAME Text Name (and optionally. 2.3 PTS File Format The PTS (path loss tuning) files contain a header and the list of measurement points.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 a.16.16.3 LOS File Format The LOS (path loss results) files are binary files with a standard row-column structure.Atoll 3. see the Technical Reference Guide. Data are stored starting from the southwest to the northeast corner of the area.2.1. +32767] with a 1/16 dB precision.16.dbf File Format See "Pathloss. These coordinates enable Atoll to determine the area of calculation for each transmitter.losses 4 bytes: global error 4 bytes: rx height 4 bytes: frequency 8 bytes: tx Position List of measurement points: • • • • 4 bytes: X 4 bytes: Y 4 bytes: measurement value 4 bytes: incidence angle.dbf File Format" on page 107. In order to benefit from the calculation sharing feature. This can be done using the Open from database command for a new document or the Refresh command for an existing one.2. Atoll generates different model_ID (even if same parameters are applied on the same kind of model) and calculation sharing become unavailable due to inconsistency. Otherwise. 2. path) of .2. 111 . Measuremment paths that are used for path loss tuning are stored as a catalogue in a folder containing a pathloss. 5 dB accuracy and probability values are calculated and stored with an accuracy of 0. Column4 C/I threshold C/I value. 2. its value is identical to the one of the line above.1. If the column is null. there can be up to three additional values (this number depends on the probability variation between the fixed values). This column cannot be null. Between two fixed C/I value. 8. and the columns 1 and 2 of the CLC file must contain the names of the interfered and interfering transmitters instead of their identification numbers.002 for probabilities between 1 and 0. Commented lines start with #.3.05. The C/I values have 0. 1. Interference matrices can be imported and exported using the following formats: • • GSM: CLC. and 22 dB. and WiMAX AFPs (automatic frequency planning tools). if the power offset of a subcell is X dB. Atoll does not check their validity and imports interference probability values for loaded transmitters only. and not in precentage (between 0 and 100%). # Version 1. In GSM interference matrices: • • • The interferer TRX type is not specified and is always considered to be BCCH.2 CLC File Format The CLC file consists of two parts.16. 2. Atoll saves probabilities for several C/I values (6 to 24 values).05. This field must not be empty.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 2. If no power offset is defined on the interfered TRX type.16. IM2 LTE and WiMAX: IM2. TXT. When interference matrix files are imported. its value is identical to the one of the line above. IM0.1 CLC Format (One Value per Line) The CLC format uses wo ASCII text files: a CLC file and a DCT file. then its interference histogram will be shifted by X dB with respect to the BCCH interference histogram. it is possible to set "All".3. 112 . and 3 must be defined only in the first line of each histogram. they must have the following format: <Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline> The 5 tab-separated columns are defined in the table below: Column Name Description Column1 Interfered transmitter Identification number of the interfered transmitter. The columns 1.16. Atoll assumes that the transmitter identifiers are the transmitter names. Atoll looks for the associated DCT file in the same directory and uses it to decode transmitter identifiers. LTE. Interference matrices are imported by selecting the CLC file to import. IM1. If the column is null. Subcells have different powers defined as offsets with respect to the BCCH. 14. For each interfered subcell-interferer subcell pair. The lines after the header are considered as comments if they start with "#". If the column is empty. If no DCT file is available.3 Interference Matrix File Formats Interference matrices are used by GSM. The second part details interference histogram of each interfered subcell-interfering subcell pair.Atoll 3. In order to save storage. Tab separated format. BCCH. If not. It must start with and contain the following lines: # Calculation Results Data File. Column5 Probability C/I > Threshold Probability to have C/I the value specified in column 4 (C/I threshold).g. Column2 Interfering transmitter Identification number of the interferer transmitter. and with an accuracy of 0. lines starting with the "#" are considered as comments. 2. all subcells with no power offset are not duplicated (e. In the following format descriptions and samples.0001 for probabilities lower than 0.3. Column3 Interfered TRX type Interfered subcell. its value is identical to the one of the line above. TCH). including five fixed ones: –9. The first part is a header used for format identification. CSV Interference matrix files must contain interference probability values between 0 and 1. For subcells other than the BCCH. 752 22 0. The parameter settings of this header can be wrong if # the "export" is performed following an "import". It must start with and contain the following lines: # Calculation Results Dictionary File. # # Service Zone Type is "Best signal level of the highest priority HCS layer".904 13 0. 2 BCCH.944 10 0.1. # Cell edge coverage probability 75%. they must have the following format: <Column1><tab><Column2><newline> 113 .316 25 0.772 If the TCH and BCCH histograms are the same.844 15 0.812 17 0. The second part provides information about transmitters taken into account in AFP. The first part is a header used for format identification.3 DCT File Format The . # Traffic spreading was Uniform ##---------------------------------------------------------------------# 1 2 TCH_INNER 8 9 1 a.16. # Version 1. They # are correct when the "export" follows a "calculate".1. 2.944 10 .3. The lines after the header are considered as comments if they start with "#".Atoll 3. Tab separated format. # Remark: Tab separated format.3. # Version 1.dct file is divided into two parts. they are not repeated. # Margin is 5. A single record indicates that the histograms belong to TCH and BCCH both. # Fields are: ##------------#------------#------------#-----------#------------------# #| Interfered | Interfering| Interfered | C/I #| Transmitter| Transmitter| Trx type | Probability | | Threshold | C/I >= Threshold | ##------------#------------#------------#-----------#------------------# # # Warning.904 11 0. C/I results do not incorporate power offset values.84 17 0.892 14 0.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Sample # Calculation Results Data File. If not. Commented lines start with #. Commented lines start with #.832 16 0.TCHa 1 0.872 14 0.292 8 1 9 0. # Version 1.im0 file consists of two parts. One transmitter per line is described separated with a tab character. they must have the following format: <Column1><tab><Column2><tab><Column3><tab><Column4><newline> The 4 tab-separated columns are defined in the table below: 114 . Tab separated format. Tab separated format. The parameter settings of this header can be wrong if # the "export" is performed following an "import".3. If not.1. Commented lines start with #. # Traffic spreading was Uniform (percentage of interfered area) ##---------------------------# Site0_0 1 -1 -1 100 100 Site0_1 2 -1 -1 100 100 Site0_2 3 -1 -1 100 100 Site1_0 4 -1 -1 100 100 Site1_1 5 -1 -1 100 100 Site1_2 6 -1 -1 100 100 Site2_0 7 -1 -1 100 100 Site2_1 8 -1 -1 100 100 2.4 IM0 Format (One Histogram per Line) This file contains one histogram per line for each interfered/interfering subcell pair. They # are correct when the "export" follows a "calculate". Sample # Calculation Results Dictionary File. # # Service Zone Type is "Best signal level per HCS layer".16. It must start with and contain the following lines: # Calculation Results Data File.1.3. # Fields are: ##-----------#-----------#-----------#-----------#---------#---------# #|Transmitter|Transmitter|BCCH during|BSIC during|% of vic'|% of int'| #|Name |Identifier |calculation|calculation|coverage |coverage | ##-----------#-----------#-----------#-----------#---------#---------# # # Warning. The .0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Column Name Type Description Column1 Transmitter name Text Name of the transmitter Column2 Transmitter Identifier Integer Identification number of the transmitter Column3 BCCH during calculation Integer BCCH used in calculations Column4 BSIC during calculation Integer BSIC used in calculations Column5 % of vic’ coverage Float Percentage of overlap of the victim service area Column6 % of int’ coverage Float Percentage of overlap of the interferer service area The last four columns describe the interference matrix scope. The first part is a header used for format identification. The second part details interference histogram of each interfered subcell-interferer subcell pair. The histogram is a list of C/I values with associated probabilities. Commented lines start with #. # Cell edge coverage probability is 75%. # Margin is 5. The lines after the header are considered as comments if they start with "#". # Version 2.Atoll 3. Commented lines start with #.94 1 0. The second part details interference histogram of each interfered subcell-interferer subcell pair.644 2. Column2 Interfering transmitter Name of the interferer transmitter.964 -1 0.844 11 0.692 18 0.996 -6 0. Column3 Interfered TRX type Interfered subcell.804 14 0.TCH-10 1 -9 0. Commented lines start with #. In order to save storage.16. # Version 1.804 13 0. If not.928 4 0.86 9 0.932 1 0.932 4 0.3. they must have the following format: 115 .5 IM1 Format (One Value per Line.996 -6 0. # Margin is 5.98 -3 0.Atoll 3.976 -4 0.744 14 0.816 10 0.972 -1 0.94 1 0.848 9 0. # Version 1.948 0 0. # # Service Zone Type is "Best signal level of the highest priority HCS layer".636 15 0.712 17 0.824 11 0.996 -6 0.im1 file consists of two parts.TCH-9 1 -6 0.856 8 0.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Column Name Description Column1 Interfered transmitter Name of the interfered transmitter.964 -2 0. The parameter settings of this header can be wrong if # the "export" is performed following an "import". The first part is a header used for format identification. The histogram is a list of C/I values with associated probabilities.1.852 8 0.896 7 0.832 10 0. all subcells with no power offset are not duplicated (e.66 Site0_2 Site0_3 BCCH.948 0 0.564 Site0_3 Site0_2 BCCH.932 1 0. TCH). They # are correct when the "export" follows a "calculate". The .1. # Fields are: #-----------------------------------------------------------------------#Transmitter Interferer TRX type {C/I Probability} values #-----------------------------------------------------------------------# # Warning.716 15 0. Tab separated format. Tab separated format.TCH-10 1 -9 0.864 8 0. Sample # Calculation Results Data File.876 8 0. This entry cannot be null.764 14 0.644 15 0.84 11 0.976 -4 0.TCH-10 1 -9 0.608 18 0.924 4 0. # Cell edge coverage probability 75%.g.972 -3 0.772 13 0.784 11 0.556 Site0_3 Site0_1 BCCH.688 14 0.892 7 0.924 4 0. BCCH.832 9 0. Column4 C/I probability C/I value and the probability associated to this value separated by a space character.936 0 0.616 18 0. # Remark: C/I results do not incorporate power offset values.896 7 0.3.904 7 0. # Traffic spreading was Uniform ##---------------------------------------------------------------------# # Site0_2 Site0_1 BCCH.96 0 0. It must start with and contain the following lines: # Calculation Results Data File. TX Name Repeated) This file contains one C/I threshold and probability pair value per line for each interfered/interfering subcell pair. The lines after the header are considered as comments if they start with "#". Column3 Interfered TRX type Interfered subcell. Commented lines start with #. They # are correct when the "export" follows a "calculate".TCH -1 0. # # Service Zone Type is "Best signal level of the highest priority HCS layer".896 Site0_2 Site0_1 BCCH.3.TCH 10 0.964 Site0_2 Site0_1 BCCH.TCH -10 Site0_2 Site0_1 BCCH. In GSM.TCH 1 0.16.. # Version 1. Tab separated format.1. # Traffic spreading was Uniform ##---------------------------------------------------------------------# Site0_2 Site0_1 BCCH.932 Site0_2 Site0_1 BCCH.TCH 8 0. TCH).996 1 Site0_2 Site0_1 BCCH. This field must not be empty.TCH -6 0. Each line must have the following format: <Column1><SEP><Column2><SEP><Column3><SEP><Column4><newline> 116 .TCH -9 0.TCH 4 0. 2.6 IM2 Format (Co.936 Site0_2 Site0_1 BCCH.3.824 .g. The parameter settings of this header can be wrong if # the "export" is performed following an "import".976 Site0_2 Site0_1 BCCH. BCCH. This column cannot be null. there is only one set of values for all the subcells of the interfered transmitter.TCH 7 0. # Fields are: #-----------------------------------------------------------------------#Transmitter Interferer TRX type C/I Probability #-----------------------------------------------------------------------# # Warning.TCH 9 0.TCH -4 0. # Cell edge coverage probability 75%.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 <Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline> The 5 tab-separated columns are defined in the table below: Column Name Description Column1 Interfered transmitter Name of the interfered transmitter.TCH 0 0. Sample # Calculation Results Data File.848 Site0_2 Site0_1 BCCH..Atoll 3.and Adjacent-channel Probabilities) IM2 files contain co-channel and adjacent-channel interference probabilities for each interfered transmitter – interfering transmitter pair.864 Site0_2 Site0_1 BCCH.924 Site0_2 Site0_1 BCCH. Column5 Probability C/I > Threshold Probability to have C/I the value specified in column 4 (C/I threshold). Column4 C/I threshold C/I value.832 Site0_2 Site0_1 BCCH. In order to save storage. all subcells with no power offset are not duplicated (e. Column2 Interfering transmitter Name of the interferer transmitter. # Margin is 5. # Remark: C/I results do not incorporate power offset values. 0. the results are stored in two files for the entire prediction: one HDR file and one BIL file (both identified by the prediction name).028 The columns in the sample above are separated with a tab.226.024 Site0_3 Site0_1 0..0.0. • When a calculation is "per transmitter".0. They # are correct when the "export" follows a "calculate". when a Cell Identifier Collision Zones (DL) prediction is calculated by value intervals with the display type set "No.276.Site0_3. 117 . an additional DBF file is created for the entire prediction (identified by the prediction name). The calculation of the coverage prediction is either global or "per transmitter".0.3. The DBF file contains information on each transmitter and a pointer to each transmitter’s specific HDR and BIL files.024 Site0_3.e.276 0.16.226667.1.0. Commented lines start with #. # Fields are: #-----------------------------------------------------------------------#Transmitter Interferer Co-channel Adjacent channel #-----------------------------------------------------------------------# # Warning. # Margin is 5. of interferers per cell". # Version 1. outside the ATL file.024 Site0_2. i.02 Site0_3.0. A corresponding ’<doc_name>\{<GUID>}’ folder is actually created where the ATL document is located.226 0. # # Service Zone Type is "Best signal level of the highest priority HCS layer". # Remark: C/I results do not incorporate power offset values.27. In LTE.Site0_2.028 2. one HDR file and one BIL file are created for each transmitter in the prediction (both identified by the transmitter’s name). as soon as the latter is saved. These columns can also be separated with a semilcolon: Site0_2.226667 0. The four columns are defined in the table below: Column Name Description Column1 Interfered transmitter Name of the interfered transmitter Column2 Interfering transmitter Name of the interferer transmitter Column3 Co-channel probability Co-channel interference probability Column4 Adjacent-channel probability Adjacent channel interference probability Sample # Calculation Results Data File. Tab separated format. # Cell edge coverage probability 75%.0.27 0. The parameter settings of this header can be wrong if # the "export" is performed following an "import".Site0_1.02 Site0_3 Site0_2 0.Site0_1.Atoll 3. In some "per transmitter" predictions. the HDR file and the BIL file are created for each cell in the prediction (both identified by the cell’s name).0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Where the separator (<SEP>) can either be a tab or a semicolon.4 "Per Transmitter" Prediction File Format When a coverage prediction is calculated by value intervals it is stored externally. • When the calculation is global. # Traffic spreading was Uniform ##---------------------------------------------------------------------# Site0_2 Site0_1 0.024 Site0_2 Site0_3 0. dbf’ files is identical to the format described in "Pathloss.e.4.16. Here.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 The format and the content of the DBF file is described here.16.2 <per_transmitter_prediction>.1 <per_transmitter_prediction>.16. same as ’xdim’ and ’ydim’ in the HDR file AREA_XMIN Float Same as ’ulxmap’ in the HDR file AREA_XMAX Float Same as ’ulxmap’ + ’xdim’ * ’ncols’ in the HDR file AREA_YMIN Float Same as ’ulymap’ in the HDR file AREA_YMAX Float Same as ’ulymap’ + ’ydim’ * ’nrows’ in the HDR file NBITS Float Same as ’nbits’ in the HDR file NBANDS Float Same as ’nbands’ in the HDR file BYTE_ORDER Float Same as ’byteorder’ in the HDR file BAND_ROW_BYTES Float Same as ’bandrowbytes’ in the HDR file TOTAL_ROW_BYTES Float Same as ’totalrowbytes’ in the HDR file SKIP_BYTES Float Same as ’skipbytes’ in the HDR file DATA_TYPE Text Same as ’datatype’ in the HDR file NO_DATA_VALUE Same as ’nodatavalue’ in the HDR file 2.5. surrounding pixels.5 Coverage Prediction Export and Reports 2. i. 118 . Field Type Description TX_NAME Text Name of the transmitter FILE_NAME Text Name of the transmitter’s BIL result file RESOLUTION Float Resolution of the calculation. the pixel will be coloured by the most representative value within this bounding box. In both cases.. Predictions are filtered by setting the colour of a pixel to the dominant colour of the bounding box. In other words.1 Filtering Coverage Predictions at Export Raster and vector coverage predictions can be filtered at export in order to exclude holes and islands.dbf File Format" on page 107.4. using a dispersion factor: 2 exp  – D   X  2   . an XML file describing the prediction is also created in the corresponding ’<doc_name>\{<GUID>}’ folder. 2.dbf File Format The format of ‘<per_transmitter_prediction>.dbf File Contents The ‘<per_transmitter_prediction>.16.3. D is the distance from the pixel to be coloured to each pixel within the bounding box and X is the value at that pixel.dbf’ files generated in specific ’{<GUID>}’ folders provide information that is needed to check the validity of each "per transmitter" prediction> calculated by value intervals.Atoll 3. 2. 3. Smoothing by Percentage 2 Z The user-defined smoothing percentage Z gives the approximation tolerance: ------.28: Smoothing Tolerance Definition For example.29 on page 119. Tolerance is the interval within which Atoll tries to reduce the number of points.16.2 Smoothing Coverage Predictions at Export Vector coverage predictions can be smoothed at export in orer to simplify its contours.. R  -----. A1. the bounding box is increased by one pixel every 10 % (since Y is defined as a percentage). A2 will be deleted if within this tolerance (and A1 and A3 will be directly linked) and A2 will be conserved if outside this tolerance. A2 outside the tolerance interval A2 inside the tolerance interval Figure 2.5. 2.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Figure 2.Atoll 3. Figure 2. for three successive points. where R is the user-defined 2 20 export resolution.27: Bounding box for prediction filtering The user-defined filtering percentage Y gives the size of the bounding box: Y  10 pixels in each direction. Predictions are smoothed by reducing the number of points defining the contours of the polygons using a vertex reduction routine that successively reduces the number of closely clustered vertices (vertex reduction within tolerance of prior vertex cluster. A2. and A3 as shown in Figure 2. Two smoothing methods exist for defining the degree of coverage smoothing: smoothing by percentage and smoothing by the maximum number of points. Douglas-Peucher polyline simplification). In other words.29: Smoothing by Percentage 119 . 32 on page 121.16. This number of points helps the algorithm to determine the optimised tolerance (see "Smoothing by Percentage" on page 119) such that. If this number is greater than the maximum number of points defined by the user. Starting from the maximum possible tolerance.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation ©Forsk 2015 Smoothing by Number of Points The second method consists in defining a maximum number of points to be deleted. The first the number of points respecting the constraint is obtained.3. 1 2 3 4 Figure 2.31 on page 121 shows the original signal level coverage prediction whose filtered and smoothed exported results are presented in Figure 2.5.30: Smoothing by Number of Points 2. the number of points to be filtered out are estimated (circled in red in the following example ( 2 )).3 Examples of Prediction Export Filtering and Smoothing Figure 2. the number of points to be deleted will be lower than this value.Atoll 3. smoothing is applied by deleting these points and linking the remaining closest points ( 4 ). with this obtained tolerance. 120 . Let’s consider the following example ( 1 ). Atoll reduces the tolerance until reaching the requested maximum number of points or less ( 3 ). or the covered area. if neither zone exists.32: Exported prediction with filtering and smoothing 2.4 Coverage Prediction Reports Over Focus/Computation Zones Statistics are calculated in coverage prediction reports over the focus zone or the computation zone.31: Bounding box for prediction filtering Filtering Percentage: 0 % Smoothing Percentage: 0 % Filtering Percentage: 0 % Smoothing Percentage: 100 % Filtering Percentage: 100 % Smoothing Percentage: 0 % Filtering Percentage: 100 % Smoothing Percentage: 100 % Figure 2. the pixel is considered outside the zone. a pixel is considered inside the zone if its centre is inside. calculated from resolution of the coverage prediction. Otherwise.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation AT330_TRR_E1 Figure 2. This estimation may give rise to inaccuracies. 121 .3. If the reference surface area for the statistics is based on a focus or computation zone. The surface area of a coverage predictions is calculated by counting the number of covered pixels and multiplying this number with the area of one pixel.16. there may be minute inaccuracies in the calculated statistics because of the difference in the surface area calculation methods: • • The surface areas of the zones (polygons) are calculated by triangulation.5. if no focus zone exists.Atoll 3. At the border of the focus or computation zone. 3.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 2: Radio Propagation 122 ©Forsk 2015 . Chapter 3 GSM GPRS EDGE Networks This chapter covers the following topics: • "Signal Level Calculation" on page 125 • "Interference-based DL Calculations" on page 131 • "GPRS/EDGE Calculations" on page 136 • "Codec Mode Selection and CQI Calculations" on page 145 • "UL Coverage Predictions" on page 152 • "Traffic Analysis" on page 159 • "Network Dimensioning" on page 171 • "Key Performance Indicators Calculation" on page 181 • "Simulations" on page 184 • "Automatic Neighbour Allocation" on page 191 • "AFP Appendices" on page 196 . 0 Technical Reference Guidefor Radio Networks © Forsk 2015 124 .Atoll 3.3. • • All the calculations are performed on TBC (to be calculated) transmitters.3.1. network dimensioning. 3. • L ant Tx is the transmitter antenna attenuation (from antenna patterns). and KPI calculation processes.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3 GSM GPRS EDGE Networks This chapter describes all the calculations performed in Atoll GSM/GPRS/EDGE documents. • max P Term is the maximum terminal allowed power.2 UL Signal Level Two parameters can be studied in UL signal level-based coverage predictions: Studied Parameter Formulas Signal level received from a terminal at a transmitter Term UL Signal level ( P rec ) Term UL Total losses ( L total ) Term P rec = Term max P Term Txi + G ant – L path – M Shadowing – model – L Indoor +  G ant Tx Term – L Tx – UL – L Term  Txi L total =  L path + M Shadowing – model + L Indoor + L Tx – UL + L Term  –  G ant + G ant Tx Term  Here.1 Signal Level Calculation 3. The first four sections describe the signal level.Atoll 3.1. and CQI calculations. • • EIRP is the effective isotropic radiated power of the transmitter. • L Term are the terminal losses. Term is the terminal antenna gain. 3. L model is the loss on the transmitter-terminal path (path loss) calculated by the propagation model. The following three sections explain the traffic analysis. This parameter is taken into account when the option “Shadowing taken • into account” is selected. tt is the TRX type (in the GSM GPRS EDGE. there are three possible TRX types. TCH and inner TCH). • L Tx – DL are the total transmitter DL losses. • G ant • • P is the power offset defined for the selected TRX type in the transmitter property dialogue. L Indoor are the indoor losses. • M Shadowing – model is the shadowing margin.1 DL Signal Level Three parameters can be studied in point analysis (Profile tab) and in signal level-based coverage predictions: Studied Parameter Formulas Signal level received from a transmitter on a TRX type DL Signal level Txi ( P rec ) Txi Txi P rec  tt  = EIRP  tt  – P  tt  – L path – M Shadowing – model – L Indoor +  G ant Txi – L Term  Txi Txi L path = L model + L ant Path loss ( L path ) DL Total losses ( L total ) Term Txi Tx Txi L total =  L path + M Shadowing – model + L Indoor + L Tx – DL + L Term  –  G ant + G ant Tx Term  Here. GPRS/EDGE-specific. BCCH. For the definition of TBC transmitters please refer to "Path Loss Matrices" on page 103.mdb document template. Logarithms used in this chapter (Log function) are base-10 unless stated otherwise. respectively. 125 . taken into account when the option “Indoor coverage” is selected. The last section describes the neighbour allocation process in GSM. interference. the path loss. The number of displayed bars depends on the signal level received from the best server. 3. No max range is set.1.4. L total – DL . identical in DL and in UL (see above).1. belongs to a Hierarchical Cell Structure (HCS) layer. In other words. • L Tx – UL are the total transmitter UL losses. Therefore. the received signal level from the selected transmitter will be the same for all TRX types. Coverage prediction parameters to be set are: • • The coverage conditions in order to determine the service area of each TBC transmitter. with a defined priority and a defined reception threshold. Atoll calculates the selected parameter on each pixel inside the Txi calculation area.4. Txi Txi For each transmitter. 3. the path loss.1.3. If the power reduction values defined for all the subcells are the same. • G ant • G ant Tx is the transmitter antenna gain.1. If the power reduction values defined for all the subcells are the same. each pixel inside the Txi calculation area is considered a probe (non-interfering) receiver.1. Bars are only displayed for transmitters whose signal level is within a 30 dB margin from the best server signal level.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Txi • L path is the path loss on the transmitter-terminal link. Txi or the DL total losses. • L Term are the terminal losses. L path .3.3 Point Analysis 3.1.2 Reception Tab Analysis provided in the Reception tab is based on path loss matrices.1. Atoll can display the signal level received from a TRX type ( P rec  tt  ). This parameter is taken into account when the option “Shadowing taken • into account” is selected.1 DL Service Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine the areas coverage will be displayed. Txi. L total – DL . the received signal level from the selected transmitter will be the same for all TRX types. k. Txi. L path .1 Profile Tab Txi Txi For a selected transmitter. or the total Txi losses. 3. 3.3. We can distinguish eight cases as below. Term is the terminal antenna gain. All Servers The service area of Txi corresponds to the pixels : 126 . Reception level bars are displayed in the order of decreasing signal level. taken into account when the option “Indoor coverage” is selected.1 Each transmitter. Path loss and DL total losses are the same for all TRX types. for example a smaller value for improving the calculation speed. and The display settings to select the displayed parameter and its shading levels.4 Signal Level-based DL Coverage Predictions For each TBC transmitter. Let us assume that: • • 3. • M Shadowing – model is the shadowing margin. Path loss and total losses are the same for all TRX types. For more information on defining a different value for this margin.Atoll 3. see the Administrator Manual. it is possible to display the signal levels received from TBC transmitters for which path loss matrices have been calculated over their calculation areas. You can use a value other than 30 dB for the margin from the best server signal level. L Indoor are the indoor losses. it is possible to display the signal level received from a TRX type ( P rec  tt  ). Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 3rd best servers.1. nd Txi Txj And P rec  tt   2 Best  P rec  tt   – M ji M is the specified margin (dB). Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin from the second highest. If M = -2 dB. If M = 2 dB.1. Atoll considers pixels the received signal level from Txi is the highest. • • • 3. If M = 2 dB. Atoll considers pixels the received signal level from Txi is the second highest.2 Best Signal Level and a Margin The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)).1. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. P rec  tt  can be replaced Txi Txi with L total – DL or L path .3 If M = 0 dB. If M = -2 dB. Best Signal Level per HCS Layer and a Margin For each HCS layer.4 If M = 0 dB. The 2nd Best function considers the second highest value from a list of values. k. the service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)).Atoll 3.4. P rec  tt  can be replaced Txi Txi with L total – DL or L path . Second Best Signal Level and a Margin The service area of Txi corresponds to the pixels: Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). P rec  tt  can be replaced Txi Txi with L total – DL or L path .0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). P rec  tt  can be replaced Txi Txi with L total – DL or L path . Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji 127 . Txi Txj And P rec  tt   Best  P rec  tt   – M ji M is the specified margin (dB).4. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. The Best function considers the highest value from a list of values. • • • 3.1.4.1.3. 3.1. 1. P rec  tt  can be replaced Txi Txi with L total – DL or L path . • • • 3. the service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). • • • 3.6 If M = 0 dB. HCS Servers and a Margin The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)).1. Highest Priority HCS Server and a Margin The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)).5 If M = 0 dB. Atoll considers pixels the received signal level from Txi is the highest.4. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. Second Best Signal Level per HCS Layer and a Margin For each HCS layer.1.1.Atoll 3. P rec  tt  can be replaced Txi Txi with L total – DL or L path . Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. The Best function considers the highest value from a list of values. 128 . Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. If M = 2 dB. k. If M = -2 dB. nd Txi Txj And P rec  BCCH   2 Best  P rec  BCCH   – M ji M is the specified margin (dB). Atoll considers pixels the received signal level from Txi is the second highest. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. • • • 3. If M = -2 dB.1.4. If M = 2 dB.3. Atoll considers pixels the received signal level from Txi is the highest. Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin from the second highest. If M = -2 dB. The Best function considers the highest value from a list of values. The 2nd Best function considers the second highest value from a list of values. If M = 2 dB. P rec  tt  can be replaced Txi Txi with L total – DL or L path .1.7 If M = 0 dB. M is the specified margin (dB). Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji Txi And the received P rec  tt  exceeds the reception threshold defined per HCS layer.4.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 M is the specified margin (dB). Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 3rd best servers. Txi And the received P rec  tt  exceeds the reception threshold defined per HCS layer.3.4. the transmitter with the highest C2 value is kept. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers.2 Display Types It is possible to display the coverage predictions with colours depending on any transmitter attribute or other criteria such as: 129 .1. If M = 2 dB. M is the specified margin (dB). Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. And C2 Txi Txj  BCCH  = Best  C2  BCCH   j The Best function considers the highest value from a list of values.1.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji And Txi belongs to the HCS layer with the highest priority.4. The Best function considers the highest value from a list of values.4. P rec  tt  can be replaced Txi Txi with L total – DL or L path . 3. If M = -2 dB.2. Best Idle Mode Reselection Criterion (C2) Such type of coverage is useful : • • To compare idle and dedicated mode best servers for voice traffic Display the GPRS/EDGE best server map (based on GSM idle mode) The path loss criterion C1 used for cell selection and reselection is defined by: Txi C1 = P rec  BCCH  – MinimumThreshold  BCCH  The path loss criterion (GSM03.1 Coverage Resolution The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction.22) is satisfied if C1  0 . Atoll considers pixels the received signal level from Txi is the highest. The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  BCCH   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). the traffic is served by the transmitter for which the difference between the received signal strength and the HCS threshold is the highest. therefore. The way the competition is managed between layers with the same priority can be modified.Atoll 3.2 Coverage Display 3.4. In the case two layers have the same priority. C2 is defined as an integer in the 3GPP specifications. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. 3.8 If M = 0 dB.1.1. • • • 3. see the Administrator Manual.1. see "Path Loss Calculation Prerequisites" on page 57 for more information).2. On each pixel. the C2 values in the above calculations are rounded down to the nearest integer. It corresponds to the best server in idle mode. The highest priority is defined by the priority field (0: lowest). For more information. The reselection criterion C2 is used for cell reselection only and is defined by: C2 = C1 + CELL_RESELECT_OFFSET CELL_RESELECT_OFFSET is the Cell Reselect Offset defined for the transmitter. Each layer shows the different total losses levels in the transmitter service area. the coverage corresponds to the pixels the signal level from this transmitter fulfils signal conditions defined in Conditions tab with different cell edge coverage probabilities. A pixel of a service area is coloured if the total losses is greater than or equal to the defined thresholds (pixel colour depends on total losses). Atoll determines the best transmitter and evaluates path loss from the best transmitter. A pixel of a service area is coloured if the C2 value is greater than or 130 . A pixel of a service area is coloured if total losses is greater than or equal to the defined minimum thresholds (pixel colour depends on total losses). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as transmitter service areas. Total Losses (dB) Atoll calculates DL total losses from the transmitter on each pixel of each transmitter service area. Path Loss (dB) Atoll calculates path loss from the transmitter on each pixel of each transmitter service area. Atoll determines the best transmitter and evaluates DL total losses from the best transmitter. Each layer corresponds to an area the path loss from the best server exceeds a defined minimum threshold. Best Signal Level (in dBm. Number of Servers Atoll evaluates how many service areas cover a pixel in order to determine the number of servers. Each layer corresponds to an area the number of servers is greater than or equal to a defined minimum threshold. Coverage consists of several independent layers whose visibility in the workspace can be managed. Each layer corresponds to an area the signal level from the best server exceeds a defined minimum threshold. dBµV/m) Atoll calculates signal level received from the transmitter on each pixel of each transmitter service area. Coverage consists of several independent layers whose visibility in the workspace can be managed. The pixel colour depends on the number of servers. service areas overlap the studied one.3. Coverage consists of several independent layers whose visibility in the workspace can be managed. Atoll chooses the highest value. Each layer shows the different path loss levels in the transmitter service area.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Signal Level (in dBm. Best Cell Edge Coverage Probability (%) On each pixel of each transmitter service area. Cell Edge Coverage Probability (%) On each pixel of each transmitter service area. the coverage corresponds to the pixels the best signal level received fulfils signal conditions defined in Conditions tab. dBµV. There are as many layers as defined thresholds. There is one coverage area per transmitter in the explorer. Best Server Path Loss (dB) Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Each layer corresponds to an area the total losses from the best server exceed a defined minimum threshold. Each layer shows the different signal levels available in the transmitter service area. There are as many layers as service areas. There are as many layers as service areas. There are as many layers as defined thresholds. When other service areas overlap the studied one. When other service areas overlap the studied one.Atoll 3. There are as many layers as defined thresholds. Coverage consists of several independent layers whose visibility in the workspace can be managed. When other service areas overlap the studied one. Coverage consists of several independent layers whose visibility in the workspace can be managed. Coverage consists of several independent layers whose visibility in the workspace can be managed. A pixel of a service area is coloured if the signal level is greater than or equal to the defined thresholds (the pixel colour depends on the signal level). A pixel of a service area is coloured if path loss is greater than or equal to the defined minimum thresholds (pixel colour depends on path loss). dBµV. Best C2 (dBm) Atoll calculates C2 values received from transmitters on each pixel of each transmitter service area. Best Server Total Losses (dB) Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Atoll chooses the highest value. dBµV/m) Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. There are as many layers as defined thresholds. There is one coverage area per cell edge coverage probability in the explorer. A pixel of a service area is coloured if the signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on signal level). A pixel of a service area is coloured if the path loss is greater than or equal to the defined thresholds (pixel colour depends on path loss). 3) Therefore. Examples: • • • Non-hopping (NH): An MSA is the channel assigned to a TRX used by a mobile.54. For non-hopping (NH) mode. are related to a transmitter.2) Synthesised frequency hopping (SFH): An MSA is the Mobile Allocation List (MAL) and the Mobile Allocation Index Offset (MAIO). the channel list corresponds to the mobile allocation list (MAL).54. MSAS(v) is the set of MSAs (Mobile Station Allocations) associated to v.56]. it is interesting to consider the Mobile Station Allocation.-) Baseband hopping (BBH): An MSA is the Mobile Allocation List (MAL) and the TRX index. Atoll calculates the DL C/I  ------------- Iv  m   transmitter v with MSA m (m  MSAS(v)).54.55]. TRX index Channel list MAIO MSA 1 53 54 55 56 2 ([53.-) 2 54 - (54. m. Therefore.2) 2 53 54 55 56 3 ([53. There are as many layers as defined thresholds.1 DL Carrier-to-Interference Ratio Calculation MSA (Mobile Station Allocation) Definition In order to understand the difference between each frequency hopping mode from the point of view of a mobile. In the following.0) 2 54 * ([53. 3. Notations and Assumptions In the following description: • • v is a victim transmitter. You may study a given TRX type tt (there will be as many MSA(v) as TRXs allocated to the subcell (v. BBH and SFH work in the same way. TRX index Channel list MAIO MSA 1 53 - (53.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 equal to the defined thresholds (the pixel colour depends on the C2 value). MSA is characterised by the pair (Channel list. Coverage consists of several independent layers whose visibility in the workspace can be managed. The number of MSAS(v) depends on TRX types to be analysed. For base-band hopping (BBH) or synthesized frequency hopping (SFH). we will use this notion to characterise the interference and resources set of a mobile. For BBH. Each layer corresponds to an area the best C2 value exceeds a defined minimum threshold. An MSA will be attached to each mobile considered during the simulation and the level of interference will be evaluated on this MSA.54. 131 .1) 3 55 * ([53. the channel list is 1 channel. channels of MAL belong to the same TRX type. MAIO).54.  C v  m  - for each victim Several MSAs. TRX index Channel list MAIO MSA 1 53 * ([53.56].2 Interference-based DL Calculations Interference-based calculations include all the calculations that involve the calculation of interference received from interfering transmitters in addition to the signal level received from the server.55].55.55]. from the point of view of a mobile station.55.tt)) or all the TRX types (the number of MSA(v) will correspond to the number of TRXs allocated to v).2.3.Atoll 3. 3. • C  m  is the carrier power level received from v on m. Each interference component is explained below. respectively. If the Detailed Results check box is selected. therefore. given by:  v I adj  m  =     i  INT  v  n  MSAS  i  i P rec  n   v i . The total noise is the sum of the thermal noise N thermal (-121 dBm by default or user-defined). T i  n  p m n  ---------------F  adj i Here.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015  C v  m  C - or Atoll considers the most interfered MSA. • • • i is any potential interfering transmitter (TBC transmitters whose calculation areas intersect the service area of v). Co. the C/I values for k  I v  m  + N Term  I + N Term tot  tot v all MSAs are displayed. P rec  n  is the carrier power level received from i on n. The interference levels are not changed. INT(v) is the set of transmitters that interfere v. Interference may also be received from the transmitters of another technology. • I  m  corresponds to the interference received from interfering transmitters i on m. v v v DL i v Therefore. 132 . MSAS(i) is the set of MSAs related to potential interferers i. It can be set to 100% in the coverage prediction properties. and the inter- . the terminal noise figure NF v DL NR inter – techno log y technology downlink noise rise Term N tot = N thermal + NF Term Term . Ti(n) is occupancy of the MSA n: i i T i  n  = L traffic  n   f act  n  i L traffic  n  is the traffic load defined for the MSA n or i.Atoll 3. I m = I co m + I adj m + I inter – techno log y – G PC – G Div i v G PC is the average power control gain defined for the interfering transmitter i and G Div is the diversity gain defined for the considered subcell. the displayed C/I or C/(I+N) are  --- = Min  ------------- Iv k  Iv  m   v   C - C  m  - --------------------------------= Min  --------------------   . • M Shadowing used in the C/I calculation is based on the C/I standard deviation. Atoll takes the total noise N tot into account. v DL + NR inter – techno log y Interference can be received from interfering transmitters i on co-channel and adjacent channels. v v The C/I shadowing margin is applied on the carrier power level. v v C  m  = P rec  m  Term If the interference conditions are based on C/(I+N).3. given by:  v I co  m  =     i  INT  v  n  MSAS  i  v I adj  m   v i i p m n  P rec  n   T i  n   co is the interference received at v on m on adjacent channels.and Adjacent Channel Interference: v I co  m  is the interference received at v on m on co-channel. Calculations The carrier power level is the power received from the victim transmitter at the terminal. or adjacent channel collision between MSAs n and m. There can be two cases: i. Example: Schematic view of hopping sequences MSA m of v ([34 37 39]. and MAIO’. DTX and traffic loads do not impact the interference i i from BCCH. • Collision Probability for Non Hopping Mode: v i p m n = 1 • Collision Probability for BBH and SFH Modes: MSA m of v can be defined as the pair ([f1.or adjacent channels: v i v i Collision = OCCUR  f m f' n  such that f m – f' n = 0 or 1 The probability of collision is the ratio of the number of collisions to the number of occurences: n collision v i p m n = ---------------------n occurence The probibility of collision depends on the correlation between m and n.and  p m n  adj = 1 --3 3 ii. v i p m n is the probability of having a co. MSAs m and n are not correlated m and n do not have identical HSN and synchronisation. MAIO’=2) 38 36 34 Here. If the subcell (i.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 i f act  n  is the activity factor defined for the MSA n of i.f’2. depending on the used frequency hopping mode.f2. The number of occurrences depends on the MAL length. the value specified in the coverage prediction properties is used. Otherwise. v i An occurence OCCUR  f m f' n  refers to the event when a channel f of m encounters a channel f’ of n during hopping. MSAs m and n are correlated m and n must have identical HSN and synchronisation. the number of co-channel collisions is 1. MAIO’=2) 38 36 34 Here. and the number of adjacent channel collisions is 1. Example: Schematic view of hopping sequences MSA m of v ([34 37 39]. MAIO=0) 34 37 39 MSA n of i ([38 36 34]. the number of occurrences is 3.Atoll 3. Therefore. f act  n  = 1 and L traffic  n  = 1 for the BCCH TRXs of the interferers. MAIO) and MSA n of i as the pair ([f’1. MAIO’) ( f and f’ are channels). MAIO. the activity factor is 1. and the number of adjacent channel collisions is 3. v i v i  p m n  co = 1 --.…. The probability of collision is the same for all the channels.3. the number of co-channel collisions is 1. Therefore. Therefore. In other words.f’n]. BCCH TRXs are always on.fn]. MAIO=0) 34 37 39 MSA n of i ([38 36 34].tt) supports DTX.…. A collision occurs when f and f’ are co. 133 . the number of occurrences is 9. 2 Point Analysis Analysis provided in the Interference tab is based on path loss matrices. Two types of diversity modes can be defined. and ICP ic  f is i the inter-technology channel protection between the frequencies used by the transmitter Tx and the victim transmitter v. the signal is transmitted as many times that there are antennas. adjacent channel.and  p m n  adj = 1 --9 3 Diversity gain: v G Div is the diversity gain defined for the victim subcell. Therefore.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 v i v i  p m n  co = 1 --. Therefore. In the ICP. the signal is successively transmitted on the various antennas. • Neither DTX nor traffic load of TRXs are taken into account to evaluate interference i i levels. Interferers are sorted in the order of descending carrier power levels.3. L total are the total losses between the transmitter Tx and the terminal. In. • • In case of frequency hopping.2. The interference from a transmitter Tx in a linked Atoll document is given as: Tx DL I inter – techno log y = P Transmitted  ic i   ------------------------------------Tx Tx L  ICP ni ic i is the i th ic i f total Tx frequency used by the transmitter Tx within its list of frequencies. the frequency gap is based on the defined base frequency for each technology (e. please refer to Signal to noise calculation: noise calculation part). The interference levels are not changed.. it is possible to display the interference levels received from TBC transmitters for which path loss matrices have been calculated over their calculation areas.Atoll 3. For Tx Diversity mode. In Tx Diversity. 935 MHz in GSM 900) 3.g. Co-channel. the diversity gain is defined as: v Tx_Div G Div = 3dB + Gclutter Tx_Div G clutter is the additional transmit diversity gain defined for the clutter class on which is located m.and adjacent channel interference received from interfering transmitters i on MAS m (for further information about noise calculation. the diversity gain is defined as: v Ant_Div G Div = G clutter Ant_Div G clutter is the antenna hopping gain defined for the clutter class on which is located m. P Transmitted  ic i  is the total Tx Tx transmitted Tx power on ic i . For Antenna Hopping mode. Inter-technology Downlink Interference: DL I inter – techno log y is the total inter-technology interference level on m due to transmitters in a linked Atoll document. we have T i  n  = L traffic  n   f act  n  = 1 . or both co. Atoll displays the following at the terminal: • • The carrier power level received from the victim transmitter v on the most interfered MAS m. the ICP value is weighted according to the fractional load. . • 134 The C/I shadowing margin is applied on the carrier power level. For each TBC transmitter. the coverage area of a transmitter Txi corresponds to the pixels : C Minimum threshold   --- I v TRX j C  Maximum threshold or Minimum threshold   ----------- I+N v  Maximum threshold TRX j . Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. This value can be modified by the user. determines the pixels where the calculated C/I exceeds the defined minimum threshold.1 Interference Condition Satisfied by At Least One TRX In this case. and colours these pixels depending on colour of the interfered transmitter. 3.3.2.3.2.1 Coverage Resolution The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction.2. and colours these pixels depending on C/I value.2.2. see "Path Loss Calculation Prerequisites" on page 57 for more information). Atoll calculates the C/I on each pixel within the service area of studied transmitters.2.3.2 Coverage Area Determination C C For each victim transmitter v. and The display settings to select the displayed parameter and its shading levels.Atoll 3.3. In other words.2 Display Types It is possible to display the coverage predictions with colours depending on any transmitter attribute or other criteria such as: 135 . TRXj is the TRX (belonging to Txi) with the worst C/I or C/(I+N) at the pixel. 3. The thermal noise (N = -121 dBm. 3. Coverage prediction parameters to be set are: • • • The coverage conditions in order to determine the service area of each TBC transmitter.3.3. each pixel inside the Txi calculation area is considered a probe (non-interfering) receiver.2. determines the pixels where the calculated C/I is lower than the defined maximum threshold.3.3 Interference-based DL Coverage Predictions Two interference-based DL coverage predictions are available: • Coverage by C/I Level (DL): Provides a global analysis of the network quality. Atoll calculates the C/I on each pixel within the service area of studied transmitters.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3.2 Interference Condition Satisfied by The Worst TRX In this case. The two options defining the thresholds are explained below. 3. by default) is used in the calculations if the coverage prediction is based on C/(I+N).3 Coverage Display 3.3.2.3.2.2.3. Txi. 3. coverage area corresponds to pixels where DL  --- or  ----------- is between the lower and upper I v I+N v thresholds defined in the coverage prediction properties. TRXj is any TRX belonging to Txi. The interference conditions to meet for a pixel to be covered. 3. the coverage area of a transmitter Txi corresponds to the pixels : C Minimum threshold   ---  Iv TRX j C  Maximum threshold or Minimum threshold   -----------  I + N v  Maximum threshold TRX j . See "DL Service Area Determination" on page 126 for more information. Atoll calculates the selected parameter on each pixel inside the Txi calculation area. Service areas are determined in the same manner as for signal level-based coverage predictions. • Interfered Zones: Shows the areas a transmitter is interfered.1 Service Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine the areas coverage will be displayed. Coding schemes may be selected using ideal link adaptation or without it. A pixel of a coverage area is coloured if the DL C/I (or C/(I+N)) level is greater than or equal to the specified thresholds (the pixel colour depends on the DL C/I (or C/(I+N)) level). Each layer shows the different DL C/I levels available in the transmitter coverage area. either the noise figure defined for the calculations or that of the selected terminal type is used. A pixel of a coverage area is coloured if the DL C/I (or C/(I+N)) level is greater than or equal to the specified thresholds (the pixel colour depends on the DL C/I (or C/(I+N)) level). There are as many layers as defined thresholds. . When the calculations are based on C/I and C/(I+N): • • 136 Atoll calculates the carrier-to-interference ratio for all the GPRS/EDGE TBC transmitters but takes into account all the TBC transmitters (GSM and GPRS/EDGE) to evaluate the interference. If no terminal type is defined for the calculation. or 32QAM modulations. C/I thresholds are also indexed by the C/(I+N) value. or on C/(I+N).0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 C/I Level Each pixel of the transmitter coverage area is coloured if the calculated DL C/I (or C/(I+N)) level is greater than or equal to the specified minimum thresholds (pixel colour depends on DL C/I (or C/(I+N)) level).Atoll 3. Max C/I Level Atoll compares calculated DL C/I (or C/(I+N)) levels received from transmitters on each pixel of each transmitter coverage area coverage areas overlap the studied one and chooses the highest value. Coverage consists of several independent layers whose visibility in the workspace can be managed. no coding scheme selection and throughput calculation is carried out. i. In this case. Coverage consists of several independent layers whose visibility in the workspace can be managed.e. Txi • P Backoff  TRX  is the Power Backoff defined for the subcell for 8PSK. Min C/I Level Atoll compares DL C/I (or C/(I+N)) levels received from transmitters on each pixel of each transmitter coverage area the coverage areas overlap the studied one and chooses the lowest value. with and without ideal link adaptations. Each layer corresponds to an area the highest received DL C/I level exceeds a defined minimum threshold.3. There are as many layers as transmitter coverage areas. There are as many layers as defined thresholds. The following sections describe the two categories of calculations. • • CS is the set of all available coding schemes. Once coding schemes have been selected. and gives priority to the thresholds defined in the terminal configuration. Coverage consists of several independent layers whose visibility in the workspace can be managed. Atoll only selects the coding schemes that are common in the two. on C/I. Different GPRS/EDGE configurations may be defined for transmitter and terminals. GPRS/EDGE calculations may be based on signal levels (C) alone. 3.  Reception Threshold  CS are the values of reception thresholds for the coding schemes available in the GPRS/EDGE configuration. The reception thresholds given for signal level C are internally converted to C/N thresholds ( N is the thermal noise defined in the document database at -121 dBm by default) in order to be indexed by C/(I+N) values..3 GPRS/EDGE Calculations GPRS/EDGE calculations include coding scheme selection and throughput calculation. Each layer corresponds to an area the lowest received DL C/I level exceeds a defined minimum threshold. The priorities of the coding scheme lists are as follows: DBS > DAS > MCS > CS. If the transmitter does not have any GPRS/EDGE configuration assigned to it. throughputs corresponding to these coding schemes are readily determined from the look-up tables. C  ----------  I + N Threshold CS are the values of C/(I+N) thresholds for the coding schemes available in the GPRS/EDGE configuration. or if the terminal type does not have any GPRS/EDGE configuration assigned to it. • • • C  -- I Threshold CS are the values of C/I thresholds for the coding schemes available in the GPRS/EDGE configuration. Atoll only uses the GPRS/EDGE configuration of the transmitter. For calculating the noise. Ideal link adaptation implies that the selected coding scheme corresponds to the highest available throughput under the given radio conditions. 16QAM. In the following calculations. we assume that: • Txi P rec  TRX  is the signal level received from the selected TRX type (tt) or on all the TRXs of Txi on each pixel of the Txi coverage area. corresponding to the signal level is determined from the TP = f(C) graph.3. Nevertheless. 3. and csC/I is the coding scheme determined from the C/I level. from among the coding schemes available in the GPRS/EDGE configuration. 3. Throughput Calculation Based on the Worst Case Between C and C/I For the coding scheme csC determined above. such that: For each TRX type.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 The selection of coding schemes is mainly based on the radio conditions mentionned above.1. tt.2 Calculations Based on C/I Coding Scheme Selection Atoll selects two coding schemes from among the coding schemes available in the GPRS/EDGE configuration. you can model the gain due to longer MALs in coding scheme selection.1 Calculations Based on C Coding Scheme Selection Atoll selects a coding scheme. tt.Threshold I  CS I csC is the coding scheme determined from the signal level. you can optionally define some specific coding scheme graphs accoding to a specific hopping mode. is the coding scheme with the lower coding scheme number among csC and csC/I: cs = Min  cs C cs C  I  .3 Calculations Based on C/(I+N) Coding Scheme Selection Atoll selects two coding schemes from among the coding schemes available in the GPRS/EDGE configuration. mobility type. TPC. As an example. TPC/I. cs C = Lowest  CS  And. 3. see "DL Carrier-to-Interference Ratio Calculation" on page 131. such that: For each TRX type. cs.1. a throughput value.3. cs. a throughput value. The selected coding scheme. The resulting throughput TP is the lower of the two values.Atoll 3. cs. For more information on interference (I) calculation. corresponding to the C/I is determined from the TP = f(C/I) graph. Both coding schemes are the coding schemes with the lowest coding scheme number from the lowest priority coding scheme list. cs = Lowest  CS   Txi Txi  P rec  TRX  – P Backoff  TRX    Reception Threshold  CS The selected coding scheme. --. is the coding scheme with the lowest coding scheme number from the lowest priority coding scheme list.3. frequency band and MAL.3.3. TPC and TPC/I: TP = Min  TP C TP C  I  . such that: 137 . Atoll reads the corresponding throughput value for the received signal level from the Throughput=f(C) graph associated with cs.1. Throughput Calculation Once the coding scheme cs is selected.1 Coding Scheme Selection and Throughput Calculation Without Ideal Link Adaptation 3. For the coding scheme csC/I determined above. cs C  I   = Lowest  CS    Txi Txi  P rec  TRX  – P Backoff  TRX    Reception Threshold  CS    Txi Txi P  TRX  – P  TRX   rec Backoff C   ------------------------------------------------------------------------. pN is the thermal noise power (value in Watts). is the one with the lowest coding scheme number from the lowest priority coding scheme list. corresponding to the C/(I+N) is determined from the TP = f(C/N) graph..2 Calculations Based on C/I Throughput Calculation Based on Worst Case Between C and C/I For the received signal level. TPC/N.3. corresponding to the C/(I+N) is determined from the TP = f(C/(I+N)) graph.3. the TP = f(C) graph is internally converted to TP = f(C/N) graph. The selected coding scheme. and coding schemes whose reception thresholds are lower than the received signal level.2 Coding Scheme Selection and Throughput Calculation With Ideal Link Adaptation 3. TPC/(I+N). For the coding scheme csC/(I+N) determined above. 3.1 Calculations Based on C Throughput Calculation For the received signal level.3. cs. The final throughput is computed by interpolating between the throughput values obtained from these two graphs. Atoll determines the highest throughput from the TP=f  C  graphs available in the GPRS/EDGE configuration.3.2. tt. The throughput interpolation method consists in interpolating TPC/N and TPC/(I+N) according to the respective weights of I and N values. Both coding schemes are the coding schemes with the lowest coding scheme numbers from the lowest priority coding scheme list. cs C   I + N    = Lowest  CS   ©Forsk 2015    Txi Txi P  TRX  – P  TRX   rec Backoff C ------------------------------------------------------------------------.Atoll 3. If there are more than one coding schemes providing the highest throughput at the pixel. cs. cs. the TP = f(C/I) graph is internally converted to TP = f(C/(I+N)) graph. is the one corresponding to the highest throughput calculated above. 138 .2. cs C  N = Lowest  CS   And. is the coding scheme with the higher coding scheme number among csC/N and csC/(I+N): cs = Max  cs C  N cs C   I + N   . 3. Atoll determines the highest throughput from the TP=f  C  graphs available in the GPRS/EDGE configuration. and p(I+N) is the interferences + thermal noise power (value  = ------------------pI + N in Watts).Threshold  I + N  CS N    Txi Txi P rec  TRX  – P Backoff  TRX   C ------------------------------------------------------------------------.  --------- I + N Threshold CS I+N csC/N is the coding scheme determined from the C/N. and csC/(I+N) is the coding scheme determined from the C/(I+N) level. A throughput value. Txi Txi TP C = Highest  TP=f  C = P rec  TRX  – P Backoff  TRX     CS   Txi Txi  P rec  TRX  – P Backoff  TRX    Reception Threshold  CS Coding Scheme Selection The selected coding scheme.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks   For each TRX type. The resulting throughput TP is given by: TP =   TP C  N +  1 –    TP C   I + N  pN .  ----------. Throughput Calculation Based on Interpolation Between C/N and C/(I+N) For the coding scheme csC/N determined above. A throughput value. and coding schemes whose reception thresholds are lower than the received signal level. the selected coding scheme. cs.3. is the one corresponding to the lower of the two highest throughputs calculated above. TPC and TPC/I. and coding schemes whose C/ (I+N) thresholds are lower than the received C/(I+N).3. TP C   I + N   Txi Txi P rec  TRX  – P Backoff  TRX    C   = Highest  TP=f  ----------. and coding schemes whose C/I thresholds are lower than the received C/I.Atoll 3. cs.– TP TP Application = TP RLC  MAC  -------Offset 100 139 .3 Calculations Based on C/(I+N) Throughput Calculation Based on Interpolation Between C/N and C/(I+N) Atoll internally converts the TP = f(C) graphs into TP = f(C/N) graphs.= ----------------------------------------------------------------   CS I+N N      Txi Txi P rec  TRX  – P Backoff  TRX   C ------------------------------------------------------------------------.  ----------. Atoll determines the highest throughput from the TP = f(C/N) graphs available in the GPRS/EDGE configuration.  ----------. pN is the thermal noise power (value in Watts). is the one with the highest coding scheme number from the highest priority coding scheme list. If there are more than one coding schemes providing the highest throughputs at the pixel. is the one with the lowest coding scheme number from the lowest priority coding scheme list.= ----------------------------------------------------------------   CS I+N I+N      Txi Txi P rec  TRX  – P Backoff  TRX   C ------------------------------------------------------------------------. --.2. cs. If there are more than one coding schemes providing the highest throughputs at the pixel. Atoll determines the highest throughput from the TP=f  C  I  graphs available in the GPRS/EDGE configuration.Threshold I  CS I The resulting throughput TP is the lower of the two values. and p(I+N) is the interferences + thermal noise power (value pI + N in Watts). 3. Atoll determines the highest throughput from the TP = f(C/(I+N)) graphs available in the GPRS/EDGE configuration. Coding Scheme Selection The selected coding scheme.3. The resulting throughput TP is given by: TP =   TP C  N +  1 –    TP C   I + N  pN  = ------------------. TP = Min  TP C TP C  I  Coding Scheme Selection The selected coding scheme.Threshold  I + N  CS I+N Atoll internally converts the TP = f(C/I) graphs into TP = f(C/(I+N)) graphs. The throughput interpolation method consists in interpolating TPC/N and TPC/(I+N) according to the respective weights of I and N values. For the received C/(I+N).3 Application Throughput Calculation Application throughput is calculated from the effective RLC throughput as follows: SF. 3.Threshold  I + N  CS I+N The final throughput is computed by interpolating between the throughput values obtained from these two graphs. the selected coding scheme. and coding schemes whose C/(I+N) thresholds are lower than the received C/(I+N). is the one corresponding to the higher of the two highest throughputs calculated above. TP C  N  Txi Txi P rec  TRX  – P Backoff  TRX    C   = Highest  TP=f  --. For the received C/(I+N).0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Txi Txi TP C = Highest  TP=f  C = P rec  TRX  – P Backoff  TRX     CS  P Txi Txi   TRX  – P  TRX    Reception Threshold  rec Backoff CS For the received C/I. cs. TP C  I  Txi Txi P rec  TRX  – P Backoff  TRX      = Highest  TP=f  C  I = ----------------------------------------------------------------   CS I      Txi Txi P rec  TRX  – P Backoff  TRX   C   ------------------------------------------------------------------------. the selected coding scheme. 5. belongs to a Hierarchical Cell Structure (HCS) layer. In other words. • • • 140 If M = 0 dB.Atoll 3.1. Txi.1 Service Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine the areas coverage will be displayed. Atoll calculates the selected parameter on each pixel inside the Txi calculation area.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 TP RLC  MAC is the effective RLC throughput. Each transmitter.3. Coverage prediction parameters to be set are: • • • The coverage conditions in order to determine the service area of each TBC transmitter.5 GPRS/EDGE Coverage Predictions Two GPRS/EDGE coverage predictions are available: • Coverage by GPRS/EDGE Coding Scheme: Shows the areas various coding schemes are available. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest.4 BLER Calculation Block error rate is calculated as follows:  TP  --------------. with a defined priority and a defined reception threshold. k. Txi.3. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. All Servers The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  tt  3.3. Atoll considers pixels the received signal level from Txi is the highest. No max range is set.5.3. each pixel inside the Txi calculation area is considered a probe (non-interfering) receiver. • Packet Throughput and Quality Analysis: Shows the throughputs corresponding to the coding schemes available. If M = 2 dB.5. This value can be modified by the user.1.2 Best Signal Level and a Margin The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  tt  Txi Txj And P rec  tt   Best  P rec  tt   – M ji M is the specified margin (dB). and The display settings to select the displayed parameter and its shading levels. and TP Offset and SF are the throughput offset (kbps) and the throughput scaling factor (%) defined for the selected service. 3. 3. Txi. is GPRS/EDGE-capable. The thermal noise (N = -121 dBm. For each TBC transmitter. We can distinguish eight cases as below. 3. by default) is used in the calculations if the coverage prediction is based on C/(I+N). Let us assume that: • • • 3.If  TP  TP MAX  BLER =  TP MAX  0 If  TP  TP MAX   TP is the throughput per timeslot calculated for a pixel and TPMAX is the maximum throughput per timeslot read from the GPRS/EDGE configuration used for the calculations.3. The interference conditions to meet for a pixel to be covered. If M = -2 dB.3.1 Each transmitter. The Best function considers the highest value from a list of values. . 4 If M = 0 dB. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. k. • • • 3. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 3rd best servers. • • • 3. the service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  tt  nd Txi Txj And P rec  BCCH   2 Best  P rec  BCCH   – M ji M is the specified margin (dB).5. k.3.3.6 If M = 0 dB.5.5. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 3rd best servers. Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin from the second highest. If M = 2 dB.3. If M = -2 dB. • • • 3.3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3. HCS Servers and a Margin The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  tt  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji Txi And the received P rec  tt  exceeds the reception threshold defined per HCS layer. Best Signal Level per HCS Layer and a Margin For each HCS layer. M is the specified margin (dB). 141 .1. The Best function considers the highest value from a list of values. If M = -2 dB. Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin from the second highest. Atoll considers pixels the received signal level from Txi is the highest.3. If M = 2 dB. Atoll considers pixels the received signal level from Txi is the second highest.1. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. Atoll considers pixels the received signal level from Txi is the highest. If M = -2 dB. • • If M = 0 dB. Atoll considers pixels the received signal level from Txi is the second highest. The 2nd Best function considers the second highest value from a list of values.1. the service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  tt  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji M is the specified margin (dB).1.3 Second Best Signal Level and a Margin The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  tt  nd Txi Txj And P rec  tt   2 Best  P rec  tt   – M ji M is the specified margin (dB).Atoll 3. Second Best Signal Level per HCS Layer and a Margin For each HCS layer. If M = 2 dB.5 If M = 0 dB. The Best function considers the highest value from a list of values. If M = 2 dB.5. The 2nd Best function considers the second highest value from a list of values. 1 Coverage Resolution The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. see "Path Loss Calculation Prerequisites" on page 57 for more information).22) is satisfied if C1  0 . see the Administrator Manual. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers.5.Atoll 3. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers.3. The highest priority is defined by the priority field (0: lowest). Highest Priority HCS Server and a Margin The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  tt  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji And Txi belongs to the HCS layer with the highest priority.5. 142 . M is the specified margin (dB). On each pixel. Atoll considers pixels the received signal level from Txi is the highest.5. Txi And the received P rec  tt  exceeds the reception threshold defined per HCS layer.2 Coverage Display 3.7 ©Forsk 2015 If M = -2 dB. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. In the case two layers have the same priority.3. 3. The way the competition is managed between layers with the same priority can be modified.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks • 3. the C2 values in the above calculations are rounded down to the nearest integer.3. the traffic is served by the transmitter for which the difference between the received signal strength and the HCS threshold is the highest. The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  BCCH  And C2 Txi Txj  BCCH  = Best  C2  BCCH   j The Best function considers the highest value from a list of values. The reselection criterion C2 is used for cell reselection only and is defined by: C2 = C1 + CELL_RESELECT_OFFSET CELL_RESELECT_OFFSET is the Cell Reselect Offset defined for the transmitter. The Best function considers the highest value from a list of values. It corresponds to the best server in idle mode.1. Best Idle Mode Reselection Criterion (C2) Such type of coverage is useful: • • To compare idle and dedicated mode best servers for voice traffic Display the GPRS/EDGE best server map (based on GSM idle mode) The path loss criterion C1 used for cell selection and reselection is defined by: Txi C1 = P rec  BCCH  – MinimumThreshold  BCCH  The path loss criterion (GSM03. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes.3.2. therefore. For more information.1. • • • 3. If M = -2 dB.3. the transmitter with the highest C2 value is kept. C2 is defined as an integer in the 3GPP specifications. If M = 2 dB.8 If M = 0 dB.5. Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many layers as throughput display thresholds. Packet Throughput and Quality Analysis: Effective RLC Throughput (kbps) A pixel of the coverage area is coloured if the calculated effective RLC throughput from any transmitter covering that pixel exceeds the defined minimum threshold. Each layer shows the application throughput that a transmitter can provide on one timeslot. Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many layers as throughput display thresholds. Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many layers as transmitter coverage areas and throughput display thresholds. Each layer shows the coding schemes available in the transmitter coverage area.3. The pixel colour depends on the highest application throughput per timeslot. There are as many layers as transmitter coverage areas. The pixel colour depends on the application throughput per timeslot.2. There are as many layers as throughput display thresholds.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3. Packet Throughput and Quality Analysis: Max Effective RLC Throughput/Timeslot (kbps) A pixel of the coverage area is coloured if the calculated highest effective RLC throughput per timeslot from any transmitter covering that pixel exceeds the defined minimum threshold. Each layer shows the average effective RLC throughput that all the transmitters can provide on one timeslot. Packet Throughput and Quality Analysis: Application Throughput/Timeslot (kbps) A pixel of the coverage area is coloured if the calculated application throughput per timeslot from any transmitter covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the effective RLC for all the timeslots supported by the 143 . Only the pixels with a coding scheme assigned are coloured. There are as many layers as possible coding schemes. Each layer shows the best effective RLC throughput that any transmitter can provide on one timeslot.Atoll 3. Packet Throughput and Quality Analysis: Average Effective RLC Throughput/Timeslot (kbps) A pixel of the coverage area is coloured if the calculated average effective RLC throughput per timeslot from all the transmitters covering that pixel exceeds the defined minimum threshold. Each layer shows the areas a given coding scheme can be used. The pixel colour depends on the effective RLC throughput per timeslot.3. Coverage consists of several independent layers whose visibility in the map window can be managed. Coverage consists of several independent layers whose visibility in the map window can be managed. Packet Throughput and Quality Analysis: Average Application Throughput/Timeslot (kbps) A pixel of the coverage area is coloured if the calculated average application throughput per timeslot from all the transmitters covering that pixel exceeds the defined minimum threshold. Packet Throughput and Quality Analysis: Effective RLC Throughput/Timeslot (kbps) A pixel of the coverage area is coloured if the calculated effective RLC throughput per timeslot from any transmitter covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the assigned coding scheme. Coverage by GPRS/EDGE Coding Scheme: Max Coding Schemes On each pixel. The pixel colour depends on the average effective RLC throughput per timeslot. Coverage consists of several independent layers whose visibility in the map window can be managed.2 Display Types It is possible to display the coverage predictions with colours depending on criteria such as: Coverage by GPRS/EDGE Coding Scheme: Coding Schemes Only the pixels with a coding scheme assigned are coloured. Coverage consists of several independent layers whose visibility in the map window can be managed. Each layer shows the best application throughput that any transmitter can provide on one timeslot. There are as many layers as transmitter coverage areas and throughput display thresholds. Packet Throughput and Quality Analysis: Best Application Throughput/Timeslot (kbps) A pixel of the coverage area is coloured if the calculated highest application throughput per timeslot from any transmitter covering that pixel exceeds the defined minimum threshold. Coverage consists of several independent layers whose visibility in the map window can be managed. The pixel colour depends on the average application throughput per timeslot. The pixel colour depends on the highest effective RLC throughput per timeslot. Each layer shows the effective RLC throughput that a transmitter can provide on one timeslot. Atoll chooses the highest coding scheme available from the TRXs of different transmitters covering that pixel.5. There are as many layers as throughput display thresholds. Each layer shows the average application throughput that all the transmitters can provide on one timeslot. The pixel colour depends on the assigned coding scheme. Packet Throughput and Quality Analysis: Average Effective RLC Throughput (kbps) A pixel of the coverage area is coloured if the calculated average effective RLC throughput from all the transmitters covering that pixel exceeds the defined minimum threshold.Atoll 3. Packet Throughput and Quality Analysis: Max Application Throughput (kbps) A pixel of the coverage area is coloured if the calculated highest application throughput from any transmitter covering that pixel exceeds the defined minimum threshold.3. There are as many layers as transmitter coverage areas and throughput display thresholds. The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. The pixel colour depends on the average application throughput for all the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many layers as transmitter coverage areas and throughput display thresholds. Each layer shows the application throughput that a transmitter can provide on all available timeslots in the terminal. The pixel colour depends on the highest effective RLC throughput for all the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). Each layer shows the highest application throughput that any transmitter can provide on all available timeslots in the terminal. Coverage consists of several independent layers whose visibility in the map window can be managed. Packet Throughput and Quality Analysis: Average Application Throughput (kbps) A pixel of the coverage area is coloured if the calculated average application throughput from all the transmitters covering that pixel exceeds the defined minimum threshold. determined using the selected dimensioning model. Coverage consists of several independent layers whose visibility in the map window can be managed. 144 . The pixel colour depends on the throughput per user for all the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The pixel colour depends on the application throughput for all the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). Each layer shows the application throughput that a transmitter can provide on all available timeslots in the terminal. The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. The throughput per user is calculated by applying the throughput reduction factor. The pixel colour depends on the average effective RLC throughput for all the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). Coverage consists of several independent layers whose visibility in the map window can be managed. Packet Throughput and Quality Analysis: Max Effective RLC Throughput (kbps) A pixel of the coverage area is coloured if the calculated highest effective RLC throughput from any transmitter covering that pixel exceeds the defined minimum threshold. Packet Throughput and Quality Analysis: Throughput per User (kbps) A pixel of the coverage area is coloured if the calculated throughput per user from any transmitter covering that pixel exceeds the defined minimum threshold. Packet Throughput and Quality Analysis: Application Throughput (kbps) A pixel of the coverage area is coloured if the calculated application throughput from any transmitter covering that pixel exceeds the defined minimum threshold. The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. Each layer shows the throughput per user that a transmitter can provide on all available timeslots in the terminal. The pixel colour depends on the highest application throughput for all the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. Each layer shows the average application throughput that all the transmitters can provide on all available timeslots in the terminal. There are as many layers as transmitter coverage areas and throughput display thresholds. Coverage consists of several independent layers whose visibility in the map window can be managed. Coverage consists of several independent layers whose visibility in the map window can be managed. Each layer shows the highest effective RLC throughput that any transmitter can provide on all available timeslots in the terminal.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). There are as many layers as throughput display thresholds. There are as many layers as throughput display thresholds. to the application throughput. There are as many layers as throughput display thresholds. There are as many layers as throughput display thresholds. Each layer shows the average effective RLC throughput that all the transmitters can provide on all available timeslots in the terminal. Coverage consists of several independent layers whose visibility in the map window can be managed. The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. HR. The pixel colour depends on the average throughput per user for all the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). Packet Throughput and Quality Analysis: BLER (%) A pixel of the coverage area is coloured if the calculated BLER from any transmitter exceeds the defined minimum threshold. and codec mode adaptation thresholds (calculated from the FER vs. determined using the selected dimensioning model. The pixel colour depends on the BLER. There are as many layers as transmitter coverage areas and BLER display thresholds. Each layer shows the average throughput per user that all the transmitters can provide on all available timeslots in the terminal.4 Codec Mode Selection and CQI Calculations Atoll supports FR. There are as many layers as BLER display thresholds. The pixel colour depends on the highest throughput per user for all the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots).3. These frames are usually discarded. Each layer shows the highest throughput per user that any transmitter can provide on all available timeslots in the terminal. C/I graphs for all codec modes at 5 % FER). EFR. The throughput per user is calculated by applying the throughput reduction factor. to the application throughput. or the perceived end-user voice quality. such as frequency hopping. Atoll has the following circuit quality indicators included by default: • • • FER or Frame Erasure Rate: The number of frames in error divided by the total number of frames. The pixel colour depends on the BLER.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Packet Throughput and Quality Analysis: Max Throughput per User (kbps) A pixel of the coverage area is coloured if the calculated highest throughput per user from any transmitter covering that pixel exceeds the defined minimum threshold. will impact the correlation between BER and FER. A codec configuration contains codec mode adaptation thresholds and quality graphs for circuit quality indicators. Coverage consists of several independent layers whose visibility in the map window can be managed. The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. in which case this can be called the Frame Erasure Rate. 3. the more sensitive it becomes to frame erasures. 145 . MOS values can only be measured in a test laboratory environment. BER or Bit Error Rate: BER is a measurement of the raw bit error rate in reception before the decoding process begins. The number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. MOS or Mean Opinion Score: Voice quality can be quantified using mean opinion score (MOS). Coverage consists of several independent layers whose visibility in the map window can be managed. and MOS quality graphs with respect to the carrier to interference ratio. Packet Throughput and Quality Analysis: Max BLER (%) A pixel of the coverage area is coloured if the calculated highest BLER from all the transmitters exceeds the defined minimum threshold. The throughput per user is calculated by applying the throughput reduction factor. Each layer shows the BLERs that the covered pixels experience on one timeslot. and AMR codec modes. determined using the selected dimensioning model. Each layer shows the BLER that the covered pixels experience on one timeslot. Packet Throughput and Quality Analysis: Average Throughput per User (kbps) A pixel of the coverage area is coloured if the calculated average throughput per user from all the transmitters covering that pixel exceeds the defined minimum threshold. Different voice codecs have slightly different FER to MOS correlation since the smaller the voice codec bit rate is. BER. Coverage consists of several independent layers whose visibility in the map window can be managed. Any factor that impacts the decoding performance. MOS values range from 1 (bad) to 5 (excellent). There are as many layers as throughput display thresholds. The default codec configurations in Atoll include default FER. to the application throughput. Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many layers as throughput display thresholds. C/I Graphs Figure 3.Atoll 3.2: BER vs.1: FER vs. C/I Graphs Figure 3. C/I Graphs 146 .3: MOS vs.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Figure 3.3. Atoll selects the higher priority codec mode. Melero. J. J. which may be different from the CQI being calculated. the defined graphs are used as is. either the noise figure defined for the calculations or that of the selected terminal type is used. Otherwise. Halonen. but takes into account all the TBC transmitters (with and without codec configurations) to evaluate the interference. see "DL Carrier-to-Interference Ratio Calculation" on page 131. The following sections describe the two categories of calculations. 3. A simple mapping from C/I to FER and BER for a GSM type of air interface. Performance characterization of the Adaptive Multi-Rate (AMR) speech codec (Release 6) 3. we assume that: Txi • P rec  TRX  is the signal level received from the selected TRX type (tt) or on all the TRXs of Txi on each pixel of the Txi • • coverage area. For more information on interference (I) calculation.3. for the transmitter under study.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 The graphs are based on: [1] T. Mogensen.Atoll 3.1 Calculations Based on C/N Atoll selects the highest priority codec mode. the codec mode is selected based on the codec adaptation thresholds. and gives priority to the thresholds defined in the terminal configuration. with and without ideal link adaptations. If the transmitter does not have any codec configuration assigned to it. If the values are the same.0. [2] J. Different codec configurations may be defined for transmitter and terminals.1 Circuit Quality Indicator Calculations Circuit quality indicator calculations include codec mode selection and CQI calculation. Atoll will use the adaptation thresholds defined in the Adaptation Thresholds tab to determine the codec mode to be used in the studies. frequency band and MAL.e. Once codec modes have been selected. Romero. If more than one codec modes satisfy the C/N or C/I conditions. or if the terminal type does not have any codec configuration assigned to it.2 CQI Calculation Without Ideal Link Adaptation 3. The selection of codec modes is mainly based on the radio conditions mentionned above. For calculating the noise. you can optionally define some specific codec mode graphs accoding to a specific hopping mode. The computed noise N is compared to the codec configuration reference noise N Ref . cm. from among the codec modes available in the codec configuration: 147 . CQI corresponding to these codec modes are determined from the look-up tables.4. CM is the set of all available codec modes. As an example.. GSM. Atoll will select the codec mode. Codec modes may be selected using ideal link adaptation or without it. [3] 3GPP Specifications TR 26. CQI calculations may be based on C/N or on C/(I+N). i.2. In this case. John Wiley and Sons Ltd. Atoll only uses the codec configuration of the transmitter. Nevertheless. If the ideal link adaptation option is checked. no codec mode selection and CQI calculation is carried out. GPRS and EDGE performance – Evolution towards 3G/UMTS. mobility type. otherwise the graphs are downshifted by the difference N – N Ref . In the following calculations. Wigard. Atoll only selects the coding schemes that are common in the two.975 V6.  Adaptation Threshold  CM are the values of adaptation thresholds for the codec modes available in the codec configuration. When the calculations are based on C/(I+N): • Atoll calculates the carrier-to-interference ratio for all the TBC transmitters with codec configurations assigned.4.0. according to the codec quality graphs (CQI = f(C/N) and CQI = f(C/I)) related to the defined reference CQI. Ideal link adaptation for circuit quality indicator studies is defined at the codec configuration level. Without ideal link adaptation. Ideal link adaptation implies that the selected codec mode corresponds to the best value of the reference CQI under the given radio conditions. P.4. If no terminal type is defined for the calculation. you can model the gain due to longer MALs in codec mode selection. The selected codec mode among these filtered codec modes will be.  Adaptation Threshold  CM I+N Txi P rec  TRX  . 148 .  Adaptation Threshold   CM N Txi P rec  TRX  For -----------------------. The selected codec mode among these filtered codec modes will be. tt.2.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015  For each TRX type. highest or the lowest value at the received C/N level. Atoll calculates the C/I level received from the transmitter on each pixel of Txi coverage area..for the selected codec mode. ----------------------N tot If more than one codec mode graphs give the same value for reference CQI. Atoll filters all the codec modes that satisfy the C/N criterion (defined by the CQI = f(C/N) graphs for the reference CQI) and are common between the transmitter and the terminal type codec configuration. cm is the codec mode with the highest priority among the set of codec modes CM for which the reference CQI gives the Txi P rec  TRX  ..   For each TRX type. N 3. cm. Then. tt.3. cm = Highest Priority  CM       Txi P rec  TRX  --------------------------. Atoll calculates signal level received from Txi on each pixel of Txi coverage area and converts it into C/N values as described earlier.Atoll 3. cm. For ----------------------I+N 3.. for MOS  Txi  P    TRX  C rec  CQI Ref = Highest  CQI=f  ---= ---------------------------  N  N tot          .4.1 Calculations Based on C/N Ideal link adaptation is used by a codec configuration according to a defined reference CQI (MOS by default). tt.2 Calculations Based on C/(I+N) Atoll selects the highest priority codec mode. cm = Highest Priority  CM       . N tot 3.3 CQI Calculation With Ideal Link Adaptation 3. for BER and FER  Txi   P   TRX  C rec  CQI Ref = Lowest  CQI=f  ---= ---------------------------   N N tot    . Then. from among the codec modes available in the codec configuration:  For each TRX type. Atoll determines the CQI from the CQI=f(C/I) graph associated to the selected codec mode.4.2 Calculations Based on C/(I+N) Ideal link adaptation is used by a codec configuration according to a defined reference CQI (MOS by default).4. cm = Highest Priority  CM    Txi  P  TRX  rec --------------------------. Atoll determines the CQI from the CQI=f(C/N) graph associated to the selected codec mode. cm = Highest Priority  CM      Or. cm. Atoll evaluates the CQI for which the study was Txi P rec  TRX  performed corresponding to -----------------------.3.4. From the CQI = f(C/N) graph associated to the selected codec mode cm.3. Atoll filters all the codec modes that satisfy the C/(I+N) criteria (defined by the CQI = f(C/I) graphs for the reference CQI) and are common between the transmitter and the terminal type codec configuration. then Atoll selects the codec mode with the highest priority. for each TRX and converts it into C/(I+N). The thermal noise (N = -121 dBm. FER. and MOS values in the transmitter coverage areas. Txi.1. for MOS    P Txi  TRX   rec C  CQI = Highest  CQI=f  --.4. tt. All Servers The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  BCCH  3.3. Txi.4. Let us assume that: • • • 3.4. In other words.4. for BER and FER  Txi   P  TRX   C rec  CQI Ref = Lowest  CQI=f  --.= ---------------------------  Ref  I I+N  tot       . highest or the lowest value at the received C/(I+N) level. No max range is set.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1   For each TRX type. then Atoll selects the codec mode with the highest priority. with a defined priority and a defined reception threshold. Txi. Coverage prediction parameters to be set are: • • • The coverage conditions in order to determine the service area of each TBC transmitter. This value can be modified by the user.for the selected codec mode. CQI for which the study was performed corresponding to ----------------------I + N tot 3. From the CQI = f(C/I) graph associated to the selected codec mode cm (indexed with the C/(I+N) values).Atoll 3.4. Atoll calculates the selected parameter on each pixel inside the Txi calculation area. cm is the codec mode with the highest priority among the set of codec modes CM for which the reference CQI gives the Txi P rec  TRX  -. 149 .4 Circuit Quality Indicators Coverage Predictions The Circuit Quality Indicators coverage predictions show the areas BER.1 Each transmitter.4.1 Service Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine the areas coverage will be displayed. The interference and quality indicator conditions to meet for a pixel to be covered. k. 3. cm = Highest Priority  CM       . has a codec configuration assigned. and The display settings to select the displayed parameter and its shading levels.= ---------------------------   I I+N  tot    . Each transmitter. ----------------------I + N tot If more than one codec mode graphs give the same value for reference CQI. each pixel inside the Txi calculation area is considered a probe (non-interfering) receiver.2 Best Signal Level and a Margin The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  BCCH  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji M is the specified margin (dB). belongs to a Hierarchical Cell Structure (HCS) layer. For each TBC transmitter. Atoll evaluates the Txi P rec  TRX  . Atoll considers pixels the received signal level from Txi is the highest. The Best function considers the highest value from a list of values. We can distinguish seven cases as below.1. • If M = 0 dB.4. cm = Highest Priority  CM      Or. by default) is used in the calculations if the coverage prediction is based on C/(I+N). Atoll considers pixels the received signal level from Txi is the highest.4.3. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers.6 If M = 0 dB.1. the service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  BCCH  nd Txi Txj And P rec  BCCH   2 Best  P rec  BCCH   – M ji M is the specified margin (dB). If M = 2 dB. Second Best Signal Level and a Margin The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  BCCH  nd Txi Txj And P rec  BCCH   2 Best  P rec  BCCH   – M ji M is the specified margin (dB). If M = -2 dB. The 2nd Best function considers the second highest value from a list of values. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 3rd best servers. Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin from the second highest. If M = 2 dB. the service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  BCCH  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji M is the specified margin (dB).4.4 If M = 0 dB.1. If M = -2 dB. k.4. If M = -2 dB. If M = -2 dB. The 2nd Best function considers the second highest value from a list of values. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest.Atoll 3.4. Best Signal Level per HCS Layer and a Margin For each HCS layer. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. Atoll considers pixels the received signal level from Txi is the second highest. If M = 2 dB.4.4.3 ©Forsk 2015 If M = 2 dB. k. 150 . Atoll considers pixels the received signal level from Txi is the second highest.5 If M = 0 dB. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 3rd best servers.1.4.1. • • • 3. Second Best Signal Level per HCS Layer and a Margin For each HCS layer.4.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks • • 3. HCS Servers and a Margin The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  BCCH  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji Txi And the received P rec  BCCH  exceeds the reception threshold defined per HCS layer. The Best function considers the highest value from a list of values. • • • 3. Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin from the second highest. • • • 3. There are as many layers as transmitter coverage areas and BER display thresholds. For more information. If M = 2 dB. The Best function considers the highest value from a list of values.4. M is the specified margin (dB). Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. see the Administrator Manual. MOS Only the pixels with a codec mode assigned are coloured. There are as many layers as transmitter coverage areas and FER display thresholds.4.4. Coverage consists of several independent layers whose visibility in the map window can be managed. see "Path Loss Calculation Prerequisites" on page 57 for more information).4. Txi And the received P rec  BCCH  exceeds the reception threshold defined per HCS layer. The Best function considers the highest value from a list of values.7 If M = 0 dB. FER Only the pixels with a codec mode assigned are coloured.2 Coverage Display 3.4. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. 3. • • • 3.2 Display Types It is possible to display the coverage predictions with colours depending on criteria such as: BER Only the pixels with a codec mode assigned are coloured.4.4.1 Coverage Resolution The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. In the case two layers have the same priority. There are as many layers as transmitter coverage areas and MOS display thresholds. If M = -2 dB. Each layer shows the FER in the transmitter coverage area. The pixel colour depends on the FER value.2. Each layer shows the MOS in the transmitter coverage area. Atoll considers pixels the received signal level from Txi is the highest. • • • If M = 0 dB. Coverage consists of several independent layers whose visibility in the map window can be managed. Coverage consists of several independent layers whose visibility in the map window can be managed. The pixel colour depends on the MOS value. The way the competition is managed between layers with the same priority can be modified. Highest Priority HCS Server and a Margin The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  P rec  BCCH  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji And Txi belongs to the HCS layer with the highest priority. If M = 2 dB.3. Atoll considers pixels the received signal level from Txi is the highest. Each layer shows the BER in the transmitter coverage area.2.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 M is the specified margin (dB).4.Atoll 3. 3. The highest priority is defined by the priority field (0: lowest). If M = -2 dB. The pixel colour depends on the BER value. 151 . Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes.1. the traffic is served by the transmitter for which the difference between the received signal strength and the HCS threshold is the highest. Each layer shows the BER value. Txi. No max range is set. Let us assume that: • • Each transmitter. Txi. The pixel colour depends on the highest MOS value among the MOS values for all the transmitters covering the pixel. Max MOS Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the highest FER value among the FER values for all the transmitters covering the pixel.5. The pixel colour depends on the highest BER value among the BER values for all the transmitters covering the pixel. Each layer shows the MOS value.Atoll 3.1. 152 . Two interfaced predictions are available: • • One prediction which shows on each pixel UL losses or UL signal levels One prediction which shows on each pixel UL C/I levels. transmitters are here (non-interfering) receivers. Hence. Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many layers as MOS display thresholds. Results are shown on each pixel.1 DL Service Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine the service areas of the TBC transmitters.3. There are as many layers as BER display thresholds. There are as many layers as FER display thresholds. 3.5 UL Coverage Predictions For each TBC transmitter.5. k.2 Best Signal Level and a Margin The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). Max FER Only the pixels with a codec mode assigned are coloured. 3. P rec  tt  can be replaced Txi Txi with L total – DL or L path . Coverage consists of several independent layers whose visibility in the map window can be managed. and The display settings to select the displayed parameter and its shading levels. We can distinguish eight cases as below. 3. belongs to a Hierarchical Cell Structure (HCS) layer.1. with a defined priority and a defined reception threshold. Coverage consists of several independent layers whose visibility in the map window can be managed.5. P rec  tt  can be replaced Txi Txi with L total – DL or L path . Coverage prediction parameters to be set are: • • The coverage conditions in order to determine the DL service area of each TBC transmitter. Additional studies such as codec modes and coding schemes predictions are used during simulations but are not graphically available.1 All Servers The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)).0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Max BER Only the pixels with a codec mode assigned are coloured. each pixel acting as a transmitting terminal. Atoll calculates the selected parameter at each Txi inside its calculation area. 3. Each layer shows the FER value. 5. • • • If M = 0 dB. P rec  tt  can be replaced Txi Txi with L total – DL or L path . Atoll considers pixels the received signal level from Txi is the highest. k. Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji M is the specified margin (dB). The Best function considers the highest value from a list of values. 3. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers.1. Atoll considers pixels the received signal level from Txi is the highest. k. 3.5. • • • If M = 0 dB.3. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 3rd best servers.1. If M = -2 dB. If M = 2 dB. Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin from the second highest. 3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Txi Txj And P rec  tt   Best  P rec  tt   – M ji M is the specified margin (dB). Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers. P rec  tt  can be replaced Txi Txi with L total – DL or L path .Atoll 3. nd Txi Txj And P rec  tt   2 Best  P rec  tt   – M ji M is the specified margin (dB). If M = 2 dB. If M = -2 dB. The 2nd Best function considers the second highest value from a list of values. the service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold 153 . • • • If M = 0 dB. If M = -2 dB.5 Second Best Signal Level per HCS Layer and a Margin For each HCS layer.5. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. If M = 2 dB. Atoll considers pixels the received signal level from Txi is the second highest. the service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)).1.3 Second Best Signal Level and a Margin The service area of Txi corresponds to the pixels: Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest.4 Best Signal Level per HCS Layer and a Margin For each HCS layer. The Best function considers the highest value from a list of values. For more information. If M = 2 dB. P rec  tt  can be replaced Txi Txi with L total – DL or L path . • • • If M = 0 dB. see the Administrator Manual. Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin from the second highest.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Txi For pure signal level-based calculations (not C/I or C/(I+N)). M is the specified margin (dB). • • • If M = 0 dB. Atoll considers pixels the received signal level from Txi is the second highest. Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji And Txi belongs to the HCS layer with the highest priority. If M = -2 dB. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 3rd best servers. In the case two layers have the same priority.6 HCS Servers and a Margin The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). The way the competition is managed between layers with the same priority can be modified. nd Txi Txj And P rec  BCCH   2 Best  P rec  BCCH   – M ji M is the specified margin (dB). If M = 2 dB.3. Atoll considers pixels the received signal level from Txi is the highest.5.7 Highest Priority HCS Server and a Margin The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  tt   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). The highest priority is defined by the priority field (0: lowest). Txi Txj And P rec  BCCH   Best  P rec  BCCH   – M ji Txi And the received P rec  tt  exceeds the reception threshold defined per HCS layer. The Best function considers the highest value from a list of values. P rec  tt  can be replaced Txi Txi with L total – DL or L path . P rec  tt  can be replaced Txi Txi with L total – DL or L path .1. the traffic is served by the transmitter for which the difference between the received signal strength and the HCS threshold is the highest. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. If M = -2 dB. 3. 3. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers.1. The 2nd Best function considers the second highest value from a list of values.5. 154 . Txi And the received P rec  tt  exceeds the reception threshold defined per HCS layer. Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from transmitters which are 2nd best servers.2.5. dBµV/m) Atoll calculates the signal level received at each transmitter on its service area from surrounding pixels.2.5.8 Best Idle Mode Reselection Criterion (C2) Such type of coverage is useful : • • To compare idle and dedicated mode best servers for voice traffic Display the GPRS/EDGE best server map (based on GSM idle mode) The path loss criterion C1 used for cell selection and reselection is defined by: Txi C1 = P rec  BCCH  – MinimumThreshold  BCCH  The path loss criterion (GSM03. 3. dBµV/m) Atoll calculates the signal level received at each transmitter on its service area from surrounding pixels. dBµV.2 Best UL Signal Level (in dBm. 3.3. If M = -2 dB.5. On each pixel. • • • If M = 0 dB. If M = 2 dB. A pixel of a service area is coloured if the UL signal level is 155 .2. Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from the highest. C2 is defined as an integer in the 3GPP specifications. dBµV. the C2 values in the above calculations are rounded down to the nearest integer. see "Path Loss Calculation Prerequisites" on page 57 for more information).22) is satisfied if C1  0 . And C2 Txi Txj  BCCH  = Best  C2  BCCH   j The Best function considers the highest value from a list of values. Atoll considers pixels the received signal level from Txi is the highest.2. Coverage consists of several independent layers whose visibility in the workspace can be managed.2 Display Types UL signal levels and UL losses calculations are explained in "UL Signal Level" on page 125.5. The service area of Txi corresponds to the pixels : Txi MinimumThreshold  P rec  BCCH   MaximumThreshold Txi For pure signal level-based calculations (not C/I or C/(I+N)). The Best function considers the highest value from a list of values. therefore. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes.2 Coverage by UL Signal Level 3. A pixel of a service area is coloured if the UL signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on signal level).2. It corresponds to the best server in idle mode. When other service areas overlap the studied one.1 Coverage Resolution The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction.1 UL Signal Level (in dBm.1.Atoll 3. 3.5. There are as many layers as transmitter service areas. It is possible to display the coverage by UL signal level with colours depending on any transmitter attribute or other criteria such as: 3. the transmitter with the highest C2 value is kept.5. Atoll chooses the highest value. Each layer shows the different UL signal levels at the transmitter on its service area. 3. P rec  tt  can be replaced Txi Txi with L total – DL or L path .0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 M is the specified margin (dB). The reselection criterion C2 is used for cell reselection only and is defined by: C2 = C1 + CELL_RESELECT_OFFSET CELL_RESELECT_OFFSET is the Cell Reselect Offset defined for the transmitter.2. 5. 3. Atoll takes the TRX UL noise rise in case of non-hopping or extracts a mean noise rise from the several TRXs composing the MSA in case of Base Band Hopping or Synthesized Frequency Hopping. If --I UL not. Coverage consists of several independent layers whose visibility in the workspace can be managed.2 UL C/I Evaluation The UL C/I level can be computed as follows. is the transmitter noise figure is the intra-technology UL noise rise at the considered MSA. Each layer shows the different UL signal levels at the transmitter on its service area.5.2. Since UL noise rise are defined per TRX.5. When other service areas overlap the studied one. for a given MSA C --I MSA Term Tx = P rec – N tot UL . A pixel of a service area is coloured if UL total losses are greater than or equal to the defined minimum thresholds (pixel colour depends on total losses). A pixel of a service area is coloured if UL total losses are greater than or equal to the defined minimum thresholds (pixel colour depends on UL total losses).5. 3. Coverage consists of several independent layers whose visibility 156 . Each layer shows the different total losses levels in the transmitter service area.4. 3. It provides the UL C/I level at the transmitter level caused by surrounding uplink traffic. 3. the worst results (the min C/I per transmitter) are retained. Tx Tx MSA DL • N tot = N thermal + NF • P rec • N thermal is the thermal noise (-121 dBm by default or user-defined) • NF • MSA DL NRIntra – techno log y Term Tx + NR Intra – techno log y is the UL total noise at transmitter on the considered MSA is the received signal level at the transmitter.4 Minimum UL Total Losses (dB) Atoll calculates total losses from the transmitter on each pixel of each transmitter service area.2. see "Path Loss Calculation Prerequisites" on page 57 for more information).3. Coverage consists of several independent layers whose visibility in the workspace can be managed. MSA For a given transmitter having several MSAs.3. MSA is between the lower and upper thresholds defined in the UL 3.3 Coverage by UL C/I An UL C/I coverage predictions is available.3 UL Total Losses (dB) Atoll calculates total losses from the terminal at each transmitter on its service area.1 C/I Level Each pixel of the transmitter coverage area is coloured if the calculated UL C/I level is greater than or equal to the specified minimum thresholds (pixel colour depends on UL C/I level). Each layer shows the different UL total losses at the transmitter on its service area. 3.Atoll 3.5.2.5.2. There are as many layers as service areas. There are as many layers as service areas. Coverage consists of several independent layers whose visibility in the workspace can be managed.4 Display Types It is possible to display the coverage predictions with colours depending on any transmitter attribute or other criteria such as: 3. Atoll chooses the lowest value. There are as many layers as transmitter service areas.3. all possible C are displayed in case the detailed results box is selected.5.1 Coverage Resolution The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. 3.3.3. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes.3 Coverage Area Determination For each MSA.3.5.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 greater than or equal to the defined thresholds (the pixel colour depends on the signal level). coverage area corresponds to pixels where C --I coverage prediction properties. Coverage consists of several independent layers whose visibility in the workspace can be managed.4 Coverage by UL Coding Schemes An UL Coding Scheme coverage prediction is implemented in order to be used in simulations. There are as many layers as defined thresholds. If no terminal type is defined for the calculation. The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  Prec  tt  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – 4dB ji Txi And the received P rec  BCCH  exceeds the reception threshold defined per HCS layer. 3.4. Each layer corresponds to an area the highest received UL C/I level exceeds a defined minimum threshold. and gives priority to the thresholds defined in the transmitter configuration. In that case. For more information on UL C/I calculation. There are as many layers as transmitter service areas. The reception thresholds given for signal level C are internally converted to C/N thresholds (N is the thermal noise defined in the document database at -121 dBm by default) in order to be indexed by C/(I+N) values. These calculations are based on C/(I+N). A pixel of a coverage area is coloured if the UL C/I level is greater than or equal to the specified thresholds (the pixel colour depends on the UL C/I level). we assume that: Txi • P rec  TRX  is the DL signal level received from the BCCH of Txi on each pixel of the Txi coverage area.3. Atoll only selects the coding schemes that are common in the two.5.2 Max C/I Level Atoll compares calculated UL C/I levels received from transmitters on each pixel of each transmitter coverage area coverage areas overlap the studied one and chooses the highest value.5. see "Coverage by UL C/I" on page 156. Coding schemes are selected without using ideal link adaptation.5. 3.5. Each layer shows the different UL C/I levels available in the transmitter coverage area.Atoll 3. Different GPRS/EDGE configurations may be defined for transmitter and terminals. There are as many layers as defined thresholds. In this case.1 Service Area Determination Atoll uses hard-coded parameters for simulations. 157 . the DL service area is based on the option "HCS servers" with a margin of 4 dB.4. • P rec • • CS is the set of all available coding schemes.3. A pixel of a coverage area is coloured if the UL C/I level is greater than or equal to the specified thresholds (the pixel colour depends on the UL C/I level). 3. for the simulations. no coding scheme selection and throughput calculation is carried out. The priorities of the coding scheme lists are as follows: DBS > DAS > MCS > CS.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 in the workspace can be managed. or if the terminal type does not have any GPRS/EDGE configuration assigned to it. C/I thresholds are also indexed by the C/(I+N) value. Each layer corresponds to an area the lowest received UL C/I level exceeds a defined minimum threshold. Atoll only uses the GPRS/EDGE configuration of the transmitter. settings are hard coded and are described hereafter. 3.3.  Reception Threshold  CS are the values of reception thresholds for the coding schemes available in the GPRS/EDGE Term is the UL the signal level received at each transmitter on its service area from surrounding pixels configuration. Coverage consists of several independent layers whose visibility in the workspace can be managed. If the transmitter does not have any GPRS/EDGE configuration assigned to it. • • C  ----------Threshold I + N  CS are the values of C/(I+N) thresholds for the coding schemes available in the GPRS/EDGE configuration. Since the calculations are based on C/I and C/(I+N): • • Atoll calculates the UL C/I to all the GPRS/EDGE TBC transmitters.3 Min C/I Level Atoll compares UL C/I levels received from transmitters on each pixel of each transmitter coverage area the coverage areas overlap the studied one and chooses the lowest value. The prediction itself does not have any interface.4. In the following calculations. no codec mode selection and CQI calculation is carried out.  ---------Threshold I + N  CS N    Term P rec  C ----------------. The resulting throughput TP is given by: TP =   TP C  N +  1 –    TP C   I + N  pN  = ------------------. The computed noise N is compared to the codec configuration reference noise N Ref . The prediction itself does not have any interface. • P rec • • CM is the set of all available codec modes. cs C  N = Lowest  CS     And. If the values are the same. or if the terminal type does not have any codec configuration assigned to it.3 Throughput Calculation For the coding scheme csC/N determined above.4.  ----------. 3. TPC/N.5. corresponding to the C/(I+N) is determined from the TP = f(C/N) graph.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 3.5 in case of HR) corresponding to these codec modes are determined from the look-up tables.  Adaptation Threshold  CM are the values of adaptation thresholds for the codec modes available in the codec Term is the UL the signal level received at each transmitter on its service area from surrounding pixels configuration.4.2 Coding Scheme Selection Atoll selects two coding schemes from among the coding schemes available in the GPRS/EDGE configuration. For the coding scheme csC/(I+N) determined above. Different codec configurations may be defined for transmitter and terminals. 158 . cs C   I + N  = Lowest  CS      Term P rec  C ----------------. the TP = f(C) graph is internally converted to TP = f(C/N) graph.5 Coverage by UL Codec Modes An UL Codec Mode coverage prediction is implemented in order to be used in simulations.5. otherwise the graphs are downshifted by the difference N – N Ref . cs. and p(I+N) is the interferences + thermal noise power (value pI + N in Watts). In this case. A throughput value. and gives priority to the thresholds defined in the transmitter configuration. the defined graphs are used as is.Atoll 3. for the simulations. TPC/(I+N). corresponding to the C/(I+N) is determined from the TP = f(C/(I+N)) graph. and csC/(I+N) is the coding scheme determined from the C/(I+N) level. The final throughput is computed by interpolating between the throughput values obtained from these two graphs. The throughput interpolation method consists in interpolating TPC/N and TPC/(I+N) according to the respective weights of I and N values. the TP = f(C/I) graph is internally converted to TP = f(C/(I+N)) graph. Atoll only uses the codec configuration of the transmitter. such that:   For each MSA . If no terminal type is defined for the calculation. settings are hard coded and are described hereafter. CQI and number of used timeslots (0. Once codec modes have been selected. we assume that: Txi • P rec  TRX  is the DL signal level received from the BCCH of Txi on each pixel of the Txi coverage area. Atoll only selects the coding schemes that are common in the two. is the coding scheme with the higher coding scheme number among csC/N and csC/(I+N): cs = Max  cs C  N cs C   I + N   . 3.Threshold  I + N  CS I+N csC/N is the coding scheme determined from the C/N. pN is the thermal noise power (value in Watts). If more than one codec modes satisfy the quality conditions. If the transmitter does not have any codec configuration assigned to it. Atoll selects the higher priority codec mode. A throughput value.3. Codec modes are selected according to C/(I+N) quality without using ideal link adaptation. Both coding schemes are the coding schemes with the lowest coding scheme numbers from the lowest priority coding scheme list. The selected coding scheme. In the following calculations. Circuit quality indicator calculations include codec mode selection and CQI calculation.5. For more information on UL C/I calculation. t. t. works on the frequency band used by the TCH subcell.5. m.2 The terminal. will be distributed to the BCCH and TCH subcells of a transmitter if: • • 3. m. No HCS Layer) 3. The terminal. services. mobility type. Transmitters considered in traffic analysis are the active and filtered transmitters that belong to the focus zone.6 Traffic Analysis When starting a traffic analysis. cm.5.1.1 Traffic Distribution 3.5. t.1. The terminal.6. 3. t. and a mobility type. Atoll determines the CQI from the CQI=f(C/I) graph associated to the selected codec mode. the DL service area is based on the option "HCS servers" with a margin of 4 dB. works on the frequency band used by the BCCH subcell.1. works on the frequency band used by the BCCH subcell. will be distributed to the BCCH and TCH subcells of a transmitter if: • • • • The transmitter is an GPRS/EDGE station (option specified in the transmitter property dialogue). c. cm. t. • • If no focus zone exists in the . The terminal.1 Normal Cells (Nonconcentric. The service area of Txi corresponds to the pixels : Txi SubcellReceptionThreshold  Prec  BCCH  Txi Txj And P rec  BCCH   Best  P rec  BCCH   – 4dB ji Txi And the received P rec  BCCH  exceeds the reception threshold defined per HCS layer.1 Service Area Determination Atoll uses hard-coded parameters for simulations. works on the frequency band used by the TCH subcell.6. Packet Switched Services A user with a given packet switched service. cm = Highest Priority  CM     Term  P rec ----------------. a terminal. from among the codec modes available in the codec configuration:  For each MSA.3. Atoll takes into account the computation zone.6. Atoll distributes the traffic from maps to transmitters of each layer according to the compatibility criteria defined in the transmitter. In that case.atl document. p. For details of the average timeslot capacity calculation. 3. 3.Atoll 3. see the Network Dimensioning section (calculation of minimum reduction factor).1. For -----------I+N 3. t.  Adaptation Threshold   CM I+N Term P rec .2 Codec Mode Selection Atoll selects the highest priority codec mode. t. see "Coverage by UL C/I" on page 156. is technologically compatible with the transmitter.1 Circuit Switched Services A user with a given circuit switched service. and a mobility type. The terminal. terminal type properties.5.6.1.. a terminal. 159 .0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Since the calculations are based on C/I and C/(I+N): • Atoll calculates the UL C/I to all the GPRS/EDGE TBC transmitters. m. m. p. k. which are determined using the option “Best signal level per HCS layer” with a 0dB margin and the subcell reception threshold as lower threshold. works on the frequency band(s) used by the TCH_INNER and TCH subcells. 3.3. The terminal. The terminal.Atoll 3. works on the frequency band(s) used by the TCH_INNER and TCH subcells. t.2. Circuit Switched Services For a circuit switched service c.6.1 Normal Cells (Nonconcentric. BCCH and TCH subcells of a transmitter if: • • • • The transmitter is an GPRS/EDGE station (option specified in the transmitter property dialogue). c. Atoll calculates the probability for the user to be connected with a given service using a terminal t. t. The terminal. t.1. works on the frequency band used by the BCCH subcell. 3. Atoll distributes traffic on subcell service areas.6. 3. Same traffic is distributed to the BCCH and TCH subcells. a terminal. t. For each behaviour described in the user profile up. mobility. No HCS Layer) Number of subscribers ( X up m ) for each TCH subcell (Txi.m is the TCH service area containing the user profile up with the mobility m and D is the user profile density.1.2 Concentric Cells In case of concentric cells. t.1. t. per user profile up with a given mobility m. c.6. a terminal. works on the frequency band used by the BCCH subcell. works on the frequency band(s) used by the TCH_INNER and TCH subcells.3. and a mobility type.1 Circuit Switched Services A user with a given circuit switched service. t.1 User Profile Traffic Maps 3. we have: 160 . t. m. The terminal.2 The terminal. a terminal.2 The terminal.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 3. t.1. t. you may specify the maximum mobile speed supported by the transmitters of the layer. t. a terminal.6. The terminal. 3.6. will be distributed to the BCCH and TCH subcells (and TCH_INNER in case of concentric cells) of a transmitter if: • • • 3. p. and a mobility type. k. 3.1 Circuit Switched Services A user with a given circuit switched service.6. is less than the maximum speed supported by the layer. terminal properties as explained above. k. will be distributed to the BCCH and TCH subcells (and TCH_INNER in case of concentric cells) of a transmitter if: • • • • • The transmitter is an GPRS/EDGE station (option specified in the transmitter property dialogue).1.2 Calculation of the Traffic Demand per Subcell Here we assume that: • • • Users considered for evaluating the traffic demand fulfil the compatibility criteria defined in the transmitter. The user’s mobility. m. t. Packet Switched Services A user with a given packet switched service. is technologically compatible with the transmitter. t. works on the frequency band used by the BCCH subcell. m.6. works on the frequency band(s) used by the TCH_INNER and TCH subcells. The terminal. will be distributed to the TCH_INNER.1.2.2. TCH). and a mobility type.2. is inferred as: X up m  Txi TCH  = S up m  Txi TCH   D Sup. t.6.3. TCH_INNER TRX type has the highest priority to carry traffic.1. will be distributed to the TCH_INNER.3 HCS Layers For each HCS layer. works on the frequency band used by the BCCH subcell. Packet Switched Services A user with a given packet switched service. The user mobility. is technologically compatible with the transmitter. and a mobility type.6. The terminal. The terminal. m. is less than the maximum speed supported by the layer. BCCH and TCH subcells of a transmitter if: • • 3. services. Atoll distributes a part of traffic on the TCH_INNER service area (TCH_INNER is the highest priority traffic carrier) and the remaining traffic on the outer ring served by the TCH subcell. per user profile up with a given mobility m. Figure 3. D up  p t  m  Txi TCH  = X up m  Txi TCH   p up  p t  Packet Switched Services (Constant Bit Rate) For a constant bit packet switched service p. TCH) service area. D up  p t  m .TCH_INNER   D X up m  Txi. Atoll evaluates the traffic demand.TCH_INNER  = S up m  Txi. In this case.TCH  respectively refer to the TCH_INNER and TCH subcell service areas containing the user profile up with the mobility m. in Erlangs for the subcell (Txi.2. Number of subscribers ( X up m ) for each TCH_INNER (Txi. TCH_INNER) and TCH (Txi. TCH) subcell. D up  c t  m .TCH_INNER  and S up m  Txi. The traffic spread over the TCH_INNER subcell may overflow to the TCH subcell.TCH  – S up m  Txi.2 Concentric Cells In case of concentric cells. TCH) service area. D up  p t  m .3. in kbits/s for the subcell (Txi.4: Representation of a Concentric Cell TXi 161 .Atoll 3. in kbits/s for the subcell (Txi. • Traffic overflowing from the TCH_INNER to the TCH is not uniformly spread over the TCH service area.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 N call  d p up  c t  = ------------------3600 Ncall is the number of calls per hour and d is the average call duration (in seconds). D is the user profile density. Then. TCH) service area. D up  p t  m  Txi TCH  = X up m  Txi TCH   p up  p t  3.TCH_INNER    D S up m  Txi.1.TCH  =  S up m  Txi. we have: N call  V  8 p up  p t  = ----------------------------3600 Ncall is the number of calls per hour and V is the transmitted data volume per call (in Kbytes). Then. D up  c t  m  Txi TCH  = X up m  Txi TCH   p up  c t  Packet Switched Services (Max Rate) For a max rate packet switched service p. is inferred as: X up m  Txi. It is still located on the TCH_INNER service area. the traffic demand is the same on the TCH_INNER subcell but increases on the TCH subcell. Then. Atoll evaluates the traffic demand.6. we have: N call  d p up  p t  = ------------------3600 Ncall is the number of calls per hour and d is the average call duration (in seconds). Atoll evaluates the traffic demand. TCH_INNER   p up  p t  D up  p t  m  Txi. TCH_INNER) and (Txi. TCH) subcell service areas. in kbits/s in the (Txi.TCH  = X up m  Txi. Atoll evaluates the traffic demand. Normal Cells Atoll distributes traffic on the TCH service areas. In this case. The traffic contained in the input traffic map can be assigned to all the HCS layers. in Erlangs in the (Txi. D up  p t  m . probability of the user being connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159.3.Atoll 3. Calculations are detailed in "Circuit Switched Services" on page 159.3 HCS Layers We assume two HCS layers: the micro layer has a higher priority than the macro layer. Packet Switched Services (Constant Bit Rate) For each user of the user profile up using a constant bit packet switched service p with a terminal t. TCH_INNER) and (Txi. 3.2.TCH_INNER  = X up m  Txi.TCH  = X up m  Txi.TCH  = X up m  Txi. probability of the user being connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159. Let S overlapping  Txj TCH  denote this area (TCH service area of the macro layer overlapped by the TCH service area of the micro layer).TCH_INNER  = X up m  Txi. the traffic demand is the same on the TCH subcell of the micro layer but increases on the TCH subcell of the macro layer. TCH_INNER) and (Txi.TCH_INNER   O max  Txi. D up  p t  m  Txi. Traffic overflowing to the macro layer is not uniformly spread over the TCH service area of Txj.1. Txi belongs to the micro layer and Txj to the macro.TCH_INNER   O max  Txi. D up  p t  m . traffic of the micro layer may overflow to the macro layer. Atoll evaluates the traffic demand.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Circuit Switched Services For each user of the user profile up using a circuit switched service c with a terminal t. D up  c t  m  Txi.TCH_INNER  is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell.TCH   p up  p t  + D up  p t  m  Txi.TCH   p up  p t  + D up  p t  m  Txi.TCH_INNER   p up  p t  D up  p t  m  Txi. It is only located on the overlapping area. D up  c t  m . TCH) subcell service areas. Packet Switched Services (Max Rate) For each user of the user profile up using a max rate packet switched service p with a terminal t.TCH_INNER  is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell. TCH) subcell service areas. Then. On this area.TCH_INNER  is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell.TCH_INNER   p up  c t  D up  c t  m  Txi.TCH_INNER  O max  Txi. 162 . The traffic capture is calculated with the option “Best signal level per HCS macro layer” meaning that there is an overlap between HCS layers service areas.TCH   p up  c t  + D up  c t  m  Txi.TCH_INNER  = X up m  Txi. Atoll evaluates the traffic demand.TCH_INNER   O max  Txi. Traffic on the overlapping area is distributed to the TCH subcell of the micro layer because it has a higher priority.6.TCH_INNER  O max  Txi.TCH_INNER  O max  Txi. in kbits/s in the (Txi. D up  p t  m  Txi. Atoll calculates the probability ( p up  c t  ) of the user being connected. macro Then. macro S upm – overlapping  Txj TCH  macro macro micro D up  c t m  Txj TCH  = X up m Txj TCH   p up  c t  + D up  c t m Txi TCH   ----------------------------------------------------------------- Omax Txi TCH  micro S up m  Txi TCH  For each user described in the user profile up with the packet switched service p and the terminal t.TCH_INNER  macro S 3 = S up m – overlapping –  Txi TCH   Txj. The traffic capture is calculated with the option “Best signal level per HCS layer”. S1. is inferred as: macro macro macro X up m  Txj TCH  =  S up m  Txj TCH  – S up m – overlapping  Txj TCH    D macro S up m  Txj TCH  is the TCH service area of Txj containing the user profile up with the mobility m and D is the profile density.Atoll 3. macro Then. please refer to formulas given in case of concentric cells. Then. TCH) of the macro layer.TCH_INNER   Txj. Atoll evaluates the traffic demand. it proceeds with the macro layer (lower priority HCS layer).TCH_INNER  and another overlapped by macro the TCH_INNER service area of the micro layer S overlapping –  Txi. Let us consider three areas. Atoll evaluates the traffic demand. macro S upm – overlapping  Txj TCH  macro macro micro D up  p t m  Txj TCH  = X up m Txj TCH   p up  p t  + D up  p t m Txi TCH   ----------------------------------------------------------------- Omax Txi TCH  micro S up m  Txi TCH  O max  Txi TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi (micro micro layer) and S up m  Txi TCH  is the TCH service area of Txi containing the user profile up with the mobility m. the probability for the user being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 159. TCH) service area. For each user described in the user profile up with the circuit switched service c and the terminal t. On these areas. probability for the user to be connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159. in Erlangs in the subcell (Txj.TCH_INNER  . in kbits/s in the subcell (Txj. traffic of the higher priority layer may overflow.TCH_INNER  – S 2 163 .TCH_INNER  – S up m – overlapping –  Txi TCH   Txj.TCH_INNER   Txj.3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Figure 3. per user profile up with the mobility m. Concentric Cells Atoll evaluates the traffic demand on the micro layer (higher priority HCS layer) as explained above. S2 and S3. It means that there are overlapping areas between HCS layers traffic is spread according to the layer priority.5: Representation of Micro and Macro Layers Atoll evaluates the traffic demand on the micro layer (higher priority) as explained above. macro Number of subscribers ( X up m ) for each TCH subcell (Txj. For further details. D up  p t  m . it proceeds with the macro layer (lower priority). For further details. please refer to formulas for normal cells.TCH_INNER  macro S 2 = S up m – overlapping –  Txi. This area consists of two parts: an area macro overlapped by the TCH service area of the micro layer S overlapping –  Txi TCH   Txj. macro macro S 1 = S up m  Txj. D up  c t  m . The TCH_INNER service area of the macro layer is overlapped by the micro layer. Then. TCH) service area. 0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Figure 3.TCH_INNER   O  Txi.Atoll 3. the probability for the user being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 159. Atoll evaluates the macro traffic demand. stated in kbits/s in the subcell (Txj. D up  p t  m .TCH  and S up m  Txi. macro X up m  Txj. S2 R 2 = ------------------------------------------------------micro S up m  Txi.TCH_INNER  is the TCH_INNER subcell service area of Txj containing the user profile up with the mobility m.TCH_INNER  = R  D micro  Txi. in Erlangs in the subcell (Txj.3. TCH_INNER) service area.TCH_INNER   p up  p t  + macro D up  p t  m  Txj. probability for the user to be connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159. macro On S1.TCH_INNER  = S 1  D D is the user profile density.TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively containing the user profile up with the mobility m. S3 R 3 = -------------------------------------------------------------------------------------------------micro micro S up m  Txi. macro X up m  Txj.TCH  + micro R 3  X up m  Txi TCH   p up  p t   O max  Txi TCH  164 . For each user described in the user profile up with a circuit switched service c and a terminal t.TCH_INNER   p up  c t  + macro D up  c t  m  Txj. macro Then. We only consider the overlapping areas containing the user profile up with the mobility m.TCH  – S up m  Txi. Atoll evaluates the traffic demand. TCH_INNER) service area.TCH_INNER   O max  Txi.TCH_INNER   O  Txi. Then. The traffic overflowing to the TCH subcell is located on the TCH_INNER service area.TCH  + 2 up  c t  m max max micro R 3  X up m  Txi TCH   p up  c t   O max  Txi TCH  For each user described in the user profile up with a packet switched service p and a terminal t.TCH_INNER  micro micro S up m  Txi. The traffic spread over the TCH_INNER service area of the micro layer may overflow on the TCH subcell.TCH_INNER   O max  Txi.TCH_INNER  = R  D micro 2 up  p t  m  Txi. On S2. the number of subscribers per user profile up with a given mobility m ( X up m ) is inferred: macro X up m  Txj.TCH_INNER  The traffic spread over the ring served by the TCH subcell of the micro layer only may overflow on S3 proportional to R3. the TCH subcell traffic coming from the TCH_INNER subcell traffic overflow may overflow proportional to R2. D up  c t  m .6: Concentric Cells macro S up m  Txj. TCH) service area.TCH_INNER  – S' 2 macro macro S up m  Txj. S' 2 R' 2 = ------------------------------------------------------micro S up m  Txi. TCH) service area. macro Then. the probability for the user being connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159. The traffic overflowing on the TCH subcell is located on the TCH_INNER service area.TCH  and S up m  Txi. We only consider the overlapping areas containing the user profile up with the mobility m.TCH_INNER  and another one by the TCH_INNER macro service area of the micro layer S overlapping –  Txi. The area of the TCH ring of the macro layer is overlapped by the micro layer. S’2 and S’3.TCH -. macro On S’1.TCH_INNER  – S up m – overlapping –  Txi TCH   Txj.3. D up  p t  m . macro X up m  Txj TCH   p up  p t  + macro D up  p t  m  Txj TCH  = macro D up  p t  m  Txj.TCH_INNER  + micro R' 2  D up  c t  m  Txi.TCH_INNER  + micro R' 2  D up  p t  m  Txi. S' 3 R' 3 = -------------------------------------------------------------------------------------------------micro micro S up m  Txi.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 O max  Txi TCH  and O max  Txi. macro X up m  Txj TCH   p up  c t  + macro D up  c t  m  Txj TCH  = macro D up  c t  m  Txj.TCH -. The traffic spread over the TCH_INNER service area of the micro layer may overflow on the TCH subcell.TCH  = S' 1  D D is the user profile density. D up  c t  m .TCH_INNER  macro S' 3 = S up m – overlapping –  Txi TCH   Txj.TCH_INNER  micro micro S up m  Txi. macro Then.TCH_INNER   O max  Txi.TCH_INNER   O max  Txj.TCH_INNER  The traffic spread over the ring served by the TCH subcell of the micro layer only may overflow on S’3 proportional to R’3.TCH  + micro R' 3  X up m  Txi TCH   p up  c t  m  O max  Txi TCH  For each user described in the user profile up with a packet switched service p and a terminal t. Let us consider three areas.TCH_INNER   Txj.TCH_INNER  macro S' 2 = S up m – overlapping –  Txi.TCH_INNER  are the TCH and TCH_INNER subcell service areas of Txj respectively.TCH_INNER   Txj.TCH_INNER   O max  Txi.TCH_INNER   O max  Txj. Atoll evaluates the traffic demand. On S’2.Atoll 3.TCH -.TCH  + micro R' 3  X up m  Txi TCH   p up  p t  m  O max  Txi TCH  165 . macro macro macro S' 1 = S up m  Txj.TCH  and S up m  Txj. There are two parts: an area overlapped by the macro TCH service area of the micro layer S overlapping –  Txi TCH   Txj.TCH  – S up m  Txj.TCH_INNER  . the TCH subcell traffic coming from the TCH_INNER subcell traffic overflow may overflow proportionally to R’2. S’1. the number of subscribers per user profile up with a given mobility m ( X up m ) is inferred: macro X up m  Txj.TCH -.TCH -. Atoll evaluates the traffic demand.TCH  – S up m  Txi. in Erlangs in the subcell (Txj. the probability for the user being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 159.TCH_INNER   O max  Txi. For each user described in the user profile up with a circuit switched service c and a terminal t.TCH_INNER  are the maximum rates of traffic overflow (stated in %) specified for the TCH and TCH_INNER subcells of Txi respectively. in kbits/s in the subcell (Txj.TCH_INNER   O max  Txi.TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively containing the user profile up with the mobility m. TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi (micro layer).6.3. p. For each circuit switched service. E c  Txi TCH  S  Txi TCH  and D c t m  Txi. and the mobility type.m. the traffic demand is the same on the TCH_INNER subcell and rises on the TCH subcell. Dp.2 Concentric Cells In case of concentric cells. you have to specify the traffic demand per transmitter and per service (throughput for a max rate packet switched service and Erlangs for a circuit switched or constant bit rate packet switched service) and the global distribution of terminals and mobility types. TCH). service areas. TCH).TCH_INNER   E c  Txi TCH  + S  Txi TCH  D c t m  Txi. on the TCH subcell of Txi. No HCS Layer) For each circuit switched service. 3.m.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 O max  Txi.2 Sector Traffic Maps We assume that the traffic map is built from a coverage by transmitter prediction calculated for the TCH subcells with options: • • “HCS Servers” and no margin if the network only consists of normal cells and concentric cells. in kbits/s in the subcell (Txi. in Erlangs in the subcell (Txi. TCH) service area. Atoll evaluates the traffic demand.t. Atoll evaluates the traffic demand. D c t m  Txi TCH  = E c  Txi TCH  For each packet switched service (Max Bit Rate). Atoll evaluates the traffic demand. p. O max  Txi. “Highest Priority HCS Server” and no margin in case of HCS layers. in kbits/s in the subcell.TCH  = -------------------------------------------------------------------------------S  Txi. D p t m  Txi TCH  = T p  Txi TCH  For each packet switched service (Constant Bit Rate). It is only located on the TCH_INNER service area.TCH_INNER  D c t m  Txi. p.TCH  – S  Txi.m. Atoll distributes a part of traffic on the TCH_INNER service area (TCH_INNER is the highest priority traffic carrier) and the remaining traffic. The traffic spread over the TCH_INNER subcell may overflow to the TCH subcell. Let E c  Txi TCH  denote the Erlangs for the circuit switched service. Dp. on the ring served by the TCH subcell only. In this case.2.6. c.TCH_INNER   O max  Txi. service areas.TCH_INNER  the maximum rate of traffic overflow indicated for the TCH_INNER subcell of Txi (macro layer). When creating the traffic map. p.2. Let T p  Txi TCH  denote the throughput of the packet switched service (Max Bit Rate).m.6. We assume that 100% of users have the terminal. m. Let E p  Txi TCH  denote the Erlangs for the packet switched service (Constant Bit Rate).t. Dc.2. Atoll evaluates the traffic demand. Dc.t.TCH_INNER  = -------------------------------------------. in Erlangs in the subcell.1 Normal Cells (Nonconcentric.t. TCH_INNER) and (Txi. t.2.TCH_INNER  the maximum rate of traffic overflow indicated for the TCH_INNER subcell of Txj (macro layer) and micro X up m  Txi TCH  the number of subscribers with the user profile up and mobility m on the TCH service area of Txi (as explained in "Concentric Cells" on page 160). in kbits/s in the subcell (Txi. Atoll evaluates the traffic demand. p.Atoll 3. c. on the TCH subcell of Txi. D p t m  Txi TCH  = E p  Txi TCH   TP p GBR TP p GBR is the guaranteed bit rate of the constant bit rate packet switched service p. TCH) service area. (Txi. 3. O max  Txj. on the TCH subcell of Txi. TCH) service area. 3. TCH_INNER) and (Txi.TCH_INNER  For each packet switched service (Max Bit Rate). Traffic overflowing from the TCH_INNER to the TCH is not uniformly spread over the TCH service area. 166 . c.2. S  Txi.t. (Txi.m. Dp. TCH_INNER) and (Txi. Atoll starts evaluating the traffic demand on the micro layer (highest priority HCS layer). Normal Cells Atoll distributes traffic on the TCH service areas. micro D c t m  Txi TCH  = E c  Txi TCH  micro For each packet switched service (Max Bit Rate). S  Txi.2. TCH). service areas.TCH_INNER   --------------------------------------------------------------------------------.TCH_INNER  O max  Txi. D p t m . Traffic on the overlapping area is distributed to the TCH subcell of the micro layer (higher priority layer).TCH  and S  Txi. p. D c t m . TCH) service area. S  Txi. T p  Txi TCH  + S  Txi TCH  D p t m  Txi. It is only located on the overlapping area. in kbits/s in the subcell (Txi. Atoll proceeds with the macro layer (lower priority HCS layer).TCH_INNER  D p t m  Txi. micro D p t m  Txi TCH  = T p  Txi TCH  micro For each packet switched service (Constant Bit Rate). micro D p t m  Txi TCH  = E p  Txi TCH   TP p GBR Then. micro For each circuit switched service.TCH_INNER  D p t m  Txi. Atoll calculates the traffic demand. Let S overlapping  Txj TCH  denote the TCH service area of the macro layer overlapped by the TCH service area of the micro layer.6.TCH  – S  Txi. in kbits/s in the subcell. Atoll evaluates the traffic demand.TCH  =  S  Txi. T p  Txi TCH  S  Txi TCH  and D p t m  Txi. the traffic demand is the same on the TCH subcell of the micro layer but rises on the TCH subcell of the macro layer. c.TCH_INNER   E p  Txi TCH   TP p GBR + S  Txi TCH  D p t m  Txi. (Txi. For each packet switched service (Constant Bit Rate). In this case. in Erlangs in the subcell (Txi. Atoll calculates the traffic demand. TCH) service area. macro For each circuit switched service.2. Dp. Atoll calculates the traffic demand. p. Traffic overflowing on the macro layer is not uniformly spread over the TCH service area of Txj. D p t m . Txi belongs to the micro layer and Txj to the macro one.3.TCH  – S  Txi.TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell.TCH_INNER   O max  Txi.TCH_INNER  = -------------------------------------------.TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell. c. D c t m . On this area.m.3 HCS Layers We assume we have two HCS layers: the micro layer has a higher priority and the macro layer has a lower one. in Erlangs in the subcell (Txj. It means that macro there is an overlapping area between HCS layers. The traffic capture is calculated with the option “HCS Servers”.TCH_INNER  O max  Txi. TCH) service area. 3.TCH_INNER   O max  Txi. traffic of the micro layer may overflow to the macro layer. TCH) service area. S  Txi.TCH  and S  Txi. 167 . E p  Txi TCH   TP p GBR S  Txi TCH  and D p t m  Txi.Atoll 3.t. p.TCH_INNER  = -------------------------------------------.TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively. Atoll calculates the traffic demand.TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 S  Txi. The traffic contained in the input traffic map can be assigned to all the HCS layers.TCH  = -------------------------------------------------------------------------------S  Txi. in kbits/s in the subcell (Txi. Let us consider three areas.TCH_INNER   Txj. TCH) service area. You can restrict the traffic assignement of each traffic map to a specific HCS layer in the running options of the traffic capture. D p t m .Atoll 3. no overflow occurs between HCS layers and the only overflow which is considered occurs within concentric cells (See "Concentric Cells" on page 160). Atoll calculates the traffic demand.TCH_INNER  . If you do so. D p t m . TCH) service area. Atoll calculates the traffic demand. traffic of the higher priority layer may overflow. Concentric Cells Atoll evaluates the traffic demand on the micro layer as explained above in case of concentric cells and then proceeds with the macro layer (lower priority layer). O max  Txi TCH  D c t m  Txj TCH  = E c  Txj TCH  + D c t m  Txi TCH   ---------------------------------------------------micro S  Txi TCH  macro For each packet switched service (Max Bit Rate). O max  Txi TCH  D p t m  Txj TCH  = T p  Txj TCH  + D p t m  Txi TCH   ---------------------------------------------------micro S  Txi TCH  O max  Txi TCH  is the maximum rate of traffic overflow (in %) specified for the TCH subcell of Txi (micro cell) and S micro  Txi TCH  the TCH service area of Txi. p. in kbits/s in the subcell (Txj. p. O max  Txi TCH  D p t m  Txj TCH  = E p  Txi TCH   TP p GBR + D p t m  Txi TCH   ---------------------------------------------------micro S  Txi TCH  O max  Txi TCH  is the maximum rate of traffic overflow (in %) specified for the TCH subcell of Txi (micro cell) and S micro  Txi TCH  the TCH service area of Txi.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 macro S overlapping  Txj TCH  macro micro . It means that there is overlapping areas between HCS layers traffic is spread over according to the layer priority.TCH_INNER  and another overlapped by macro the TCH_INNER service area of the micro layer S overlapping –  Txi. macro S overlapping  Txj TCH  macro micro . The traffic capture is calculated with the option “HCS Servers”. macro For each packet switched service (Constant Bit Rate). 168 . S1.3. Figure 3. On these areas. This area consists of two parts: an area macro overlapped by the TCH service area of the micro layer S overlapping –  Txi TCH   Txj.7: Concentric Cells The TCH_INNER service area of the macro layer is overlapped by the micro layer. S2 and S3. macro S overlapping  Txj TCH  macro micro . in kbits/s in the subcell (Txj. 0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 S1 = S macro macro  Txj. R 1  T p  Txj TCH  + macro D p t m  Txj. in Erlangs in the subcell (Txj.TCH_INNER  – S overlapping –  Txi TCH   Txj.TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi and S micro  Txi TCH  is the TCH subcell service area of Txi. in kbits/s in the subcell (Txj. O max  Txi. p.TCH_INNER  -  ---------------------------------------------------------------------------------------------------------micro   Txi TCH  S R3      E  Txi  TCH   TP  p p GBR  O max  Txi TCH         169 . TCH_INNER) service area. the TCH subcell traffic coming from the TCH_INNER subcell traffic overflow may overflow proportional to R2. in kbits/s in the subcell (Txj. The traffic specified for Txj in the map description ( E c  Txj TCH  ) is spread over S1 proportionally to R1.TCH_INNER  macro S 3 = S overlapping –  Txi TCH   Txj. macro For each packet switched service (Constant Bit Rate). c.TCH_INNER   O max  Txi TCH  + micro micro S  Txi TCH  – S  Txi. Atoll calculates the traffic demand.3.TCH_INNER  – S 2 S macro  Txj.TCH_INNER   O max  Txi TCH  + micro micro S  Txi TCH  – S  Txi. E c  Txi TCH   O max  Txi TCH  R 3  ---------------------------------------------------------------------------------------------------------micro S  Txi TCH  macro For each packet switched service (Max Bit Rate).Atoll 3. TCH_INNER) service area. p.TCH_INNER   .TCH_INNER  = micro R 2  D c t m  Txi. The traffic overflowing to the TCH subcell is located on the TCH_INNER service area.TCH_INNER   O max  Txi. S2 R 2 = ------------------------------------------------------micro S  Txi.TCH_INNER  = O max  Txi TCH  micro R 2  D p t m  Txi. R 1  E p  Txi TCH   TP p GBR + micro R 2  D p t m  Txi. TCH_INNER) service area.TCH_INNER   Txj.TCH_INNER   O max  Txi.TCH_INNER   O max  Txi TCH  + macro D p t m  Txj.TCH_INNER  The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S3 proportional to R3. D p t m .TCH_INNER  - R 3  --------------------------------------------------------------------------------------------------------- T p  Txi TCH   O max  Txi TCH  micro  Txi TCH  S is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi. S3 R 3 = -------------------------------------------------------------------------------------------------micro micro S  Txi. The traffic spread over the TCH_INNER service area of the micro layer may overflow to the TCH subcell. Atoll calculates the traffic demand.TCH_INNER  =  micro micro S  Txi TCH  – S  Txi.TCH_INNER   O max  Txi.TCH_INNER  is the TCH_INNER subcell service area of Txj. S1 R 1 = ------------------------------------map S  Txj TCH  map S  Txj TCH  is the TCH service area of Txj in the traffic map with the option “Best signal level of the highest priority layer”. R 1  E c  Txj TCH  + macro D c t m  Txj. On S2.TCH  – S  Txi. D p t m . D c t m .TCH_INNER  macro For each circuit switched service. Atoll calculates the traffic demand.TCH_INNER  macro S 2 = S overlapping –  Txi. The traffic overflowing to the TCH subcell is located on the TCH_INNER service area. D p t m . E c  Txi TCH   O max  Txi TCH  R' 3  ------------------------------------------------------------------------------------------------------micro S  Txi. The area of the TCH ring of the macro layer is overlapped by the micro layer.TCH_INNER  . T p  Txi TCH   O max  Txi TCH  R' 3  ------------------------------------------------------------------------------------------------------micro S  Txi.TCH_INNER   O max  Txi TCH  + micro micro S  Txi. O max  Txi.3.TCH_INNER  – S overlapping –  Txi TCH   Txj.TCH and another overlapped by the -.TCH_INNER  – S' 2 S macro  Txj TCH  and S macro  Txj.TCH_INNER  macro For each circuit switched service. S' 1 R' 1 = ------------------------------------map S  Txj TCH  map S  Txj TCH  is the TCH service area of Txj in the traffic map with the option “Best signal level of the highest priority layer”.TCH_INNER  macro S' 2 = S overlapping –  Txi. TCH) service area.TCH_INNER  TCH_INNER service area of the micro layer macro S overlapping –  Txi.TCH_INNER  The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S’3 proportional to R’3.TCH_INNER  + micro R' 2  D c t m  Txi. the TCH subcell traffic coming from the TCH_INNER subcell traffic overflow may overflow proportional to R’2.TCH_INNER   O max  Txi. S' 2 R' 2 = ------------------------------------------------------micro S  Txi. in kbits/s in the subcell (Txj.TCH -.TCH -. D c t m . p. in Erlangs in the subcell (Txj.TCH_INNER   .TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi and S micro  Txi TCH  is the TCH subcell service area of Txi.TCH  – S  Txi.TCH_INNER   O max  Txj. S' 3 R' 3 = -------------------------------------------------------------------------------------------------micro micro S  Txi.TCH_INNER   Txj.TCH -.Atoll 3. R' 1  T p  Txj TCH  + macro macro D p t m  Txj TCH  = D c t m  Txj.TCH  170 . There are two parts: an area overlapped by the TCH service area of the micro layer macro S overlapping –  Txi TCH   Txj.TCH  – S  Txi.TCH_INNER   Txj.TCH_INNER   O max  Txi.TCH  macro For each packet switched service (Max Bit Rate).TCH_INNER   O max  Txj. Let us consider three areas. c.TCH  – S  Txi. The traffic specified for Txj in the map description ( E c  Txj TCH  ) is spread over S’1 proportional to R’1.TCH -. On S’2.TCH_INNER   . The traffic spread over the TCH_INNER service area of the micro layer may overflow to the TCH subcell. S’2 and S’3. S' 1 = S macro  Txj TCH  – S macro macro  Txj.TCH_INNER  are the TCH and TCH_INNER subcell service areas of Txj respectively. R' 1  E c  Txj TCH  + macro macro D c t m  Txj TCH  = D c t m  Txj.Atoll calculates the traffic demand.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks O max  Txi TCH  ©Forsk 2015 is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi. S’1.TCH_INNER  macro S' 3 = S overlapping –  Txi TCH   Txj. Atoll calculates the traffic demand.TCH_INNER  + micro R' 2  D p t m  Txi. TCH) service area.TCH_INNER   O max  Txi TCH  + micro micro S  Txi. O max  Txi TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi. the Grade of Service is the percentage of incoming calls that are placed in a waiting queue when there are no resources available. a dimensioning model is an entity utilized by the dimensioning engine along with other inputs (traffic.1 Dimensioning Models and Quality Graphs In Atoll.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 O max  Txj. micro S Txi. D p t m . R' 1  E p  Txi TCH   TP p GBR + macro D c t m  Txj. The user can define either to use Erlang B or Erlang C queuing model for circuit switched traffic and can define which KPIs to consider when dimensioning the network for packet switched traffic. Therefore.TCH_INNER  is the TCH_INNER subcell service area of macro For each packet switched service (Constant Bit Rate).TCH_INNER  + micro macro D p t m  Txj TCH  = R' 2  D p t m  Txi. This queuing system has no lost calls.1 Circuit Switched Traffic The network dimensioning for circuit switched traffic is performed using the universally accepted and adopted Erlang B and Erlang C formulas. For example. criteria. Wireless Communications Principles and Practice by Theodore S. O max  Txi. 171 .7. micro S Txi. The KPIs not selected are supposed to be either already satisfactory or not relatively important.1. These formulas and their details are available in many books. as explained later in the network dimensioning steps.TCH  is the TCH subcell service area of Txi and S micro  Txi. Atoll first performs network dimensioning according to the circuit switched traffic present in the subcell in order to ensure the higher priority service availability before performing the same for the packet switched traffic.  Txi.3.  Txi. 3.TCH_INNER   O max  Txi TCH  +  micro micro S  Txi. TCH) service area. In the Erlang C approach.TCH_INNER   O max  Txj.TCH_INNER  is the TCH_INNER subcell service area of 3. the average waiting time in the queue also increases.TCH  R' 3       E p  Txi TCH   TP p GBR  O max  Txi TCH         O max  Txj. p.7 Network Dimensioning Atoll is capable of dimensioning a GSM GPRS EDGE network with a mixture of circuit and package switched services. This section describes the technical details of Atoll’s dimensioning engine. Rappaport. in kbits/s in the subcell (Txj. The blocked calls are supposed to be lost and the caller has to reinitiate it.TCH  is the TCH subcell service area of Txi and S micro  Txi.Atoll calculates the traffic demand.TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txj. The dimensioning criterion in these formulas is the Grade of Service or the allowed blocking probability of the circuit switched traffic.) in the process of dimensioning. Prentice Hall.Atoll 3. A dimensioning model defines the QoS KPIs to be taken into account when dimensioning a network for both circuit and packet switched traffic.TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txj. O max  Txi. The dimensioning engine will only utilize the quality curves of the KPI selected.TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi. In the Erlang B approach. limitations. until some resources or timeslots are liberated.TCH_INNER    ------------------------------------------------------------------------------------------------------micro  S  Txi. O max  Txi TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi. etc. this Grade of Service is defined as the percentage of incoming circuit switched calls that are blocked due to lack of resources or timeslots. network dimensioning in Atoll is based on the principle that a voice or GSM call has priority over data transmission.TCH  – S  Txi.TCH_INNER   O max  Txi.7. 3. Following the common practice. This formula implies a loss system.TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi. As the load on the system increases. this information has been considered as a part of the network dimensioning process in this document. It represents the average throughput from the network point of view. the throughput is reduced as the number of active users increases. how efficiently the hardware resources are being utilized by the network. It may also depend on the RLC protocol efficiency. Networks with timeslot utilisation close to 100% are close to saturation and the end-user performance is likely to be very poor.1. and consequently. etc. But since this information is displayed in the network dimensioning results (only due to relevance). In GPRS. In Atoll this parameter is termed as the Load (Traffic load for circuit switched traffic and packet switched traffic load for packet switched traffic). Each network will have a different throughput probability distribution depending on the load and network configuration.Atoll 3. Instead of using the precise probability distributions. The throughput experienced by a user accessing a particular service can be calculated as: User throughput = Number of allocated timeslots x Timeslot capacity x Reduction Factor Or User throughput per allocated timeslot = Timeslot capacity x Reduction Factor Timeslot Capacity The timeslot capacity is the average throughput per fully utilized timeslot. Data Erlangs or data traffic is given by: Data Erlangs = L P  N S 172 . and hence each user.7. 3. The figure below shows how the peak throughput available per timeslot is reduced by interference and sharing.). a Temporary Block Flow (TBF) is initiated for transferring these packets.2. it is more practical to compute the average and percentile throughput values. number of timeslots available in the system (Ns) and the average system packet switched traffic load (Lp) (utilization of resources in the system). Timeslot Utilisation Timeslot utilization takes into account the average number of timeslots that are available for packet switched traffic. Packet traffic is intermittent and bursty. it is much more complicated to study than circuit switched traffic.1 Throughput Throughput is defined as the amount of data delivered to the Logical Link Control Layer in a given unit of time.2 Packet Switched Traffic Since packet switched traffic does not occupy an entire timeslot the whole time. Each temporary block flow (TBF). It is a measure of how much the network is loaded with data services. carrier-to-interference distribution. has an associated measured throughput sample in a given network. or in other words. It is described in more detail in the Network dimensioning steps section. Multiple TBFs can be multiplexed on the same timeslot. This reduction in the throughput is more perceivable when the system traffic load is high. The following parts describe the three most important Key Performance Indicators in GPRS/EDGE networks and how they are modelled in Atoll.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 3. This multiplexing of a number of packet switched service users over the same timeslots incurs a certain reduction in the throughput (data transfer rate) for each multiplexed user.3. It is a measure of how much data the network is able to transfer with 1 data Erlang. Reduction Factor Reduction factor takes into account the user throughput reduction due to timeslot sharing among many users.1. This implies that there can be many packet switched service users that have the same timeslots assigned for packet data transfer but at different intervals of time. Atoll computes the average timeslot capacity during the traffic analysis and is used to determine the minimum throughput reduction factor. It mainly depends on the network’s propagation conditions and criteria in the coverage area of a transmitter (carrier power. the resources are shared between the users being served.Reduction factor is a function of the number of timeslots assigned to a user (Nu).7. Whenever there is packet data to be transferred. This reduction in user perceived throughput is modelled through a reduction factor. 8: Reduction of Throughput per Timeslot More precisely. the probability function can be written as: n  LP  NP  ----------------------n! P  X = n  = --------------------------------------------------------------------------------------N P  LP  NP  i   LP  NP  ----------------------------- i – NP  + 1 N P!  N P +   --------------------i! i=0 i=N P if 0  n  N P i n  LP  NP  -----------------------------i – N  P N P!  N P P  X = n  = --------------------------------------------------------------------------------------N P  LP  NP  i   LP  NP  ----------------------------- i – NP  + 1 N P!  N P +   --------------------i! i=0 i = NP if n > N P i Hence the reduction factor can finally be written as: 173 . The formula for reduction factor can be derived following the same hypotheses followed by Erlang in the derivation of the blocking probability formulas (Erlang B and Erlang C). The throughput reduction factor is defined as:  P X= n  X  -------------------P X= 0 RF  n=0 Or. P(X=n) is the probability function of having n connections in the system. For example.3. Under the same assumptions as those of the Erlang formulas. a 24-timeslot system with each user assigned 3 timeslots per connection can be modelled by a single timeslot connection system with 8 timeslots in total.Atoll 3. Np models the equivalent timeslots that are available for the packet switched traffic in the system.if n > NP n n is the instantaneous number of connections in the system.  RF = PX= n   X  --------------------------n=0  P X= i i 0 Here. the reduction factor is a function of the ratio Ns/Nu (Np).0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Figure 3. Let X be a random variable that measures the reduction factor in a certain system state: 0 if n = 0 1 if 0 < n  NP X N -----P. TBF blocking and reduction factor. -------------N P! 1 – LP n! The default quality curves for the Reduction Factor have been derived using the above formula. implying a saturated system with no data throughput. hence it is difficult to model and is currently under study. X-axis) Each curve in the above figure represents an equivalent number of packet switched timeslots. and the minimum can be 0. As the delay is a function of the delays and the losses incurred at the packet level. The Maximum reduction factor can be 1.1. such as the size of the LLC PDU.Atoll 3.– ------------------.2 Delay Delay is the time required for an LLC PDU to be completely transferred from the SGSN to the MS. Hence. no default curve is presently available for delay in Atoll.3. the network parameters.+ i! i=1  i  LP  NP  -----------------------------i – N  P + 1 N P!  N P  i = NP This formula is not directly applicable in any software application due to the summations up to infinity. ln  1 – L P  + ----P  N P! n! n   n = 1 n = 1 RF = -----------------------------------------------------------------------------------------------------------------------N P  n=1 NP n LP  LP  NP   LP  NP  ----------------------. such as the packet queue length. or vice versa. become important.7.2.+ i! ©Forsk 2015 i N  LP  NP    -----P- -----------------------------i – N   i  P i=1 i = N + 1 N P!  N P P RF = ------------------------------------------------------------------------------------------------------N  P   i  LP  NP  --------------------. timeslot capacity. like throughput per cell. It is also quite dependent upon the radio access round trip time (RA RTT) and has a considerable impact on the application level performance viewed by the user. Atoll uses the following version of this formula that is exactly the same formula without the summation overflow problem. NP. NP N    NP + 1  P n n  NP L  LP  NP   ----------------------. The delay parameter is a user level parameter rather than being a network level quantity.9: Reduction Factor for Different Packet Switched Traffic Loads (Lp. and different protocol properties.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks NP  i  LP  NP  --------------------. 174 . 3. Each curve is for a fixed number of timeslots available for packet switched traffic (Np) describing the reduction factor at different values of packet switched traffic load (Lp). implying a maximum throughput.+ -------------------------. Figure 3. The figure below contains all the reduction factor quality curves in Atoll. The figure below contains all the blocking probability curves for packet switched traffic dimensioning in Atoll. there is no blocking as in circuit switched connections. the system starts to get more and more loaded. the request is queued in the system to be established later when resources become available. M being the maximum limit of multiplexing per timeslot (Multiplexing factor).+ i! i=0  i  LP  NP  ----------------------------- i – NP  + 1 N P!  N P  i = NP So. n  LP  NP  ----------------------------- i – NP  N P!  N P P  X = n  = --------------------------------------------------------------------------------------N P  i  LP  NP  --------------------. ------------N P! 1 – LP i! The above formula has been used to generate the default quality curves for blocking probability in Atoll.7.3 Blocking Probability In GPRS. This implies that if a new TBF is requested when there are already M * Np users active. Each curve represents an equivalent number of packet switched timeslots. P X= n for n =  M  N P  + 1 as in this case n is always greater than Np.3. hence there is higher probability of having a temporary block flow placed in a waiting queue. The blocking probability increases with the packet switched traffic load. The curves depict the blocking probabilities for different number of available connections (Np) at different packet switched traffic loads (Lp) for a fixed user multiplexing factor of 8. it will be blocked and placed in a queue. So the blocking probability is the probability of having M * Np + 1 users in the system or more.Atoll 3.1. Supposing that M number of users can be multiplexed over a single timeslot (PDCH). ------------ M  NP – NP  1 – L P N P!  N P BP = ---------------------------------------------------------------------------------N P  i=0 NP i LP  LP  NP   LP  NP  --------------------.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3. 175 . the Blocking Probability can be given as:    BP = n  LP  NP  ------------------------------ 1 – NP  N !  N n = MN +1 P P P P  X = n  = ---------------------------------------------------------------------------------------N  P n = MN+1  i  LP  NP  --------------------. we have. which implies that as the packet switched traffic increases for a given number of timeslots. These graphs are generated for a user multiplexing factor of 8 users per timeslot. meaning.2.+ i! i=0   i = NP i  LP  NP  ------------------------------ 1 – NP  + 1 N P!  N P Eliminating the summations to infinity. NP.+ -------------------------. If a new temporary block flow (TBF) establishment is requested and there are already M users per timeslot. we can have a maximum of M * Np users in the system. the blocking probability can be stated in a simpler form: M  NP  LP  NP  LP ----------------------------------------. delay and blocking probability). Then it calculates the number of timeslots to add in order to satisfy the demand of packet switched traffic.10: Blocking Probability for Different Packet Switched Traffic Loads (Lp. and • The blocking probability is less than the maximum allowable blocking probability defined in the service properties. This is performed using the quality curves entered in the dimensioning model used. the number of timeslots to be added is calculated such that: • The throughput reduction factor is greater than the minimum throughput reduction factor.7.7. the best method is to detail each process individually in form of steps of the global dimensioning process. GPRS and EDGE performance – Evolution towards 3G/UMTS.2. GSM. • Delay is less than the maximum permissible delay defined in the service properties. J. with outputs of a preceding one being the inputs to the next. Melero.11: Network Dimensioning Process 176 . J. If the dimensioning model has been indicated to take all three KPIs in to account (throughput reduction factor. As the whole dimensioning process is in fact a chain of small processes that have there respective inputs and outputs.2 Network Dimensioning Process The network dimensioning process is described below in detail. Halonen. 3. X-axis) Reference: T. Atoll first computes the number of timeslots required to accommodate the circuit switched traffic.1 Network Dimensioning Engine During the dimensioning process. The figure below depicts a simplified flowchart of the dimensioning engine in Atoll.3. Figure 3.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Figure 3. 3. Romero. John Wiley and Sons Ltd. OTarget. following are the inputs and outputs of the network dimensioning process: 3. and their quality curves. and computes the required number of timeslots to satisfy this demand using the Erlang B or Erlang C formula (as defined by the user). then the effective number of equivalent full-rate circuit switched timeslots TSeff. it is going to be taken as the Grade of Service required for that subcell instead of the maximum blocking rate of the service. TS reqC  k  TD C  ---------------k! k=0 Atoll considers the effect of half-rate circuit switched traffic by taking into account a user-defined percentage of half-rate traffic. Atoll takes the circuit switched traffic demand (Erlangs) either user-defined or calculated in the traffic analysis and assigned to the current subcell and the maximum blocking probability defined for the circuit switched service.Atoll 3. we have: TS reqC  TD C  GoS = -----------------------------------------------------------------------------------------------------------------------------------TS –1 reqC  TD C  TS reqC TD C  +  TS reqC !   1 – -------------. KPIs to consider.1 Step 1: Timeslots Required for CS Traffic Atoll computes the number of timeslots required to accommodate the circuit switched traffic assigned to each subcell. In fact. For Erlang B. If the number of timeslots required to accommodate the full-rate circuit switched traffic is TSreq. Outputs • • • • • • • Number of required TRXs per transmitter Number of required shared.7.2.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 On the whole.7. number of minimum dedicated packet switched timeslots per transmitter. For the blocking probability GoS and circuit switched traffic demand TDC. Atoll computes the effective equivalent number of full-rate timeslots that will be required to carry the total traffic with the defined percentage of half-rate traffic. maximum allowable blocking probability etc. and the percentage of half-rate traffic within the subcell is defined by HR.1 Inputs • • • • • • 3. maximum number of TRXs added for packet switched services. circuit switched and packet switched timeslots Traffic load Served circuit switched traffic Served packet switched traffic Effective rate of traffic overflow Actual KPI values: throughput reduction factor. FR. that can carry this traffic mix is calculated by: TS eff = TS reqFR   1 – HR -------  2 177 . 3. Atoll determines the required number of timeslots TSreq.2.2 Circuit switched traffic demand Packet switched traffic demand Timeslot configurations defined for each subcell Target traffic overflow rate and Half-rate traffic ratio for each subcell Service availability criteria: minimum required throughput per user.2. Details of the calculations of the parameters that are calculated during each step are described as well. Dimensioning model parameters: Maximum number of TRXs per transmitter. is greater than the maximum blocking rate defined in the services properties. C for each subcell using formulas described below.3.2.2.1. If the user-defined target rate of traffic overflow per subcell.1. dimensioning model for circuit switched traffic. maximum permissible delay. C value until the defined grade of service is reached.2 Network Dimensioning Steps This section describes the entire process step by step as it is actually performed in Atoll. delay and blocking probability 3. Atoll searches for TSreq. we have: TSreqC  TD C  --------------------------- TS reqC ! GoS = TS -----------------------------reqC  k  TD C  --------------k! k=0 For Erlang C.7.7. The total numbers of timeslots that carry circuit switched and packet switched traffic respectively are the sums of respective dedicated and shared timeslots: TS P = TS S + TS P dedicated and TS C = TS S + TS C dedicated 3.2.2. This is computed by dividing the total number of timeslots 178 . Delay and Blocking Probability is the equivalent number of available timeslots for packet switched traffic. This calculation is in fact performed in the traffic analysis process or is userdefined in the subcells table. those timeslots will not be considered capable of carrying circuit switched traffic and hence will not be allocated. It then calculates the effective traffic overflow rate.3 Step 3: Effective CS Blocking.2 Step 2: TRXs Required for CS Traffic and Dedicated PS Timeslots This stage of the network dimensioning process computes the number of TRXs required to carry the circuit switched traffic demand through the number of required timeslots calculated above and the timeslot configuration defined by the user in the network settings. Knowing the traffic demand per pixel of the covered area in terms of kbps and the maximum attainable throughput per pixel (according to the C and/or C/I conditions and the coding scheme curves in the GPRS/EDGE configuration). NP. First of all. following the principle that shared timeslots are potential carriers of both traffic types. the circuit switched traffic is supposed to be placed in an infinitelength waiting queue. the timeslots required for packet switched traffic and their respective distributions according to the timeslot configurations being known. Atoll computes the number of TRXs to be added to carry the packet switched traffic demand. From this data. These timeslots can be dedicated packet switched timeslots and the shared ones. TS P = TS S + TS P dedicated TS C = TS S + TS C dedicated The packet switched traffic load is calculated by the formula:  ST C – TS C dedicated + TD P  Timeslots L P = -----------------------------------------------------------------------------------TS P The second important parameter for the calculation of Reduction Factor. For example. This is the number of TRXs that contain dedicated packet switched and shared timeslots.Atoll 3.7.2. if 4 timeslots have been marked as packet switched timeslots in the first TRX and Atoll computes 8 timeslots for carrying a certain circuit switched traffic demand. then the number of TRXs to be allocated cannot be 1 even if there is no packet switched traffic considered yet. So. and the circuit switched traffic demand of TDC. Effective CS Traffic Overflow and Served CS Traffic In this step. This is performed by using the Erlang B or Erlang C formula with the circuit switched traffic demand and the number of required TRXs as inputs and computing the Grade of Service (or blocking probability). Hence. Atoll calculates the number of timeslots available for carrying the packet switched traffic demand.e.. for an effective traffic overflow rate of Oeff. While in case of the Erlang C model. This is the difference of the circuit switched traffic demand and the percentage of traffic that overflows from the subcell to other subcells calculated above. Oeff.3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Atoll employs this simplified approach to integrating half-rate circuit switched traffic. Atoll calculates the average traffic demand in packet switched timeslots by: TD P Timeslots =  Traffic demand per pixel (kbps)-------------------------------------------------------------------------Throughput per pixel (kbps) pixel The average timeslot capacity of a transmitter is calculated by dividing the packet switched traffic demand over the entire coverage area (in kbps) by the packet switched traffic demand in timeslots calculated above. the effective rate of traffic overflow for the circuit switched traffic is the same as the circuit switched blocking rate.4 Step 4: TRXs to Add for PS Traffic This step is the core of the dimensioning process for packet switched services.2. i. With the number of timeslots required to serve the circuit switched traffic. In case of Erlang B formula. it also computes the served circuit switched traffic. This implies that there is no overflow in this case.7.2. Atoll calculates the equivalent average packet switched traffic demand in timeslots by studying each pixel covered by the transmitter. To determine this number of TRXs. 3. which provides approximately the same results as obtained by using the half-rate traffic charts. the previously calculated number of required TRXs is used to compute the effective blocking rate for the circuit switched traffic. If a timeslot configuration defines a certain number of dedicated packet switched timeslots pre-allocated in certain TRXs.7.2. Atoll distributes the number of required circuit switched timeslots calculated in Step 1 taking into account the presence of dedicated packet switched timeslots in each TRX according to the timeslot configurations. the served circuit switched traffic STC is computed as: ST C = TD C   1 – O eff  3. depending on the user-defined traffic analysis RF conditions criteria. It then recalculates the packet switched traffic load. The throughput per timeslot per pixel TPTS. To calculate the minimum throughput reduction factor for the transmitter. If dimensioning is not based on a traffic analysis. and the equivalent number of available timeslots. the total number of downlink timeslots defined in the properties of the mobile terminal (See TS Max TerminalType defintion above) and the required availability defined in the service properties. The number timeslots that a terminal can use in packet switched calls is the product of the number of available DL timeslots for packet-switched services (on a frame) and the number of simultaneous carriers (in case of EDGE evolution). the min(X. TP TS Pixel = f  C  Or C TP TS Pixel = f  C  and TP TS Pixel = f  ---  i The required availability parameter defines the percentage of pixels within the coverage area of the transmitter that must satisfy the minimum throughput condition. The calculated values of all the KPIs are compared with the ones defined in the service properties. LP. It is at the stage of calculating the average timeslot capacity per transmitter that Atoll studies each covered pixel for carrier power or carrier-to-interference ratio. then Atoll increases the number of TRXs calculated for carrying packet switched traffic by 1 (each increment adding 8 more timeslots for carrying packet switched traffic as the least unit that can be physically added or removed is a TRX) and resumes the computations from Step 3. Atoll deduces the maximum throughput available on that pixel through the throughput vs. C/I curves of the GPRS/EDGE configuration. NP. it means that the dimensioning process has acceptable results. Atoll computes the minimum throughput reduction factor for each pixel using the formula: 179 .Atoll 3. TSTerminal is the number of timeslots that a terminal will use in packet switched calls. C or throughput vs. but since it is relevant to the dimensioning procedure. According to the measured carrier power or carrier-to-interference ratio. Otherwise. LP.Y) function yields the lower value among X and Y as result.3. TS Terminal = min  TS Max Service TS Max TerminalType  and TS Max TerminalType = TS DL TerminalType  Carriers DL TerminalType Here. This particular part of this step can be iterative if the KPIs to consider in dimensioning are not satisfied in the first try. So. Atoll performs another iteration to find the best possible results. If these KPIs are not satisfied. The number of timeslots that a terminal will use in packet switched calls is determined by taking the lower of the maximum number of timeslots on a carrier for packet switched service defined in the service properties and the maximum number of timeslots that a mobile terminal can use for packet switched services (see above) on acarrier. This parameter renders user-manageable flexibility to the throughput requirement constraint. and the equivalent number of available timeslots. the average throughput per timeslot deduced from the throughput curves stored in the GPRS/EDGE configuration properties for each coding scheme. knowing the packet switched traffic load. Pixel can be either a function of carrier power C. The values for maximum Delay and Blocking probability are defined directly in the properties but the minimum throughput reduction factor is calculated by Atoll using the user’s inputs: minimum throughput per user and required availability. Atoll finds out the KPIs that have been selected before launching the dimensioning process using the quality curves stored in the dimensioning model. Minimum Throughput Reduction Factor Calculation The minimum throughput reduction factor is computed using the input data: minimum required throughput per user defined in the service properties. This calculation is in fact performed during the traffic analysis process. it is displayed in a column in the dimensioning results so that the user can easily compare the minimum requirement on the reduction factor KPI with the resulting one. Therefore. the minimum throughput reduction factor is a user-defined parameter. NP is calculated at this stage as: TS P N P = ----------------------TS Terminal . or carrier power C and the carrierto-interference ratio C/I.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 available for carrying packet switched traffic by the number of downlink timeslots defined in the mobile terminal properties. If the KPIs calculated above are within acceptable limits as defined by the user. NP. Then it recomputes the KPIs with these new values of LP and NP. Now. If the KPIs are within satisfactory limits the results are considered to be acceptable. in turn.e. This calculation is performed for each service type available in the subcell coverage area. This implies that the required minimum throughput for the given service will be available at 90% of the pixels covered. i. Atoll computes the minimum reduction factor at each pixel using the formula mentioned above.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 TP user min RF min Pixel = ---------------------------------------------------TP TS Pixel  TS Terminal Once the minimum reduction factor for each pixel is known.3.2.2. Atoll always finds the best possible answer in terms of number of timeslots required to carry the packet switched traffic demand unless the requirement exceeds the maximum limit on the number of the packet switched traffic timeslots defined in the dimensioning model properties.36 57 0.2. These will be the pixels that do not provide satisfactory service. The minimum throughput reduction factor RFmin value is a minimum requirement that must be fulfilled by the network dimensioning process when the Reduction Factor KPI is selected in the dimensioning model. if the actual reduction factor in that subcell becomes less than a minimum required. there is no packet traffic overflow unless the packet switched traffic demand requires more TRXs than the maximum allowed 3. the total traffic load. 945 pixels.98. the corresponding RFmin will be the one provided at least 90% of the pixels covered. implies that there will be a certain limit on the reduction factor. Example: Let the total number of pixels. This.5 Step 5: Served PS Traffic Atoll calculates the served packet switched traffic using the number of timeslots available to carry the packet switched traffic demand. i. Hence. The total traffic load L is calculated as: 180 . The corresponding value of the resulting RFmin in this example hence turns out to be 0. As the result of the above iterative step.Atoll 3.6 200 0. The final minimum throughput reduction factor is the highest one amongst all calculated for each service separately.3 189 0.5 20 0.7. the service will not be satisfactory. covered by a subcell S.12: Minimum Throughput Reduction Factor 3.e.9 23 0.98 87 So for a reliability level of 90%. The reliability level set to 90%. The following example may help in understanding the concept and calculation method. be 1050. Only 87 of the covered pixels imply an RFmin of 0.6 Step 6: Total Traffic Load This step calculates the final result of the dimensioning process. Figure 3. since this value covers 962 pixels in total.72 473 0.7.2.e. Atoll calculates the global minimum reduction factor that is satisfied by the percentage of covered pixels defined in the required availability.9. i. and outputs the following results: RFmin Number of pixels 0. the respective numbers of shared and dedicated timeslots and the circuit switched and packet switched traffic demands. The concept of this computation is the inverse of that of the dimensioning process. TDC. TSC. • • • TSC. OC. under the current conditions of circuit switched traffic demand.3. dedicated is the number of timeslots dedicated to the circuit switched traffic. the blocking probability can be easily computed using the Erlang formulas or tables.1. The computation algorithm utilizes the parameters set in the dimensioning model properties and the quality curves for the throughput reduction factor. TSS is the number of shared timeslots for a transmitter. dedicated is the number of dedicated packet switched timeslots TSS is the number of shared timeslots 3. TDC. and the number of timeslots available for circuit switched traffic. • • • • • STC is the served circuit switched traffic STP is the served packet switched traffic TSC.1. delay and the blocking probability. Also knowing the circuit switched traffic demand. It can be used to evaluate an already dimensioned network in which recent traffic changes have been made in limited regions to infer the possible problematic areas and then to improve the network dimensioning with respect to these changes. dedicated is the number of timeslots dedicated to the packet switched traffic.8. is the same as the percentage of traffic blocked by the subcell calculated above. and the number of timeslots available for the circuit switched traffic. the circuit switched traffic overflow rate. Atoll has the results of the dimensioning process already committed and known. 3. is given by: TS P = TS S + TS P dedicated 3. TSC. TSC. the percentage of blocked circuit switched traffic can be computed through: TS C  TD C  -------------------- TS C ! % of blocked traffic = -------------------------TS C  k  TD C  ---------------k! k=0 In a network dimensioning based on Erlang B model. TSP. Then.1 Erlang B Under the current conditions of circuit switched traffic demand.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 ST C + ST P L = --------------------------------------------------------------------------------TS C dedicated + TS P dedicated + TS S . TSP.8. the number of timeslots available for the circuit switched traffic.Atoll 3.8 Key Performance Indicators Calculation This feature calculates the current values for all circuit switched and packet switched Key Performance Indicators as a measure of the current performance of the network. is defined as: TS C = TS S + TS C dedicated And the number of timeslots available for the packet switched traffic.2 Erlang C Similarly. Atoll has already calculated the effective traffic overflow rate and the blocking rate during the dimensioning process. the percentage of delayed circuit switched traffic can be computed through: 181 . TDC. dedicated is the number of dedicated circuit switched timeslots TSP. In this case. The following conventional relations apply: If. TSC. and the number of timeslots available for the circuit switched traffic. Atoll then computes the current values for all the KPIs knowing the number of required TRXs.8.1 Circuit Switched Traffic For each subcell. 3. 4 Delay Again for a 100% loaded or saturated subcell. This implies that the system will be loaded to the maximum and even saturated. OC.1. 182 .2.8.1 Traffic Load The traffic load will be 100%.2. TDC. there will be packet switched data traffic that will be rejected by the subcell as it will not be able to accommodate it. as the subcell will have more traffic to carry than it can. 3. Hence the user level quality of service is bound to be very unsatisfactory. This overflow rate is calculated as show below:   TS C dedicated + TS P dedicated + TS S  – ST C  O P = 1 – ----------------------------------------------------------------------------------------------------------.Atoll 3.2.1.2 Packet Switched Traffic Identifying the total traffic demand. 3. If. This circuit switched traffic overflow rate.1. Hence there will be a perceptible packet switched traffic overflow in this subcell. is calculated as: ST C = TD C   1 – O C  3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 TS C  TD C  % of traffic delayed = ---------------------------------------------------------------------------------------------------------TS – 1 C  TD C  TS C TD +  TS C !   1 – --------C-  TS C  k  TD C  --------------k! k=0 If the circuit switched traffic demand.8. TDC.8. Hence.2.8. 100 TD P 3.1. is less than the number of timeslots available for the circuit switched traffic. Hence.1.3 Served Circuit Switched Traffic The result of the above two processes will be a traffic overflow rate for the circuit switched traffic for each subcell. The following results are expected in this case: 3. 3. the circuit switched traffic demand. STC. is higher than the number of timeslots available to accommodate circuit switched traffic.2 Packet Switched Traffic Overflow In a 100% loaded. is calculated as: TD C – TS C O C = ----------------------TD C 3. the column for this result will be empty signifying that there is a percentage of circuit switched traffic actually being rejected rather than just being delayed under the principle of Erlang C model.1 Case 1: Total Traffic Demand > Dedicated + Shared Timeslots In the case the total number of timeslots available is less than the total traffic demand. (circuit switched traffic demand + packet switched traffic demand) as: TD T = TD C + TD P The following two cases can be considered. the throughput perceived by the packet switched service user will be 0. OC. TDT. on the other hand. TSC. or even saturated subcell. OP. will be 0 if the circuit switched traffic demand. the packet switched data calls will start being rejected because of shortage of available resources.8. 3. OC. The circuit switched traffic overflow rate. the blocking probability will be 100%.3.8.2.3 Throughput Reduction Factor The resulting throughput reduction factor for a 100% loaded or saturated subcell will be 0. then there will be a certain percentage of circuit switched traffic that will overflow from the subcell. The served circuit switched traffic.8.8.5 Blocking Probability All the data packets will be rejected by the system since it is saturated and has no free resources to allocate to incoming data packets. the delay at the packet switched service user end will be infinite as there is no data transfer (throughput = 0).2. is higher than the number of timeslots available to carry the circuit switched traffic. TSC.1. TDC. implying a very bad quality of service. 8.1. the served packet switched traffic will be the same as the packet switched traffic demand: ST P = TD P 183 .2.4 Delay The resulting delay the subcell is calculated through the delay quality curve for given packet switched traffic load.3 Throughput Reduction Factor The resulting throughput reduction factor for a normally loaded subcell is calculated through the throughput reduction factor quality curve for given packet switched traffic load. and number of equivalent timeslots.2.2. NP. LP.8.2.8. 3.8.1 Traffic Load The traffic load is computed knowing the total traffic demand and the total number of timeslots available to carry the entire traffic demand: TD T Traffic Load = --------------------------------------------------------------------------------TS C dedicated + TS P dedicated + TS S 3.8. no packet switched data calls will be rejected. 3.6 Served Packet Switched Traffic With the packet switched data traffic overflowing from the subcell.5 Blocking Probability The resulting blocking probability for a normally loaded subcell is calculated through the blocking probability quality curve for given packet switched traffic load. and number of equivalent timeslots. be 0. The packet switched traffic overflow will.8.8.2 Packet Switched Traffic Overflow In a normally loaded subcell. which is calculated in the same manner as in the dimensioning process as well: TS P N P = ----------------------TS Terminal These parameters calculated.6 Served Packet Switched Traffic As there is no overflow of the packet switched traffic demand from the subcell under consideration. the subcell will not be saturated and there will be some deducible values for all the data KPIs. NP.2.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3. there will be a within limits packet switched traffic load. STP. NP. 3. and number of equivalent timeslots. calculated as under:  ST C – TS C dedicated + TD P  Timeslots L P = -----------------------------------------------------------------------------------TS P The second parameter for computing the KPIs from the quality curves of the dimensioning model is the number of equivalent timeslots available for the packet switched data traffic. This is due to the fact that the packet switched data traffic is rather placed in a waiting queue than be rejected. LP.2. there will be a part of that traffic that is not served. Therefore. The served packet switched data traffic. LP. 3. In a normally loaded subcell. the packet switched data traffic will have no overflow percentage.2 Case 2: Total Traffic Demand < Dedicated + Shared Timeslots In the case the total traffic demand is less than the number of timeslots available to carry the traffic.2.2. 3. now Atoll can compute the required KPIs through their respective quality curves. therefore.2.8.2.2.2. NP.3. LP.2. is calculated on the same principle as the served circuit switched traffic: ST P = TD P   1 – O P  3.2. a mobility type. The specific simulation process in GSM consists of the following steps: 184 . Both active and inactive users consume radio resources and create interference. Obtaining a realistic user distribution: Atoll generates a user distribution using a Monte Carlo algorithm. A user can be either active or inactive.13 shows the GSM simulation algorithm. Finally.3. Each user is assigned a service. the following are explained: • • "MSA (Mobile Station Allocation) Definition" on page 131 "GSM Simulation Process" on page 184. and an activity status by random trial. The user activity status is an important output of the random trial and has direct consequences on the next step of the simulation and on the network interferences. The simulation process consists of two steps: 1.9. another random trial determines user positions in their respective traffic zone (possibly according to the clutter weighting and the indoor ratio per clutter class).Atoll 3.9. according to a probability law that uses the traffic database. you can create simulations. this user distribution is based on the traffic database and traffic maps and is weighted by a Poisson distribution between simulations of the same group.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 3. Modelling network regulation mechanisms: Atoll manages the GSM resources as described in "Radio Resource Management in GSM" on page 184 3.1 GSM Simulation Process Figure 3. 3.9 Simulations Once you have modelled the network services and users and have created traffic maps. 2.1.1 Radio Resource Management in GSM In this section. b.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Figure 3.3. 1. Determines the downlink C/(I+N) for each of these mobiles as explained in "DL Carrier-to-Interference Ratio Calculation" on page 131. Determines the uplink C/(I+N) for each of these mobiles as explained "UL C/I Evaluation" on page 156 d. 185 . Determines MSA codec modes in uplink as explained in "Codec Mode Selection" on page 159. Determines MSA codec modes in downlink as explained in "Calculations Based on C/(I+N)" on page 148 part of "CQI Calculation Without Ideal Link Adaptation" on page 147. Cell traffic loads for each MSA and transmitter are set to their average current value in the Transmitters table (one traffic load value per subcell). e. For each circuit-switched mobile a. Mobile transmission power is set to the maximum mobile power. the simulation process. f. Determines the server and the MSA to which the circuit-switched mobile is attached. It sets initial values for the following parameters: a. c.Atoll 3. Performs the corresponding target power controls. b. the simulation process 2.13: GSM simulation algorithm For each simulation. For each iteration k. "DL Power Control Gain Management" on page 189 and "DTX DL Gain Management" on page 190. k = Max Number of Iterations defined Subcell Subcell i k Subcell i Subcell i TL UL – GSM  TL UL – GSM k 186 TRX i Req Subcell i when creating the simulation. "Half-Rate Traffic Ratio Management" on page 189. See detailed information in "Servers Selection" on page 187 and "Coding Scheme Assignment. c. Throughput Evaluation and DL Power Control" on page 188. Determines MSA coding scheme in uplink as explained in "Coding Scheme Selection" on page 158. Performs the corresponding target power controls. Performs the convergence test to see whether the differences between the current and the new loads are within the convergence thresholds. if: TL DL – GSM  TL DL – GSM Req TRX i OR NR UL – GSM  NR UL – GSM k Req .e. Convergence: Simulation has converged between iteration k . Resources and throughputs are finally assigned to each packet-switched user. PCG DL – GSM  k – 1 Subcell i Req  k – 1 k – 1 i Subcell i If TL DL – GSM Subcell i = k Subcell  k – 1 i i  PCG DL – GSM k – PCG DL – GSM  All Subcell = k TRX i NR UL – GSM Max i Subcell i PCG DL – GSM Subcell i i  TL – TL DL – GSM  DL – GSM k All Subcell = . i. and can be written as follows: Subcell i TL DL – GSM k Subcell i TL UL – GSM k Subcell Max Subcell Subcell i i  TL – TL UL – GSM  UL – GSM k All Subcell Max = TRX  TRX i i Max NR UL – GSM – NRUL – GSM k All TRX i Subcell i Req . 5. 4.3. The convergence criteria are evaluated at the end of each iteration k. b. For each packet-switched mobile a. g. TL UL – GSM TRX i Req and NR UL – GSM k are the simulation convergence thresholds defined when creating the simulation. Subcell i i OR PCGDL – GSM  PCG DL – GSM k Req OR . DL power control gains and DTX gains of all the subcells according to the resources in use and the total resources. It updates the UL traffic loads of all the subcells and the UL noise rises of all the TRXs according to the resources in use and the total resources. The number of timeslots in DL and UL are obviously not linked. Half-Rate traffic ratios. Atoll stops the simulation in the following cases. f. It updates the traffic loads. Determines the server and the MSA to which the packet-switched mobile is attached. Determines the downlink C/(I+N) for each of these mobiles as explained in "DL Carrier-to-Interference Ratio Calculation" on page 131. 7. See detailed information in "Codec Mode Assignment and DL Power Control" on page 187.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 See detailed information in "Servers Selection" on page 187 and "Codec Mode Assignment and DL Power Control" on page 187. It equally shares the remaining resources to packet-switched users who did not reach their maximum throughput demands. Determines the uplink C/(I+N) for each of these mobiles as explained "UL C/I Evaluation" on page 156 d.1 and k if: Subcell i Subcell i TL DL – GSM  TL DL – GSM k TRX i Subcell i Req Subcell i AND PCG DL – GSM  PCG DL – GSM k Subcell i Req Subcell i AND TL UL – GSM  TL UL – GSM k Req AND TRX i NRUL – GSM  NR UL – GSM k Req No convergence: Simulation has not converged even after the last iteration. 6. e. Evaluates the number of necessary timeslots to reach the minimum downlink and uplink throughput demands (defined in the requested service) of the users randomly ranked.Atoll 3. Determines MSA coding scheme in downlink as explained in "Calculations Based on C/(I+N)" on page 137 part of "Coding Scheme Selection and Throughput Calculation Without Ideal Link Adaptation" on page 137.. 3. See detailed information in "Subcell Traffic Loads Management" on page 189. the user will use either an entire timeslot (any AMR. several MSAs are candidate.2 Servers Selection For a given network. For each candidate MSA. the resource element assigned to a mobile station is an MSA. the algorithm will try to attach this mobile to a lower priority layer until no solution can be found. to optimise the resource allocation. depending on the computed radio conditions at the server location. EFR or FR codec mode) or half a timeslot (HR codec mode). and then to serve packet services with the remaining resources.) for the iteration k+1 using the new calculated loads as the current loads until convergence. 3.Atoll 3. Finally the frequency band(s) in use in the transmitter have to be supported by the user terminal. as a result. In none of these conditions are fulfilled. the level of quality at the mobile might be different. the mobile selects the highest priority transmitter. as a consequence. The maximum allowed power reduction is set at 30 dB by default. even after convergence. For each mobile list. for a given mobile distribution. 187 . On the very first iteration of the simulation.9. a mobile might be served by a transmitter if its mobility (as assigned by Atoll at the beginning of the simulation) does not exceed the maximum speed permitted on that layer. if the mobile is regularly rejected from the highest priority transmitter. each server being on a different HCS layer. For each mobile. a codec mode study is run. Due to the radio conditions. When serving a circuit-switched user. For a given user.3 Codec Mode Assignment and DL Power Control Two types of services can be assigned to users: circuit-switched and packet-switched ones. within his MSA list. the service areas of each transmitter are evaluated in the same way than an HCS server study with 0 dB margin. For any other TRX type. This mechanism is then reproduced for all the users requesting a circuit-switched service. The network has been set up and dimensioned in order to first serve circuit services. For each MSA. no traffic is attached to MSAs.3. the MSA having currently the lowest load is selected and. Their load starts from 0 and is increased as traffic increases and mobiles are attached to them. no power k control is applied. it will select the second highest priority transmitter and so on. During the iterative process. In addition. and using k the victim max power. Atoll sorts the potential servers according to their HCS layer priority in decreasing order. If these conditions are fulfilled. If  C  I  Max   C  I  Target . a codec mode is assigned to a user.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 8. and consequently. In other words. Atoll selects the MSAs the most optimised in term of timeslot occupancy. Atoll evaluates the minimum required power P Min in order to reduce the quality i at the user’s terminal to  C  I  Target for the assigned MSA k. each pixel. Depending on the assigned MSA. a codec mode costing half a timeslot might be chosen instead of a codec mode costing an entire timeslot in the case the MSA with the lowest cost would have been chosen. Assuming a server is selected for each mobile. Repeats the above steps (from step 2. 3. the list of candidate codec modes is saved. if the mobile is rejected from a layer. the mobile is rejected with the condition "No Coverage". assuming the received signal strength is greater than the reception threshold defined on that layer. At the beginning of a simulation iteration. Depending on this codec mode. is covered by the best server of each HCS layer. As explained in "GSM Simulation Process" on page 184. until convergence. i For each MSA k.1. each mobile then has a list of potential servers. When MSAs are almost full. The power control is considered achieved when the final C/(I+N) is at less than 1 dB from i the  C  I  Target . the assigned codec mode i corresponds to a quality target:  C  I  Target . its served codec mode so as the required number of timeslots. no codec mode can be served and the mobile is rejected with the condition "No Service". In addition to the coverage condition above. As an example. the corresponding codec mode is assigned to the mobile.9. a  C  I  Max is obtained. The effect of this mechanism results in a load balancing of MSAs within a transmitter. the load of this MSA is now increased. a codec mode is obtained.1. This means that the power cannot be reduced by more than 30 dB from the initial to final C/(I+N). using the computed C/(I+N) and based on the user terminal and mobility (See "Calculations Based on C/(I+N)" on page 148 for more information). k i k i If  C  I  Max   C  I  Target . If the MSA is on the BCCH. after power control. j For each MSA k. each packet-switched service has a coding scheme and. no coding scheme can be served and the mobile is rejected with the condition "No Service". If this MSA is defined over a TCH subcell.1. the user is rejected with the condition "Resource Saturation". In order to reach their maximum throughput demand. from which we get a throughput per timeslot. As a consequence. 0. the resource element assigned to a mobile station is an MSA. This mechanism of equally share of remaining resources is then applied for all the connected packet-switched service users over all their MSAs. For each candidate MSA.4 TS have been used. Considering the minimum DL throughput demand for the service.8 TS per user. its capacity is 8 TS. Then. 4. let’s imagine than a MSA is already occupied as follows: • • 2 TS for circuit-switched service users (3 users: 2 HR codec modes + 1 FR codec mode) 2. Throughput Evaluation and DL Power Control After having served the circuit traffic over one iteration.8 TS remain unused for that MSA.5 or 1) and a corresponding minimum required power to get the  C  I  Target of the served MSA. If. a coding scheme is assigned to a user and a throughput per timeslot is obtained. For each MSA. 3. a corresponding number of timeslots i (0.1 TS in order to get his maximum demand.4 Coding Scheme Assignment. the MSA having currently the lowest load is selected and. for each mobile. Depending on the assigned MSA. 188 . and 3. and consequently. the algorithm now tries to serve packet-switched traffic. the timeslot Assignment is realised in two steps.Atoll 3. Then. using remaining resources (timeslots). If a user cannot get its minimum throughput demand for insufficient number of available timeslots. When serving a packet-switched user. In the same way than for circuit traffic. one can estimate how many timeslots are needed to get that throughput on each MSA. As explained in "MSA (Mobile Station Allocation) Definition" on page 131. Atoll tries to allocate more throughput up to the maximum throughput demand of the service. at this step. if the user has been dropped as active at the beginning of the simulation. Then some timeslots are assigned to each packet-switched service user in order to obtain a throughput between the min and the max DL throughput demand per user defined in the considered service properties. The two packet-switched users have obtained their minimum throughput demand. k j If  C  I  Max   C  I  Target .3. the remaining 0.9. In other words. the level of quality at the mobile might be different. and k using the victim max power. Atoll tries to allocate the minimum throughput demand of the service. a coding scheme is obtained. As an example. Then. Finally. This mechanism is then reproduced for all the users requesting a packet-switched service. the effect of this mechanism results in a load balancing of MSAs within a transmitter. depending on the computed radio conditions at the server location. if the user has been dropped as inactive at the beginning of the simulation. a  C  I  Max is obtained. the maximum of timeslots the user can benefit is the minimum between the number of DL timeslots defined in the selected terminal and service. is supposed to be served his DL minimum throughput demand. If the first user can get his maximum throughput demand with only 1. For packet-switched traffic. its served coding scheme so as the required number of timeslots to get a certain throughput demand.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 To summarise. Atoll only keeps the MSAs for which this number of timeslots is lower than the number of timeslots supported (see above) and for which there is enough remaining timeslots. the list of candidate coding schemes is saved. For a given user. As explained in "Packet Throughput and Quality Analysis: Application Throughput (kbps)" on page 144. Due to the radio conditions. within his MSA list. In the first step. this second user could benefit of 2. the assigned coding scheme j corresponds to a quality target:  C  I  Target .3 TS will be able to be used by the user. ideally. his corresponding number of timeslots is consumed but no DL power is considered for this specific user. a codec mode. Inactive users only participate in the timeslot management but do not affect DL power. finally. At this step. the remaining TS are equally shared between the two connected users: 1. Assuming a server is selected for each mobile.3 TS to get this demand.4 TS for packet-switched service users after the first step (2 users).6 TS remain. The second step of resources allocation for packet-switched traffic is to share the remaining resources between connected users in order they get their maximum throughput demand. as a consequence. a coding scheme study is run. each circuit-switched user is assigned a MSA. In the second step. using the computed C/(I+N) and based on the user terminal and mobility (See "Calculations Based on C/(I+N)" on page 139 for more information). the load of this MSA is now increased. he only needs 1. both timeslots and powers have to be considered to make him connected.5 TS. several MSAs are candidate. Considering that subcell loads are values which are unique per traffic pool (e. a corresponding number of j timeslots (which might not be an integer value) and a corresponding minimum required power to get the  C  I  Target of the served MSA. each packet-switched user is assigned a MSA. This means that the power cannot be reduced by more than 30 dB from the initial to final C/(I+N). the number of timeslots necessary to connect the traffic have to be summed up over the several MSAs over a same traffic pool.5 Subcell Traffic Loads Management When circuit-switched and packet-switched traffic have been served or rejected.1. 3.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 k j If  C  I  Max   C  I  Target . This value referring to voice traffic only.where the number of UL TS available for a BCCH subcell is 7 and 8 for any other subcell. TS DL available  MSA TP i The DL traffic load value is then assigned to all the subcells of a same traffic pool.  HR users MSA TP  i HR RATIO TP = ------------------------------------. i users MSATP  MSA TP i i The Half-Rate traffic ratio is then assigned to all the subcells of a same traffic pool. The maximum allowed power reduction is set at 30 dB by default. For the traffic pool TP i . The first parameter to be updated are the subcell DL and UL traffic loads. no power k control is applied. 189 . Atoll evaluates the minimum required power P Min in order to reduce the quality j at the user’s terminal to  C  I  Target for the assigned MSA k. This value is used to calculate the number of timeslots required to respond to the voice traffic demand and is evaluated per traffic pool. To summarise.g. TS UL available  MSA TP i The UL traffic load value is then assigned to all the subcells of a same traffic pool. a coding scheme. For any other TRX type. the corresponding coding scheme is assigned to the mobile.9.where the number of DL TS available for a BCCH subcell is 7 and 8 for any other subcell.3. at this step. The power control is considered achieved when the final C/(I+N) is at less than 1 dB from i the  C  I  Target . Atoll performs an update on several parameters.6 Half-Rate Traffic Ratio Management The second parameter at the end of an iteration is the Half-rate traffic ratio. This is the percentage of half-rate voice traffic in the subcell. the subcell UL traffic load is computed as follows:  MSA TL TP i UL TS UL used TP i = ------------------------------------------------. If the MSA is on the BCCH.1. 3. after power control.9. BCCH and TCH subcells belong to the same traffic pool because they are in charge of the same traffic area) in DL and in UL. circuit-switched users only are taken into account in its evaluation. For the traffic pool TP i . users represents HR and FR circuit-switched service users. the subcell DL traffic load is computed as follows:  TL TP TS DL used MSA TP i DL i = ------------------------------------------------. The DTX gain models the fact that inactive circuit-switched users. due to active and inactive users. Considering only the connected traffic at the end of the GSM part of the simulation process. even if they are connected to the network.(in dB).8 DTX DL Gain Management A certain gain representing inactive circuit-switched service users has also to be evaluated.= P Moy TS i  S active where i active are the circuit-switched active mobiles connected to the subcell S.1. an active user can be connected in DL if: • • • he has a serving cell assigned. In "DL Power Control Gain Management" on page 189. If a user is rejected during server determination. do not produce the same level of interference than active circuit-switched users.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 3.1.1. It is essential to note that there is no power control on the BCCH and. the cause of rejection is "Resource Saturation.e. where P Max P Moy S is the max power of the considered subcell.Atoll 3. all of the cell’s resources were used up by other users. the main results obtained are: • At the subcell level • • • 190 Subcell traffic loads (UL and DL) DL Power control gains DTX gains . where P Moy P Moy S active S is average requested power defined in "DL Power Control Gain Management" on page 189 above.9. If a user is rejected because quality is too low to obtain any codec mode or coding scheme. he has a codec mode (resp. over active i active its MSAs P Moy S The ratio -------------------------.9 GSM Simulation Results At the end of the simulations. represents the DTX gain. consequently.. which can be assigned to the subcell. For a circuit-switched (resp. the mean DL power control gain concerns both active and inactive users. over its MSAs i The ratio PCG S P Max = ---------------S (in dB). which can be assigned to the subcell. From the minimum required powers evaluated at the end of "Codec Mode Assignment and DL Power Control" on page 187 in order to get the appropriate codec modes without any excess of unneeded power. If a user is rejected because he cannot be allocated a sufficient number of resources to obtain its codec mode or coding scheme. the cause of rejection is "No Service". coding scheme) corresponding to his activity status.9. the cause of rejection is "No Coverage". an average minimum required power is obtained for each mobile connected to the subcell S as follows: k  PMin  TSi i i-------------------------------S -  = P Moy TS i S where i are the mobiles connected to the subcell S. represents the mean power S control gain." i. Throughput Evaluation and DL Power Control" on page 188 in order to get respectively the appropriate codec modes and coding schemes without any excess of unneeded power. 3.7 DL Power Control Gain Management At the end of each iteration. he is not rejected due to resource saturation. 3. the subcell DL power control gain is evaluated by taking into account all the connected users: • active and inactive circuit-switched service users (assuming each inactive user does not cost any DL power but only some timeslots) all packet users • From the minimum required powers evaluated at the end of "Codec Mode Assignment and DL Power Control" on page 187 and "Coding Scheme Assignment. due to circuit-switched active users. an average minimum required power is obtained for each circuit-switched active mobile connected to the subcell S as follows:  ki active P Min  TS i i active  S active -----------------------------------------------------------.9.3. the mean power control gain on the BCCH is 0. packet-switched) service. 0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 • • Half-rate traffic ratios At the TRX level • intra-technology UL noise rises Subcell traffic loads. which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. They must fulfil the following conditions: • • • • They are active. • The calculation options: • Force co-site transmitters as neighbours: This option enables you to force transmitters located on the reference transmitter site in the candidate neighbour list. 3. see "Appendix: Calculation of the InterTransmitter Distance" on page 195.Atoll 3. 3. They are located inside the focus zone. It means that all the TBC transmitters of the . Atoll takes into account the computation zone. • Force adjacent transmitters as neighbours: This option enables you to force transmitters geographically adjacent to the reference transmitter in the candidate neighbour list. This folder can be either the Transmitters folder or a group of transmitters or a single transmitter.1 Neighbour Allocation for All Transmitters We assume a reference transmitter A and a candidate neighbour. They satisfy the filter criteria applied to the Transmitters folder. transmitter B. The transmitters to be allocated will be called TBA transmitters.3. This constraints can be weighted among the others and ranks the neighbours through the importance field. If no focus zone exists in the . Atoll calculates the effective distance. When automatic allocation starts. If the distance between the reference transmitter and the candidate neighbour is greater than this value.10 Automatic Neighbour Allocation The intra-technology neighbour allocation algorithm takes into account all the TBC transmitters.atl document. This constraint can be weighted among others and ranks the neighbours through the importance field.10. For information on the effective distance calculation.atl document are potential neighbours. Atoll checks following conditions: • The distance between both transmitters must be less than the user-definable maximum inter-site distance. Only TBA transmitters may be assigned neighbours. They belong to the folder on which allocation has been executed. then the candidate neighbour is discarded. 191 . DL Power control gains and intra-technology UL noise rises can be used as input for GSM quality-based coverage predictions. SA is the coverage area of reference transmitter A restricted between two boundaries. adjacent cells are sorted and listed from the most adjacent to the least. The weight of this constraint is always the average of the Min and Max values defined for the adjacency factor. This weight is used to calculate the rank of each neighbour and its importance. the existing neighbours are kept. You can force Atoll to keep that transmitter in the reference transmitter’s neighbours list by adding the following option in the Atoll.Atoll 3. Force adjacent layers as neighbours: If selected. Force symmetry: This option enables user to force the reciprocity of a neighbourhood link. The figure below shows the above concept. The overlapping zone ( S A  S B ) is defined as follows: • • 192 SA is the area where the received signal level from transmitter A is greater than a minimum signal level. there must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability. . depending on the above criterion. a transmitter TXi is considered adjacent to another transmitter TXj if there exists at least one pixel of TXi Best Server coverage area TXj is the 2nd Best Server. If the neighbours list of a transmitter is full. you may force/forbid a transmitter to be candidate neighbour of the reference transmitter. The ranking of the adjacent neighbour transmitter increases with the number of these pixels. only the distance criterion is taken into account. More precisely. SB is the coverage area where the candidate transmitter B is the best server. Adjacency is relative to the number of pixels satisfying the criterion. Transmitters are considered adjacent across layers if they belong to different layers and have a coverage overlap of at least one pixel. the reference transmitter will not be added as a neighbour of that transmitter and that transmitter will be removed from the reference transmitter’s neighbours list. Atoll deletes all the current neighbours and carries out a new neighbour allocation. the first boundary represents the start of the handover area (best server area of A plus handover margin named “handover start”) and the second boundary shows the end of the handover area (best server area of A plus the margin called “handover end”). Therefore.ini file: [Neighbours] DoNotDeleteSymmetrics = 1 • • • Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. If not selected. the latter will be considered as candidate neighbour of the reference transmitter.3. Atoll adds all the transmitters adjacent across network layers to the reference transmitter to the candidate neighbour list. • • • When the adjacency option is checked. If the Use Coverage Conditions check box is selected. Therefore.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Adjacency criterion: Geographically adjacent transmitters are determined on the basis of their Best Server coverages in 2G (GSM GPRS EDGE) projects. Delete existing neighbours: When selecting the Delete existing neighbours option. Otherwise. if the reference transmitter is a candidate neighbour of another transmitter. intra-HCS and inter-HCS adjacent. Therefore.14: Overlapping Zones Atoll uses traffic map(s) selected in the default traffic analysis in order to determine the percentage of traffic covered in the overlapping area. the percentage of area meeting the adjacency conditions and the corresponding surface area (km2).3. Among these 15 candidate neighbours.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 SA  SB .Atoll 3. For neighbours accepted for co-site. it provides the list of neighbours. if cells have previous allocations in the list. a neighbour may be marked as exceptional pair. the candidate neighbour B is discarded. In addition. If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference transmitter is 8. intra-HCS or inter-HCS adjacency and coverage reasons. Atoll only displays the transmitters for which it finds new neighbours. For these transmitters. If this percentage is not exceeded. Finally. co-site. Then. The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see number 4 below). or the percentage of traffic covered on the overlapping area S A  S B for the option “Take into account Covered Traffic”. coverage or symmetric. Atoll displays the percentage of area meeting the coverage conditions (or the percentage of covered traffic on this area) and the corresponding surface area (km2) (or the traffic covered on the area in Erlangs). it indicates the importance (in %) of each neighbour and the allocation reason. • The importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason. only 8 (having the highest importances) will be allocated to the reference transmitter. neighbours are marked as existing. it compares this value to the % minimum covered area (minimum percentage of covered area for the option “Take into account Covered Area” or minimum percentage of covered traffic for the option “Take into account Covered Traffic”). In the Results part. Figure 3. 100 ) if the option “Take into account Covered Area” is Atoll calculates either the percentage of covered area ( ----------------SA selected. and to quantify the neighbour importance. Atoll lists all neighbours and ranks them by importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be allocated to each transmitter is exceeded. 193 . the number of neighbours and the maximum number of neighbours allowed for each transmitter. there can be cases where the real distance between assigned neighbours is higher than the Max inter-site distance. priority assigned to each neighbourhood cause is determined using the Importance Function (IF). you can force Atoll to prioritise the individual distances between reference transmitters and their respective neighbour candidates by adding the following lines in Atoll. You can force Atoll to compare the Max inter-site distance with the real inter-site distance by adding the following lines in Atoll. Neighbours of TBA transmitters that satisfy coverage conditions.Atoll 3. d Di  = 1 – ---------d max d is the effective distance (in m).ini: [Neighbours] CandidatesMaxDistanceInImportanceCalculation=1 3. The IF considers the following factors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. Neighbours of TBA transmitters marked as exceptional pair. see "Appendix: Calculation of the Inter-Transmitter Distance" on page 195. the automatic neighbour allocation compares the defined Max intersite distance with the effective inter-transmitter distance. the neighbour importance calculated with respect to distance is based on the global Max inter-site distance setting for all neighbour candidates.3 Neighbour Importance Calculation The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf.ini: [Neighbours] RealInterSiteDistanceCondition=1 • By default. this value varies between 0 and 100%. because the effective distance is smaller. Automatic neighbour allocation parameters are described in "Neighbour Allocation for All Transmitters" on page 191.10.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks • ©Forsk 2015 By default.2 Neighbour Allocation for a Group of Transmitters or One Transmitter In this case. For information on the effective distance calculation.3. It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. there can be cases where the calculated importance is different when the global Max inter-site distance is modified. 194 . 3. intra-HCS or inter-HCS adjacent and symmetric. To avoid that.10. Atoll allocates neighbours to: • • • TBA transmitters. Neighbourhood cause When Importance value Existing neighbour Only if the Delete existing neighbours option is not selected and in case of a new allocation Existing importance Exceptional pair Only if the Force exceptional pairs option is selected 100 % Co-site transmitter Only if the Force co-site cells as neighbours option is selected Importance Function (IF) Adjacent transmitters Only if the Force intra-HCS adjacents as neighbours option is selected Importance Function (IF) Adjacent layer Only if the Force inter-HCS adjacents as Neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions Only if the % minimum covered area is exceeded Importance Function (IF) Symmetric neighbourhood relationship Only if the Force neighbour symmetry option is selected Importance Function (IF) Except the case of forced neighbours (importance = 100%). table below). As a consequence. As a consequence. The minimum and maximum importance assigned to each of the above factors can be defined. If the Min and Max value ranges of the importance function factors do not overlap. Figure 3. • • • The co-site factor (C): a Boolean.3% so that the maximum D variation does not exceed 1%. There can be a mix of the neighbourhood causes. With the default values for minimum and maximum importance fields. neighbours will be ranked in this order: co-site neighbours. The adjacency factor (A): the percentage of adjacency. 3.3.4 Appendix: Calculation of the Inter-Transmitter Distance Atoll takes into account the real distance ( D in m) and azimuths of antennas in order to calculate the effective intertransmitter distance ( d in m). the neighbours may be ranked differently.Atoll 3. The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 30% Adjacency factor (A) Min(A) 30% Max(A) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The Importance Function is evaluated as follows: Neighbourhood cause Importance Function Resulting IF using the default values from the table above Coverage Min(O)+Delta(O){Max(Di)(Di)+(100%Max(Di))(O)}+Min(Di)+Delta(Di)(Di) 10%+20%{10%(Di)+90%(O)}+1%+9%(Di) Adjacent layer (Min(A)+Max(A))/2 45% Adjacent transmitters Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 30%+30%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Co-site transmitters Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 60%+40%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Delta(X)=Max(X)-Min(X) • • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. adjacent neighbours.15: Inter-Transmitter Distance Computation 195 . neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping. With a value of Min(O) = 0%. the neighbours will be ranked by neighbour cause.10. and neighbours allocated based on coverage overlapping. If the Min and Max value ranges of the importance function factors overlap.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 d max is the maximum distance between the reference transmitter and a possible neighbour. The overlapping factor (O): the percentage of overlapping. d = D   1 + x  cos  – x  cos   x = 0. (|g| = number of TRXs in TRGi).11 AFP Appendices 3. and in the case of BBH. • When i is BBH. |g| . each using the entire group g and having a virtual MAIO [0. the same MAL length and the same hopping mode.1 The AFP Cost Function The notations listed hereafter are used to describe the cost function: • • • • • TRG: TRGs: : g : ARFCN: • 2 ARFCN • x : • A i g : Group of TRXs Set of all the TRGs If and only if Size of any group g Set of all the frequencies : Set of all the subsets of frequencies The largest integer x Number of times a group g  2 ARFCN is assigned to TRGi in the assignment A For example: • When i is NH. A i g = 1  g is a single member group containing one of the frequencies assigned at TRGi. ATOM  i   ATOM  k  i and k are synchronised. For synchronised TRGs. then A i g = 0 . 3.Atoll 3. A i g = Number of TRXs in TRGi  g is the set of frequencies assigned to TRXs of TRGi. A i g can be either 0 or equal to the number of TRXs in TRGi. The probability to be interfered is denoted by P i i' g  A  (i’ is the TRX index). It is a function of the whole frequency assignment as well. Different TRX indexes may have different MAIOs.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 The formula above implies that two cells facing each other will have a smaller effective distance than the real physical distance. If two TRGs interfere but are not in the same atom. The probability penalty due to violating a separation constraint is P i i' g  A  . A i g = n  g is the set of frequencies assigned to n TRXs of TRGi.3.1]. When we talk about "TRXs of i using g". A i g must be less than or equal to the umber of TRXs in TRGi. The quality of unsynchronised TRGs is a function of all possible frequency combinations. • When i is SFH. The precise definition of the term “to be interfered” is provided afterwards. If |g| is not 1 or if g does not contain a frequency assigned at i. • • TSi: TLi: Number of timeslots available for each TRX in TRGi Traffic load of TRGi (calculated or user-defined) TL i = #Erlangs of a single TRX in TRGi divided by TSi • • • • • TSUi: CFi: QMINi: PMAXi: REQi: Downlink timeslot use ratio (due to DTX) at TRGi Cost factor of TRGi (AFP Weight) Minimum required quality (in C/I) at TRGi Percentage permitted to have quality lower than QMINi at TRGi Required number of TRXs at TRGi A communication uses the group g in TRGi if its mobile allocation is g. P i i' g  A  is a function of the whole frequency assignment. then there are |g| such virtual TRXs. i and k. NH TRGs or BBH TRGs are always in separate atoms.11. The term “Atom” will be used in the following context: For two TRGs. It is this effective distance that will be taken into account rather than the real distance. have the same HSN. 196 . these can be taken as unsynchronised. We assume all the groups assigned to TRGi to have the same length. pairs of frequencies emitted at the same time are known. • MIS_TRX i is the number of missing TRXs for the subcell i.  is the cost value of a corrupted TRX. DOM_TRX i is the number of TRXs. This value can vary between 0 and 10. The default cost value is set to 1 and can be modified in the AFP module properties dialogue.3. MIS_TRX i = MAX 0 REQi –  g2 • • • • • A i g ARFCN  is the cost value for a missing TRX. This value can vary between 0 and 10. This cost has two components. If i’ is valid. It counts the impaired traffic of the network TRXs in weighted Erlangs.  comp .0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3. the algorithm evaluates the cost of a valid TRX.  mis represents the missing TRX cost component  sep represents the separation component  comp represents the additional cost component (interference. The default cost value is set to 10 and can be modified in the AFP module properties dialogue.1 Cost Function The Atoll AFP cost function is a TRX based cost and not an interference matrix entry based cost.Atoll 3.11..  =  mis +  sep +  comp +  corr +  dom . cost of changing a TRX)  corr represents the corrupted TRX cost component  dom represents the out-of-domain frequency assignment cost component  mis =   MIS_TRX i     TL i  CF i  TS i   CORR_TRX i     TL i  CF i  TS i   DOM_TRX i     TL i  CF i  TS i i  TRGs  corr = i  TRGs  dom = i  TRGs  sep    =   i  TRGs    comp   ARFCN g2 i'  TRXs of i using g    =   i  TRGs       ' i i' g  A   TL i  CF i  TS i     ARFCN g2 i'  TRXs of i using g    '' i i' g  A   TL i  CF i  TS i    In the above equations. This value can vary between 0 and 1.  sep . • i’ is the TRX index belonging to  0 1 . A i g – 1  .1. 197 .. for the subcell i.5 and can be modified in the AFP module properties dialogue. having out-of-domain frequencies assigned. CORR_TRX i is the number of corrupted TRXs for the subcell i. as mentioned earlier. ' i i' g  A  and '' i i' g  A  . The cost function  is reported to the user during the AFP progress with the help of its 5 components:  mis .  is the cost value of a TRX with out-of-domain frequencies assigned. The default cost value is set to 0.  corr and  dom . And. a virtual TRX is considered in case of BBH. 1. Let Cost s z denote the user defined separation penalty for a required separation “s” and actual separation “z”.  ii'kgg'k' is considered to be the effect of a separation violation on the i' th TRX of TRGi assigned the group g. Hence. It can be modified in the AFP module properties dialogue. Co-site 4. Neighbours For example. Estimation is based on costs specified for the required separations. set to 1 by default. If only a neighbour relation exists between two 198 . 3. The AFP module properties dialogue takes probability percentages as inputs while this document deals in probability values. ©Forsk 2015 If the option “Take into account the cost of all the TRXs” available in the AFP module properties dialogue is selected.1 Separation Violation Cost Component The separation violation cost component is evaluated for each TRX. then.2 Cost Components Separation violation and interference cost components are described hereafter. if a pair of subcells are co-site and neighbours at the same time. This value can be between 0 and 1. P MAX : If P' i i' g  A  + P'' i i' g  A   P MAX .11. while others (such as separation costs and diversity gains) can be managed through the properties dialogue of the Atoll AFP module.3. ik of these subcells will be the weight of co-site relations. P' i i' g  A  is the same as ' i i' g  A  (separation violation probability penalty) and P'' i i' g  A  the same as '' i i' g  A  (complementary probability penalty due to interference and the cost of modifying a TRX) in most cases.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks • ' i i' g  A  is the separation violation probability penalty.11. they will be considered as co-site because higher priority.2. SEP i k v is used instead of Cost SEP_CONSTR i k z as abbreviation. ik represents the weight of the specific separation constraint between i and k. Exceptional pairs 2. These are explained in detail in the next sections. percentage of interference allowed per subcell).  denotes the overall weight of the separation violation cost component. Then ' i i' g  A  = P' i i' g  A  and '' i i' g  A  = P'' i i' g  A  . caused by the k' th TRX of TRGk assigned the group g' . 3.Atoll 3. Let SEP_CONSTR i k denote the required separation constraint between TRGi and TRGk. Otherwise. the algorithm compares P' i i' g  A  + P'' i i' g  A  with the quality target specified for i. Parameters considered in the cost function components can be fully controlled by the user. This specific weight depends on the type of separation violation and follows the following priority rule: 1. Some of these parameters are part of the general data model (quality requirements. Co-transmitters 3.1. Both ' i i' g  A  and '' i i' g  A  will be equal 0. • '' i i' g  A  is complementary probability penalty due to interference and the cost of modifying a TRX. ' i i' g  A  = P' i i' g  A  and '' i i' g  A  = P'' i i' g  A  Or if the option “Do not include the cost of TRXs having reached their quality target” available in the AFP module properties dialogue is selected. 199 .0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 subcells. the SEP_GAIN  i k g  g'  becomes. Many “small” Gˆ i k g g' and ' ii'kgg' have to be combined to form one cost element. The default weights of each type of separation are available in the Separation cost tab. the effects of DTX and traffic load become more significant (due to the consideration of the average case instead of the worst case)..  F_N – 1  ----------------------------------------------------------------------------- F_N In the above equations. If ATOM  i   ATOM  k   SEP i k f – f' fg f'  g' Then  ii'kgg'k' =   ik  ----------------------------------------.Atoll 3. In addition. the cost is weighted by an interferer diversity gain. For this reason. This sum is naturally limited to 100% of the TRX traffic. Therefore. of the hopping mode). Let f_n denote the instantaneous frame number from 0 to F_N . which is same for all values of k. if this option is selected. It is used in P i i' g  A  definition as well. F_N  g  is the number of frames in the MAL g. the P' i i' g  A  . This is done through iterating over all violating assignments and by summing up an equivalent to the probability of not being violated while considering each separation violation as an independent probability event. On the other hand.1  SEP_GAIN  i k g  g'   10 The separation gain... F_N  g  = g . then ik will be further weighted by the neighbour relation importance.  2 + ASYN_GAIN  i k g'   SEP_GAIN  i k g  g'  = I_DIV  g  +0. of course. TL k  TSU k And ASYN_GAIN  i k g'  = 0 I_DIV( g'  if ATOM(i) = ATOM(k) Otherwise More than one separation violations may exist for a TRX. While  =  f_n + MAIO A And  =  f_n + MAIO A th i g i' k g' k'  modulo F_N and g  is the  frequency in g. Without this option. Since F_N  g  = F_N  g'  . we shortly denote the two as F_N .  modulo F_N and g'  is the  th frequency in g’. The value of ik remains between 0 and 1. 1 Gˆ i k g g' = ---------------------------------------------------------- 0. the SEP_GAIN  i k g  g'  is: SEP_GAIN  i k g  g'  = I_DIV  g  I_DIV  g  is the user defined interferer diversity gain (dB) for a given MAL length. frequencies belonging to a MAL with a low fractional load. denoted by SEP_GAIN  i k g  g'  is basically a function of the MAL length (and.5  TSU_GAIN  k   min  10 4 +   2 + I_DIV  g    -----------------------------------------------------------------   4 1 TSU_GAIN  k  = log 10  -------------------------- . and is given by.3. it is possible to consider these effects in SEP_GAIN  i k g  g'  through the relevant option available in the Advanced tab of the AFP module properties dialogue. should not be weighted equally as in a non-hopping separation breaking case. and breaking a separation constraint. With frequency hopping. g  g' If ATOM  i  = ATOM  k   Then  ii'kgg'k' =   ik  SEP i k g  – g'  f_n   0  1  . For further description. then MAIO A k g' j = 0. As said earlier. we consider g' virtual TRXs. the jth TRX has the MAIO j. caused by the k' th TRX of TRGk assigned the group g' . then MAIO A k g' j If TRGk is BBH. GSM GPRS EDGE network optimisation. ' ii'kgg'k' is the effect interference on the i' th TRX of TRGi assigned the group g. is the simplest possible as it solves the first problem by linear summation and truncation at the value of 1 and it solves the second problem by averaging and adding the two diversity gains: • F_DIV  g  . and • I_DIV  g  .2.1. In case of a MAL containing more than one frequencies. then  k'  i'  . the frequency diversity gain. GSM GPRS EDGE generic AFP management). Let MAIO A k g' j be the j’th MAIO of A k g' . the following problems are encountered: • • The QMINi C/I quality indicator corresponds to the accumulated interference level of all interferers while the C/I interference histograms correspond to pair-wise interferences. If ATOM  i   ATOM  k  Then ' ii'kgg'k' =  f  g f'  g' C Probability  ----.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015    1 –  ii'kgg'k'  Gˆ i k g g'  1 –  P' i i' g  A  =  k  TRGs  ARFCN  g'  2  k'  TRXs of k using g'            In the above formula.35 by default.  denotes the global weight of interference cost component. j is one of the  0 1 . it takes into account frequency and interferer diversity gains and models frequency hopping and gain due to DTX. in case of BBH. 3. the interferer diversity gain. Interference histograms are described in User Manual (GSM GPRS EDGE project management. In addition.2 Interference Cost Component The interference cost component is evaluated for each TRX.. if  k = i  . interferences on several different frequencies of a MAL must be combined. Q_UB i k f f'  I ik  -------------------------------------------------------------------------g  g' Q_UB i k f f' = QMIN i – If ATOM  i  = ATOM  k  200 f – f'  ADJ_SUP + INTERF_GAIN  i k g  g'  . so that interference with itself is not taken into account.3. ' ii'kgg'k' is defined as an interference event. When estimating P'' i i' g  A  . g'  k g' j If TRGk is NH. presented below. The value of MAIO A is one of  0 1 . This value can vary between 0 and 1 and is set to 0. Let g i be the ith frequency in the group g. Let F_N  g  be the number of frames in the MAL g. which can be modified in the AFP module properties dialogue. A k g' – 1  TRXs. Let f_n denote the instantaneous frame number from 0 to F_N .. Both QMINi and the histograms correspond to a single frequency. Its estimation is based on interference histograms calculated for pairs of subcells. F_N  g  = g . Interference histograms can also be exported to files.11. Similar to the definition of  ii'kgg'k' .Atoll 3. This estimation. Hereafter. refer to "Interference" on page 205... = j . The Advanced properties tab shown in the figure below facilitates modelling these effects. so that interference with itself is not taken into account. Since F_N  g  = F_N  g'  . the following considerations should be taken into account: 1. f = g . if  i = k  . The fact that long MALs with synthesized hopping permit discarding the worst case estimation and include a gain due to DTX and low traffic load at the interferer end. Q_UB i k f f' = QMIN i – f – f'  ADJ_SUP + INTERF_GAIN  i k g  g'     Therefore.11. 3. these are both represented by F_N . P'' i i' g  A  = 1 –   1 – P' i i' g  A            1 –  ii'kgg'k'   – P' i i' g  A     k  TRGs ARFCN g'  2 k'  TRXs of k using g' In the above formula. 2. This factor represents that the effect of average negative effects over user geographic location are directly proportional to the MAL length. Non-linearity of Frame Error Rate (FER) with respect to average C/I conditions and MAL length. F_DIV and Other Advanced Cost Parameters When combining interference effects (or separation violation effects) on different frequencies belonging to a MAL.  ii'kgg'k' C  Probability  ----. Frequency Diversity Gain. 201 . f' = g'  . Interference Diversity Gain.. 3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 Then. added to SEP_GAIN  i k g  g'  .3. k g' k'  modulo F_N .3 I_DIV.2.. The sum is limited to 100% of the TRX traffic. 4. F_DIV  g  . then  k'  i'  .1.Atoll 3. This factor models the gain due to diversity of multi-path effects and should be applied to the interference cost component only. we have. Q_UB i k f f'    I  ik - =  ------------------------------------------------------------------------F_N   f_n   0 1 . The only difference is the frequency diversity gain. INTERF_GAIN  i k g  g'  is quite similar to SEP_GAIN  i k g  g'  . F_N – 1     .  =  f_n + MAIO A  =  f_n + MAIO A i g i'  modulo F_N . This equation generates values somewhat lower than empirical best-found values (this is because we prefer a slightly pessimistic cost function to be on the safe side). For concentric cells.3. Numbers of traffic timeslots. The notations listed hereafter are used for the description. Number of required TRXs of each TRX type in the pool. a traffic pool may include additional subcells. …. it provides a means to model the non-linear FER effects and interference diversity both. …. The default values in this table correspond to the curve y = 2  log 10  x  . CF(1). The BCCH subcell may not always be part of the pool’s TRX types. The other table contains the F_DIV values. ts(n)}: {L(0). Overall circuit-switched traffic demand of the traffic pool (Subcells table or traffic analysis results). its value is set to 1 in order to avoid working with transmitters carrying no traffic. L(2). d(n)}: {ts(0). Number of TRXs in the frequency plan. HR(1). 202 • {BCCH. CF(n)}: CS (Erlangs): • PS (Data Timeslots): • • {nb(0). In most cases. HR(2).0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Figure 3. Traffic loads. ts(2). …. Overall packet-switched traffic demand of the traffic pool (Subcells table or traffic analysis results). and the recalculation of traffic loads. L(1). TCH HR use ratios.11. This optimisation is performed for each traffic pool in the network. nb(n)}: {HR(0). which are the same as the I_DIV values by default.Atoll 3. …. d(2).16: The Advanced tab of the AFP module Properties dialogue The Interference Diversity Gain table lists the values of I_DIV provided as a functions of MAL length. TCH(n)}: • • • • • {d(0). …. . …. CF(2). The cost component described below. at least two traffic pools exist per transmitter. AFP cost factors. In more complex cases. the traffic pool is equivalent to a transmitter and corresponds to the BCCH and TCH subcells.2 The AFP Blocked Traffic Cost This section provides additional information on the AFP cost components used for the optimisation of the number of TRXs. and more than one traffic pools may exist per transmitter. nb(2). …. d(1). L(n)}: {CF(0). nb(1). is only used when the AFP performs the oprimisation of the number of TRXs. This gain is applied to the interference cost component and to the separation constraint violation cost component. HR(n)}: Subcells of a traffic pool. ts(1). If CS or PS is less than 1. 3. TCH(2). TCH(1). Therefore. P Blocking = ErlangB  CS TTS  The above equations give the number of served circuit-switched timeslots (SCS): HR   CS   1 – P SCS =  1 – ------------Blocking  2  The number of served packet-switched timeslots (SPS) is obtained as follows: n         SPS = Min  PS Max 1 nb  i   ts  i  – SCS       i=0     L  nb  is given by: SCS + SPS L  nb  = -------------------------------------------------------------n   Max  1 nb  i   ts  i     i=0   BL  nb  is given by: HR  PS + CS   1 – ------------ 2  BL  nb  = --------------------------------------------------------------. L  nb  is obtained from the Erlang B equation applied to the traffic pool demand and the total number of timeslots (TTS): n    nb  i   ts  i - TTS = Max  1 ---------------------------- HR      i = 0  1 – ------------2   The Max() function above gives 1 timeslot when there is no TRX.1 Calculation of New Traffic Loads Including Blocked Traffic Loads During the optimisation of the number of TRXs. ….0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3.3. nb(1). traffic loads are calculated in order to determine the blocked traffic loads BL  nb  . Without the optimisation of the number of required TRXs. They are calculated at traffic pool level for the vector {nb(0). nb(2).11.– L  nb  n     Max  1 nb  i   ts  i   i=0   203 . nb(n)} as follows:  HR  PS +  CS   1 – -------------  2   BL  nb  + L  nb  = --------------------------------------------------------------n   Max  1 nb  i   ts  i     i=0   HR n = Max i = 0  HR  i   BL  nb  is determined from the above equation once L  nb  is known.2.Atoll 3. the network’s weighted Erlangs are calculated as follows: n WE =  d  i   ts  i   L  i   CF  i  i=0 With the optimisation of the number of TRXs. the network’s weighted Erlangs are calculated as follows: n WE =  nb  i   ts  i    BL  nb  + L  nb    CF  i  i=0 BL  nb  and L  nb  represent the load estimation and the blocked load estimation of the AFP. The blocked traffic load is then multiplied by the AFP cost weight and the number of timeslots to calculate the blocked traffic cost. 0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 Once L  nb  and BL  nb  are known. In other words. and R’. which are calculated for the actual traffic loads.11. MB = ErlangB  CS' TTS'  PS' R' = ------------------------------------------------------HR  PS' + CS'   1 – ------------ 2  PS' . In order to guide the AFP to generate it from the loads. and calculate the cost. The equations are solvable due to the monotone nature of the Erlang B function. however. the ratio is calculated so that the worst case blocking rate is BM.Atoll 3. …. d(1). 204 . we assume that R’ = R. PS R = ---------------------------------------------------HR   PS + CS   1 – ------------- 2 P Blocking = ErlangB  CS TTS'  HR   CS   1 – P SCS =  1 – ------------Blocking   2  n       SPS = Min  PS Max  1 d  i   ts  i  – SCS     i=0     n SCS + SPS =  d  i   ts  i   L  i  i=0 The above five equations are solved to get the values of the five variables PS. and BL  nb  is used to generate a new cost component. d(n)}: L: Number of required TRXs of each TRX type in the pool Traffic load • TTS' : n     d  i   ts  i  TTS' = Max  1 ---------------------------- HR      i = 0  1 – ------------2  • MB: Maximum blocking rate (between 0 and 1). Here. P Blocking . Inputs for a given traffic pool: • • {d(0). giving a load of 1. The following equations are solved to find PS’.2.2 Recalculation of CS and PS From Traffic Loads In earlier versions. and R. we assume that a traffic load of 1 is generated by a demand of (1+MB)*TTS’ which generates a blocking rate of MB. d(2). SPS. L  nb  replaces TLi in the cost function (See "The AFP Cost Function" on page 196). PS is different from PS’ and CS is different from CS’. the detailed traffic demand information is not available. The following equations are solved to find PS. SCS. which are calculated for a traffic load of 1. CS’. the following two equations with three variables must be solved. the blocked Erlangs of the pool: n  nb  i   ts  i   BL  nb   CF  i  i=0 3. This assumption implies that R is more or less the same as MB for big traffic pools and considerably larger than MB for smaller pools.+ CS'  1 + MB   TTS' = --------------------------HR   1 – ------------ 2  When the traffic load of a pool is not 1. PC.3. CS.  The ratio of packet-switched demand is given by: PS R = ---------------------------------------------------HR  PS + CS   1 – ------------ 2  Here. BCCH] and [TX2.2. Otherwise. BCCH] and [TX2. BCCH] be the interfered and interfering subcells respectively. Atoll looks up the cumulative density function at the value corresponding to X .Y dB. The HR ratios are the same within the subcells of a traffic pool.3. In Atoll. Then.11.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3. 205 . Y dB being the adjacency suppression value.3 Testing the Blocked Cost Using Traffic Analysis As long as the conditions below hold truw.1 Using Interferences If interferences are to be taken into account by the AFP. In order to deduce the adjacent interference probability value.11.5%). Atoll integrates these C/I values to determine a C/I distribution and transforms this distribution function into a cumulative density function in the normal way. 3. The following example may be helpful in further clarifying this concept: Example: Let [TX1. the user should previously decide to take interferences into account (enabling the loading of all the potential interferers). BCCH] In this case.11.11. 3. All transmitters belong to the same HCS Layer. Atoll does not allow performing their computation by disabling the histogram part in the corresponding dialogue.1%.17: The cumulative density of C/I levels between [TX1. interfering subcell] pair. Atoll calculates a C/I value on each pixel of the interfered subcell service area (as if the two subcells share the same channel). Figure 3.5% (as this requirement is fulfilled with a probability of 93. The service areas for both have been defined by Best Server with 0 dB margin. The dimensioning model is based on Erlang B.3.3 Interference This appendix provides a high-level overview of interference taken into account by the AFP. The interference probability is stated in percentage of interfered area. The conditions are: • • • • • • The AFP cost factors are 1. the blocked cost calculation in the AFP and the effective overflow calculation in the KPI calculation and dimensioning use the same algorithm. L  nb  Effective Overflow rate = 1 – -------------------------------------L  nb  + BL  nb  Output: New values for CS and PS.Atoll 3. the interference probability is 6. we observe that the probability for C/I (BCCH of TX2 effecting the BCCH of TX1) being greater than 0 is 100% (which is normal because TX1 is the Best Server). The probability of having a C/I value at least equal to 31 dB is 31. The timeslot configurations are the default ones. they must be calculated or imported beforehand. 3. There exists at least one TRX in the traffic pool (and at least one Erlang of traffic). For a required C/I level of 12 dB on the BCCH of TX1. In order to do this.3.2 Cumulative Density Function of C/I Levels For each [interfered subcell. both the IMco and IMadj are represented by this Cumulative Density function This implies that each query for the probability to have C/I conditions worse than X dB requires a single memory access: the co-channel interference probability at X dB. 3. and creates more generic interference information. and when one or more frequencies are common (or adjacent) in two interfering MAL sequences. assignment strategies. introduces very little (or no) overhead.Atoll 3. its influence on the subcell service zone is taken into account in the . and the number of common (and adjacent) frequencies in the two MALs. frequency domains.1 improves the AFP result. In the case of precise distribution information. synchronisation information. the interference information keeps its original probability units and is easier to check and validate. We assume C_I values to be discrete and in dB.11.4.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015 The subcell power offset does not enter the calculation results in the . …). Once Atoll finishes the counting for an [interfered. 5. 206 . Separation requirements and/or neighbour relations. the heaviest part of the task. the user spends less time on interference recalculations than in the case of a two-entry matrix ( “everything” is included). All these factors cannot be pre-calculated since it is the AFP that determines the MAL length and the MAL frequencies. C/I weighting (the interference levels above and below the C/I target).11.clc file. the number of operations is the same: one increment of an integer value.3. traffic load levels. Therefore.2 Efficient Calculation and Storage of Interference Distribution In the innermost loop of the calculation process Atoll increments a counter each time a C/I level has a certain value.clc file. cost function parameters. Remove equipment By not mixing any of the elements above. These two overheads are negligible and do not affect the calculations. or 6.3. number of required TRXs. 4. 3.3.11. CDF(Pci) is the cumulative density function of Pci: CDF  Pci  v n C_I   =  Pci  v n x  xC I 3.3 Precise Definition Pci  v n C_I  is defined to be the probability of a communication (call) occupying a timeslot in subcell v (victim) to have C/ I conditions of C_I with respect to a co-channel interference from the BCCH TRX of cell n (neighbour).4. 3.4. the only overheads are the read / write times to the files and the memory occupation at running time. 3. Hopping gain values.3 Robustness of the IM By having precise C/I distributions calculated and exported. access to interference probability at a certain level is instantaneous.3.3.11. Direct Availability of Precise Interference Distribution to the AFP In the presence of frequency hopping. In the case of a two-entry IM. On the other hand. DTX level. interferer] pair. Quality requirements of network elements (required C/I.11. it compresses the information from the counters to a Cumulative Density Function (CDF) representation. 3. Thus. HSNs. 2. In both cases. …). interferer] pair. It is added later by the AFP interface.4 Precise Interference Distribution Strategy Why does Atoll calculate and maintain precise interference distributions. Any frequency assignment setting (MAL length directives. there are about 40 counters per pair. the user is free to change the following settings without the need for recalculating their interference distributions: 1. there are only two counters for each [interfered. the traffic load on the interferer TRX. the hopping gain depends on following factors: • • • • the MAL length. % Probability Max. while the most common solution (used by most other tools) is rather to compress the information into two values: the co-channel and adjacent-channel interference probabilities? The reason is simply that it. DTX activities. • • • 3. In this way. • The gain introduced by the traffic load of the interferer depends on the hopping mode and the MAL length. It is facilitates merging IMs with different traffic units. an IM calculated externally to Atoll.5 Traffic Load and Interference Information Discrimination Atoll maintains the traffic load separate from the interference information.11. 2. being a subset. the AFP must know the traffic loads in order to choose a low load TRX to be removed. Let us look at the possible alternatives to this strategy: 1. It does not create any overhead (the size of the additional information is negligible compared to the size of the IM). in detail. Incorporating this gain in the IM (as a result of the mixed option) means that the IMs become hopping-mode and MALsize dependent. this option is so practically useless due to its inefficiency. in order to be able to not violate site constraints. does not contain option 2 (i. The traffic information can be used for weighting the separation violation component. A third option also exists. with a nonhopping BCCH can be used for the hopping TCH. But option 1. The reasons for implementing this strategy are explained here.3. The traffic load can be used in deciding whether a TRX can be left uncreated. It helps keeping the unit definitions simpler. This is again a bad idea because of the unit definition and the variety of IM sources. The mixed option: The interference information contains the traffic information as well.e. Though. this has been done because: • • • • • • Option 2 is a superset that contains option 1. This is a bad idea since the AFP should be able to change the MAL.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks AT330_TRR_E1 3. It involves less benefits than the option chosen in Atoll. if there are too many TRXs at a site and the user wishes that the AFP remove one of them.3. For example.Atoll 3. The separated option: The AFP has separate access to traffic load information and to interference probabilities (As in Atoll). However. In addition. 207 . It consists in mixing IM and traffic but still keeping the traffic in its isolated form. Knowing the difference between the two alternative solutions explains why the second strategy has been opted for for Atoll. each IM entry will contain the quantity of traffic interfered if a co-channel / adjacent channel reuse exists. And the user should be able to change the hopping mode without recalculating the IM. In this way. once the information are mixed they cannot be separated). 3.0 Technical Reference Guide for Radio Networks Chapter 3: GSM GPRS EDGE Networks 208 ©Forsk 2015 .Atoll 3. Chapter 4 UMTS HSPA Networks This chapter covers the following topics: • "General Prediction Studies" on page 211 • "Definitions" on page 214 • "Simulations" on page 225 • "UMTS HSPA Prediction Studies" on page 284 • "Automatic Neighbour Allocation" on page 312 • "Primary Scrambling Code Allocation" on page 319 • "Automatic GSM-UMTS Neighbour Allocation" on page 329 . 0 Technical Reference Guidefor Radio Networks © Forsk 2015 210 .Atoll 3.3. taken into account when the option “Indoor coverage” is selected. Atoll considers that G term and L term equal zero. For the definition of TBC transmitters please refer to "Path Loss Matrices" on page 103. 4. M Shadowing – model is the shadowing margin. All the calculations are performed on TBC (to be calculated) transmitters.1. L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model. see "UMTS. 211 .2. Study criteria are detailed in the table below: Study criteria Formulas Signal level ( P rec ) in dBm Signal level received from a transmitter on a carrier (cell) P rec  ic  = EIRP  ic  – L path – M Shadowing – model – L Indoor + G term – L term L path = L model + L ant Path loss ( L path ) in dBm Total losses ( L total ) in dBm Tx L total =  L path + L Tx + L term + L indoor + M Shadowing – model  –  G Tx + G term  where. L Tx is the transmitter loss ( L Tx = L total – DL ).1. Atoll takes the highest pilot power of carriers to calculate the signal level received from a transmitter.1.3. G term is the receiver antenna gain. and LTE Documents" on page 30. TDSCDMA. L Indoor are the indoor losses. For information on calculating transmitter loss. • 4. L term are the receiver losses. In this case. CDMA2000. you can consider the best carrier of all bands or the best carrier of a particular frequency band (Best (All Bands/Specific Band) option).1 Profile Tab Atoll displays either the signal level received from the selected transmitter on a carrier ( P rec  ic  ). WiMAX. G Tx is the transmitter antenna gain.1 Calculation Criteria Three criteria can be studied in point analysis (Profile tab) and in common coverage studies. L ant Tx is the transmitter antenna attenuation (from antenna patterns).1 General Prediction Studies 4. ic is a carrier rank. This parameter is taken into account when the option “Shadowing taken into account” is selected.Atoll 3. or the highest signal level received from the selected transmitter on the best carrier. EIRP is the effective isotropic radiated power of the transmitter.2 Point Analysis 4.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4 UMTS HSPA Networks This chapter describes all the calculations performed in Atoll UMTS HSPA documents. • When you make the prediction. • EIRP  ic  = P pilot  ic  + G Tx – L Tx ( P pilot  ic  is the cell pilot power). L path .1.1 All Servers The service area of Txi corresponds to the bins where: Txi Txi Txi MinimumThreshold  P rec  ic   or L total or L path   MaximumThreshold 4.3 Coverage Studies For each TBC transmitter. or the total losses. • For a selected transmitter.Atoll 3.1. 4. see the Administrator Manual. L total . Txi.1. Atoll displays either the signal level received on a carrier.1 Service Area Determination Atoll uses parameters entered in the Condition tab of the coverage study property dialogue to predetermine areas where it will display coverage. • You can use a value other than 30 dB for the margin from the best serving cell signal level. In fact.3. Atoll determines the selected criterion on each pixel inside the Txi calculation area. or the total losses. it is also possible to study the path loss.1. Path loss and total losses are the same on any carrier. L path . L total .3. We can distinguish three cases: 4. 4.2.3. For each transmitter. The maximum number of reception bars depends on the signal level received from the best serving cell. ( P rec  ic  ). each pixel within the Txi calculation area is considered as a potential (fixed or mobile) receiver. 212 .3. 4. Reception bars are displayed in a decreasing signal level order.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 For a selected transmitter. it is also possible to study the path loss. Only reception bars of cells whose signal level is within a 30 dB margin from the best serving cell can be displayed. So.1.1. Path loss and total losses are the same on any carrier. or the highest signal level received on the best carrier.1. Coverage study parameters to be set are: • • The study conditions in order to determine the service area of each TBC transmitter. Best function: considers the highest value. you can study reception from TBC transmitters for which path loss matrices have been computed on their calculation areas. The display settings to select how to colour service areas. for example a smaller value for improving the calculation speed.2 Reception Tab Analysis provided in the Reception tab is based on path loss matrices. For more information on defining a different value for this margin.2 Best Signal Level and a Margin The service area of Txi corresponds to the bins where: Txi Txi Txi MinimumThreshold  P rec  ic   or L total or L path   MaximumThreshold And Txi Txj P rec  ic   Best  P rec  ic   – M ji M is the specified margin (dB). 2 Coverage Display 4.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 • If the margin equals 0 dB.1 Plot Resolution Prediction plot resolution is independent of the matrix resolutions and can be defined on a per study basis. A pixel of a service area is coloured if the signal level is greater than or equal to the defined thresholds (the pixel colour depends on the signal level). If the margin is set to -2 dB. Coverage consists of several independent layers whose visibility in the workspace can be managed. which are 2nd best servers.1.2.1.3 Second Best Signal Level and a Margin The service area of Txi corresponds to the bins where: Txi Txi Txi MinimumThreshold  P rec  ic   or L total or L path   MaximumThreshold And nd Txi Txj P rec  ic   2 Best  P rec  ic   – M ji M is the specified margin (dB).2. Best Signal Level (in dBm. Atoll will consider bins where the signal level received from Txi is either the second highest or 2dB lower than the second highest.3. Each layer corresponds to an area where the signal level from the best server exceeds a defined minimum threshold. Atoll chooses the highest value. dBµV/m) Atoll calculates signal level received from the transmitter on each pixel of each transmitter service area.1. 4. A pixel of a service area is coloured if path loss is greater than or equal to the defined minimum thresholds (pixel colour depends on path loss).1. There are as many layers as defined thresholds.Atoll 3. dBµV/m) Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Atoll will consider bins where the signal level received from Txi is 2dB higher than the signal levels from transmitters. Where other service areas overlap the studied one. which are 3rd best servers. If the margin is set to 2 dB.2 Display Types It is possible to display the transmitter service area with colours depending on any transmitter attribute or other criteria such as: Signal Level (in dBm. Coverage consists of several independent layers whose visibility in the workspace can be managed.3. dBµV. dBµV. Each layer shows the different signal levels available in the transmitter service area. 2nd Best function: considers the second highest value. • • • If the margin equals 0 dB. Atoll will consider bins where the signal level received from Txi is 2dB higher than the signal levels from transmitters. If the margin is set to -2 dB. If the margin is set to 2 dB. Atoll will consider bins where the signal level received from Txi is the second highest. There are as many layers as service areas. 4. Prediction plots are generated from multi-resolution path loss matrices using bilinear interpolation method (similar to the one used to evaluate site altitude). Atoll will consider bins where the signal level received from Txi is either the highest or 2dB lower than the highest. • • 4. 213 . Each layer shows the different path loss levels in the transmitter service area.3. A pixel of a service area is coloured if the signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on signal level).1. Path Loss (dB) Atoll calculates path loss from the transmitter on each pixel of each transmitter service area. There are as many layers as transmitter service areas.3. Coverage consists of several independent layers whose visibility in the workspace can be managed. Atoll will consider bins where the signal level received from Txi is the highest.3. Each layer corresponds to an area where the number of servers is greater than or equal to a defined minimum threshold. There are as many layers as service areas. DC-HSDPA users: The dual-cell HSDPA users. Each layer corresponds to an area where the path loss from the best server exceeds a defined minimum threshold. Packet (HSPA . Coverage consists of several independent layers whose visibility in the workspace can be managed. the A-DPCH radio bearer) and an HSDPA bearer.Best Effort). 4. the input parameters as well as the formulas used in simulations and predictions (coverage predictions and point analysis).3. HSPA users require an R99 bearer (i.e. There are as many layers as defined thresholds. They have an HSDPA-capable terminal and one of these services: • • Packet (HSDPA .e. Atoll determines the best transmitter and evaluates total losses from the best transmitter. the coverage corresponds to the pixels where the signal level from this transmitter fulfils signal conditions defined in Conditions tab with different Cell edge coverage probabilities. we will use the following terms to describe the users and the services: R99 users: The Circuit (R99) and Packet (R99) service users.Variable Bit Rate). There are as many layers as defined thresholds. They have an HSPA-capable terminal and one of these services: • • • Packet (HSPA . A pixel of a service area is coloured if total losses is greater than or equal to the defined minimum thresholds (pixel colour depends on total losses).2 Definitions This section details the terms that describe the users and the services. Coverage consists of several independent layers whose visibility in the workspace can be managed. The pixel colour depends on the number of servers.1 Glossary In this chapter.Best Effort). They require an R99 bearer.Atoll 3. Packet (HSDPA . an HSDPA bearer and an HSUPA bearer. There is one coverage area per cell edge coverage probability in the explorer. Where other service areas overlap the studied one. A pixel of a service area is coloured if the total losses is greater than or equal to the defined thresholds (pixel colour depends on total losses). the coverage corresponds to the pixels where the best signal level received fulfils signal conditions defined in Conditions tab.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Total Losses (dB) Atoll calculates total losses from the transmitter on each pixel of each transmitter service area. HSPA users: The users that support both HSDPA and HSUPA. There are as many layers as defined thresholds. A pixel of a service area is coloured if the path loss is greater than or equal to the defined thresholds (pixel colour depends on path loss). the E-DPCCH/A-DPCH radio bearer). Best Cell Edge Coverage Probability (%) On each pixel of each transmitter service area. Coverage consists of several independent layers whose visibility in the workspace can be managed.2. Best Server Path Loss (dB) Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. There is one coverage area per transmitter in the explorer. Number of Servers Atoll evaluates how many service areas cover a pixel in order to determine the number of servers. HSDPA users: The users that only support HSDPA. The R99 A-DPCH bearer is transmitted on one of the cells.Variable Bit Rate). which is 214 . Users with dual-cell HSDPA-capable terminals that can simultaneously connect to two HSDPA cells of the transmitter for data transfer.Constant Bit Rate). Best Server Total Losses (dB) Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. 4. Packet (HSPA . Coverage consists of several independent layers whose visibility in the workspace can be managed. HSDPA users require an R99 bearer (i. Atoll determines the best transmitter and evaluates path loss from the best transmitter. Each layer corresponds to an area where the total losses from the best server exceed a defined minimum threshold. Where service areas overlap the studied one. Each layer shows the different total losses levels in the transmitter service area. Cell Edge Coverage Probability (%) On each pixel of each transmitter service area. CBR services do not support multi-cell HSDPA mode. 4. MC-HSDPA users: The multi-cell HSDPA users.3. which is called the anchor cell. Users with dual-band multi-cell HSDPA-capable terminals that can simultaneously connect to several HSDPA cells on co-site transmitters using different frequency bands. which is called the anchor cell. CIO  Txi ic  Cell parameter Cell Individual Offset for a transmitter on a given carrier ic. Users with multi-cell HSDPA-capable terminals that can simultaneously connect to several HSDPA cells of the transmitter for data transfer.). Name Value Unit Description F ortho Clutter parameter None Orthogonality factor Tx Site equipment parameter None MUD factor F MUD Terminal parameter . The user can be assigned an HSDPA bearer in each of the cells. quality targets. The R99 A-DPCH bearer is transmitted on one of the cells. None Used for best serving cell selection in UMTS and HSPA specific predictions. The user can be assigned an HSDPA bearer in each of the cells. DB-MC-HSDPA users: The dual-band multi-cell HSDPA users. The user can be assigned an HSDPA bearer in each of the cells. RSCP min  Txi ic  Cell parameter or Global parameter W The minimum pilot RSCP required for a user to be connected to the transmitter on a given carrier  E----c  I 0  threshold Mobility parameter None Ec/I0 target on downlink for the best serving cell Global parameter None Pilot RSCP threshold for compressed mode activation Global parameter None Ec/I0 threshold for compressed mode activation AS_Th  Txi ic  M HO  Txi ic  req Q pilot CM – activation RSCP pilot CM – activation Q pilot 215 . The R99 A-DPCH bearer is transmitted on one of the cells.2. etc. active set management conditions. None Used for best serving cell selection in UMTS and HSPA specific predictions.HSDPA properties None MUD factor cn first Frequency band parameter None First carrier number cnlast Frequency band parameter None Last carrier number cn Frequency band parameter None Carrier number step F MUD Term ic Frequency band parameter None Carrier rank of the current carrier calculated as follows: cn – cn first . CBR services: Constant Bit Rate services. VBR services: Variable Bit Rate services.2 Inputs This table lists simulation and prediction inputs (calculation options.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 called the anchor cell.– cn lower ic =  ------------------------ cn  Where cn lower is the number of carrier numbers lower than cn including excluded carriers and carriers of other frequency bands Cell parameter Threshold for macro diversity None specified for a transmitter on a given carrier ic Cell parameter Handover margin for a transmitter on a given carrier ic.Atoll 3. BE services: Best Effort services. site equipment) parameter None Number of channel elements consumed by the HSUPA bearer on uplink Max Site parameter kbps Maximum Iub backhaul throughput for a site in the uplink TP Iub – DL  N I  Max Site parameter kbps Maximum Iub backhaul throughput for a site in the downlink TP Iub – UL  N I  Simulation result kbps Iub backhaul throughput for a site in the uplink TP Iub – DL  N I  Simulation result kbps Iub backhaul throughput for a site in the downlink Site equipment parameter kbps Iub throughput required by the cell for common channels in the downlink HSDPA Site equipment parameter % HSDPA Iub backhaul overhead E1  T1  Ethernet Site equipment parameter kbps Throughput carried by an E1/T1/ Ethernet link R99 – T CH (R99 bearer.UL overhead resources for common channels/cell None Number of channel elements used by the cell for common channels on uplink Site equipment parameter . Mobility) parameter  N t req None Eb/Nt target on downlink Global parameter None Downlink Eb/Nt target increase due to compressed mode activation E b  --- N t req (Reception equipment. site equipment) parameter kbps Iub backhaul throughput consumed by the R99 bearer in the uplink R99 – T CH (R99 bearer.3. site equipment) parameter kbps Iub backhaul throughput consumed by the HSUPA bearer in the uplink N Codes  Txi ic  Simulation constraint None Maximum number of 512 bit-length OVSF codes available per cell (512) N Codes  Txi ic  Simulation result None Number of 512 bit-length OVSF codes used by the cell DL DL Q req DL Q req UL UL Q req UL Q req N CE – UL  N I  Overhead N CE – UL Overhead N CE – DL N CE – UL N CE – DL TP Iub – UL  N I  Overhead TP Iub – DL  N I  Overhead Iub TP TPIub – UL TPIub – DL TP Iub Max 216 ©Forsk 2015 .DL overhead resources for common channels/cell None Number of channel elements used by the cell for common channels on downlink R99 – T CH (R99 bearer.Atoll 3. R99 bearer. Mobility) parameter None Eb/Nt target on uplink Global parameter None Uplink Eb/Nt target increase due to compressed mode activation Max Site parameter None Number of channel elements available for a site on uplink N CE – DL  N I  Max Site parameter None Number of channel elements available for a site on downlink N CE – UL  N I  Simulation result None Number of channel elements of a site consumed by users on uplink N CE – DL  N I  Simulation result None Number of channel elements of a site consumed by users on downlink Site equipment parameter .0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks Name Value Unit Description E b  ---(Reception equipment. site equipment) parameter None Number of channel elements used for R99 traffic channels on downlink N CE HSUPA (HSUPA bearer. site equipment) parameter None Number of channel elements used for R99 traffic channels on uplink R99 – T CH (R99 bearer. R99 bearer. site equipment) parameter kbps Iub backhaul throughput consumed by the R99 bearer in the downlink HSUPA (HSUPA bearer. Atoll 3. it is assumed that there is no inter-technology downlink interferences due to external transmitters None Inter-technology downlink noise rise None Inter-technology uplink noise rise None Interference reduction factor between two adjacent carriers ic and ic adj Inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the None frequency gap between ic i (external network) and ic UL Cell parameter (user-defined or simulation result) % Total uplink load factor of the cell UL Simulation result % Uplink cell load contribution due to R99 traffic X HSUPA Cell parameter % Uplink cell load contribution due to HSUPA traffic UL Simulation constraint (global parameter or cell parameter) % Maximum uplink load factor of the cell Simulation constraint (global parameter or cell parameter) % Maximum percentage of used power W Thermal noise at transmitter W Thermal noise at terminal bps Chip rate X X R99 UL X max DL %Power max Tx UL Tx NF Tx  K  T  W  NR inter – techno log y Term NF Term  K  T  W  NR inter – techno log y N0 N0 Rc Tx DL W  10 –3 W UL Site equipment parameter f rake efficiency DL Terminal parameter None Downlink rake receiver efficiency factor TP P – DL R99 R99 bearer parameter kbps R99 bearer downlink peak throughput F spreading  Active user  R99 bearer parameter None Downlink spreading factor for active users DL R99 bearer parameter None Downlink spreading factor for inactive users f rake efficiency DL F spreading  Inactive user  None Uplink rake receiver efficiency factor 217 .0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Name Overhead N Codes HSPDSCH – Min N Codes Unit Description Site equipment parameter .DL overhead resources for common channels/cell None Number of 256 bit-length OVSF codes used by the cell for common channels  Txi ic  Cell parameter (for HSDPA only) Maximum number of 16 bit-length None OVSF codes available per cell for HSPDSCH  Txi ic  Cell parameter (for HSDPA only) Minimum number of 16 bit-length None OVSF codes available per cell for HSPDSCH HSPDSCH – Max N Codes Value NF term Terminal parameter None Terminal Noise Figure NF Tx Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) None Transmitter Noise Figure K 1.84 MHz Hz Spreading Bandwidth NRinter – techno log y Cell parameter Tx DL NRinter – techno log y Cell parameter Only used in uplink interference-based calculations of the MonteCarlo simulation RF  ic ic adj  Network parameter If not defined. it is assumed that there is no inter-carrier interference Tx UL Tx m ICP ic  ic i Network parameter If not defined.38 10-23 J/K Boltzman constant T 293 K Ambient temperature W 3.3. ic) in the downlink DL Gp UL DL TP P – RLC  I HSDPABearer  DL Without MIMO: TP P – RLC  I HSDPABearer  DL TPP – RLC  Tx ic  DL With MIMO (transmit diversity): TP P – RLC  I HSDPABearer  With MIMO (spatial multiplexing): DL TP P – RLC  I HSDPABearer  218 Max   1 + f SM – Gain   G SM – 1    .Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Name Value Unit Description rc DL R99 bearer parameter None Ratio between DPCCH and DPCH transmission duration on downlink DPCCH and DPCH respectively refer to the Dedicated Physical Control Channel and Dedicated Physical Channel TP max – DL Cell parameter kbps Maximum connection throughput per user on downlink TP P – UL R99 R99 bearer parameter kbps R99 bearer uplink peak throughput f act UL Service parameter kbps Uplink activity factor for the service f act DL Service parameter kbps Downlink activity factor for the service f act –ADPCH UL Service parameter kbps Uplink activity factor on E-DPCCH channels f act –ADPCH DL Service parameter kbps Downlink Activity factor on A-DPCH channel TPD min – UL Service parameter kbps Minimum required bit rate that the service should have in order to be available in the uplink TPD min – DL Service parameter kbps Minimum required bit rate that the service should have in order to be available in the downlink TPD max – UL Service parameter kbps Maximum bit rate that the service can require in the uplink TPD max – DL Service parameter kbps Maximum bit rate that the service can require in the downlink Ratio between the DPCCH and DPCH powers transmitted on uplink DPCCH and DPCH respectively refer None to the Dedicated Physical Control Channel and Dedicated Physical Channel rc UL R99 bearer parameter TP max – UL Cell parameter kbps Maximum connection throughput per user on uplink W ----------------R99 TP P – DL None Service downlink processing gain Gp W ----------------R99 TP P – UL None Service uplink processing gain I HSDPABearer HSDPA bearer parameter None Index of the HSDPA bearer obtained by the user in the cell (Txi.ic) HSDPA bearer parameter kbps Peak RLC throughput supported by the HSDPA bearer kbps Peak RLC throughput provided to the user in the cell (Txi.3. 3.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Name Value Unit Description kbps Peak RLC throughput provided to the user in the downlink TPD min – DL ----------------------------------------------------DL TP P – RLC  I HSDPABearer  % HSDPA bearer consumption for a CBR service user HSDPA study result kbps Effective RLC throughput supported by the HSDPA bearer HSDPA study result kbps Average effective RLC throughput supported by the HSDPA bearer DL HSDPA study result kbps Peak MAC throughput supported by the HSDPA bearer DL HSDPA study result kbps Effective MAC throughput supported by the HSDPA bearer DL HSDPA study result kbps User application throughput on downlink TP A UL HSUPA study result kbps User application throughput on uplink TP Av – A UL HSUPA study result kbps User average application throughput on uplink I HSUPABearer HSUPA Bearer parameter None Index of the HSUPA bearer obtained in the cell (Txi.ic) HSDPA study result For single-carrier HSDPA users DL TP P – RLC  Tx ic  DL TP P – RLC For DC-HSDPA users  DL TP P – RLC  Tx ic  ic  Tx C HSDPABearer DL TP E – RLC DL TP Av – E – RLC TP P – MAC TP E – MAC TP A N Rtx  I HSUPABearer  HSUPA bearer selection parameter Maximum number of retransmissions a HARQ process will kbps perform for a block of data before moving on to a new block of data.ic) in the uplink HSUPA study result TP P – RLC TP P – RLC  I HSUPABearer  C HSUPABearer TPD min – UL ----------------------------------------------------UL TP P – RLC  I HSUPABearer  % HSUPA bearer consumption for a CBR service user HSUPA study result kbps Minimum effective RLC throughput supported by the HSUPA bearer TP Av – E – R LC HSUPA study result kbps Average effective RLC throughput supported by the HSUPA bearer TP P – M AC UL HSUPA study result kbps Peak MAC throughput supported by the HSUPA bearer TP Offset Service parameter (for HSDPA only) kbps Throughput offset f TP – Scaling Service parameter (for HSDPA only) % Scaling factor P max  Txi  Transmitter parameter W Maximum shared power Available only if the inter-carrier power sharing option is activated P SCH  Txi ic  Cell parameter W Cell synchronisation channel power UL TP Min –E – R LC UL UL 219 . for the HSUPA bearer index UL TP P – RLC  I HSUPABearer  UL HSUPA bearer parameter kbps Peak RLC throughput supported by the HSUPA bearer kbps Peak RLC throughput provided to the user in the cell (Txi. 0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Name Value Unit Description P OtherCCH  Txi ic  Cell parameter W Cell other common channels (except CPICH and SCH) power P pilot  Txi ic  Cell parameter W Cell pilot power P HSDPA  Txi ic  Cell parameter (user-defined or simulation result) (for HSDPA only) P HS – PDSCH  Txi ic  + n HS – SCCH  P HS – SCCH  Txi ic  W Available cell HSDPA power HSDPA: High Speed Downlink Packet Access P HS – PDSCH  Txi ic  Simulation result (for HSDPA only) W Cell HS-PDSCH power HS-PDSCH: High Speed Physical Downlink Shared Channel P HS – SCCH  Txi ic  Cell parameter (for HSDPA only) W Cell HS-SCCH power HS-SCCH: High Speed Shared Control Channel n HS – SCCH Cell parameter (for HSDPA only) P Headroom  Txi ic  Cell parameter (for HSDPA only) W Cell headroom power P max  Txi ic  Cell parameter W Maximum Cell power P tch  Txi ic  Simulation result W R99 traffic channel power transmitted on carrier ic min R99 bearer parameter W Minimum power allowed on R99 traffic data channel P tch max R99 bearer parameter W Maximum power allowed on R99 traffic data channel P HSUPA  Txi ic  Cell parameter W Cell HSUPA power HSUPA: High Speed Uplink Packet Access P tx –H SDPA  Txi ic  Simulation result W Transmitter HSDPA power transmitted on carrier ic W Transmitter R99 power transmitted on carrier ic P tch number of HS-SCCH channels managed by the cell Simulation result P pilot  Txi ic  + P SCH  Txi ic  + P OtherCCH  Txi ic  + P tx – R99  Txi ic    P tch  Txi ic  + tch(ic) used for R99 users DL P tch  Txi ic   f act –ADPCH tch(ic) used for HSUPA users P tx  Txi ic  Cell parameter (user-defined or simulation result) P tx – R99  Txi ic  + P tx – H SDPA  Txi ic  + P HSUPA  Txi ic  W Transmitter total power transmitted on carrier ic P term – R99 Calculated in the simulation but not displayed W Terminal power transmitted to obtain the R99 radio bearer P term – HSUPA Calculated in the simulation but not displayed W Terminal power transmitted to obtain the HSUPA radio bearer W Total power transmitted by the terminal Simulation result P term P term – R99  UL f act – ADPCH + P term – HSUPA for HSPA users P term – R99 for R99 and HSDPA users 220 P term min Terminal parameter W Minimum terminal power allowed P term max Terminal parameter W Maximum terminal power allowed  BTS BTS parameter % Percentage of BTS signal correctly transmitted  term Terminal parameter % Percentage of terminal signal correctly transmitted  Clutter parameter % Percentage of pilot finger percentage of signal received by the terminal pilot finger .3.Atoll 3. DL Result calculated from cell edge coverage probability and DL Eb/Nt standard deviation None DL Eb/Nt Shadowing margin Only used in prediction studies UL Result calculated from cell edge coverage probability and UL Eb/Nt standard deviation None UL Eb/Nt Shadowing margin Only used in prediction studies None UL quality gain due to signal diversity in soft handoffc. None Random shadowing error drawn during Monte-Carlo simulation Only used in simulations G Div G Div DL DL M Shadowing –  Eb  Nt  n=2 or 3 UL UL G macro – diversity E Shadowing npaths G macro – diversity = M Shadowing – Ec  Io – M Shadowing – Ec  Io DL G macro – diversity M Shadowing –  Eb  Nt  Indoor loss npaths G macro – diversity = M Shadowing –  Eb  Nt  UL – M Shadowing – Eb  Nt  n=2 or 3 Global parameter (default value) Simulation result UL 221 .Depends on the transmitter Rx diversity None Gain due to receive diversity G SM Max MIMO configuration parameter dB Maximum spatial multiplexing gain for a given number of transmission and reception antennas G TD DL MIMO configuration parameter dB Downlink Transmit Diversity gain for a given number of transmission and reception antenna ports f SM – Gain Clutter parameter None Spatial multiplexing gain factor G TD Clutter parameter dB Additional diversity gain in downlink L Tx Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) None Transmitter lossa L body Service parameter None Body loss L Term Terminal parameter None Terminal loss L indoor Clutter parameter L path Propagation model result None Path loss M Shadowing – model Result calculated from cell edge coverage probability and model standard deviation None Model Shadowing margin Only used in prediction studies M Shadowing – Ec  Io Result calculated from cell edge coverage probability and Ec/I0 standard deviation None Ec/I0 Shadowing margin Only used in prediction studies None DL gain due to availability of several pilot signals at the mobile b.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Name Value Unit Description G Tx Antenna parameter None Transmitter antenna gain G Term Terminal parameter None Terminal gain DL R99 bearer parameter .3.Depends on the transmitter Tx diversity None Gain due to transmit diversity UL R99 bearer parameter .Atoll 3. b.3. TD-SCDMA. Therefore. WiMAX. Name Value I intra  txi ic  P SCH  txi ic   DL DL P tot  txi ic  –  BTS     P tot  txi ic  – ---------------------------- L DL 222 T or Unit Description W Downlink intra-cell interference at terminal on carrier ic . In uplink prediction studies. CDMA2000. npaths c.2. M Shadowing –  Eb  Nt  d. For information on calculating transmitter losses on uplink and downlink.3 Ec/I0 Calculation This table details the pilot quality ( Q pilot or Ec  Io ) calculations.Atoll 3. only carrier power level is downgraded by the shadowing margin ( M Shadowing –  Eb  Nt  UL corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in UL ). npaths M Shadowing –Ec  Io corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in case of downlink Ec/I0 modelling. M Shadowing –  Eb  Nt  DL DL or M Shadowing – Ec  Io is set to 1 in downlink extra-cell interference calculation. see "UMTS.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Name Value Unit Description None Transmitter-terminal total loss P pilot  Txi ic  -------------------------------LT W Chip power received at terminal DL P tch  Txi ic  ----------------------------LT W Bit power received at terminal on carrier ic DL P tx  Txi ic  --------------------------LT W Total power received at terminal from a transmitter on carrier ic P tch  Txi ic  ----------------------------LT W Total power received at terminal from traffic channels of a transmitter on carrier ic P term -----------LT W Bit power received at transmitter on carrier ic used by terminal P term – R99 -----------------------LT W Bit power received at transmitter on carrier ic used by terminal W Bit power received at transmitter on DPDCH from a terminal on carrier ic In prediction studiesd For Ec/I0 calculation L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io ------------------------------------------------------------------------------------------------------------------------------------G Tx  G term For DL Eb/Nt calculation L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  DL ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term LT For UL Eb/Nt calculation L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term In simulations L path  L Tx  L term  L body  L indoor  E Shadowing ------------------------------------------------------------------------------------------------------------------G Tx  G term P c  Txi ic  P b  Txi ic  P tot  Txi ic   DL P traf  Txi ic  tch  ic  UL P b  ic  UL P b – R99  ic  UL UL P b – DPDCH  ic  a. 4. and LTE Documents" on page 30. carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing –  Eb  Nt  M Shadowing – Ec  Io ) while extra-cell interference level is not. case of uplink soft handoff modelling. UL P b – R99  ic    1 – r c  L Tx = L total – UL on uplink and L Tx = L total – DL on downlink. In downlink prediction studies. Atoll 3. 4. 223 . In the case of an interfering GSM external network in frequency hopping.2.4 DL Eb/Nt Calculation Eb DL This table details calculations of downlink traffic channel quality ( Q tch or  ------ ).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Name Value  DL I extra  ic  DL P tot  txj ic  Unit Description W Downlink extra-cell interference at terminal on carrier ic W Downlink inter-carrier interference at terminal on carrier ic txj j  i  Ptot  txj icadj  DL DL I inter – carrier  ic  txj j --------------------------------------------- RF  ic ic adj   DL I inter – techno log y  ic  ni Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic W Downlink inter-technology interference at terminal on carrier ic a i Without Pilot: DL DL DL DL I intra  txi ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  Term + N0 DL I 0  ic  –  1 –     BTS  P c  txi ic  DL Total noise: DL W Total received noise at terminal on carrier ic None Quality level at terminal on pilot for carrier ic DL P tot  txi ic  + I extra  ic  + I inter – carrier  ic  DL Term + I inter – techno log y  ic  + N 0  BTS    P c  txi ic  -------------------------------------------------DL I 0  ic  E Q pilot  txi ic    ----c I0 a. G DL Div  G p DL DL N tot  ic  –  1 – F ortho    BTS  P b  txi ic  DL  BTS  P b  txi ic  DL . G DL Total Noise: ------------------------------------------Div  G p DL N tot  ic  DL Q  ic  DL f rake efficiency   tx k  ActiveSet DL Q tch  tx k ic  Quality level at terminal using carrier ic due to combination of all None transmitters of the active set (Macro-diversity conditions).3. Nt DL Name Value Unit Description I intra  txi ic  P SCH  txi ic   DL DL - P tot  txi ic  –  BTS  F ortho   P tot  txi ic  – ----------------------------L W Downlink intra-cell interference at terminal on carrier ic W Downlink extra-cell interference at terminal on carrier ic W Downlink inter-carrier interference at terminal on carrier ic DL T  DL I extra  ic  DL P tot  txj ic  txj j  i  Ptot  txj icadj  DL DL I inter – carrier  ic  txj j --------------------------------------------RF  ic ic adj  Tx P Transmitted  ic i   ------------------------------------Tx Tx m L  ICP DL I inter – techno log y  ic  DL N tot  ic  DL DL ic i ic total ni W DL DL Downlink inter-technology interference at terminal on carrier ic a Term I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 W Total received noise at terminal on carrier ic None Quality level at terminal on a traffic channel from one transmitter on carrier ic b Without useful signal: DL E DL Q tch  txi ic    ----b-  N t DL  BTS  P b  txi ic  DL -----------------------------------------------------------------------------------------------------. the ICP value is weighted according to the fractional load. 0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Name Value Unit Description None Soft handover gain on downlink W Required transmitter traffic channel power to achieve Eb/Nt target at terminal on carrier ic DL G SHO Q  ic  --------------------------------------------------DL Q tch  BestServer ic  req P tch  txi Q req ----------------. ic  In the case of an interfering GSM external network in frequency hopping. G UL Total noise: --------------------------------------------------Div  G p UL N tot  txi ic  . P tch  txi ic  DL Q  ic  DL DL a. 4.2. Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt. G UL Div  G p UL Tx UL N tot  txi ic  –  1 – F MUD    term  P b  ic  UL  term  P b – DPDCH  ic  UL . Nt UL Name Value  Pb UL UL intra I tot UL extra I tot  txi ic   ic  term txi   txi ic  Unit Description W Total power received at transmitter from intra-cell terminals using carrier ic W Total power received at transmitter from extra-cell terminals using carrier ic W Uplink inter-carrier interference at terminal on carrier ic UL P b  ic  term txj j  i  Pb UL UL I inter – carrier  txi ic  UL I tot  txi ic  UL N tot  txi ic   ic adj  term txj j ----------------------------------RF  ic ic adj  UL extra I tot UL intra Tx  txi ic  +  1 – F MUD   term  I tot UL I tot  txi ic  + UL  txi ic + I inter – carrier  txi W ic  tx N0 Total received interference at transmitter on carrier ic W Total noise at transmitter on carrier ic (Uplink interference) None Quality level at transmitter on a traffic channel for carrier ic a Without useful signal: UL E UL Q tch  txi ic    ----b-  N t UL 224  term  P b – DPDCH  ic  UL -------------------------------------------------------------------------------------------------------.5 UL Eb/Nt Calculation Eb UL This table details calculations of uplink traffic channel quality ( Q tch or  ------ ). The chosen option will be taken into account only in simulations. the ICP value is weighted according to the fractional load. b. In point analysis and coverage studies.Atoll 3. Calculation option may be selected in the Global parameters tab.3. Power control simulation 4.3 Simulations The simulation process consists of two steps: 1. Atoll randomly assigns a shadowing error to each user using the probability distribution that describes the shadowing effect. Both active and inactive users consume radio resources and create interference. tx  othersite l UL  G macro – diversity UL Q  ic  --------------------------------------------------UL Q tch  BestServer ic  UL G SHO None Soft handover gain on uplink W Required terminal power to achieve Eb/Nt target at transmitter on carrier ic UL Q req ----------------. a mobility type.3. 4. Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt.3. The resulting user distribution complies with the traffic database and maps provided to the algorithm. according to a probability law that uses the traffic database. The user activity status is an important output of the random trial and has direct consequences on the next step of the simulation and on the network interferences.Atoll 3. a first random trial is performed to determine the number of users and their activity status. A user may be either active or inactive. Finally. 2. Calculation option may be selected in the Global parameters tab. softer/soft HO (No MRC): Max  Q UL tch  tx k tx k  ActiveSet UL ic    G macro – diversity UL Q  ic  Softer/soft HO (MRC): Quality level at site using carrier ic due to combination of all transmitters of the active set located at the same site and taking into None account increasing of the quality due to macro-diversity (macro-diversity gain).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Name Value Unit Description UL No HO: Q tch  txi ic   UL Softer HO: f rake efficiency  UL Q tch  txk ic  tx k  ActiveSet  samesite  Soft. Four activity status are modelled: • Active UL: the user is active on UL and inactive on DL • Active DL: the user is active on DL and inactive on UL • Active UL+DL: the user is active on UL and on DL • Inactive: the user is inactive on UL and on DL 225 . and an activity status by random trial. Each user is assigned a service. Obtaining a realistic user distribution Atoll generates a user distribution using a Monte-Carlo algorithm. another random trial determines user positions in their respective traffic zone and whether they are indoors or outdoors (according to the clutter weighting and the indoor ratio per clutter class defined for the traffic maps). The chosen option will be taken into account only in simulations. Then. In point analysis and coverage studies. which requires traffic maps and data as input. P term UL Q  ic  req P term  ic  a.txl  ActiveSet  f rake efficiency    tx k  samesite   tx k Max UL  In simulations G macro – diversity = 1 .1 Generating a Realistic User Distribution During the simulation.    UL  UL UL Q tch  tx k ic  Q tch  tx l ic  txk . • The average duration of a call (seconds) d .3. This may lead to slight variations in the total numbers of users in different simulations. If the map is composed of points.1 Simulations Based on User Profile Traffic Maps User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of user profiles with an associated mobility type and a given density (number of subscribers with the same profile per km²). • Calculation of activity probabilities: UL DL Probability of being inactive on UL and DL: p inactive =  1 – f act    1 – f act  UL DL DL UL Probability of being active on UL only: p UL = f act   1 – f act  Probability of being active on DL only: p DL = f act   1 – f act  UL DL Probability of being active both on UL and DL: p UL + DL = f act  f act UL DL Where. For each behaviour described in a user profile. according to the service. The user profile models the behaviour of the different subscriber categories. f act and f act are respectively the UL and DL activity factors defined for the circuit switched service i. the number of subscribers (X) per user profile is calculated from the line length (L) and the user profile density (D) (nb of subscribers per km) as follows: X = L  D The number of subscribers (X) is an input when a user profile traffic map is composed of points.ini file: [Simulation] RandomTotalUsers=0 4. add the following lines in the Atoll.1 Circuit Switched Service (i) User profile parameters for circuit switched services are: • • The used terminal (equipment used for the service (from the Terminals table)).3.1.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 The determination of the number of users and the activity status allocation depend on the type of traffic cartography used. Atoll calculates the probability for the user being active in uplink and in downlink at an instant t.Atoll 3. The average number of calls per hour N call . User profile traffic maps: Each polygon and line of the map is assigned a density of subscribers with given user profile and mobility type. Atoll determines the total number of users attempting connection in each simulation based on the Poisson distribution. X = SD • • When user profile traffic maps are composed of lines. frequency use and exchange volume. The number of users and their distribution per activity status is determined as follows: • Calculation of the service usage duration per hour ( p 0 : probability of a connection): N call  d p o = ------------------3600 • Calculation of the number of users trying to access the service i ( n i ): ni = X  p0 Next. we can take into account activity periods during the connection in order to determine the activity status of each user.1. 4. • 226 Calculation of number of users per activity status: .1. a number of subscribers (X) per user profile is inferred. Each user profile contains a list of services and their associated parameters describing how these services are accessed by the subscriber. To have the same total number of users in each simulation of a group.3. each point is assigned a number of subscribers with given user profile and mobility type. From environment (or polygon) surface (S) and user profile density (D). DL S packet : Packet size (Bytes) on downlink.3. or active on UL only. UL T packet – call : Average time (millisecond) between two packets calls on the uplink . DL T packet : Average time (millisecond) between two packets on the downlink . DL T packet – call : Average time (millisecond) between two packets calls on the downlink . Each packet call is defined by its size and may be divided in packets of fixed size (1500 Bytes) separated by an inter arrival time. a packet session is described by following parameters: UL N packet –c all : Average number of packet calls on the uplink during a session. 4. A packet session consists of several packet calls separated by a reading time.1. The average number of packet sessions per hour N sess . Figure 4. or inactive on both links. UL S packet : Packet size (Bytes) on uplink. or active on DL only.1: Description of a Packet Session The number of users and their distribution per activity status is determined as follows: • Calculation of the average packet call size (kBytes): V UL V DL UL DL S packet –c all = ---------------------------------------and S packet –c all = ---------------------------------------UL UL DL DL N packet –c all  f eff N packet –c all  f eff UL DL Where f eff and f eff are the UL and DL efficiency factors defined for the packet switched service j. In Atoll.3.2 Packet Switched Service (j) User profile parameters for packet switched services are: • • The used terminal (equipment used for the service (from the Terminals table)). a user when he is connected can have four different activity status: either active on both links.1.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 inactive Number of inactive users on UL and DL: n i = n i  p inactive Number of users active on UL and inactive on DL: n i  UL  = n i  p UL Number of users active on DL and inactive on UL: n i  DL  = n i  p DL Number of users active on UL and DL both: n i  UL + DL  = n i  p UL + DL Therefore. 227 . • The volume (in kbytes) which is transferred on the downlink V DL and the uplink V UL during a session. UL T packet : Average time (millisecond) between two packets on the uplink . DL N packet –c all : Average number of packet calls on the downlink during a session. packet are downloaded (no packet is uploaded). the probability of being connected is: UL DL p Connection   1 – p Connection  UL p Connected = ----------------------------------------------------------------------p Connected • 3rd case: At a given time.and  D DL  D Activity  session = N packet –c all  -----------------------------------------------Activity  session = N packet – c all  -----------------------------------------------UL DL TP Av  1000 TP Av  1000 UL DL Where TP Av and TPAv are the uplink and downlink average requested throughputs defined for the service j. the number of users who want to get the service j is: n j = X  p Connected As you can see on the picture above. f eff and f eff are set to 1. we have to consider three possible cases when a user is connected: • 1st case: At a given time.3. In this case. Here.and  D Inactivity  packet – call = --------------------------------------------------------1000 1000 • Calculation of the average duration of inactivity in a session (s): UL UL UL DL DL DL  D Inactivity  session = N packet –c all   D Inactivity  packet – call and  D Inactivity  session = N packet –c all   D Inactivity  packet – call • Calculation of the average duration of activity in a session (s): UL UL DL DL N packet  S packet  8 N packet  S packet  8 UL UL DL . the average duration of a connection (in s) is: UL UL UL DL DL DL D Connection =  D Activity  session +  D Inactivity  session and D Connection =  D Activity  session +  D Inactivity  session • Calculation of the service usage duration per hour (probability of a connection): N sess N sess UL UL DL DL p Connection = -----------. D Connection and p Connection = -----------. D Connection 3600 3600 • Calculation of the probability of being connected: UL DL p Connected = 1 –  1 – p Connection    1 – p Connection  Therefore. • Calculation of the average number of packets per packet call: UL DL  S packet –c all   S packet – c all  UL - + 1 and N DL - + 1 N packet = int  ------------------------------packet = int  ------------------------------UL  S packet  1024  S DL packet  1024 1kBytes = 1024Bytes. • Calculation of the average duration of inactivity within a packet call (s): UL UL DL DL  N packet – 1   T packet  N packet – 1   T packet UL DL  D Inactivity  packet – call = --------------------------------------------------------.Atoll 3. the probability of being connected is: DL UL p Connection   1 – p Connection  DL p Connected = ----------------------------------------------------------------------p Connected 228 .0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 UL DL For packet (HSDPA) and packet (HSPA) services. packets are downloaded and uploaded. the probability of being connected is: UL DL p Connection  p Connection UL + DL p Connected = --------------------------------------------------------p Connected • 2nd case: At a given time. In this case. packet are uploaded (no packet is downloaded). Therefore. will correspond to calculated distributions. And the service and the activity status of each user are randomly drawn in each simulation. packet are downloaded (no packet is uploaded). active on DL and active on UL and DL users. this probability is: 2 p UL = f UL UL  p Connected The user can be inactive on both links. The user distribution per service and the activity status distribution between the users are average distributions. But if you check each simulation. this probability is: 3 p DL = f DL DL  p Connected The user can be inactive on both links.3.Atoll 3. the user distribution between services as well as the activity status distribution between users is different in each of them. Therefore. respectively. the average number of users per service and average numbers of inactive. this probability is: 1 p UL = f UL DL UL + DL   1 – f   p Connected The user can be active on DL and inactive on UL. or active on DL only.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Now. active on UL. a user when he is connected can have four different activity status: either active on both links. this probability is: 1 p UL + DL = f UL f DL UL + DL  p Connected The user can be inactive on both links. this probability is: UL 1 DL UL + DL p inactive =  1 – f    1 – f   p Connected • 2nd case: At a given time. The user can be active on DL and inactive on UL. packets are downloaded and uploaded. this probability is: UL 2 UL p inactive =  1 – f   p Connected • 3rd case: At a given time. we have to take into account activity periods during the connection in order to determine the activity status of each user. The user can be active on UL and inactive on DL. The user can be active on UL and inactive on DL. this probability is: DL 3 DL p inactive =  1 – f   p Connected • Calculation of number of users per activity status inactive Number of inactive users on UL and DL: n j 1 2 3 = n j   p inactive + p inactive + p inactive  1 2 1 3 Number of users active on UL and inactive on DL: n j  UL  = n j   p UL + p UL  Number of users active on DL and inactive on UL: n j  DL  = n j   p DL + p DL  1 Number of users active on UL and DL: n j  UL + DL  = n j  p UL + DL Therefore. or inactive on both links. 229 . • f UL Calculation of the probability of being active: UL DL  D Activity  session  D Activity  session DL = -------------------------------------------------------------------------------------------and f = -------------------------------------------------------------------------------------------UL UL DL DL   D Inactivity  session +  D Activity  session    D Inactivity  session +  D Activity  session  Therefore. packet are uploaded (no packet is downloaded). we have: • 1st case: At a given time. if you compute several simulations at once. or active on UL only. this probability is: 1 p DL = f DL UL UL + DL   1 – f   p Connected The user can be active on both links. 2 Simulations Based on Sector Traffic Maps Sector traffic maps can be based on live traffic data from OMC (Operation and Maintenance Centre).1. Then. Rt Rt DL TP Av is the downlink average requested throughput defined for the service. And users active in both links (ni(UL+DL)).and UL TP Av DL Rt N DL = ---------DL TP Av UL is the kbits per second transmitted in UL in the Txi cell to supply the service. Users active in downlink and inactive in uplink (ni(DL)). or active in UL only. Atoll takes into account activity periods during the connection in order to determine the activity status of each user. NUL and NDL values include: • • • Users active in uplink and inactive in downlink (ni(UL)). we have: N UL  p UL + DL N DL  p UL + DL Number of users active in UL and DL both: n i  UL + DL  = min  --------------------------------- -------------------------------- p UL + p UL + DL p DL + p UL + DL Number of users active in UL and inactive in DL: n i  UL  = N UL – n i  UL + DL  Number of users active in DL and inactive in UL: n i  DL  = N DL – n i  UL + DL  inactive Number of inactive users in UL and DL: n i  n j  UL  + n j  DL  + n j  UL + DL   .3. UL TP Av is the uplink average requested throughput defined for the service. 230 .Atoll 3. a connected user can have four different activity status: either active in both links.1 Throughputs in Uplink and Downlink When selecting Throughputs in Uplink and Downlink. f act and f act are respectively the UL and DL activity factors defined for the service i.2. you can input the throughput demands in the uplink and downlink for each sector and for each listed service.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 4.3. or inactive in both links. 4. or active in DL only. Atoll calculates the number of users per activity status: We have:  p UL + p UL + DL    n j  UL  + n j  DL  + n j  UL + DL   = N UL  p DL + p UL + DL    n j  UL  + n j  DL  + n j  UL + DL   = N DL Therefore. DL is the kbits per second transmitted in DL in the Txi cell to supply the service. p inactive = -----------------------------------------------------------------------------1 – p inactive Therefore.1.3. Activity probabilities are calculated as follows: UL DL Probability of being inactive in UL and DL: p inactive =  1 – f act    1 – f act  UL DL DL UL Probability of being active in UL only: p UL = f act   1 – f act  Probability of being active in DL only: p DL = f act   1 – f act  UL DL Probability of being active both in UL and DL: p UL + DL = f act  f act UL DL Where. Atoll calculates the number of users active in uplink and in downlink in the Txi cell using the service (NUL and NDL) as follows: UL Rt N UL = ---------. Traffic is spread over the best server coverage area of each transmitter and each coverage area is assigned either the throughputs in the uplink and in the downlink or the number of users per activity status or the total number of users (including all activity statuses). Atoll takes into account activity periods during the connection in order to determine the activity status of each user.2 Power Control Simulation The power control algorithm simulates the way a UMTS network regulates itself by using uplink and downlink power controls in order to minimize interference and maximize capacity. you can directly input the number of inactive users ( n i ). active set and handoff status for each terminal. Atoll calculates the number of users per activity status: inactive Number of inactive users in UL and DL: n i = n i  p inactive Number of users active in UL and inactive in DL: n i  UL  = n i  p UL Number of users active in DL and inactive in UL: n i  DL  = n i  p DL Number of users active in UL and DL both: n i  UL + DL  = n i  p UL + DL Therefore.2 Total Number of Users (All Activity Statuses) When selecting Total Number of Users (All Activity Statuses). HSDPA users are linked to the A-DPCH radio bearer (an R99 radio bearer). 4.2. The activity status distribution between users is an average distribution. 231 . for each sector and for each service. if you compute several simulations at once. But if you check each simulation. in each simulation.3. i.3.1. you can input the number of connected users for each sector and for each listed service ( n i ). the activity status distribution between users is different in each of them. average numbers of inactive.Atoll 3. or active in UL only. the network uses a A-DPCH power control on UL and DL and then it performs fast link adaptation on DL in order to select an HSDPA radio bearer. For HSPA users. Activity probabilities are calculated as follows: UL DL Probability of being inactive in UL and DL: p inactive =  1 – f act    1 – f act  UL DL DL UL Probability of being active in UL only: p UL = f act   1 – f act  Probability of being active in DL only: p DL = f act   1 – f act  UL DL Probability of being active both in UL and DL: p UL + DL = f act  f act UL DL Where. Therefore.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4. checks that there is an HSDPA connection on downlink and then carries out noise rise scheduling in order to select an HSUPA radio bearer on uplink.2. the network first uses a E-DPCCH/A-DPCH power control on UL and DL. until the convergence criteria (on UL and DL) are satisfied. In fact.3 Number of Users per Activity Status inactive When selecting Number of Users per Activity Status. active on UL. f act and f act are respectively the UL and DL activity factors defined for the service i. active on DL and active on UL and DL users correspond to the calculated distribution.3. Then. 4.1. network parameters such as cell power..e. Therefore. or active in DL only. Atoll simulates these network regulation mechanisms with an iterative algorithm and calculates. mobile terminal power. a connected user can have four different activity status: either active in both links. or inactive in both links. in the downlink ( n i  DL  ) and in the uplink and downlink ( n i  UL + DL  ). During each iteration of the algorithm. for each user distribution. the activity status of each user is randomly drawn. The process is repeated until the network is balanced. the number of users active in the uplink ( n i  UL  ). all the users selected during the user distribution generation (1st step) attempt to connect one by one to network transmitters.3. are then evaluated by the HSUPA part of the algorithm. P Tx  txi ic m  . The steps of this algorithm are detailed below.3. 232 UL extra  txi ic m  .Atoll 3.2. unless they have been rejected during the R99 or HSDPA parts of the algorithm. no connected mobile). I tot are initialised to 0 W (i. All users are evaluated by the R99 part of the algorithm.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Figure 4. the simulation algorithm is divided in three parts. Finally. HSDPA and HSPA users. I tot UL  txi ic m  and I inter – carrier  txi ic m  .1 Algorithm Initialization The total power transmitted by the base station txi on the carrier ic m . is initialised to P pilot  txi ic m  + P SCH  txi ic m  + P otherCCH  txi ic m  + P HSDPA  txi ic m  + P HSUPA  txi ic  . are then evaluated by the HSDPA part of the algorithm.2: UMTS HSPA Power Control Algorithm As shown in Figure 4. unless they have been rejected during the R99 part of the algorithm. HSPA users.e.2 on page 232.3. UL intra Uplink powers received by the base station txi on carrier ic m . 4. ic) for which the pilot RSCP exceeds the minimum pilot RSCP: P c  txi M b ic   RSCP min  txi ic  . For each mobile Mb. this carrier is referred to as the "anchor" carrier.ic).ic). 233 . we have:  X R99  txi ic m   k = ------------------------------UL N tot  txi ic m  4.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 UL I tot  txi ic m  UL . the UL DL thresholds. If a given carrier is specified for the service requested by Mb ic BS  M b  is the carrier specified for the service Else the carrier selection mode defined for the site equipment is considered. Mb has failed to be connected to the network and is rejected. For each mobile (Mb).= 0 Therefore. The algorithm applies to single frequency band networks and to multi-band networks. For each NodeB having candidate cells. k Analysis of candidate cells. For MC-HSDPA and DB-MC-HSDPA users. within the set of candidate cells of the NodeB.ic). we have the following steps: Determination of Mb’s Best Serving Cell For each transmitter txi containing Mb in its calculation area and working on a frequency band supported by the Mb’s terminal ). For each pair (txBS.3. then (txBS. calculation of the uplink load factor: UL I tot  tx BS ic  UL UL  X R99  tx BS ic  k = ------------------------------. selection of the transmitter with the highest Q pilot  txi M b ic  . depend on the user mobility type and are defined in the R99 bearer selection table.3. determination of the best carrier. icBS. (txBS. Q req and Q req . All variables are described in Definitions and formulas part.ic) is rejected by Mb k UL UL If  X R99  tx BS ic   k  X max . The bearer downgrading is not dealt with. Atoll only considers the cells (txi.    BTS  P c  txi M b ic  Calculation of Q pilot  txi ic Mb  = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------Term k DL DL DL DL P tot  txi ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 If user selects “without Pilot”    BTS  P c  txi M b ic  Q pilot  txi ic Mb  = -----------------------------------------------------------------------------------------------------------------------------------------------------------k DL DL DL  DL   I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic     Term + N0 –  1 –     BTS  P c  txi M b ic    Determination of the candidate cells. For information on how this parameter is calculated.  tx BS ic   M i  .2. In the algorithm. Xk is the value of the X (variable) at the iteration k. see "Admission Control in the R99 Part" on page 276. (txBS.+ X UL N tot  tx BS ic  UL X corresponds to the load rise due to the mobile. Rejection of bad candidate cells if the pilot is not received or if the uplink load factor is exceeded during the admission load control (if simulation respects a loading factor constraint and Mb was not connected in previous iteration) pilot If Q pilot  tx BS M b ic   Q req  Mobility  M b   then (txBS.2 R99 Part of the Algorithm The algorithm is detailed for any iteration k.ic) as good candidate cell If no good candidate cell has been selected.ic) is rejected by Mb Else Keep (txBS.Atoll 3. For each carrier ic. neighbour of BestCell k  M b     BTS  P c  txi M b ic  Calculation of Q pilot  txi M b ic  = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------DL DL DL DL Term k P tot  txi ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 If user selects “without Pilot”    BTS  P c  txi M b ic  Q pilot  txi M b ic  = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------k DL DL DL  DL  I  txi  ic  + I  ic  + I  ic  + I  ic  extra inter – carrier inter – techno log y  intra    Term + N0 –  1 –     BTS  P c  txi M b ic    Rejection of txi from the active set if difference with the best server is too high max If Q pilot  M b  – Q pilot  txi M b ic   AS_Th  BestCell k  M b   then txi is rejected k k Else txi is included in the Mb active set Rejection of a station if the mobile active set is full Station with the lowest Q pilot in the active set is rejected k EndFor Uplink Power Control R99 – req Calculation of the terminal power required by Mb to obtain the R99 radio bearer: P term For each cell (txi. ic2. and. among the adjacent carriers. selection of the second carrier. UL Load Factor” UL ic BS  M b  is the carrier where we obtain the lowest  X R99  tx BS ic   k Else if carrier selection mode is “Min. if neighbours are used. using ic .ic) of the Mb active set 234  M b ic  k . DL Total Power” ic BS  M b  is the carrier where we obtain the lowest P tx  tx BS ic  k Else if carrier selection mode is “Random” ic BS  M b  is randomly selected Else if carrier selection mode is "Sequential" UL UL ic BS  M b  is the first carrier where  X R99  tx BS ic   k  X max Endif Determination of the best serving cell. calculation of Q pilot  tx BS ic p M b  k Selection of the carrier. we will consider ic as the carrier used by the best serving cell Selection of the second serving cell for DC-HSDPA users MC-HSDPA and DB-MC-HSDPA users are processed as DC-HSDPA users.ic2) is rejected by Mb k Else Keep (txBS. with the highest Q pilot  tx BS ic p M b  k pilot If Q pilot  tx BS ic 2 M b   Q req  Mobility  M b   then (txBS. (txBS. If txBS supports multi-cell HSDPA and if it has several carriers. icp. ic2.3.icBS) max (tx BS.ic BS) k  M b  is the best serving cell ( BestCell k  M b  ) and its pilot quality is Q pilot  M b  k In the following lines.ic2) as second serving cell Active Set Determination For each station txi containing Mb in its calculation area.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 If carrier selection mode is “Min. For each carrier adjacent to the best serving carrier.Atoll 3. with the minimum power allowed on traffic channel for the Mb service req P term – R99  M b ic  k – 1 UL P b – R99  txi M b ic  = ---------------------------------------------------L T  txi M b  UL UL UL UL UL P b – DPDCH  txi M b ic  = P b – R99  txi M b ic    1 – r c  UL P b – DPCCH  txi M b ic  = P b – R99  txi M b ic   r c UL UL UL UL UL P b – R99  txi M b ic  = P b – DPCCH  txi M b ic  + P b – DPDCH  txi M b ic  if the user is active. Compressed mode is operated if Mi and Sj support compressed mode. P b – R99  txi M b ic  = P b – DPCCH  txi M b ic  if the user is inactive.3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Calculation of quality level on Mb traffic channel at (txi.ic). and Resulting CM – activation  txi M b ic   Q pilot • Either Q pilot • Or P c  txi M b ic   RSCP pilot k CM – activation if the Ec/I0 Active option is selected. UL  term  P b – DPDCH  txi M b ic  k UL UL UL Q tch  txi M b ic  k = ------------------------------------------------------------------------- G p  Service  Mb    G div UL N tot  txi ic  End For If (Mb is in not in handoff) UL UL Q k  M b  = Q tch  txi M b ic  k Else if (Mi is in softer handoff) UL UL Q k  M b  = f rake efficiency   UL Q tch  txi M b ic  k txi  ActiveSet Else if (Mb is in soft. G UL Q tch  txi M b ic  k = --------------------------------------------------------------------------------------------------------------------------------------------p  Service  M b    G div UL Tx UL N tot  txi ic  –  1 – F MUD    term  P b – R99  txi M b ic  k – 1 If user selects "Total noise". UL  term  P b – DPDCH  txi M b ic  k UL UL .Atoll 3. 235 . P req P term – R99  M b ic  k = ------------------------------------------------------------------------------------term – R99  M b ic  k – 1 UL Qk  Mb  If compressed mode is operated. or softer/soft without MRC) UL Qk  Mb  = UL UL Max  Q tch  txi M b ic  k    G macro – diversity  2 links txi  ActiveSet Else if (Mb is in soft/soft) UL Qk  Mb  = UL UL Max  Q tch  txi M b ic  k    G macro – diversity  3 links txi  ActiveSet Else if (Mb is in softer/soft with MRC) UL Qk  Mb     UL  UL UL UL = Max  f rake efficiency  Q tch  ic  Q tch  ic    G macro – diversity  2 links other site   txi  ActiveSet    samesite   End If UL Q req  Service  M b  Mobility  M b   req . if the RSCP Active option is selected. ic) and Mb: DL Q req  Service  M b  Mobility  M b   req .ic) traffic channel at Mb with the minimum power allowed on traffic channel for the Mb service min P tch  Service  M b   DL P b  txi M b ic  = ----------------------------------------------L T  txi M b  DL  BTS  P b  txi M b ic  k DL DL . if the RSCP Active option is selected. DL DL Q req  Service  M b  Mobility  M b    Q req   Service  Mb  Mobility  M b    req . G DL Q tch  txi M b ic  k = -------------------------------------------------------p  Service  M b    G div DL N tot  ic  End For DL DL Q k  M b  = f rake efficiency   DL Q tch  txi M b ic  k txi  ActiveSet Do For each cell (txi. max If P tch  txi M b ic  k  P tch  Service  M b   then  txi ic  is set to P tch DL max Recalculation of a decreased Q req (a part of the required quality is managed by the cells set to P tch ) req P tch  Service  M b   DL P b  txi M b ic  = ---------------------------------------------L T  txi M b  236 . G DL Q tch  txi M b ic  k = -------------------------------------------------------------------------------------------------------------------------p  Service  M b    G div DL DL N tot  ic  –  1 – F ortho    BTS  P b  txi M b ic  k – 1 If the user selects the option "Total noise" DL  BTS  P b  txi M b ic  k DL DL .ic) in Mb active set Calculation of the required power for DL traffic channel between (txi.3.Atoll 3. P req P term – R99  M b ic  k = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------term – R99  M b ic  k – 1 UL Qk  Mb  req min req max req min If P term – R99  M b ic  k  P term  M b  then P term – R99  M b ic  k = P term  txi M b  If P term – R99  M b ic  k  P term  M b  then Mb cannot select any cell and its active set is cleared R99 If TP P – UL  M b   TP Max – UL  txi ic  then Mb cannot be connected Endif Downlink Power Control If (mobile does not use a packet switched service that is inactive on the downlink) For each cell (txi. and req Resulting CM – activation  txi M b ic   Q pilot • Either Q pilot • Or P c  txi M b ic   RSCP pilot k CM – activation max if the Ec/I0 Active option is selected. P min P tch  txi M b ic  k = ------------------------------------------------------------------------------------tch  Service  M b   DL Qk  Mb  If compressed mode is operated. P min P tch  txi M b ic  k = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------tch  Service  M b   DL Qk  Mb  Compressed mode is operated if Mi and Sj support compressed mode.ic) in Mb active set Calculation of quality level on (txi.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 UL UL Q req  Service  M b  Mobility  M b    Q req   Service  M b  Mobility  M b    req . Ni While N CE – DL CE – DL  N i  k  N max  Ni  Rejection of the mobile with the lowest service priority starting from the last admitted While N CE – UL CE – UL  N i  k  N max  Ni  Rejection of the mobile with the lowest service priority starting from the last admitted EndFor For each NodeB. then his contribution to interference in the calculation of N tot  ic  is P b  txi M b ic   r c . Iub Backhaul Throughput) For each cell (txi.Atoll 3. G DL Q tch  txi M b ic  k = ----------------------------------------------------------------------------------------------------------------p  Service  M b    G div DL DL N tot  ic  –  1 – F ortho    BTS  P b  txi M b ic  DL DL DL If the user is inactive.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 DL  BTS  P b  txi M b ic  DL DL .3.ic) P tx  txi ic  DL While ----------------------------k  %Powermax P max Rejection of the mobile with the lowest service priority starting from the last admitted EndFor For each cell (txi. For each cell (txi. EndFor DL  DL Q k  M b  = f rake efficiency  DL Q tch  txi M b ic  k txi  ActiveSet DL DL While Q k  M b   Q req  Service  M b  Mobility  M b   and Mb active set is not empty R99 If TP P – DL  M b   TP Max – DL  txi ic  then Mb cannot be connected Endif Uplink and Downlink Interference Update Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones). Channel Elements.ic) UL Update of N tot  txi ic  EndFor For each mobile Mi DL Update of N tot  ic  EndFor EndFor Control of Radio Resource Limits (OVSF Codes. Ni Max While TP Iub – DL  N I  k  TP Iub – DL  N I  Rejection of the mobile with the lowest service priority starting from the last admitted Max While TP Iub – UL  N I  k  TPIub – UL  N I  Rejection of the mobile with the lowest service priority starting from the last admitted 237 .ic) While N Codes Codes  txi ic  k  N max  txi ic  Rejection of the mobile with the lowest service priority starting from the last admitted EndFor For each NodeB. Cell Power. 2. HSPA BE and HSPA VBR service user consumes one HS-SCCH channel. all HSDPA bearer users. Therefore. Here.1 HSDPA Power Allocation The total transmitted power of the cell ( P tx  ic  ) is the sum of the transmitted R99 power.. 4. HSPA BE and HSPA VBR service users active on DL as well as all HSPA CBR service users (i. CBR service users have the highest priority and are processed first. are then evaluated by the HSDPA part of the algorithm. This parameter is not taken into account for CBR service users as HS-SCCH-less operation (i.3. 18 HSDPA BE and HSPA BE service users active on DL. Atoll checks in the simulation that: DL P tx – R99  ic  + P HSUPA  ic   P max  ic   %Power max And it calculates the available HSDPA power as follows: P HSDPA  ic  = P max  ic  – P Headroom  ic  – P tx – R99  ic  – P HSUPA  ic  4. After processing the CBR service users.3.e. HSDPA VBR. HSDPA VBR. Finally. VBR 238 . are taken into consideration. i. HSPA BE and HSPA VBR and HSPA CBR service users. HS-DSCH transmissions without any accompanying HS-SCCH) is performed.e.. Atoll processes the remaining HSDPA bearer users (i. the number of these users connected to an HSDPA bearer cannot exceed the number of HSSCCH channels per cell.e. Each HSDPA BE. The maximum number of HSDPA bearer users ( n max ) corresponds to the maximum number of HSDPA bearer users that the cell can support. HSDPA BE. if the maximum DL load is set to 100%. we have: P tx  ic   P max  ic  • In case of dynamic HSDPA power allocation strategy.2 Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users The number of HS-SCCH channels ( n HS – SCCH ) is the maximum number of HS-SCCH channels that the cell can manage.Atoll 3. The scheduler manages the maximum number of users within each cell. Therefore. All users are connected to the A-DCH R99 bearer. Atoll checks in the simulation that: DL P tx  ic   P max  ic   %Power max where: DL %Power max is the maximum DL load allowed. HSDPA VBR. HSPA VBR. HSDPA VBR.2. This parameter is used to manage the number of BE and VBR service users simultaneously connected to an HSDPA bearer. 2 HSDPA VBR service users active on DL... the HSUPA power and the transmitted HSDPA power. Let us assume there are 30 users in the cell: • • • 10 HSPA CBR service users with any activity status. HSDPA BE and HSPA BE service users).3. the number of HS-SCCH channels and the maximum number of HSDPA bearer users respectively equal 4 and 25. 4.2.3.3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 EndFor Uplink Load Factor Control UL UL For each cell (txi.ic) with X R99  txi ic   X max Rejection of the mobile with the lowest service priority starting from the last admitted EndFor UL UL While at least one cell with X R99  txi ic   X max exists. P tx  ic  = P tx – R99  ic  + P tx –H SDPA  ic  + P HSUPA  ic  • In case of a static HSDPA power allocation strategy. in the order established during the generation of the user distribution.3. unless they have been rejected during the R99 part of the algorithm. active and inactive).3 HSDPA Part of the Algorithm HSDPA BE. at a time (over a transmission time interval).e. Iub backhaul throughput and OVSF codes available in order for them to obtain an HSDPA bearer. These power values are assigned one by one by the scheduler. PHSDPA(B(MX)) = 0. Their connection status will be "HSDPA Scheduler Saturation". And. where X = 0 to 10. the 10 CBR service users are represented by Mj. Iub backhaul throughput and OVSF codes available in order for them to obtain an HSDPA bearer that provides a peak RLC throughput higher or equal to the minimum throughput demand defined for the service.e. Atoll calculates the HSDPA bearer consumption ( C in %) for each user and takes into account this parameter when it determines the resources consumed by the user (i.3. The two VBR service users may be simultaneously served if there are enough HSDPA power. In this case. Iub backhaul throughput and OVSF codes available in order for them to obtain the lowest HSDPA bearer that provides a peak RLC throughput higher or equal to the minimum throughput demand defined for the service. they will be delayed. the HSDPA power used. they will be rejected.3: Connection status of HSDPA bearer users • • • All CBR service users may be served if there are enough HSDPA power. VBR service users have the highest priority and are managed before BE service users. the last five users will be rejected because the maximum number of HSDPA bearer users has been fixed to 25.Atoll 3. the initial values of their respective HSDPA powers is 0. looped back to the starting point. they will be connected. CBR Service Users Let us focus on the ten CBR service users mentioned in the example of the previous paragraph "Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users" on page 238.3. Several CBR service users can share the same HSDPA bearer. with j = 1 to 10. the number of OVSF codes and the Iub backhaul throughput). they will be connected. Fast link adaptation is carried out on these users in order to determine if they can obtain an HSDPA bearer that provides a peak RLC throughput higher or equal to the service minimum throughput demand. Their connection status will be "HS-SCCH Channels Saturation".0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 service users have the highest priority and are managed before BE service users. The next eleven ones will be delayed since there are no longer HS-SCCH channels available. the scheduler ranks the remaining users (i. After processing the CBR service users.3. Then. In the bearer allocation process shown below. Finally... only HSDPA bearers using the QPSK modulation and two HSPDSCH channels at the maximum can be selected and allocated to the users. Users are treated as described in the figure below. As HS-SCCH less operation is performed. Else. Figure 4. As explained before. i. Else. so that with their allocated values. Else. in the order established during the generation of the user distribution. they will be connected. VBR and BE service users) and shares the cell radio resources between them. HSDPA Bearer Allocation Process The HSDPA bearer allocation process depends on the type of service requested by the user. Then.2. they will be rejected. among the BE service users: • • • 4. 239 . CBR service users have the highest priority and are processed first. are used in successive steps. The users are processed in the order established during the generation of the user distribution and the cell’s available HSDPA power is shared between them as explained below. For each type of service. In this case.e. the scheduler ranks the users according to the selected scheduling technique.3 The first two users may be simultaneously served if there are enough HSDPA power. In this case.e. the initial values of their respective HSDPA powers is 0. PHSDPA(B(MX)) = 0. 240 . are used in successive steps. In the bearer allocation process shown below. the 2 VBR service users are represented by Mj. Let us focus on the two HSDPA . where X = 0 to 2. They are processed in the order defined by the scheduler and the cell’s HSDPA power available after all CBR service users have been served is shared between them as explained below. so that with their allocated values.3.Atoll 3. A new fast link adaptation is carried out on these users in order to determine if they can obtain an HSDPA bearer that provides a peak RLC throughput higher or equal to the service minimum throughput demand. with j = 1 to 2. "Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users" on page 238. i. Mj. with j varying from 1 to 10: PHSDPA  j 1  (PHSDPA X 0 (M X )) served Sufficient HS-SCCH power to reach the minimum quality threshold? No Mj is rejected Yes Enough 16-bit OVSF codes available to support the lowest HSDPA bearer allocated? No Mj is rejected Yes Sufficient Iub backhaul throughput to support the lowest HSDPA bearer allocated? No Mj is rejected Yes Sufficient HSDPA power to obtain the lowest HSDPA bearer allocated? No Mj is rejected Yes Determination of the Best HSDPA Bearer BB(Mj) Cell and UE both capable of supporting BB(Mj)? Yes BB(Mj) selected B(Mj) = BB(Mj) No Bearer Downgrading B(Mj) RLC Peak Rate of B(Mj) > Mj Min Throughput Demand? No Mj is rejected Yes Allocation of Min Throughput Demand to Mj Mj connected with B(Mj) (PHSDPA(Mj))served=PHS-PDSCH(B(Mj)) x C(B(Mj)) Update of Available Radio Resources No Mj = M10? Yes Resource allocation for Variable Bit Rate and Best Effort service users Figure 4. These power values are assigned one by one by the scheduler. looped back to the starting point. the scheduler shares the cell’s remaining resources between HSDPA and HSPA VBR service users.4: HSDPA Bearer Allocation Process for CBR Service Users VBR Service Users After processing the CBR service users.VBR service users mentioned in the example of the previous paragraph.e. And.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 For the user. the 4 BE service users are represented by Mj. PHSDPA(B(MX)) = 0. They are processed in the order defined by the scheduler and the cell’s HSDPA power available after all CBR and VBR service users have been served is shared between them as explained below.Atoll 3. the initial values of their respective HSDPA powers is 0.3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 For the user. Mj. especially on the first four users mentioned in the example of the previous paragraph. with j varying from 1 to 2: PHSDPA  j 1  (PHSDPA X 0 (M X )) served Sufficient HS-SCCH power to reach the minimum quality threshold? No Mj is rejected Yes Enough 16-bit OVSF codes available to support the lowest HSDPA bearer allocated? No Mj is rejected Yes Sufficient Iub backhaul throughput to support the lowest HSDPA bearer allocated? No Mj is rejected Yes Sufficient HSDPA power to obtain the lowest HSDPA bearer allocated? No Mj is rejected Yes Determination of the Best HSDPA Bearer BB(Mj) Bearer Downgrading B(Mj) until: 1. These power values are assigned one by one by the scheduler. 241 . In the bearer allocation process shown below. A new fast link adaptation is carried out on these users in order to determine if they can obtain an HSDPA bearer. looped back to the starting point. Let us focus on the HSDPA and HSPA BE service users. the scheduler shares the cell’s remaining resources between BE service users. i.e.5: HSDPA Bearer Allocation Process for VBR Service Users BE Service Users After processing the VBR service users. are used in successive steps. so that with their allocated values. And. with j = 1 to 4. RLC Peak Rate of B(Mj) > Mj Min Throughput Demand RLC Peak Rate of B(Mj) > Mj Min Throughput Demand? No Mj is rejected Yes Mj connected with B(Mj) (PHSDPA(Mj))served=PHS-PDSCH(B(Mj)) +nHS-SCCHxPHS-SCCH(Mj) No Mj = M2? Yes Resource allocation for Best Effort service users Figure 4. Cell and UE both capable of supporting B(Mj) And 2. where X = 0 to 4. "Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users" on page 238. 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 For the user. it determines the HS-PDSCH CQI.6: HSDPA Bearer Allocation Process for BE Service Users 4. The power on the HS-DSCH channel is transmitted at a constant power while the modulation. Atoll finds the highest downlink throughput that can be provided to the user and may calculate the application throughput. Once the bearer selected.3. with j varying from 1 to 4: Figure 4.2. the coding and the number of codes during a communication. Atoll calculates for each user either the best pilot quality (CPICH Ec/Nt) or the best HS-PDSCH quality (HS-PDSCH Ec/Nt). the coding and the number of codes are changed to adapt to the radio conditions variations. 64QAM). 16QAM.Atoll 3. this depends on the option selected in Global parameters (HSDPA part): CQI based on CPICH quality or CQI based on HS-PDSCH quality (CQI means channel quality indicator).3. Mj. the node-B may change every 2ms the modulation (QPSK. Then.4 Fast Link Adaptation Modelling Fast link adaptation (or Adaptive Modulation and Coding) is used in HSDPA. 242 . calculates the best bearer that can be used and selects the suitable bearer so as to comply with cell and terminal user equipment HSDPA capabilities. Based on the reported channel quality indicator (CQI). are defined in "Inputs" on page 215. defined between ic and icadj and set to a value different from 0. ic  corresponds to the CPICH quality. In the HSDPA coverage prediction. CPICH Quality Calculation Ec Let us assume the following notation:  -----. Nt pilot Two options.  and N 0 3.3.Atoll 3.( ) G Tx  G term term  BTS .for the total noise option. we have:  BTS    P c  ic  Eci  ---- ic  = ----------------------------------------. 1. Atoll proceeds as follows. RF  ic ic adj  is the interference reduction factor. L T is calculated as follows: L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io -) L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx  G term 243 .  Nt  pilot DL N tot  ic  And  BTS    P c  ic  Eci  ----for the without useful signal option. DL I inter – techno log y  ic  =  ni ic i is the i th Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic i interfering carrier of an external transmitter Tx m ICPic  ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the i frequency gap between ic i (external network) and ic . may be used to calculate Nt: option Without useful signal or option Total noise. Therefore.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 CQI Based on CPICH Quality When the option “CQI based on CPICH quality” is selected. available in Global parameters. P pilot  ic  P c  ic  = --------------------i LT i L path  L Tx  L term  L body  L indoor  E Shadowing 3 L T = ------------------------------------------------------------------------------------------------------------------.  ic  = -------------------------------------------------------------------------------- Nt  pilot DL N tot  ic  –  1 –     BTS  P c  ic  i With DL DL DL DL DL term N tot  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0  DL  DL DL P SCH  ic  P SCH  ic  DL term I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 –     P tot  ic  – ------------------- –  BTS   P tot  ic  – ------------------LT LT  txi txi txi     DL I extra  ic  =  DL P tot  ic  txj j  i  Ptot  icadj  DL DL  j I inter – carrier  ic  = txj -----------------------------------RF  ic ic adj  icadj is a carrier adjacent to ic. DL I inter – techno log y  ic  is the inter-technology interference at the receiver on ic. P HS – SCCH  ic  is the HS-SCCH power on carrier ic. ic  ). It is Nt HS – SCCH specified in mobility properties. req Ec In this case. HS-PDSCH Quality Calculation Atoll proceeds as follows: 1st step: Atoll calculates the HS-SCCH power ( P HS – SCCH ). defined between ic and icadj and set to a value different from 0.  Nt  HS – SCCH DL N tot  ic  And  BTS  P c  ic  Eci  ----  ic  = ---------------------------------------------------------------------------------------------------------------------------. Atoll considers the following formula: P max  ic  – P SCH  ic  P max  ic  DL term - I intra  ic  = -------------------+  BTS  1 – F MUD  1 –    --------------------------------------------- LT LT P max  ic  – P SCH  ic  - – BTS   ---------------------------------------------  LT 2. ic   Nt pilot 3.3.  CQI  pilot is read in the table Ec  . CPICH CQI Determination Let us assume the following notation:  CQI  pilot corresponds to the CPICH CQI.  CQI  pilot = f   -----. We have:  BTS  P c  ic  Eci  ---- ic  = ------------------------------. It is either fixed by the user (when the option “HS-SCCH Power Dynamic Allocation”in the cell property dialogue is unchecked) or dynamically calculated (when the option “HS-SCCH Power Dynamic Allocation” is selected).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Atoll performs intra-cell interference computations based on the total power. This table is defined for the terminal reception equipment and the selected mobility. 244 . RF  ic ic adj  is the interference reduction factor.  Nt  HS – SCCH DL term N tot  ic  –  1 – F ortho    1 – F MUD    BTS  P c  ic  i With DL DL DL DL DL term N tot  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0  DL   DL  DL P SCH  ic  P SCH  ic  DL term I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – ------------------- –  BTS   P tot  ic  – ------------------- LT LT txi txi txi     DL I extra  ic  =  DL P tot  ic  txj j  i  Ptot  icadj  DL DL  j -----------------------------------I inter – carrier  ic  = txj RF  ic ic adj  icadj is a carrier adjacent to ic. DL I inter – techno log y  ic  is the inter-technology interference at the receiver on ic. You can instruct Atoll to use maximum power by adding the following lines in the Atoll. the HS-SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (noted  -----.Atoll 3.ini file: [CDMA] PmaxInIntraItf = 1 In this case.for the without useful signal option.for the total noise option. req EcDL   ---- ic   N tot  ic    Nt  HS – SCCH P HS – SCCH  ic  =  ------------------------------------------------------------------  L T for the total noise option. req i Ecterm   1 + 1 – F  ----     1 – F    ic  BTS ortho MUD    Nt  HS – SCCH  2nd step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ). Atoll evaluates the HS-PDSCH quality Ec Let us assume the following notation:  -----. i  BTS     And req EcDL  ----    ic   Nt  HS – SCCH  N tot  ic    ------------------------------------------------------------------------------------------------------------------------------------------P HS – SCCH  ic  =    L T for the without useful signal option.for the without useful signal option.Atoll 3. P HSDPA  ic  = P HS – PDSCH  ic  + n HS – SCCH  P HS – SCCH  ic  Therefore. P HSDPA  ic  is the power available for HSDPA on the carrier ic. P HS – SCCH  ic  P c  ic  = ------------------------------i LT i and L path  L Tx  L term  L body  L indoor  E Shadowing 4 L T = ------------------------------------------------------------------------------------------------------------------.  Nt  ic  HS – PDSCH = ---------------------------------------------------------------------------------------------------------------------------P c  ic  DL term i N tot  ic  –  1 – F ortho    1 – F MUD    BTS  --------------n 4.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 DL I inter – techno log y  ic  =  n ic i is the i th i Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic i interfering carrier of an external transmitter Tx m ICPic  ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the i frequency gap between ic i (external network) and ic . Therefore.3.( ) G Tx  G term term term  BTS . ic  corresponds to the HS-PDSCH quality. F ortho .  Nt  HS – PDSCH DL N tot  ic  And  BTS  P c  ic  Eci  ---- . 3rd step: Then. F MUD and N 0 are defined in "Inputs" on page 215. In the HSDPA coverage prediction.for the total noise option.  Nt  HS – PDSCH We have:  BTS  P c  ic  Eci  ---- ic  = ------------------------------. L T is calculated as follows: L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io -) L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx  G term 245 . or a user-defined cell input. we have: P HS – PDSCH  ic  = P HSDPA  ic  – n HS – SCCH  P HS – SCCH  ic  n HS – SCCH is the number of HS-SCCH channels. This parameter is either a simulation output. –  BTS   P tot  ic  – ------------------  LT  LT  txi txi txi     DL I extra  ic   =  DL P tot  ic  txj j  i  Ptot  icadj  DL DL  j -----------------------------------I inter – carrier  ic  = txj RF  ic ic adj  icadj is a carrier adjacent to ic. P HS – PDSCH  ic  P c  ic  = ---------------------------------i LT i And L path  L Tx  L term  L body  L indoor  E Shadowing 5 . In the HSDPA coverage prediction. Atoll works on the assumption that five HS-PDSCH channels are used (n=5). Atoll calculates  CQI  HS – PDSCH as follows:  CQI  HS – PDSCH =  CQI  pilot – P pilot + P HS – PDSCH 5. With DL DL DL DL DL term N tot  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0  DL   DL  DL P SCH  ic  P SCH  ic  DL term I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – ------------------. RF  ic ic adj  is the interference reduction factor. L T is calculated as follows: L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io -) L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx  G term 246 .( ) L T = ------------------------------------------------------------------------------------------------------------------G Tx  G term term term  BTS . Atoll performs intra-cell interference computations based on the total power. HS-PDSCH CQI Determination The best bearer that can be used depends on the HS-PDSCH CQI. DL I inter – techno log y  ic  is the inter-technology interference at the receiver on ic. defined between ic and icadj and set to a value different from 0. Atoll considers the following formula: P max  ic  – P SCH  ic  P max  ic  – P SCH  ic  P max  ic  DL term I intra  ic  = -------------------+  BTS  1 – F MUD  1 –    ----------------------------------------------- –  BTS   ----------------------------------------------- LT LT LT 4. DL I inter – techno log y  ic  =  ni ic i is the i th Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic i interfering carrier of an external transmitter Tx m ICP ic  ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the i frequency gap between ic i (external network) and ic . You can instruct Atoll to use maximum power by adding the following lines in the Atoll. Let us assume the following notation:  CQI  HS – PDSCH corresponds to the HS-PDSCH CQI. F ortho .Atoll 3. F MUD and N 0 are defined in "Inputs" on page 215.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Here.3.ini file: [CDMA] PmaxInIntraItf = 1 In this case. the UE is not informed about the transmission format and has to revert to blind decoding of the transport format used on the HS-DSCH. When several HSDPA bearers can supply the same RLC peak throughput.. Complexity of blind detections in the UE is decreased by limiting the transmission formats that can be used (i. Additionally.. The number of HS-PDSCH channels required by the bearer does not exceed the maximum number of HS-PDSCH channels that the terminal can use. • HSDPA bearers using 64QAM modulation (improvement introduced by the release 7 of the 3GPP UTRA specifications.e. referred to as HSPA+): These HSDPA bearers can be allocated to VBR and BE service users connected to cells with HSPA+ capabilities only. When there are several HSDPA bearers compatible. Atoll considers an HSDPA bearer as compatible with the user equipment if: • • • The transport block size does not exceed the maximum transport block size supported by the user equipment. For VBR service users. the HSDPA bearers available). For CBR service users. 64QAM modulation in the DL and MIMO systems) and the maximum number of HS-PDSCH channels is 15.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 5. HS-DSCH transmissions without any accompanying HS-SCCH) is performed. For VBR service users. Atoll chooses the HSDPA bearer with the highest modulation scheme. In this case. the selected HSDPA must provide a peak RLC throughput between the minimum and the maximum throughput demands defined for the service.e. Finally. The modulation is supported by the user equipment. only HSDPA bearers using the QPSK modulation and two HS-PDSCH channels at the maximum can be selected and allocated to these users. HSDPA Bearer Selection Atoll selects the HSDPA bearer associated to this CQI (in the table Best Bearer=f(HS-PDSCH CQI) defined for the terminal reception equipment and the user mobility) and compatible with the user equipment and cell capabilities. These HSDPA bearers cannot be allocated to CBR service users. Atoll selects the HSDPA bearer that provides the highest RLC peak throughput.e. Example1: One HSDPA BE service user with category 13 user equipment and a 50km/h mobility.7: HSDPA UE Categories Table The cell to which the user is connected supports HSPA+ functionalities (i. HS-SCCH-less operation (i. the selected HSDPA bearer must provide a peak RLC throughput higher or equal to the minimum throughput demand defined for the service. Therefore. if no HSDPA bearer is compatible.3. The number of HS-PDSCH channels required by the bearer must not exceed the maximum number of HS-PDSCH codes available for the cell.Atoll 3. 247 . The user equipment capabilities are: • • • • Maximum transport block size: 35280 bits Maximum number of HS-PDSCH channels: 15 Highest modulation supported: 64QAM MIMO Support: No Figure 4. The number of HS-PDSCH channels required by the bearer must not exceed the maximum number of HS-PDSCH codes available for the cell. HSDPA bearers can be classified into two categories: • HSDPA bearers using QPSK and 16QAM modulations: They can be selected for all users connected to HSPA and HSPA+ capable cells. Let’s consider the following examples. the selected HSDPA bearer must provide a peak RLC throughput between the minimum and the maximum throughput demands defined for the service. Atoll allocates a lower HSDPA bearer compatible with the user equipment and cell capabilities which needs fewer resources. Therefore. Therefore. Atoll can choose between two HSDPA bearers. Atoll can choose between two HSDPA bearers. the bearer index 26.32 Mb/s .0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 1st case: The CQI experienced by the user equals 26. the bearer indexes 26 and 31. i. Characteristics of the bearer index 26 are: • • • • Transport block size: 17237 bits Number of HS-PDSCH channels used: 12 16QAM modulation is used Peak RLC Throughput: 8. the bearer index 32. Atoll can choose between two HSDPA bearers. Atoll selects the HSDPA bearer using the highest modulation scheme. i.e.32 Mb/s Characteristics of the bearer index 31 are: • • • • Transport block size: 15776 bits Number of HS-PDSCH channels used: 10 64QAM modulation is used Peak RLC Throughput: 7.Atoll 3. the bearer indexes 26 and 31.24 Mb/s Both HSDPA bearers are compatible with the user equipment and cell capabilities and the peak RLC throughput they provide is the same. Example 2: One HSDPA BE user experiencing a CQI of 26. Characteristics of the bearer index 26 are: • • • • 248 Transport block size: 17237 bits Number of HS-PDSCH channels used: 12 16QAM modulation is used Peak RLC Throughput: 8.8: HSDPA Radio Bearers Table 2nd case: The CQI experienced by the user equals 27.24 Mb/s Characteristics of the bearer index 32 are: • • • • Transport block size: 21768 bits Number of HS-PDSCH channels used: 12 64QAM modulation is used Peak RLC Throughput: 10.e. Therefore. the bearer indexes 27 and 32. Characteristics of the bearer index 27 are: • • • • Transport block size: 21754 bits Number of HS-PDSCH channels used: 15 16QAM modulation is used Peak RLC Throughput: 10.36 Mb/s Both HSDPA bearers are compatible with the user equipment and cell capabilities. Atoll selects the HSDPA bearer that provides the highest RLC peak throughput.3. Figure 4. The user equipment characteristics are the following: • • • • Maximum transport block size: 20251 bits Maximum number of HS-PDSCH channels: 15 Highest modulation supported: 16QAM MIMO Support: No The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the terminal.36 Mb/s 1st case: The user equipment category is 9. it is allocated. On the other hand. 1.e. Atoll selects a lower HSDPA bearer compatible with cell and UE category capabilities. HS-PDSCH Quality Calculation Atoll proceeds as follows: 1st step: Atoll calculates the HS-SCCH power ( P HS – SCCH ).3. 6. The user equipment capabilities are: • • • • Maximum transport block size: 35280 bits Maximum number of HS-PDSCH channels:15 Highest modulation supported: 64QAM MIMO Support: No The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the cell. The cell to which the user is connected supports HSPA+ functionalities (i. it may recalculate the HS-PDSCH quality with the real number of HS-PDSCH channels (A default value (5) was taken into account in the first HS-PDSCH quality calculation). the number of HS-PDSCH channels (12) exceeds the maximum number of HS-PDSCH channels the terminal can use (10). In the HSDPA Radio Bearer table. Atoll proceeds as follows. CQI Based on HS-PDSCH Quality When the option “CQI based on HS-PDSCH quality” is selected. 64QAM modulation in the DL and MIMO systems) and the maximum number of HS-PDSCH channels is 15. The transport block size (14411 bits) does not exceed the maximum transport block size (14411 bits) the terminal can carried. 249 .e. The cell to which the user is connected supports HSPA+ functionalities (i. • • • The number of HS-PDSCH channels (10) does not exceed the maximum number of HS-PDSCH channels the terminal can use (10) and the maximum number of HS-PDSCH channels available at the cell level (15). Only the bearer index 26 is compatible with the user equipment capabilities. Therefore. Therefore. 2nd case: The user equipment category is 8. With the bearer index 26. 16QAM modulation is supported by the terminal and the cell.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Characteristics of the bearer index 31 are: • • • • Transport block size: 15776 bits Number of HS-PDSCH channels used: 10 64QAM modulation is used Peak RLC Throughput: 7. when the method “Without useful signal” is used. and the transport block size (17237 bits) exceeds the maximum transport block size (14411 bits) the terminal can carried. 3rd case: The user equipment category is 13. The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the terminal. none of HSDPA bearers are compatible with the user equipment capabilities. 64QAM modulation in the DL and MIMO systems) and the maximum number of HS-PDSCH channels is 15. It selects the bearer index 25. The user equipment characteristics are the following: • • • • Maximum transport block size: 14411 bits Maximum number of HS-PDSCH channels: 10 Highest modulation supported: 16QAM MIMO Support: No Here. The cell to which the user is connected supports HSPA functionalities and the maximum number of HS-PDSCH channels is 15. Atoll exactly knows the number of HS-PDSCH channels. HS-PDSCH Quality Update Once the bearer selected. Atoll selects it. the bearer index 26 is compatible with cell and UE category capabilities. for the without useful signal option. ic  ). Therefore.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 P HS – SCCH  ic  is the HS-SCCH power on carrier ic. defined between ic and icadj and set to a value different from 0. F ortho . It is either fixed by the user (when the option “HS-SCCH Power Dynamic Allocation”in the cell property dialogue is unchecked) or dynamically calculated (when the option “HS-SCCH Power Dynamic Allocation” is selected). F MUD and N 0 are defined in "Inputs" on page 215. the HS-SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (noted  -----. P HS – SCCH  ic  P c  ic  = ------------------------------i LT i And L path  L Tx  L term  L body  L indoor  E Shadowing 6 .–  BTS   P tot  ic  – ------------------  LT  LT  txi txi txi     DL I extra  ic  =  DL P tot  ic  txj j  i  Ptot  icadj  DL DL  j -----------------------------------I inter – carrier  ic  = txj RF  ic ic adj  icadj is a carrier adjacent to ic. We have:  BTS  P c  ic  Eci  ---- ic  = ------------------------------. L T is calculated as follows: L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io -) L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx  G term 250 .Atoll 3. In the HSDPA coverage prediction.for the total noise option. RF  ic ic adj  is the interference reduction factor. DL I inter – techno log y  ic  =  ni ic i is the i th Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic i interfering carrier of an external transmitter Tx m ICP ic  ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the i frequency gap between ic i (external network) and ic .  Nt  HS – SCCH DL term N tot  ic  –  1 – F ortho    1 – F MUD    BTS  P c  ic  i With DL DL DL DL DL term N tot  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0  DL   DL  DL P SCH  ic  P SCH  ic  DL term I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – ------------------. 6. It is Nt HS – SCCH specified in mobility properties. DL I inter – techno log y  ic  is the inter-technology interference at the receiver on ic.3.  Nt  HS – SCCH DL N tot  ic  And  BTS  P c  ic  Eci  ----  ic  = ---------------------------------------------------------------------------------------------------------------------------.( ) L T = ------------------------------------------------------------------------------------------------------------------G Tx  G term term term  BTS . req Ec In this case. We have:  BTS  P c  ic  Eci  ---- . available in Global parameters. DL I inter – techno log y  ic  is the inter-technology interference at the receiver on ic. i  BTS     And req EcDL  ----    Nt  ic  HS – SCCH  N tot  ic    -  L T for the without useful signal option. Atoll exactly knows the number of HS-PDSCH channels and recalculates the HS-PDSCH quality with the real number of HS-PDSCH channels.for the without useful signal option. ic  corresponds to the HS-PDSCH quality. 3rd step: Then. Then.for the total noise option. it calculates the HS-PDSCH CQI and the bearer to be used. defined between ic and icadj and set to a value different from 0.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 req EcDL    ----  ic   HS – SCCH  N tot  ic    Nt P HS – SCCH  ic  =  ------------------------------------------------------------------  L T for the total noise option.  Nt  HS – PDSCH Two options. may be used to calculate Nt: option Without useful signal or option Total noise. With DL DL DL DL DL term N tot  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0  DL   DL  DL P SCH  ic  P SCH  ic  DL term I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – ------------------- –  BTS   P tot  ic  – ------------------- LT LT txi txi txi     DL I extra  ic   =  DL P tot  ic  txj j  i  Ptot  icadj  DL DL  j -----------------------------------I inter – carrier  ic  = txj RF  ic ic adj  icadj is a carrier adjacent to ic. P HSDPA  ic  = P HS – PDSCH  ic  + n HS – SCCH  P HS – SCCH  ic  Therefore. Atoll works on the assumption that five HS-PDSCH channels are used (n=5). Atoll evaluates the HS-PDSCH quality Ec Let us assume the following notation:  -----. ic   BTS    Nt HS – SCCH 2nd step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ) P HSDPA  ic  is the power available for HSDPA on the carrier ic. or a user-defined cell input.  Nt  ic  HS – PDSCH = ------------------------------DL N tot  ic  And  BTS  P c  ic  Eci  ----= ---------------------------------------------------------------------------------------------------------------------------. P HS – SCCH  ic  =  ------------------------------------------------------------------------------------------------------------------------------------------req i Ec term   1 + 1 – F    ortho    1 – F MUD    -----. we have: P HS – PDSCH  ic  = P HSDPA  ic  – n HS – SCCH  P HS – SCCH  ic  n HS – SCCH is the number of HS-SCCH channels. 251 .Atoll 3.3. This parameter is either a simulation output. Once the bearer selected. RF  ic ic adj  is the interference reduction factor.  ic   Nt  HS – PDSCH P c  ic  DL term i N tot  ic  –  1 – F ortho    1 – F MUD    BTS  --------------n Here. 3.3. You can instruct Atoll to use maximum power by adding the following lines in the Atoll. Atoll performs intra-cell interference computations based on the total power. HSDPA Bearer Selection The bearer is selected as described in "HSDPA Bearer Selection" on page 246. a terminal with an HSDPA UE category supporting MIMO).2.ini file: [CDMA] PmaxInIntraItf = 1 In this case. DL G TD is the additional diversity gain in downlink (in dB). he will benefit from downlink diversity gain on the HS-PDSCH Ec/Nt. EcEc DL DL  ----=  -----.3. ic    Nt  HS – PDSCH mobility.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks DL I inter – techno log y  ic  =  n ic i is the i th i ©Forsk 2015 Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic i interfering carrier of an external transmitter Tx m ICP ic  ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the i frequency gap between ic i (external network) and ic . HS-PDSCH CQI Determination Let us assume the following notation:  CQI  HS – PDSCH corresponds to the HS-PDSCH CQI. 7.  CQI  HS – PDSCH is read in the table Ec  . Atoll considers the following formula: P max  ic  – P SCH  ic  P max  ic  – P SCH  ic  P max  ic  DL term - –  BTS   ---------------------------------------------- I intra  ic  = -------------------+  BTS  1 – F MUD  1 –    ---------------------------------------------    LT LT LT 2. ic  + G TD + G TD in dB  ic   Nt  HS – PDSCH  Nt  HS – PDSCH Where DL G TD is the downlink transmit diversity gain (in dB) corresponding to the numbers of transmission and reception antenna ports (respectively defined in the transmitter and terminal properties).3.Atoll 3.. It is defined for the clutter class of the user.5 MIMO Modelling MIMO .Transmit Diversity If the user is connected to a cell that supports HSPA+ with transmit diversity and if he has a MIMO-capable terminal (i.e. F MUD and N 0 are defined in "Inputs" on page 215. P HS – PDSCH  ic  P c  ic  = ---------------------------------i LT i And L path  L Tx  L term  L body  L indoor  E Shadowing 7 L T = ------------------------------------------------------------------------------------------------------------------. 4. L T is calculated as follows: L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io -) L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx  G term 252 .( ) G Tx  G term term term  BTS . In the HSDPA coverage prediction. F ortho . This table is defined for the terminal reception equipment and the specified  CQI  HS – PDSCH = f   -----. Spatial Multiplexing If the user is connected to a cell that supports HSPA+ with spatial multiplexing and if he has a MIMO-capable terminal (i.4 Delayed M15 11 1440 3.e.4 Delayed M8 14 2080 3. nor VBR service user in the cell and the number of HS-SCCH channels and the maximum number of HSDPA bearer users have been respectively set to 4 and 15.4 Connected M4 9 2080 3. they are sorted in descending order by the channel quality indicator (CQI). All of them are active on DL and connected to the A-DCH R99 bearer. Round Robin and Proportional Fair. Three scheduling algorithms are available. f SM – Gain is the spatial multiplexing gain factor defined for the clutter 4.4 Delayed M6 12 2080 3. VBR service users have the highest priority and are managed before BE service users.4 Delayed M14 6 1600 3. Let us consider a cell with 16 HSDPA and HSPA BE service users. Max C/I.3.4 Connected M3 8 2080 160+3. CBR service users have the highest priority and are processed first.3..4 Delayed M12 4 1600 3.4 Delayed M13 5 1600 3. 253 .e. in the order established during the generation of the user distribution. Mobiles Simulation Rank Best Bearer (kbps) DL Obtained Throughput (kbps) Connection Status M1 2 2400 2400+3.3. ic). the peak RLC throughput obtained by the user is the following: DL DL Max TP P – R LC = TP P –R LC  Index HSDPABearer    1 + f SM – Gain   G SM – 1   Where DL TP P – R LC  Index HSDPABearer  is the peak RLC throughput that the selected HSDPA bearer ( Index HSDPABearer ) can provide in the cell (Txi. the scheduler ranks the users according the scheduling technique. Max G SM is the maximum spatial multiplexing gain (in dB) for a given number of transmission and reception antennas (respectively defined in the transmitter and terminal properties). he will benefit from the spatial multiplexing gain in its peak RLC throughput. In this case.4 Delayed M10 1 1600 3. a terminal with an HSDPA UE category supporting MIMO). in a best bearer descending order. For each type of service. It is read in the HSDPA Radio Bearer table. i.4 Delayed M11 3 1600 3.4 Delayed M16 16 2080 0 Scheduler Saturation Round Robin Users are taken into account in the same order than the one in the simulation (random order).4 Connected M2 15 2400 1440+3. the scheduler processes the remaining users (i.Atoll 3. After processing the CBR service users.4 Delayed M9 7 1920 3..e. Then. Max C/I 15 users (where 15 corresponds to the maximum number of HSDPA bearer users defined) enters the scheduler in the same order as in the simulation.6 Scheduling Algorithms The scheduler manages the maximum number of users within each cell. VBR and BE service users). Impact they have on the simulation result is described in the tables below. There is neither CBR service user.4 Delayed M5 10 2080 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 MIMO .4 Delayed M7 13 2080 3.2. 4 Connected M2 2 2400 960+3.4 Connected M3 3 1600 3. Then.4 Delayed M4 4 1600 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Mobiles Simulation Rank Best Bearer (kbps) DL Obtained Throughput (kbps) Connection Status M1 1 1600 1600+3. Simu Ri CQI Ri is the user rank in the simulation.ini file.4 Delayed M10 10 2080 3.4 Delayed M16 16 2080 0 Scheduler Saturation Proportional Fair 15 users (where 15 corresponds to the maximum number of HSDPA bearer users defined) enters the scheduler in the same order as in the simulation.4 Delayed M6 10 5 750 2080 3. You can change the default weights by editing the atoll. For a user i. is the user rank according to the CQI. see the Administrator Manual.4 Delayed M10 5 13 900 1600 3.4 Delayed M7 4 12 800 1600 3.4 Delayed Mobiles Simulation Rank M1 M2 Connection Status M9 15 2 850 2400 3.4 Delayed M5 3 11 700 1600 3.4 Delayed M13 13 7 1000 2080 3.4 Delayed M7 7 1920 3.4 Delayed M9 9 2080 3.4 Delayed M14 14 2080 3.4 Delayed M11 11 1440 3. the random parameter RP i is calculated as follows: Simu RP i = 50  R i CQI + 50  R i Where.4 Delayed M8 8 2080 3.4 Delayed M13 13 2080 3.4 Delayed M5 5 1600 3. 254 CQI Rank RP Best Bearer (kbps) DL Obtained Throughput (kbps) 2 1 150 2400 2400 Connected 1 10 550 1600 960 Connected M3 8 3 550 2080 160 Connected M4 9 4 650 2080 3.Atoll 3.4 Delayed M12 6 14 1000 1600 3.4 Delayed M8 7 9 800 1920 3.4 Delayed . For more information.3.4 Delayed M11 12 6 900 2080 3. they are sorted in an ascending order according to a new random parameter which corresponds to a combination of the user rank in the simulation and the channel quality indicator (CQI).4 Delayed M12 12 2080 3.4 Delayed M6 6 1600 3.4 Delayed M15 15 2400 3. Only site-level resources (such as the Iub throughput and the channel elements) are shared between the users of the two cells. DC-HSDPA and single-carrier HSDPA users.. Round Robin and Proportional Fair). as he may be assigned two different HSDPA bearers in the two cells. the number of HS-SCCH channels and the maximum number of HSDPA bearer users have been respectively set to 4 and 7. 255 . After the users have been ranked. Therefore. the scheduler allocates HSDPA resources to each user following the calculated order as long as there are resources available.e. maximum number of HSDPA bearer users. VBR service users have the highest priority and are managed before BE service users. For each type of service. Impact the scheduling algorithms have on the simulation results is described in the tables below. the scheduler manages a single queue of users at the Node B.Atoll 3. i. the scheduler ranks the users according the scheduling technique (Max C/I.2. Even if there is a unique list of users at the transmitter level. are ranked together in a unique list.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4. number of OVSF codes). The transmitter supports the multi-cell HSDPA mode.e. the scheduler manages the users ranked 1st to 11th (i.e. 4 singlecarrier users connected to the second carrier and 3 DC-HSDPA users). terminal with UE categories 21 to 24). DC-HSDPA users are considered twice in the list as they may be assigned two different HSDPA bearers in the two cells.e. The scheduling algorithms defined for the two cells are the same as the one selected for the transmitter.3. Simulation Rank DC-HSDPA Support Carriers Comments 1 Yes 1 and 2 Anchor carrier: 2 2 No 2 3 No 1 4 Yes 1 and 2 5 No 1 6 No 2 7 No 1 8 No 2 9 Yes 1 and 2 10 No 1 11 No 2 12 Yes 1 and 2 13 No 2 14 Yes 1 and 2 15 No 1 16 Yes 1 and 2 Anchor carrier: 2 Anchor carrier: 1 Anchor carrier: 1 Anchor carrier: 1 Anchor carrier: 2 In each cell. Among the users. Each DC-HSDPA user is counted twice. once in each cell. Otherwise. Users ranked 12th to 16th are rejected because the maximum number of HSDPA bearer users that the scheduler can manage in a cell is exceeded. VBR and BE service users). There is neither CBR service user. MC-HSDPA and DB-MC-HSDPA users are processed as DC-HSDPA users if they are connected to two carriers. 4 single-carrier users connected to the first carrier. All users are active in DL and connected to the A-DCH R99 bearer..7 Mobiles Simulation Rank CQI Rank RP Best Bearer (kbps) DL Obtained Throughput (kbps) Connection Status M14 14 8 1100 2080 3. the resources of each cell are not shared and each carrier has its own pool of resources (number of HS-SCCH channels. there are 6 DC-HSDPA users (i. After processing the CBR service users.3. the scheduler processes the remaining users (i. nor VBR service users. Let us consider a transmitter with 16 BE service users. CBR service users have the highest priority and are processed first. All users belonging to the transmitter. in the order established during the generation of the user distribution.. they are considered as single-cell HSDPA users.4 Delayed M16 16 - - 2080 0 Scheduler Saturation Dual-Cell HSDPA For transmitters that support multi-cell HSDPA mode.3.4 Delayed M15 11 15 1300 1440 3. HSDPA power. . i.Atoll 3. a total of 14 users enter the scheduler in the same order as in the simulation.4 Connected M2 (DC-HSDPA) 2 4 19 2400 2400+3.4 kbps (963.4 Delayed M9 (DC-HSDPA) 2 1 13 960 3. Then.4 Connected M3 2 8 18 2080 1440+3.4 Delayed M12 (DC-HSDPA) 1 2 12 14 15 1120 1440 0 Scheduler Saturation M13 2 13 17 1920 0 Scheduler Saturation M14 (DC-HSDPA) 1 2 14 13 15 960 1440 0 Scheduler Saturation M15 1 15 17 1920 0 Scheduler Saturation M16 (DC-HSDPA) 1 2 16 12 14 800 1120 0 Scheduler Saturation The scheduled DC-HSDPA users have the following status: • • • The user ranked 4th (here M2) is connected to an HSDPA bearer in each cell. The first user (here M9) is delayed in the two cells. He obtains a total DL throughput of 4323.4 Connected M5 1 3 16 1600 3. in a best bearer descending order.4 kbps (2403.4 Delayed M8 1 10 14 1120 3.e. Round Robin 7 users from each cell (where 7 corresponds to the maximum number of HSDPA bearer users defined for each cell). they are sorted in the order of decreasing channel quality indicator (CQI). i.4 Connected M2 (DC-HSDPA) 1 4 17 1920 1920 Connected M4 (DC-HSDPA) 1 9 17 1920 960+3. The user ranked 9th (here M4) is connected to an HSDPA bearer in each cell.. He obtains a total DL throughput of 3.4 Delayed M4 (DC-HSDPA) 2 9 16 1600 1120 Connected M6 2 2 15 1440 3. i.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Max C/I 7 users from each cell (where 7 corresponds to the maximum number of HSDPA bearer users defined for each cell).4+1920). He obtains a total DL throughput of 2083.e.e.3.4 Delayed M10 2 6 13 960 3.4 kbps.4 Delayed M7 1 7 14 1120 3.4+1120). Mobiles Carrier Simulation Rank CQI Best Bearer (kbps) DL Obtained Throughput (kbps) Connection Status M1 1 5 21 3040 3040+3. a total of 14 users enter the scheduler in the same order as in the simulation.4 Delayed M9 (DC-HSDPA) 1 1 12 800 0 Delayed M11 2 11 12 800 3. 256 Mobiles Carrier Simulation Rank CQI Best Bearer (kbps) DL Obtained Throughput (kbps) Connection Status M1 (DC-HSDPA) 1 1 12 800 800 Connected . He obtains a total DL throughput of 1763. • • Proportional Fair 7 users from each cell (where 7 corresponds to the maximum number of HSDPA bearer users defined for each cell).4 Delayed M8 2 8 18 2080 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Mobiles Carrier Simulation Rank CQI Best Bearer (kbps) DL Obtained Throughput (kbps) Connection Status M1 (DC-HSDPA) 2 1 13 960 960+3..Atoll 3. Simu Ri CQI Ri is the user rank in the simulation. The user ranked 4th (here M4) is connected to an HSDPA bearer in each cell.4 Connected M6 2 6 13 960 160+3.4 Delayed M12 (DC-HSDPA) 1 2 12 14 15 1120 1440 0 Scheduler Saturation M13 2 13 17 1920 0 Scheduler Saturation M14 (DC-HSDPA) 1 2 14 13 15 960 1440 0 Scheduler Saturation M15 1 15 17 1920 0 Scheduler Saturation M16 (DC-HSDPA) 1 2 16 12 14 800 1120 0 Scheduler Saturation The scheduled DC-HSDPA users have the following status: • The first user (here M1) is connected to an HSDPA bearer in each cell.4). He obtains a total DL throughput of 2563. For a user i.e.4 Connected M7 1 7 14 1120 3. a total of 14 users enter the scheduler in the same order as in the simulation. i.4 kbps.4 Connected M3 1 3 16 1600 1600+3.4+960).4 Connected M4 (DC-HSDPA) 1 4 17 1920 960 Connected M5 1 5 21 3040 480+3. He obtains a total DL throughput of 3.4 Delayed M10 1 10 14 1120 3.4 Connected M4 (DC-HSDPA) 2 4 19 2400 1600+3.4 kbps (1603.3. they are sorted in an ascending order according to a new random parameter which corresponds to a combination of the user rank in the simulation and the channel quality indicator (CQI).4 Delayed M9 (DC-HSDPA) 2 9 16 1600 0 Delayed M9 (DC-HSDPA) 1 9 17 1920 3. the random parameter RPi is calculated as follows: Simu RPi = 50  R i CQI + 50  R i Where. The user ranked 9th (here M9) is delayed in the two cells.4 Connected M2 2 2 15 1440 1440+3.4 Delayed M11 2 11 12 800 3.4 kbps (800+963. Then. 257 . is the user rank according to the CQI. 4 Connected M2 1 5 21 1 300 3040 3040+3.3. The user ranked 9th (here M7) is delayed in the two cells. see the Administrator Manual.4 Connected M6 DC-HSDPA 1 1 12 13 700 800 0 Delayed M7 DC-HSDPA 1 9 17 5 700 1920 3.e..2. Then.4 kbps.ini file. CBR service users have the highest priority and are processed first. in the order established during the generation of the user distribution. Atoll manages the maximum number of users within each cell.4 HSUPA Part of the Algorithm HSPA VBR and BE service users active in the UL as well as all HSPA CBR service users (i. For more information. He obtains a total DL throughput of 483.4 Delayed M7 DC-HSDPA 2 9 16 7 800 1600 0 Delayed M9 2 6 13 12 900 960 3.4+1440).4+0).4 Connected M4 2 2 15 8 500 1440 1120+3. 4. unless they have been rejected during the R99 or HSDPA parts of the algorithm. The first user (here M6) is connected to an HSDPA bearer in his anchor cell and delayed in the other cell. He obtains a total DL throughput of 3.4 kbps (2403. it processes BE service users in the order established during the generation of the user distribution. are then evaluated by the HSUPA part of the algorithm.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 You can change the default weights by editing the atoll.4 Connected M1 DC-HSDPA 1 4 17 4 400 1920 1440 Connected M3 1 3 16 6 450 1600 800+3. active and inactive).Atoll 3.4 Connected M6 DC-HSDPA 2 1 13 11 600 960 480+3.4 Delayed M11 2 11 12 14 1250 800 3.4 Delayed M12 (DC-HSDPA) 1 2 12 14 15 1120 1440 0 Scheduler Saturation 0 Scheduler Saturation M13 2 13 17 1920 0 Scheduler Saturation 0 Scheduler Saturation M14 (DC-HSDPA) 1 2 14 13 15 960 1440 0 Scheduler Saturation 0 Scheduler Saturation M15 1 15 17 1920 0 Scheduler Saturation 0 Scheduler Saturation M16 (DC-HSDPA) 1 2 16 12 14 800 1120 0 Scheduler Saturation 0 Scheduler Saturation CQI CQI Rank RP The scheduled DC-HSDPA users have the following status: • • • The user ranked 4th (here M1) is connected to an HSDPA bearer in each cell. Let us assume there are 12 HSPA users in the cell: 258 .4 Delayed M8 1 7 14 9 800 1120 3.3.4 Delayed M10 1 10 14 10 1000 1120 3.4 kbps (483.4 Connected M5 2 8 18 3 550 2080 800+3. DL Best Bearer Obtained Connection Throughput (kbps) Status (kbps) Mobiles Carrier Simulation Rank M1 DC-HSDPA 2 4 19 2 300 2400 2400+3. He obtains a total DL throughput of 3843. Atoll considers VBR service users in the order established during the generation of the user distribution and lastly. This user is connected to one cell only.2. The capabilities of the category 3 user equipment are: • • • • • • Maximum Number of E-DPDCH codes: 2 TTI 2 ms: No so it supports 10 ms TTI Minimum Spreading Factor: 4 Maximum Block Size for a 2ms TTI: no value Maximum Block Size for a 10ms TTI: 14484 bits Highest Modulation Supported: QPSK Figure 4. 7 packet BE service users active on UL.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 • • • 3 CBR service users with any activity status.9: HSUPA UE Categories Table HSUPA bearer characteristics are provided in the HSUPA Bearer table.3. The cell supports HSPA+ functionalities. Atoll selects a list of HSUPA bearers for each user.1 Evaluation by the HSDPA HSUPA part of the Connection Status algorithm Mobiles Service Simulation Rank M1 CBR 4 Connected Yes M2 CBR 7 Connected Yes M3 CBR 9 Connected Yes M4 VBR 3 Connected Yes M5 VBR 5 Connected Yes M6 BE 1 Connected Yes M7 BE 2 Connected Yes M8 BE 6 Delayed Yes M9 BE 8 Delayed Yes M10 BE 10 Delayed Yes M11 BE 11 Delayed No M12 BE 12 Rejected No Admission Control During admission control. HSUPA user equipment categories are provided in the HSUPA User Equipment Categories table. An HSUPA bearer is described with following characteristics: • Radio Bearer Index: The bearer index number. The first two users have been connected to an HSDPA bearer. the last one has been rejected and the remaining four have been delayed in the HSDPA part. They have been connected to an HSDPA bearer. In this case. Atoll will consider the first ten HSPA users only and will reject the last two users in order not to exceed the maximum number of HSUPA bearer users allowed in the cell (their connection status is "HSUPA scheduler saturation").3.e the cell supports QPSK and 16QAM modulations in the UL.4. 4. the list of compatible bearers is restricted to HSUPA bearers that provide a peak RLC throughput between the maximum and the minimum throughput demands. For VBR service users. 2 packet VBR service users. The selected HSUPA bearers have to be compatible with the user equipment and capabilities of each HSUPA cell of the active set. the maximum number of HSUPA bearer users equals 10. For CBR service users. All of them have been connected to an HSDPA bearer.Atoll 3. Let us focus on one HSPA-BE service user with category 3 user equipment and a 50km/h mobility. i. the list is restricted to HSUPA bearers that provide a peak RLC throughput higher than the minimum throughput demand. 259 . Finally. etc. The minimum spreading factor used by the bearer is not less than the smallest spreading factor supported by the terminal (4). Ec req req Then.10: HSUPA Radio Bearers Table Then. Among the compatible HSUPA bearers.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks • • • • • • ©Forsk 2015 TTI Duration (ms): The TTI duration in ms.( ) L T = ------------------------------------------------------------------------------------------------------------------G Tx  G term 260 UL tx  ic  + I inter – carrier  ic  + N 0 . • HSUPA bearers using 16QAM modulation (improvement introduced by the release 7 of the 3GPP UTRA specifications. The TTI can be 2 or 10 ms. Here. Minimum Spreading Factor: The smallest spreading factor used. P term – HSUPA . during admission control. from the required Ec/Nt threshold. The HSUPA bearers compatible with category 3 user equipment are framed in red: Figure 4.3. addressing. overhead. Atoll considers an HSUPA bearer as compatible with the category 3 user equipment if: • • • • • The TTI duration used by the bearer is supported by the user equipment (10 ms). The modulation required by the bearer is supported by the terminal.). Number of E-DPDCH Codes: The number of E-DPDCH channels used. The transport block size does not exceed the maximum transport block size supported by the user equipment (14484 bits): The number of E-DPDCH channels required by the bearer does not exceed the maximum number of E-DPDCH channels that the terminal can use (2). referred to as HSPA+).  Nt E – DPDCH Ec req req UL P term – HSUPA =  ------  L T  N tot Nt E – DPDCH With UL tx UL intra N tot  ic  =  1 – F MUD   term   I tot UL extra  ic  + I tot L path  L Tx  L term  L body  L indoor  E Shadowing 8 . Transport Block Size (Bits): The transport block size in bits.Atoll 3.  ------ . the required Ec/Nt threshold to obtain this bearer is -21. HSUPA bearers can be classified into two categories: • HSUPA bearers using QPSK modulation: They can be selected for users connected to HSPA and HSPA+ capable cells.7dB. Atoll calculates the required terminal power. this is the index 1 HSUPA bearer. These HSUPA bearers can be allocated to users connected to cells with HSPA+ capabilities only. Atoll chooses the one with the lowest required Ec/Nt threshold. Atoll checks that the lowest compatible bearer in terms of the required E-DPDCH Ec⁄Nt does not require a terminal power higher than the maximum terminal power allowed. Atoll uses the HSUPA Bearer Selection table. Modulation: the modulation used (QPSK or 16QAM) Peak RLC Throughput (bps): The RLC peak throughput represents the peak throughput without coding (redundancy. 4. Noise rise scheduling and radio resource control are carried out on each user in order to determine the best HSUPA bearer that the user can obtain. Then. I tot UL tx .3. Several CBR service users can share the same HSUPA bearer. As explained before. Then. it repeats the same steps on VBR service users first.3. After the admission control on CBR service users. the number of channel elements and the Iub backhaul throughput). and lastly on BE service users.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 tx UL intra  term .2.2 HSUPA Bearer Allocation Process The HSUPA bearer allocation process depends on the type of service requested by the user. L T is calculated as follows: L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL -) L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term 261 . In the HSUPA coverage prediction. In the bearer allocation process shown below.11: HSUPA Bearer Selection Table req Atoll rejects the user if the terminal power required to obtain the lowest compatible HSUPA bearer ( P term – HSUPA ) exceeds the maximum terminal power (his connection status is "HSUPA Admission Rejection").4.e. the number of non-rejected HSUPA bearer users is n HSUPA . in the order established during the generation of the user distribution.. the terminal power used. CBR Service Users Let us focus on the three CBR service users mentioned in the example of the previous paragraph "HSUPA Part of the Algorithm" on page 258. I tot UL extra . followed by a radio resource control. All of them will be connected to an HSUPA bearer at the end. F MUD . with j = 1 to 3. CBR service users have the highest priority and are processed first. the 3 CBR service users are represented by Mj.Atoll 3. 8. in the order established during the generation of the user distribution. We assume that all of them have been admitted. I inter – carrier and N 0 are defined in "Inputs" on page 215. Atoll calculates the HSUPA bearer consumption ( C in %) for each user and takes into account this parameter when it determines the resources consumed by the user (i. Figure 4. At the end of this step. Atoll performs a noise rise scheduling. with j varying from 1 to 2: Determination of the best HSUPA bearer Sufficient Iub backhaul throughput to support the HSUPA bearer? No Is there a lower HSUPA bearer available? Yes No Yes Enough channel elements available to support the HSUPA bearer? Downgrading to lower HSUPA bearer Mj is rejected Yes No Is there a lower HSUPA bearer available? No Yes Mj is rejected Pterm-HSUPA recalculation and interference update No Mj = M2? Yes Resource allocation for packet (HSPA – Best Effort) service users Figure 4. We assume that all of them have been admitted. the 2 VBR service users are represented by Mj. In the bearer allocation process shown below. Mj. Mj. with j = 1 to 5. For the user. In the bearer allocation process shown below.3. with j varying from 1 to 3: Determination of the best HSUPA bearer B(Mj) Allocation of the minimum throughput demand to Mj Calculation of C(B(Mj)) Sufficient Iub backhaul throughput to support the HSUPA bearer? No Is there a lower HSUPA bearer available? Yes No Yes Enough channel elements available to support the HSUPA bearer? Downgrading to lower HSUPA bearer Mj is rejected Yes No Is there a lower HSUPA bearer available? No Yes Mj is rejected Pterm-HSUPA recalculation and interference update No Mj = M3? Yes Resource allocation for packet (HSPA – Variable Bit Rate) service users Figure 4. the 5 BE service users are represented by Mj. Noise rise scheduling and radio resource control are carried out on each user in order to determine the best HSUPA bearer that the user can obtain. 262 . We assume that all of them have been admitted.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 For the user.12: HSUPA Bearer Allocation Process for CBR Service Users VBR Service Users Let us focus on the two VBR service users mentioned in the example of the previous paragraph "HSUPA Part of the Algorithm" on page 258. Noise rise scheduling and radio resource control are carried out on each user in order to determine the best HSUPA bearer that the user can obtain.Atoll 3.13: HSUPA Bearer Allocation Process for VBR Service Users BE Service Users Let us focus on the five BE service users mentioned in the example of the previous paragraph "HSUPA Part of the Algorithm" on page 258. with j = 1 to 2. CBR service users have the highest priority and are processed first. For further  Nt E – DPDCH information on the calculation.14: HSUPA Bearer Allocation Process for BE Service Users 4. see "Uplink Load Factor Due to One User" on page 281. Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. with j varying from 1 to 5: Determination of the best HSUPA bearer Sufficient Iub backhaul throughput to support the HSUPA bearer? No Is there a lower HSUPA bearer available? Yes No Yes Downgrading to lower HSUPA bearer Mj is rejected Enough channel elements available to support the HSUPA bearer? Yes No Is there a lower HSUPA bearer available? No Yes Mj is rejected Pterm-HSUPA recalculation and interference update No Mj = M5? Figure 4.Atoll 3. each user is allocated a right to produce interference. Therefore. Ec- max 1  ----. UL X HSPA – CBR  txi ic  UL X user  txi ic  = -----------------------------------------------n HSPA – CBR Ec max From this value.– 1 UL X user  txi ic  UL X user Ec- max  ----for the Total noise option = -------------- Nt E – DPDCH UL F Then. Atoll selects the one with the lowest  ------ . This is the HSUPA bearer ( Index HSUPABearer ) UL TP P – R LC  Index HSUPABearer  with the highest potential throughput ( ----------------------------------------------------------------.3. it selects an HSUPA bearer. Atoll evenly shares the remaining cell load factor between the CBR service users admitted during the previous step ( n HSPA – CBR ). the noise rise scheduling algorithm attempts to evenly share the remaining cell load between the CBR service users admitted in admission control.for the Without useful signal option  Nt E – DPDCH = ------------------------------------------UL  txi ic  F ---------------------------------.3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 For the user. The allocation depends on the maximum E-DPDCH Ec⁄Nt allowed and on UE and cell capabilities. in terms of HSUPA.2. The remaining cell load factor on uplink UL ( X HSPA – CBR  txi ic  ) depends on the maximum load factor allowed on uplink and how much uplink load is produced by the served R99 traffic. Nt E – DPDCH 263 .3 Noise Rise Scheduling Determination of the Obtained HSUPA Bearer The obtained HSUPA radio bearer is the bearer that the user obtains after noise rise scheduling and radio resource control. It can be expressed as follows: UL UL UL X HSPA – CBR  txi ic  = X max  txi ic  – X R99  txi ic  Then. Mj. after the admission control.4. Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ) for each CBR service user.) where: N Rtx  Index HSUPABearer  • Ec- req Ec max  ----  ------  Nt E – DPDCH  Nt E – DPDCH • And P term – HSUPA  P term req max req Ec When several HSUPA bearers are available. During the noise rise scheduling. Atoll carries out radio resource control. the UL load factor allotted to each user is 0.5.5 dB) and a terminal power lower than the maximum terminal power allowed. The remaining cell load factor equal to 0. UL X HSPA – VBR  txi ic  UL X user  txi ic  = -----------------------------------------------n HSPA – VBR Ec max From this value. Therefore. UL X HSPA  txi ic  UL X user  txi ic  = -----------------------------------n HSPA Ec max From this value. While HSDPA sends data from one cell only. verifying if enough channel elements and Iub backhaul throughput are available for the HSUPA bearer assigned to the user.9 2 192 96 5 -13 2 256 128 6 -10. For information on radio resource control. Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ) as explained above and selects an  Nt E – DPDCH HSUPA bearer for each VBR service user. During the noise rise scheduling. Atoll carries out noise rise scheduling and radio resource control on VBR service users. Atoll carries out noise rise scheduling and radio resource control on BE service users. uplink soft handover impacts the scheduling operation. It provides a potential throughput of 128 kbps and requires E-DPDCH Ec⁄Nt of -13 dB (lower than -11. 264 . see "Radio Resource Control" on page 267.6 is shared between the BE service users. Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (the Without useful signal option is selected). Atoll distributes the remaining cell load factor available after all CBR and VBR service users have been served. Atoll carries out radio resource control on VBR service users. After the noise rise scheduling. HSUPA Bearers Index Required Ec/Nt Threshold (dB) Nb of Retransmissions Peak RLC Throughput (kbps) Potential Throughput (kbps) 1 -21.1 2 128 64 4 -13. For information on radio resource control.1. Ec max We have:  ------ = -11.5 dB  Nt E – DPDCH Here. with HSUPA all cells in the active set receive the transmission from the terminal. the obtained HSUPA bearer is the index 5 HSUPA bearer. Example: We have a cell with six BE service users. It can be expressed as follows: UL UL UL UL X HSPA – VBR  txi ic  = X max  txi ic  – X R99  txi ic  – X HSPA – CBR  txi ic  The remaining cell load factor is shared equally between the admitted VBR service users ( n HSPA – VBR ). For information on radio resource control. and neither CBR user nor VBR user. All BE service users have been admitted. see "Radio Resource Control" on page 267. After processing VBR service users. Atoll carries out radio resource control on BE service users. Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ) as explained above and selects an  Nt E – DPDCH HSUPA bearer for each BE service user.3. After the noise rise scheduling.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 After the noise rise scheduling. Let’s take the cell UL reuse factor equal to 1. all the cells are impacted by the transmission in terms of noise rise. see "Radio Resource Control" on page 267. . It can be expressed as follows: UL UL UL UL UL X HSPA  txi ic  = X max  txi ic  – X R99  txi ic  – X HSPA – CBR  txi ic  – X HSPA – VBR  txi ic  The remaining cell load factor is shared equally between the admitted BE service users ( n HSPA ). Atoll distributes the remaining cell load factor available after all CBR service users have been served.7 2 32 16 2 -19 2 64 32 3 -16. Therefore. After processing all CBR service users.1 2 512 256 7 -8 2 768 384 8 -7 2 1024 512 AtollAtollNoise Rise Scheduling in Soft Handover With HSUPA. As HSUPA bearer users in soft handover use the lowest granted noise rise. Atoll chooses the lowest of maximum terminal power allowed for each cell of the active set  tx k ic  . Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed max Ec  tx  ic  ) as explained in "HSUPA Bearer Allocation Process" on page 261. Atoll calculates the maximum terminal power allowed to obtain an HSUPA radio bearer max ( P term – HSUPA  tx k ic  ). (  ------ Nt E – DPDCH k For each cell of the active set  tx k ic  . Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ) after signal recombination of all HSUPA capable  Nt E – DPDCH cells of the active set 10.Atoll 3. it recalculates the maximum E-DPDCH Ec⁄Nt allowed Ec max (  ------  tx  ic  ) for each HSUPA-capable cell of the active set.( ) L T = ------------------------------------------------------------------------------------------------------------------G Tx  G term tx UL intra  term . F MUD .3. we have: 9. If selected. we have: max Ec- UL  ----= f rake efficiency   Nt E – DPDCH  max Ec-  ---- tx  ic   Nt E – DPDCH k txk  ActiveSet  samesite  Ec max = For soft (2/2) and soft-soft (3/3) handovers. max P term – HSUPA = min tx  AS k max  P term – HSUPA  tx k ic   max Once Atoll knows the selected maximum terminal power ( P term – HSUPA ).  Nt E – DPDCH k max P term – HSUPA Ec- max  ---- Nt E – DPDCH  tx k ic  = ----------------------------UL L T  N tot max Ec Then. we have:  ------  Nt E – DPDCH Ec- max Max   ---- tx  ic    Nt E – DPDCH k  txk  ActiveSet For softer-soft handover (2/3). In the HSUPA coverage prediction. I inter – carrier and N 0 are defined in "Inputs" on page 215.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 For each HSPA-capable cell of the active set  tx k ic  . I tot UL extra . I tot UL tx . For softer (1/2) and softer-softer (1/3) handovers. it depends on if the MRC option is selected (option available in Global parameters). max Ec max UL max P term – HSUPA  tx k ic  = min    ------  tx  ic   L T  N tot  P term  Nt E – DPDCH k With UL UL tx intra N tot  ic  =  1 – F MUD   term   I tot UL extra  ic  + I tot tx UL  ic  + I inter – carrier  ic  + N 0 L path  L Tx  L term  L body  L indoor  E Shadowing 9 . L T is calculated as follows: L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL -) L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term 265 . Nt E – DPDCH 10. In HSUPA coverage predictions. we have:  ------ = Nt E – DPDCH 266 Ec- max UL  Max   ----  Nt E – DPDCH  tx k ic    G macro – diversity  2links txk  ActiveSet . Atoll selects an HSUPA bearer as previously explained in "HSUPA Bearer Allocation Process" on page 261. Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. we have the following: Ec max UL For softer (1/2) and softer-softer (1/3) handovers:  ------ = f rake efficiency  Nt E – DPDCH  max Ec-  ---- Nt E – DPDCH  tx k ic  txk  ActiveSet  samesite  Ec max For soft handover (2/2):  ------ =  Nt E – DPDCH Ec- max UL Max   ---- tx  ic    G macro – diversity  2links   Nt E – DPDCH k  txk  ActiveSet Ec max For soft-soft handover (3/3):  ------ =  Nt E – DPDCH tx k Ec- max UL Max   ---- tx  ic    G macro – diversity  3links   Nt E – DPDCH k   ActiveSet For softer-soft handover (2/3). it depends on if the MRC option is selected (option available in Global parameters). This is the HSUPA bearer ( Index HSUPABearer ) with the highest potential throughput UL TP P – R LC  Index HSUPABearer  ( ----------------------------------------------------------------.tx  ActiveSet  f rake efficiency    Nt Nt k l E – DPDCH E – DPDCH   tx  samesite   tx k k Max  tx  othersite l max Ec Else.3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks max Ec-  ---- Nt E – DPDCH = ©Forsk 2015    UL  Ec- max Ec- max   -------- tx k ic     tx l ic  tx . Atoll selects the one with the lowest  ------ . The allocation depends on the maximum E-DPDCH Ec⁄Nt allowed and on UE and cell capabilities. If selected. we have: max Ec-  ----=  Nt E – DPDCH    UL  Ec- max Ec- max   --------f   tx  ic    tx  ic   txk .Atoll 3.) where: N Rtx  Index HSUPABearer  Ec- req Ec- max  ---- ---- Nt E – DPDCH   Nt E – DPDCH • Ec req When several HSUPA bearers are available.txl  ActiveSet  rake efficiency  Nt E – DPDCH l  Nt E – DPDCH k   tx k  samesite   tx k Max  tx  othersite l UL   G macro – diversity  2links max Ec Else. we have:  ------ =  Nt E – DPDCH Ec- max Max   ---- tx  ic    Nt E – DPDCH k  txk  ActiveSet Then. 4. do not change during 15 successive iterations: Atoll stops the algorithm at the 56th iteration without reaching convergence.  UL and/or  DL are still higher than their respective thresholds and from the 30th iteration. 2nd case: After 30 iterations. 100   UL = max  int  ----------------------------------------------------------------------------------UL UL    I tot  ic  k N user  ic  k      Atoll stops the algorithm if: 1st case: Between two successive iterations.8 dB and the requested HSUPA bearer is the index 7 HSUPA bearer. the maximum E-DPDCH Ec⁄Nt allowed is equal to -1. Atoll determines the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. Therefore. On the same hand.2. 267 . UL and DL convergence thresholds are set to 5. The simulation has reached convergence.4 Radio Resource Control Atoll checks to see if enough channel elements are available and if the Iub backhaul throughput is sufficient for the HSUPA bearer assigned to the user (taking into account the maximum number of channel elements defined for the site and the maximum Iub backhaul throughput allowed on the site in the uplink).3. The user is treated as if he is the only user in the cell. UL and DL convergence thresholds are set to 5.Atoll 3.  UL and  DL are lower than their respective thresholds (defined when creating a simulation). Example: Let us assume that the maximum number of iterations is 100. Convergence has not been reached. 100  = max int  ----------------------------------------------------------------------------------. they start decreasing slowly until the 40th iteration (without going under the thresholds) and then. If no channel elements are available. Examples: Let us assume that the maximum number of iterations is 100. Atoll allocates a lower HSUPA bearer ("downgrading") which needs fewer channel elements and consumes lower Iub backhaul throughput.2. 4.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Determination of the Requested HSUPA Bearer The requested HSUPA radio bearer is selected from the HSUPA bearers compatible with the user equipment. The simulation has not reached convergence (specific divergence symbol). 1.3 Results 4.8 dB) and a terminal power lower than the maximum terminal power allowed.3. 3rd case: After the last iteration. 4.5 Convergence Criteria The convergence criteria are evaluated for each iteration.  UL and/or  DL equal 80. After the 30th iteration. 2. 100  int --------------------------------------------------------------------------------------------DL     P tx  ic  k  N user  ic  k    UL UL UL UL   max   max  I tot  ic  k – I tot  ic  k – 1 N user  ic  k – N user  ic  k – 1 Stations Stations .4.3. After the 30th iteration. the simulation has not reached convergence (specific divergence symbol). It requires E-DPDCH Ec⁄Nt of -8 dB (lower than -1. If  UL and  DL are lower than their respective thresholds. the user is rejected. the user is rejected. if the maximum Iub backhaul throughput allowed on the site in the uplink is still exceeded even by using the lowest HSDPA bearer.  UL and/or  DL do not decrease during the next 15 successive iterations. the simulation has reached convergence. Convergence has been reached.3.3.3.1 R99 Related Results This table contains some R99 specific simulation results provided in the Cells and Mobiles tabs of the simulation property dialogue. If  UL  5 and  DL  5 between the 4th and the 5th iteration.  UL and/or  DL equal 100 and do not decrease during the next 15 successive iterations: Atoll stops the algorithm at the 46th iteration. If not. 100  int  --------------------------------------------------------------------------------------------. Atoll stops the algorithm after the 5th iteration. if we go on with the previous example. If  UL and/or  DL are still strictly higher than their respective thresholds. and can be written as follow:  DL DL DL   max  max  P tx  ic  k – P tx  ic  k – 1 N user  ic  k – N user  ic  k – 1  Stations Stations   . Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Name Value Unit Description Nb E1  T1  Ethernet E1  T1  Ethernet     TPIub – DL  N I    TP   RoundUp  Max     E1  T1  Ethernet  TP Iub – UL  N I   TP   None Number of E1/T1/Ethernet links required by the site None Downlink intra-cell interference at terminal on carrier ic W Downlink extra-cell interference at terminal on carrier ic W Downlink inter-carrier interference at terminal on carrier ic DL I intra  txi  DL  SCH  txi ic   P  txi ic  – P DL ----------------------------P tot  txi ic  – F ortho   BTS   tot  LT txi   ic  DL –  1 – F ortho    BTS  P b  txi ic   DL I extra  ic  DL P tot  txj ic  txj j  i  Ptot  txj icadj  DL DL I inter – carrier  ic  txj  j --------------------------------------------- RF  ic ic adj   DL I inter – techno log y  ic  ni DL DL I tot  ic  DL Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP n  ic DL DL DL DL Term I tot  ic  + N 0  Pb UL I tot  txi ic   ic  term W Total effective interference at terminal on carrier ic (after unscrambling) W Total received noise at terminal on carrier ic W Total power received at transmitter from intra-cell terminals using carrier ic W Total power received at transmitter from extra-cell terminals using carrier ic W Uplink inter-carrier interference at terminal on carrier ic txi UL extra I tot   txi ic  UL P b  ic  term txj j  i  Pb UL UL I inter – carrier  txi UL I tot  txi ic  UL N tot  txi ic  ic   ic adj  term txj j ----------------------------------RF  ic ic adj  UL extra I tot UL intra Tx  txi ic +  1 – F MUD   term  I tot UL I tot  txi ic  + Downlink inter-technology interference at terminal on carrier ic a i I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  N tot  ic  UL intra W UL  txi ic  +I inter – carrier  txi icW tx N0 Total received interference at transmitter on carrier ic W Total noise at transmitter on carrier ic (Uplink interference) None Cell uplink load factor on carrier ic UL UL I tot  txi ic  ---------------------------UL N tot  txi ic  UL I tot  txi ic  --------------------------------------------------------------------------------------UL intra Tx I tot  txi ic    1 – F MUD   term  UL 1 --------------------------UL F  txi ic  X  txi ic  UL F  txi ic  E  txi ic  268 None Cell uplink reuse factor on carrier ic None Cell uplink reuse efficiency factor on carrier ic .3. the lowest HSDPA bearer the user can obtain does not provide a peak RLC throughput higher than the minimum throughput demand. BE service users can be: • • • Either connected if they obtain an HSDPA bearer. 269 . the lowest HSDPA bearer the user can obtain does not provide a peak RLC throughput higher than the minimum throughput demand.3. HS-SCCH channels. Or delayed in case of lack of resources (HSDPA power. Statistics Tab In the Statistics tab. At the end of the HSDPA part. 4. the maximum Iub backhaul throughput allowed on the site in the uplink is exceeded. VBR service users can be: • • Either connected if they obtain an HSDPA bearer. there are no more channel elements available. Or rejected for the following reasons: the maximum number of HSDPA bearer users per cell is exceeded.2 HSPA Related Results At the end of the R99 part. the lowest compatible HSUPA bearer they can obtain does not provide a peak RLC throughput higher than the minimum throughput demand (only for CBR and VBR service users).3. HS-SCCH power.3. Atoll processes HSPA service users who are connected to an HSDPA bearer or were delayed in the previous step.+ 1 – F ortho   BTS DL P Tx  txi ic  --------------------------------------------------------------------------------------------------------------------------------1 . Or rejected for the following reasons: the maximum number of HSUPA bearer users per cell is exceeded. Atoll displays as results: • The number of rejected users. the users can be: • • Either connected and in this case. None Downlink reuse factor on a carrier ic DL dB Noise rise on downlink UL dB Noise rise on uplink In the case of an interfering GSM external network in frequency hopping. At the end.3. the HS-SCCH signal quality is not sufficient.3. the maximum Iub backhaul throughput allowed on the site in the downlink is exceeded.1 Either connected if they obtain an HSUPA bearer.2. Or rejected if the maximum number of HSDPA bearer users per cell is exceeded. the maximum Iub backhaul throughput allowed on the site in the downlink is exceeded.+ 1 – F ---------  ortho DL CI req DL X  txi ic  BTS DL Q req DL with CI req = --------DL Gp Simulation result available per mobile DL I tot  ic  -----------------DL N tot  ic  DL I tot  ic  ----------------------------DL I intra  txi ic  DL F  txi ic  DL – 10 log  1 – X  txi ic   UL – 10 log  1 – X  txi ic   NR  txi ic  NR  txi ic  a. there are no more OVSF codes available. they obtain the requested R99 bearer.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Name Value Unit Description None Downlink load factor on carrier ic Simulation result available per cell DL  I extra  ic   tch DL + I inter – carrier  ic    L T --------------------------------------------------------------------------------. OVSF codes). Or rejected for the following reasons: the maximum number of HSDPA bearer users per cell is exceeded. there are no more OVSF codes available. CBR service users can be: • • Either connected if they obtain an HSDPA bearer. they can be: • • 4. the HS-SCCH signal quality is not sufficient. In the HSUPA part. Only connected HSDPA and HSPA users are considered in the HSDPA part. the ICP value is weighted according to the fractional load. the terminal power required to obtain the lowest compatible HSUPA bearer exceeds the maximum terminal power. Or rejected exactly for the same reasons as R99 users.Atoll 3. Atoll determines the HSDPA bearer the user would obtain by considering the entire HSDPA power available of the cell. the uplink requested throughput corresponds to the peak throughput of ADPCH R99 radio bearer and the downlink requested throughput is the sum of the ADPCH radio bearer peak throughput and the peak RLC throughput(s) that the selected HSDPA radio bearer(s) can provide.3. In addition.Atoll 3. • The downlink and uplink peak throughputs ( TP P – D L and TP P – U L ) generated by their connection to R99 bearers. This figure includes R99 users as well as HSDPA and HSPA users since all of them request an R99 bearer. The number of R99 bearer users per activity status. Atoll indicates the uplink peak RLC throughput generated by active users connected with an HSUPA bearer UL ( TP HSUPA ): UL TP HSUPA =  UL TP P – RLC Active users UL TP P – RLC is the peak RLC throughput provided in the uplink. The requested HSUPA radio bearer is selected from the HSUPA 270 . R99 R99 Only active users are considered. Only HSPA service users are considered.  R99 TP P –D L = R99 R99 TP P – DL  R99 Bearer  and TP P – U L = Active users  R99 TP P – UL  R99 Bearer  Active users R99 R99 TP P – DL  R99 Bearer  is the downlink peak throughput of the user R99 radio bearer and TP P – UL  R99 Bearer  is the uplink peak throughput of the user R99 radio bearer.  DL TP HSDPA = DL TP P – RLC Active users DL TP P – RLC is the peak RLC throughput provided in the downlink. HSDPA and HSPA service users are considered since they all request an HSDPA bearer. only active users are taken into consideration in the downlink throughput calculation ( TP HSDPA ).3. Here. 4.2Mobiles Tab In the Mobiles tab. TP requested  M b  and TP requested  M b  ) For R99 users. • • The number of R99 bearer users per frequency band.3. The number of R99 bearer users connected to a cell (result of the R99 part). Atoll indicates for each user: • UL DL The uplink and downlink total requested throughputs in kbps (respectively. DL R99 UL R99 TP requested  M b  = TP P – DL  R99 Bearer  TP requested  M b  = TP P – UL  R99 Bearer  For HSDPA users. the DL and UL total requested throughputs correspond to the DL and UL peak throughputs of the R99 bearer associated to the service.2. the uplink requested throughput is equal to the sum of the ADPCH-EDPCCH radio bearer peak throughput and the peak RLC throughput of the requested HSUPA radio bearer. • The number of connected users with an HSDPA bearer (result of the HSDPA part) and the downlink peak RLC throughput they generate. • The number of connected users with an HSUPA bearer (result of the HSUPA part).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks • • ©Forsk 2015 The number of delayed users. the user is treated as if he is the only user in the cell and then. On DL the other hand. DL R99 DL R99 DL TP requested  M b  = TP P – DL  ADPCH R99 Bearer  + TP P – RLC for single-carrier users TP requested  M b  = TP P – DL  ADPCH R99 Bearer  AnchorCell +  DL TP P – RLC  c  for dual-carrier users c  Serving Cells UL R99 TP requested  M b  = TP P – UL  ADPCH R99 Bearer  For HSPA users. the downlink obtained throughput corresponds to the downlink peak throughput of the ADPCH-EDPCCH radio bearer in the anchor cell. In the downlink. the uplink and downlink total obtained throughputs are the sum of the ADPCHEDPCCH radio bearer peak throughput and the minimum throughput demand defined for the service. the downlink obtained throughput corresponds to the instantaneous throughput. For dual-carrier HSPA VBR and BE service users connected to two HSDPA bearers. the uplink obtained throughput is zero. Finally. If the user is delayed (he is only connected to an R99 radio bearer). if the user is rejected. If the user is rejected either in the R99 part or in the HSDPA part (i. Finally. the downlink obtained throughput is the sum of the peak throughput provided by the ADPCH-EDPCCH radio bearer in the anchor cell and the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource control. the uplink obtained throughput is the sum of the ADPCH-EDPCCH radio bearer peak throughput and the peak RLC throughput provided by the selected HSUPA radio bearer after noise rise scheduling. If the user is rejected. HSDPA bearer users can be connected to a single cell or to two adjacent cells of the same transmitter when the user has a DC-HSDPA-capable terminal and when the transmitter supports the multi-cell HSDPA mode. For a single-carrier HSDPA service user connected to an HSDPA bearer. the downlink obtained throughput is "0".3. the downlink obtained throughput corresponds to the downlink peak throughput of ADPCH-EDPCCH radio bearer. TP obtained  M b  and TP obtained  M b  ) For R99 service users. For a connected HSPA CBR service user. this is the sum of the peak throughput provided by the A-DPCH radio bearer in the anchor cell and the peak RLC throughputs provided by the selected HSDPA radio bearers after scheduling and radio resource control.e.. The requested HSDPA radio bearer is determined as explained in the previous paragraph.Atoll 3. If the user is connected to one cell and delayed in the other cell. If the user is delayed in the two cells (he is only connected to an R99 radio bearer in the anchor cell). Atoll determines the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. because the HSDPA scheduler is saturated). Finally. this is the sum of the peak throughput provided by the ADPCH-EDPCCH radio bearer in the anchor cell and the peak RLC throughputs provided by the selected HSDPA radio bearers after scheduling and radio resource control. HSDPA service users can only have a single-carrier connection. DL R99 DL R99 DL TP requested  M b  = TP P – DL  ADPCH – EDPCCH R99 Bearer  + TP P – RLC for single-carrier users TP requested  M b  = TP P – DL  ADPCH – EDPCCH R99 Bearer  AnchorCell +  DL TP P – RLC  c  for dual-carrier users c  Serving cells UL R99 UL TP requested  M b  = TP P – UL  ADPCH – EDPCCH R99 Bearer  + TP P – RLC • UL DL The uplink and downlink total obtained throughputs in kbps (respectively. If the user is delayed in the two cells (he is only connected to an R99 radio bearer in the anchor cell). if the user is rejected either in the R99 part or in the HSDPA part (i. the downlink obtained throughput corresponds to the instantaneous throughput. HSPA VBR and BE service users can only have a single-carrier connection. the user is treated as if he is the only user in the cell and then. downlink obtained throughput corresponds to the downlink peak throughput of the ADPCH radio bearer. the downlink obtained throughput is zero. If the user is rejected. the uplink and downlink total obtained throughputs are "0". the downlink obtained throughput is zero. If the user is rejected. Here. Otherwise.. The instantaneous throughput is the sum of the ADPCHEDPCCH radio bearer peak throughput and the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource control. this is the sum of the A-DPCH radio bearer peak throughput and the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource control. the uplink obtained throughput is zero. the downlink obtained throughput corresponds to the downlink peak throughput of the ADPCH radio bearer in the anchor cell. the obtained throughput is zero. • The mobile total power ( P term ) 271 . the downlink obtained throughput is the sum of the peak throughput provided by the A-DPCH radio bearer in the anchor cell and the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource control. the obtained throughput is the same as the requested throughput if he is connected without being downgraded.. if the user is connected to an HSDPA bearer. because the HSDPA scheduler is saturated). on downlink. if the user is rejected either in the R99 part or in the HSDPA part (i. the downlink obtained throughput is zero. because the HSDPA scheduler is saturated).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 bearers compatible with the user equipment. the downlink obtained throughput corresponds to the instantaneous throughput. the obtained throughput is lower (it corresponds to the peak throughput of the selected R99 bearer). In the uplink. For single-carrier HSPA VBR and BE service users.e. If the user is rejected. If the user is connected to one cell and delayed in the other cell. If the user is delayed. In uplink. the uplink obtained throughput corresponds to the uplink peak throughput of the ADPCH radio bearer.e. the downlink obtained throughput corresponds to the instantaneous throughput. When the user is connected to an HSUPA bearer. For a dual-carrier HSDPA service user connected to two HSDPA bearers. The downlink requested throughput is the sum of the ADPCH-EDPCCH radio bearer peak throughput and the peak RLC throughput(s) that the requested HSDPA radio bearer(s) can provide. When the user is either connected or delayed. etc. TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used. • The served HSDPA power in dBm (  P HSDPA  served ) This is the HSDPA power required to provide the HSDPA bearer user with the downlink obtained throughput. If the HSDPA bearer allocated to the user is the best one. if the HSDPA bearer has been downgraded in order to be compliant with cell and UE capabilities or for another reason. Knowing the HS-PDSCH Ec/Nt. of HSUPA Retransmissions (Required) .Atoll 3. • The number of OVSF codes This is the number of 512-bit length OVSF codes consumed by the user. f act –EDPCCH = 0.). The downlink obtained rate is the throughput experienced by the user after scheduling and radio resource control. HSPA BE and VBR service users. Atoll calculates the corresponding BLER. TP Offset and f TP – Scaling represent the scaling factor between the application throughput and the RLC (Radio Link Control) throughput and the throughput offset respectively. addressing.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 UL P term = P term – R99  f act –EDPCCH + P term – HSUPA for HSPA VBR and BE service users. overhead. UL For HSPA CBR service users. They are defined in the service properties. • The required HSDPA power in dBm (  P HSDPA  required ) It corresponds to the HSDPA power required to provide the HSDPA bearer user with the downlink requested throughput. • 272 The No.1 .  P HSDPA  required =  P HS – PDSCH  used + n HS – SCCH  P HS – SCCH  P HS – PDSCH  used is the HS-PDSCH power required to obtain the selected HSDPA bearer (in dBm). And  P HSDPA  served =  P HS – PDSCH  used  C HSDPABearer for HSPA CBR service users Where  P HS – PDSCH  used is the HS-PDSCH power required to obtain the selected HSDPA bearer.  DL TP P – RLC  c    1 – BLER HSDPA  DL c  Serving cells TP A  M b  = ------------------------------------------------------------------------------------------------------- f TP – Scaling – TP Offset TTI Where: DL TP P – RLC is the peak RLC throughput provided to the user by the selected HSDPA radio bearer after scheduling and radio resource control. And P term = P term – R99 for R99 and HSDPA users. In this case. UL P term = P term – R99  f act –EDPCCH + P term – HSUPA  C HSDPABearer for HSPA CBR service users. it is defined in the terminal user equipment category properties.  P HS – PDSCH used corresponds to the available HS-PDSCH power of the cell. Atoll determines the HSDPA bearer the user would obtain by considering the entire HSDPA power available of the cell. The downlink requested throughput is the throughput the user would obtain if he was the only user in the cell. This graph describes the variation of BLER as a function of the measured quality (HS-PDSCH Ec/Nt).  P HS – PDSCH  used will be lower than the available HS-PDSCH power of the cell. • DL The HSDPA application throughput in kbps ( TP A  M b  ) This is the net HSDPA throughput without coding (redundancy. On the other hand.3.  P HSDPA  served =  P HS – PDSCH  used + n HS – SCCH  P HS – SCCH for HSDPA users. BLER HSDPA is read in the quality graph defined for the triplet “reception equipment-selected bearer-mobility” (HSDPA Quality Graphs tab in the Reception equipment properties). These two parameters model the header information and other supplementary data that does not appear at the application level. if the user is connected to an HSUPA bearer. BLER HSUPA is the residual BLER after N Rtx retransmissions. 4. Atoll gives: • The available HSDPA power in the cell. It is read in the quality graph defined for the quartet “reception equipment-selected bearer-number of retransmissions-mobility” (HSUPA Quality Graphs tab in the Reception equipment properties). Knowing the EDPDCH Ec/Nt. This figure is read in the HSUPA Bearer Selection table. • The No.Atoll 3. The requested HSUPA radio bearer is selected from the HSUPA bearers compatible with the user equipment. The following columns appear if. Atoll determines the HSDPA bearer he would obtain by considering the entire HSDPA power available of the cell. the uplink obtained peak RLC throughput is not calculated.). For HSPA users. the user is treated as if he is the only user in the cell and then. For HSDPA users. the obtained uplink peak RLC throughput is the throughput provided by the selected HSUPA radio bearer after noise rise scheduling.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 The maximum number of retransmissions in order to have the requested HSUPA radio bearer with a given BLER.3 Cells Tab In the Cells tab. the uplink and downlink obtained peak RLC throughputs are the uplink and downlink minimum throughput demands defined for the service. you select "Detailed information about mobiles": • The uplink and downlink requested peak RLC throughputs (kbps) Downlink and uplink requested peak RLC throughputs are not calculated for R99 users. the downlink obtained peak RLC throughput is the throughput provided by the selected HSDPA radio bearer(s) after scheduling and radio resource control. On downlink. Or a simulation result when the option "HSDPA Power Dynamic Allocation" is selected. when creating the simulation.2. For a connected HSPA CBR service user. We have: P HSDPA  c  = P max  c  – P Headroom  c  – P tx – R99  c  – P HSUPA  c  273 . of HSUPA Retransmissions (Obtained) The maximum number of retransmissions in order to have the obtained HSUPA radio bearer with a given BLER.3. in dBm ( P HSDPA  c  ): This is: • • Either a fixed value in case of a static HSDPA power allocation strategy. If the user is connected to one or two HSDPA bearers in the downlink. the uplink peak RLC throughput is not calculated and the downlink requested peak RLC throughput is the throughput that the selected HSDPA radio bearer(s) can provide. and the downlink obtained peak RLC throughput is the throughput provided by the selected HSDPA radio bearer(s) after scheduling and radio resource control. the user is treated as if he is the only user in the cell and then. For connected HSPA BE and VBR service users. the requested uplink peak RLC throughput is the throughput of the requested HSUPA radio bearer. Atoll calculates the corresponding BLER. For HSDPA users connected to one or two HSDPA bearers. overhead.3. N Rtx is the maximum number of retransmissions for the obtained HSUPA bearer. the downlink requested peak RLC throughput is the throughput that the requested HSDPA radio bearer(s) can provide. • UL The HSUPA application throughput in kbps ( TP A  M b  ) This is the net HSUPA throughput without coding (redundancy. addressing. TP Offset and f TP – Scaling respectively represent the scaling factor between the application throughput and the RLC (Radio Link Control) throughput and the throughput offset. Atoll determines the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. They are defined in the service properties. These two parameters model the header information and other supplementary data that does not appear at the application level. Here.3. on uplink. c. This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt). UL TP P – RLC  M b    1 – BLER HSUPA   f TP – Scaling – TP Offset UL TP A  M b  = ---------------------------------------------------------------------------------------------------------------------------------------N Rtx Where: UL TP P – RLC is the peak RLC throughput provided by the selected HSUPA radio bearer after noise rise scheduling. if the user is connected to one or two HSDPA bearers. The requested HSDPA radio bearer is determined as explained in the previous paragraph. • The uplink and downlink obtained peak RLC throughput (kbps) Downlink and uplink obtained peak RLC throughputs are not calculated for R99 users. etc. Here. TTI  M b  is the minimum number of TTI (Transmission Time Interval) between two TTI used.e. • DL DL The instantaneous HSDPA Effective MAC Throughput in the cell. All these users are connected to the cell at the end of the HSDPA part of the simulation. it is defined for each HSDPA bearer in the HSDPA Radio Bearers table.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015  P tx – R99  c  = P pilot  c  + P SCH  c  + P OtherCCH  c  + with tch used for R99 users •  P tch  c  + DL P tch  c   f act –ADPCH tch used for HSPA users The transmitted HSDPA power in the cell. it is defined in the terminal user equipment category properties. c.Atoll 3.3. within one transmission time interval. DL R99 DL TP requested  M b  = TP P – DL  ADPCH R99 Bearer  + TP P – RLC   DL TP obtained  M s  + Ms  c R99 DL  TP P – DL  R99 Bearer  + TP P – RLC  M d – HSDPA   M d – HSDPA  c c is the anchor cell  + DL TP P – RLC  M d – HSDPA  + M c d – HSDPA c is the secondary cell DL TP Inst  cell  =  R99 DL  TP P – DL  R99 Bearer  + TP P – RLC  M d – HSPA   + M d – HSPA  c c is the anchor cell  DL M d – HSPA  c TP P – RLC  M d – HSPA  c is the secondary cell DL TP P – RLC is the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource control. We will differentiate single-carrier users (Ms) from DC-HSDPA users (Md-HSDPA stands for HSDPA BE and VBR users. HSDPA and HSPA users are considered since they all request an HSDPA bearer. they have a connection with the R99 bearer and an HSDPA bearer. in dBm ( P tx –H SDPA  c  ): It corresponds to the HSDPA power used to serve HSDPA bearer users. in kbps ( TPE –M AC  c  ) TP E – M AC  c  =  Mb  c S block  M b  --------------------------------------T TTI   TTI  M b  Where. it corresponds to the ADPCH-EDPCCH radio bearer. R99 TP P – DL  R99 Bearer  is the peak throughput of the ADPCH radio bearer if the user is an HSDPA user. i.  P tx –H SDPA  c  =  P HSDPA  M b   served Mb  c • The number of HSDPA users in the cell They are the connected and delayed HSDPA bearer users. in kbps ( TP Inst  c  ) This is the number of kilobits per second that the cell supports on downlink to provide simultaneous connected HSDPA bearer users with an HSDPA bearer. DC-HSDPA users are accounted for once in each serving cell. c. c. • DL The instantaneous HSDPA throughput in the cell. • The number of simultaneous HSDPA users in the cell ( n M ) b It corresponds to the number of connected HSDPA bearer users that the cell supports at a time. S block  M b  is the transport block size (in kbits) of the HSDPA bearer selected by the user. 274 . For HSPA users. DC-HSDPA users are accounted for once in each serving cell. and Md-HSPA refers to HSPA BE and VBR service users). 3. in kbps ( TP Av – Inst  c  ) DL TP Inst  c  DL TP Av – Inst  c  = -------------------nM b • DL The HSDPA application throughput in the cell. • DL The average instantaneous HSDPA throughput in the cell. it is defined in the terminal user equipment category properties. in kbps ( TP A  c  ) UL TP A  c  =  UL TP A  M b  Mb  c • UL The uplink cell load factor due to HSUPA traffic ( X HSUPA  c  ): UL  I tot  c   HSUPA UL X HSUPA  c  = --------------------------------UL N tot  c  Where UL  I tot  c   HSUPA is the total interference at transmitter received from HSUPA bearer users. f TP – Scaling and TP Offset respectively represent the scaling factor between the application throughput and the RLC (Radio Link Control) throughput and the throughput offset.if the scheduling algorithm is Round Robin or TTI Proportional Fair.if the scheduling algorithm is Max C/I.4 Sites Tab In the Sites tab. Knowing the HS-PDSCH Ec/Nt. This value is specified by the 3GPP.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 –3 T TTI is the TTI duration. c.3. BLER HSDPA is read in the quality graph defined for the triplet “reception equipment-selected bearer-mobility” (HSDPA Quality Graphs tab in the Reception equipment properties).2. in kbps ( TP A  c  ) DL Either TP A  c  =  DL M c b TP P – RLC  M b    1 – BLER HSDPA   f TP – Scaling – TP Offset -------------------------------------------------------------------------------------------------------------------------------------. Atoll calculates the corresponding BLER. These two parameters model the header information and other supplementary data that does not appear at the application level. DL TP P – RLC  M b  maxC  I     1 – BLER HSDPA   f TP – Scaling – TP Offset DL .3. DL TP P – RLC is the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource control. TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used. Or TPA  c  = -------------------------------------------------------------------------------------------------------------------------------------------------------------TTI M b  maxC  I  is the user with the highest C  I in the cell. • The maximum HSDPA RLC peak throughput in kbps ( DL max  TP P – RLC  M b   ) M b  cell It corresponds to the highest of RLC peak throughputs obtained by HSDPA bearer users connected to the cell. c.Atoll 3. 4. • The minimum HSDPA RLC peak throughput in kbps ( DL min  TP P – RLC  M b   ) M b  cell It corresponds to the lowest of RLC peak throughputs obtained by HSDPA bearer users connected to the cell. 2 10 s (2000 TTI in one second). Atoll displays: • DL The instantaneous HSDPA throughput carried by the site in kbps ( TP Inst  site  ) 275 . c.e. They are defined in the service properties. • The number of HSUPA users in the cell ( n M ): c They are the HSUPA bearer users connected to the cell. This graph describes the variation of BLER as a function of the measured quality (HS-PDSCH Ec/Nt). • UL The HSUPA application throughput in the cell. i. Therefore. during the resource control. Therefore.4 Appendices 4. The load rise ( X X UL UL ) is calculated as follows: 1 = ---------------------------------------------W 1 + -----------------------------------UL UL Q req  R nominal 4. Therefore.4. for each cell. if one channel needs 1 length-k/2 OVSF code. Atoll reserves for potential HSDPA bearer users: • HS – PDSCH – Min The minimum number of HS-PDSCH codes defined for the cell. To calculate the cell UL load factor. Then. In the R99 part. it can be rejected due to cell load saturation.2 Resources Management 4. thirty-two 512 bit-length OVSF codes). four 512 bit-length OVSF codes). If the cell supports HSDPA.e. 512 512-bit-length codes per cell are available in UMTS HSPA projects. It determines the number of codes that will be consumed by each cell. activity status assigned to users is not taken into account.e. Therefore. Here.3. Atoll will take four 512-bit-length codes.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks  DL TP Inst  site  = ©Forsk 2015 DL TP Inst  c  c  site • DL The instantaneous HSDPA Effective MAC Throughput carried by the site in kbps ( T MAC  site  in kbps) DL TP E – M AC  site  =  DL TP E –M AC  c  c  site • UL UL The HSUPA throughput carried by the site in kbps ( TP  site  ) TP  site  =  UL TP obtained  M c  M c  site 4. Therefore.e. OVSF codes form a binary tree. Atoll will take 4  n HS – SCCH 512-bit-length codes. for each cell. for TCH (traffic channels). Atoll will take 32  N Codes • 512-bit-length codes.3.4. two 512 bit-length OVSF codes). or it estimates a load rise due to the mobile and adds it to the current load. N Codes . So even if the mobile is not active on UL. first in the R99 part and then in the HSDPA part. either Atoll takes into account the mobile power determined during power control if mobile was connected in previous iteration. Atoll calculates the uplink load factor of a considered cell assuming the mobile concerned is connected to it. A code per cell-receiver link. it is equivalent to use 2 length-k OVSF codes.3.3. Length-k OVSF codes are generated from length-k/2 OVSF codes. or 4 length-2k OVSF codes and so on. Atoll allocates codes for the DL channels used for HSUPA: • • A 128 bit-length code for the E-HICH and E-RGCH channels (i. Atoll will take two 512-bit-length codes. for each cell.1 Admission Control in the R99 Part During admission control in the R99 part of the simulation.4. We have: . for each cell. two 512 bit-length OVSF codes). The length of code to be allocated.Atoll 3.1 OVSF Codes Management OVSF codes are managed in the downlink during the simulation since this resource is downlink limited only.e. Therefore. They are 16-bit length OVSF codes HS – PDSCH – Min (i. Code_Length. Codes of longer lengths are generated from codes of a shorter length. Atoll checks the availability of this resource during the simulation. Atoll will Overhead take 2  N Codes • 276 512-bit-length codes. A 256 bit-length code for the E-AGCH channel (i. it allocates to the cell OVSF codes to support R99 bearers required by users: • A 256 bit-length code per common channel (i.3. Atoll determines the number of 512 bit-length codes that will be consumed for each cell. A 128 bit-length code per HS-SCCH channel (i.2. If the cell supports HSPA.e. depends on the user activity. four 512 bit-length OVSF codes). one DC-HSDPA user consumes OVSF codes in both cells. This user is active on DL while connected to a cell (which does not support HSDPA). In DC-HSDPA. The spreading factor for active users has been set to 64 and site equipment requires four overhead downlink channel elements per cell.15: OVSF Code Tree Indices (Not OVSF Code Numbers) The OVSF code allocation follows the “Buddy” algorithm. ….3. all of its ancestors with lengths k/2.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 DL Either Code_Length = F spreading  Active user  when the user is active. HS – PDSCH For HSPA CBR service users. 4k. HSPA VBR and BE service users. the OVSF code allocation follows the mobile connection order (mobile order in the Mobiles tab). DC-HSDPA users have two HSDPA bearers. Atoll allocates OVSF codes for each cell-mobile link while it globally assigns channel elements to a site. N Codes 512-bit-length codes for each user  C HSDPABearer 512-bit-length codes for each user connected is the number of HS-PDSCH channels required by the HSDPA bearer. Atoll needs 32  N Codes HS – PDSCH to the cell. k/4.e. cannot be used as they will not be orthogonal. In the HSDPA part. TCH The number of 512 bit-length OVSF codes needed N Codes is calculated from the length of the code to be allocated as follows: TCH 512 N Codes = ------------------------------Code_Length Figure 4.Atoll 3. which guarantees that: • • If a k-length OVSF code is used. eight additional 512 bit-length OVSF codes). one for each serving cell. • • • In the R99 part. Therefore. all of its children with lengths 2k. If a k-length OVSF code is used. This figure depends on the HSDPA bearer assigned to the user and on the type of service. Therefore. a DCHSDPA user requires R99 resources in the best serving cell only and consumes the same amount of R99 resources as a single-cell HSDPA user. Atoll will consume four 256 bit-length OVSF codes for common channels (i. DL Or Code_Length = F spreading  Inactive user  if the user is inactive. 277 . The OVSF code and channel element management is differently dealt with in case of “softer” handover. for HS-PDSCH. HS – PDSCH For HSDPA users. Example: We consider a user with a service requiring the UDD64 R99 radio bearer. Atoll needs 32  N Codes HS – PDSCH connected to the cell. HSDPA and HSPA users are assigned an HSDPA bearer (Fast link adaptation). Atoll allocates to the cell: • 16-bit length OVSF codes per cell-receiver. eight 512 bit-length OVSF codes) and a 64 bit-length OVSF code for traffic channels (i. Therefore. ….e. cannot be used as they will not be orthogonal. N Codes is the number of HS-PDSCH channels required by the HSDPA bearer. A-DPCH is only transmitted in the anchor carrier. per cell-receiver link. if no HSDPA bearer user is connected. in the HSDPA part. This figure includes: • Channel elements for R99 bearers Overhead • N CE – DL • N CE – DL R99 – T CH channel elements for control channels (Pilot channel. N CE – UL  N I  . is: N CE – UL  N I  =  NCE – UL  j  j  NI In the downlink. In the uplink. Atoll carries out a resource control in the uplink after allocating HSUPA bearers.2 Channel Elements Management Channel elements are controlled in the R99 and the HSUPA parts of the simulation. Atoll still keeps the codes for E-HICH. HSUPA  C HSUPABearer per cell-receiver link. TP Iub – UL  j  . includes: 278 . the number of channel elements required in the uplink at the site level.4. In case of “softer” handover (the mobile has several links with co-site cells). the HSDPA and the HSUPA parts of the simulation. This is the same with HSUPA bearer users. It takes into account the channel elements consumed by HSUPA bearer users in the uplink and recalculates the number of channel elements required by each site in the uplink. for packet (HSPA .Atoll 3. in the HSUPA part.2. for R99 TCH (traffic channels). On the other hand. in the HSUPA part. is: N CE – DL  N I  =  NCE – DL  j  j  NI • • 4. This figure includes: • • Channel elements for R99 bearers: Overhead • N CE – UL • R99 – T CH N CE – UL channel elements for control channels. Therefore. Even if no HSUPA bearer user is connected to the cell. N CE – DL  N I  . Atoll consumes N CE – UL  j  channel elements for each cell j on a site NI. Then. In the R99 part. It takes into account the Iub backhaul throughput consumed by HSDPA bearer users in the downlink and recalculates the Iub backhaul throughput required by each site in the downlink. E-RGCH and EAGCH channels. Therefore. In the R99 part.3. Iub Backhaul Throughput The Iub backhaul throughput is controlled in the R99. Atoll determines the number of channel elements required by each site for R99 bearers in the uplink and downlink. Finally.3. per cell-receiver link.BE) and packet (HSPA . common channels).VBR) service users.2. Atoll still keeps these codes and the codes for HS-SCCH too.3. In the uplink. Atoll checks the availability of this resource in the uplink and downlink.4. Channel elements for HSUPA bearers: HSUPA per cell-receiver link. A-DPCH is only transmitted on the anchor carrier. the number of channel elements required in the downlink at the site level. during the resource control.3 In DC-HSDPA. Atoll gives the cell HS – PDSCH – Min back the minimum number of OVSF codes reserved for HS-PDSCH ( N Codes ).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 When HSDPA bearer users (at least one) are connected to the cell. the Iub backhaul throughput consumed by each cell j on a site NI. Atoll determines the Iub throughput required by each site for R99 bearers in the uplink and downlink. Atoll carries out another resource control after allocating HSUPA bearers. for R99 TCH (traffic channels). It takes into account the Iub backhaul throughput consumed by HSUPA bearer users in the uplink and updates the Iub backhaul throughput required by each site in the uplink. Atoll checks the availability of this resource in the uplink and downlink. Atoll consumes N CE – DL  j  channel elements for each cell j on a site NI. for CBR service users. 4. Synchronisation channel. Atoll allocates channel elements for the best serving cell-mobile link only. a DCHSDPA user requires R99 resources in the best serving cell only and consumes the same amount of R99 resources as a single-cell HSDPA user. Then. during the resource control. Atoll performs a resource control in the downlink after allocating HSDPA bearers. • N CE • N CE Therefore. 4. for HSPA CBR service users.4. HSPA BE and VBR service users. TP Iub – DL  j  .Atoll 3. the Iub backhaul throughput required on downlink at the site level. the Iub backhaul throughput consumed by each cell j on a site NI. The Iub backhaul throughput required for HSUPA bearers: HSUPA • TP Iub per cell-receiver link.4.3. required quality is limited to the effective contribution of the transmitter. • HSUPA TP Iub  C HSUPABearer per cell-receiver link. for R99 TCH (traffic channels). 4. TP Iub – DL  N I  . for HSDPA. Let CI req = --------DL Gp DL DL G p and Q req are the processing gain on downlink and the Eb/Nt target on downlink respectively.3.be the required quality. The Iub backhaul throughput required for HSDPA bearers: • TP Iub HSDPA per cell-receiver link. In case of soft-handoff. Therefore. In case of “softer” handover (the mobile has several links with co-site cells). is: TP Iub – UL  N I  =  TPIub – UL  j  j  NI In the downlink. per cell-receiver link. • HSDPA TP Iub  C HSDPABearer per cell-receiver link. is: TP Iub – DL  N I  =  TPIub – DL  j  j  NI • • In DC-HSDPA. Synchronisation channel.3 Downlink Load Factor Calculation Atoll calculates a downlink load factor for each cell (available in the Cells tab of any simulation result) and each connected mobile (available in the Mobiles tab of any given simulation result). Therefore. On the other hand. Iub backhaul throughput is consumed by the best serving cell-mobile link only. common channels). for HSPA BE and VBR service users. for HSPA CBR service users. includes: • • The Iub backhaul throughput required for R99 bearers: Overhead • TP Iub – DL • R99 – T CH TP Iub – DL for R99 control channels (Pilot channel. for R99 TCH (traffic channels). the DC-HSDPA user has two HSDPA bearers (one for each serving cell) and consumes HSDPA resources in both cells. HSDPA With TP Iub DL HSDPA = TP P – RLC + Overhead Iub DL  TP P – RLC Therefore. a DCHSDPA user requires R99 resources in the best serving cell only and consumes the same amount of R99 resources as a single-cell HSDPA user.3. DL P tx  c  = P pilot  c  + P SCH  c  + P otherCCH  c  +  Ptch  c  tch DL ortho nonOrtho P tx  c  = P CCH  c  + P CCH c +  Ptch  c  tch where 279 .3. the Iub backhaul throughput required on uplink at the site level.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 • The Iub backhaul throughput required for R99 bearers: • • R99 – T CH TP Iub – UL per cell-receiver link. DL Q req . TP Iub – UL  N I  .1 Downlink Load Factor per Cell Approach for downlink load factor evaluation is highly inspired by the downlink load factor defined in the book “WCDMA for UMTS by Harry Holma and Antti Toskala”. A-DPCH is only transmitted on the anchor carrier. In case of an HSDPA bearer user.+ N term +  1 – F ortho   BTS   ------------------------------------------------------------------------------0 LT LT    L r T    DL  I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r +  1 – F ortho   BTS   P tx  c   r + nonOrtho term F ortho   BTS  P CCH  c   r + N0  LT  r P tch  ic  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------1 -------------------. DL r = f act – ADPCH . I inter – carrier  c  is the inter-carrier interference received at the terminal.Atoll 3. Ptch: term P tch  c  = CI req   I extra  c  + I inter – carrier  c  + I inter – techno log y  c  + I intra  c  + N 0   LT  r DL With r = 1 when the user is active on the downlink and r = r c when the user is inactive. we have a required power.+ 1 tch -------------------.3.+  1 – F ortho   BTS  tch CI req  r  .+ -----------------------------.   P tch  c  = CI req     I extra  c  + I inter – carrier  c  + I inter – techno log y  c  DL nonOrtho nonOrtho  P tx  c  – P CCH  c  – P tch  c   P CCH c .+ CCH CCH 1 -------------------.  P DL  ic  P tx  c  –   tx 1 . P DL tx  c  DL DL P tx  c  P tx  c  = P ortho  c  + P nonOrtho  c  + ---------------------------------------------------------------------------------------------------------------------------------------------------------------. I extra  c  is the total power received at the receiver from other cells.1 F ------------------+  – ortho   BTS  tch CI req  r   I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r -----------------------------------------------------------------------------------------------------------------------------------------. We have: ortho nonOrtho P CCH  c  + P CCH DL P tx  c  = c    I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r    DL nonOrtho term   + 1 – F     P  c   r + F    P  c   r + N  L  r  ortho BTS tx ortho BTS CCH 0 T  ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- +  1   ------------------1 F +  –    ortho BTS tch   CI req  r       I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r ------------------------------------------------------------------------------------------------------------------------------------------.+ 1 – F ------------------ ortho   BTS  tch  CI req  r    ortho nonOrtho = P CCH  c  + P CCH 280 c + nonOrtho term F ortho   BTS  P CCH  c   r + N0  LT  r ----------------------------------------------------------------------------------------------------------------1 -------------------.+  1 – F ortho   BTS  CIreq  r I intra  c  is the total power received at the receiver from the cell with which it is connected.+  1 – F ortho   BTS  CI req  r  nonOrtho term F ortho   BTS  P CCH  c   r + N0  LT  r ----------------------------------------------------------------------------------------------------------------1 . I inter – techno log y  c  is the inter-technology interference received at the terminal from an external transmitter.+  1 – F ortho   BTS  tch CI req  r  DL  1 – F ortho   BTS   P tx  c   r --------------------------------------------------------------------------.+ 1 – F ortho   BTS  r  DL   P  c  DL tx -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 ortho P CCH  c  = P pilot  c  + P otherCCH  c  nonOrtho P CCH  c  = P SCH  c  At mobile level. we have: K 281 . I intra  F UL Q UL Q req  k   ------------------------req  k   -------------------------.+ ----------------------------------------------------------------R99 R99 TP P – UL  k  TP P – UL  k  UL UL 1 + Q req  k   -------------------------1 + Q req  k   -------------------------W W R99 req TP P – UL  k  Ec UL We note  -----.+ 1 – F ------------------ ortho   BTS  tch   r CI req   ortho nonOrtho P CCH  c  + P CCH c +   Therefore. k     Nt. Without Useful Signal Option UL  P b  k   req W .3. N 0 W W = ----------------------------------------------------------------------------------.+ -----------------------------------------------------        1 1 .---------------------------------------------------------------------------------------------------------------UL  Q req  k  = ------------------------R99 UL tx TP P – UL  k  I intra –  P b  k   req + I extra + I inter – carrier + N 0 UL  P b  k   req W UL .  I intra  F W W   R99 UL  P b  k   req R99 TP P – UL  k  TP P – UL  k  UL tx . k  = Q req  k   ------------------------ Nt  E – DPDCH W UL tx I intra  F N0 UL  P b  k   req = -----------------------------------------------------.3.1 F ------------------+  – ortho   BTS  tch CIreq  r DL P tx  c  = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r -----------------------------------------------------------------------------------------------------------------------------------------.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 nonOrtho term  c   r + N0  LT  r F ortho   BTS  P CCH ----------------------------------------------------------------------------------------------------------------1 . the downlink load factor can be expressed as: X DL  I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r ------------------------------------------------------------------------------------------------------------------------------------------. we assume that the cell UL reuse factor ( F  txi ic  ) is constant.3.+ 1 – F ortho   BTS  r DL P tx  c  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------1 . UL In this calculation.4.Atoll 3.3.4 Uplink Load Factor Due to One User UL This part details how Atoll calculates the contribution of one user to the UL load factor ( X k ).+ 1 – F ------------------   tch ortho BTS CI req  r  The downlink load factor represents the signal degradation in relation to the reference interference (thermal noise plus synchronisation channel power).+ 1  --------------------------------------. The result depends on the option used to calculate Nt (Without useful signal or Total noise that you may select in Global parameters).2 Downlink Load Factor per Mobile Atoll evaluates the downlink load factor for any connected mobile as follows: X DL DL I tot  c  = ---------------DL N tot  c  4. 4. -----------------------------------------------------------------------Q req  k  = ------------------------UL R99 UL tx TP P – UL  k  I intra  F –  P b  k   req + N 0 R99 R99  TP P – UL  k  TP P – UL  k  UL UL UL tx - = Q UL  P b  k   req   1 + Qreq  k   ------------------------+ N0  req  k   -------------------------.+ 1  --------------------------------------req req Ec  Ec    ----   ----   Nt.4.+ 1 – F ortho   BTS  r  DL   P  c  tx ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 1–   1 . k   E – DPDCH E – DPDCH As I intra =   Pb UL  k   req . + 1  --------------------------------------req Ec    ----   Nt. req   ----   Nt  k   E – DPDCH K   1 . req   ----   Nt  k   E – DPDCH 1  -----------------------------------------------------    1 .+ 1  --------------------------------------req Ec    ----   Nt.  I intra  F UL + N tx  P b  k   req = Q req  k   ------------------------0  W R99 req TP P – UL  k  Ec UL We note  -----. k   E – DPDCH So. k    I intra  F + N 0   Nt  E – DPDCH 282 .+ N 0   ------------------------------------------------------ -----------------------------------------------------    K N0  ©Forsk 2015 tx   1 .--------------------------------------UL Q req  k  = ------------------------ UL R99 tx TP P – UL  k  I intra  F + N 0 R99 TP P – UL  k  UL UL .+ 1  --------------------------------------req Ec    ----   Nt. k   E – DPDCH  X UL UL I intra + I extra + I inter – carrier I intra  F 1 = ------------------------------------------------------------------------------= --------------------------------------= ---------------------------------UL tx tx tx I intra + I extra + I inter – carrier + N 0 I intra  F + N 0 N0 1 + ------------------------UL I intra  F Therefore. we have: X UL = F UL  1  -----------------------------------------------------  K   1 .Atoll 3. k   E – DPDCH K I intra  UL tx N0  F I intra = -------------------------------------------------------------------------------------1 ----------------------------------------------------------------------------–1 UL 1 F  -----------------------------------------------------   K  1 . k   E – DPDCH = -------------------------------------------------------------------------------------UL 1 1–F  -----------------------------------------------------   K  1 .3.+ 1  --------------------------------------Ec.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks I intra = I intra  F UL  1 tx 1 .+ 1  --------------------------------------req Ec    ----   Nt. k   E – DPDCH Total Noise Option UL  P b  k   req W UL . we can conclude that the contribution of one user to the UL load is defined as: UL X k = F UL 1  ------------------------------------------------------    1 .+ 1  --------------------------------------Ec.+ 1  --------------------------------------req Ec    ----   Nt. ------------------------------------------------------------------------------Q req  k  = ------------------------R99 tx TP P – UL  k  I intra + I extra + I inter – carrier + N 0 UL  P b  k   req W . k  = Q req  k   ------------------------ Nt  E – DPDCH W req Ec UL UL tx  P b  k   req =  -----. best serving cell determination used to be performed by selecting the best carrier within transmitters according to the selected method (site equipment) and then the best transmitter using the best carrier. we have: X UL = F UL   req Ec. we will consider the most common scenario.1dBm . P HSDPA  Tx c 2  = P max  Tx  – P tx – R99  Tx c 1  – P tx – R99  Tx c 2  – P Headroom  Tx c 2  = 44. In this case. we have: P max  Tx c 1  = 43dBm and P tx – R99  Tx c 1  = 39. the HSDPA cell can use 100% of the available power. P tx – R99  Tx c 2  = 36.6 Best Serving Cell Determination in Monte Carlo Simulations . k    ----Nt  E – DPDCH Ec K tx N0  req .1dBm and P Headroom  Tx c 2  = 0dB .0.e.3. we have: P max  Tx c 2  = 43dBm . we have: P max  Tx c 1  = 43dBm and P tx – R99  Tx c 1  = 39.1dBm . In this part. To switch back to this method. • 1st case: Inter-carrier power sharing is not activated On c1.3.3. k    ----Nt  E – DPDCH Ec K I intra = ------------------------------------------------------------UL 1–F X UL UL I intra + I extra + I inter – carrier I intra  F 1 = ------------------------------------------------------------------------------= --------------------------------------= ---------------------------------UL tx tx tx I intra  F + N 0 N0 I intra + I extra + I inter – carrier + N 0 1 + ------------------------UL I intra  F Therefore. Let’s take the following example to measure the impact of the inter-carrier power sharing.5 Inter-carrier Power Sharing Modelling Inter-carrier power sharing enables the network to dynamically allocate available power from R99-only and HSDPA carriers among HSDPA carriers. On c2. we have: K I intra =  I intra  F UL tx + N0   req .  ----k  Nt  E – DPDCH K So.ini file: [CDMA] MultiBandSimu = 0 The method is described below: 283 . k  Nt E – DPDCH 4. we can conclude that the contribution of one user to the UL load is defined as: UL X k = F UL req Ec   -----. all of the R99-only cell’s unused power can be allocated to the HSDPA cell.Atoll 3.8. i.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1   Pb UL As I intra =  k   req . the maximum power of the HSDPA cell must be set to the same value as the maximum shared power in order to use power sharing efficiently. Therefore. a network consisting of an R99-only carrier (c1) and an HSDPA carrier with dynamic power allocation (c2) (c2 does not support HSUPA). P tx – R99  Tx c 2  = 36.1dBm and P Headroom  Tx c 2  = 0dB . On c2. Therefore.Old Method Before Atoll 2. As explained in The User Manual. add the following lines in the Atoll. we have: P max  Tx c 2  = 46dBm . P HSDPA  Tx c 2  = P max  Tx c 2  – P tx – R99  Tx c 2  – P Headroom  Tx c 2  = 42dBm • 2nd case: Inter-carrier power sharing is activated and P max  Tx  = 46dBm On c1.4.4.4dBm 4. 284 .+ X X k  txi ic  = ---------------------------UL N tot  txi ic  EndFor UL BestCarrier k  txi M b  is the carrier with the lowest X k  txi ic  Else if carrier selection mode is “Min. we calculate current loading factor: UL I tot  txi ic  UL UL . If carrier selection mode is “Min.3. DL Total Power” BestCarrier k  txi M b  is the carrier with the lowest P tx  txi ic  k Else if carrier selection mode is “Random” BestCarrier k  txi M b  is randomly selected Else if carrier selection mode is "Sequential" UL UL BestCarrier k  txi M b  is the first carrier so that X k  txi ic   X max Calculation of    BTS  P c  txi M b BestCarrier  Q pilot  txi BestCarrier  = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------k DL DL   P tot  txi BestCarrier k  txi M b   + I extra  BestCarrier k  txi M b   +    DL DL Term  I  BestCarrier  txi  M   + I  BestCarrier  txi  M   + N  inter – carrier  k b inter – techno log y k b 0 If user selects “without Pilot”    BTS  P c  txi M b BestCarrier  Q pilot  txi BestCarrier  = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------k   DL DL P tot  txi BestCarrier k  txi M b   + I extra  BestCarrier k  txi M b       DL  + I DL  BestCarrier k  txi M b   + I inter – techno log y  BestCarrier k  txi M b    inter – carrier   Term   N + –  1 –      P  txi  M  BestCarrier  0 BTS c b   Rejection of station txi if the pilot is not received pilot If Q pilot  txi M b BestCarrier   Q req  Mobility  M b   then txi is rejected by Mb k max If Q pilot  txi M b BestCarrier   Q pilot  M b  k k Admission control (If simulation respects a loading factor constraint and Mb was not connected in previous iteration).Atoll 3. Determination of BestCarrier k  txi M b  . UL UL If X k  txi BestCarrier  txi M b    X max . UL Load Factor” For each carrier ic used by txi.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 For each station txi containing Mb in its calculation area and using a frequency band supported by the Mb’s terminal. then txi is rejected by Mb Else max Q pilot  M b  = Q pilot  txi M b BestCarrier  k k Tx BS  M b  = txi Endif EndFor If no TxBS has been selected. Mb has failed to be connected to the network and is rejected. If a given carrier is specified for the service requested by Mb and if it is used by txi BestCarrier k  txi M b  is the carrier specified for the service. Else the carrier selection mode defined for txi is considered. These parameters can be results of a given simulation. no default value is used for the HSDPA power. this parameter must be defined by the user. Uplink reuse factor = 1 Uplink load factor due to HSUPA = 0% Maximum uplink load factor = 75% On the other hand.4. They must also belong to layers supported by the service and the terminal. average values calculated from a group of simulations. For more information about setting options in the Atoll. The cell UL reuse factor. Atoll first selects the cells which belong to the highest priority layer and then. Atollcalculates the best server indicator ( I BS ) for the best serving cell candidate and the other potential serving cells ( c OC ): I BS  c BC  = Q pilot  c BC  + M HO  c BC  + C IO  c BC  for the best serving cell candidate. a service and a mobility type. You can make the prediction for a specific carrier. The active set may consist of one or more cells depending on whether the service supports soft handover and on the terminal active set size. For a DB-MC-HSDPA user. MC-HSDPA and DB-MC-HSDPA users. I BS  c OC  = Q pilot  c OC  + C IO  c OC  for the other potential serving cells.3. The cell with the highest I BS is selected as the best serving cell if its best server indicator ( I BS ) exceeds the Ec/I0 threshold defined in the properties of the mobility type.1 Best Serving Cell and Active Set Determination The mobile active set is the list of the cells to which the mobile is connected. Atoll ranks the potential serving cells according to the best server indicator ( I BS ). select "Best (All bands)" as the carrier or layers associated with several carriers on different frequency bands. In addition. the pilot signal level received from these cells must exceed the defined minimum RSCP threshold. The analysis is based on the following parameters: • • • The uplink load factor and the downlink total power of cells. selecting one specific carrier or one layer associated with one unique carrier is not suitable. or for all layers. the quality of the pilot ( Q pilot ). This receiver does not create any interference.Atoll 3.2.1.ini file. Results are displayed for any point of the map where the pilot signal level exceeds the defined minimum RSCP. the cell UL load factor due to HSUPA and the maximum cell UL load factor for HSUPA bearer users. The pilot quality difference between the cell and the best serving cell must not exceed the AS-threshold set per cell. For DC-HSDPA. select "Best (All/Specific band)" as the carrier or layers associated with several carriers. and these layers must support a speed higher than the user mobility. Atoll uses the following default values: • • • • • Total transmitted power = 50% of the maximum power (i. It must belong to the neighbour list of the best serving cell if it is located on a site where the equipment imposes this restriction (the “restricted to neighbours” option selected in the equipment properties). by setting an option in the Atoll.e. In the last case. or user-defined cell inputs. Among all potential serving cells.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4. for a specific layer.AS Analysis Tab Let us suppose a receiver with a terminal. If you have selected a DC-HSDPA user or a MC-HSDPA user. This cell is referred to as the best serving cell candidate ( c BC ). Potential serving cells must use a frequency band with which the terminal is compatible. The best serving cell and other cells of the active set are selected among a list of potential serving cells which fulfil a set of conditions. You can return to the old best serving cell selection mechanism as in Atoll 3. see the Administrator Manual. The layer priority. the handover margin ( M HO ) and the cell individual offset ( CIO ) are considered to rank all potential serving cells and determine the best serving cell. when no value is defined in the Cells table.4 UMTS HSPA Prediction Studies 4.2 Point Analysis . the one with the highest RSCP. The available HSDPA power of the cell in case of an HSDPA bearer user. 40 dBm if the maximum power is set to 43 dBm) Uplink load factor = 50%. Each other cell of the active set is selected among the other potential serving cells as follows: • • • It must use the same carrier as the best serving cell. 4. for all carriers. 285 .4.ini file. Then. 1 Bar Graph and Pilot Sub-Menu Atoll performs a first selection of potential serving cells depending on if you have chosen "Carrier" or "Layer". If the frequency band is fixed ("Best (Specific band)"). The potential serving cells are selected among all transmitters i that contain the receiver in their calculation areas and that have cells using the selected layer.Atoll 3. Option Without pilot: Atoll considers the total noise deducting the pilot signal. ic) P c  i ic  is the pilot power of a transmitter i on carrier ic at the receiver. They must also belong to layers supported by the service and the terminal. 3rd case: Analysis based on the best layer The layer that can be used is fixed.4. is defined between ic and icadj and set to a value different from 0. and these layers must support a speed higher than the user mobility. Calculation option may be selected in Global parameters. In addition. The potential serving cells are selected among all transmitters i that contain the receiver in their calculation areas and that use the selected carrier ic. we have:  BTS    P c  i ic  Q pilot  i ic  = --------------------------------------------DL I 0  ic  With. The pilot signal level received from these cells must exceed the defined minimum RSCP threshold. The interference reduction factor. We can consider the following cases: 1st case: Analysis based on a specific carrier The carrier that can be used is fixed.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 4. Option Total noise: Atoll considers the noise generated by all the transmitters and the thermal noise. 4th case: Analysis based on all layers The potential serving cells are selected among all transmitters i that contain the receiver in their calculation areas. the potential serving cells are selected among all transmitters i that contain the receiver in their calculation areas. Two ways may be used to calculate I0. 1st step: P c  i ic  calculation for each potential serving cell (i. the potential serving cells are selected among all transmitters i that contain the receiver in their calculation areas and that use a carrier of the selected frequency band. RF  ic ic adj  . And DL DL DL DL DL term I 0  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 pilot option. DL DL DL DL DL term I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 for the total noise option. 2nd case: Analysis based on all carriers of all frequency bands/a specific frequency band If you have selected "Best (All bands)".2.3. Ec/I0 (or Q pilot  ic  ) Evaluation Atoll calculates the pilot quality for all potential serving cells (i. Therefore. P pilot  i ic  P c  i ic  = ------------------------LT I L T is the total loss between transmitter i and receiver. I 286 –  1 –     BTS  P c  i ic  for the without . ic). ic is the studied carrier and icadj is another carrier adjacent to ic. potential serving cells must satisfy the following conditions: • • • They must use a frequency band with which the terminal is compatible. Total power transmitted by each cell is either a simulation result (provided in Simulation properties (Cells tab)) or a value user-defined in Cell properties. This parameter is determined from cell edge coverage probability and Ec/I0 standard deviation. P tot  ic  is the total power received at the receiver from the transmitter on the carrier ic. the macro-diversity gain equals 0. P tot  i ic  and P tot  j ic adj  calculations We have: DL I extra  ic  =  DL P tot  j ic  txj j  i P SCH  ic  DL DL DL I intra  ic  = P tot  i ic  –  BTS     P tot  i ic  – ------------------ LT   Ptot  j icadj  DL DL txj j I inter – carrier  ic  = ---------------------------------------RF  ic ic adj  and Tx DL I inter – techno log y  ic  = P Transmitted  ic i   ------------------------------------Tx Tx m L  ICP ni ic i ic total DL For each transmitter of the network. P tot  ic adj  is the total power received at the receiver from the transmitter on the carrier icadj. Total power transmitted by each cell is either a simulation result (provided in Simulation properties (Cells tab)) or a value user-defined in Cell properties.3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 L Tx  L path  L term  L body  L Indoor  M Shadowing – Ec  Io L T = -----------------------------------------------------------------------------------------------------------------------------------I G Tx  G term DL DL DL 2nd step: P tot  j ic  . term 3rd step: N 0 term N0 calculation Tx DL = NF Term  K  T  W  NR inter – techno log y DL 4th step: I 0  ic  and Q pilot  i ic  evaluation using formulas described above DL 5th step: G macro – diversity calculation DL The macro-diversity gain.Atoll 3. When the Ec/I0 standard deviation is set to 0. models the decrease in shadowing margin due to the fact there are several available pilot signals at the mobile. P Tx  ic  DL P tot  ic  = ---------------LT P Tx  ic  is the total power transmitted by the transmitter on the carrier ic. 6th step: Determination of the best serving cell 287 . G macro – diversity . P Tx  ic adj  DL P tot  ic adj  = ---------------------LT P Tx  ic adj  is the total power transmitted by the transmitter on the carrier icadj. DL npaths G macro – diversity = M Shadowing – Ec  Io – M Shadowing – Ec  Io npaths M Shadowing – Ec  Io is the shadowing margin when the mobile receives n pilot signals (not necessarily from transmitters belonging to the mobile active set). DL For each transmitter of the network. In this case. And DL DL DL DL DL term I 0  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 –  1 –     BTS  P c  i ic  for the without pilot option. Same formulas and calculation method are used to update DL I 0  ic BS  value and determine Q pilot  i ic BS  . The cell enters the active set as best serving cell. Other cells (i. Q pilot Resulting Resulting . Pilot is available. the one with the highest RSCP. Resulting If Q pilot req  Q pilot .0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Among all potential serving cells. it informs that the pilot signal of the best serving cell (BS) is deducted from the total noise. I BS  c OC  = Q pilot  c OC  + C IO  c OC  for the other potential serving cells. Then. no cell among the potential serving cells can enter the active set.icBS) in the active set must satisfy the following criteria: Q pilot  i ic BS  – Q pilot  BS   AS_threshold  BS   i ic BS   neighbour list  BS  (optionally) Number of Cells in Active Set This is a user-specified input in the Terminal properties. Atoll takes the cell with the highest best server indicator ( c max  I Resulting cell edge coverage probability. Q pilot BS  ) and calculates the best pilot quality received with a fixed DL = G macro – diversity  Q pilot  c max  I req Resulting  Q pilot . 7th step: Determination of active-set Then. This cell is referred to as the best serving cell candidate ( c BC ). BS. it means pilot quality at the receiver exceeds Q pilot If Q pilot BS   x% of time (x is the fixed cell edge coverage probability). Pilot is unavailable. The notation “Best server” refers to the best serving cell of active set. This is relevant when using the calculation option “Without pilot”.Atoll 3. Downlink Macro-Diversity Gain This parameter is calculated as described above (5th step). Its carrier (icBS) will be used by other transmitters of the active set (when active set size is greater than 1). DL DL DL DL DL term I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 for the total noise option. It corresponds to the active set size. Atoll calculates the best server indicator ( I BS ) for the best serving cell candidate ( c BC ) and the other potential serving cells ( c OC ): I BS  c BS  = Q pilot  c BS  + M HO  c BS  + C IO  c BS  for the best serving cell candidate. We have:  BTS    P c  i ic  Q pilot  i ic  = --------------------------------------------DL I 0  ic  With. I0 (Best Server) I0 (Best server) is the total noise received at the receiver on icBS. Atoll first selects the cells which belong to the highest priority layer and then. 288 . pilot qualities received from all potential serving cells other than BS ( Q pilot  i ic BS  ) are recalculated to determine the entire receiver active set (when active set size is greater than 1). Thermal Noise This parameter is calculated as described above (3rd step).3. icBS).3. DL N tot  ic BS  is the total noise at the receiver on the carrier of the best serving cell. the downlink Eb/Nt target is increased by the value DL user-defined for the DL Eb/Nt target increase field (Global parameters). Compressed mode is operated when amobile supporting compressed mode is connected to a cell located on a site with a compressed-mode-capable equipment.2 Downlink R99 Sub-Menu The Downlink R99 sub-menu contains R99-related results. Atoll calculates the traffic channel quality from each cell (k. No power control is performed as in simulations. 289 . This parameter is available in the R99 Bearer Selection table.2. When compressed mode is activated. Q req .Atoll 3. a mobility type and a reception equipment. Required transmitter power on traffic channels req The calculation of the required transmitter power on traffic channels ( P tch ) may be divided into three steps. we have: DL  BTS  P b – max  k ic BS  DL DL . DL 1st step: Q max  k ic BS  evaluation for each cell DL Let us assume the following notation: Eb/Nt max corresponds to Q max Therefore. DL I intra  ic BS  is the intra-cell interference at the receiver on the carrier of the best serving cell. This parameter is user-defined in the R99 Radio Bearers table.4.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4. Here. Then. P SCH  k ic BS  DL I intra  ic BS  = P DL  k ic  –  BTS  F ortho   P DL  k ic  – ----------------------------- tot BS tot BS L T DL I extra  ic BS  is the extra-cell interference at the receiver on the carrier of the best serving cell. and • Either the received Ec/I0 is lower than the Ec/I0 activation threshold (Global Resulting parameters): Q pilot • CM – activation  Q pilot .icBS) of the receiver’s active set at the receiver. the total downlink traffic channel quality is evaluated and compared with the specified target quality. for each cell (k. G DL Q max  k ic BS  = -----------------------------------------------------p  G Div DL N tot  ic BS  max P tch DL With P b – max  k ic BS  = ---------LT k DL DL DL DL DL term and N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + I inter – techno log y  ic BS  + N 0 Where max P tch is the maximum power allowed on traffic channels. DL I extra  ic BS  =  Ptot  j icBS  DL j j  k DL I inter – carrier  ic BS  is the inter-carrier interference at the receiver on the carrier of the best serving cell. after combination. Atoll determines the downlink traffic channel quality at the receiver for the maximum allowed traffic channel power per transmitter. Eb/Nt Target DL Eb/Nt target ( Q req ) is defined for a given R99 bearer. Or the pilot RSCP is lower than the pilot RSCP activation threshold (Global CM – activation parameters): P c  RSCP pilot . we have: P tch = -----------------------------tch DL Q MAX  ic BS  Max Eb/Nt for Each Cell of Active Set For each cell (k. DL I inter – techno log y  ic BS  is the inter-technology interference at the receiver on the carrier of the best serving cell. P max P tch = -------------------------tch DL Q MAX  ic BS  Compressed mode is operated when a mobile supporting compressed mode is connected to a cell located on a site with a compressed-mode-capable equipment. req 3rd step: P tch calculation DL Q req req . we have: DL DL Q MAX  ic BS  = Q max  k ic BS  For any other handoff status. we have: DL DL Q MAX  ic BS  = f rake efficiency   Qmax  k icBS  DL k Where DL f rake efficiency is the downlink rake efficiency factor defined in Terminal properties. Or the pilot RSCP is lower than the pilot RSCP activation threshold (Global CM – activation parameters): P c  RSCP pilot When compressed mode is activated. P max case. we have: DL  BTS  P b – max  k ic BS  DL DL .3.  DL I inter – techno log y  ic BS  = ni ic i is the i Tx m ICP ic  ic i BS th Tx P Transmitted  ic i  ----------------------------------------Tx Tx m L total  ICP ic  ic i BS interfering carrier of an external transmitter is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the frequency gap between ic i (external network) and ic BS .0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015  Ptot  j icadj  DL DL txj j I inter – carrier  ic BS  = ---------------------------------------RF  ic BS ic adj  icadj is a carrier adjacent to icBS. RF  ic BS ic adj  is the interference reduction factor.icBS). Q req . the downlink Eb/Nt target is increased by the value DL user-defined for the DL Eb/Nt target increase field (Global parameters). and • Either the received Ec/I0 is lower than the Ec/I0 activation threshold (Global Resulting parameters): Q pilot • CM – activation  Q pilot . G DL Q max  k ic BS  = -----------------------------------------------------p  G Div DL N tot  ic BS  290 . In this DL DL Q req  Q req req . On downlink. if there is no handoff. 2nd step: Calculation of the total traffic channel quality DL Q MAX is the traffic channel quality at the receiver on icBS after signal combination of all the transmitters k of the active set. defined between ic and icadj and set to a value different from 0.Atoll 3. 0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 max P tch DL With P b – max  k ic BS  = ---------LT k DL DL DL DL DL term N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + I inter – techno log y  ic BS  + N 0 max req P tch – P tch P SCH  k ic BS  DL I intra  ic BS  = P DL  k ic  –  BTS  F ortho   P DL  k ic  – -----------------------------–  1 –  BTS   max (--------------------------. On downlink. the service on the downlink traffic channel is available if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req when compressed mode is activated). Downlink Soft Handover Gain DL G SHO corresponds to the DL soft handover gain. DL DL DL DL DL DL DL Q eff = min  Q MAX Q req  (or Q eff = min  Q MAX Q req  Q req  when compressed mode is activated).4. DL Q MAX  ic BS  DL G SHO = -----------------------------------------------DL max  Qmax  k ic BS   DL DL max  Qmax  k ic BS   corresponds to the highest Q max  k ic BS  value.0)  tot  tot BS BS L L T DL I extra  ic BS  = Tk  Ptot  j icBS  DL j j  k  Ptot  j icadj  DL DL  j ---------------------------------------I inter – carrier  ic BS  = txj RF  ic BS ic adj  DL I inter – techno log y  ic BS  =  ni Tx P Transmitted  ic i  ----------------------------------------Tx Tx m L total  ICP ic  ic i BS Where req P tch is the required transmitter power on traffic channels. Effective Eb/Nt DL Q eff is the effective traffic channel quality at the receiver on icBS. DL DL DL DL DL Therefore. 4. if there is no handoff. Max Eb/Nt DL Q MAX is the traffic channel quality at the receiver on icBS after signal combination of all the transmitters k of the active set. we have: DL DL Q MAX  ic BS  = f rake efficiency   Qmax  k icBS  DL k Where DL f rake efficiency is the downlink rake efficiency factor defined in Terminal properties.Atoll 3. 291 . we have: DL DL Q MAX  ic BS  = Q max  k ic BS  For any other handoff status.3.2.3 Uplink R99 Sub-Menu The Uplink R99 sub-menu contains R99-related results. For softer and softer-softer handoffs (1/2 and 1/3): UL UL Q MAX  ic BS  = f rake efficiency    Qmax  k icBS   UL k For softer-soft handoffs (2/3). Then. there are two possibilities. Here. If the MRC option is selected (option available in Global parameters). Atoll determines the uplink traffic channel quality at the cell for the maximum terminal power allowed. UL UL If there is no handoff (1/1): Q MAX  ic BS  = Q max  k ic BS  For soft handoff (2/2): UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  k ic BS   UL  G macro – diversity  2 links is the uplink macro-diversity gain. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation.3. No power control is performed as in simulations. tx N0 UL N tot  k ic BS  = -----------------------------------UL 1 – X  k ic BS  tx N 0 is the transmitter thermal noise. For soft-soft handoffs (3/3): UL UL UL Q MAX  ic BS  =  G macro – diversity  3 links  max  Q max  k ic BS   UL  G macro – diversity  3 links is the uplink macro-diversity gain. When the option “Shadowing taken into account” is not selected (Prediction properties). Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. UL UL max  Q max  k ic BS   corresponds to the highest Q max  k ic BS  value. UL 1st step: Q max  k ic BS  evaluation for each cell For each cell (k.icBS) in the receiver’s active set.Atoll 3. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. Atoll calculates uplink traffic channel quality from receiver. G UL Q max  k ic BS  = -------------------------------------------------------p  G Div UL N tot  k ic BS  max UL P term   1 – r c  UL With P b – max  k ic BS  = --------------------------------------LT k UL N tot  k ic BS  is the total noise at the transmitter on the carrier of the best serving cell. the total uplink traffic channel quality is evaluated with respect to the receiver handover status. Max Terminal Power max Max terminal power ( P term ) is an input user-defined for each terminal. we have: 292 .0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 For each cell (k. This value is calculated from the cell UL uplink load factor X  k ic BS  . 2nd step: Calculation of the total traffic channel quality UL Q MAX  ic BS  is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters k of the active set. Atoll calculates the terminal power required to obtain the R99 bearer and compares it to the maximum terminal power allowed. Required Terminal Power req The calculation of the terminal power required to obtain an R99 bearer ( P term – R99 ) may be divided into three steps.icBS) in the receiver’s active set. When the option “Shadowing taken into account” is not selected (Prediction properties). It corresponds to the terminal’s maximum power. we have: UL  term  P b – max  k ic BS  UL UL . From this value. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. a given R99 bearer and a given mobility type. UL Q MAX  ic BS  is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters k of the active set.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1  UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency    UL UL  Q max  k ic BS   Q max k on the same site k on the same site   k ic BS   Else. Eb/Nt Max For each cell (k. The pilot RSCP is lower than the pilot RSCP activation threshold (Global parameters): CM – activation P c  RSCP pilot When compressed mode is activated.0) UL LT 1 – X  k ic BS  k tx N 0 is the transmitter thermal noise. Compressed mode is operated when a mobile supporting compressed mode is connected to a cell located on a site with a compressed-mode-capable equipment.3. the service on the uplink traffic channel is available if P term – R99  P term . Q req . UL UL If there is no handoff (1/1): Q MAX  ic BS  = Q max  k ic BS  For soft handoff (2/2): UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  k ic BS   293 . we have: UL  term  P b – max  k ic BS  UL UL . P max P term – R99 = -------------------------term UL Q MAX  ic BS  UL Q req is the uplink traffic quality target defined by the user for a given reception equipment. This value is calculated from the cell UL uplink load factor X  k ic BS  . In this UL UL Q req  Q req req . UL Q req req . tx max req N0 P term – P term – R99 UL . and • The received Ec/I0 is lower than the Ec/I0 activation threshold (Global parameters): Resulting Q pilot • CM – activation  Q pilot . This parameter is available in the R99 Bearer Selection table.Atoll 3. UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  k ic BS   req 3rd step: P term – R99 calculation req P term – R99 is the required terminal power.+  1 –  term   max (-----------------------------------------N tot  k ic BS  = -----------------------------------. P max case. G UL Q max  k ic BS  = -------------------------------------------------------p  G Div UL N tot  k ic BS  max UL P term   1 – r c  UL With P b – max  k ic BS  = --------------------------------------LT k UL N tot  k ic BS  is the total noise at the transmitter on the carrier of the best serving cell. the uplink Eb/Nt target is increased by the value UL user-defined for the UL Eb/Nt target increase field (Global parameters).icBS) in the receiver’s active set. we have: P term – R99 = -----------------------------term UL Q MAX  ic BS  req max Therefore. 3.Atoll 3. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. For further information on MC-HSDPA user modelling. In each cell. When the option “Shadowing taken into account” is not selected (Prediction properties). UL Q MAX  ic BS  UL G SHO = -----------------------------------------------UL max  Q max  k ic BS   UL UL max  Q max  k ic BS   corresponds to the highest Q max  k ic BS  value. UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  k ic BS   Effective Eb/Nt UL Q eff is the effective traffic channel quality at the transmitter on icBS. For further information on DB-MC-HSDPA user modelling.4. see "DB-MC-HSDPA Users" on page 296. we have:  UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency    UL UL  Q max  k ic BS   Q max k on the same site k on the same site   k ic BS   Else. Atoll determines the serving cells and the best HSDPA bearer obtained in each serving cell. The HSDPA bearer user is processed as if he is the only user in the cell. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. Uplink Soft Handover Gain UL G SHO corresponds to the uplink soft handover gain. the frequency band used by the transmitter. Atoll determines the best HSDPA bearer that the user can obtain.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 UL  G macro – diversity  2 links is the uplink macro-diversity gain.4 HSDPA Sub-Menu The HSDPA sub-menu contains HSDPA-related results for HSDPA and HSPA users when the HS-SCCH quality is sufficient and if the user can obtain an HSDPA bearer. For soft-soft handoffs (3/3): UL UL UL Q MAX  ic BS  =  G macro – diversity  3 links  max  Q max  k ic BS   UL  G macro – diversity  3 links is the uplink macro-diversity gain.2. UL UL max  Q max  k ic BS   corresponds to the highest Q max  k ic BS  value. When modelling MC-HSDPA users (including DC-HSDPA users) and DB-MC-HSDPA users.e. For further information on the fast link adaptation modelling. see "MCHSDPA Users" on page 296. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. UL UL UL UL UL UL UL Q eff = min  Q MAX Q req  (or Q eff = min  Q MAX Q req  Q req  when compressed mode is activated). When the option “Shadowing taken into account” is not selected (Prediction properties). Atoll details the results for each cell to which the user is connected. 4. the selected 294 . i. there are two possibilities. he uses the entire HSDPA power available in the cell. the user is processed as if he is the only user in the cell. For softer and softer-softer handoffs (1/2 and 1/3): UL UL Q MAX  ic BS  = f rake efficiency    Qmax  k icBS   UL k For softer-soft handoffs (2/3). If the MRC option is selected (option available in Global parameters). General Results Atoll displays the name of the cell to which the user is connected. see "Fast Link Adaptation Modelling" on page 242. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation.  Nt  HS – SCCH DL N tot  ic  And  BTS  P c  ic  Eci  ---- = ---------------------------------------------------------------------------------------------------------------------------. For further details on the HS-PDSCH power calculation. the HS-SCCH Ec/Nt is calculated from the fixed HS-SCCH power. defined between ic and icadj and set to a value different from 0. RF  ic ic adj  is the interference reduction factor. HS-PDSCH Power Atoll calculates the available HS-PDSCH power.–  BTS   P tot  ic  – ------------------  LT  LT  txi txi txi     DL I extra  ic  =  DL P tot  ic  txj j  i  Ptot  icadj  DL DL  j I inter – carrier  ic  = txj -----------------------------------RF  ic ic adj  icadj is a carrier adjacent to ic. and the maximum available HSDPA power of the cell. When the HS-SCCH power allocation strategy is static. DL I inter – techno log y  ic  =  n ic i is the i th i Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic i interfering carrier of an external transmitter 295 . When the HS-SCCH power allocation strategy is dynamic. HS-SCCH Ec/Nt Threshold Atoll displays the HS-SCCH Ec/Nt threshold set for the selected mobility type. DL I inter – techno log y  ic  is the inter-technology interference at the receiver on ic. see either "HS-PDSCH Quality Calculation" on page 244 if the selected option is "CQI based on CPICH quality" or "HS-PDSCH Quality Calculation" on page 249 if the selected option is "CQI based on HS-PDSCH quality". For further details on the HS-PDSCH quality calculation. We have:  BTS  P c  ic  Eci  ---- ic  = ------------------------------. HS-SCCH Ec/Nt Atoll displays the obtained HS-SCCH quality.3. The way of calculating it depends on the selected option in the transmitters global parameters (HSDPA part): CQI based on CPICH quality or CQI based on HS-PDSCH quality.for the total noise option. this parameter corresponds to the HS-SCCH Ec/Nt threshold defined for the selected mobility type.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 carrier. HS-PDSCH Ec/Nt Atoll calculates the best HS-PDSCH quality (HS-PDSCH Ec/Nt). see either "HS-PDSCH Quality Calculation" on page 244 if the selected option is "CQI based on CPICH quality" or "HS-PDSCH Quality Calculation" on page 249 if the selected option is "CQI based on HS-PDSCH quality".  Nt  ic  HS – SCCH DL term N tot  ic  –  1 – F ortho    1 – F MUD    BTS  P c  ic  i With DL DL DL DL DL term N tot  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0  DL  DL DL P SCH  ic  P SCH  ic  DL term I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – ------------------.for the without useful signal option.Atoll 3. see "HSDPA Bearer Selection" on page 246. Peak RLC Throughput DL Once the bearer selected. For further details on the HSDPA bearer selection. C HSDPABearer . The bearer consumption expressed in %.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Tx m ICP ic  ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the i frequency gap between ic i (external network) and ic . Therefore. Knowing the HS-PDSCH Ec/Nt. The Effective RLC throughput is calculated as follows: DL TP P –RLC DL TP E – RLC = ----------------TTI Where TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used. HSDPA Bearer Parameters Knowing the HS-PDSCH CQI. see either "HS-PDSCH CQI Determination" on page 246 if the selected option is "CQI based on CPICH quality" or "HS-PDSCH CQI Determination" on page 252 if the selected option is "CQI based on HS-PDSCH quality". F MUD and N 0 are defined in "Inputs" on page 215. is calculated as follows: TPD Min – DL C HSDPABearer = --------------------------------------------------DL TP P –R LC  I HSDPABearer  296 . Atoll displays the parameters of the selected HSDPA bearer: • • • The transport block size. Atoll calculates the best HSDPA bearer that can be used and selects a bearer compatible with cell and terminal user equipment HSDPA capabilities. CQI It corresponds to the HS-PDSCH CQI. The way of calculating it depends on the selected option in the transmitters global parameters (HSDPA part): CQI based on CPICH quality or CQI based on HS-PDSCH quality. Atoll determines the peak RLC throughput that can be provided to the user TP P –R LC .Atoll 3. Effective RLC Throughput DL Atoll displays the Effective RLC throughput ( TP E – RLC ) provided to the user. it finds the corresponding BLER. The minimum throughput demand required by the service is allocated to these users.3. it is defined in the terminal user equipment category properties. they partly consume the HSDPA bearer. Bearer Consumption Atoll provides this result for HSPA CBR service users only. The number of HS-PDSCH channels used. BLER Atoll reads the BLER in the quality graph BLER = f(HS-PDSCH Ec/Nt) that is defined for the selected bearer and mobility type. F ortho . For further details on the HS-PDSCH quality calculation. The modulation scheme used. P HS – SCCH  ic  P c  ic  = ------------------------------i LT i And L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io L T = -----------------------------------------------------------------------------------------------------------------------------------G Tx  G term term term  BTS . e. The secondary cells are taken in the same band as the best carrier (i. • The Application Throughput  DL  TP P –RLC  c    1 – BLER HSDPA   DL  Serving cells . the user is processed as if he is the only user in the cell. Results for MC-HSDPA and DB-MC-HSDPA Users When the user is simultaneously connected to several HSDPA cells. the secondary cells. Atoll details the results for each cell. Atoll selects the carriers in a transmitter using the second frequency band. Atoll determines the best HSDPA bearer obtained. according to the CQI value. Within one frequency band. In addition. This graph describes the variation of BLER as a function of the measured quality (HS-PDSCH Ec/Nt). For information on how the best serving cell is selected... it is defined in the terminal user equipment category properties.e. the best cell and the secondary cells). DB-MC-HSDPA Users When multi-cell HSDPA and dual-band HSDPA modes are active. In each cell. Atoll determines the best serving cell using the best serving cell selection algorithm. The user is connected to a cell if he obtains an HSDPA bearer. When two adjacent carriers are available. Atoll selects the other serving cells. Atoll determines the best HSDPA bearer obtained. When two adjacent carriers are available. Atoll takes the one with the highest CQI value. The maximum number of cells to which the user can simultaneously connect depends on the DL multi-cell mode set for the HSDPA UE category of the terminal. The maximum number of cells to which the user can simultaneously connect depends on the DL multi-cell mode set for the HSDPA UE category of the terminal. see "Best Serving Cell and Active Set Determination" on page 284. TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used.. i. The user is connected to a cell if he obtains an HSDPA bearer. as long as carriers are available.e. then the users can only connect to the HSDPA cells of one transmitter. In each serving cell (i. If the two co-site transmitters work on the same frequency band. if additional carriers are required and if there are no more carriers available in this transmitter.3. DB-MC-HSDPA users can simultaneously connect to HSDPA cells of two co-site transmitters using different frequency bands. Atoll determines the best serving cell using the best serving cell selection algorithm. In each serving cell (i. If the best carrier belongs to a transmitter that supports the multi-cell HSDPA mode and if the transmitter has several HSDPA carriers. f TP – Scaling – TP Offset TP A = c---------------------------------------------------------------------------------------------------------TTI Where: BLER HSDPA is read in the quality graph defined for the triplet “reception equipment-selected bearer-mobility” (HSDPA Quality Graphs tab in the Reception equipment properties).Atoll 3. Atoll selects secondary cells as long as HSDPA carriers are available in the transmitter and the maximum number of cells to which the user can simultaneously connect is not exceeded.e. Let’s consider the following configuration: • • • • A site with transmitters working on two different frequency bands. Then. the best cell and the secondary cells). The site equipment supports the dual-band HSDPA mode. 297 . Each transmitter has several HSDPA carriers. see "Best Serving Cell and Active Set Determination" on page 284. Atoll finds the corresponding BLER. the user is processed as if he is the only user in the cell. they belong to the same transmitter). In each cell. The multi-cell HSDPA mode is active for each transmitter.. The secondary cells belong to the same transmitter and are chosen among the adjacent carriers according to the CQI. MC-HSDPA users can simultaneously connect to several HSDPA cells of the transmitter for data transfer. it is defined in the terminal user equipment category properties. Knowing the HS-PDSCH Ec/Nt. For information on how the best serving cell is selected.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 MC-HSDPA Users When multi-cell HSDPA is active. the secondary cells are first selected according to an adjacency criterion and then. Atoll takes the one with the highest CQI value. it displays the following results under Total: • DL The Peak RLC Throughput TP P – RLC =  DL TP P –RLC  c  c  Serving cell • The Effective RLC Throughput DL TP P –RLC DL TP E – RLC = ----------------TTI Where TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used. Atoll displays the parameters of the selected HSUPA bearer: • • • • 298 The radio bearer index. This value is defined for an HSUPA Nt E – DPDCH bearer ( Index HSUPABearer ) and a number of retransmissions ( N Rtx ) in the HSUPA Bearer Selection table. see "HSUPA Bearer Allocation Process" on page 261. These two parameters model the header information and other supplementary data that does not appear at the application level.2. HSUPA Bearer Parameters Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. P term – HSUPA . Required Terminal Power Ec req req From  ------ . The HSUPA bearer user is processed as if he is the only user in the cell. I tot UL tx . I tot UL extra . he uses the entire remaining load of the cell. . Atoll determines the best HSUPA bearer that the user can obtain.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 f TP – Scaling and TP Offset respectively represent the scaling factor between the application throughput and the RLC (Radio Link Control) throughput and the throughput offset. For further details on the calculation. see "Max E Nt E – DPDCH DPDCH Ec/Nt" on page 298.4.e.Atoll 3.3. The TTI duration. They are defined in the service properties.) where: throughput ( -----------------------------------------------------------------N Rtx  Index HSUPABearer  • Ec  req Ec max  ----  ------  Nt E – DPDCH  Nt E – DPDCH • And P term – HSUPA  P term req max With max P term : the maximum terminal power allowed. Atoll calculates the terminal power required to obtain the HSUPA bearer. For further information on the HSUPA bearer selection. 4. I inter – carrier and N 0 are defined in "Inputs" on page 215. This is the HSUPA bearer with the highest potential UL TP P – RLC  Index HSUPABearer  . The modulation scheme used. F MUD . Max E-DPDCH Ec/Nt Ec max Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ). i.  Nt E – DPDCH req Ec req UL P term – HSUPA =  ------  L T  N tot Nt E – DPDCH With UL UL intra tx N tot  ic  =  1 – F MUD   term   I tot UL extra  ic  + I tot UL tx  ic  + I inter – carrier  ic  + N 0 L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term tx UL intra  term . Required E-DPDCH Ec/Nt req Ec It corresponds to the E-DPDCH Ec/Nt required to obtain the HSUPA bearer (  ------ ).5 HSUPA Sub-Menu The HSUPA sub-menu contains HSUPA-related results for HSPA users if the user can obtain an HSUPA bearer. The number of E-DPDCH codes used. UL TP P – RLC   1 – BLER HSUPA  UL TP Min – E – RLC  M b  = -----------------------------------------------------------------N Rtx Where: BLER HSUPA is the residual BLER after N Rtx retransmissions. Therefore. Potential serving cells are filtered depending on the prediction definition (selected layers or carriers. The bearer consumption expressed in %. Atoll determines the corresponding RLC peak throughput. Peak RLC Throughput/No. is calculated as follows: TPD min – UL C HSUPABearer = ----------------------------------------------------UL TP P – RLC  I HSUPABearer  4. for all carriers.4. they parly consume the HSUPA bearer. Min Effective RLC Throughput UL From the RLC peak throughput. Atoll considers the N Rtx  Index HSUPABearer  ratio to select the HSUPA bearer when several HSUPA bearers meet the selection criteria. see "Bar Graph and Pilot Sub-Menu" on page 285. Bearer Consumption Atoll provides this result for CBR service users only.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Peak RLC Throughput UL After selecting the HSUPA bearer. For further information on formulas.3. layers supported by the service and the terminal.3.). Application Throughput UL Atoll displays the provided application throughput ( TP A ). Atoll calculates the minimum effective RLC throughput. For information on the best serving cell selection and pilot quality calculation. This one is calculated as follows: UL TP P – RLC   1 – BLER HSUPA   f TP – Scaling – TP Offset UL TP A  M b  = --------------------------------------------------------------------------------------------------------------------------N Rtx BLER Knowing the E-DPDCH Ec/Nt. C HSUPABearer . The application throughput represents the net throughput after deduction of coding (redundancy. 4. etc. 299 .Atoll 3. or for all layers. Let us assume each pixel on the map corresponds to a probe receiver with a terminal. of RTX UL TPP – RLC  Index HSUPABearer  Atoll displays the peak RLC throughput to number of retransmissions ratio ( ------------------------------------------------------------------. overhead. or userdefined cell inputs. Coverage predictions are based on parameters that can be either simulation results. addressing. This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt). You can make the coverage prediction for a specific carrier.). This receiver does not create any interference. TP Min – E – RLC . for a specific layer.1 Pilot Quality Analysis Atoll determines the best serving cell (BS) for each pixel and calculates the pilot quality received with a fixed cell edge Resulting coverage probability. mobility type) and the pilot signal level which must exceed the defined minimum RSCP threshold. TP P – RLC . see "Definitions" on page 214. a mobility type and a service. Q pilot  BS  .3 Coverage Studies Atoll calculates UMTS-specific coverage studies on each pixel where the pilot signal level exceeds the minimum RSCP threshold.4. The minimum bit rate required by the service is allocated to these users. Atoll finds the corresponding BLER. It is read in the quality graph defined for the quartet “reception equipment-selected bearer-number of retransmissions-mobility” (HSUPA Quality Graphs tab in the Reception equipment properties). There is a layer per userdefined cell edge coverage probability. 4. There is a layer per userdefined pilot signal level defined in the Display tab (Prediction properties). There is a layer per userdefined quality margin defined in the Display tab (Prediction properties). For each layer.4. For each layer.3. For each layer.4. For each layer.4.3.e. Single colour Resulting Atoll displays a coverage if Q pilot req req  BS   Q pilot . or a user-defined cell input.2 Downlink Service Area Analysis As in point predictions. Coverage consists of a single layer with a unique colour. Colour per quality level (Ec/I0) Coverage consists of several independent layers that can be displayed and hidden on the map. 40 dBm if the maximum power is set to 43 dBm). There is a layer per user-defined probability level defined in the Display tab (Prediction Resulting properties). There is a layer per transmitter.1. receiver is not completely defined and no mobility is assigned. Coverage consists of several independent layers that can be displayed and hidden on the map.Atoll 3. area is covered if Resulting Q pilot  BS    Q pilot  threshold . area is covered if Q pilot req  BS   Q pilot in the required number of simulations. Colour per pilot signal level (Ec) Coverage consists of several independent layers that can be displayed and hidden on the map.1 ©Forsk 2015 Prediction Study Inputs The Pilot Quality Analysis depends on the downlink total transmitted power of cells. For each layer.e. p. This parameter can be either a simulation output. In the last case.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks 4.3. No power control is performed as in simulations. Colour per mobility In this case. area is covered if Resulting Q pilot req  BS p   Q pilot . For each layer. Atoll considers 50% of the maximum power as default value (i. Atoll calculates traffic channel quality at the receiver for each cell (k. Atoll determines downlink traffic channel quality at the receiver for a 300 . if you select a group of simulations and the “All” option in the Condition tab of prediction properties). Q pilot is a target value defined in the Mobility table by the user. area is covered if Q pilot req  BS   Q pilot . There is a layer per userdefined quality threshold defined in the Display tab (Prediction properties). Coverage consists of several independent layers that can be displayed and hidden on the map.2 Study Display Options Atoll displays the best pilot quality received with a fixed cell edge coverage probability. Colour per transmitter Resulting Atoll displays a coverage if Q pilot req  BS   Q pilot .icBS) in the receiver’s active set. There is a layer per userResulting defined mobility defined in the Mobility Types sub-folder. Colour per quality margin (Ec/I0 margin) Coverage consists of several independent layers that can be displayed and hidden on the map. area is covered if Resulting Q pilot req  BS  – Q pilot   Q pilot  m arg in . Here. defined in the Display tab (Prediction properties). Layer colour is the colour assigned to the transmitter of the best serving cell (BS). Coverage consists of several independent layers that can be displayed and hidden on the map. area is covered if Resulting Q pilot  BS    Q pilot  threshold . 4.3.1. when no value is defined in the Cells table for the total transmitted power. Colour per cell edge coverage probability Coverage consists of several independent layers that can be displayed and hidden on the map. Colour per probability This display option is available only if analysis is based on all simulations in a group (i. 2 Study Display Options Single colour DL DL DL DL DL Atoll displays a coverage with a unique colour if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). 40 dBm if the maximum power is set to 43 dBm).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 DL maximum allowed traffic channel power for transmitters. This parameter can be either a simulation output.1 Prediction Study Inputs The Downlink Service Area Analysis depends on the downlink total transmitted power of cells. see "Downlink R99 Sub-Menu" on page 288. area is covered if Q MAX  ic BS   Q req in the required number of simulations. For each layer. 301 . see "Best Serving Cell and Active Set Determination" on page 284. For information on best serving cell selection and active set determination. area is covered if DL DL DL DL DL Q MAX  ic BS p   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). Atoll considers 50% of the maximum power as default value (i. p.3. For each DL DL DL DL DL layer.2. 4. area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). receiver is not completely defined and no service is assigned. 4. Colour per service In this case. area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). Colour per mobility In this case. For each layer. Then. Layer colour is the colour assigned to best serving transmitter. Coverage consists of several independent layers that can be displayed and hidden on the map.Atoll 3. if you select a group of simulations and the “All” option in the Condition tab of prediction properties). There is a layer per user-defined service defined in Services sub-folder. For DL DL DL DL DL each layer.3.3. Atoll displays the total traffic channel quality in the downlink. this parameter is user-defined in the Global parameters. In the last case.e.2. area is covered if DL Q MAX  ic BS   Threshold . There is a layer per user-defined mobility defined in Mobility sub-folder.4. a given R99 bearer and a given mobility type. DL Q req is the DL Eb/Nt target increase. receiver is not completely defined and no mobility is assigned. when no value is defined in the Cells table for the total transmitted power. Colour per probability This display option is available only if analysis is based on all simulations in a group (i. For further details on calculations. Colour per cell edge coverage probability Coverage consists of several independent layers that can be displayed and hidden on the map. defined in the Display tab (Prediction properties). There is a layer per userdefined quality threshold defined in the Display tab (Prediction properties).e. Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined cell edge coverage probability. There is a layer per user-defined probability level defined in the Display tab (Prediction DL DL properties). or a user-defined cell input.4. Coverage consists of several independent layers that can be displayed and hidden on the map. For each layer. This parameter is available in the R99 Bearer Selection table. Colour per maximum quality level (max Eb/Nt) Coverage consists of several independent layers that can be displayed and hidden on the map. the total downlink traffic channel quality ( Q MAX  ic BS  ) is evaluated after recombination. There is a layer per transmitter. Coverage consists of several independent layers that can be displayed and hidden on the map. Colour per transmitter DL DL DL DL DL Atoll displays a coverage if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). DL Q req is the downlink traffic quality target defined by the user for a given reception equipment. Then. area is covered if req P tch  ic BS   Threshold . 4. It corresponds to the maximum allowable traffic channel power for a transmitter. a given reception equipment and a mobility type.3.3. DL Q MAX  ic BS  Coverage consists of several independent layers that can be displayed and hidden on the map.4. 302 . area is covered if DL DL DL DL DL Q MAX  ic BS  – Q req  M arg in (or Q MAX  ic BS  – Q req  Q req  M arg in when compressed mode is activated). For information on best serving cell selection and active set determination. is user-defined for a given R99 bearer. There is a layer per userdefined required power threshold defined in the Display tab (Prediction properties). or a user-defined cell input. P max When compressed mode is activated. Atoll displays the total traffic channel quality in the uplink. For each layer.4. available in the R99 Bearer Selection table.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Colour per effective quality level (Effective Eb/Nt) Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per userdefined quality margin defined in the Display tab (Prediction properties). No power control simulation is performed. There is a layer per userdefined power margin defined in the Display tab (Prediction properties). In the last case. see "Uplink R99 Sub-Menu" on page 291. Colour per required power margin Coverage consists of several independent layers that can be displayed and hidden on the map. max P tch is a user-defined input for each bearer related to a service. There is a layer per userdefined quality threshold defined in the Display tab (Prediction properties). For further details on calculations.icBS) in receiver active set. see "Best Serving Cell and Active Set Determination" on page 284. area is covered if DL Q eff  ic BS   Threshold . the total uplink traffic channel quality ( Q MAX  ic BS  ) is evaluated with respect to receiver handover status. DL DL DL Q eff  ic BS  = min  Q MAX  ic BS  Q req  (or DL DL DL DL Q eff  ic BS  = min  Q MAX  ic BS  Q req  Q req  when compressed mode is activated). Atoll uses 50% as default value.Atoll 3.3 Uplink Service Area Analysis As in point prediction. Colour per quality margin (Eb/Nt margin) Coverage consists of several independent layers that can be displayed and hidden on the map. This parameter can be either a simulation output. For each layer. Colour per required power req Atoll calculates the downlink required power. For each layer. P max P tch  ic BS  = -------------------------tch DL Q MAX  ic BS  Where DL Q req is the Eb/Nt target on downlink.3. P tch  ic BS  . DL DL Q req  Q req req . Atoll determines uplink traffic channel quality at the transmitter for the UL maximum terminal power allowed. For each layer. area is covered if req max P tch  ic BS  – P tch  M arg in .3. Atoll calculates uplink traffic channel quality from the receiver for each cell (k. when no value is defined in the Cells table for the uplink load factor. This parameter.1 Prediction Study Inputs The Uplink Service Area Analysis depends on the UL load factor of cells. as follows: DL Q req req . we have: P tch  ic BS  = -----------------------------tch . 4. Coverage consists of several independent layers that can be displayed and hidden on the map. For each layer. For UL UL UL UL UL each layer. a R99 bearer and a mobility type. area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). Colour per quality margin (Eb/Nt margin) Coverage consists of several independent layers that can be displayed and hidden on the map. UL Q req is defined for a reception equipment.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4. There is a layer per user-defined probability level defined in the Display tab (Prediction UL UL UL UL UL properties). 303 .3. area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). Colour per effective quality level (Effective Eb/Nt) Coverage consists of several independent layers that can be displayed and hidden on the map. Coverage colour is unique. if you select a group of simulations and the “All” option in the Condition tab of prediction properties). There is a layer per user-defined mobility defined in Mobility sub-folder.4. Coverage consists of several independent layers that can be displayed and hidden on the map.3. There is a layer per userdefined quality threshold defined in the Display tab (Prediction properties). Colour per maximum quality level (Max Eb/Nt) Coverage consists of several independent layers that can be displayed and hidden on the map. area is covered if UL UL UL UL UL Q MAX  ic BS  – Q req  M arg in (or Q MAX  ic BS  – Q req  Q req  M arg in if compressed mode is activated). For each layer. this parameter is user-defined in the Global parameters. Colour per mobility In this case. For each layer. receiver is not completely defined and no service is assigned. Colour per probability This display option is available only if analysis is based on all simulations in a group (i. UL UL UL Q eff  ic BS  = min  Q MAX  ic BS  Q req  (or UL UL UL UL Q eff  ic BS  = min  Q MAX  ic BS  Q req  Q req  when compressed mode is activated). Colour per transmitter UL UL UL UL UL Atoll displays a coverage if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated) in the required number of simulations. Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per transmitter. area is covered if UL Q effective  ic BS   Threshold . receiver is not completely defined and no mobility is assigned.e.Atoll 3. This parameter is available in the R99 Bearer Selection table. For each layer. Colour per service In this case. There is a layer per userdefined quality margin defined in the Display tab (Prediction properties).3. For each UL UL UL UL UL layer. area is covered if UL Q MAX  ic BS   Threshold . Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per user-defined service defined in Services sub-folder. UL Q req is the UL Eb/Nt target increase.2 Study Display Options Single colour UL UL UL UL UL Atoll displays a coverage if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). Layer colour is the colour assigned to best server transmitter. There is a layer per userdefined quality threshold defined in the Display tab (Prediction properties). Atoll 3. area is covered if DL minN tot  ic   Threshold . or a user-defined cell input. or for all layers. Atoll determines DL total noise or DL noise rise on this carrier. There is a layer per userdefined noise level defined in the Display tab (Prediction properties). is calculated from the downlink total noise. ic 304 . Colour per required power margin Coverage consists of several independent layers that can be displayed and hidden on the map. area is covered if req max P term – R99  ic BS  – P term  M arg in .2 Display Options The following display options are available for the prediction: Colour per minimum noise level Coverage consists of several independent layers that can be displayed and hidden on the map. as follows: term  N0  - NR DL  ic  = – 10 log  ----------- N DL tot  4. This parameter can be either a simulation output. For each layer.e. Atoll considers 50% of the maximum power as default value (i. For each layer.+ N 0 Tx Tx m L total  ICP ic  ic i DL Downlink noise rise. For each layer. NR DL  ic  .4. 40 dBm if the maximum power is set to 43 dBm). area is covered if G SHO  Threshold . Atoll determines the DL total noise level on each carrier supported by the user service. a mobility type and a service.4. area is covered if DL maxN tot  ic   Threshold . For each layer.3. You can make the coverage prediction for a specific carrier. For each layer. N tot . We assume that each pixel on the map corresponds to a probe receiver with a terminal. There is a layer per userdefined noise level defined in the Display tab (Prediction properties).4 Downlink Total Noise Analysis Atoll determines the downlink total noise level generated by cells. When you select "Best (All/Specific band)" as the carrier or layers associated with several carriers.1 Study Inputs The Downlink Total Noise Analysis depends on the downlink total transmitted power of cells. When only one carrier is analysed.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Colour per required power Coverage consists of several independent layers that can be displayed and hidden on the map. area is covered if req P term – R99  ic BS   Threshold . ic Colour per maximum noise level Coverage consists of several independent layers that can be displayed and hidden on the map.4. There is a layer per userdefined power threshold defined in the Display tab (Prediction properties). 4.3.4. for all carriers. Colour per soft handover gain Coverage consists of several independent layers that can be displayed and hidden on the map. for a specific layer. There is a layer per userdefined power margin defined in the Display tab (Prediction properties). when no value is defined in the Cells table for the total transmitted power. 4.3.3. There is a layer per soft UL handover gain value defined in the Display tab (Prediction properties).  Ptot  icadj  DL DL N tot  ic  =  txj j DL txj j P tot  ic  + -----------------------------------+ RF  ic ic adj   ni Tx P Transmitted  ic i  term -------------------------------------. In the last case.4. 4. Coverage consists of several independent layers that can be displayed ic and hidden on the map.Atoll 3.e. You can make the coverage prediction for a specific carrier. you have to select several carriers ("Best HSPA (All/Specific band)" as the carrier) or layers associated with several carriers. maximum). 305 . The number of HSDPA bearer users within the cell if the study is calculated for several users. or user-defined cell inputs. Here we assume that each pixel on the map corresponds to one or several users with HSDPA capable terminal. or for all layers. Coverage consists of several independent layers that can be ic displayed and hidden on the map. Atoll determines DL total noise or DL noise rise on this carrier. This type of analysis is not relevant when modelling MC-HSDPA and DB-MC-HSDPA users. selecting one specific carrier or one layer associated with one unique carrier is not suitable. mobility and HSDPA service.4. When studying a certain HSDPA radio bearer. this parameter must be defined by the user. area is covered if DL averageNtot  ic   Threshold . MC-HSDPA and DB-MC-HSDPA users. either you can take all the possible HSDPA radio bearers into consideration.4. There is a layer per user-defined noise rise threshold defined in the Display tab. In this case. Coverage consists of several independent layers that can be displayed ic and hidden on the map.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Colour per average noise level Coverage consists of several independent layers that can be displayed and hidden on the map. you can set display parameters: • • • To analyse the uplink and downlink A-DPCH qualities on the map. Then. 40 dBm if the maximum power is set to 43 dBm) Number of HSDPA bearer users = 1 On the other hand. When considering all the HSDPA radio bearers. Colour per maximum noise rise Atoll displays bins where maxNR DL  ic   Threshold . For information on the best serving cell and secondary cells selection. the displayed coverage is the same for any selected display option (average. no default value is used for the available HSDPA power. minimum. for a specific layer.5 HSDPA Prediction Study When calculating the HSDPA coverage prediction. There is a layer per user-defined noise rise threshold defined in the Display tab. the probability of having a certain RLC peak throughput). The downlink total transmitted power of the cell.1 Prediction Study Inputs Parameters used as input for the HSDPA prediction study are: • • • The available HSDPA power of the cell. To display the global throughput. see "MC-HSDPA Users" on page 296 and "DB-MCHSDPA Users" on page 296. ic Colour per minimum noise rise Atoll displays bins where minNR DL  ic   Threshold . Colour per average noise rise Atoll displays bins where averageNR DL  ic   Threshold . In the last case. or you can study a certain HSDPA radio bearer. These parameters can be either simulation outputs.3.5.The user does not create any interference. you can display areas where a certain RLC peak throughput is available with different cell edge coverage probabilities (i. There is a layer per user-defined noise rise threshold defined in the Display tab. Note that the HSDPA service area is limited by the pilot quality. the A-DPCH quality and the HS-SCCH quality. when no value is defined in the Cells table for the total transmitted power and the number of HSDPA bearer users. Atoll uses the following default values: • • Total transmitted power = 50% of the maximum power (i. 4. for all carriers. available display options depend on what you have selected.3.e. For DC-HSDPA.3. When only one carrier is analysed. There is a layer per userdefined noise level defined in the Display tab (Prediction properties). For each layer. To analyse the HS-SCCH quality/power. To model fast link adaptation for a single HSDPA bearer user or for a defined number of HSDPA bearer users. There is a layer per threshold. see "Fast Link Adaptation Modelling" on page 242. To model fast link adaptation for a defined number of HSDPA bearer users. Atoll displays on each pixel the HSSCCH quality per HS-SCCH channel.e. For further information on the fast link adaptation modelling.3. For each layer. There is a layer per userdefined quality threshold defined in the Display tab (Prediction properties). see "Downlink R99 Sub-Menu" on page 288. Atoll considers one user on each pixel and determines the best HSDPA bearer that the user can obtain. Coverage consists of several independent layers that can be displayed and hidden on the map. Atoll determines uplink traffic channel quality at the receiver for a maximum terminal power allowed. Ec There is a layer per threshold. For each layer. several display options are available in the study properties dialogue. Atoll determines downlink traffic channel quality at the receiver for a maximum traffic channel power allowed for the best serving cell. In this case. area is covered if DL Q MAX  BS   Threshold . No power control is performed as in simulations. To analyse the HS-SCCH quality/power.5. Atoll displays on each pixel the HS-SCCH power per HS-SCCH channel. On each pixel. he uses the entire HSDPA power available in the cell. For each layer. There is a layer per userdefined quality threshold defined in the Display tab (Prediction properties). Analysis of UL And DL A-DPCH Qualities • Colour per Max A-DPCH Eb/Nt DL DL Atoll displays the A-DPCH quality at the receiver ( Q MAX  BS  ) for the best serving cell (BS). For each layer.  Nt  HS – SCCH  Threshold Fast Link Adaptation Modelling For A Single User When you calculate the study with the following display options. When studying a certain HSDPA radio bearer. Atoll determines the best HSDPA bearers that the user can obtain in each serving cell.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks 4. area is covered if P HS – SCCH  BS   Threshold .2 ©Forsk 2015 Study Display Options When considering all the HSDPA radio bearers. No power control is performed as in simulations. the user is processed as if he is the only user in the cell i. For MC-HSDPA and DB-MC-HSDPA users. Here. • Colour per HS-SCCH Ec/Nt This display option is relevant in case of static HS-SCCH power allocation only. To model fast link adaptation for a single HSDPA bearer user. area is covered if  -----. Here. • Colour per Max A-DPCH Eb/Nt UL UL Atoll displays the A-DPCH quality at the best serving cell ( Q MAX  BS  ). In this case. area is covered if UL Q MAX  BS   Threshold . Coverage consists of several independent layers that can be displayed and hidden on the map. It allows you to display where a certain RLC peak throughput is available with different cell edge coverage probabilities.3.  Nt  HS – PDSCH  Threshold • 306 Colour per CQI . it corresponds to the HS-PDSCH Ec/Nt of the best serving cell. BS  . Coverage consists of several independent layers that can be displayed and hidden on the map.Atoll 3. see "Uplink R99 Sub-Menu" on page 291. Coverage consists of several independent layers that can be displayed and hidden on the Ec map. For information on calculation methods. • Colour per HS-PDSCH Ec/Nt Atoll displays on each pixel the HS-PDSCH quality. They can be regrouped in four categories according to the objective of the study: • • • • To analyse the uplink and downlink A-DPCH qualities on the map. area is covered if  -----. For MC-HSDPA and DB-MC-HSDPA users. Coverage consists of several independent layers that can be displayed and hidden on the map. Analysis of The HS-SCCH Quality/Power • Colour per HS-SCCH Power This display option is relevant in case of dynamic HS-SCCH power allocation only. only one display option is available. For information on calculation methods. For each layer.4. There is a layer per threshold. BS  . area is covered if the Effective MAC throughput exceeds the userdefined thresholds. For each layer. • Colour per Peak MAC Throughput DL Atoll displays the Peak MAC throughput ( TPP –M AC ) provided on each pixel. • Colour per Peak RLC Throughput After selecting the bearer. S block  c  is the transport block size (in kbits) of the HSDPA bearer selected in the cell. For each layer. we have: TP P – RLC = TP P –RLC  c  For MC-HSDPA and DB-MC-HSDPA users. S block  c  is the transport block size (in kbits) of the selected HSDPA bearer in the cell. There is a layer per CQI threshold (  CQI  threshold ). for the user. Coverage consists of several independent layers that can be displayed and hidden on the map.e. i.1. There is a layer per possible DL Effective MAC throughput ( TP E – M AC ). in the downlink.Atoll 3.2. area is covered if the Peak MAC throughput exceeds the user-defined thresholds. The Peak MAC throughput is calculated as follows: DL TP P – M AC =  c  Serving cells S block  c  --------------------T TTI Where. This value is specified by the 3GPP. DL DL For an HSDPA user. There is a layer per possible DL Peak MAC throughput ( TPP –M AC ). Coverage consists of several independent layers that can be displayed and hidden on the map. it is defined for each HSDPA bearer in the HSDPA Radio Bearers table. it determines the peak RLC throughput provided by the serving DL cell. area is covered if CQI   CQI  threshold . • Colour per Effective RLC Throughput 307 . –3 T TTI is the TTI duration. There is a layer per possible DL RLC peak throughput ( TP P – RLC ). 2 10 s (2000 TTI in one second). Coverage consists of several independent layers that can be displayed and hidden on the map. it corresponds to the CQI of the best serving cell. i. • Colour per Effective MAC Throughput DL Atoll displays the Effective MAC throughput ( TP E – M AC ) provided on each pixel.2) when considering the CQI based on HS-PDSCH quality option. The Effective MAC throughput is calculated as follows: DL TP E – M AC =  c  Serving cells S block  c  -------------------------T TTI   TTI Where. For MC-HSDPA and DB-MC-HSDPA users. For each layer. or the HS-PDSCH CQI (see the calculation detail in the section 10. area is covered if the peak RLC throughput can be provided. This value is specified by the 3GPP. c.3. it is defined in the terminal user equipment category properties. Then. Atoll reads the corresponding RLC peak throughput ( TP DL  I ). it is defined for each HSDPA bearer in the HSDPA Radio Bearers table.7. This is the highest P – RLC HSDPABearer  throughput that the bearer can provide on each pixel. TP P – RLC  c  . –3 T TTI is the TTI duration. 2 10 s (2000 TTI in one second). c. c. the peak RLC throughput provided to the user is calculated as follows: DL TP P – RLC =  DL TP P –RLC  c  c  Serving cell Coverage consists of several independent layers that can be displayed and hidden on the map.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Atoll displays either the CPICH CQI (see the calculation detail in "CPICH CQI Determination" on page 244) when the selected option in Global parameters (HSDPA part) is CQI based on CPICH quality. TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used.e. For each layer. f TP – Scaling and TP Offset respectively represent the scaling factor between the application throughput and the RLC (Radio Link Control) throughput and the throughput offset.Atoll 3. it is defined in the terminal user equipment category properties. 308 . Atoll finds the corresponding BLER. There is a layer per possible DL Effective RLC throughput ( TP E – RLC ). For each layer. TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used. area is covered if the average effective RLC throughput exceeds the user-defined thresholds. BLER HSDPA is read in the quality graph defined for the triplet “reception equipment-selected bearer-mobility” (HSDPA Quality Graphs tab in the Reception equipment properties). Atoll finds the corresponding BLER. • Colour per Average Effective RLC Throughput DL Atoll displays the average effective RLC throughput ( TP Av – E – RLC ) provided on each pixel. This graph describes the variation of BLER as a function of the measured quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt. we have: DL TP P –RLC  c    1 – BLER HSDPA  DL TP Av –E –RLC = -----------------------------------------------------------------------TTI For MC-HSDPA and DB-MC-HSDPA users. overhead. This graph describes the variation of BLER as a function of the measured quality (HS-PDSCH Ec/Nt). These two parameters model the header information and other supplementary data that does not appear at the application level. • Colour per Application Throughput DL Atoll displays the application throughput ( TP A ) provided on each pixel. addressing. It is calculated as follows: DL DL TP A = TP Av –E – RLC  f TP – Scaling – TP Offset Where: DL TP Av –E –RLC is the average effective RLC throughput. it is defined in the terminal user equipment category properties. area is covered if the Effective RLC throughput exceeds the user-defined thresholds. There is a layer per possible DL application throughput ( TPA ).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 DL Atoll displays the Effective RLC throughput ( TP E – RLC ) provided on each pixel. Coverage consists of several independent layers that can be displayed and hidden on the map. area is covered if the application throughput exceeds the user-defined thresholds. For an HSDPA user. TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used. BLER HSDPA is read in the quality graph defined for the triplet “reception equipment-selected bearer-mobility” (HSDPA Quality Graphs tab in the Reception equipment properties). etc. it is defined in the terminal user equipment category properties. Coverage consists of several independent layers that can be displayed and hidden on the map. Knowing the HS-PDSCH Ec/Nt. we have:  DL DL  TP P –RLC  c    1 – BLER HSDPA    Serving cells TP Av –E –RLC = c---------------------------------------------------------------------------------------------------------TTI Where. There is a layer per possible DL average effective RLC throughput ( TP Av –E –RLC ). The application throughput represents the net throughput after deduction of coding (redundancy. The Effective RLC throughput is calculated as follows: DL TP P –RLC DL TP E – RLC = ----------------TTI Where TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used. Coverage consists of several independent layers that can be displayed and hidden on the map. They are defined in the service properties.3.). For each layer. For each layer. • Colour per Effective MAC Throughput per User DL Atoll displays the average Effective MAC throughput per user (  TP E –M AC  Av ) provided on each pixel. The average effective RLC throughput per user is calculated as follows: n HSDPA  DL TP E –R LC  x  DL x=1  TP E –R LC  Av = ----------------------------------------n HSDPA Where. There is a layer per possible DL average Effective MAC throughput per user (  TP E –M AC  Av ). see "Fast Link Adaptation Modelling" on page 242. For each layer. For each layer.3. the number of HSDPA bearer users is taken from the cell properties.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Fast Link Adaptation Modelling For Several Users When you calculate the study with the following display options. DL TP E – R LC  x  is the Effective RLC throughput of each HSDPA bearer user. n HSDPA is the number of HSDPA bearer users within the cell. The displayed results of the coverage prediction will be an average result for one user. 309 . n HSDPA  c  is the number of HSDPA bearer users within the cell. see "Colour per Effective MAC Throughput" on page 306. • Colour per RLC Throughput per User DL Atoll displays the average effective RLC throughput per user (  TP E –R LC  Av ) provided on each pixel. Coverage consists of several independent layers that can be displayed and hidden on the map. The average application throughput per user is calculated as follows: n HSDPA  DL TP A  x  DL x=1  TP A  Av = ---------------------------------n HSDPA Where. the cell available HSDPA power is shared between HSDPA bearer users. DL TP E – M AC  x  is the Effective MAC throughput of each HSDPA bearer user.Atoll 3. The average Effective MAC throughput per user is calculated as follows: n HSDPA  DL TP E –M AC  x  DL x=1  TP E –M AC  Av = ---------------------------------------------Max  n HSDPA  c   c  Serving cells  x  Where. area is covered if the average Effective MAC throughput per user exceeds the user-defined thresholds. In this case. see "HSDPA Bearer Allocation Process" on page 239 For further information on the fast link adaptation modelling. • Colour per ApplicationThroughput per User DL Atoll displays the average application throughput per user (  TPA  Av ) provided on each pixel. Coverage consists of several independent layers that can be displayed and hidden on the map. For further information on the calculation of the Effective RLC throughput. There is a layer per possible DL average effective RLC throughput per user (  TP E –R LC  Av ). For further information on the calculation of the Effective MAC throughput. c. area is covered if the average effective RLC throughput per user exceeds the user-defined thresholds. n HSDPA is the number of HSDPA bearer users within the cell. see "Colour per Effective RLC Throughput" on page 307. Atoll considers several users per pixel and determines the best HSDPA bearer that each user can obtain. When the coverage prediction is not based on a simulation. For further information on the HSDPA bearer allocation process when there are several users. the remaining load of the cell will be shared equally between all the HSUPA bearer users.Atoll 3. Coverage consists of several independent layers that can be displayed and hidden on the map.3. mobility and HSUPA service. The cell UL load factor due to HSUPA.6. These parameters can be either simulation outputs.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 DL TP A  x  is the application throughput of each HSDPA bearer user. area is covered if the selected HSDPA radio bearer is available. it can consider either a single HSPA user or several ones on each pixel. HSUPA resources can be shared by HSUPA users defined or calculated per cell: Atoll considers several HSUPA bearer users per pixel. Atoll determines on each pixel the best HSUPA bearer that can be obtained.2 Uplink load factor = 50% Uplink reuse factor = 1 Uplink load factor due to HSUPA = 0% Maximum uplink load factor = 75% Number of HSUPA beare users = 1 Calculation Options Atoll can calculate the HSUPA coverage prediction in one of two ways: • • 4.4. There is a layer per cell edge coverage probability defined in the Display tab. We assume that each pixel on the map corresponds to one or several users with HSUPA capable terminal.4. When no value is defined in the Cells table.3 HSUPA resources can be dedictated to a single user: On each pixel. In the last case. The displayed results of the coverage prediction will be an average result for one user. For each layer. the user is processed as if he is the only user in the cell i. for a specific layer. For further information on the HSUPA bearer selection. You can make the coverage prediction for a specific carrier. Coverage consists of several independent layers that can be displayed and hidden on the map. The cell UL reuse factor. 4. for all carriers. or user-defined cell inputs. The user does not create any interference. 4. Probability of Having a Certain Peak RLC Throughput This result can be obtained only if you have selected an HSDPA radio bearer in the Condition tab. To analyse the required E-DPDCH quality. or for all layers. The number of HSUPA bearer users within the cell if the study is calculated for several users. the number of HSUPA bearer users is taken from the cell properties.3. After allocating capacity to all R99 users. For each layer. see "Colour per Application Throughput" on page 308. Atoll uses the following default values: • • • • • 4. Note that the HSUPA service area is limited by the pilot quality and the A-DPCH-EDPCCH quality.3.4. area is covered if the average application throughput per user exceeds the user-defined thresholds.6 HSUPA Prediction Study A dedicated HSUPA study is available with different calculation and display options.3. There is a layer per possible DL average application throughput per user (  TP A  Av ).1 Prediction Study Inputs Parameters used as input for the HSUPA prediction study are: • • • • • The cell UL load factor. To analyse peak and effective throughputs. The maximum cell UL load factor. By calculating this study with suitable display options. • Colour per Cell Edge Coverage Probability Atoll shows areas where the selected HSDPA radio bearer is available with different cell edge coverage probabilities. it is possible: • • • To analyse the power required by the selected terminal. see "HSUPA Bearer Allocation Process" on page 261.e he will use the entire remaining load after allocating capacity to all R99 users.6. 310 . Display Options The following display options are available in the prediction property dialogue.6. When the coverage prediction is not based on a simulation.4.3. For further information on the calculation of the application throughput. Atoll reads the corresponding RLC peak throughput. UL S block is the transport block size (in kbits) for the selected HSUPA bearer. For each layer. Knowing the E-DPDCH Ec/Nt. it is defined for each HSUPA bearer in the HSUPA Radio Bearers table. T TTI is the duration of one TTI for the selected HSUPA bearer. Coverage consists of several independent layers that can be displayed and hidden on the map. The minimum effective RLC throughput corresponds to the RLC throughput obtained for a given BLER and the maximum number of retransmissions. Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per possible UL minimum effective RLC throughput ( TP Min – E –RLC ). Coverage consists of several independent layers that can be displayed and hidden on the map. This is the highest throughput that the selected HSUPA bearer can provide on each pixel. Coverage consists of several independent layers that can be displayed and hidden on the map. The Peak MAC throughput is calculated as follows: UL S block UL TP P – M AC = -----------T TTI Where. There is a layer per possible UL Peak MAC throughput ( TPP –M AC ). it is defined for each HSUPA bearer in the HSUPA Radio Bearers table.3. Coverage consists of several independent layers that can be displayed and req hidden on the map. This figure is read in the HSUPA Bearer Selection table.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Colour per Required E-DPDCH Ec/Nt Atoll displays on each pixel the E-DPDCH Ec/Nt required to obtain the selected HSUPA bearer. There is a layer per threshold. For each layer. There is a layer per threshold. BLER HSUPA is the residual BLER for the selected uplink transmission format (HSUPA bearer with N Rtx retransmissions). if  ------  Nt E – DPDCH Colour per Required Terminal Power Atoll displays on each pixel the terminal power required to obtain the selected HSUPA bearer. For each layer. Colour per Peak MAC Throughput UL Atoll displays the Peak MAC throughput ( TPP –M AC ) provided on each pixel. N Rtx is the maximum number of retransmissions for the selected HSUPA bearer. The required terminal power is calculated from the required E-DPDCH Ec/Nt. There is a layer per possible UL RLC peak throughput ( TP P – RLC ). Colour per Minimum Effective RLC Throughput UL Atoll displays the minimum effective RLC throughput ( TP Min – E – RLC ) provided on each pixel. For each layer. area is covered if P term  Threshold . area is covered Ec req  Threshold . This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt). area is covered if the Peak MAC throughput exceeds the user-defined thresholds. For each layer. area is covered if the minimum effective RLC throughput exceeds the user-defined thresholds. 311 .Atoll 3. Colour per Peak RLC Throughput After selecting the HSUPA bearer. Atoll finds the corresponding BLER. It is calculated as follows: UL TP P – RLC   1 – BLER HSUPA  UL TP Min – E – RLC = ---------------------------------------------------------------N Rtx Where. It is read in the quality graph defined for the quartet “reception equipment-selected bearer-number of retransmissions-mobility” (HSUPA Quality Graphs tab in the Reception equipment properties). area is covered if the peak RLC throughput can be provided. For each layer. There is a layer per possible UL application throughput ( TP A ).3. Coverage consists of several independent layers that can be displayed and hidden on the map. Colour per Application Throughput UL Atoll displays the application throughput ( TP A ) provided on each pixel. This is the RLC throughput obtained for a given BLER and the average number of retransmissions. This one is calculated as follows: UL UL TP A  M b  = TP Min – E –RLC  f TP – Scaling – TP Offset Where: f TP – Scaling and TP Offset respectively represent the scaling factor between the application throughput and the minimum RLC (Radio Link Control) throughput and the throughput offset. the required average number of retransmissions is smaller and the UL Effective RLC throughput is an average effective RLC throughput ( TP Av –E –RL C ). 312 . Knowing the E-DPDCH Ec/Nt. Colour per Average Application Throughput UL Atoll displays the average application throughput ( TP Av – A ) provided on each pixel. It is calculated as follows: UL TP P –RLC   1 – BLER HSUPA  UL TP Av –E –RL C = --------------------------------------------------------------- N Rtx  av BLER HSUPA is the residual BLER for the selected uplink transmission format (HSUPA bearer with N Rtx retransmissions). Atoll calculates the average number of retransmissions (  N Rtx  av ) as follows: N  Rtx max  N  p  N Rtx  – p  N Rtx – 1    N Rtx =1 Rtx  N Rtx  av = ----------------------------------------------------------------------------------------------p   N Rtx  max  Coverage consists of several independent layers that can be displayed and hidden on the map. Atoll finds the corresponding BLER. For each layer. It is read in the quality graph defined for the quartet “reception equipment-selected bearer-number of retransmissions-mobility” (HSUPA Quality Graphs tab in the Reception equipment properties). These two parameters model the header information and other supplementary data that does not appear at the application level.Atoll 3. It is calculated as follows: UL UL TP Av – A  M b  = TP Av –E –RL C  f TP – Scaling – TP Offset Where: f TP – Scaling and TP Offset respectively represent the scaling factor between the average application throughput and the average RLC (Radio Link Control) throughput and the throughput offset. addressing. They are defined in the service properties. This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt). area is covered if the minimum effective RLC throughput exceeds the user-defined thresholds. The application throughput represents the net throughput after deduction of coding (redundancy. Coverage consists of several independent layers that can be displayed and hidden on the map. overhead. The Early Termination Probability graph shows the probability of early termination ( p ) as a function of the number of retransmissions ( N Rtx ). The average number of retransmissions (  N Rtx  av ) is determined from early termination probabilities defined for the selected HSUPA bearer (in the HSUPA Bearer Selection table). etc. area is covered if the application throughput exceeds the user-defined thresholds. There is a layer per possible UL average application throughput ( TP Av – A ). There is a layer per possible UL average effective RLC throughput ( TP Av – E – RL C ).0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Colour per Average Effective RLC Throughput When HARQ (Hybrid Automatic Repeat Request) is used. area is covered if the average application throughput exceeds the user-defined thresholds. For each layer.). They are defined in the service properties. These two parameters model the header information and other supplementary data that does not appear at the application level. For information on the effective distance calculation. • The calculation options: • Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. cell A.5.3. cell B. They must fulfil following conditions: • • • • They are active. the following are explained: • • • "Neighbour Allocation for All Transmitters" on page 312. Atoll checks the following conditions: • The distance between both cells must be less than the user-definable maximum inter-site distance. and a candidate neighbour.1 Neighbour Allocation for All Transmitters We assume that we have a reference.atl document are potential neighbours. The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. They are located inside the focus zone. Atoll will allocate neighbours only to the cells using the selected carriers. When the automatic neighbour allocation starts. They satisfy the filter criteria applied to the Transmitters folder. This constraint can be weighted among the others and ranks the neighbours through the importance field (see after). Two allocation algorithms are available. In this section. For inter-carrier neighbours. you can select the carrier(s) of potential neighbours. Only TBA cells may be assigned neighbours. • • Force co-site cells as neighbours: This option enables you to force cells located on the reference cell site in the candidate neighbour list. If no focus zone exists in the . The cells to be allocated will be called TBA cells. "Neighbour Allocation for a Group of Transmitters or One Transmitter" on page 316. Atoll takes into account the computation zone.5 Automatic Neighbour Allocation Atoll permits the automatic allocation of intra-technology neighbours in the current network.atl document. Atoll calculates the effective distance. then the candidate neighbour is discarded. This folder can be either the Transmitters folder or a group of transmitters or a single transmitter.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4. It means that all the cells of TBC transmitters of your . "Importance Calculation" on page 316. in addition to the carrier(s) on which you want to run the allocation. which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. 4. You may choose one or more carriers. Force adjacent cells as neighbours (only for intra-carrier neighbours): This option enables you to force cells geographically adjacent to the reference cell in the candidate neighbour list. 313 .Atoll 3. If the distance between the reference cell and the candidate neighbour is greater than this value. one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours. see "Calculation of the Inter-Transmitter Distance" on page 319. They belong to the folder on which allocation has been executed. This constraint can be weighted among the others and ranks the neighbours through the importance field (see below). Atoll adds all the cells adjacent across network layers to the reference cell to the candidate neighbour list. This weight is used to calculate the rank of each neighbour and its importance. You can force Atoll to keep that cell in the reference cell’s neighbours list by adding the following option in the Atoll. adjacent cells are sorted and listed from the most adjacent to the least adjacent. SA is the area where the cell A is the best serving cell. this one will be considered as candidate neighbour of the reference cell. Therefore.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Adjacency criterion: Let CellA be a candidate neighbour cell of CellB. the existing neighbours are kept. Adjacency is relative to the number of pixels satisfying the criterion. • Delete existing neighbours: When selecting the Delete existing neighbours option. the reference cell will not be added as a neighbour of that cell and that cell will be removed from the reference cell’s neighbours list. 314 . • I BS  A  is the highest one.Atoll 3. if the reference cell is a candidate neighbour of another cell. Therefore. The best server indicator of A ( I BS  A  ) exceeds the minimum pilot quality (min Ec/I0). see "Best Serving Cell Determination in Monte Carlo Simulations . • • The pilot signal level received from A is greater than the minimum pilot signal level (min RSCP). If the Use Coverage Conditions check box is selected.3. only the distance criterion is taken into account. It means that the cell A is the first one in the active set. Cells are considered adjacent across layers if they belong to different layers and have a coverage overlap of at least one pixel. When Force adjacent cells as neighbours is selected. Force symmetry: This option enables user to force the reciprocity of a neighbourhood link. The reference cell A and the candidate cell B are using the carrier c1 (c1 is the selected carrier on which you run the allocation). Atoll deletes all the current neighbours and carries out a new neighbour allocation. CellA is considered adjacent to CellB if there exists at least one pixel in the CellB Best Server coverage area where CellA is Best Server (if several cells have the same best server value) or CellA is the second best server that enters the Active Set (respecting the HO margin of the allocation). The overlapping zone ( S A  S B ) is defined as follows: • Intra-carrier neighbours: intra-carrier handover is a soft handover. depending on the above criterion. you may force/forbid a cell to be candidate neighbour of the reference cell. Otherwise. The weight of this constraint is always the average of the Min and Max values defined for the adjacency factor.Old Method" on page 283. • • Force adjacent layers as neighbours: If selected. If not selected. there must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability. If the neighbours list of a cell is full.ini file: [Neighbours] DoNotDeleteSymmetrics = 1 • • Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. For information on the best server indicator calculation. Atoll 3. 2nd case: The cell A is not the best serving cell among all cells using c1 but it can enter the active set of a user connected to c1. SA is the area where the reference cell A is either the best serving cell among all cells using c1 (1st case) or a cell that can enter the active set of a user connected to c1 (2nd case). It is strictly lower than the best server indicator of the best serving cell and greater than the best server indicator of the best serving cell minus the handover margin. The best server indicator of B ( I BS  B  ) exceeds the minimum pilot quality (min Ec/I0). The reference cell A is using the carrier c1 (c1 is the carrier selected in Source) and the candidate cell B is using the carrier c2 (c2 is the carrier selected in Destination). Figure 4. • I BS  B  is the highest one. The best server indicator of A ( I BS  A  ) exceeds the minimum pilot quality (min Ec/I0). • • The pilot signal level received from B is greater than the minimum pilot signal level (min RSCP). 315 .16: Overlapping Zone for Intra-carrier Neighbours • Inter-carrier neighbours: inter-frequency handover is a hard handover. I BS  A  is not the highest one. • I BS  A  is lower than the minimum pilot quality (min Ec/I0) plus the handover margin.3. • • The pilot signal received from the cell B is greater than the minimum pilot signal level (min RSCP). • • • The pilot signal level received from A is greater than the minimum pilot signal level (min RSCP). The pilot quality from B is greater than the pilot quality from A minus the Ec/I0 margin (AS Threshold). triggered in multi-carrier W-CDMA networks for coverage reasons (1st case) and to balance the load between carriers (2nd case). • • The pilot signal level received from A is greater than the minimum pilot signal level (min RSCP). SB is the area where the cell B is the best serving cell among all cells using c2. The best server indicator of A ( I BS  A  ) is the highest one.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 SB is the area where the cell B can enter the active set. • • 1st case: The cell A is the best serving cell among all cells using c1 but its pilot quality starts significantly decreasing. . • 316 The importance of neighbours. 100 ) and compares this value to the % minimum covered area.3.Atoll 3. The service and terminal are selected such that the selection gives the largest possible coverage areas for the cells. • Two ways enable you to determine the I0 value: 1. SA  SB Atoll calculates the percentage of covered area ( -----------------. Atoll uses the service with the lowest body loss. if the option is selected.17: Overlapping Zone for Inter-carrier Neighbours .0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Figure 4. Defined per Cell: Atoll takes into account the total downlink power defined per cell. Atoll takes into account the pilot power of the cell.18: Overlapping Zone for Inter-carrier Neighbours . if %  Pmax  P pilot . 2.1st case Figure 4.e. If SA this percentage is not exceeded. I0 represents the sum of total transmitted powers. • For calculating the overlapping coverage areas. the candidate neighbour B is discarded. Then. i.Old Method" on page 283. the terminal that has the highest difference between gain and losses or the lowest noise figure when all terminals have the same (gain-losses) value. If the % of maximum power is too low. see "Best Serving Cell Determination in Monte Carlo Simulations .1st Case For information on the best server indicator calculation. Global Value: A percentage of the cell maximum power is considered. I0 represents the sum of values calculated for each cell. and the shadowing margin calculated using the defined cell edge coverage probability. 4. In the Results part.ini: [Neighbours] RealInterSiteDistanceCondition=1 • By default. the percentage of area meeting the adjacency conditions and the corresponding surface area (km2). As a consequence. coverage or symmetric. co-site. neighbours are marked as existing. if cells have previous allocations in the list. see "Importance Calculation" on page 316. Therefore. Among these 15 candidate neighbours. adjacent and symmetric. Neighbours of TBA cells that satisfy coverage conditions. 4. a neighbour may be marked as exceptional pair.5.1 Importance of Intra-carrier Neighbours The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf. there can be cases where the real distance between assigned neighbours is higher than the Max inter-site distance. because the effective distance is smaller. To avoid that. there can be cases where the calculated importance is different when the global Max inter-site distance is modified. it provides the list of neighbours. Neighbourhood cause When Importance value Existing neighbour Only if the Delete existing neighbours option is not selected and in case of a new allocation Existing importance Exceptional pair Only if the Force exceptional pairs option is selected 100 % Co-site cell Only if the Force co-site cells as neighbours option is selected Importance Function (IF) 317 .3. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be allocated to each transmitter is exceeded. the neighbour importance calculated with respect to distance is based on the global Max inter-site distance setting for all neighbour candidates. Atoll displays the percentage of area meeting the coverage conditions and the corresponding surface area (km2). Finally. You can force Atoll to compare the Max inter-site distance with the real inter-site distance by adding the following lines in Atoll. Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason. Neighbours of TBA cells marked as exceptional pair. For these cells. If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8.5. it indicates the importance (in %) of each neighbour and the allocation reason.3. In addition.2 Neighbour Allocation for a Group of Transmitters or One Transmitter Atoll allocates neighbours to: • • • TBA cells. If defined there. and to quantify the neighbour importance. the number of neighbours and the maximum number of neighbours allowed for each cell. Note that specific maximum numbers of neighbours (maximum number of intra-carrier neighbours. As a consequence. this value varies between 0 and 100%. Atoll only displays the cells for which it finds new neighbours. the automatic neighbour allocation compares the defined Max intersite distance with the effective inter-cell distance.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 For information on the importance calculation.3 Importance Calculation Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason and the distance. maximum number of inter-carrier neighbours) can be defined at the cell level (property dialogue or cell table). only 8 (having the highest importance values) will be allocated to the reference cell. Automatic neighbour allocation parameters are described in "Neighbour Allocation for All Transmitters" on page 312. table below).Atoll 3. this value is taken into account instead of the default one available in the Neighbour Allocation dialogue. adjacent.ini: [Neighbours] CandidatesMaxDistanceInImportanceCalculation=1 4. you can force Atoll to prioritise the individual distances between reference cells and their respective neighbour candidates by adding the following lines in Atoll. For neighbours accepted for co-site. adjacency and coverage reasons. • By default.5. d max is the maximum distance between the reference transmitter and a possible neighbour. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 30% Adjacency factor (A) Min(A) 30% Max(A) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The Importance Function is evaluated as follows: Neighbourhood cause Importance Function Resulting IF using the default values from the table above Coverage Min(O)+Delta(O){Max(Di)(Di)+(100%Max(Di))(O)}+Min(Di)+Delta(Di)(Di) 10%+20%{10%(Di)+90%(O)}+1%+9%(Di) Adjacent layer (Min(A)+Max(A))/2 45% Adjacent cells Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 30%+30%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Co-site cells Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 60%+40%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Where: Delta(X)=Max(X)-Min(X) 318 . For information on the effective distance calculation. d Di  = 1 – ---------d max d is the effective distance (in m). The overlapping factor (O): the percentage of overlapping.Atoll 3. It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Neighbourhood cause When Importance value Adjacent layer Only if the Force adjacent layers as neighbours option is selected Importance Function (IF) Adjacent cell Only if the Force adjacent cells as neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions Only if the % minimum covered area is exceeded Importance Function (IF) Symmetric neighbourhood relationship Only if the Force neighbour symmetry option is selected Importance Function (IF) Except the case of forced neighbours (importance = 100%). The adjacency factor (A): the percentage of adjacency.3. The minimum and maximum importance assigned to each of the above factors can be defined. The IF considers fourfactors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. • • • The co-site factor (C): a Boolean. see "Calculation of the Inter-Transmitter Distance" on page 319. priority assigned to each neighbourhood cause is determined using the Importance Function (IF). d Di  = 1 – ---------d max d is the effective distance (in m). It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. and neighbours allocated based on coverage overlapping. If the Min and Max value ranges of the importance function factors overlap. The IF is user-definable using the Min importance and Max importance fields. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The IF evaluates importance as follows: Co-site Neighbourhood cause IF Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)} Yes Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+ Max(O)(O)/(Max(Di)+Max(O))} 60%+40%{1/7%(Di)+6/7%(O)} Where Delta(X)=Max(X)-Min(X) 319 . the neighbours may be ranked differently. There can be a mix of the neighbourhood causes.5. this value varies between 0 to 100%. d max is the maximum distance between the reference transmitter and a possible neighbour.2 Importance of Inter-carrier Neighbours As indicated in the table below.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 • • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation.Atoll 3. see "Calculation of the Inter-Transmitter Distance" on page 319. 4. the neighbour importance depends on the distance and on the neighbourhood cause. Neighbourhood cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site cell If the Force co-site cells as neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions If the % minimum covered area is exceeded Importance Function (IF) Symmetric neighbourhood relationship If the Force neighbour symmetry option is selected Importance Function (IF) Except the case of forced neighbours (importance = 100%). For information on the effective distance calculation. neighbours will be ranked in this order: co-site neighbours. If the Min and Max value ranges of the importance function factors do not overlap. • • The co-site factor (C): a Boolean. With a value of Min(O) = 0%. neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping. The IF considers threefactors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. The overlapping factor (O): the percentage of overlapping. With the default values for minimum and maximum importance fields. the neighbours will be ranked by neighbour cause. priority assigned to each neighbourhood cause is determined using the Importance Function (IF).3. The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. adjacent neighbours.3. 4. the neighbours may be ranked differently.Atoll 3. It is this effective distance that will be taken into account rather than the real distance. 4.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks • • • • ©Forsk 2015 Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation.4 Appendices 4. The cells to which Atoll allocates scrambling codes are referred to as the TBA cells (cells to be allocated).6 Primary Scrambling Code Allocation Downlink primary scrambling codes enable you to distinguish cells from one another (cell identification). With the default values for minimum and maximum importance fields. With a value of Min(O) = 0%. 320 . The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. They satisfy the filter criteria applied to the Transmitters folder. 4. Figure 4. neighbours will be ranked in this order: co-site neighbours and neighbours allocated based on coverage overlapping.5.19: Inter-Transmitter Distance Computation The formula above implies that two cells facing each other will have a smaller effective distance than the real physical distance. They are located inside the focus zone. They belong to the folder on which allocation has been executed. By default. TBA cells fulfil following conditions: • • • • They are active.atl document. neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping. This folder can be either the Transmitters folder or a group of transmitters or a single transmitter.3% so that the maximum D variation does not exceed 1%. Atoll takes into account the computation zone. d = D   1 + x  cos  – x  cos   where x = 0.3.1 Calculation of the Inter-Transmitter Distance Atoll takes into account the real distance ( D in m) and azimuths of antennas in order to calculate the effective intertransmitter distance ( d in m). There can be a mix of the neighbourhood causes. If the Min and Max value ranges of the importance function factors do not overlap. the neighbours will be ranked by neighbour cause.5.. there are 512 primary scrambling codes numbered (0.. If the Min and Max value ranges of the importance function factors overlap. If no focus zone exists in the .511). its second order neighbours and its third order neighbours. • Domains of scrambling codes.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4. scrambling codes will be distributed in 128 clusters). Cells fulfilling a criterion on Ec/I0 (option “Additional Overlapping Conditions”). For a reference cell “A”. When this parameter is not specified in the cell properties.atl document. However. you can define another value (e. Reuse distance. Third order neighbours: The neighbour’s neighbour’s neighbours. 3GPP specifications define 64 clusters consisting of 8 scrambling codes (in this case. A code cannot be reused at a cell that is not at least as far away as the reuse distance from the cell allocated with the particular code.6. you must make the Transmitters folder of the GSM . Atoll can take into account inter-technology neighbour relations as constraints to allocate different scrambling codes to the UMTS neighbours of a GSM transmitter. Atoll considers the 512 primary scrambling codes available. Atoll considers symmetry relationship between a cell. Atoll considers all the cells “B” that can enter the active set on the area where the reference cell is the best server (area where (Ec/I0)A exceeds the minimum Ec/I0 and is the highest one and (Ec/I0)B is within a Ec/I0 margin of (Ec/I0)A). For information on making links between GSM and UMTS . The defined interval is applied by adding the following lines in the Atoll. see the User Manual. • The number of primary scrambling codes per cluster. • • Exceptional pairs. Second order neighbours: The neighbours of neighbours. clusters are numbererd from 0 to 63). if you set the number of codes per cluster to 4.3. In this case. In order to consider inter-technology neighbour relations in the scrambling code allocation.ini file: [PSC] 321 .1 Options and Constraints The scrambling code allocation algorithm can take into account following constraints and options: • Neighbourhood between cells. and to calculate their importance. When the allocation is based on a Distributed strategy (Distributed per Cell or Distributed per Site). In Atoll. the term "neighbours" refers to intra-carrier neighbours. Atoll uses 50% of the maximum power. Scrambling code reuse distance can be defined at cell level.1. • • Atoll considers either a percentage of the cell maximum powers or the total downlink power used by the cells in order to evaluate I0. Atoll reuses the intra-carrier neighbour allocation algorithm to determine the list of cells which cannot be allocated the same scrambling code.g.1 Automatic Allocation Description 4. When no domain is assigned to cells. we call "cluster". You may consider: • • • First order neighbours: The neighbours of TBA cells listed in the Intra-technology neighbours table.Atoll 3.atl document accessible in the UMTS . If this value is not defined. • Reuse distance is a constraint on the allocation of scrambling codes.6. then Atoll will use the default reuse distance defined in the Scrambling Code Automatic Allocation dialogue. I0 equals the sum of total transmitted powers. this parameter can also be used to define the interval between the primary scrambling codes assigned to cells on a same site.atl documents. its first order neighbours. • • • • In the context of the primary scrambling code allocation. a group of scrambling codes as defined in 3GPP specifications. one code from the cluster to each cell of each transmitter. see the Administrator Manual. The neighbour cells cannot share the same cluster (for the "Distributed per site" allocation strategy only).1 Single Carrier Network The allocation process depends on the selected strategy.1. 322 . Distributed per cell allocation: This strategy consists in using as many clusters as possible. then. In this case. When a cell has too many constraints and there are not anymore scrambling codes available. For information on the cost generated by each constraint. During the allocation. In the Results table. Atoll will preferentially allocate codes from different clusters. Atoll tries to assign different scrambling codes to the TBA cell and its near cells. It allocates scrambling codes starting with the highest priority cell and its near cells. 4. this information is required to start allocation based on this strategy. and continuing with the lowest priority cells not allocated yet and their near cells. The cells with distance from the TBA cell less than the reuse distance. the existing scrambling codes are kept.atl document is accessible in the UMTS . The neighbours of its neighbours (options “Existing neighbours” and “Second Order”). Atoll lists all cells which have constraints with the cell. scrambling codes among a minimum number of clusters.6. then. In addition. These constraints have a certain weight taken into account to determine the TBA cell priority during the allocation process and the cost of the scrambling code plan. or it allocates the same primary scrambling code to each carrier of a transmitter if the option "Allocate carriers identically" is selected.3. see "Cell Priority" on page 323. Allocation strategies can be: • • • • Clustered allocation: The purpose of this strategy is to choose for a group of mutually constrained cells. Atoll will preferentially allocate all the codes within the same cluster. In this case.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 ConstantStep = 1 For more information about setting options in the atoll.1. Distributed per site allocation: This strategy allocates a group of adjacent clusters to each site. either Atoll independently plans scrambling codes for the different carriers. Atoll reuses the clusters as far as possible at another site.Atoll 3. Atoll only displays scrambling codes allocated to TBA cells. one cluster to each transmitter on the site according to its azimuth and finally. The third order neighbours (options “Existing neighbours” and “Third Order”). If not selected. One cluster per site allocation: This strategy allocates one cluster to each site. • The carrier on which the allocation is run: It can be a given carrier or all of them. the cost of the scrambling code plan is 0. • The possibility to use a maximum of codes from the defined domains (option "Use a Maximum of Codes"): Atoll will try to spread the scrambling code spectrum the most. one code from the cluster to each cell of each site. see "Cell Priority" on page 323. Additional constraints are considered when: • • The cell and its near cells are neighbours of a same GSM transmitter (only if the Transmitters folder of the GSM . The cells that make exceptional pairs with the TBA cell. For information on calculating cell priority.atl document). The number of adjacent clusters per group depends on the number of transmitters per site you have in your network. • The "Delete All Codes" option: When selecting this option. Algorithm works as follows: Strategies: Clustered and Distributed per Cell Atoll processes TBA cells according to their priority. If it respects all the constraints. Atoll deletes all the current scrambling codes and carries out a new scrambling code allocation. it depends on the selected allocation strategy. When all the groups of adjacent clusters have been allocated and there are still sites remaining to be allocated. They are referred to as near cells. 4.2 Allocation Process For each TBA cell. The cells that fulfil Ec/I0 condition (option “Additional Overlapping Conditions”). Atoll reuses the groups of adjacent clusters as far as possible at another site. When all the clusters have been allocated and there are still sites remaining to be allocated.ini file. The near cells of a TBA cell may be: • • • • • • Its neighbour cells: the neighbours listed in the Intra-technology neighbours table (options “Existing neighbours” and "First Order").6.2. Atoll breaks the constraint with the lowest cost so as to generate the scrambling code plan with the lowest cost. Therefore. Determination of Groups of Adjacent Clusters In order to determine the groups of adjacent clusters to be used. Let us assume that we have a domain consisted of 12 clusters: clusters 1 to 8 and clusters 12 to 15. Then. Atoll can use all these groups for the allocation.3. see "Site Priority" on page 325. When all the clusters have been allocated and there are still sites remaining to be allocated. 7 and 8.. 2 and 15 will not be used. Strategy: Distributed per Site All sites which have constraints with the studied site are referred to as near sites. Then. 64 clusters are supposed to be available). the user is warned through the 'Event Viewer'. Otherwise. Atoll assigns a cluster to each site. 4. When all the groups of adjacent clusters have been allocated and there are still sites remaining to be allocated.2. see "Cell Priority" on page 323. Group 6 with cluster 12. 4 and 5. the algorithm tries to assign reused groups of adjacent clusters as spaced out as possible. In this case. Atoll reuses the clusters at another site. independently of the defined domain. When the Reuse Distance option is selected. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet and their near cells. When the Reuse Distance option is selected. when the option is not selected.Atoll 3.6. considering the 512 primary scrambling codes available and the specified number of codes per cluster (if this one is set to 8. 323 . Group 21 . Atoll assigns each cluster of the group to each transmitter of the site according to the transmitter azimuth and selected neighbourhood constraints (options "Neighbours in Other Clusters" and "Secondary Neighbours in Other Clusters"). the allocation process depends on the allocation strategy as detailed above and in addition. On the other hand..2 Multi-Carrier Network In case you have a multi-carrier network and you run the scrambling code allocation on all the carriers. wether the option "Allocate Carriers Identically" is selected or not. the number of possible groups. when the option is not selected. Cluster 61 Cluster 62 Cluster 63 If no domain is assigned to cells. Atoll reuses the groups of adjacent clusters at another site. the tool compares adjacent clusters really available in the assigned domain to the theoretical groups and only keeps adjacent clusters mapping the theoretical groups.. For information on calculating site priority.1. For information on calculating cell priority. Let us assume that the number of codes per cluster is set to 8 and the maximum number of transmitters per site in the network is 3. starting with the highest priority site and its near sites. It starts the division in group from the cluster 0 (hard coded) and takes into account the maximum number of transmitters per site user-specified in order to determine the number of clusters in each group and then. we have the following theoretical groups: Group 1 Group 2 Group 3 Group 4 Cluster 0 Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Cluster 6 Cluster 7 Cluster 8 Cluster 9 Cluster 10 Cluster 11 . and continuing with the lowest priority sites not allocated yet and their near sites. Atoll proceeds as follows: It defines theoretical groups of adjacent clusters. the algorithm reuses the clusters as soon as the reuse distance is exceeded. and continuing with the lowest priority sites not allocated yet and their near sites. The clusters 1. the algorithm tries to assign reused clusters as spaced out as possible. if a domain is used. Then. see "Site Priority" on page 325. 13 and 14. For information on calculating cell priority. If a domain does not contain any adjacent clusters. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet and their near cells. starting with the highest priority site and its near sites. Otherwise. see "Cell Priority" on page 323.. For information on calculating site priority.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Strategy: One Cluster per Site All sites which have constraints with the studied site are referred to as near sites. Atoll will be able to use the following groups of adjacent clusters: • • • • Group 2 with cluster 3. Group 3 with cluster 6. Atoll assigns a group of adjacent clusters to each site. Atoll allocates a primary scrambling code to each cell located on the transmitters (codes belong to the assigned clusters). the algorithm reuses the groups of adjacent clusters as soon as the reuse distance is exceeded. Atoll allocates a primary scrambling code from the cluster to each cell located on the sites (codes belong to the assigned clusters). 0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 When the option is not selected. For information on calculating transmitter priority. as explained above. Atoll assigns a group of adjacent clusters to each site. On the other hand. When cells. Atoll starts scrambling code allocation with the highest priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters.1 Cell Priority Scrambling code allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities assigned to cells depend upon how much constrained each cell is and the cost defined for each constraint. It starts with the highest priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. the less will be the cost due to this criterion. The same scrambling code is assigned to each cell of the transmitter. The total cost due to exceptional pair constraint is given as: 324 . A cell without any constraint has a default cost.1.3. the higher the priority it has for the scrambling code allocation process. see "Transmitter Priority" on page 325.1. allocates a scrambling code to each transmitter. There are six criteria employed to determine the cell priority: • Scrambling Code Domain Criterion The cost due to the domain constraint. when the option is selected. The same scrambling code is assigned to each cell of the transmitter. The higher the number of codes available in the domain.6. the weight for co-site cells is 1 and the weight for two cells spaced out 2100m apart is 0. Atoll assigns a cluster to each site and then.3 Priority Determination 4. This weight is inversely proportional to the inter-cell distance. All transmitters which have constraints with the studied transmitter will be referred to as near transmitters. each unavailable scrambling code generates a cost. 512 scrambling codes are available and we have: C i  Dom  = 0 When domains of scrambling codes are assigned to cells. 4.6. It is no longer based on the cell priority but depends on the transmitter priority.Atoll 3. C i  Dom  . algorithm works for each strategy. The same scrambling code is assigned to each cell of the transmitter. allocates a scrambling code to each transmitter. depends on the number of scrambling codes available for the allocation. This value can be defined in the Constraint Cost dialogue.25. It starts with the highest priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. The total cost due to the distance constraint is given as: C i  Dist  =  Cj  Dist  i   j Each cell j within the reuse distance generates a cost given as: C j  Dist  i   = w  d ij   c dis tan ce Where w  d ij  is a weight depending on the distance between i and j. the weight for an inter-cell distance of 1500m is 0. C . In case of the "One cluster per site" strategy. c dis tan ce is the cost of the distance constraint. The cost is given as: C i  Dom  = 512 – Number of scrambling codes in the domain • Distance Criterion The constraint level of any cell i depends on the number of cells (j) present within a radius of "reuse distance" from its centre. For a reuse distance of 2000m. The higher the cost on a cell. In case of the "Distributed per site" strategy. When no domain is assigned to cells. transmitters or sites have the same priority. The domain constraint is mandatory and cannot be broken. In case of a "Per cell" strategy (Clustered and Distributed per cell).3. equal to 0. • Exceptional Pair Criterion The constraint level of any cell i depends on the number of exceptional pairs (j) for that cell. processing is based on an alphanumeric order. allocation order changes. then a cluster to each transmitter and finally. 20: Neighbourhood Constraints The total cost due to the neighbour constraint is given as:  Ci  N  =         Cj  N1  i   +  Cj – j  N1  i   +   Ck  N2  i   +  Ck – k  N2  i   +   Cl  N3  i   +  Cl – l  N3  i   j j k k l l Each first order neighbour cell j generates a cost given as: C j  N1  i   = I j  c N1 Where I j is the importance of the neighbour cell j.Atoll 3. C k  N2  i   + C k  N2  i   C k – k  N2  i   = ------------------------------------------------------2 Each third order neighbour cell l generates a cost given as:  C  N1  i    C k  N1  j    C l  N1  k   C j  N1  i    C k  N1  j    C l N1  k   C l  N3  i   = Max  j   c N3   C j  N1  i    C k  N1  j     C l N1  k  C j  N1  i    C k  N1  j    C l N1  k   Where c N3 is the cost of the third order neighbour constraint. Because two second order neighbours must not have the same scrambling code. This value can be defined in the Constraint Cost dialogue. Atoll considers the cost created by two first order neighbours to be each other. c N1 is the cost of the first order neighbour constraint. Because two third order neighbours must not have the same scrambling code. This value can be defined in the Constraint Cost dialogue. C j  N1  i    C k  N1  j   )  c N2 Where c N2 is the cost of the second order neighbour constraint. C l  N3  i   + C l  N3  i   C l – l  N3  i   = ----------------------------------------------------2 325 . Atoll considers the cost created by two third order neighbours to be each other.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 C i  EP  =  cEP  i – j  j Where c EP is the cost of the exceptional pair constraint. Atoll considers the cost created by two second order neighbours to be each other. This value can be defined in the Constraint Cost dialogue. • Neighbourhood Criterion The constraint level of any cell i depends on the number of its neighbour cells j. Because two first order neighbours must not have the same scrambling code. the number of second order neighbours k and the number of third order neighbours l. Let’s consider the following neighbour schema: Figure 4. This value can be defined in the Constraint Cost dialogue. C j  N1  i   + C j  N1  i   C j – j  N1  i   = ----------------------------------------------------2 Each second order neighbour cell k generates a cost given as: C k  N2  i   = Max ( C j  N1  i    C k  N1  j   .3. If the cell i is neighbour of a GSM transmitter. Let us consider a transmitter Tx with two cells using carriers 0 and 1.2 Transmitter Priority In case you have a multi-carrier network and you run scrambling code allocation on "all" the carriers with the option "allocate carriers identically". the higher the priority it has for the scrambling code allocation process. the domain available for the transmitter is the intersection of domains assigned to cells of the transmitter. In this case.1. The domain constraint is mandatory and cannot be broken. the higher the priority it has for the scrambling code allocation process. This value can be defined in the Constraint Cost dialogue. Priorities assigned to sites depend on how much constrained each site is and the cost defined for each constraint. This value can be defined in the Constraint Cost dialogue. Therefore.Atoll 3. its first order neighbours and its second order neighbours must be assigned scrambling codes from different clusters).atl document is made accessible in the UMTS . The higher the cost on a site. • Cluster Criterion When the "Distributed per Site" allocation strategy is used. 4. The total cost due to the cluster constraint is given as: C i  Cluster  =  Cj  N1  i    cCluster +  Ck  N2  i    cCluster j k Where c Cluster is the cost of the cluster constraint. In this case.3.3. The higher the cost on a transmitter. the total cost due to constraints on any cell i is defined as: C i = C i  Dom  + C i  U  With C i  U  = C i  Dist  + C i  EP  + C i  N  + C i  N 2G  + C i  Cluster  4.3 Site Priority In case of "Per Site" allocation strategies (One cluster per site and Distributed per site).C j  N1  k    C i  N1  j  )  c N2 • GSM Neighbour Criterion This criterion is considered when the co-planning mode is activated (i. The cost due to constraints on the transmitter is given as: C Tx = C Tx  Dom  + C Tx  U  With C Tx  U  = Max  C  U   and C  Dom  = 512 – Number of scrambling codes in the domain Tx i  Tx i Here.3. we have: C j  N1  i   = Max  I i – j I j – i   c N1 And C k  N2  i   = Max (C j  N1  i    C k  N1  j  .6. algorithm in Atoll allots priorities to sites.atl document) and inter-technology neighbours have been allocated. the Transmitters folder of the GSM . The total cost due to GSM neighbour constraint is given as: C i  N 2G  =  cN2G  j – Tx2G  j Where cN 2G is the cost of the GSM neighbour constraint. you can consider additional constraints on allocated clusters (one cell. the constraint level of any cell i depends on the number of first and second order neighbours.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 Atoll considers the highest cost of both links when a neighbour relation is symmetric and the importance value is different. the cell constraint level depends on how many cells j are neighbours of the same GSM transmitter. j and k.e.1. Priorities assigned to transmitters depend on how much constrained each transmitter is and the cost defined for each constraint. algorithm in Atoll allots priorities to transmitters. 326 .6. Site2 and Site3 be four sites with 3 cells using carrier 0 whom scrambling codes have to be allocated out of three clusters consisted of 8 primary scrambling codes.1. 4.2. let us consider the following sample scenario: Figure 4.3. Site1. 4. the domain considered for the site is the intersection of domains available for transmitters of the site. This implies that the domain of scrambling codes for the four sites is from 0 to 23 (cluster 0 to cluster 2). 327 . Then. every cell has the same priority.6. The following section lists the results of each combination of options with explanation where necessary.21: Primary Scrambling Codes Allocation Let Site0.1 Allocation Strategies and Use a Maximum of Codes In order to understand the differences between the different allocation strategies and the behaviour of algorithm when using a maximum of codes or not. The domain constraint is mandatory and cannot be broken.Atoll 3. The reuse distance is supposed to be less than the inter-site distance. Only co-site neighbours exist. The cost due to constraints on the site is given as: C S = C S  U  + C S  Dom  With C S  U  = Max  C  U   and C  Dom  = 512 – Number of scrambling codes in the domain S Tx  S Tx Here.6. each of them has two cells using carriers 0 and 1.2.1 Strategy: Clustered Since the restrictions of neighbourhood only apply to co-sites with the same importance and sites distances are greater than reuse distances.2 Allocation Examples 4.6.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Let us consider a site S with three transmitters. scrambling code allocation to cells is performed in an alphanumeric order. 0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks Without ‘Use a Maximum of Codes’ ©Forsk 2015 With ‘Use a Maximum of Codes’ Atoll starts allocating the codes from the start of cluster 0 at As it is possible to use a maximum of codes.2 Strategy: Distributed Since the restrictions of neighbourhood only apply to co-sites with the same importance and sites distances are greater than reuse distances. allocation at the start of a different cluster at each site.1. Then. cells. same codes can be allocated to each site’s allocates the codes so that there is least repetition of codes. Under given constraints of neighbourhood and cells. Atoll reuse distance.6.Atoll 3. and there are non allocated codes left in the cluster.2. As it is possible to use a maximum of codes. When a cluster is reused. every cell has the same priority. scrambling code allocation to cells is performed in an alphanumeric order. 328 . 4. Atoll starts each site. Without ‘Use a Maximum of Codes’ With ‘Use a Maximum of Codes’ Atoll allocates codes from different clusters to each cell of Atoll allocates codes from different clusters to each site’s the same site. Atoll first allocates those codes before reusing the already used ones.3. 3 Strategy: ‘One Cluster per Site Since the restrictions of neighbourhood only apply to co-sites with the same importance and sites distances are greater than reuse distances. every site has the same priority. Then.1. the group of adjacent clusters allocation to sites is performed in an alphanumeric order. 1 and 2) is available). 329 .2. a group of adjacent clusters is limited to be When it is possible to use a maximum of codes. Then. In this case Atoll reuses the clusters as far as possible at another site. In this case (here only one group of adjacent clusters (clusters 0. remaining to be allocated. Atoll can used at just one site at a time unless all codes and groups of allocate different codes from a reused group of adjacent adjacent clusters have been allocated and there are still sites cluster at another site.2. Without ‘Use a Maximum of Codes’ With ‘Use a Maximum of Codes’ In this strategy. allocated and there are still sites remaining to be allocated.1. 4. cluster allocation to sites is performed in an alphanumeric order. Atoll reuses the group at another site. Without ‘Use a Maximum of Codes’ With ‘Use a Maximum of Codes’ In this strategy.6. Atoll can one site at a time unless all codes and clusters have been allocate different codes from a reused cluster at another site. every site has the same priority.4 Strategy: ‘Distributed per Site Since the restrictions of neighbourhood only apply to co-sites with the same importance and sites distances are greater than reuse distances.Atoll 3.6. a cluster of codes is limited to be used at just When it is possible to use a maximum of codes.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4.3. it allocates a code from the cluster to one code to each transmitter so as to use a maximum of codes. They satisfy the filter criteria applied to Transmitters folder. transmitter. you must have: • • An . Site1. Inter-technology handover is used in two cases: • • When the UMTS coverage is not continuous. Scrambling codes have to be allocated out of 3 clusters consisted of 8 primary scrambling codes. In this case. Only co-site neighbours exist.7 Automatic GSM-UMTS Neighbour Allocation 4. They are located inside the focus zone. Note that the automatic inter-technology neighbour allocation algorithm takes into account both cases. The external neighbour allocation algorithm takes into account all the GSM TBC transmitters. the same code is given to each cell of the each cell of the site so as to use a maximum of codes.6. Then. They belong to the folder for which allocation has been executed.atl. GSM. 4. This implies that the domain of scrambling codes for the five sites is from 0 to 23 (cluster 0 to cluster 2).atl document containing the GSM network.2. Then. Then.atl. Without ‘Allocate Carriers Identically’ With ‘Allocate Carriers Identically’ Atoll allocates one cluster at each site as detailed in the In this case. 330 . every site has the same priority. Site2 and Site3 be four sites with 3 cells using carrier 0 and 3 cells using carrier 1. satisfy following conditions: • • • • They are active. cluster allocation to sites is performed in an alphanumeric order. Only UMTS TBA cells may be assigned neighbours. The reuse distance is supposed to be less than the inter-site distance.3. An existing link on the Transmitters folder of GSM.1 Overview You can automatically calculate and allocate neighbours between GSM and UMTS networks. In both cases (with and without ’Allocate Carriers Identically’). It means that all the TBC transmitters of GSM. and another one describing the UMTS network. The cells to be allocated will be called TBA cells which.atl into UMTS. being cells of UMTS.2 Allocate Carriers Identically In order to understand the behaviour of algorithm when using the option "Allocate Carriers Identically" or not. UMTS. previous section. the UMTS coverage is extended by UMTS-GSM handover into the GSM network.Atoll 3. Atoll allocates one cluster at each site and then.atl are potential neighbours. In Atoll.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 4. Allocation algorithm will be based on the "One Cluster per Site" strategy and the option "Use a Maximum of Codes" is selected. This folder can be either the Transmitters folder or a group of transmitters subfolder.atl.atl.7. let us consider the following sample scenario: Let Site0. And in order to balance traffic and service distribution between both networks. it is called intertechnology neighbour allocation. In order to be able to use the inter-technology neighbour allocation algorithm. In addition. Finally. For neighbours accepted for distance reasons. Therefore. or distance. A. • The importance of neighbours.3.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 4. The maximum number of neighbours fixed. If the distance between the UMTS reference cell and the GSM neighbour is greater than this value. Atoll will allocate neighbours to cells using the selected carriers. it indicates the importance (in %) of each neighbour and the allocation reason. co-site.1 Algorithm Based on Distance When the automatic allocation starts. transmitter B. • The calculation options. In the Results part. and a GSM candidate neighbour.2 Automatic Allocation Description The allocation algorithm takes into account criteria listed below: • • • • The inter-transmitter distance. As indicated in the table below. If defined there.Atoll 3. neighbours are marked as existing.2. The selected allocation strategy. which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. For information on the effective distance calculation. Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. Two allocation strategies are available: the first one is based on distance and the second one on coverage overlapping. see "Calculation of the Inter-Transmitter Distance" on page 319. then the candidate neighbour is discarded. Atoll checks the following conditions: • The distance between the UMTS reference cell and the GSM neighbour must be less than the user-definable maximum inter-site distance. 4. Note that the maximum number of inter-technology neighbours can be defined at the cell level (property dialogue or cell table).7. a neighbour may be marked as exceptional pair.7. If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8. Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Importance values are used by the allocation algorithm to rank the neighbours. We assume we have a UMTS reference cell. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be allocated to each cell is exceeded. Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site as the reference UMTS cell in the candidate neighbour list. if cells have previous allocations in the list. Among these 15 candidate neighbours. Delete existing neighbours: When selecting the Delete existing neighbours option. Therefore. only 8 (having the highest importance values) will be allocated to the reference cell. If not selected. Atoll displays the distance from the reference cell (m). This option is automatically selected. Atoll provides the list of neighbours. You may choose one or more carriers. Atoll calculates the effective distance. this value varies between 0 to 100%. the neighbour importance depends on the distance and on the neighbourhood cause. this value is taken into account instead of the default one available in the Neighbour Allocation dialogue. existing neighbours are kept. 331 . the number of neighbours and the maximum number of neighbours allowed for each cell. you may force/forbid a GSM transmitter to be candidate neighbour of the reference UMTS cell. Atoll deletes all the current neighbours and carries out a new neighbour allocation. Allocation options. Neighbourhood cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter If the Force co-site cells as neighbours option is selected 100 % Neighbourhood relationship that fulfils distance conditions If the maximum distance is not exceeded d1 – ---------d max Where d is the effective distance between the UMTS reference cell and the GSM neighbour and d max is the maximum intersite distance. 2nd case: The margin is different from 0dB and SB is the area where: • The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is within a margin from the best BCCH signal level. • The pilot quality from A is within a margin from the best Ec/I0 (where the best Ec/I0 exceeds the minimum Ec/ I0) and lower than the maximum Ec/I0. Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site as the reference UMTS cell in the candidate neighbour list.2 Algorithm Based on Coverage Overlapping When automatic allocation starts. If Atoll calculates the percentage of covered area ( ----------------SA this percentage is not exceeded. 4th case: SA represents the area where: • The pilot signal received from A is greater than the minimum pilot signal level. • There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability. • • • The pilot signal received from A is greater than the minimum pilot signal level. the candidate neighbour B is discarded. If not selected. For information on the effective distance calculation. then the candidate neighbour is discarded. 100 ) and compares this value to the % minimum covered area. In this case. the margin must be set to 0dB. the max Ec/I0 option selected and a maximum Ec/I0 userdefined. • 2nd case: SA represents the area where the pilot quality from the cell A strats decreasing but the cell A is still the best serving cell of the UMTS network. see "Calculation of the Inter-Transmitter Distance" on page 319. • The pilot quality from A is the highest one.2. 3rd case: SA represents the area where the cell A is not the best serving cell but can enter the active set. Carriers: This option enables you to select the carrier(s) on which you want to run the allocation.3. In this case. • The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/I0. Delete existing neighbours: When selecting the Delete existing neighbours option. Four different cases may be considered for SA: • 1st case: SA is the area where the cell A is the best serving cell of the UMTS network. you may force/forbid a GSM transmitter to be candidate neighbour of the reference UMTS cell. • The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0) and is the highest one. The Ec/I0 margin must be equal to 0dB. SA  SB . This option is automatically selected. • The pilot signal received from A is greater than the minimum pilot signal level. Here. the Ec/I0 margin must be equal to 0dB and the max Ec/I0 option disabled. The pilot quality from A is within a margin from the best Ec/I0. which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. • The calculation options.Atoll 3. Atoll deletes all the current neighbours and carries out a new neighbour allocation.7. 332 . Atoll calculates the effective distance. existing neighbours are kept. the Ec/I0 margin has to be different from 0dB and the max Ec/I0 option disabled. Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. the max Ec/I0 option selected and a maximum Ec/I0 user-defined. If the distance between the UMTS reference cell and the GSM neighbour is greater than this value. In this case. where the best Ec/I0 exceeds the minimum Ec/ I0. Two different cases may be considered for SB: • 1st case: SB is the area where the cell B is the best serving cell of the GSM network. You may choose one or more carriers. • • The pilot signal received from A is greater than the minimum pilot signal level. • • The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is the highest one. the margin must be different from 0dB.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks ©Forsk 2015 4. Atoll checks following conditions: • The distance between the UMTS reference cell and the GSM neighbour must be less than the user-definable maximum inter-site distance. Therefore. Atoll will allocate neighbours to cells using the selected carriers. Importance values are used by the allocation algorithm to rank the neighbours according to the distance and the allocation reason. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be allocated to each cell is exceeded. If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8. the max Ec/I0 option must be disabled. this value is taken into account instead of the default one available in the Neighbour Allocation dialogue. • • The co-site factor (C): a Boolean. A second allocation in order to complete the previous list with handovers motivated for reasons of traffic and service distribution. The overlapping factor (O): the percentage of overlapping. For information on the effective distance calculation.3. Note that the maximum number of inter-technology neighbours can be defined at the cell level (property dialogue or cell table). priority assigned to each neighbourhood cause is determined using the Importance Function (IF). The IF considers threefactors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. Here. we recommend you to perform two successive automatic allocations: • • • A first allocation in order to find handovers due to non-continuous UMTS coverage.Atoll 3. When the automatic allocation is based on coverage overlapping.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks AT330_TRR_E1 Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to % of covered area. The IF is user-definable using the Min importance and Max importance fields. d  Di  = 1 – ----------d max d is the effective distance (in m). In this case. this value varies between 0 to 100%. The importance of neighbours. Among these 15 candidate neighbours. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The IF evaluates importance as follows: Co-site Neighbourhood cause IF Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)} Yes Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+ Max(O)(O)/(Max(Di)+Max(O))} 60%+40%{1/7%(Di)+6/7%(O)} Where Delta(X)=Max(X)-Min(X) 333 . d max is the maximum distance between the reference transmitter and a possible neighbour. As indicated in the table below. see "Calculation of the Inter-Transmitter Distance" on page 319. Neighbourhood reason When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter If the Force co-site cells as neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions If the % minimum covered area is exceeded Importance Function (IF) Except the case of forced neighbours (importance = 100%). only 8 (having the highest importance values) will be allocated to the reference cell. the neighbour importance depends on the distance and on the neighbourhood cause. you have to select the max Ec/I0 option and define a high enough value. If defined there. It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. atl.2. Atoll displays a warning in the Event viewer indicating that the constraint on the forbidden neighbour will be ignored by algorithm because the neighbour already exists. the neighbours may be ranked differently. a neighbour may be marked as exceptional pair. Therefore. A new GSM TBC transmitter can enter the TBA cell neighbour list if allocation criteria are satisfied. • • • No prediction study is needed to perform an automatic neighbour allocation. Atoll determines the neighbour list of the cell i. neighbours will be ranked in this order: co-site neighbours and neighbours allocated based on coverage overlapping. With the default values for minimum and maximum importance fields. Atoll displays the percentage of area meeting the coverage conditions and the corresponding surface area (km2).3 Appendices 4. Therefore.1 Delete Existing Neighbours Option As explained above. Atoll automatically calculates the path loss matrices if not found. if you start a new allocation without selecting the Delete existing neighbours option. When starting an automatic neighbour allocation. 4. it indicates the importance (in %) of each neighbour and the allocation reason.0 Technical Reference Guide for Radio Networks Chapter 4: UMTS HSPA Networks • • • ©Forsk 2015 Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. if a TBA cell has already reached its maximum number of neighbours before starting the new allocation.7. If you change some allocation criteria (e. it examines the neighbour list of TBA cells and checks allocation criteria if there is space in their neighbour lists.Atoll 3.7. We assume that we have an existing allocation of inter-technology neighbours.2. In addition. A forbidden neighbour must not be listed as neighbour except if the neighbourhood relationship already exists and the Delete existing neighbours option is unchecked when you start the new allocation. the neighbours will be ranked by neighbour cause. it will not appear in the Results table.3. Atoll keeps the existing inter-technology neighbours when the Delete existing neighbours option is not checked. increase the maximum number of neighbours or create a new GSM TBC transmitter) and start a new allocation without selecting the Delete existing neighbours option. 334 . A new TBA cell i is created in UMTS. Atoll provides the list of neighbours. if cells have previous allocations in the list. the number of neighbours and the maximum number of neighbours allowed for each cell. neighbours are marked as existing. In this case. In the Results.3. Therefore. It will be the first one in the neighbour list. There can be a mix of the neighbourhood causes. For neighbours accepted for co-site and coverage reasons. co-site or coverage. Atoll displays only the cells for which it finds new neighbours.g. In the Results part. Finally. If the Min and Max value ranges of the importance function factors overlap. If the Min and Max value ranges of the importance function factors do not overlap. Chapter 5 CDMA2000 Networks This chapter covers the following topics: • "General Prediction Studies" on page 337 • "Definitions and Formulas" on page 340 • "Active Set Management" on page 358 • "Simulations" on page 358 • "CDMA2000 Prediction Studies" on page 392 • "Automatic Neighbour Allocation" on page 423 • "PN Offset Allocation" on page 430 • "Automatic GSM-CDMA Neighbour Allocation" on page 438 . Atoll 3.3.0 Technical Reference Guidefor Radio Networks © Forsk 2015 336 . Atoll 3. Atoll takes the highest power of both cells for each transmitter (i.1 Calculation Criteria Three criteria can be studied in point analysis (Profile tab) and in common coverage studies. L Tx is the transmitter loss ( L Tx = L total – DL ). and LTE Documents" on page 30. WiMAX. Study criteria are detailed in the table below: Study criteria Formulas Signal level ( P rec ) in dBm Signal level received from a transmitter on a carrier (cell) P rec  ic  = EIRP  ic  – L path – M Shadowing – model – L Indoor + G term – L term L path = L model + L ant Path loss ( L path ) in dBm Total losses ( L total ) in dBm Tx L total =  L path + L Tx + L term + L indoor + M Shadowing – model  –  G Tx + G term  where. EIRP  ic  = P pilot  ic  + G Tx – L Tx (where. G term is the receiver antenna gain.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 5 CDMA2000 Networks This chapter describes all the calculations performed in Atoll CDMA2000 documents. L Indoor are the indoor losses. For the definition of TBC transmitters please refer to "Path Loss Matrices" on page 103. 5. CDMA2000. EIRP is the effective isotropic radiated power of the transmitter. For CDMA2000 1xEV-DO systems. L ant Tx is the transmitter antenna attenuation (from antenna patterns).1 General Prediction Studies 5. M Shadowing – model is the shadowing margin. 337 . P pilot  ic  • is the cell pilot power). • For CDMA2000 1xRTT systems. Atoll considers that G term and L term equal zero. L term are the receiver losses. L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model. taken into account when the option “Indoor coverage” is selected. if the network consists of 1xRTT and 1xEV-DO carriers. TD-SCDMA.1. "UMTS. the highest value between the pilot power of the 1xRTT cell and the maximum power of the 1xEV-DO cell) to calculate the received signal level. When you make the prediction. G Tx is the transmitter antenna gain. you can consider the best carrier of all bands or the best carrier of a particular frequency band (Best (All Bands/Specific Band) option). ic is a carrier rank. EIRP  ic  = P max  ic  + G Tx – L Tx (where P max  ic  • • is the maximum cell power). All the calculations are performed on TBC (to be calculated) transmitters. Therefore.e. Atoll displays the best signal level received from a transmitter. This parameter is taken into account when the option “Shadowing taken into account” is selected.3. For information on calculating transmitter loss. In this case. each bin within the Txi calculation area is considered as a potential (fixed or mobile) receiver.1 Profile Tab Atoll displays either the signal level received from the selected transmitter on a carrier ( P rec  ic  ). • For a selected transmitter. it is also possible to study the path loss. L path .2 Best Signal Level and a Margin The service area of Txi corresponds to the bins where: Txi Txi Txi Minimum threshold  P rec  ic   or L total or L path   Maximum threshold And Txi Txj P rec  ic   Best  P rec  ic   – M ji M is the specified margin (dB). for example a smaller value for improving the calculation speed. 5. Reception bars are displayed in a decreasing signal level order. 5.2. The display settings to select how to colour service areas.3.1.3 Coverage Studies For each TBC transmitter. • For a selected transmitter. Best function: considers the highest value. ( P rec  ic  ).Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 5.3. see the Administrator Manual. For more information on defining a different value for this margin.1. or the total losses. • You can use a value other than 30 dB for the margin from the best server signal level. In fact.1.3.1 Service Area Determination Atoll uses parameters entered in the Condition tab of the coverage study property dialogue to predetermine areas where it will display coverage.2 Reception Tab Analysis provided in the Reception tab is based on path loss matrices. The maximum number of reception bars depends on the signal level received from the best server.3.1 All Servers The service area of Txi corresponds to the bins where: Txi Txi Txi Minimum threshold  P rec  ic   or L total or L path   Maximum threshold 5. Coverage study parameters to be set are: • • The study conditions in order to determine the service area of each TBC transmitter. 5. Only reception bars of transmitters whose signal level is within a 30 dB margin from the best server can be displayed. Path loss and total losses are the same on any carrier. or the highest signal level received on the best carrier. L total . it is also possible to study the path loss.1. We can distinguish three cases: 5. Path loss and total losses are the same on any carrier.2. or the total losses. So. For each transmitter.2 Point Analysis 5. L path . Txi. you can study reception from TBC transmitters for which path loss matrices have been computed on their calculation areas.1.1. Atoll displays either the signal level received on a carrier.1.1. 338 . or the highest signal level received from the selected transmitter on the best carrier.1. L total . Atoll determines the selected criterion on each bin inside the Txi calculation area. There are as many layers as defined thresholds.Atoll 3. Each layer shows the different signal levels available in the transmitter service area.3 If the margin equals 0 dB.1.1 Plot Resolution Prediction plot resolution is independent of the matrix resolutions and can be defined on a per study basis. There are as many layers as transmitter service areas. Each layer corresponds to an area where the signal level from the best server exceeds a defined minimum threshold. Atoll will consider bins where the signal level received from Txi is either the second highest or 2dB lower than the second highest.1. Where other service areas overlap the studied one. If the margin is set to -2 dB.2 Display Types It is possible to display the transmitter service area with colours depending on any transmitter attribute or other criteria such as: Signal Level (in dBm. Path Loss (dB) Atoll calculates path loss from the transmitter on each bin of each transmitter service area. Coverage consists of several independent layers whose visibility in the workspace can be managed. dBµV.3.3.2.1.1. Best Signal Level (in dBm. If the margin is set to 2 dB. There are as many layers as service areas. Each layer shows the different path loss levels in the transmitter service area. 5.1. Atoll chooses the highest value.3. 5. which are 2nd best servers. Atoll will consider bins where the signal level received from Txi is 2dB higher than the signal levels from transmitters. dBµV/m) Atoll calculates signal level received from the transmitter on each bin of each transmitter service area. Atoll will consider bins where the signal level received from Txi is either the highest or 2dB lower than the highest. 339 . Prediction plots are generated from multi-resolution path loss matrices using bilinear interpolation method (similar to the one used to evaluate site altitude).2.2 Coverage Display 5. A bin of a service area is coloured if the signal level is greater than or equal to the defined minimum thresholds (bin colour depends on signal level).3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 • • • 5. dBµV/m) Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. A bin of a service area is coloured if the signal level is greater than or equal to the defined thresholds (the bin colour depends on the signal level). If the margin is set to -2 dB. Atoll will consider bins where the signal level received from Txi is 2dB higher than the signal levels from transmitters. Coverage consists of several independent layers whose visibility in the workspace can be managed. If the margin is set to 2 dB. A bin of a service area is coloured if path loss is greater than or equal to the defined minimum thresholds (bin colour depends on path loss). • • • If the margin equals 0 dB. dBµV. Atoll will consider bins where the signal level received from Txi is the highest.3. Coverage consists of several independent layers whose visibility in the workspace can be managed. which are 3rd best servers. Second Best Signal Level and a Margin The service area of Txi corresponds to the bins where: Txi Txi Txi Minimum threshold  P rec  ic   or L total or L path   Maximum threshold And Txi nd Txj P rec  ic   2 Best  P rec  ic   – M ji M is the specified margin (dB). 2nd Best function: considers the second highest value. Atoll will consider bins where the signal level received from Txi is the second highest. Each layer corresponds to an area where the path loss from the best server exceeds a defined minimum threshold. quality targets.3. etc. Best Server Total Losses (dB) Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. the coverage corresponds to the pixels where the best signal level received fulfils signal conditions defined in Conditions tab. A bin of a service area is coloured if total losses is greater than or equal to the defined minimum thresholds (bin colour depends on total losses).1 Parameters Used for CDMA2000 1xRTT Modelling 5.Atoll 3.1 Inputs This table lists simulation and prediction inputs (calculation options. There is one coverage area per cell edge coverage probability in the explorer. A bin of a service area is coloured if the total losses is greater than or equal to the defined thresholds (bin colour depends on total losses). Where service areas overlap the studied one. active set management conditions. the coverage corresponds to the pixels where the signal level from this transmitter fulfils signal conditions defined in Conditions tab with different Cell edge coverage probabilities.) 340 Name Value Unit Description F ortho Clutter parameter None Orthogonality factor F MUD Tx Site equipment parameter None MUD factor cn first Frequency band parameter None First carrier number cn last Frequency band parameter None Last carrier number cn Frequency band parameter None Carrier number step . Best Server Path Loss (dB) Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. There are as many layers as defined thresholds. Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds.2.2. Each layer shows the different total losses levels in the transmitter service area. Coverage consists of several independent layers whose visibility in the workspace can be managed. 5. There is one coverage area per transmitter in the explorer. Number of Servers Atoll evaluates how many service areas cover a bin in order to determine the number of servers.2 Definitions and Formulas 5. There are as many layers as service areas. Each layer corresponds to an area where the number of servers is greater than or equal to a defined minimum threshold. Cell Edge Coverage Probability (%) On each bin of each transmitter service area. Atoll determines the best transmitter and evaluates total losses from the best transmitter. A bin of a service area is coloured if the path loss is greater than or equal to the defined thresholds (bin colour depends on path loss). Each layer corresponds to an area where the total losses from the best server exceed a defined minimum threshold. Best Cell Edge Coverage Probability (%) On each bin of each transmitter service area.1. Coverage consists of several independent layers whose visibility in the workspace can be managed. Where other service areas overlap the studied one. Coverage consists of several independent layers whose visibility in the workspace can be managed. Atoll determines the best transmitter and evaluates path loss from the best transmitter.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Total Losses (dB) Atoll calculates total losses from the transmitter on each bin of each transmitter service area. The bin colour depends on the number of servers. There are as many layers as defined thresholds. Mobility) parameter SCH – DL E b  --- N t req DL  Q req  SCH (Service. Mobility) parameter SCH – UL UL  Q req  SCH E b  --- N t req (Service. Ec/I0 . Terminal. Terminal. Mobility.3.Cell parameter min T_Drop . SCH throughput multiple) parameter Max Site parameter None Number of channel elements available for a site on uplink N CE –D L  N I  Max Site parameter None Number of channel elements available for a site on downlink N CE –U L  N I  Simulation result None Number of channel elements of a site consumed by users on uplink N CE –D L  N I  Simulation result None Number of channel elements of a site consumed by users on downlink Overhead Site equipment parameter None Number of channel elements used by the cell for common channels on uplink Overhead Site equipment parameter None Number of channel elements used by the cell for common channels on downlink N CE –U L  N I  N CE –U L N CE –D L 341 .Cell parameter Q pilot Q pilot  txi ic  req req min min Active set upper threshold None (used to determine the best server in the active set) None Active set lower threshold (used to determine other members of the active set) Minimum Ec/I0 required from the None cell to be the best server in the active set None Minimum Ec/I0 required from the cell not to be rejected from the active set Variation of the minimum Ec/I0 None required from the cell to be the best server in the active set req Delta Min. Terminal.Mobility parameter Q pilot min Delta T_Drop .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name Value Unit Frequency band parameter ic None Description Carrier rank of the current carrier calculated as follows: cn – cn first .– cn lower ic =  ------------------------ cn  Where cn lower is the number of carrier numbers lower than cn including excluded carriers and carriers of other frequency bands Q pilot req Q pilot  txi ic  + Q pilot min Q pilot  txi ic  + Q pilot Q pilot  txi ic  req Min.Atoll 3.Mobility parameter None Variation of the minimum Ec/I0 required from the cell not to be rejected from the active set RSCP min  Txi ic  Cell parameter or Global parameter W The minimum pilot RSCP required for a user to be connected to the transmitter on a given carrier None Eb/Nt target for FCH channel on downlink None Eb/Nt target for SCH channel on downlink None Eb/Nt target for FCH channel on uplink None Eb/Nt target for SCH channel on uplink Q pilot DL  Q req  FCH FCH – DL E b  --- N t req (Service. Ec/I0 . Mobility. SCH throughput multiple) parameter UL  Q req  FCH FCH – UL E b  --- N t req (Service. Terminal. 23 MHz Hz Spreading Bandwidth N CE –D L Max Tx DL Cell parameter None Inter-technology downlink noise rise NR inter – techno log y Cell parameter None Inter-technology uplink noise rise RF  ic ic adj  Network parameter If not defined. it is assumed that there is no inter-carrier interference None Interference reduction factor between two adjacent carriers ic NR inter – techno log y Tx UL Tx m ICP ic  ic i Network parameter If not defined.Atoll 3. site equipment) parameter None Number of channel elements used for FCH on downlink N Codes  txi ic  Simulation constraint None Maximum number of Walsh codes available per cell (128) N Codes  txi ic  Simulation result None Number of Walsh codes used by the cell NF term Terminal parameter None Terminal Noise Figure NF Tx Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) None Transmitter Noise Figure K 1.38 10-23 J/K Boltzman constant T 293 K Ambient temperature W 1.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Name Value Unit Description N CE –U L FCH (Terminal. site equipment) parameter None Number of channel elements used for FCH on uplink FCH (Terminal. it is assumed that there is no inter-technology downlink interferences due to external transmitters and ic adj Inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the None frequency gap between ic i (external network) and ic UL X max DL %Powermax % Maximum uplink load factor Simulation constraint (global parameter or cell parameter) % Maximum percentage of used power W Thermal noise at transmitter Tx UL Tx NF Tx  K  T  W  NR inter – techno log y Term NF Term  K  T  W  NR inter – techno log y W Thermal noise at terminal Rc W bps Chip rate f rake efficiency UL Equipment parameter DL Terminal parameter N0 N0 f rake efficiency SCH TPF DL FCH TPP – DL SCH TPP – DL SCH TPF UL FCH TP P – UL SCH TP P – UL 342 Simulation constraint (global parameter or cell parameter) Tx DL Simulation result Terminal parameter FCH SCH TP P – DL  TPF DL Simulation result Terminal parameter FCH SCH TPP – UL  TPF UL None Uplink rake receiver efficiency factor None Downlink rake receiver efficiency factor SCH throughput factor (drawn None following the SCH probabilities of the service) bps Downlink FCH peak throughput bps Downlink SCH bit rate SCH throughput factor (drawn None following the SCH probabilities of the service) bps Uplink FCH peak throughput bps Uplink SCH bit rate .3. Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name Value Unit Description W ----------------FCH TP P – DL None Downlink service processing gain on FCH W ----------------SCH TP P – DL None Downlink service processing gain on SCH W ----------------FCH TP P – UL None Uplink service processing gain on FCH W ----------------SCH TP P – UL None Uplink service processing gain on SCH DL Service parameter None Downlink activity factor on FCH AF FCH UL Service parameter None Uplink activity factor on FCH P Sync  txi ic  Cell parameter W Cell synchronisation channel power P paging  txi ic  Cell parameter W Cell other common channels (except CPICH and SCH) power P pilot  txi ic  Cell parameter W Cell pilot power P max  txi ic  Cell parameter W Maximum cell power M pooling  txi ic  Cell parameter dB Maximum amount of power reserved for pooling P FCH min Service parameter W Minimum power allowed for FCH P FCH max Service parameter W Maximum power allowed for FCH P SCH min Service parameter W Minimum power allowed for SCH P SCH max Service parameter W Maximum power allowed for SCH P FCH  txi ic tch  Simulation result including the term AF FCH  Serv  W Cell FCH power for a traffic channel on carrier ic W Total FCH power on carrier ic Simulation result W Transmitter SCH power for a traffic channel on carrier ic  W Total SCH power on carrier ic W Transmitter total transmitted power on carrier ic FCH – DL Gp SCH – DL Gp FCH – UL Gp SCH – UL Gp AF FCH P FCH  txi ic  DL  P FCH  txi ic tch  tch  FCH  ic   P SCH  txi ic tch  P SCH  txi ic  P SCH  ic tch  tch  SCH  ic   P tx  txi ic  P pilot  txi ic  + P Sync  txi ic  + P paging  txi ic  + P SCH  txi ic  + P FCH  txi ic  P term min Terminal parameter W Minimum terminal power allowed max Terminal parameter W Maximum terminal power allowed P term FCH Simulation result including the term AF FCH  Serv  W Terminal FCH power transmitted in carrier ic P term  ic  SCH Simulation result W Terminal SCH power transmitted on carrier ic  BTS BTS parameter % Percentage of BTS signal correctly transmitted  term Terminal parameter % Percentage of terminal signal correctly transmitted P term  ic  UL 343 .3. frequency band) parameter L path Propagation model result None Path loss f Terminal parameter None Number of fingers p Terminal parameter % Pilot power percentage M Shadowing – model Result calculated from cell edge coverage probability and model standard deviation None Model Shadowing margin Only used in prediction studies M Shadowing – Ec  Io Result calculated from cell edge coverage probability and Ec/I0 standard deviation None Ec/I0 Shadowing margin Only used in prediction studies None DL gain due to availability of several pilot signals at the mobile b. DL M Shadowing –  Eb  Nt  npaths G macro – diversity = M Shadowing – Ec  Io – M Shadowing –Ec  Io DL G macro – diversity M Shadowing –  Eb  Nt  Indoor loss n=2 or 3 DL Result calculated from cell edge coverage probability and DL Eb/Nt standard deviation None DL Eb/Nt Shadowing margin Only used in prediction studies UL Result calculated from cell edge coverage probability and UL Eb/Nt standard deviation None UL Eb/Nt Shadowing margin Only used in prediction studies None UL quality gain due to signal diversity in soft handoffc.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Name Value Unit Description  Clutter parameter % Percentage of pilot finger percentage of signal received by the terminal pilot finger G Tx Antenna parameter None Transmitter antenna gain G Term Terminal parameter None Terminal gain L Tx Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) None Transmitter lossa L body Service parameter None Body loss L Term Terminal parameter None Terminal loss L indoor Clutter (and.Atoll 3. None Random shadowing error drawn during Monte-Carlo simulation Only used in simulations None Transmitter-terminal total loss P pilot  txi ic  ------------------------------LT W Chip power received at terminal P FCH  txi ic tch  ----------------------------------------LT W Bit received power at terminal for FCH on carrier ic UL UL G macro – diversity E Shadowing npaths G macro – diversity = M Shadowing –  Eb  Nt  UL – M Shadowing –  Eb  Nt  n=2 or 3 Global parameter (default value) Simulation result UL In prediction studiesd For Ec/I0 calculation L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io ------------------------------------------------------------------------------------------------------------------------------------G Tx  G term LT For DL Eb/Nt calculation L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  DL ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term For UL Eb/Nt calculation L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term In simulations L path  L Tx  L term  L body  L indoor  E Shadowing -------------------------------------------------------------------------------------------------------------------G Tx  G term P c  txi ic  FCH – DL Pb 344  txi ic tch  .3. optionally. 2. and LTE Documents" on page 30. Therefore. M Shadowing –  Eb  Nt  d. M Shadowing –  Eb  Nt  DL DL or or M Shadowing – Ec  Io is set to 1 in downlink extra-cell interference calculation. b. Name Value I intra  txi ic  P tot  txi ic  DL DL  DL I extra  ic  DL P tot  txj ic  Unit Description W Downlink intra-cell interference at terminal on carrier ic W Downlink extra-cell interference at terminal on carrier ic W Downlink inter-carrier interference at terminal on carrier ic txj j  i  Ptot  txj icadj  DL DL I inter – carrier  ic  DL I inter – techno log y  ic  txj j --------------------------------------------RF  ic ic adj   ni DL I 0  ic  Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic W Downlink inter-technology interference at terminal on carrier ic a i Term DL DL DL DL I intra  txi ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 W Total received noise at terminal on carrier ic b 345 .  txi ic tch  p  Ptot  ic  L Tx = L total – UL on uplink and L Tx = L total – DL on downlink. npaths M Shadowing –Ec  Io corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in case of downlink Ec/I0 modelling. In uplink prediction studies. CDMA2000. TD-SCDMA. see "UMTS. npaths c.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name SCH – DL Pb  txi ic tch  FCH – DL DL P b  txi ic tch  Pb Value Unit Description P SCH  txi ic tch  ----------------------------------------LT W Bit received power at terminal for SCH on carrier ic W Bit received power at terminal for FCH+SCH on carrier ic W Total received power at terminal from a transmitter on carrier ic W Total power received at terminal from traffic channels of a transmitter on carrier ic W Bit received power at transmitter for FCH on carrier ic W Bit received power at transmitter for SCH on carrier ic W Bit received power at transmitter for SCH+FCH on carrier ic W Total power transmitted by the terminal on carrier ic W Chip received power at transmitter SCH – DL  txi ic tch  + P b P tx  txi ic  -------------------------LT DL P tot  txi ic   DL P traf  txi ic  tch  ic  P FCH  txi ic  + P SCH  txi ic  -----------------------------------------------------------------LT FCH FCH – UL  ic  P term -----------LT SCH – UL  ic  P term -----------LT Pb Pb SCH FCH – UL UL P b  ic  Pb SCH – UL  ic  + P b  ic  UL P b  ic  UL UL P b  ic  + P c  ic  = ---------------1 – p UL P tot  ic  UL UL P c  ic  a.1. For information on calculating transmitter losses on uplink and downlink.2 Ec/I0 Calculation This table details the pilot quality ( Q pilot or Ec  Io ) calculations. case of uplink soft handoff modelling. carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing –  Eb  Nt  M Shadowing – Ec  Io ) while extra-cell interference level is not.3. 5. In downlink prediction studies. only carrier power level is downgraded by the shadowing margin ( M Shadowing –  Eb  Nt  UL corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in UL ). WiMAX. the ICP value is weighted according to the fractional load. DL DL  G SHO  FCH Q FCH  ic  ---------------------------------------------------DL Q FCH  BestServer ic  None Downlink soft handover gain for FCH channel on carrier ic None Downlink soft handover gain for SCH channel on carrier ic DL DL  G SHO  SCH 346 Q SCH  ic  ---------------------------------------------------DL Q SCH  BestServer ic  Quality level at terminal on a traffic channel from one transmitter for a SCH channel on carrier icc . Without useful signal: SCH – DL Pb  txi E DL DL Q SCH  txi ic    ----b-  N t SCH DL Q SCH  ic   BTS  ic tch  – DL -----------------------------------------------------------------------------------------------------.Atoll 3. G FCH Total noise: -----------------------------------------------------------------p DL N tot  ic  DL f rake efficiency   DL Q FCH  tx k ic  tx k  ActiveSet  FCH  Quality level at terminal for FCH using carrier ic due to combination of None all transmitters of the active set (Macro-diversity conditions).3 DL Eb/Nt Calculation Eb DL This table details calculations of downlink traffic channel quality ( Q tch (tch could be FCH or SCH) or  ------ ). G SCH p DL DL N tot  ic  –  1 – F ortho    BTS  P b  txi ic  None SCH – DL  BTS  P b  txi ic tch  – DL . In an active set. G FCH p DL DL N tot  ic  –  1 – F ortho    BTS  P b  txi ic  FCH – DL  BTS  P b  txi ic tch  – DL .1. In the case of an interfering GSM external network in frequency hopping.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Name Value Unit Description E Q pilot  txi ic    ----c  I0   BTS    P c  txi ic  -------------------------------------------------DL I 0  ic  None Quality level at terminal on pilot for carrier ic a.2. G SCH Total noise: -----------------------------------------------------------------p DL N tot  ic  DL f rake efficiency   DL Q SCH  tx k ic  tx k  ActiveSet  SCH  Quality level at terminal for SCH using carrier ic due to combination of None all transmitters of the active set (Macro-diversity conditions). N 0 Term is calculated for all its members with Inter-technology downlink noise rise of the best server.  Nt DL Name Value Unit Description I intra  txi ic   1 –  BTS  F ortho   P DL  txi ic  tot W Downlink intra-cell interference at terminal on carrier ic W Downlink extra-cell interference at terminal on carrier ic W Downlink inter-carrier interference at terminal on carrier ic DL  DL I extra  ic  DL P tot  txj ic  txj j  i  Ptot  txj icadj  DL DL I inter – carrier  ic  txj  j --------------------------------------------- RF  ic ic adj   DL I inter – techno log y  ic  ni DL N tot  ic  DL DL Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic W a i DL Downlink inter-technology interference at terminal on carrier ic Term DL I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 W Total received noise at terminal on carrier ic None Quality level at terminal on a traffic channel from one transmitter for a FCH channel on carrier ic b Without useful signal: FCH – DL Pb  txi DL Q FCH  txi E DL ic    ----b-  N t FCH DL Q FCH  ic   BTS  ic tch  – DL -----------------------------------------------------------------------------------------------------. 5. b.3. Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt. the ICP value is weighted according to the fractional load. req P FCH  txi ic  + P SCH  txi ic  In the case of an interfering GSM external network in frequency hopping. In point analysis and coverage studies. In point analysis and coverage studies. The chosen option will be taken into account only in simulations. 5. The chosen option will be taken into account only in simulations.1.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name Value Unit Description  Q req  FCH ----------------------. Calculation option may be selected in the Global parameters tab.3. P FCH  txi ic  DL Q FCH  ic  W Required transmitter FCH traffic channel power to achieve Eb/Nt target at terminal on carrier ic W Required transmitter SCH traffic channel power to achieve Eb/Nt target at terminal on carrier ic W Required transmitter traffic channel power on carrier ic DL req P FCH  txi ic  DL  Q req SCH ----------------------. G FCH p UL Tx UL N tot  txi ic  –  1 – F MUD    term  P b  ic  FCH – UL  term  P b  ic  FCH – UL Total noise: ----------------------------------------------- Gp UL N tot  txi ic  Without useful signal: SCH – UL E UL Q SCH  txi ic    ----b- N t UL  term  P b  ic  – UL -------------------------------------------------------------------------------------------------------. G SCH p UL Tx UL N tot  txi ic  –  1 – F MUD    term  P b  ic  SCH – UL  term  P b  ic  SCH – UL Total noise: ----------------------------------------------- Gp UL N tot  txi ic  347 . b.  Nt UL Name Value   Pb UL UL intra I tot  txi UL extra I tot ic   UL term txj j  i UL UL I inter – carrier  txi ic  W Total power received at transmitter from intra-cell terminals using carrier ic W Total power received at transmitter from extra-cell terminals using carrier ic W Uplink inter-carrier interference at terminal on carrier ic UL  P b  ic  + P c  ic     Pb Description UL  ic  + P c  ic   term txi  txi ic  Unit UL  ic adj  + P c  ic adj   term txj j ---------------------------------------------------------------------- RF  ic ic adj  UL I tot  txi ic  UL N tot  txi ic  UL extra I tot UL intra Tx  txi ic  +  1 – F MUD   term I tot UL UL W  txi ic  +I inter – carrier  txi ic tx I tot  txi ic  + N 0 Total received interference at transmitter on carrier ic W Total noise at transmitter on carrier ic (Uplink interference) a None Quality level at transmitter on a traffic channel for the FCH channel on carrier icb None Quality level at transmitter on a traffic channel for the SCH channel on carrier icc Without useful signal: FCH – UL E UL Q FCH  txi ic    ----b-  N t UL  term  P b  ic  – UL -------------------------------------------------------------------------------------------------------. Calculation option may be selected in the Global parameters tab.2.Atoll 3. c. P SCH  txi ic  DL Q SCH  ic  req P SCH  txi ic  req req P tch  txi ic  a. Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt.4 UL Eb/Nt Calculation Eb UL This table details calculations of uplink traffic channel quality ( Q tch (tch could be FCH or SCH) or  ------ ). The chosen option will be taken into account only in simulations. The chosen option will be taken into account only in simulations.tx l  ActiveSet  f rake efficiency    tx k  samesite   tx k Max Quality level at site using carrier ic due to combination of all transmitters of the active set located at the same site and taking into account increase of the quality due None to macro-diversity (macro-diversity gain). In an active set. Calculation option may be selected in the Global parameters tab. P SCH term  ic  UL Q SCH  ic  P term UL P term FCH – req req P term  ic  P term SCH – req  ic  + P term  ic  tx a.Atoll 3. In point analysis and coverage studies. N 0 is calculated for all its members with Inter-technology uplink noise rise of the best server. Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt. In point analysis and coverage studies.3. UL  G macro – diversity UL Q FCH  ic  ---------------------------------------------------UL Q FCH  BestServer ic  UL  G SHO  FCH None Uplink soft handover gain for FCH channel on carrier ic None Uplink soft handover gain for SCH channel on carrier ic W Required terminal power to achieve Eb/Nt target at transmitter for FCH on carrier ic W Required terminal power to achieve Eb/Nt target at transmitter for SCH on carrier ic W Required terminal power on carrier ic UL Q SCH  ic  ---------------------------------------------------UL Q SCH  BestServer ic  UL  G SHO  SCH UL FCH – req  ic   Q req  FCH ---------------------. c. Calculation option may be selected in the Global parameters tab. P FCH term  ic  UL Q FCH  ic  SCH – req  ic   Q req  SCH ----------------------. tch could be FCH or SCH  In simulations.2.5 Simulation Results This table contains some simulation results provided in the Cells and Mobiles tabs of the simulation property dialogue.1. Softer/Soft HO (No MRC): Max  Q UL tch  tx k tx k  ActiveSet UL ic    G macro – diversity Softer/Soft HO (MRC): UL Q tch  ic     UL  UL UL Q tch  tx k ic  Q tch  tx l ic  tx k .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Name Value Unit Description UL No HO: Q tch  txi ic   UL Softer HO: f rake efficiency  UL Q tch  txk ic  tx k  ActiveSet  samesite  Soft. Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt. 5. Name DL I intra  txi ic  Value Unit Description DL DL P tot  txi ic  – F ortho   BTS  P tot  txi ic  None Downlink intra-cell interference at terminal on carrier ic W Downlink extra-cell interference at terminal on carrier ic DL –  1 – F ortho   BTS   P b  txi ic  DL I extra  ic  348  txj j  i DL P tot  txj ic  . UL G macro – diversity tx  othersite l = 1. b. Atoll 3. 100  P max  txi ic  None Percentage of max transmitter power used. None Downlink load factor on carrier ic None Downlink reuse factor on a carrier ic F  txi ic  %Power  txi ic  Simulation result available per cell DL  I extra  ic  DL + I inter – carrier  ic    L T --------------------------------------------------------------------------------.+ 1 – F ortho   BTS P tx  txi ic  --------------------------------------------------------------------------------------------------------------------------------1 .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name Value Unit Description  Ptot  txj icadj  W Downlink inter-carrier interference at terminal on carrier ic DL DL I inter – carrier  ic  txj j --------------------------------------------- RF  ic ic adj  Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic  DL I inter – techno log y  ic  ni DL DL I tot  ic  DL DL DL DL DL Term I tot  ic  + N 0   Pb UL I tot  txi ic  UL extra I tot  txi ic  UL term txj j  i   Pb ic  Total effective interference at terminal on carrier ic (after unscrambling) W Total received noise at terminal on carrier ic W Total power received at transmitter from intra-cell terminals using carrier ic W Total power received at transmitter from extra-cell terminals using carrier ic W Uplink inter-carrier interference at terminal on carrier ic UL  P b  ic  + P c  ic   UL UL I inter – carrier  txi W UL  ic  + P c  ic   term txi  Downlink inter-technology interference at terminal on carrier ic a i I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  N tot  ic  UL intra W UL  ic adj  + P c  ic adj   term txj j ---------------------------------------------------------------------- RF  ic ic adj  UL I tot  txi ic  UL extra I tot UL Tx intra  txi ic  +  1 – F MUD   term I tot UL UL N tot  txi ic  UL W  txi ic  +I inter – carrier  txi ic tx I tot  txi ic  + N 0 Total received interference at transmitter on carrier ic W Total noise at transmitter on carrier ic (Uplink interference) None Cell uplink load factor on carrier ic UL I tot  txi ic  ---------------------------UL N tot  txi ic  UL X  txi ic  UL UL I tot  txi ic  --------------------------------------------------------------------------------------UL intra Tx I tot  txi ic    1 – F MUD   term  None Cell uplink reuse factor on carrier ic E  txi ic  UL 1 -------------------------UL F  txi ic  None Cell uplink reuse efficiency factor on carrier ic DL P tx  txi ic    ----------------------------.3.+ 1 – F ---------tch ortho   BTS  DL CI req  DL X  txi ic  with DL CI req SCH – DL FCH – DL Q req Q req = -------------------+ -------------------SCH – DL FCH – DL Gp Gp DL I tot  ic  Simulation result available per mobile: -----------------DL N tot  ic  DL DL F  txi ic  I tot  ic  ----------------------------DL I intra  txi ic  349 . Unit In the case of an interfering GSM external network in frequency hopping.1 Inputs This table lists simulation and prediction inputs (calculation options.Cell parameter Q pilot Q pilot  txi ic  req min min Active set upper threshold None (used to determine the best server in the active set) None Active set lower threshold (used to determine other members of the active set) Minimum Ec/I0 required from the None cell to be the best server in the active set None Minimum Ec/I0 required from the cell not to be rejected from the active set Variation of the minimum Ec/I0 None required from the cell to be the best server in the active set req Delta Min. etc. 5.Mobility parameter Q pilot min Delta T_Drop .2. the ICP value is weighted according to the fractional load. quality targets. active set management conditions. 0 users None Minimum pilot quality required in the uplink to operate EV-DO Rev.2.) Name Value Unit Description F ortho Clutter parameter None Orthogonality factor F MUD Tx Site equipment parameter None MUD factor cn first Frequency band parameter None First carrier number cn last Frequency band parameter None Last carrier number cn Frequency band parameter None Carrier number step ic Frequency band parameter None Carrier rank of the current carrier calculated as follows: cn – cn first .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Name Value Description DL dB Noise rise on downlink UL dB Noise rise on uplink DL – 10 log  1 – X  txi ic   UL – 10 log  1 – X  txi ic   NR  txi ic  NR  txi ic  a.Mobility parameter None Variation of the minimum Ec/I0 required from the cell not to be rejected from the active set RSCP min  Txi ic  Cell parameter or Global parameter W The minimum pilot RSCP required for a user to be connected to the transmitter on a given carrier Ec  --- N t min – Rev0 Mobility parameter for 1xEV-DO Rev.2. Ec/I0 .Cell parameter min T_Drop . 0 Ec  --- N t min – RevB Transmitter parameter None Minimum pilot quality required in the uplink to operate multi-carrier EV-DO Q pilot UL UL 350 req .2 Parameters Used for CDMA2000 1xEV-DO Modelling 5.Atoll 3. Ec/I0 .3.– cn lower ic =  ------------------------ cn  Where cn lower is the number of carrier numbers lower than cn including excluded carriers and carriers of other frequency bands Q pilot req Q pilot  txi ic  + Q pilot min Q pilot  txi ic  + Q pilot Q pilot  txi ic  req Min. Atoll 3. B users None Minimum pilot quality level required to obtain a radio bearer in the downlink n TS 1xEV-DO Radio Bearer Selection (Downlink) table None Number of timeslots associated with the 1xEV-DO radio bearer in the downlink DL Downlink 1xEV-DO Radio Bearer Table None Downlink RLC peak throughput provided by the 1xEV-DO radio bearer N EVDO – CE  N I  Site parameter None Number of EVDO channel elements available for a site on uplink and downlink N EVDO – CE  N I  Simulation result None Total number of EVDO channel elements of a site consumed by users on uplink and downlink N CE – UL TCH (Terminal.23 MHz Hz Spreading Bandwidth NRinter – techno log y Cell parameter None Inter-technology downlink noise rise NRinter – techno log y Cell parameter None Inter-technology uplink noise rise RF  ic ic adj  Network parameter If not defined.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name Ec  --- N t min UL Value Unit Description Parameter read in the 1xEV-DO Radio Bearer Selection (Uplink) table Minimum pilot quality level required None for 1xEV-DO Rev. A and Rev. A and Rev. site equipment) parameter None Number of channel elements used for TCH on uplink N MacIndexes  txi ic  Simulation constraint None Maximum number of MAC indexes available per cell (59 for Rev0 and 114 for RevA) N MacIndexes  txi ic  Simulation result None Number of MAC indexes used by the cell n EVDO  txi ic  Simulation constraint (cell parameter) None Maximum number of EVDO users that can be connected to the cell n EVDO  txi ic  Simulation result None Number of EVDO users connected to the cell NF term Terminal parameter None Terminal Noise Figure NF Tx Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) None Transmitter Noise Figure K 1. it is assumed that there is no inter-carrier interference None Interference reduction factor between two adjacent carriers ic DL TP P – R LC Max Max Max Tx DL Tx UL Tx m ICP ic  ic i Network parameter If not defined.3.38 10-23 J/K Boltzman constant T 293 K Ambient temperature W 1. it is assumed that there is no inter-technology downlink interferences due to external transmitters and ic adj Inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the None frequency gap between ic i (external network) and ic UL X max Simulation constraint (global parameter or cell parameter) % Maximum uplink load factor 351 . B users to obtain a radio bearer in the uplink n SF 1xEV-DO Radio Bearer Selection (Uplink) table Number of subframes associated None with the 1xEV-DO radio bearer in the uplink TP P – R LC UL 1xEV-DO Radio Bearer Selection (Uplink) table None Uplink RLC peak throughput provided by the 1xEV-DO radio bearer Ec  --- N t min Parameter read in the 1xEV-DO Radio Bearer Selection (Downlink) table for 1xEV-DO Rev. 3.Guaranteed Bit Rate) service user Gp W---------UL TP None Uplink service processing gain on FCH G idle – power Cell parameter None Idle power gain G MU Cell parameter None Multi user gain P max  txi ic  Cell parameter W Max cell power P tx  txi ic b pilot  P max  txi ic  W Pilot burst transmitted by the transmitter on carrier ic.Atoll 3. TP UL DL DL UL P tx  txi ic b traffic  352 ©Forsk 2015 P max  txi ic  if users to support P max  txi ic   G idle – power if no user to support ER DRC Cell parameter % Error rate on the DRC channel TS BCMCS Cell parameter % Pourcentage of EVDO timeslots dedicated to Broadcast/Multicast services TS EVDO – CCH Cell parameter % Pourcentage of EVDO timeslots dedicated to control channels . A .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks Name Value Tx UL Unit Description W Thermal noise at transmitter Tx NF Tx  K  T  W  NR inter – techno log y Term NF Term  K  T  W  NR inter – techno log y W Thermal noise at terminal Rc W bps Chip rate f rake efficiency Equipment parameter N0 N0 UL UL Tx DL None Uplink rake receiver efficiency factor Simulation result bps Uplink throughput TP TCP – ACK Simulation result bps Uplink throughput due to TCP aknowledgements TPBCMCS Cell parameter bps Downlink throughput for Broadcast/ Multicast services TP max – DL Simulation result bps Downlink peak throughput supplied to the terminal TP avg Simulation result bps Downlink average cell throughput TPD min – UL Service parameter kbps Minimum required bit rate that the service should have in order to be available in the uplink TPD min – DL Service parameter kbps Minimum required bit rate that the service should have in order to be available in the downlink TPA f TP – Scaling  TP max – DL – TP Offset bps Downlink user application throughput f TP – Scaling Service parameter % Scaling factor TPOffset Service parameter kbps Offset C DL – Bearer TPD min – DL -----------------------------------------------------------DL TP P – R LC  Index DL – Bearer  % Downlink radio bearer consumption for a (1xEV-DO Rev. A .Guaranteed Bit Rate) service user C UL – Bearer TPD min – UL -----------------------------------------------------------UL TP P – R LC  Index UL – Bearer  % Uplink radio bearer consumption for a (1xEV-DO Rev. W Traffic burst transmitted by the transmitter on carrier ic. optionally. frequency band) parameter L path Propagation model result None Path loss G ACK Terminal parameter None Acknowledgement Channel gain G RRI Terminal parameter (for 1xEV-DO Rev A terminals only) None Reverse Rate Indicator Channel gain G DRC Terminal parameter None Data Rate Control Channel gain G Auxiliary – pilot Terminal parameter (for 1xEV-DO Rev A terminals only) None Auxiliary Pilot Channel gain G TCH Terminal parameter None Traffic data Channel gain carriers Terminal parameter None Maximum number of carriers in multi-carrier mode M Shadowing – model Result calculated from cell edge coverage probability and model standard deviation None Model Shadowing margin Only used in prediction studies M Shadowing – Ec  Io Result calculated from cell edge coverage probability and Ec/I0 standard deviation None Ec/I0 Shadowing margin Only used in prediction studies None DL gain due to availability of several pilot signals at the mobile b.3.Atoll 3. None UL Eb/Nt Shadowing margin Only used in prediction studies None UL quality gain due to signal diversity in soft handoffc. None Random shadowing error drawn during Monte-Carlo simulation Only used in simulations n max DL n=2 or 3 UL Result calculated from cell edge coverage probability and UL Eb/Nt standard deviation UL UL G macro – diversity E Shadowing npaths G macro – diversity = M Shadowing – Ec  Io – M Shadowing –Ec  Io DL G macro – diversity M Shadowing –  Eb  Nt  Indoor loss npaths G macro – diversity = M Shadowing –  Eb  Nt  UL – M Shadowing – Eb  Nt  n=2 or 3 Global parameter (default value) Simulation result UL 353 .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name Value Unit Description P term  ic  Simulation result W Terminal power transmitted on carrier ic P term min Terminal parameter W Minimum terminal power allowed P term max Terminal parameter W Maximum terminal power allowed  BTS BTS parameter % Percentage of BTS signal correctly transmitted  term Terminal parameter % Percentage of terminal signal correctly transmitted  Clutter parameter % Percentage of pilot finger percentage of signal received by the terminal pilot finger G Tx Antenna parameter None Transmitter antenna gain G Term Terminal parameter None Terminal gain L Tx Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) None Transmitter lossa L body Service parameter None Body loss L Term Terminal parameter None Terminal loss L indoor Clutter (and. 2 Ec/I0 and Ec/Nt Calculations E E E This table details ----c  txi ic b pilot  . npaths c. 5. txi ic b traffic  calculations. txi ic b pilot  and ----c.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Name Value Unit Description None Transmitter-terminal total loss P tx  txi ic b pilot  ----------------------------------------LT W Pilot burst received at terminal from a transmitter on carrier ic P tx  txi ic b traffic  --------------------------------------------LT W Traffic burst received at terminal from a transmitter on carrier ic P b  ic  P term -----------LT W Bit received power at transmitter on carrier ic NRthreshold  txi ic  Cell parameter dB Cell uplink noise rise threshold Cell parameter dB Cell uplink noise rise upgrading/ downgrading delta In prediction studiesd For Ec/I0 and Ec/Nt calculations L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io ------------------------------------------------------------------------------------------------------------------------------------G Tx  G term For UL Eb/Nt calculation L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term LT In simulations L path  L Tx  L term  L body  L indoor  E Shadowing -------------------------------------------------------------------------------------------------------------------G Tx  G term DL P tot  txi ic b pilot  DL P tot  txi ic b traffic  UL UL UL NR threshold  txi ic  a.3. only carrier power level is downgraded by the shadowing margin ( M Shadowing –  Eb  Nt  UL corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in UL ). In downlink prediction studies. M Shadowing – Ec  Io is set to 1 in downlink extra-cell interference calculation.2. ----c. M Shadowing –Ec  Io corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in case npaths of downlink Ec/I0 modelling. I0 Nt Nt Name Value Unit Description  txi ic DL  I intra    b pilot or b traffic  0 W Downlink intra-cell interference at terminal on carrier ic (only one mobile is served at a time) W Downlink extra-cell interference based on pilot at terminal on carrier ic DL I extra  ic b pilot  DL I extra  ic b traffic   P tot  txj ic b pilot   P tot  txj ic b traffic  W Downlink extra-cell interference based on traffic at terminal on carrier ic  Ptot  txj icadj bpilot  W Downlink inter-carrier interference based on pilot at terminal on carrier ic DL txj j  i DL txj j  i DL DL I inter – carrier  ic 354 b pilot  txj j ------------------------------------------------------------RF  ic ic adj  . M Shadowing –  Eb  Nt  d. Therefore. carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing – Ec  Io ) while extra-cell interference level is not. L Tx = L total – UL on uplink and L Tx = L total – DL on downlink. In uplink prediction studies.2. case of uplink soft handoff modelling.Atoll 3. b. 3 UL Eb/Nt Calculation This table details calculations of uplink quality ( Q Name intra I tot  txi UL extra I tot  txi Eb or  ------ ). 5. txi ic b pilot  I0 Downlink inter-technology interference at terminal on carrier ic a i P tot  txi ic b pilot  + I extra  ic b pilot  + I inter – carrier  ic b pilot  DL I 0  ic b pilot  a.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name Value Unit Description  Ptot  txj icadj btraffic  W Downlink inter-carrier interference based on traffic at terminal on carrier ic DL DL I inter – carrier  ic b traffic  txj j ---------------------------------------------------------------RF  ic ic adj   DL I inter – techno log y  ic  ni DL W DL + DL W Total noise based on pilot received at terminal on carrier ic W Total noise based on traffic received at terminal on carrier ic I extra  ic b pilot  + N 0 W Total noise based on pilot received at terminal on carrier ic DL W Total noise based on traffic received at terminal on carrier ic None Pilot quality level at terminal on carrier ic None Pilot quality level at terminal on carrier ic None Traffic quality level at terminal on carrier ic DL I inter – techno log y  ic  DL + DL term N0 DL P tot  txi ic b traffic  + I extra  ic b traffic  + I inter – carrier  ic b traffic  DL I 0  ic b traffic  + DL DL I inter – techno log y  ic  DL N tot  ic b pilot  DL N tot  ic b traffic  + term N0 term term I extra  ic b traffic  + N 0 Q pilot  txi ic  DL  BTS    P tot  txi ic b pilot  ---------------------------------------------------------------------DL I 0  ic b pilot  Ec  ---.3.2. txi ic b pilot  Nt  BTS    P tot  txi ic b pilot  ---------------------------------------------------------------------------------------------------------------DL DL N tot  ic b pilot  +  1 –  BTS   P tot  txi ic b pilot  E ----c.  Nt UL Value  Pb UL UL UL ic  term txi  ic  term txj j  i  Pb ic  Description W Total power received at transmitter from intra-cell terminals using carrier ic W Total power received at transmitter from extra-cell terminals using carrier ic W Uplink inter-carrier interference at terminal on carrier ic UL P b  ic  UL UL I inter – carrier  txi  ic  Unit  ic adj  term txj j ----------------------------------- RF  ic ic adj  UL I tot  txi ic  UL N tot  txi ic  UL extra I tot UL intra Tx  txi ic  +  1 – F MUD   term I tot UL tx I tot  ic  + N 0 UL W  txi ic  +I inter – carrier  txi ic W Total received interference at transmitter on carrier ic Total noise at transmitter on carrier ic (Uplink interference) 355 .2. the ICP value is weighted according to the fractional load. Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic DL E ----c. txi ic b traffic  Nt  BTS    P tot  txi ic b traffic  ----------------------------------------------------------------------------------------------------------------------DL DL N tot  ic b traffic  +  1 –  BTS   P tot  txi ic b traffic  DL In the case of an interfering GSM external network in frequency hopping.Atoll 3. Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt.3.Atoll 3.  In simulations. tx l  othersite UL  G macro – diversity UL Q total  ic  ------------------------------------------------UL Q  BestServer ic  UL G SHO None Uplink soft handover gain on carrier ic None Eb/Nt target on uplink W Required terminal power to achieve Eb/Nt target at transmitter on carrier ic For 1xEV-DO Rev 0 terminal UL E UL  ----c-  G p   1 + G ACK + G DRC + G TCH   N t min For 1xEV-DO Rev A terminalb When the acknoledgement signal is considered UL Q req UL Ec  UL  --- G p   1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary – Pilot   N t min When the acknoledgement signal is not considered UL E UL  ----c-  G p   1 + G RRI + G DRC + G TCH + G Auxiliary – Pilot   N t min UL Q req ---------------------.4 Simulation Results This table contains some simulation results provided in the Cells and Mobiles tabs of the simulation property dialogue.2. The chosen option will be taken into account only in simulations. b. P term UL Q total  ic  req P term  ic  a. Calculation option may be selected in the Global parameters tab. In point analysis and coverage studies.tx  ActiveSet  f rake efficiency  k l   tx k  samesite   tx k Max Quality level at site using carrier ic due to combination of all transmitters of the active set located at the same site and taking into account increase of the quality due None to macro-diversity (macro-diversity gain). the uplink Eb/Nt target is calculated without considering the acknowledgement signal. G UL p UL Tx UL N tot  txi ic  –  1 – F MUD    term  P b  ic  E UL Q  txi ic    ----b-  N t UL UL  term  P b  ic  UL Total noise: ----------------------------------- Gp UL N tot  txi ic  UL No HO: Q  txi ic   UL Softer HO: f rake efficiency  UL Q tch  txk ic  tx k  ActiveSet  samesite  Soft. 5. Softer/Soft HO (No MRC): Max  Q UL tch  tx k tx k  ActiveSet ic    UL G macro – diversity Softer/Soft HO (MRC): UL Q total  ic     UL  UL UL Q tch  tx k ic  Q tch  tx l ic  tx . 356 . In simulations. UL G macro – diversity = 1.2.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Name Value Unit Description None Quality level at transmitter on carrier ica Without useful signal: UL  term  P b  ic  -------------------------------------------------------------------------------------------------------. A Guaranteed bit rate) service users connected to transmitter txi on carrier ic N VBR –m obiles  txi ic  Simulation result None Number of (1xEV-DO .Variable bit rate) service users connected to transmitter txi on carrier ic DL X  txi ic  DL I tot  ic b traffic  ------------------------------------DL N tot  ic b traffic  None Cell downlink load factor on carrier ic UL UL I tot  txi ic  ---------------------------UL N tot  txi ic  UL I tot  txi ic  --------------------------------------------------------------------------------------UL intra Tx I tot  txi ic    1 – F MUD   term  X  txi ic  None Cell uplink load factor on carrier ic None Cell uplink reuse factor on carrier ic UL F  txi ic  357 .Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Name Value Unit Description I intra  txi ic b traffic   1 – F ortho   BTS   P tot  txi ic b traffic  = 0 DL W Downlink intra-cell interference at terminal on carrier ic (only one mobile is served at a time) W Downlink extra-cell interference based on traffic at terminal on carrier ic W Downlink inter-carrier interference based on traffic at terminal on carrier ic DL  DL I extra  ic b traffic  DL P tot  txj ic b traffic  txj j  i  Ptot  txj icadj btraffic  DL DL I inter – carrier  ic b traffic  txj j ---------------------------------------------------------------RF  ic ic adj  Tx P Transmitted  ic i   ------------------------------------Tx Tx m L  ICP DL I inter – techno log y  ic  DL DL I tot  ic b traffic  n i ic total ni DL DL DL term I tot  ic b traffic  + N 0  Pb UL intra I tot  txi UL extra I tot ic   ic  term txi   txi ic  Total effective interference based on traffic at terminal on carrier ic (after unscrambling) W Total noise based on traffic received at terminal on carrier ic W Total power received at transmitter from intra-cell terminals using carrier ic W Total power received at transmitter from extra-cell terminals using carrier ic W Uplink inter-carrier interference at terminal on carrier ic UL P b  ic  term txj j  i  Pb UL UL I inter – carrier  txi W DL + I inter – techno log y  ic  DL ic  Downlink inter-technology interference at terminal on carrier ic a I intra  ic b traffic  + I extra  ic b traffic  + I inter – carrier  ic b traffic  N tot  ic b traffic  UL W  ic adj  term txj j ----------------------------------- RF  ic ic adj  UL I tot  txi ic  UL extra I tot UL intra Tx  txi ic  +  1 – F MUD   term I tot UL W  txi ic  +I inter – carrier  txi ic Total received interference at transmitter on carrier ic N tot  txi ic  I tot  txi ic  + N 0 W Total noise at transmitter on carrier ic (Uplink interference) N mobiles  txi ic  Simulation result None Number of mobiles connected to transmitter txi on carrier ic UL UL tx N GBR –m obiles  txi ic  Simulation result None Number of (1xEV-DO Rev. however. It is equal to the sum of T_Drop defined in the properties of the best server and the Delta T_Drop defined in the properties of the mobility type. The upper threshold is set for the carrier as defined in the cell properties and can also take into account the user mobility type if the Delta minimum Ec/I0 defined in the mobility type is different from 0. When the locked mode is used. Cells entering the mobile’s active set must fulfill the following conditions: • The best server (first cell entering active set) In order for a given transmitter to enter the mobile active set as best server. the cell must be a neighbour of the best server (the "restricted to neighbours” option is selected in the equipment properties).B users. • In order for a transmitter to enter the active set (other cells of active set): • They must use the same carrier as the best server cell. The carrier used by the transmitters in the active set corresponds to the best carrier of the best server. It is. 5. each one being associated with one carrier. the quality of this transmitter’s pilot must be the highest one and it must exceed an upper threshold equal to the sum of the minimum Ec/I0 defined in the properties of the best serving cell and the Delta minimum Ec/I0 defined in the properties of the mobility type. 5. the serving transmitters may be different from one sub-active set to another. the uplink Ec⁄Nt received by the best server on the best carrier and on the studied carrier determines whether or not a carrier can have a subactive set.3 Active Set Management The mobile active set is the list of the transmitters to which the mobile is connected. and an activity status by random trial. With the unlocked mode. For information on the best carrier selection.Atoll 3. The resulting user distribution complies with the traffic database and maps provided to the algorithm. The lower threshold depends both on the type of carrier and the mobility type. which requires traffic maps and data as input. a mobility type. Each user is assigned a service. the quality of the pilot (Ec⁄I0) that finally determines whether or not a transmitter can belong to the active set. • The pilot quality from other candidate cells must exceed a lower threshold. As detailed above. see the Technical Reference Guide. The active set may consist of one or more transmitters. • If you have selected to restrict the active set to neighbours. For multi-carrier EVDO Rev. depending on whether the service supports soft handoff and on the terminal active set size. the quality of the pilot (Ec⁄I0) determines whether or not a transmitter can belong to a sub-active set. 358 . according to a probability law that uses the traffic database. the serving transmitters must be the same in all sub-active sets.4 Simulations The simulation process is divided into two steps: 1. the active set may consist of several sub-active sets.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks Name UL E  txi ic  Value Unit Description 1 --------------------------UL F  txi ic  None Cell uplink reuse efficiency factor on carrier ic DL dB Noise rise on downlink UL dB Noise rise on uplink DL – 10 log  1 – X  txi ic   UL – 10 log  1 – X  txi ic   NR  txi ic  NR  txi ic  a. For the other carriers. ©Forsk 2015 In the case of an interfering GSM external network in frequency hopping.3. Obtaining a realistic user distribution Atoll generates a user distribution using a Monte-Carlo algorithm. the ICP value is weighted according to the fractional load. The number of sub-active sets depends on the maximum number of carriers supported by the terminal. Transmitters in the mobile active set must use a frequency band with which the terminal is compatible and the pilot signal level received from these transmitters must exceed the defined minimum RSCP threshold. The Ec/Nt received by the best serving transmitter on the studied carrier must exceed the minimum uplink Ec/Nt defined in the properties of the transmitter. The sub-active set associated with the best carrier is the same as the active set of a single-carrier user. and the transmitters in the sub-active sets depend on the mode supported by the terminal (locked mode or unlocked mode): • • • The Ec/Nt received by the best serving transmitter on the best carrier must exceed the minimum uplink Ec/Nt defined in the properties of the transmitter. X = SD • • In case of user profile traffic maps composed of lines.1 Generating a Realistic User Distribution 5.4 to 76. • Calculation of the service usage duration per hour ( p 0 : probability of a connection): N call  d p 0 = ------------------3600 where N call is the number of calls per hour and d is the average call duration (in second). the number of subscribers (X) per user profile is calculated from the line length (L) and the user profile density (D) (nb of subscribers per km) as follows: X = L  D The number of subscribers (X) is an input when a user profile traffic map is composed of points.1 Simulations Based on User Profile Traffic Maps User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of user profiles with an associated mobility type and a given density (number of subscribers with the same profile per km²). 38. another random trial determines user positions in their respective traffic zone and whether they are indoors or outdoors (according to the clutter weighting and the indoor ratio per clutter class defined for the traffic maps).2 to 9.8 kbps. Both active and inactive users consume radio resources and create interference.8 to 153. frequency use and exchange volume. and 76.2 kbps. 359 . A user may be either active or inactive.1. 19.4 kbps. 5.4 to 19. This may lead to slight variations in the total numbers of users in different simulations.4. add the following lines in the Atoll. 76.4. Finally. These transition flags are based on the throughput downgrading and upgrading probabilities.6 to 76. Additionally.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 The user activity status is an important output of the random trial and has direct consequences on the next step of the simulation and on the network interferences. Atoll calculates the probability for the user being connected in uplink and in downlink at an instant t. From environment (or polygon) surface (S) and user profile density (D). User Activity Status and User Throughput During the simulation.Atoll 3.4 kbps.3. For each behaviour described in a user profile.1. which mixes throughput control on downlink and power control on uplink. To have the same total number of users in each simulation of a group. 38. Atoll uses a power control algorithm in case of CDMA2000 1xRTT networks and a different algorithm.1 Number of Users. 0 user is assigned a transition flag ("True" or "False") for each possible throughput transition (from 9.ini file: [Simulation] RandomTotalUsers=0 5. a first random trial is performed to determine the number of users and their activity status. Then.6 kbps for throughput downgrading). for CDMA2000 1xEV-DO networks.8 to 38. Atoll determines the total number of users attempting connection in each simulation based on the Poisson distribution. Atoll randomly assigns a shadowing error to each user using the probability distribution that describes the shadowing effect.2 kbps. 2. a number of subscribers (X) per user profile is inferred.1. and 19. If the map is composed of points. If a transition flag is "True." the user throughput can be downgraded or upgraded if necessary.2 to 38.6 to 19. each 1xEV-DO Rev. User profile traffic maps: Each polygon and line of the map is assigned a density of subscribers with given user profile and mobility type. Modelling the network regulation mechanism This algorithm depends on the network. The user profile models the behaviour of the different subscriber categories. each point is assigned a number of subscribers with given user profile and mobility type. The determination of the number of users and the activity status allocation depend on the type of traffic cartography used. according to the service.4.6 kbps for throughput upgrading and from 153.8 kbps. Each user profile contains a list of services and their associated parameters describing how these services are accessed by the subscriber. 4x. Possible throughputs are detailed in the table below: Only FCH is used SCH throughput factor rk On UL - TP P – UL  AF FCH 2x TP P – UL   AF FCH + 2  4x TP P – UL   AF FCH + 4  8x TP P – UL   AF FCH + 8  16x TP P – UL   AF FCH + 16  Both FCH and SCH are used FCH Allocated throughputs FCH On DL UL DL FCH UL TPP – DL   AF FCH + 2  FCH UL TPP – DL   AF FCH + 4  FCH UL TPP – DL   AF FCH + 8  FCH UL FCH TP P – UL and TP P – DL are respectively the uplink and downlink FCH peak throughputs. Atoll calculates the total number of users trying to access a certain service. AF FCH and AF FCH . Therefore. 360 FCH TP P – DL  AF FCH FCH DL FCH DL FCH DL FCH DL TP P – DL   AF FCH + 16  . • Calculation of number of users per activity status: This steps depends on the type of service (Voice. SCH may be allocated with four possible throughputs (2x. This is modelled by UL DL the FCH activity factor. • Calculation of the number of users trying to access the service j ( n j ): nj = X  p0 The next step determines the activity status of each user. Atoll determines the distribution of users between the different possible throughputs. we have: Probability of being active on UL: p UL = 0 Probability of being active on DL: p DL = 0 Probability of being active both on UL and DL: p UL + DL = 1 Probability of being inactive: p inactive = 0 Thus. TP P – UL  AF FCH on uplink and TP P – DL  AF FCH on downlink. 1xEV-DO data…). Therefore. FCH FCH TP P – UL and TP P – DL are respectively the uplink and downlink FCH peak throughputs. • Data Users Data service users are active on uplink and downlink. • CDMA2000 1xRTT Services Activity status of voice and data service users is determined as follows. all voice service users try to access the service with the following FCH FCH UL FCH DL throughputs. Then. for voice and data services.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Then. FCH is always allocated but can have inactivity periods on both links. AF FCH and AF FCH . Therefore. 1xRTT data. the FCH can have inactivity periods on both links. this UL DL is modelled by the FCH activity factor. However. 8x and 16xFCH peak throughput). data service users can access the service with different throughputs.3. Users are always active on FCH in both directions.Atoll 3. we have: Number of inactive users: n j  inactive  = n j  p inactive = 0 Number of users active on UL: n j  UL  = n j  p UL = 0 Number of users active on DL: n j  DL  = n j  p DL = 0 Number of users active on UL and DL both: n j  UL + DL  = n j  p UL + DL = n j n j = n j  UL  + n j  DL  + n j  UL + DL  + n j  inactive  = n j  UL + DL  • Voice Users Voice users are active on uplink and downlink. uplink and downlink. 19.8 and 153. 76. the number of users n j FCH nj = nj – FCH UL with the throughput. r k .4.6. TP P – UL  AF FCH . The number of users active on uplink ( n j  UL  ) and the number of inactive users ( n j  inactive  ) are calculated as follows: Probability of being active on UL: p UL =  Pk UL UL  TP k  UL Rk Probability of being inactive: p inactive = 1 –  Pk UL UL  TP k  UL Rk Probability of being active on DL: p DL = 0 Probability of being active on UL and DL both: p UL + DL = 0 Therefore. r k . P k . these probabilities are 0. 153.4. On uplink. can be assigned to different throughput factors. 614. B service users can access the service with uplink throughputs of 4.6 kbps. is: rk  nj r k On downlink. a random trial compliant with throughput probabilities is performed for each link in order to determine the throughput for each user. 1. 38.228. 1xEV-DO data service users will be considered either active in the uplink or inactive.4.6.8. r k .3. 460.Atoll 3. we have: Number of users active on UL: n j  UL  = n j  p UL Number of inactive users: n j  inactive  = n j  p inactive Number of users active on DL: n j  DL  = n j  p DL = 0 361 . several throughput probabilities.6. j. 307. 921. n j with the throughput. can be assigned to different throughputs TP k . A and Rev. for SCH channel. UL UL For each service.2. P k and P k .8 and 1.2.8.6. 38.2.8. For non-data services. rk UL nj = Pr  nj k FCH Therefore. For data service users.2. 19. TP P – DL   AF FCH + r k  . is calculated as follows. the number of users. we have: rk FCH UL FCH DL For each SCH throughput factor. the number of users n j with the throughput TP P – UL   AF FCH + r k  is calculated as follows. rk DL nj = Pr  nj k FCH Therefore. TP P – DL  AF FCH . 9.2 kbps. 230. several data throughput probabilities. we have: rk For each SCH throughput factor.4. the number of users n j FCH nj = nj – r  nj FCH DL with the throughput. 1xEV-DO data Rev. 1xEV-DO data Rev. 115. j. is: k rk • CDMA2000 1xEV-DO Services As power control is performed in the uplink only. 0 service users can access the service with uplink throughputs of 9.848. 76.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 UL DL In case of a data service. the user distribution between services is different in each of them. CDMA2000 1xRTT Services • Voice Service (j) For each transmitter. P r   k   rk  P r  .3. • • • The user distribution per service is an average distribution and the service of each user is randomly drawn in each simulation. Atoll  considers normalised throughput probabilities values. Traffic is spread over the best server coverage area of each transmitter and each coverage area is assigned either the throughputs in the uplink and in the downlink. Txi. Therefore. instead of k  specified throughput probabilities P r .1. you can input the throughput demands in UL ( TPD DL ( TPD DL UL ) and ) for each sector. k 5. or the number of users per activity status or the total number of users (including all activity statuses). we assume that the sum of throughput probabilities is less than or equal to 1. Atoll proceeds as follows: • When selecting Throughputs in Uplink and Downlink. Therefore. TP k . Therefore. the average number of users per service will correspond to the calculated distribution.2 Simulations Based on Sector Traffic Maps Sector traffic maps can be based on live traffic data from OMC (Operation and Maintenance Centre). UL TP j DL and TP j correspond to the UL and DL throughputs of a user. Atoll determines the distribution of users between the different possible throughputs. Atoll calculates the number of users active in UL and DL using the voice service in the Txi cell as follows: UL DL N UL = TPD --------------. But if you check each simulation.Atoll 3. is calculated as follows: UL UL n j  TP k  = P k  n j Inactive users have a requested throughput equal to 0. The number of users with UL UL the throughput TP k . In calculations detailed above.4. UL TPD is the number of kbits per second transmitted in UL in the Txi cell to provide the service j to the users (userdefined value in the traffic map properties) DL TPD is the number of kbits per second transmitted in DL in the Txi cell to provide the service j to the users (userdefined value in the traffic map properties). Probability of being active in UL: p UL = 0 362 . we have TP j FCH UL FCH = TP P – UL  AF FCH (where TP P – UL is the UL service FCH peak throughput on UL and AF FCH corresponds to the FCH activity factor on UL) and DL TP j FCH DL FCH DL = TPP – DL  AF FCH (where TP P – DL is the service FCH peak throughput on DL and AF FCH corresponds to the FCH activity factor on DL). It is the same for the SCH throughput distribution between 1xRTT data service users and the traffic throughput distribution between 1xEV-DO data service users. n j  TP k  . If the sum of throughput probabilities exceeds 1.and N DL = TPD --------------UL DL TP j TP j Where.1.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Number of users active on UL and DL both: n j  UL + DL  = n j  p UL + DL = 0 n j = n j  UL  + n j  DL  + n j  UL + DL  + n j  inactive  = n j  UL  + n j  inactive  UL Then. FCH is always allocated to active users but UL can have inactivity periods on both links. Users are always active on FCH for both links. if you compute several simulations at once. we have following activity probabilities. Txi. Users are always active on FCH for both links.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Probability of being active in DL: p DL = 0 Probability of being active in UL and DL both: p UL + DL = 1 Probability of being inactive: p inactive = 0 Then. • When selecting Number of Users per Activity Status. SCH may be allocated with four possible throughputs (2x. all connected voice users ( n j ) are active in both links. This is modelled by the FCH activity factors. a second random trial is performed to obtain their geographical positions.N DL) Number of users active in UL and inactive in DL: n j  UL  = 0 Number of users active in DL and inactive in UL: n j  DL  = 0 inactive Number of inactive users in UL and DL: n j = 0 Therefore. FCH UL FCH DL Voice service users try to access the service with the FCH throughputs. r k . 4x. Atoll calculates the number of users per activity status: inactive Number of inactive users in UL and DL: n j = n j  p inactive = 0 Number of users active in UL and inactive in DL: n j  UL  = n j  p UL = 0 Number of users active in DL and inactive in UL: n j  DL  = n j  p DL = 0 Number of users active in UL and DL both: n j  UL + DL  = n j  p UL + DL = n j Therefore. • Data Service Users (j) FCH is always allocated to active users but can have inactivity periods on both links. you can input the number of connected users for each sector ( n j ). UL DL AF FCH and AF FCH . for SCH channel. Therefore. 8x.3. for each sector. P k and P k . these probabilities are 0. all connected users ( n j ) are active in both links. we have following activity probabilities. • When selecting Total Number of Users (All Activity Statuses). Atoll calculates the number of users active in UL and DL using the service in the Txi cell as follows: 363 . Probability of being active in UL: p UL = 0 Probability of being active in DL: p DL = 0 Probability of being active in UL and DL both: p UL + DL = 1 Probability of being inactive: p inactive = 0 Then. you can input the throughput demands in UL ( TPD DL ( TPD DL UL ) and ) for each sector.Atoll 3. TPP – UL  AF FCH on uplink and TP P – DL  AF FCH on downlink. can be assigned to different throughputs factor. Atoll calculates the number of users per activity status: Number of users active in UL and DL both: n j  UL + DL  = max (N UL. For non-data services. For each transmitter. 16xFCH peak throughput). All user characteristics determined. Several UL DL throughput probabilities. Atoll proceeds as follows: • When selecting Throughputs in Uplink and Downlink. you can directly input the number of users active in the uplink and downlink ( n j  UL + DL  ). k Users are always active on FCH for both links.Atoll 3. • In calculations detailed above. we assume that the sum of throughput probabilities is less than or equal to 1. you can input the number of connected users for each sector ( n j ). Probability of being active in UL: p UL = 0 Probability of being active in DL: p DL = 0 Probability of being active in UL and DL both: p UL + DL = 1 Probability of being inactive: p inactive = 0 Then. UL TP j UL Pj = DL and TP j  rk DL Pj = correspond to uplink and downlink throughputs of a user. UL TPD is the number of kbits per second transmitted in UL in the Txi cell to provide the service j to the users (userdefined value in the traffic map properties) DL TPD is the number of kbits per second transmitted in DL in the Txi cell to provide the service j to the users (userdefined value in the traffic map properties). all connected users ( n j ) are active in both links.3.N DL) Number of users active in UL and inactive in DL: n j  UL  = 0 Number of users active in DL and inactive in UL: n j  DL  = 0 inactive Number of inactive users in UL and DL: n j = 0 Therefore. Atoll calculates the number of users per activity status and the total number of users: Number of users active in UL and DL both: n j  UL + DL  = max (N UL. Therefore. If the sum of throughput probabilities exceeds 1. Probability of being active in UL: p UL = 0 Probability of being active in DL: p DL = 0 Probability of being active in UL and DL both: p UL + DL = 1 Probability of being inactive: p inactive = 0 364 . Therefore. Users are always active on FCH for both links.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 UL DL N UL = TPD --------------.  UL UL UL  r k + AF FCH   R FCH  P r +  1 –  k   rk  UL FCH UL P r   TP P – UL  AF FC k      rk + AFFCH   RFCH  Prk +  1 –  Prk   TPP – DL  AFFC DL DL DL  rk FCH FCH DL rk DL  FCH TP P – UL and TP P – DL are the uplink and downlink FCH peak throughputs respectively. we have following activity probabilities. we have following activity probabilities.and N DL = TPD --------------UL DL TP j TP j Where. instead of k  specified throughput probabilities P r . • When selecting Total Number of Users (All Activity Statuses). Atoll  considers normalised throughput probabilities values. P r   k    rk  P r  . 0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Then. 76.228.6. r k . 19. • When selecting Number of Users per Activity Status.4. As explained above. n j with the throughput.2.8. rk UL nj = Pr  nj k FCH Therefore. TP P – DL  AF FCH .2.2. 1. is calculated as follows. rk  nj rk On downlink. for each sector. Atoll calculates the number of users per activity status: inactive Number of inactive users in UL and DL: n j = n j  p inactive = 0 Number of users active in UL and inactive in DL: n j  UL  = n j  p UL = 0 Number of users active in DL and inactive in UL: n j  DL  = n j  p DL = 0 Number of users active in UL and DL both: n j  UL + DL  = n j  p UL + DL = n j Therefore. 76. 38.8. rk  nj rk CDMA2000 1xEV-DO Services As power control is performed in the uplink only. 614. 1xEV-DO data Rev.8 and 1. 230. 365 . 1xEV-DO data Rev. 0 service users can access the service with uplink throughputs of 9. we have.6. 460.3. k DL nj = Pk  nj FCH Therefore. we have.6.6 kbps. data service users can access the service with different throughputs. 38. 9. On uplink. is. 921.8.848. 153. the number of users n j FCH nj = nj – FCH UL with the throughput. r k . the number of users. the number of users n j FCH nj = nj – FCH DL with the throughput. TP P – UL  AF FCH . B service users can access the service with uplink throughputs of 4. rk FCH UL FCH DL For each SCH throughput factor. A and Rev. the number of users n j with the throughput TP P – UL   AF FCH + r k  is calculated as follows. TP P – DL   AF FCH + r k  . 307.2 kbps. 115.4. 19.6. Possible throughputs are detailed in the table below: SCH throughput factor rk Only FCH is used Allocated throughputs On UL FCH TP P – UL -  On DL UL AF FCH FCH TP P – DL FCH UL TP P – DL   AF FCH + 2  FCH UL TP P – DL   AF FCH + 4  FCH UL TP P – DL   AF FCH + 8  2x TPP – UL   AF FCH + 2  4x TPP – UL   AF FCH + 4  8x TPP – UL   AF FCH + 8  16x TPP – UL   AF FCH + 16  Both FCH and SCH are used DL  AF FCH FCH UL FCH DL FCH DL FCH DL FCH DL TP P – DL   AF FCH + 16  Atoll determines the distribution of users with the different possible throughputs.8 and 153.4.2. A random trial compliant with throughput probabilities is performed for each link in order to determine the throughput of each user. is. all connected users ( n j ) are active in both links. r k For each SCH throughput factor.4. you can directly input the number of users active in the uplink and downlink ( n i  UL + DL  ). 1xEV-DO data service users will be considered either active in the uplink or inactive.Atoll 3. Atoll determines the number of users active in UL using the service j in the Txi cell.3. can be assigned to different uplink throughputs TP k . The number of users active in uplink ( n j  UL  ) and the number of inactive users ( n j  inactive  ) are calculated into several steps.Atoll 3. several throughput probabilities. If the sum of throughput probabilities exceeds 1. you can input the throughput demands in UL ( TPD each sector. =  Pk UL UL  TP k k In the above calculations. we have: Number of users active in UL: n j  UL  = N UL  p UL Number of inactive users: n j  inactive  = N UL  p inactive Number of users active in DL: n j  DL  = 0 Number of users active in UL and DL both: n j  UL + DL  = 0 Total number of connected users: n j = n j  UL  + n j  inactive  • When selecting Total Number of Users (All Activity Statuses). For each transmitter. Atoll considers  normalised throughput probabilities values. UL TP j UL TP j corresponds to the uplink throughput for a user. instead of specified k  throughput probabilities P r . P k . UL ) for Atoll calculates the number of users active in UL using the service j in the Txi cell as follows: UL N UL = TPD --------------UL TP j UL Where TPD is the number of kbits per second transmitted on UL in the Txi cell to provide the service j (userdefined value in the traffic map properties). k We have the following activity probabilities: Probability of being active in UL: p UL =  Pk UL UL  TP k  UL Rk Probability of being inactive: p inactive = 1 –  Pk UL R UL  TP k  UL k Probability of being active in DL: p DL = 0 Probability of being active in UL and DL both: p UL + DL = 0 Therefore. We have the following activity probabilities: Probability of being active in UL: p UL =  Pk UL UL Rk 366 UL  TP k  . P r   k    rk  P r  . j. Txi. and each service j: • When selecting Throughputs in Uplink and Downlink. First of all. we assume that the sum of throughput probabilities is less than or equal to 1.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 UL UL For each service. you can input the number of connected users for each sector ( n j ). we have: Number of users active in UL: n j  UL  = n j  p UL Number of inactive users: n j  inactive  = n j  p inactive Number of users active in DL: n j  DL  = 0 Number of users active in UL and DL both: n j  UL + DL  = 0 • When selecting Number of Users per Activity Status.4 kbps 76. you can directly input the number of inactive users ( n j  inactive  ) and the number of users active in the uplink ( n j  UL  ). 0 services supporting throughput downgrading. each 1xEV-DO Rev. Therefore. The number of users with the UL UL throughput TP k . The total number of connected users ( n j ) is calculated as follows n j = n j  UL  + n j  inactive  Then.2 kbps 76.6 kbps During the generation of the user distribution.4 kbps 19. the transition flag for this throughput transition is set to "True" meaning that this throughput transition can be performed if necessary.8 kbps 38. The user distribution per service is an average distribution and the service of each user is randomly drawn In each simulation. 0 user is assigned a random number between 1 and 255 for each possible throughput transition.4 kbps 38.8 kbps 38. is calculated as follows: UL UL n j  TPk  = P k  n j Inactive users have a requested throughput equal to 0. It is the same for the SCH throughput distribution between 1xRTT data service users and the traffic throughput distribution between 1xEV-DO data service users.or underloaded.1. Possible Throughput Changes During Upgrading Possible Throughput Changes During Downgrading From To From To 9. for each sector. The probabilities are taken into account in order to determine if a user with a certain throughput can be upgraded or downgraded. the average number of users per service will correspond to the calculated distribution. if you compute several simulations at once.6 kbps 19. the user distribution between services is different in each of them.6 kbps 76.8 kbps 19.8 kbps 153. When this number is lower or equal to the value of the probability.6 kbps 19. Atoll determines the distribution of users with the different possible throughputs. you can define the probability of the service being upgraded UL UL UL UL UL ( P Upg – k  TP k  ) or downgraded ( P Downg – k  TP k  ) on the uplink (reverse link) for each throughput ( TP k ). 367 .4. n j  TP k  . User throughput downgrading and upgrading occur during congestion control when the cell is over.2 kbps 38. 5.2 kbps 9. The probabilities are defined with a number from 1 to 255 for each throughput. The following table shows the throughput changes that are possible when a throughput is upgraded or downgraded. But if you check each simulation.2 Transition Flags for 1xEV-DO Rev.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Probability of being inactive: p inactive = 1 –  Pk UL R UL  TP k  UL k Probability of being active in DL: p DL = 0 Probability of being active in UL and DL both: p UL + DL = 0 Therefore.0 User Throughputs For 1xEV-DO Rev.4 kbps 76.Atoll 3.3.2 kbps 153. Atoll 3.2 Network Regulation Mechanism 5.3 User Geographical Position Once all the user characteristics determined.1 CDMA2000 1xRTT Power Control Simulation Algorithm CDMA2000 1xRTT network automatically regulates itself using traffic driven uplink and downlink power control on the fundamental and supplemental channels (FCH and SCH respectively) in order to minimize interference and maximize capacity. active set and handoff status for each terminal. The process is repeated from iteration to iteration until convergence is achieved. if you compute several simulations at once. The power control simulation is based on an iterative algorithm. for each user distribution. network parameters such as base station power.3. all the mobiles selected during the user distribution generation (1st step) try to connect to network active transmitters with a calculation area. where in each iteration. Atoll simulates this network regulation mechanism with an iterative algorithm and calculates. 5. the average number of users with a certain throughput that can be downgraded or upgraded will correspond to the calculated value. Figure 5.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 UL The number of 1xEV-DO Rev.2. another random trial is performed to obtain their geographical positions and whether they are indoors or outdoors according to the percentage of indoor users per clutter class defined for the traffic maps. The algorithm steps are detailed below.4. 5. mobile terminal power. 0 users with a certain throughput that can be downgraded ( n j  TP k  Downg ) and upgraded UL ( n j  TP k  Upg ) are calculated as follows: UL UL UL P Upg – k  TP k   n j  TP k  UL n j  TP k  Upg = -----------------------------------------------------------255 And UL UL UL P Downg – k  TP k   n j  TP k  UL n j  TP k  Downg = -----------------------------------------------------------------255 The number of users with a certain throughput that can be downgraded or upgraded is an average. Therefore. But if you check each simulation.4. this number is different in each of them.4.1: CDMA2000 1xRTT Power Control Algorithm 368 .1. UL intra Uplink received powers on carrier ic. either f1 for a single frequency band network. of base station Sj is initialised to P pilot  ic  + P sync  ic  + P paging  ic  . In the algorithm. For each carrier ic.  S BS ic   M i  .3.ic) for which the pilot RSCP exceeds the minimum pilot RSCP: P c  Sj M i ic   RSCP min  Sj ic  . if no good candidate cell has been selected. k Analysis of candidate cells. f2). For each mobile (Mi). or f1 for a multi-band terminal with the configuration 2).4.1.e. then (SBS. all Q req and DL Q req thresholds depend on user mobility type and are defined in Service and Mobility parameters tables. The algorithm applies to single frequency band networks and to multi-band networks (dual-band and tri-band networks). Configuration 2: The terminal can work on f1. For each pair (SBS. we have the following steps: Determination of Mi’s Best Serving Cell For each transmitter Sj containing Mi in its calculation area and working on the main frequency band supported by the Mi’s terminal (i. selection of the transmitter with the highest Q pilot  Sj M i ic  .1. if no good candidate cell has been selected. All variables are described in Definitions and formulas part. f2 and f3 but f1 has a higher priority (select "f1" as main frequency band. "f2" as secondary frequency band and "f3" as third frequency band in the terminal property dialogue). Mi has failed to be connected to the network and is rejected. If no good candidate cell has been selected. (SBS. For multi-band terminals with the configuration 2.    BTS  P c  Sj M i ic  Calculation of Q pilot  Sj ic M i  = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------Term k DL DL DL DL P tot  Sj ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 Determination of the candidate cells.1 Algorithm Initialization Total power on carrier ic. f2 or f3 for a multi-band terminal with the configuration 1.ic) as good candidate cell For multi-band terminals with the configuration 1 or terminals working on one frequency band only. I tot UL  ic  and I inter – carrier  ic  .ic). UL I tot  S j ic  UL .Atoll 3. Atoll only considers the cells (Sj.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 5. (SBS.ic) is rejected by Mi Else Keep (SBS. Xk is the value of the variable X at the iteration k.ic).ic). P Tx  ic  . or f1.ic) is rejected by Mi k UL UL If X k  S BS ic   X max . For each mobile Mi. try to connect Mi to transmitters txi containing Mi in their calculation area and working on the secondary frequency band supported by the Mi’s terminal (i. try to connect Mi to transmitters txi containing Mi in their 369 . I tot UL extra  ic  . Multi-band terminals can have the following configurations: • • Configuration 1: The terminal can work on f1.e.2.2 Presentation of the Algorithm UL The algorithm is detailed for any iteration k. at base station Sj are initialised to 0 W (no connected mobile).2. calculation of the uplink load factor: UL I tot  S BS ic  UL UL X k  S BS ic  = ----------------------------+ X UL N tot  S BS ic  Rejection of bad candidate cells if the pilot is not received or if the uplink load factor is exceeded during the admission load control (if simulation respects a loading factor constraint and Mb was not connected in previous iteration) pilot If Q pilot  S BS M i ic   Q req then (SBS.= 0  X k  S j ic  = ------------------------UL N tot  S j ic  5.4. f2 and f3 without any priority (select "All" as main frequency band in the terminal property dialogue). UL Load Factor” UL ic BS  M i  is the cell with the lowest X k  S BS ic  Else if carrier selection mode is “Min. we will consider ic as the carrier used by the best serving cell.3.ic) present in the Mi active set Calculation of quality level on Mi traffic channel at (Sj. Mi has failed to be connected to the network and is rejected. If no good candidate cell has been selected. and if neighbours are used.icBS). P term  M i ic  k For each cell (Sj. max (S BS.and P b  M i S j ic  = --------------------------------------------- M i S j ic  = ---------------------------------------------L T  M i S j  L T  M i S j  FCH – U L  term  P b  M i S j ic  UL – UL .e.ic BS) k  M i  is the best serving cell ( BestCell k  M i  ) and its pilot quality is Q pilot  M i  . (SBS.ic). If carrier selection mode is “Min. with the minimum power allowed on traffic channel for the Mi service FCH – U L Pb FCH – r eq SCH – r eq P term  M i ic  k – 1 P term  M i ic  k – 1 SCH – U L . f3). Determination of the best carrier. neighbour of BestCell k  M i     BTS  P c  M i S j  Calculation of Q pilot  M i S j ic  = ------------------------------------------------DL k I 0  ic  Rejection of station Sj if the pilot is not received pilot If Q pilot  M i S j ic   Q min then Sj is rejected by Mi k Else Sj is included in the Mi active set Rejection of Sj if the Mi active set is full Station with the lowest Q pilot in the active set is rejected k EndFor Uplink Power Control req Calculation of the required power for Mi. If a given carrier is specified for the service requested by Mi ic BS  M i  is the carrier specified for the service Else the carrier selection mode defined for the site equipment is considered. k In the following lines. icBS. Determination of the Active Set For each station Sj containing Mi in its calculation area. DL Total Power” ic BS  M i  is the cell with the lowest P tx  S BS ic  k Else if carrier selection mode is “Random” ic BS  M i  is randomly selected Else if carrier selection mode is "Sequential" UL UL ic BS  M i  is the first carrier where X k  S BS ic X max Endif Determination of the best serving cell. using ic.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 calculation area and working on the third frequency band supported by the Mi’s terminal (i. G FCH Q FCH  M i S j ic  k = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Service  p UL FCH – U L SCH – U L N tot  ic  –  1 – F MUD    term   P b  M i S j ic  + P b  M i S j ic   370 .  G FCH Q FCH  M i S j ic  k = ------------------------------------------------------------- Service  p UL N tot  ic  SCH –U L  term  P b  M i S j ic  UL – UL .Atoll 3. P FCH  M i ic  k = -------------------------------------------------------------------------------------------------------------------------- M i ic  k – 1 term UL Q FCH  M i  k SCH – r eq P term UL  Q req  Service  Mi  Term  M i  Mobility  M i  SCH_rate_multiple   SCH – r eq . P SCH  M i ic  k = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------- M i ic  k – 1 term UL Q SCH  M i  k req FCH – r eq P term  M i ic  k = P term SCH – r eq  M i ic  k + P term  M i ic  k 371 .3. G SCH Q SCH  M i S j ic  k = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Service  p UL FCH – U L SCH – U L N tot  ic  –  1 – F MUD    term   P b  M i S j ic  + P b  M i S j ic   If the user selects the option “Total noise” FCH – U L  term  P b  M i S j ic  UL – UL .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 SCH – U L  term  P b  M i S j ic  UL – UL . G SCH Q SCH  M i S j ic  k = ------------------------------------------------------------- Service  p UL N tot  ic  End For If (Mi is not in handoff) UL UL UL UL Q FCH  M i  = Q FCH  M i S j ic  k and Q SCH  M i  = Q SCH  M i S j ic  k k k Else if (Mi is in softer handoff) UL UL Q FCH  M i  = f rake efficiency  k UL  UL Q FCH  M i S j ic  k S j  ActiveSet UL Q SCH  M i  = f rake efficiency  k  UL Q SCH  M i S j ic  k S  ActiveSet j Else if (Mi is in soft or softer/soft without MRC) UL UL UL Q FCH  M i  =  G macro – diversity  2 links  Max  Q FCH  M i S j ic  k  k UL S j  ActiveSet UL UL Q SCH  M i  =  G macro – diversity  2 links  Max  Q SCH  M i S j ic  k  k S j  ActiveSet Else if (Mi is in soft/soft) UL UL UL Q FCH  M i  =  G macro – diversity  3 links  Max  Q FCH  M i S j ic  k  k UL S j  ActiveSet UL UL Q SCH  M i  =  G macro – diversity  3 links  Max  Q SCH  M i S j ic  k  k S j  ActiveSet Else if (Mi is in softer/soft with MRC)   UL UL UL UL Q FCH  M i  = Max  f rake efficiency  Q FCH  ic  Q FCH  k other site  i AS  ActiveSet    UL  ic    G macro – diversity  2 links      UL UL UL UL Q SCH  M i  = Max  f rake efficiency  Q SCH  ic  Q SCH  k other site  i  ActiveSet AS    UL  ic    G macro – diversity  2 links     (same site)  (same site) EndIf FCH – r eq P term UL  Q req  Service  M i  Term  M i  Mobility  M i    FCH – r eq . with the minimum power allowed on FCH for the Mi service 372 i . P FCH  M i ic  k = ----------------------------- M i ic  k term req P term  M i  k min P term  M i S j  – r eq . P SCH  M i ic  k = ----------------------------- M i ic  k term req P term  M i  k EndIf FCH – r eq If P term max  M i ic  k  P term  M i  then Mi cannot select any station and its active set is cleared req max If P term  M i ic  k  P term  M i  and Mi uses SCH then: Downgrading the service SCH throughput: req max SCH FCH While P term  M i ic  k  P term  M i  and TP P – UL  Service  M i    TPP – UL  Service  M i    2 SCH TP P – UL  Service  M i   SCH TP P – UL  Service  M i    ----------------------------------------------------2 SCH – r eq SCH – r eq P term UL SCH P term  M i ic   Q req  Service  M i  Term  M i  Mobility  M i  TP P – UL  Service  Mi     SCH  M i ic  k = -----------------------------------------k  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------UL SCH 2  Q  Service  M  Term  M  Mobility  M  TP  Service  M   2    req req FCH – r eq P term  M i ic  k = P term SCH – r eq  M i ic  k + P term i i P – UL i i SCH  M i ic  k EndWhile req max If P term  M i ic  k  P term  M i  then Mi will not use SCH Endif Endif If the required number of channel elements exceeds the available quantity in the site of Sj (Best server of Mi) and Mi uses SCH then: Downgrading the service SCH throughput: Max SCH FCH While N CE –U L  M i   N CE –U L  S j  and TP P – UL  Service  M i    TP P – UL  Service  M i    2 SCH TP P – UL  Service  M i   SCH TP P – UL  Service  M i    ----------------------------------------------------2 SCH N CE –U L  M i  k SCH N CE –U L  M i  k = ----------------------------2 SCH – r eq SCH – r eq P term  M i SCH – UL SCH P term  M i ic   Service  M i  Term  M i  Mobility  M i  TP P – UL  Service  Mi    Q req ic  k = -----------------------------------------k  -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------SCH – UL SCH 2 Q  Service  M  Term  M  Mobility  M  TP  Service  M   2   req req FCH – r eq P term  M i ic  k = P term SCH SCH – r eq  M i ic  k + P term i i i P – UL  M i ic  k FCH N CE –U L  M i  k = N CE –U L  M i  k + N CE –U L  M i  k EndWhile Endif Downlink Power Control If Mi uses an SCH on the downlink For each cell (Sj.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks req ©Forsk 2015 min If P term  M i ic  k  P term  M i  then FCH – r eq P term SCH – r eq P term min P term  M i S j  – r eq .Atoll 3.ic) FCH at Mi.3.ic) in Mi FCH active set Calculation of quality level on (Sj. ic) in Mi SCH active set Calculation of quality level on (Sj.ic) best server cell of Mi): req max SCH While P SCH  M i S j ic  k  P SCH  Service  M i  TP P – DL  Service  M i    req SCH FCH Or P tx  S j ic  k + P tch  M i S j ic  k  P max  S j ic  and TPP – DL  Service  M i    TP P – DL  Service  M i    2 373 .Atoll 3.ic) and Mi: DL FCH  Q req  Service  M i  Term  M i  Mobility  M i  TP P – DL  Service  M i     FCH req . P min P SCH  M i S j ic  k = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------SCH  Service  M i   DL Q SCH  M i  k Downgrading the service SCH throughput (only for (Sj. with the minimum power allowed on SCH for the Mi service SCH – D L Pb min P SCH  Service  M i    M i S j ic  = ---------------------------------------------L T  M i S j  SCH – D L  BTS  P b  M i S j  DL – DL .ic) and Mi: DL SCH  Q req  Service  M i  Term  M i  Mobility  M i  TP P – DL  Service  M i     SCH req . G SCH Q SCH  M i S j ic  k = ------------------------------------------------------------------------------------------------------------ Service  Mi   p DL DL N tot  ic  –  1 – F ortho    BTS  P b  M i S j ic  If the user selects the option “Total noise” SCH – D L  BTS  P b  M i S j  DL Q SCH  M i S j ic  k = ----------------------------------------------------DL N tot  ic  EndIf End For Recombination of the first f active set links (f is the number of fingers of the Mi terminal): only quality levels from the first f cells (Sf.ic) of active set are recombined.ic) SCH at Mi.ic) in Mi SCH active set Calculation of the required power for DL traffic channel between (Sj.3. G FCH S j ic  k = ------------------------------------------------------------------------------------------------------------ Service  M i   p DL DL N tot  ic  –  1 – F ortho    BTS  P b  M i S j ic  If the user selects the option “Total noise” FCH – D L  BTS  P b  M i S j  DL Q FCH  M i S j ic  k = ----------------------------------------------------DL N tot  ic  If cell (Sj.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 FCH – D L Pb min P FCH  Service  M i    M i S j ic  = ---------------------------------------------L T  M i S j  FCH – D L DL Q FCH  M i  BTS  P b  M i S j  – DL . P min P FCH  M i S j ic  k = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------FCH  Service  M i   DL Q FCH  M i  k req max If P FCH  M i S j ic  k  P FCH  Service  M i   then  S j ic  is excluded from Mi active set DL Recalculation of a decreased Q req If cell (Sj. DL DL Q FCH  M i  = f rake efficiency  k  Q FCH  M i S j ic  k  Q SCH  M i S j ic  k DL S f  ActiveSet  FCH  DL DL Q SCH  M i  = f rake efficiency  k DL S f  ActiveSet  SCH  Do For each cell (Sj.ic) in Mi FCH active set Calculation of the required power for DL traffic channel between (Sj. 0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 SCH TP P – DL  Service  M i   SCH TP P – DL  Service  M i   = ----------------------------------------------------2 req DL SCH P SCH  M i S j ic  k  Q req  Service  M i  Term  M i  Mobility  M i  TPP – DL  Service  M i     SCH req  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------P SCH  M i S j ic  k = -------------------------------------DL SCH 2  Q  Service  M  Term  M  Mobility  M  TP  Service  M   2    req req req i i i P – DL i SCH req P tch  M i S j ic  k = P SCH  M i S j ic  k + P FCH  M i S j ic  k EndWhile req max req If P SCH  M i S j ic  k  P SCH  Service  M i   or P tx  S j ic  k + P tch  M i S j ic  k  P max  S j ic  then Mi will not use SCH Endif Max SCH FCH While N CE –D L  M i   N CE –D L  S j  and TP P – DL  Service  M i    TP P – DL  Service  Mi    2 SCH TP P – DL  Service  M i   SCH TP P – DL  Service  M i   = ----------------------------------------------------2 SCH N CE –D L  M i  k SCH N CE –D L  M i  k = ----------------------------2 req DL SCH P SCH  M i S j ic  k  Q req  Service  M i  Term  M i  Mobility  M i  TPP – DL  Service  M i     SCH req P SCH  M i S j ic  k = ------------------------------------- ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------DL SCH 2  Q  Service  M  Term  M  Mobility  M  TP  Service  M   2    req req req i i i P – DL i SCH req P tch  M i S j ic  k = P SCH  M i S j ic  k + P FCH  M i S j ic  k FCH SCH N CE –D L  M i  k = N CE –D L  M i  k + N CE –D L  M i  k EndWhile Max If N CE –D L  M i   N CE –D L  S j  then Mi will not use SCH Endif Max SCH FCH While N Codes  M i   N Codes  S j ic  and TP P – DL  Service  M i    TP P – DL  Service  Mi    2 SCH TP P – DL  Service  M i   SCH TP P – DL  Service  M i   = ----------------------------------------------------2 SCH N Codes  M i  k SCH N Codes  M i  k = ---------------------------2 req DL SCH P SCH  M i S j ic  k  Q req  Service  M i  Term  M i  Mobility  M i  TPP – DL  Service  M i     SCH req P SCH  M i S j ic  k = ------------------------------------- ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------DL SCH 2  Q  Service  M  Term  M  Mobility  M  TP  Service  M   2    req req req i i i P – DL i SCH req P tch  M i S j ic  k = P SCH  M i S j ic  k + P FCH  M i S j ic  k FCH SCH N Codes  M i  k = N Codes  M i  k + N Codes  M i  k EndWhile Max If N Codes  M i   N Codes  S j ic  then Mi will not use SCH Endif Endif EndFor Recombination of the first f active set links (f is the number of fingers of the Mi terminal): only quality levels from the first f cells (Sf.ic) of active set are recombined.3.Atoll 3. DL DL Q FCH  M i  = f rake efficiency  k 374  DL Q FCH  M i S f ic  k S f  ActiveSet  FCH  . Atoll 3. Cell Power and Site Channel Elements) For each cell (Sj.ic) of the site can be allocated to cells (Sj.ic) Max While N Codes  S j ic  k  N Codes  S j ic  375 .3.ic) on a site Nl P tx  S j ic  DL While -------------------------k  %Power max P max req Rejection of mobile with highest P tch  S j M b ic  k for the lowest service priority EndWhile EndFor For each site Nl The list of rejected mobiles for the site Nl is L rejected  N l  If the equipment installed on Nl supports power pooling between transmitters Activation of power pooling between transmitters for each cell (Sj.ic) Sort of all the rejected mobiles by priority in a descending order and by simulation rank in a descending order For the first mobile Mb of the list ( M b  L rejected  N l  ) req DL If P tx  S j ic  k + P tch  S j M b ic  k  %Power max  P max + M Pooling  S j ic  Mb is reconnected EndIf EndFor EndIf EndFor For each cell (Sj.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 DL DL Q SCH  M i  = f rake efficiency  k  DL Q SCH  M i S f ic  k S  ActiveSet  SCH  f DL DL While Q k  M i   Q req  Service  M i  Mobility  M i   and Mi FCH active set is not empty DL DL And Q k  M i   Q req  Service  M i  Mobility  M i   (if SCH active set is not empty) Endif Uplink and Downlink Interference Updates Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones) For each cell (Sj.ic) of the site where power pooling between transmitters is not activated If  DL  %Power max  P max – P tx  S i ic  k   0  S i ic  Si  Nl Then. the power unused by the cells (Si.ic) UL Update of N tot  S j ic  EndFor For each mobile Mi DL Update of N tot  ic  EndFor Control of Radio Resource Limits (Walsh Codes.ic) containing rejected users Control of the available power for the other cells (Si. 0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Rejection of last admitted mobile EndFor For each site (Node B) Nl Max While N CE – DL  N I  k  N CE – DL  N I  req Rejection of mobile with highest P tch  M i S j  k for the lowest service priority Max While N CE – UL  N I  k  N CE – UL  N I  req Rejection of mobile with highest P term  M i ic  k for the lowest service priority EndFor Uplink Load Factor Control UL UL For each cell (Sj. 376 .3. Examples: Let us assume that the maximum number of iterations is 100.  UL and  DL are lower (  ) than their respective thresholds (defined when creating a simulation).  UL or/and  DL are still higher than their respective thresholds and from the 30th iteration.  UL or/and  DL do not decrease during the next 15 successive iterations. After the 30th iteration.  UL and/or  DL equal 80. The simulation has reached convergence. 2nd case: After 30 iterations. If  UL and  DL are lower than their respective thresholds. After the 30th iteration.Atoll 3. Convergence has been achieved. Atoll stops the algorithm after the 5th iteration. and can be written as follow: DL DL      P tx  ic  k – P tx  ic  k – 1 N user  ic  k – N user  ic  k – 1  DL = max  int  ma x ------------------------------------------------ 100  int  ma x ----------------------------------------------------------- 100  DL Stations Stations P tx  ic  k      N  ic  user k UL UL UL UL      I tot  ic  k – I tot  ic  k – 1 N user  ic  k – N user  ic  k – 1  100  int  ma x ----------------------------------------------------------- 100   UL = max  int  ma x -------------------------------------------------UL UL   Stations   Stations  I  ic  N  ic  tot k user k Atoll stops the algorithm if: 1st case: Between two successive iterations. the simulation has not converged (specific divergence symbol). The simulation has not reached convergence (specific divergence symbol).2. UL and DL convergence thresholds are set to 5.1. If  UL  5 and  DL  5 between the 4th and the 5th iteration. they start decreasing slowly until the 40th iteration (without going under the thresholds) and then do not change during the next 15 successive iterations: Atoll stops the algorithm at the 56th iteration without achieving convergence. 3rd case: After the last iteration.  UL and/or  DL equal 100 and do not decrease during the next 15 successive iterations: Atoll stops the algorithm at the 46th iteration. UL and DL convergence thresholds are set to 5.ic) with X  S j ic   X max Rejection of a mobile with the lowest service priority EndFor UL UL While at least one cell with X  S j ic   X max exists 5. Example: Let us assume that the maximum number of iterations is 100. Convergence has not been achieved.4. 1. the simulation has converged.3 Convergence Criterion The convergence criteria are evaluated at each iteration. 2. If  UL and/or  DL are still strictly higher than their respective thresholds. 5.  N t min UL 377 . For each distribution of users. The simulation uses an iterative algorithm. In the downlink.4. where in each iteration.  ----- is the minimum pilot quality level required in the uplink to operate 1xEV-DO Rev. Ec  ---is the value defined in the 1xEV-DO Radio Bearer Selection (Uplink) table for the combination (radio bearer Index. I tot UL  ic  and I inter – carrier  ic  . Atoll simulates the power control mechanism for the UL and the data rate control for the DL. The algorithm steps are detailed below.Atoll 3. at base station Sj are initialised to 0 W (no connected mobile). all the 1xEV-DO data service users selected during the user distribution generation (1st step) try to connect to network active transmitters with a calculation area. the transmitter transmits at the full power (Pmax) when a connection is established. E UL For 1xEV-DO Rev.4.2 Presentation of the Algorithm The algorithm is detailed for any iteration k.1 Algorithm Initialization UL intra Uplink received powers on carrier ic. Ec  --- N t min – RevB is the minimum pilot quality level required in the uplink to operate EV-DO multi-carrier. UL I tot  S j ic  UL  X k  S j ic  = ------------------------.2. B users.3. 0. Atoll considers the guaranteed bit rate service users first. Ec In the algorithm. The process is repeated from iteration to iteration until convergence is achieved.2: CDMA2000 1xEVDO Power Control Algorithm In a CDMA2000 1xEV-DO system. This threshold is UL defined in the Transmitter properties dialogue.= 0 UL N tot  S j ic  5.2. it processes the variable bit rate service users.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 5.2. in the order established during the generation of the user distribution. in the order established during the generation of the user distribution. power control is performed in the uplink only. Instead of power control.2.4. Xk is the value of the variable X at the iteration k. there is a data rate control based on the C/ I ratio calculated at the mobile.2. This N t min – Rev0 UL threshold depends on the user mobility type and is defined in the Mobility parameters table. the value of  ----c- depends on the user requested throughput. This throughput can  N t min be obtained by using a certain uplink 1xEV-DO radio bearer ( Index UL – Bearer ) in a certain number of subframes ( n SF ). and then. I tot UL extra  ic  . A and Rev.2 CDMA2000 1xEV-DO Power/Data Rate Control Simulation Algorithm Figure 5. we have the following steps: Determination of Mi’s Best Serving Cell For each transmitter Sj containing Mi in its calculation area and working on the main frequency band supported by the Mi’s terminal (i. either f1 for a single frequency band network.ic). selection of the transmitter with the highest Q pilot  Sj M i ic  .ic) is rejected by Mi Else Keep (SBS.Atoll 3. (SBS. Atoll only considers the cells (Sj. f2 and f3 without any priority (select "All" as main frequency band in the terminal property dialogue). then (SBS. if no good candidate cell has been selected. f3).  S BS ic   M i  . or f1 for a multi-band terminal with the configuration 2).3. f2).b pilot  Calculation of Q pilot  Sj ic M i  = -----------------------------------------------------------------------------------------------------------------------------------------------------------------Term k DL DL DL P tot  Sj ic . If carrier selection mode is “Min. (SBS. DL Total Power” 378 . one when the service uplink mode is "Low Latency" and another one for high capacity services. if no good candidate cell has been selected. Configuration 2: The terminal can work on f1. If a given carrier is specified for the service requested by Mi ic BS  M i  is the carrier specified for the service Else the carrier selection mode defined for the site equipment is considered. UL Load Factor” UL ic BS  M i  is the cell with the lowest X k  S BS ic  Else if carrier selection mode is “Min. or f1.ic). calculation of the uplink load factor: UL I tot  S BS ic  UL UL X k  S BS ic  = ----------------------------+ X UL N tot  S BS ic  Rejection of bad candidate cells if the pilot is not received or if the uplink load factor is exceeded during the admission load control (if simulation respects a loading factor constraint and Mb was not connected in previous iteration) pilot If Q pilot  S BS M i ic   Q req then (SBS. For each mobile (Mi). "f2" as secondary frequency band and "f3" as third frequency band in the terminal property dialogue). try to connect Mi to transmitters txi containing Mi in their calculation area and working on the third frequency band supported by the Mi’s terminal (i. For multi-band terminals with the configuration 2.    BTS  P c  Sj M i ic . If no good candidate cell has been selected. Determination of the best carrier. f2 and f3 but f1 has a higher priority (select "f1" as main frequency band. Mi has failed to be connected to the network and is rejected.e.ic) as good candidate cell For multi-band terminals with the configuration 1 or terminals working on one frequency band only.ic) is rejected by Mi k UL UL If X k  S BS ic   X max . icBS.b pilot  + I inter – carrier  ic .b pilot  + N 0 Determination of the candidate cells. For each mobile Mi. All variables are described in Definitions and formulas part (see "Definitions and Formulas" on page 340). If no good candidate cell has been selected. f2 or f3 for a multi-band terminal with the configuration 1. Mi has failed to be connected to the network and is rejected. For each carrier ic.b pilot  + I extra  ic .e.ic) for which the pilot RSCP exceeds the minimum pilot RSCP: P c  Sj M i ic b pilot   RSCP min  Sj ic  .e.ic). Multi-band terminals can have the following configurations: • • Configuration 1: The terminal can work on f1. For each pair (SBS. Two values are available for this parameter. The algorithm applies to single frequency band networks and to multi-band networks (dual-band and tri-band networks). try to connect Mi to transmitters txi containing Mi in their calculation area and working on the secondary frequency band supported by the Mi’s terminal (i. k Analysis of candidate cells.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 mobility and number of subframe) providing the user requested throughput. based on the received pilot quality. max (S BS. Determination of the Active Set For each station Sj containing Mi in its calculation area.icn) Ec  --- N t UL max  term  P term  M i   S BS ic n  = -----------------------------------------UL L T  N tot  S BS ic n  379 . k In the following lines. (SBS.ic BS) k  M i  is the best serving cell ( BestCell k  M i  ) and its pilot quality is Q pilot  M i  .icBS). and if neighbours are used. we will consider ic as the carrier used by the best serving cell. icn. or it belongs to f1. Nt N t min For each transmitter Sj containing Mi in its calculation area and using other EV-DO carriers. from the highest to the lowest value. in the defined order: carriers While n max  M i  is not exceeded Determination of the best transmitter of the sub-active set.3. k Determination of the other transmitters of the sub-active set.B service users with a 1xEV-DO Rev.ic) Ec  --- N t UL max  term  P term  M i   S BS ic  = -----------------------------------------UL L T  N tot  S BS ic  E c UL E c UL If  -----  S BS ic    -----  S BS  then EV-DO multi-carrier is not activated. k For each received carrier.Atoll 3. using ic. Q pilot  Sj ic n M i  . calculation of the quality level received by the best serving cell (SBS. Q pilot  Sj ic n M i  . icn. k Calculation of the quality level received by the best serving cell (SBS. f2 or f3 for a multi-band terminal) Calculation of Q pilot  Sj ic n M i  k Ranking of carriers. neighbour of SBS(Mi) DL  BTS    P tot  M i S j ic b pilot  Calculation of Q pilot  M i S j ic  = ---------------------------------------------------------------------------DL k I 0  ic b pilot  Rejection of station Sj if the pilot is not received min If Q pilot  M i S j ic   Q pilot then Sj is rejected by Mi k Else Sj is included in the Mi active set Rejection of Sj if the Mi active set is full Station with the lowest Q pilot in the active set is rejected k EndFor Determination of the Sub-active Sets of a EVDO Multi-carrier User For multi-carrier EV-DO Rev.according to Q pilot  Sj ic n M i  . B capable terminal.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 ic BS  M i  is the cell with the lowest P tx  S BS ic  k Else if carrier selection mode is “Random” ic BS  M i  is randomly selected Else if carrier selection mode is "Sequential" UL UL ic BS  M i  is the first carrier where X k  S BS ic   X max Endif Determination of the best serving cell. icn (either icn belongs to f1 for a single frequency band network. based on the received pilot quality. ic) present in the Mi active set or sub-active set Calculation of quality level on Mi traffic channel at (Sj. with the minimum power allowed on traffic channel for the Mi service req P term  M i ic k – 1 UL P b  M i S j ic  = --------------------------------------L T  M i S j  UL  term  P b  M i S j ic  UL .3. G UL Q  M i S j ic  k = ---------------------------------------------------p  Service  UL N tot  ic  End For If (Mi is not in handoff) UL UL Q total  M i  = Q  M i S j ic  k Else if (Mi is in softer handoff) UL  UL Q total  M i  = f rake efficiency  k UL Q  M i S j ic  k S j  ActiveSet Else if (Mi is in soft or softer/soft without MRC) UL UL Q total  M i  = k UL Max  Q  M i S j ic  k    G macro – diversity  2 links I AS  ActiveSet Else if (Mi is in soft/soft) UL UL Q total  M i  = k I UL Max  Q  M i S j ic  k    G macro – diversity  3 links AS  ActiveSet Else if (Mi is in softer/soft with MRC)   UL UL UL UL Q total  M i  = Max  f rake efficiency  Q  M i S j ic  k Q  M i S j ic  k k othersite  i AS  ActiveSet   (same site)      G UL macro – diversity  2 links    EndIf UL Q req  Service  M i  Term  M i  Mobility  M i   req . then no sub-active set is associated with icn  N t min If the user terminal supports the ’Locked’ mode. If the studied sub-active set does not contain the same transmitters as the sub-active set associated with the best carrier. P req P term  M i ic  k = --------------------------------------------------------------------------------------------------------------term  M i ic  k – 1 UL Q total  M i  k 380 .Atoll 3. G UL Q  M i S j ic  k = -------------------------------------------------------------------------------------------------------------p  Service  UL Tx UL N tot  ic  –  1 – F MUD    term  P b  M i S j ic  If the user selects the option “Total noise” UL  term  P b  M i S j ic  UL . P term  M i ic  k For each cell (Sj.ic). then it is removed. then the studied sub-active set is removed. EndIf Endwhile EndFor Uplink Power Control req Calculation of the required power for Mi. analysis of the sub-active set If a transmitter of the studied sub-active set does not belong to the sub-active set associated with the best carrier.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks E If  ----c-  N t UL ©Forsk 2015 E UL  S BS ic n    ----c-  S BS  . selection of the downlink 1xEV-DO radio bearer ( Index DL – Bearer ): Index DL – Bearer DL Ec Ec where ----. Index DL – Bearer  Nt Nt min If Mi is a 1xEV-DO Rev. 0 service user.B terminals  N t min EndFor UL UL UL UL UL UL UL UL Linear interpolation of CI req  TP TCP – ACK  between CI req  TP low  and CI req  TP high  UL UL UL CI req = CI req  TP  + CI req  TP TCP – ACK  W UL UL Q req = CI req  ----------------------------------------------UL UL  TP + TPTCP – ACK  EndIf 381 . calculation of CI req  TP low  and CI req  TP high  E UL CI req =  ----c-   1 + G DRC + G TCH  for DO Rev.0 terminals N t min UL And E UL UL CI req =  ----c-   1 + G DRC + G TCH + G RRI + G Auxiliary – pilot  for DO Rev.+ N0  Ptot  Sj i c btraffic  + ---------------------------------------------------------------RF  ic ic adj  DL N tot  ic b traffic  = term j j  k Calculation of the maximum throughput supplied to Mi.A and DO Rev. M i S k ic b pilot  = -------------------------------------------------DL Nt N tot  ic b pilot  If Mi is a 1xEV-DO Rev. Index DL – Bearer  and the modulation scheme is supported by the terminal. M i S k ic b pilot   Nt  If Mi is a 1xEV-DO Rev. TP max – DL  M i S k  Calculation of pilot quality level at Mi DL E P tot  M i S k ic b pilot  ----c.  Nt  min Nt DL TP P – R LC  Index DL – Bearer  Determination of the peak throughput: TP max – DL  M i S k  = -----------------------------------------------------------n TS  Index DL – Bearer  DL TP A  M i S k  = TP max – DL  M i S k   f TP – Scaling  Service  Mi   – TP Offset  Service  Mi   UL Determination of the uplink throughput due to TCP acknowledgements.3. selection of the downlink 1xEV-DO radio bearer ( Index DL – Bearer ): Index DL – Bearer DL E E where ----c. A service user. B service user. determination of the peak throughput from the graph (Peak throughput=f(C/I)) specified for the mobility type of Mi E TP max – DL  M i S k  = f  ----c.ic) of Mi Calculation of the Mi downlink application throughput DL Calculation of N tot  ic b traffic   Ptot  txj icadj btraffic  DL DL txj j . M i S k ic b pilot    ----c. M i S k ic b pilot    ----.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 If the service of Mi uses Transmission Control Protocol (TCP) For the best server cell (Sk. TP TCP – ACK  M i S k  from the graph (UL Thr due to TCP=f(DL Thr) specified for the service of Mi UL DL TP TCP – ACK  M i S k  = f  TP A  M i S k   UL UL UL Determination of the nearest lower and higher supported throughputs ( TP low and TPhigh ) for TP TCP – ACK  M i S k  UL UL UL UL UL UL For TPlow and TP high .Atoll 3. (1xEV-DO Rev. UL TP  Service  M i    4. load balancing between carriers is performed. B service users. A . then Mi is not connected to cells of the sub-active set associated with ic 1 . A . A . 0 users.0 and DO Rev.Guaranteed bit rate) service users. TP  Service  M i   = TPD min – UL  Service  M i   EndWhile req max If P term  M i ic  k  P term  M i  then Mi is rejected For 1xEV-DO Rev. UL UL TP  Service  M i   = TP low  Service  M i   UL For (1xEV-DO Rev. P term  M i ic  = P term  M i ic  k  C UL – Bearer Endif Endif For multi-carrier 1xEV-DO Rev.Atoll 3. 382 .6kbps for 1xEV-DO Rev. A . A .Variable bit rate) service users.8kbps for single-carrier 1xEV-DO Rev. UL TP  Service  M i    4. req P term  M i ic  = P term  M i ic  k req For (1xEV-DO Rev. (1xEV-DO Rev. TP UL P term  M i ic  k = ---------------------------------------------low  Service  M i   UL TP  Service  M i   UL ( TP low  Service  M i   is the nearest lower supported throughput) For 1xEV-DO Rev. A . The available terminal power is shared between each carrier as follows: The maximum terminal power is allocated to the best carrier ( ic 1 ).Variable bit rate) and single-carrier 1xEV-DO Rev. 0.A users req max If P term  M i ic  k  P term  M i  then: Downgrading the traffic channel throughput req max While P term  M i ic  k  P term  M i  And UL TP  Service  M i    9. B service users. req P term  M i ic  k req .Guaranteed bit rate) service users. 0. B service users. UL TP  Service  M i    TPD min – UL  Service  M i   for (1xEV-DO Rev. B service users.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks req ©Forsk 2015 req min P term  M i ic  k = Max  P term  M i ic k P term  M i S j   For DO Rev. UL Calculation of the traffic channel throughput ( TP  Service   M i  ic 1   ) UL Downgrading the traffic channel throughput ( TP  Service   M i  ic 1   ) While req max P term  M i ic 1  k  P term  M i  UL and TP  Service   M i  ic 1    153 6kbps req P term  M i ic 1  k req UL .8kbps for (1xEV-DO Rev. TP UL P term  M i ic 1  k = ------------------------------------------------------------low  Service  M i   ( TP low  Service  M i   is the nearest lower supported UL TP  Service   M i  ic 1   throughput) UL UL TP  Service   M i  ic 1   = TP low  Service  M i   EndWhile req max If P term  M i ic 1  k  P term  M i  .Guaranteed bit rate) service users.Variable bit rate) and single-carrier 1xEV-DO Rev.3. ic) Max While N MacIndexes  S j ic   N MacIndexes  S j ic  Rejection of the last admitted mobile EndFor For each site (Node B) Nl Max While N EVDO – CE  N I  k  N EVDO – CE  N I  Rejection of the last admitted mobile EndFor Uplink Load Factor Control UL UL UL For each cell (Sj. then Mi is rejected. The same process is repeated for the other carriers in Mi ’s active set as long as the remaining terminal power is sufficient to obtain the lowest bearer allowed.ic) Max While n EVDO  S j ic   n EVDO  S j ic  Rejection of the last admitted mobile EndFor For each cell (Sj. MAC Indices and Site Channel Elements) For each cell (Sj. If no sub-active set can be used.ic) UL Update of N tot  S j ic  EndFor Control of Radio Resource Limits (Number of EVDO users. UL UL UL If Max  TP  Service  M i     TP high  Service  M i   ( TP high  Service  M i   is the nearest supported throughput higher than the requested throughput) Downgrading the traffic channel throughput UL UL UL While Max   TP  Service  M i     TP high  Service  M i    and TP  Service   M i  ic    153 6kbps EndWhile EndIf Endfor Uplink Interference Updates Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones) For each cell (Sj.Atoll 3. Max  TP  Service  M i    . Endif UL Calculation of TP  Service  M i   for each combination of carriers n UL TP  Service  M i   =  TP UL  Service   M i  ic   where n corresponds to the number of carriers in the combination. ic = 1 UL Selection of the configuration providing the highest throughput.ic) with NR  S j ic   NR threshold  S j ic  + NRthreshold  S j ic  UL UL UL While NR  S j ic   NR threshold  S j ic  + NRthreshold  S j ic  and there is at least one mobile that can be downgraded 383 .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Else the remaining terminal power is allocated to the second best carrier ( ic 2 ) and the traffic channel throughput UL TP  Service   M i  ic 2   is calculated.3.  Index DL – Bearer    Nt Nt min 384 . UL Update of N tot  S j ic  Endwhile UL UL For each cell (Sj. determination of the peak throughput from the graph (Peak throughput=f(C/I)) specified for the mobility type of Mi E TP max – DL  M i S k  = f  ----c.ic) DL Calculation of N tot  ic b traffic  For each cell (Sj. the upgraded throughput cannot exceed the initial user throughput drawn by the Monte-Carlo algorithm. In addition. (only 1xEV-DO Rev.ic). UL Update of N tot  S j ic  Endwhile UL UL UL For each cell (Sj. TP max – DL For the Mi’s best server cell (Sk. P tx  S j ic b traffic  = G idle – power  P max  S j ic  Else P tx  S j ic b traffic  = P max  S j ic  EndFor DL N tot  ic b traffic  =  Ptot  Sj ic btraffic  + N0 DL term j j  k EndFor Calculation of the maximum throughput supplied to Mi. selection of the downlink 1xEV-DO radio bearer ( Index DL – Bearer ) for which DL E E ----c. M i S k ic b pilot  = -------------------------------------------------DL Nt N tot  ic b pilot  If Mi is a 1xEV-DO Rev.3. 0 mobiles for which the throughput transition flag is set to "True". N mobiles  S j ic  If N mobiles  S j ic  = 0 then.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Downgrading the traffic channel throughput for all 1xEV-DO Rev. 0 mobiles for which the throughput transition flag is set to "True".Atoll 3.ic) ( k  j ) Determination of the number of mobiles connected to the cell (Sj.ic) with X  S j ic   X max Rejection of a mobile with the lowest service priority EndFor UL UL While at least one cell with X  S j ic   X max exists Downlink Data Rate Control For each mobile Mi connected to a cell (Sk. This means that only mobiles downgraded during the uplink power control step can be upgraded). A service user.ic) with NR  S j ic   NR threshold  S j ic  – NRthreshold  S j ic  UL UL UL While NR  S j ic   NR threshold  S j ic  – NRthreshold  S j ic  and there is at least one mobile that can be upgraded Upgrading the traffic channel throughput for all 1xEV-DO Rev. M i S k ic b pilot    ----c.ic) (in the active set or each sub-active set) Calculation of pilot quality level at Mi DL E P tot  M i S k ic b pilot  ----c. 0 mobiles which have not been downgraded can be upgraded. M i S k ic b pilot  Nt If Mi is a 1xEV-DO Rev. 0 service user. TP max – DL  M i  =  DL TP max – DL max  M i S k ic   S k ic  For (1xEV-DO Rev. DL TP P –R LC  Index DL – Bearer  Determination of the peak throughput: TP max – DL  M i S k ic  = -----------------------------------------------------------n TS For 1xEV-DO Rev. TP av For each cell (Sj. then G MU = 1 Else if N mobiles  S j ic   1 .Guaranteed bit rate) service users. EndIf EndFor 5. A .4. B service user. G MU is determined from the graph (MUG table=f(nb users)) specified for (Sj. TP max – DL  M i  = TPD min – DL  Service  M i   For multi-carrier 1xEV-DO Rev. B service users. calculation of C DL – Bearer EndFor DL Calculation of the average cell throughput. Mi is rejected.  Nt  min Nt DL If Mi is a (1xEV-DO Rev. 385 . and can be written as follow: UL UL UL UL      I tot  ic  k – I tot  ic  k – 1 N user  ic  k – N user  ic  k – 1  UL = max  int  ma x ------------------------------------------------- 100  int  ma x ----------------------------------------------------------- 100  UL UL Stations Stations      I  ic  N  ic  tot k user k Atoll stops the algorithm if: 1st case: Between two successive iterations.2. G MU is determined from the graph (MUG table=f(nb users)) specified for Sj. 0.Atoll 3.ic). If the transmitter supports the multi-carrier EV-DO mode. Index DL – Bearer  and the modulation is supported by Mi’s terminal.ic) G MU  N mobiles  S j ic     TP max – DL  M i S j ic       Mi  NVBR –m obiles  Sj ic    ---------------------------------------------------------------------------------------C DL – Bearer  M k S j ic     1 – N VBR –m obiles  S j ic      M k  N GBR – m obiles  Sj ic        DL TP av  S j ic  =   TPD min – DL  M k     Mk  NGBR – m obiles  Sj ic   +  ---------------------------------------------------------------------------------------------  C DL – Bearer  M k S j ic  N  S  ic    GBR – m obiles j M  N  S  ic    k GBR – m obiles j      1 –  ER  S  ic   DRC j  N mobiles    1 – TS BCMCS  S j ic  – TS EVDO – CCH  S j ic   + TP BCMCS  S j ic   TS BCMCS  S j ic   If N mobiles  S j ic  = 1 . M i S k ic b pilot    ----c.Variable bit rate) and single-carrier 1xEV-DO Rev.  UL is lower (  ) than the threshold (defined when creating a simulation). UL convergence threshold is set to 5. Atoll stops the algorithm after the 5th iteration. Convergence has been achieved. The simulation has reached convergence.3. Example: Let us assume that the maximum number of iterations is 100. TP max – DL  M i  = TP max – DL  M i S k ic  For (1xEV-DO Rev. B service users.3 Convergence Criterion The algorithm convergence is studied on uplink only.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 If Mi is a 1xEV-DO Rev. selection of the downlink 1xEV-DO radio bearer ( Index DL – Bearer ) for which DL E E ----c. If  UL  5 between the 4th and the 5th iteration.Guaranteed bit rate) service user and TP P – R LC  Index DL – Bearer   TPD min – DL  Service  M i   .Guaranteed bit rate) service users. A .2. A . (1xEV-DO Rev. A . The uplink convergence criterion is evaluated at each iteration. The simulation has not reached convergence (specific divergence symbol). UL convergence threshold is set to 5.2.  UL is still higher than the threshold and from the 30th iteration.2 Resources Management 5. Here. Examples: Let us assume that the maximum number of iterations is 100.1 Walsh Code Management Walsh codes are managed in the downlink during the simulation in case of CDMA2000 1xRTT networks. Walsh codes form a binary tree with codes of a longer length generated from codes of a shorter length. If  UL is still strictly higher than the threshold. Atoll calculates the uplink load factor of a considered cell assuming the mobile concerned is connected with it.1 Admission Control During admission control. or 4 length-2k Walsh codes and so on. if a channel needs 1 length-k/2 Walsh code.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 2nd case: After 30 iterations. To calculate the cell UL load factor. Atoll performs Walsh code allocation during the radio resource control step. we have: UL UL UL Q req =  Q req  FCH +  Q req  SCH and TP UL FCH SCH = TP P – UL + TP P – UL 5.3 Appendices 5. After the 30th iteration. 386 .3: Walsh Code Tree Indices (Not Walsh Code Numbers) 128 128-bit-length Walsh codes per cell are available in CDMA2000 documents.  UL equals 80. the simulation has not converged (specific divergence symbol). 1.4.3. 5.  UL equals 100 and do not decrease during the next 15 successive iterations: Atoll stops the algorithm at the 46th iteration. the simulation has converged. either Atoll takes into account the mobile power determined during power control if mobile was connected in previous iteration. If  UL is lower than the threshold. Convergence has not been achieved. So even if the mobile is not active on UL.Atoll 3. it is equivalent to using 2 length-k Walsh codes.3. 3rd case: After the last iteration.4. Figure 5.3. Length-k Walsh codes are generated from length-k/2 Walsh codes. or it estimates a load rise due to the mobile and adds it to the current load. 2. Therefore. it starts decreasing slowly until the 40th iteration (without going under the threshold) and then does not change during the next 15 successive iterations: Atoll stops the algorithm at the 56th iteration without achieving convergence. it can be rejected due to cell load saturation.  UL does not decrease during the next 15 successive iterations. The load rise ( X X UL UL ) is calculated as follows: 1 = ------------------------------------W 1 + --------------------------UL UL Q req  TP In case of CDMA2000 1xRTT networks. activity status assigned to users is not taken into account.4. After the 30th iteration.4. Atoll determines the number of 128-bit-length Walsh codes that will be consumed by each cell.2. FCH and SCH).3.Atoll 3. The number of common channels per cell corresponds to the value defined for the DL overhead resources for common channels per cell parameter available in the site equipment properties.e. 4k.2 The Walsh code allocation follows the mobile connection order (mobile order in the Mobiles tab). all of its ancestors with lengths k/2. the number of channel elements required on downlink at the site level.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 During the resource control.3. CDMA2000 1xRTT networks Atoll checks the availability of this resource on uplink and downlink. a 128 bit-length code) per common channel for each cell. N 128 bits Walsh codes N 128 bits . Atoll consumes N CE – DL  j  channel elements for each cell j on a site NI. Atoll allocates Walsh codes for each transmitter-receiver link while it assigns channel elements globally to a site. Atoll allocates channel elements for the best serving cell-mobile link only. RC2.4. is: N CE – DL  N I  =  NCE – DL  j  j  NI In case of “softer” handover (the mobile has several links with co-site cells).e. Where SCH TPF DL is the SCH throughput factor.e. RC3 or RC5 and only one for FCH in RC4. N CE – DL  N I  . If a k-length Walsh code is used. is: N CE – UL  N I  =  NCE – UL  j  j  NI In the downlink. FCH and SCH). Channel Element Management Channel elements are controlled in the simulation. And Walsh codes N 128 bits SCH = TPF DL for RC4. for TCH (TCH correspond to Traffic channels i. RC3 and RC5. Therefore. Atoll consumes N CE – UL  j  channel elements for each cell j on a site NI. is determined as follows: SCH = TPF DL  2 for RC1. SCH   1 + TPF UL  per cell-receiver link. • • 5. Therefore. SCH   1 + TPF DL  per cell-receiver link. the number of channel elements required on uplink at the site level. k/4. Paging channel). RC2. The Walsh code and channel element management is dealt with differently in case of “softer” handoff. On uplink. Synchronisation channel. cannot be used as they are not orthogonal. 387 . it allocates : • • • A code with the longest length (i. This figure includes: Overhead • N CE – UL • FCH N CE – UL channel elements for control channels (Pilot channel). This figure includes: Overhead • N CE – DL • FCH N CE – DL channel elements for control channels (Pilot channel. cannot be used as they are not orthogonal. Therefore. N CE – UL  N I  . The Walsh code allocation follows the “Buddy” algorithm. …. for TCH (TCH correspond to Traffic channels i. The number of 128 bit-length codes to be allocated per cell-receiver link for SCH (in case SCH is supported by the user Walsh codes radio configuration). which guarantees that: • • If a k-length Walsh code is used. Two 128 bit-length codes per cell-receiver link for FCH in RC1. all of its children with lengths 2k. …. DL ortho P tx  ic  = P pilot  ic  + P sync  ic  + P paging  ic  + P SCH  ic  + P FCH  ic  = P CCH  ic  +  Ptch  ic  tch where ortho P CCH  ic  = P pilot  ic  + P sync  ic  + P paging  ic   Ptch  ic  = P SCH  ic  + P FCH  ic  tch At mobile level. only one user can be served by a cell at a time.4.3. 388 . I inter – carrier  ic  is the inter-carrier interference received at receiver.Atoll 3. so this resource is not limited. Atoll consumes N CE – UL  j  channel elements for each cell j on a site NI. is: N CE – UL  N I  =  NCE – UL  j  j  NI In the downlink. DL – FCH DL – SCH Gp Gp FCH SCH So. • N CE – UL per cell-receiver link. etc ). required quality is limited to the effective contribution of the transmitter.Guaranteed bit rate) service users. we have a required power. for (EV-DO .Variable bit rate) service users. the number of channel elements required on uplink at the site level. TCH TCH Therefore.3. This figure includes: • 2 channel elements for control channels (Pilot channel.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 CDMA2000 1xEV-DO networks In the uplink. 5. we have CI req = CI req + CI req In case of soft handoff.3 Downlink Load Factor Calculation Atoll calculates the downlink load factor for each cell (available in the Cells tab of any given simulation results) and each connected mobile (available in the Mobiles tab of any given simulation results). Ptch: term P tch  ic  = CI req   I extra  ic  + I intra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0   P tch  ic  = CI req      I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  DL  P tx  ic  – P tch  ic  - + N term +  1 – F ortho   BTS    ----------------------------------------0 LT     LT    L T    DL term  I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T +  1 – F ortho   BTS   P tx  ic  + N 0  L T P tch  ic  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------1 ----------. This value is fixed and hardcoded.1 Downlink Load Factor per Cell The downlink load factor is calculated for each CDMA2000 1xRTT cell.3. Approach for downlink load factor evaluation is highly inspired by the downlink load factor defined in the book “WCDMA for UMTS by Harry Holma and Antti Toskala”. DL I extra  ic  is the total power received at receiver from other cells. Data Rate Control channel. for (EV-DO . N CE – UL  N I  . • N CE – UL  C UL – Bearer per cell-receiver link. 5.+  1 – F ortho   BTS  CIreq where DL I intra  ic  is the total power received at receiver from the cell to which it is connected. DL – FCH DL – SCH Q req Q req + -------------------Let CI req = -------------------be the required quality.3.4.  I  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T     extra   DL term  1 F +  –     P  ic  + N  L  ortho BTS tx 0 T  DL ortho P tx  ic  = P CCH  ic  +  -------------------------------------------------------------------------------------------------------------------------------------------------- 1   ----------.+  1 – F ortho   BTS  tch   CI req      We have:   I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T     --------------------------------------------------------------------------------------------------------------------------------------DL   P tx  ic     DL term  1 F +  –     P  ic  + N  L   ortho BTS tx 0 T  DL ortho ------------------------------------------------------------------------------------------------------------------------------------------------- P tx  ic  = P CCH  ic  +  1 .Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 I inter – techno log y  ic  is the inter-technology interference received at receiver.3. add the following lines in the Atoll.3. To switch back to this method. X DL DL I tot  ic  = -----------------DL N tot  ic  5.8.+ 1 – F ortho   BTS DL   P tx  ic  1 – -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------   1 ----------.3. the downlink load factor can be expressed as: X DL I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T  --------------------------------------------------------------------------------------------------------------------------------------+ 1 – F ortho   BTS DL   P  ic  tx  -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- =   1 .Old Method Before Atoll 2.0.4.4 Best Server Determination in Monte Carlo Simulations .2 Downlink Load Factor per Mobile Atoll evaluates the downlink load factor for any connected mobile (CDMA2000 1xRTT 1xEV-DO user) as follows.+ 1 – F   ---------   ortho BTS tch   CI req            I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T   --------------------------------------------------------------------------------------------------------------------------------------+ 1 – F ortho   BTS    DL  P tx  ic   tch DL DL P tx  ic  –  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------  P tx  ic  1   ----------.+  1 – F ortho   BTS  CI req    term  ortho  N0  LT --------------------------------------------------------------  P CCH  ic  + 1   tch ----------.ini file: [CDMA] 389 . best server determination used to be performed by selecting the best carrier within transmitters according to the selected method (site equipment) and then the best transmitter using the best carrier.+  1 – F ortho   BTS   CI req DL P tx  ic  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T  --------------------------------------------------------------------------------------------------------------------------------------.4. 5.+  1 – F ortho   BTS    tch CI req     Therefore.+ 1 – F --------- ortho   BTS  tch  CI req    The downlink load factor represents the signal degradation in relative to the reference interference (thermal noise).+  1 – F ortho   BTS    CI req      ortho =  P CCH  ic  +    term  N0  LT --------------------------------------------------------------- 1  tch ----------. If a given carrier is specified for the service requested by Mi and if it is used by Sj BestCarrier k  S j M i  is the carrier specified for the service. Mi has failed to be connected to the network and is rejected.3.e. f2 or f3 for a multi-band terminal without any priority on frequency bands. either f1 for a single frequency band network. or f1 for a multi-band terminal with f1 as main frequency band). or f1.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 MultiBandSimu = 0 The method is described below: For each station Sj containing Mi in its calculation area and using the main frequency band supported by the Mi’s terminal (i. If carrier selection mode is “Min. UL UL If X k  S j BestCarrier k  S j M i    X max . then Sj is rejected by Mi Else max Q pilot  M i  = Q pilot  M i S j BestCarrier  k k S BS  M i  = S j Endif EndFor If no SBS has been selected and Mi’s terminal can work on one frequency band only. or f2 for a multi-band terminal with f2 as secondary frequency band) If a given carrier is specified for the service requested by Mi and if it is used by Sj 390 . Determination of BestCarrier k  Sj M i  for each station txj containing Mi in its calculation area and using another frequency band supported by the Mi’s terminal (i. UL Load Factor” For each carrier ic used by Sj. If no SBS has been selected and Mi’s terminal can work on another frequency band. Determination of BestCarrier k  S j M i  . f2 or f3 for a multi-band terminal without any priority on frequency bands. Else the carrier selection mode defined for Sj is considered. we calculate current loading factor: UL I tot  S j ic  UL UL X k  S j ic  = ------------------------.e. f1.+ X UL N tot  S j ic  EndFor UL BestCarrier k  S j M i  is the carrier with the lowest X k  S j ic  Else if carrier selection mode is “Min. DL Total Power” BestCarrier k  S j M i  is the carrier with the lowest P tx  S j ic  k Else if carrier selection mode is “Random” BestCarrier k  S j M i  is randomly selected Else if carrier selection mode is "Sequential" UL UL BestCarrier k  S j M i  is the first carrier so that X k  S j ic   X max    BTS  P c  M i S j BestCarrier  Calculation of Q pilot  M i S j BestCarrier  = ----------------------------------------------------------------------------------DL k I 0  BestCarrier k  S j M i   Rejection of station Sj if the pilot is not received pilot If Q pilot  M i S j BestCarrier   Q req then Sj is rejected by Mi k max If Q pilot  M i S j BestCarrier   Q pilot  M i  k k Admission control (If simulation respects a load factor constraint and Mi was not connected in previous iteration). + X X k  S j ic  = ------------------------UL N tot  S j ic  EndFor UL BestCarrier k  S j M i  is the carrier with the lowest X k  S j ic  Else if carrier selection mode is “Min. we calculate current loading factor: UL I tot  S j ic  UL UL X k  S j ic  = ------------------------. Else the carrier selection mode defined for Sj is considered. Mi has failed to be connected to the network and is rejected. If no SBS has been selected and Mi’s terminal can work on another frequency band.+ X UL N tot  S j ic  391 . f2 or f3 for a multi-band terminal without any priority on frequency bands. Determination of BestCarrier k  Sj M i  for each station txj containing Mi in its calculation area and using another frequency band supported by the Mi’s terminal (i. UL UL If X k  S j BestCarrier k  S j M i    X max . UL Load Factor” For each carrier ic used by Sj. or f3 for a multi-band terminal with f3 as third frequency band) If a given carrier is specified for the service requested by Mi and if it is used by Sj BestCarrier k  S j M i  is the carrier specified for the service. If carrier selection mode is “Min. Else the carrier selection mode defined for Sj is considered.Atoll 3. then Sj is rejected by Mi Else max Q pilot  M i  = Q pilot  M i S j BestCarrier  k k S BS  M i  = S j Endif EndFor If no SBS has been selected and Mi’s terminal can work on two frequency bands only. we calculate current loading factor: UL I tot  S j ic  UL UL .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 BestCarrier k  S j M i  is the carrier specified for the service. DL Total Power” BestCarrier k  S j M i  is the carrier with the lowest P tx  S j ic  k Else if carrier selection mode is “Random” BestCarrier k  S j M i  is randomly selected Else if carrier selection mode is "Sequential" UL UL BestCarrier k  S j M i  is the first carrier so that X k  S j ic   X max    BTS  P c  M i S j BestCarrier  Calculation of Q pilot  M i S j BestCarrier  = ----------------------------------------------------------------------------------DL k I 0  BestCarrier k  S j M i   Rejection of station Sj if the pilot is not received pilot If Q pilot  M i S j BestCarrier   Q req then Sj is rejected by Mi k max If Q pilot  M i S j BestCarrier   Q pilot  M i  k k Admission control (If simulation respects a load factor constraint and Mi was not connected in previous iteration).3. f1. If carrier selection mode is “Min.e. UL Load Factor” For each carrier ic used by Sj. then Sj is rejected by Mi Else max Q pilot  M i  = Q pilot  M i S j BestCarrier  k k S BS  M i  = S j Endif EndFor If no SBS has been selected.1 Point Analysis: The AS Analysis Tab Let us assume a receiver with a terminal.5. 392 . Results are displayed for any point of the map where the pilot signal level exceeds the defined minimum RSCP. To switch back to this method.ini file: [CDMA] SharingEquallyPower = 1 UsingPreviousIterationPowerWeight = 1 5. radio bearer allocation for multi-carrier EVDO Rev. Mi has failed to be connected to the network and is rejected.B . or user-defined cell inputs. DL Total Power” BestCarrier k  S j M i  is the carrier with the lowest P tx  S j ic  k Else if carrier selection mode is “Random” BestCarrier k  S j M i  is randomly selected Else if carrier selection mode is "Sequential" UL UL BestCarrier k  S j M i  is the first carrier so that X k  S j ic   X max    BTS  P c  M i S j BestCarrier  Calculation of Q pilot  M i S j BestCarrier  = ----------------------------------------------------------------------------------DL k I 0  BestCarrier k  S j M i   Rejection of station Sj if the pilot is not received pilot If Q pilot  M i S j BestCarrier   Q req then Sj is rejected by Mi k max If Q pilot  M i S j BestCarrier   Q pilot  M i  k k Admission control (If simulation respects a load factor constraint and Mi was not connected in previous iteration). The type of carrier and the carriers you can select depend on the service and on the frequency band(s) supported by the terminal. UL UL If X k  S j BestCarrier k  S j M i    X max .2.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 EndFor UL BestCarrier k  S j M i  is the carrier with the lowest X k  S j ic  Else if carrier selection mode is “Min.4.3. This receiver does not create any interference.B used to be performed by equally sharing the available terminal power between the carriers.3. The analysis is based on the uplink load percentage and the downlink total power of cells. You can make the prediction for a specific carrier or for the best 1xRTT or 1xEV-DO carrier. add the following lines in the Atoll. average values calculated from a group of simulations. These parameters can be either outputs of a given simulation.Atoll 3. 5.Old Method Before Atoll 3. a mobility type and a service with certain UL and DL throughputs.5 Radio Bearer Allocation Algorithm for Multi-carrier EVDO Rev.5 CDMA2000 Prediction Studies 5.1. 1 Bar Graph and Pilot Sub-Menu We can consider the following cases: 1st case: Analysis based on a specific carrier The carrier that can be used by transmitters is fixed. DL  BTS    P tot  i ic b pilot  Q pilot  i ic  = ---------------------------------------------------------------DL I 0  ic b pilot  DL DL DL DL DL term With I 0  ic b pilot  = P tot  i ic b pilot  + I extra  ic b pilot  + I inter – carrier  ic b pilot  + I inter – techno log y  ic  + N 0 The calculation of Q pilot  i ic  can be divided into 6 steps explained in the table below. The best carrier selection depends on the option selected for the site equipment (UL minimum noise. Then. DL minimum power. 2nd case: Analysis based on the best carrier of all frequency bands Atoll determines the best carrier for each transmitter i which contains the receiver in its calculation area and uses a frequency band supported by the receiver’s terminal. sequential). 3rd case: Analysis based on the best carrier of any frequency band (for multi-band terminals with priority defined on frequency bands only) The frequency band that can be used is fixed. The best carrier selection depends on the option selected for the site equipment (UL minimum noise.  BTS    P c  i ic  Q pilot  i ic  = --------------------------------------------DL I 0  ic  DL DL DL DL term DL with I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 For CDMA2000 1xEV-DO users. For CDMA2000 1xRTT users we have. Atoll determines the best carrier for each transmitter i containing the receiver in its calculation area and using the selected frequency band. CDMA2000 1xRTT or 1xEV-DO. P pilot  i ic  P c  i ic  = ------------------------LT I P tx  i ic b pilot  DL P tot  i ic b pilot  = ----------------------------------LT I and P tx  i ic b pilot  = P max  i ic  L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io L T is the total loss between the transmitter i and the receiver: ------------------------------------------------------------------------------------------------------------------------------------I G Tx  G term 393 .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 5. In this case.5. Atoll calculates the pilot quality at the receiver from these transmitters on their best carriers (ic) and defines the best server (on its best carrier). Atoll provides the same outputs in the bar graph and pilot sub-menu whichever the studied network.ic) P c  i ic  calculation for each cell (i. Then. random. for each transmitter i containing the receiver in its calculation area and using the selected carrier.1. 1st step the receiver. Then.ic) P c  i ic  is the pilot power from a transmitter i on the carrier ic at DL P tot  i ic b pilot  is the pilot burst from the transmitter i on the carrier ic at the receiver.Atoll 3. it determines the best serving transmitter using the selected carrier ic. random. Atoll calculates the pilot quality at the receiver from these transmitters on their best carriers (ic) and defines the best server (on its best carrier). DL minimum power.3. Atoll calculates the pilot quality at the receiver on this carrier. sequential). we have. • Ec/I0 (or Q pilot  ic  ) evaluation We assume that ic is the best carrier of a transmitter i containing the receiver in its calculation radius. CDMA2000 1xRTT users CDMA2000 1xEV-DO users DL P tot  i ic b pilot  calculation for each cell (i. G macro – diversity .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 CDMA2000 1xRTT users DL DL CDMA2000 1xEV-DO users DL I extra  ic  . P tx  j ic  DL P tot  j ic  = -------------------LT DL  Ptot  j ic bpilot  DL I extra  ic b pilot  = j j  i  Ptot  j icadj bpilot  DL P tx  j ic  is the total power transmitted by the transmitter j on the DL best carrier of the transmitter i. DL G macro – diversity = npaths M Shadowing – Ec  Io – M Shadowing –Ec  Io npaths M Shadowing – Ec  Io is the shadowing margin for the mobile receiving n pilot signals (not necessarily from transmitters belonging to the mobile active set). the macro-diversity gain equals 0. and  Ptot  j icadj  DL DL DL I inter – techno log y  ic  =  j I inter – carrier  ic  = j-----------------------------------RF  ic ic adj   ni Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic i and DL I inter – techno log y  ic  =  n i Tx P Transmitted  ic i  -------------------------------------Tx Tx m L total  ICP ic  ic i 3rd step term N0 calculation Tx DL 4th step NF Term  K  T  W  NR inter – techno log y DL I 0  ic  and Q pilot  i ic  evaluation based on formulas defined above DL G macro – diversity calculation DL The macro-diversity gain.Atoll 3. Note: This parameter is determined from the fixed cell edge coverage probability and the model standard deviation. received at the receiver from the transmitter j on the best carrier ic of the transmitter i.3. j j I inter – carrier  ic b pilot  = ---------------------------------------------------RF  ic ic adj  Finally. 394 . I inter – carrier  ic  and I inter – techno log y  ic  calculation We have. P tot  j ic  is the total power We have. When the model standard deviation is set to 0. models the decrease in shadowing margin due to the fact there are several pilot signals at the 5th step mobile. we have. DL I extra  ic  =  Ptot  j ic  DL DL DL I extra  ic b pilot  and I inter – carrier  ic b pilot  calculation j j  i DL 2nd step For each transmitter of the network. 5. these results are detailed for each sub-active set. Then. Other cells (i. I0 (Best server). • Number of fingers The number of fingers. It corresponds to the active set size. Resulting If Q pilot req  ic   Q pilot . We assume that f is less than or equal to mFCH and mSCH. The cell with the highest Q pilot  i ic  enters the active set as best server ( Q pilot  BS ic  ) and the best carrier (icBS) of the 6th step best server BS will be the one used by other transmitters of active set (when active set size is greater than 1).Atoll 3. 11. No power control is performed as in simulations. no cell (i. the pilot quality from the best server and from the other servers of the sub-active set.icBS) in active set must fulfill the following criteria: pilot Q pilot  i ic BS   Q min  i ic BS   neighbour list  BS ic BS  (optional) For multi-carrier 1xEV-DO Rev. This is the maximum number of active set links that the terminal (rake) can combine.1 CDMA2000 1xRTT Let mFCH and mSCH respectively denote the number of cells in the receiver active set for the fundamental channel (FCH) and the supplemental channel (SCH) and f be the number of rake fingers defined for the terminal. are recalculated to determine the entire receiver active set (when active set is greater than 1).5. 5. Here.ic) can enter the active set. pilot qualities at the receiver from transmitters i (other than the best server) on the best carrier of the best server. the total downlink traffic channel quality on FCH is evaluated and compared with the specified target quality.5. They are calculated as described above. f. only the first f cells will be considered in order to determine the FCH availability on downlink.icBS). Atoll determines the downlink traffic channel quality on FCH at the receiver for the maximum traffic channel power per transmitter allowed on FCH.2. • Thermal noise This parameter is calculated as described above (3rd step).icBS). and the downlink macro-diversity gain. icBS. Pilot is unavailable.2 Downlink Sub-Menu Outputs calculated by Atoll depend on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO). after combination.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 CDMA2000 1xRTT users CDMA2000 1xEV-DO users Determination of active set Atoll takes the transmitter i with the highest Q pilot  i ic  and calculates the best pilot quality received with a fixed cell edge coverage Resulting probability Q pilot Resulting Q pilot  ic  Resulting Q pilot = DL G macro – diversity req  ic  . of the rake receiver. Each of these cells is noted (k. only the first f cells among the mSCH cells of the receiver active set will be considered in order to determine the SCH availability on downlink.1. In the same way. • Downlink macro-diversity gain This parameter is calculated as described above (5th step). For each carrier. CDMA2000 1xEV-DO systems do not support soft handover on downlink.1. Then. Atoll displays the thermal noise. Among the mFCH cells of the receiver active set. Same formulas and calculation method are used to update DL I 0  ic BS  and determine Q pilot  i ic BS  .3. Pilot is available. 395 . This parameter is defined in the terminal properties. It is relevant in CDMA2000 1xRTT only11.B service users. Atoll calculates the traffic channel quality on FCH from each cell (k. • I0 (Best server) I0 (Best server) is the total noise received at the receiver on icBS. • Number of cells in active set This is a user-defined input in the terminal properties.  max  Q pilot  i ic   Resulting  Q pilot means that the pilot quality at the receiver exceeds Q pilot  ic  x% of times (x is the fixed cell edge coverage probability).  G DL = -------------------------------------------------------p DL N tot  ic BS  ic BS   FCH And DL – SCH  BTS  P b –max  k ic BS  DL – SCH . for each cell (k. 1st step: Eb/Nt max for the first f (number of fingers) cells of active set DL DL Let us assume the following notations: Eb/Nt max on FCH and SCH respectively correspond to  Q max  FCH and  Q max  SCH . terminal and SCH throughput. With DL I intra  ic BS  =  1 –  BTS  F ortho   P DL  k ic  tot BS And DL I extra  ic BS  =  Ptot  j icBS  DL j j  k DL For each transmitter in the network.icBS). we have: DL – FCH DL  Q max  k  BTS  P b –max  k ic BS  – FCH . G DL  Q max  k ic BS   SCH = -------------------------------------------------------p DL N tot  ic BS  DL – FCH With P b DL max max k k P FCH DL – SCH P SCH  k ic BS  = ---------. • Required transmitter powers on FCH and SCH req req The calculation of the required transmitter powers on FCH and SCH ( P FCH and P SCH ) may be divided into three steps. max P SCH is the maximum power allowed on SCH for the specified downlink throughput. This parameter is user-defined in the Services table for a certain terminal. after combination. This value depends on the downlink throughput specified in the analysis.. No power control is performed as in simulations.3. This parameter is user-defined in the Services table for a certain terminal and SCH throughput.icBS). Atoll determines the downlink traffic channel quality on SCH at the receiver for the maximum traffic channel power per transmitter allowed on SCH. L T is the total loss between the transmitter i and the receiver. Then. This value is user-defined for a given service and terminal. P b –max  k ic BS  = ---------LT LT DL DL DL DL term And N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + I inter – techno log y  ic BS  + N 0 Where max P FCH is the maximum power allowed on FCH. k L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  DL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term DL N tot  ic BS  is the total noise at the receiver on the best carrier of the best server.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Atoll calculates the traffic channel quality on SCH from each cell (k. the total downlink traffic channel quality on SCH is evaluated and compared with the specified target quality. Here. P tot  ic BS  is the total power received at the receiver from this transmitter on icBS. This value is specified for a given service. 396 . DL I inter – carrier  ic BS  is the inter-carrier interference at the receiver on the best carrier of the best server.Atoll 3. • Eb/Nt target on FCH and Eb/Nt target on SCH DL Eb/Nt target on FCH (  Q req  FCH ) is the downlink traffic data quality target on the fundamental channel (FCH). DL Eb/Nt target on SCH (  Q req  SCH ) is the downlink traffic data quality target on the supplemental channel (SCH). Therefore. 0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1  Ptot  j icadj  DL DL  j ---------------------------------------I inter – carrier  ic BS  = txj RF  ic BS ic adj  icadj is a carrier adjacent to icBS. P max P SCH = --------------------------------------SCH DL  Q MAX  ic BS   SCH • Eb/Nt max on FCH for the first f (number of fingers) cells of active set DL Let us assume the following notation: Eb/Nt max on FCH corresponds to  Q max  FCH . G DL  Q max  k ic BS   FCH = -------------------------------------------------------p DL N tot  ic BS  397 .icBS). DL I inter – techno log y  ic BS  =  ni ic i is the i Tx m ICPic  ic i BS th Tx P Transmitted  ic i  ----------------------------------------Tx Tx m L total  ICP ic  ic i BS interfering carrier of an external transmitter is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the frequency gap between ic i (external network) and ic BS . if there is no handoff. DL I inter – techno log y  ic BS  is the inter-technology interference at the receiver on the best carrier of the best server. we have: DL DL  Q MAX  ic BS  FCH = f rake efficiency    Qmax  k icBS  FCH DL k Where DL f rake efficiency is the downlink rake efficiency factor defined in Terminal properties. we have: DL – FCH  BTS  P b –max  k ic BS  DL – FCH . for each cell (k. On downlink.icBS). DL  Q MAX  SCH is the traffic channel quality on SCH at the receiver on icBS after combining the signal from each cell (k. we have: DL DL  Q MAX  ic BS  FCH =  Q max  k ic BS  FCH For any other handoff status. RF  ic BS ic adj  is the interference reduction factor. we have: DL DL  Q MAX  ic BS  SCH =  Q max  k ic BS   SCH For any other handoff status.3. we have: DL DL  Q MAX  ic BS  SCH = f rake efficiency    Qmax  k icBS  SCH DL k req req 3rd step: P FCH and P SCH calculation DL  Q req  FCH req .icBS). 2nd step: Calculation of the total traffic channel quality on FCH and SCH DL  Q MAX  FCH is the traffic channel quality on FCH at the receiver on icBS after combining the signal from each cell (k. defined between ic and icadj and set to a value different from 0. On downlink.Atoll 3. P max P FCH = --------------------------------------FCH DL  Q MAX  ic BS   FCH DL  Q req  SCH req . if there is no handoff. Therefore. P tot  ic BS  is the total power received at the receiver from the transmitter on icBS. DL I inter – techno log y  ic BS  =  ni ic i is the i Tx m ICP ic  ic i BS th Tx P Transmitted  ic i  ----------------------------------------Tx Tx m L total  ICP ic  ic i BS interfering carrier of an external transmitter is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the frequency gap between ic i (external network) and ic BS .Atoll 3. RF  ic BS ic adj  is the interference reduction factor. we have: DL – SCH  BTS  P b –max  k ic BS  DL – SCH . DL I inter – carrier  ic BS  is the inter-carrier interference at the receiver on the best carrier of the best server. • Eb/Nt max on SCH for the first f (number of fingers) cells of active set DL Let us assume the following notation: Eb/Nt max on SCH corresponds to  Q max  SCH . defined between ic and icadj and set to a value different from 0. This parameter is user-defined in the Services table for a certain terminal.  Ptot  j icadj  DL DL txj j I inter – carrier  ic BS  = ---------------------------------------RF  ic BS ic adj  icadj is a carrier adjacent to icBS. k L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  DL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term DL N tot  ic BS  is the total noise at the receiver on the best carrier of the best server. DL I inter – techno log y  ic BS  is the inter-technology interference at the receiver on the best carrier of the best server. L T is the total loss between the transmitter i and the receiver. With max req DL P FCH – P FCH I intra  ic BS  =  1 –  BTS  F ortho   P DL tot  k ic BS  –  1 –  BTS   max (----------------------------.3. for each cell (k. G DL  Q max  k ic BS   SCH = -------------------------------------------------------p DL N tot  ic BS  max P SCH DL – SCH With P b –max  k ic BS  = ---------LT k DL DL DL DL DL term and N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + I inter – techno log y  ic BS  + N 0 Where 398 .0) LT k And DL I extra  ic BS  =  Ptot  j icBS  DL j j  k DL For each transmitter in the network.and N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + N 0 LT k Where max P FCH is the maximum power allowed on FCH. Therefore.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 max P FCH DL – FCH DL DL DL DL term With P b –max  k ic BS  = ---------.icBS). defined between ic and icadj and set to a value different from 0.  Ptot  j icadj  DL DL  j ---------------------------------------I inter – carrier  ic BS  = txj RF  ic BS ic adj  icadj is a carrier adjacent to icBS. we have: DL DL  Q MAX  ic BS  FCH = f rake efficiency    Qmax  k icBS  FCH DL k Where DL f rake efficiency is the downlink rake efficiency factor defined in Terminal properties. L T is the total loss between the transmitter i and the receiver. k L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  DL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term DL N tot  ic BS  is the total noise at the receiver on the best carrier of the best server.icBS). if there is no handoff. DL  Q MAX  SCH is the traffic channel quality on SCH at the receiver on icBS after combining the signal from each cell (k. On downlink. if there is no handoff.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 max P SCH is the maximum power allowed on SCH for the specified downlink throughput. we have: 399 .Atoll 3. This parameter is user-defined in the Services table for a certain terminal and SCH throughput. RF  ic BS ic adj  is the interference reduction factor. we have: DL DL  Q MAX  ic BS  FCH =  Q max  k ic BS  FCH For any other handoff status.icBS). P tot  ic BS  is the total power received at the receiver from the transmitter on icBS. DL I inter – techno log y  ic BS  =  ni ic i is the i Tx m ICPic  ic i BS th Tx P Transmitted  ic i  ----------------------------------------Tx Tx m L total  ICP ic  ic i BS interfering carrier of an external transmitter is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the frequency gap between ic i (external network) and ic BS .3. With max req DL P SCH – P SCH I intra  ic BS  =  1 –  BTS  F ortho   P DL tot  k ic BS  –  1 –  BTS   max (----------------------------. DL I inter – carrier  ic BS  is the inter-carrier interference at the receiver on the best carrier of the best server.0) LT k And DL I extra  ic BS  =  Ptot  j icBS  DL j j  k DL For each transmitter in the network. • Eb/Nt max on FCH and Eb/Nt max on SCH DL  Q MAX  FCH is the traffic channel quality on FCH at the receiver on icBS after combining the signal from each cell (k. On downlink. DL I inter – techno log y  ic BS  is the inter-technology interference at the receiver on the best carrier of the best server. DL  Q MAX  ic BS  FCH DL  G SHO  FCH = -------------------------------------------------------------DL max   Q max  k ic BS   FCH  k And DL  Q MAX  ic BS   SCH DL  G SHO  SCH = -------------------------------------------------------------DL max   Q max  k ic BS   SCH  k max k 5.0 and 1xEV-DO Rev.2. CDMA2000 1xEV-DO Atoll calculates the effective pilot quality level at the receiver and compares this value with the required quality level.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks DL ©Forsk 2015 DL  Q MAX  ic BS   SCH =  Q max  k ic BS   SCH For any other handoff status. 0 users.2 DL  Q max  k DL ic BS   corresponds to the highest Q max  k ic BS  value.e. the service on the downlink traffic channel is available if DL  Q MAX  ic BS   SCH •  DL  Q req  SCH DL DL  Q MAX  ic BS   FCH   Q req  FCH and .1. Effective Eb/Nt on FCH and Eb/Nt on SCH DL DL  Q eff  FCH and  Q eff  SCH are respectively effective traffic channel qualities at the receiver on icBS supplied on FCH and SCH.3. • Required C/I C For 1xEV-DO Rev. It corresponds to the value read in the graph “Peak throughput=f(C/I) (Rev0)” for the DL specified required throughput.5. is the downlink throughput selected for the analysis. the required throughput ( TP req ) is obtained by using a certain downlink transmission format (i. a 1xEV-DO radio bearer ( Index DL – Bearer ) with a certain number of timeslots ( n TS )). TP req . It is calculated as follows: DL TP P – R LC  Index DL – Bearer  DL TP req = -----------------------------------------------------------n TS 400 .Atoll 3. A users. the required C/I (  --- ) is determined from the graph “Peak throughput=f(C/I)” defined for the I req mobility type selected in the analysis.0 and 1xEV-DO Rev. 1xEV-DO Rev. A users. DL DL DL DL DL  Q eff  FCH = min   Q MAX  FCH  Q req  FCH  And DL  Q eff  SCH = min   Q MAX  SCH  Q req  SCH  • Downlink soft handover gain on FCH and downlink soft handover gain on SCH DL DL  G SHO  FCH and  G SHO  SCH respectively correspond to DL soft handover gains on FCH and SCH. we have: DL DL  Q MAX  ic BS   SCH = f rake efficiency    Qmax  k icBS  SCH DL k Therefore. TP req . A Service Users For 1xEV-DO Rev. DL For 1xEV-DO Rev. Atoll displays the following results: • Required throughput DL The required throughput. Nt For the best cell (BS. ----c.Guaranteed bit rate) service users. ic BS b pilot  be the effective C/I at the receiver on icBS. For the defined mobility type. TP req . ic BS b pilot  . B users. i. TP DL DL . A . we have:   E   1 - ----c.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 C --- is the value defined in the 1xEV-DO Radio Bearer Selection (Downlink) table for this downlink transmission format  I  req (radio bearer Index. Bearer Consumption For (1xEV-DO Rev. is determined from the graph “Peak throughput=f(C/I) (Rev0)” is the value read in the graph “Peak throughput=f(C/I) (Rev0)” for E the calculated effective C/I. ic BS b pilot    --- . Then. 0 users. the obtained throughput. For each sub-active set. B Service Users For single-carrier and multi-carrier 1xEV-DO Rev.icBS) of the receiver active set.Atoll 3.3. Atoll indicates the effective C/I and the obtained throughput: 401 . Atoll calculates the 1xEV-DO bearer consumption.e the radio bearer index ( Index DL – Bearer ) with the number of timeslots ( n TS ).– 1  Q pilot  resulting Where pilot DL Q resulting  ic BS  = G macro – diversity  Q pilot  ic BS  BS • Obtained throughput For 1xEV-DO Rev. TP req . TP DL =  TP DL  ic  ic The traffic data channel on downlink is available if TP • DL DL  TP req . Nt For 1xEV-DO Rev. It corresponds to the C/I required to obtain the defined required DL throughput. it determines the downlink obtained throughput as  I  req Nt follows: TP DL DL TP P –R LC  Index DL – Bearer  = -----------------------------------------------------------n TS The traffic data channel in downlink is available if TP • DL DL  TP req . Atoll displays the following results: • Required throughput DL The required throughput. • Effective C/I Ec Let ----. TP defined for the mobility type selected in the analysis. the obtained throughput ( TP DL ) on downlink depends on the downlink transmission format. mobility and number of timeslots). is the downlink throughput selected for the analysis. Atoll selects the E C downlink transmission format where ----c. ic BS b pilo t  =  -----------------------------Nt 1  ---------------------. A users. • Obtained throughput The obtained throughput corresponds to the sum of the obtained throughputs on each carrier. 1xEV-DO Rev. TPD min – DL C DL – Bearer = -----------------------------------------------------------DL TPP –R LC  Index DL – Bearer  Where TPD min – DL corresponds to the minimum bit rate required by the service in the downlink. Atoll deduces required terminal powers on FCH and SCH. total uplink traffic channel qualities on FCH and SCH are evaluated with respect to the receiver handover status. Atoll calculates the uplink traffic channel quality on FCH and SCH from the receiver.icBS) in the receiver active set.5. Nt I req C ---  I  req is the value defined in the 1xEV-DO Radio Bearer Selection (Downlink) table for this downlink transmission format (radio bearer Index. On uplink.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 E Let ----c. ic b pilot    --- and whose modulation scheme is supported by the terminal. i. From these values. mobility and number of timeslots). the carrier associated with the sub-active set.= ---------------------UL SCH max  Q req  SCH TP P – UL  P term  SCH Therefore. FCH and SCH channels. No power control is performed as in simulations. So. Atoll selects the downlink transmission Ec C format where ----. And UL FCH UL max  Q req  FCH TP P – UL  AF FCH  P term  FCH . R req . It corresponds to the maximum terminal power allowed. Here. • Max terminal power on FCH and SCH max The Max terminal power parameter ( P term ) is user-defined for each terminal.3 Uplink Sub-Menu Outputs calculated by Atoll depend on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO). max  1 – p   P term max  P term  FCH = ----------------------------------------------------------------------------UL SCH  Q req  SCH  TP P – UL 1 + -------------------------------------------------------------------UL FCH UL  Q req  FCH  TP P – UL  AF FCH And 402 . it evaluates the uplink traffic channel quality on SCH at the cell for the maximum terminal power allowed on SCH.3. ic b pilot  be the effective C/I at the receiver on ic.ic) of the receiver sub-active set. For the defined mobility type. Then. This parameter is user-defined in the terminal properties.1. In the same way. The downlink obtained throughput is determined as follows: DL TP P – RLC  Index DL – Bearer  DL TP  ic  = -------------------------------------------------------------n TS 5. Nt For the best cell (BS. -------------------------------------------------------------.1 CDMA2000 1xRTT For each cell (i. calculates the total terminal power required and compares this value with the maximum terminal power allowed.5.Atoll 3. ic b pilot  = ----------------------------------------pilot Nt  – Q resulting  ic  Where pilot DL Q resulting  ic  = G macro – diversity  Q pilot  ic  BS DL The obtained throughput ( TP  ic  ) on downlink depends on the downlink transmission format.1. 5. Atoll determines the uplink traffic channel quality on FCH at the cell for the maximum terminal power allowed on FCH. we have: pilot E   Q resulting  ic  ----c. the terminal power is shared between pilot. we may write: max max max max P term =  P term  pilot +  P term  FCH +  P term  SCH We have: max max  P term  pilot = p  P term Where p is the percentage of the terminal power dedicated to pilot.e the radio bearer index ( Index DL – Bearer ) with the number of timeslots ( n TS ).3. It corresponds to the C/I required to obtain the defined required DL throughput. If there is no handoff. i L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term UL N tot  i ic BS  is the total noise at the transmitter on the best carrier of the best server. for each cell of active set. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. we have: UL UL UL  Q MAX  ic BS  FCH =  G macro – diversity  2 links  max   Q max  i ic BS   FCH  i And UL UL UL  Q MAX  ic BS  SCH =  G macro – diversity  2 links  max   Q max  i ic BS   SCH  i UL  G macro – diversity  2 links is the uplink macro-diversity gain. we have: UL – FCH  term  P b –max  i ic BS  UL – FCH .icBS).Atoll 3. tx N0 UL N tot  i ic BS  = ----------------------------------UL 1 – X  i ic BS  tx N 0 is the transmitter thermal noise.and P b –max  i ic BS  = -----------------------LT LT i i L T is the total loss between the transmitter i and the receiver. For each cell (i.  Q MAX  FCH and  Q max  SCH . G UL  Q max  i ic BS   FCH = --------------------------------------------------------p UL N tot  i ic BS  And UL – SCH UL  Q max  i ic BS   SCH  term  P b –max  i ic BS  – SCH . are calculated as follows: UL 1st step: Evaluation of uplink traffic channel qualities on FCH and SCH. When the option “Shadowing taken into account” is not selected (Prediction properties). UL UL 2nd step: Calculation of FCH and SCH total traffic channel qualities at the transmitter on icBS.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 max  1 – p   P term max  P term  SCH = ----------------------------------------------------------------------------UL FCH UL  Q req  FCH  TP P – UL  AF FCH 1 + -------------------------------------------------------------------UL SCH  Q req  SCH  TP P – UL • Required terminal power on FCH and SCH req req The required terminal powers on FCH and SCH. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. based on the receiver handover status.  Q max  ic BS   i UL FCH and  Q max  ic BS   i SCH . i For soft-soft handover. we have: UL UL UL  Q MAX  ic BS  FCH =  G macro – diversity  3 links  max   Q max  i ic BS   FCH  i 403 . respectively  P term  FCH and  P term  SCH . G UL = --------------------------------------------------------p UL N tot  i ic BS  max max  P term  FCH  P term  SCH UL – FCH UL – SCH With P b –max  i ic BS  = -----------------------. This value is deduced from the cell UL uplink load factor X  i ic BS  . UL UL max  Q max  i ic BS   corresponds to the highest Q max  i ic BS  value.3. we have: UL UL UL UL  Q MAX  ic BS  FCH =  Q max  i ic BS   FCH and  Q MAX  ic BS   SCH =  Q max  i ic BS   SCH For soft handover. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. • Eb/Nt max on FCH for each cell in active set For each cell (i. G UL  Q max  i ic BS   FCH = --------------------------------------------------------p UL N tot  i ic BS  404 . UL  Q req  SCH is the user-defined uplink data traffic quality target on SCH for a given service. Atoll determines the total terminal power required ( P term ). we have: UL UL  Q MAX  ic BS   FCH = f rake efficiency    Qmax  i icBS  FCH UL i UL UL And  Q MAX  ic BS   SCH = f rake efficiency    Qmax  i icBS  SCH UL i For softer-soft handover.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 And UL UL UL  Q MAX  ic BS   SCH =  G macro – diversity  3 links  max   Q max  i ic BS   SCH  i UL  G macro – diversity  3 links is the uplink macro-diversity gain. the service on the uplink data traffic channel is available if P term  P term . req Then. we have:  UL UL UL  Q MAX  ic BS   FCH =  G macro – diversity  2 links  max  f rake efficiency    UL UL  Q max  i ic BS   FCH  Q max  i ic BS   i on the other site i on the same site   FCH And  UL UL UL  Q MAX  ic BS   SCH =  G macro – diversity  2 links  max  f rake efficiency    UL UL  Q max  i ic BS   SCH  Q max  i ic BS   i on the other site i on the same site   SCH otherwise. This parameter is available in the Services table.icBS). there are two possibilities. If the MRC option is selected (option available in Global parameters). UL UL UL  Q MAX  ic BS   FCH =  G macro – diversity  2 links  max   Q max  i ic BS   FCH  i And UL UL UL  Q MAX  ic BS   SCH =  G macro – diversity  2 links  max   Q max  i ic BS   SCH  i req req 3rd step: Calculation of  P term  FCH and  P term  SCH UL UL  Q req  FCH  Q req  SCH req req . from the required terminal power on FCH and SCH. req req req req P term =  P term  FCH +  P term  SCH +  P term  pilot req req As  P term  pilot = p  P term . When the option “Shadowing taken into account” is not selected (Prediction properties).  P max  P term  FCH = --------------------------------------term  FCH and  P term  SCH = --------------------------------------term  SCH UL UL  Q MAX  ic BS   FCH  Q MAX  ic BS   SCH Where UL  Q req  FCH is the user-defined uplink data traffic quality target on FCH for a given service and a terminal. we have: UL – FCH  term  P b –max  i ic BS  UL – FCH . terminal and SCH throughput.Atoll 3. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. we have: req req  P term  FCH +  P term  SCH req P term = -------------------------------------------------------1–p req max Therefore. For softer and softer-softer handovers.3.  P max . This parameter is available in the Services table. +  1 –  term   max (--------------------------N tot  i ic BS  = -----------------------------------.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 max  P term  FCH UL – FCH With P b –max  i ic BS  = -----------------------LT i L T is the total loss between the transmitter i and the receiver. • Eb/Nt max on FCH and SCH UL UL  Q MAX  ic BS  FCH and  Q MAX  ic BS   SCH are respectively the traffic channel qualities on FCH and SCH at the transmitter on icBS after signal combination of all the transmitters of the active set. we have: UL – SCH  term  P b –max  i ic BS  UL – SCH . G UL  Q max  i ic BS   SCH = --------------------------------------------------------p UL N tot  i ic BS  max  P term  SCH UL – SCH With P b –max  i ic BS  = -----------------------LT i L T is the total loss between the transmitter i and the receiver. • Eb/Nt max on SCH for each cell in active set For each cell (i. we have: UL UL UL UL  Q MAX  ic BS  FCH =  Q max  i ic BS   FCH and  Q MAX  ic BS   SCH =  Q max  i ic BS   SCH For soft handover. When the option “Shadowing taken into account” is not selected (Prediction properties). i 405 . If there is no handoff.0) UL LT 1 – X  i ic BS  i tx N 0 is the transmitter thermal noise. UL UL max  Q max  i ic BS   corresponds to the highest Q max  i ic BS  value.+  1 –  term   max (--------------------------UL LT 1 – X  i ic BS  i tx N 0 is the transmitter thermal noise.0) . This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. This value is deduced from the cell UL uplink load factor X  i ic BS  .3.icBS). This value is deduced from the cell UL uplink load factor X  i ic BS  . i L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term UL N tot  i ic BS  is the total noise at the transmitter on the best carrier of the best server. tx max req N0 P SCH – P SCH UL N tot  i ic BS  = -----------------------------------. tx max req N0 P FCH – P FCH UL . i L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term UL N tot  i ic BS  is the total noise at the transmitter on the best carrier of the best server. we have: UL UL UL  Q MAX  ic BS  FCH =  G macro – diversity  2 links  max   Q max  i ic BS   FCH  i And UL UL UL  Q MAX  ic BS  SCH =  G macro – diversity  2 links  max   Q max  i ic BS   SCH  i UL  G macro – diversity  2 links is the uplink macro-diversity gain. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters.Atoll 3. 1. Atoll determines the uplink quality level at the cell for the maximum terminal power 406 . Atoll calculates the uplink quality level from the receiver.and  G SHO  SCH = -----------------------------------------------------------UL UL max   Q max  i ic BS  FCH  max   Q max  i ic BS   SCH  I UL I UL max  Q max  i ic BS   corresponds to the highest Q max  i ic BS  value. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. A Service Users For each cell (l.3.icBS) in the receiver active set. we have:  UL UL UL  Q MAX  ic BS   FCH =  G macro – diversity  2 links  max  f rake efficiency    UL UL  Q max  i ic BS   FCH  Q max  i ic BS   i on the other site i on the same site   FCH And  UL UL UL  Q MAX  ic BS   SCH =  G macro – diversity  2 links  max  f rake efficiency    UL UL  Q max  i ic BS   SCH  Q max  i ic BS   i on the other site i on the same site   SCH otherwise. there are two possibilities. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. When the option “Shadowing taken into account” is not selected (Prediction properties). UL UL UL  Q MAX  ic BS   FCH =  G macro – diversity  2 links  max   Q max  i ic BS   FCH  i And UL UL UL  Q MAX  ic BS   SCH =  G macro – diversity  2 links  max   Q max  i ic BS   SCH  i • Effective Eb/Nt on FCH and SCH UL  Q eff  FCH is the uplink effective traffic channel quality on FCH at the receiver on icBS. If the MRC option is selected (option available in Global parameters).3.2 CDMA2000 1xEV-DO 1xEV-DO Rev. we have: UL UL  Q MAX  ic BS   FCH = f rake efficiency    Qmax  i icBS  FCH UL i UL UL And  Q MAX  ic BS   SCH = f rake efficiency    Qmax  i icBS  SCH UL i For softer-soft handover.0 and 1xEV-DO Rev.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 For soft-soft handover.Atoll 3. UL UL UL UL UL UL  Q eff  FCH = min   Q MAX  FCH  Q req  FCH  and  Q eff  SCH = min   Q MAX  SCH  Q req  SCH  • Uplink soft handover gain FCH and SCH UL  G SHO  FCH corresponds to the UL soft handover gain on FCH. No power control is performed as in simulations. Here. UL  Q eff  SCH is the uplink effective traffic channel quality on SCH at the receiver on icBS.5. we have: UL UL UL  Q MAX  ic BS   FCH =  G macro – diversity  3 links  max   Q max  i ic BS   FCH  i And UL UL UL  Q MAX  ic BS   SCH =  G macro – diversity  3 links  max   Q max  i ic BS   SCH  i UL  G macro – diversity  3 links is the uplink macro-diversity gain. UL UL  Q MAX  ic BS   FCH  Q MAX  ic BS   SCH UL UL  G SHO  FCH = -----------------------------------------------------------. I 5. UL  G SHO  SCH corresponds to the UL soft handover gain on SCH. For softer and softer-softer handovers. • Required terminal power with ACK req The required terminal power ( P term ) calculation may be divided into four steps: UL 1st step: Evaluation of the uplink quality. • Max terminal power max The Max terminal power parameter ( P term ) is user-defined for each terminal. For softer and softer-softer handovers. G UL Q max  i ic BS  = ----------------------------------------------------p UL N tot  i ic BS  max P term UL With P b –max  i ic BS  = -----------LT i L T is the total loss between the transmitter i and the receiver. It corresponds to the maximum terminal power allowed. UL 2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status. we have: UL UL UL Q MAX  ic BS  =  G macro – diversity  3 links  max  Q max  i ic BS   i UL  G macro – diversity  3 links is the uplink macro-diversity gain. for each cell of active set For each cell (i.Atoll 3. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. i For soft-soft handover. This value is deduced from the cell UL uplink load factor X  i ic BS  . we have: UL UL Q MAX  ic BS  = Q max  i ic BS  For soft handover. Then. Q max  i ic BS  . When the option “Shadowing taken into account” is not selected (Prediction properties). UL UL max  Q max  i ic BS   corresponds to the highest Q max  i ic BS  value.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 allowed.+  1 –  term   max (-----------------------------N tot  i ic BS  = -----------------------------------.3.This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. i L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term UL N tot  i ic BS  is the total noise at the transmitter on the best carrier of the best server. we have: UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  i ic BS   i UL  G macro – diversity  2 links is the uplink macro-diversity gain. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. Atoll calculates the required terminal power and compares it with the maximum terminal power allowed. we have: UL  term  P b – max  i ic BS  UL . This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation.icBS). tx max req N0 P term – P term UL . When the option “Shadowing taken into account” is not selected (Prediction properties). From this value. we have: UL UL Q MAX  ic BS  = f rake efficiency   Qmax  i icBS  UL i 407 .0) UL LT 1 – X  i ic BS  i tx N 0 is the transmitter thermal noise. If there is no handoff. the total uplink quality level is evaluated with respect to the receiver handover status. we have: E UL UL UL  Q req  withoutACK =  ----c-  G p   1 + G DRC + G TCH  for 1xEV-DO Rev. data rate control and traffic data gains relative to the pilot. 0 capable terminals N t min 408 . one when the service uplink mode is "Low Latency" and another one for high capacity services. This parameter is available in the Mobility types table. If the MRC option is selected (option available in Global parameters). we have: E UL UL UL Q req =  ----c-  G p   1 + G ACK + G DRC + G TCH + G RRI + G Auxiliary – pilot   N t min Where Ec UL  ---is the minimum pilot quality level required on uplink to obtain the defined throughput. G TCH . G RRI and G Auxiliary – pilot are respectively acknowledgement.  N t min UL G ACK . Two values are available for this parameter. the service on the uplink traffic data channel is available if P term  P term . req 4th step: Calculation of P term UL Q req req . They are defined in the terminal properties (1xEVDO Rev. mobility and number of subframe).e. 0 capable terminal.e. traffic data channel. the uplink throughput selected for the analysis) is obtained by using a certain uplink transmission format (i. G DRC and G TCH are respectively acknowledgement.Atoll 3. In case of a 1xEV-DO Rev. 1xEV-DO radio bearer ( Index UL – Bearer ) with a certain number of subframes ( n SF )) and calculated as follows: UL TP P – R LC  Index UL – Bearer  UL TP req = -----------------------------------------------------------n SF Ec  --- N t min is the value defined in the 1xEV-DO Radio Bearer Selection (Uplink) table for this uplink transmission format (radio UL bearer Index. data rate control. 0 tab). • Required terminal power without ACK Atoll also calculates the required terminal power without taking into account the ACK channel contribution. The required  N t min UL UL throughput. A tab). G ACK . only the evaluation of the required quality on uplink is different. Calculations are quite similar to those detailed in the previous paragraph. P max P term = -------------------------term UL Q MAX  ic BS  req max Therefore. we have:  UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency     Qmax  i icBS  Qmaxi on the other site  i icBS  UL UL i on the same site otherwise.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 For softer-soft handover. Two values of G TCH are available. A capable terminal. reverse rate indicator and auxiliary pilot channel gains relative to the pilot. In this case. TP req .3. G DRC . They are defined in the terminal properties (1xEV-DO Rev. TPreq (i. Q req In case of a 1xEV-DO Rev. we have: E UL UL UL Q req =  ----c-  G p   1 + G ACK + G DRC + G TCH   N t min Where Ec  ---is the minimum pilot quality level on uplink. UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  i ic BS   i UL 3rd step: Evaluation of the required quality level on uplink. there are two possibilities. one when the service uplink mode is "Low Latency" and another one for high capacity services. UL Q MAX  ic BS  is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters of the active set. A capable terminals  N t min UL And then. UL  Q req  withoutACK req . Q max  i ic BS  . we have: UL UL UL Q MAX  ic BS  =  G macro – diversity  3 links  max  Q max  i ic BS   i UL  G macro – diversity  3 links is the uplink macro-diversity gain. When the option “Shadowing taken into account” is not selected (Prediction properties). UL UL max  Q max  i ic BS   corresponds to the highest Q max  i ic BS  value. For softer and softer-softer handovers.3. This value is deduced from the cell UL uplink load factor X  i ic BS  . we have: 409 .0) .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 And E UL UL  Q req  withoutACK =  ----c-  G p   1 + G DRC + G TCH + G RRI + G Auxiliary – pilot  for 1xEV-DO Rev.Atoll 3. for each cell of active set. P max  P term  withoutACK = -------------------------------------term UL Q MAX  ic BS  • UL SHO gain UL 1st step: Evaluation of the uplink quality. we have: UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  i ic BS   i UL  G macro – diversity  2 links is the uplink macro-diversity gain. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. i L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term UL N tot  i ic BS  is the total noise at the transmitter on the best carrier of the best server.icBS). G UL Q max  i ic BS  = ----------------------------------------------------p UL N tot  i ic BS  max P term UL With P b –max  i ic BS  = -----------LT i L T is the total loss between the transmitter i and the receiver. If there is no handoff. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation.+  1 –  term   max (-----------------------------UL LT 1 – X  i ic BS  i tx N 0 is the transmitter thermal noise. When the option “Shadowing taken into account” is not selected (Prediction properties). UL 2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status. For each cell (i. we have: UL UL Q MAX  ic BS  = Q max  i ic BS  For soft handover. tx max req N0 P term – P term UL N tot  i ic BS  = -----------------------------------. we have: UL  term  P b – max  i ic BS  UL . i For soft-soft handover. This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. L T is the total loss between the transmitter i and the receiver. i L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  UL L T = ---------------------------------------------------------------------------------------------------------------------------------------------G Tx  G term UL UL N tot  i ic  is the total noise at the transmitter on the carrier ic. Atoll indicates the UL SHO Gain. tx max req N0 P term  ic  – P term UL .ic). It starts allocating the maximum terminal power to the best carrier and selects the highest 1xEV-DO radio bearer. If the MRC option is selected (option available in Global parameters).3. The following results are displayed: • For each carrier used in the selected configuration.0) UL LT 1 – X  i ic  i 410 . and so on for the other carriers of the active set as long as the remaining terminal power is sufficient to obtain the lowest bearer.Atoll 3. the obtained throughput and the required power. This value is deduced from the cell uplink load factor X  i ic  . B users. we have: UL  term  P b –max  i ic  UL . we have:  UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency     Qmax  i icBS  Qmaxi on the other site  i icBS  UL UL i on the same site otherwise.Guaranteed bit rate) service users. UL UL UL Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  i ic BS   i 3rd step: Calculation of the UL SHO gain UL G SHO corresponds to the uplink soft handover gain. B Service Users For multi-carrier 1xEV-DO Rev. G UL Q max  i ic  = -----------------------------------------------p UL N tot  i ic  max P term  ic  UL With P b –max  i ic  = --------------------LT i max P term  ic  is the terminal power available for the carrier (ic). UL Q MAX  ic BS  UL G SHO = ----------------------------------------------UL max  Q max  i ic BS   i • Bearer Consumption For (1xEV-DO Rev. 1xEV-DO Rev.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks UL UL Q MAX  ic BS  = f rake efficiency  ©Forsk 2015  Qmax  i icBS  UL i For softer-soft handover. A . then Atoll continues allocating the available terminal power to the second carrier. TPD min – UL C UL – Bearer = -------------------------------------------------------------UL TP P – RLC  Index UL – Bearer  Where TPD min – UL corresponds to the minimum bit rate required by the service in the uplink. Q max  i ic  . The calculations can be divided into four steps: UL 1st step: Evaluation of the uplink quality.+  1 –  term   max (---------------------------------------N tot  i ic  = -----------------------------. there are two possibilities. for each cell of the sub-active set For each cell (i. Atoll shares the available terminal power between the carriers and determines the uplink 1xEV-DO radio bearer obtained on each carrier. Atoll calculates the 1xEV-DO bearer consumption. Atoll models load balancing between carriers. If it remains terminal power after serving the first carrier. E UL UL Q req =  ----c-  G p   1 + G ACK + G DRC + G TCH + G RRI + G Auxiliary – pilot  N t min UL Where Ec  ---is the minimum pilot quality level required in the uplink to obtain the 1xEV-DO radio bearer. When the option “Shadowing taken into account” is not selected (Prediction properties). we have: UL UL Q MAX  ic  = f rake efficiency   Qmax  i ic  UL i For softer-soft handover. G TCH . This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. G ACK . They are defined in the terminal properties (1xEV- 411 . there are two possibilities. mobility and number of subframe). Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. G RRI and G Auxiliary – pilot are respectively acknowledgement.3. we have: UL UL Q MAX  ic  = Q max  i ic  For soft handover. data rate control. G DRC .Atoll 3. i For soft-soft handover. reverse rate indicator and auxiliary pilot channel gains relative to the pilot. UL UL UL Q MAX  ic  =  G macro – diversity  2 links  max  Q max  i ic   i UL 3rd step: Calculation of the UL SHO gain ( G SHO ) UL Q MAX  ic  UL G SHO = ------------------------------------------UL max  Q max  i ic   i 4th step: Selection of the uplink 1xEV-DO radio bearer UL req Atoll evaluates of the required quality level in the uplink ( Q req ) and the required terminal power ( P term  ic  ) for each 1xEVDO radio bearer. Two values are available. one when the service uplink mode is "Low Latency" and another one for high capacity services. If the MRC option is selected (option available in Global parameters). For softer and softer-softer handovers. we have:  UL UL UL Q MAX  ic  =  G macro – diversity  2 links  max  f rake efficiency     Qmax  i ic  Qmaxi on the other site  i ic  UL UL i on the same site otherwise. we have: UL UL UL Q MAX  ic  =  G macro – diversity  3 links  max  Q max  i ic   i UL  G macro – diversity  3 links is the uplink macro-diversity gain. Atoll considers the uplink macro-diversity gain defined by the user in Global parameters. UL UL max  Q max  i ic   corresponds to the highest Q max  i ic  value.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 tx N 0 is the transmitter thermal noise.This parameter is determined from the fixed cell edge coverage probability and the uplink Eb/Nt standard deviation. The values are defined  N t min UL in the 1xEV-DO Radio Bearer Selection (Uplink) table for each uplink transmission format (radio bearer Index. traffic data channel. When the option “Shadowing taken into account” is not selected (Prediction properties). UL 2nd step: Calculation of the total quality at the transmitter on ic ( Q MAX ) based on the receiver handover status. we have: UL UL UL Q MAX  ic  =  G macro – diversity  2 links  max  Q max  i ic   i UL  G macro – diversity  2 links is the uplink macro-diversity gain. If there is no handoff. These parameters can either be either simulation results. • Max terminal power max The Max terminal power parameter ( P term ) is user-defined for each terminal. Max  TP total  . It corresponds to the maximum terminal power allowed. • Required throughput UL The required throughput. Let us assume each pixel of the map corresponds to a probe receiver with a terminal. The type of carrier and the carriers you can select depend on the service and on the frequency band(s) supported by the terminal. Required terminal power m req P term =  Pterm  ic  req ic = 1 5. Atoll calculates pilot quality at the receiver on this carrier icgiven.. In this case. ic = 1 The obtained throughput ( TP UL ) corresponds to the best configuration among all combinations of carriers.Atoll 3. Atoll selects the best 1xEV-DO radio bearer. 5. it determines the best serving transmitter BS using the carrier icgiven ( Q pilot  ic given  ) and deduces the best pilot quality received with a fixed BS cell edge coverage probability. • Obtained throughput Atoll calculates the throughput for all combinations of carriers. R req . is the uplink throughput selected for the analysis. Then.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 DO Rev. m UL TP total =  TP UL  ic  where m corresponds to the number of carriers in the combination. And UL Q req req . or user-defined cell inputs. This receiver does not create any interference. i. Two values of G TCH are available.5. Coverage studies are based on the uplink load percentage and the downlink total power of cells. 412 Resulting Q pilot  ic given  . a mobility type and a service. see "Definitions and Formulas" on page 340.5. .2.2 Coverage Studies Atoll calculates CDMA-specific coverage studies on each pixel where the pilot signal level exceeds the minimum RSCP threshold. This is the 1xEV-DO radio bearer ( Index UL – Bearer ) with the highest UL TPP – RLC  Index UL – Bearer  UL obtained throughput ( TP  ic  = -------------------------------------------------------------) where: n SF  Index UL – Bearer  req max • P term  ic   P term  ic   . The traffic data channel is available in uplink if TP • UL UL  TP req .1 Pilot Quality Analysis For further details on calculation formulas. P max P term  ic  = --------------------term  ic   UL Q MAX  ic  Then.3. one when the service uplink mode is "Low Latency" and another one for high capacity services. For further details on calculations. You can make the coverage prediction for a specific carrier or for the best 1xRTT or 1xEV-DO carrier.e. n SF is the number of subframes associated with the 1xEV-DO radio bearer ( Index UL – Bearer ). or average values calculated from a group of simulations. A tab). see "Bar Graph and Pilot Sub-Menu" on page 392 1st case: Analysis based on a specific carrier The carrier that can be used by transmitters is fixed. • And the required modulation scheme is supported by the terminal. the UL combination which provides the highest throughput. for each transmitter i containing the receiver in its calculation area and using the selected carrier. area is covered if Q pilot req  ic  – Q pilot   Q pilot  m arg in ( ic = ic BS or ic given ). 2nd case: Analysis based on the best carrier of all frequency bands Atoll proceeds as in point analysis. Each layer is assigned a colour and displayed with intersections between layers. Atoll displays the best pilot quality received with a fixed cell edge coverage probability. DL minimum power. Atoll determines the best carrier of each transmitter i containing the receiver in its calculation area and using the selected frequency band. sequential) and is based on the UL load percentage and the downlink total power of cells (simulation results or cell properties). Then. Atoll calculates the pilot quality at the receiver from these transmitters on their best carrier and determines the best serving transmitter BS on its best carrier icBS Resulting ( Q pilot  ic BS  ). area is covered if Resulting Q pilot • req  ic   Q pilot ( ic = ic BS or ic given ). For each layer. Coverage consists of a single layer with a unique colour. Atoll displays the best pilot quality received with a fixed cell edge coverage probability. • Colour per quality margin (Ec/I0 margin) Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction Resulting properties).e. Colour per probability This display option is available only if analysis is based on all simulations in a group (i. p.3. Q pilot BS  ic BS  . defined in the Display tab Resulting (Prediction properties). random. if you select a group of simulations and the “All” option in the Condition tab of prediction properties). • Colour per cell edge coverage probability Coverage consists of several layers with a layer per user-defined cell edge coverage probability. There is a layer per transmitter with no intersection between layers. Coverage consists of several layers with a layer per user-defined probability level defined in the Display tab (Prediction Resulting properties).Atoll 3. Each layer is assigned a colour and displayed with intersections between layers. area is covered if Q pilot  ic    Q pilot  threshold ( ic = ic BS or ic given ). it calculates the best pilot quality received with a fixed cell edge coverage probability. For each layer. The best carrier selection depends on the option selected for the site equipment (UL minimum noise. 3rd case: Analysis based on the best carrier of any frequency band (for multi-band terminals with priority defined on frequency bands only) The frequency band that can be used is fixed. • Colour per quality level (Ec/I0) Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction Resulting properties). • Single colour Resulting Atoll displays a coverage if Q pilot req  ic   Q pilot . Then. For each layer. Then. Each layer is assigned a colour and displayed with intersections between layers. ic = ic BS or ic given • Colour per transmitter Resulting Atoll displays a coverage if Q pilot req  ic   Q pilot ( ic = ic BS or ic given ). Each layer is assigned a colour and displayed with intersections between layers. the receiver is not completely defined and no mobility assigned. • Colour per mobility In this case. For each layer. sequential) and is based on the UL load percentage and the downlink total power of cells (simulation results or cell properties). it deduces the best pilot quality received with a fixed cell edge coverage probability. random. DL minimum power.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Atoll displays the best pilot quality received with a fixed cell edge coverage probability. Coverage consists of several layers with a layer per user-defined mobility type defined in the Mobility Types sub-folder. The best carrier selection depends on the option selected for the site equipment (UL minimum noise. area is covered if Q pilot req  ic   Q pilot ( ic = ic BS or ic given ) in the required number of simulations. Atoll calculates the pilot quality at the receiver from these transmitters on their best carrier and determines the best serving transmitter BS on its best carrier icBS Resulting ( Q pilot  ic BS  ). Layer colour is the colour assigned to the best serving transmitter BS. Coverage consists of several layers with associated colours. 413 . area is covered if Q pilot req  ic p   Q pilot ( ic = ic BS or ic given ). For each layer. It determines the best carrier of each transmitter i containing the receiver in its calculation area and using a frequency band supported by the receiver’s terminal. Q pilot BS  ic BS  . Each layer is assigned a colour and displayed with intersections between layers. receiver is not completely defined and no mobility is assigned. No power control is performed as in simulations. For further details on formulas. Layer colour is the colour assigned to best serving transmitter. Coverage consists of several layers with associated colours.5.2 Downlink Service Area Analysis The downlink service area analysis depends on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO). if you select a group of simulations and the “All” option in the Condition tab of prediction properties). area is covered if  Q MAX  ic   FCH   Q req  FCH in the required number of simulations. Best server and active set determination is performed as in point prediction. area is covered if  Q MAX  ic   FCH   Q req  FCH . Atoll calculates downlink quality on FCH at the receiver for each cell (k. You may choose following display options: • Single colour DL DL DL Atoll displays a coverage with a unique colour if  Q MAX  ic   FCH   Q req  FCH . 5. Atoll displays total traffic channel quality at the receiver on the carrier ic ( ic BS or ic given ). Atoll determines the downlink quality on FCH at the receiver for a maximum traffic channel power per transmitter allowed on the fundamental channel (FCH). • Colour per transmitter DL DL Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH . Coverage consists of several layers with a layer per DL DL user-defined service defined in Services sub-folder. • Colour per probability This display option is available only if analysis is based on all simulations in a group (i. receiver is not completely defined and no service is assigned.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Colour per pilot signal level (Ec) Coverage consists of several layers with a layer per user-defined pilot signal level defined in the Display tab (Prediction Resulting properties).2. see "Downlink Sub-Menu" on page 395.5. Each layer is assigned a colour and displayed with intersections between layers. Here. For each layer.  Q req  FCH is the downlink traffic data quality target on the fundamental channel (FCH). Then. For each layer.ic) (with ic=icBS or icgiven) (these cells are the first f cells in the receiver’s active set and f is the number of fingers defined for the terminal). Each layer is assigned a colour and displayed with intersections between layers. see "Definitions and Formulas" on page 340. This parameter is user-defined for a given service and a terminal in the Services subfolder. Coverage consists of several layers with a layer per DL DL user-defined mobility defined in Mobility sub-folder.e. • Colour per service In this case. 5. area is covered if  Q MAX  ic   FCH   Q req  FCH . For each layer. some of which are dedicated to CDMA2000 1xRTT networks while others are relevant when analysing CDMA2000 1xEV-DO systems only.3. Each layer is assigned a colour and displayed with intersections between layers. the total downlink quality on FCH DL (  Q MAX  ic   FCH ) is evaluated after recombination. • 414 Colour per cell edge coverage probability . Each layer is assigned a colour and displayed with intersections between layers.Atoll 3.1 CDMA2000 1xRTT As in point analysis.2. For each layer. • Colour per mobility In this case. Several display options are available when calculating this study. Coverage consists of several layers with a layer per user-defined probability level defined in the Display tab (Prediction DL DL properties). For further details on calculation.2. area is covered if Q pilot  ic    Q pilot  threshold ( ic = ic BS or ic given ). There is a layer per transmitter with no intersection between layers. area is covered if  Q MAX  ic   FCH –  Q req  FCH  M arg in . For each layer. see "Downlink Sub-Menu" on page 395. For each possible throughput. Coverage consists of several layers with a layer per user-defined required power threshold defined in the Display tab req (Prediction properties).5.2.2 CDMA2000 1xEV-DO E As in point analysis. Atoll calculates the effective pilot quality level at the receiver from the best server cell. P FCH  ic  . • Colour per required power margin Coverage consists of several layers with a layer per user-defined power margin defined in the Display tab (Prediction max req properties). For further details on calculations. Each layer is assigned a colour and displayed with intersections between layers.3. For each layer. For each layer. TP P – DL   AF FCH + 16  ). It corresponds to the maximum traffic data power allowed on FCH for a transmitter. Then. DL TP P – DL   AF FCH + 2  . area is covered if P FCH – P FCH  ic   M arg in . 415 . Each layer is assigned a colour and displayed with intersections between layers. area is covered if  Q MAX  ic   FCH  Threshold . 5. DL Q req  TP  is the downlink traffic data quality target for the throughput. as follows: DL  Q req FCH req . Each layer is assigned a colour and displayed with intersections between layers. area is covered if  Q MAX  ic p   FCH   Q req  FCH . Each layer is assigned a colour and displayed with intersections between layers. ----c. For further details on formulas. TP DL . Each layer is assigned a colour and displayed with intersections between layers. terminal and throughput in the Services sub-folder. • Colour per maximum quality level (max Eb/Nt) Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction DL properties).ic) (with ic=icBS or icgiven). Then. defined in the Display tab DL DL (Prediction properties).0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Coverage consists of several layers with a layer per user-defined cell edge coverage probability.Atoll 3. TP DL . area is covered if P FCH  ic   Threshold . area is covered if . ic b pilot  . Each layer is assigned a colour and displayed with intersections between layers. Each layer is assigned a colour and displayed with intersections between layers. For each layer. For each layer. the total DL DL downlink traffic channel quality ( Q MAX  ic TP  ) is calculated after recombination. TPP – DL   AF FCH + 8  . see "Definitions and Formulas" on page 340. area is covered if  Q eff  ic   FCH  Threshold . This parameter is user-defined for a given service. • Colour per throughput This display option is relevant for CDMA2000 1xRTT data services only. P max P FCH  ic  = ---------------------FCH DL Q MAX  ic  max Where P FCH is a user-defined input for a given service and terminal. Best Nt server and active set determination is performed as in point prediction (AS analysis). p.2. • Colour per quality margin (Eb/Nt margin) Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction DL DL properties). TP P – DL   AF FCH + 4  . For each layer. • Colour per required power req Atoll calculates the downlink power required on FCH. Atoll calculates traffic channel quality at the receiver for each cell (k. TP DL Q MAX  ic DL DL TP   DL DL Q req  TP  DL . from this value. Downlink traffic channel quality at the receiver is evaluated from a maximum traffic channel power per transmitter allowed for the corresponding throughput. TP FCH DL FCH DL FCH DL FCH DL FCH DL ( TP P – DL  AF FCH . For each layer. it determines the effective downlink throughput received. Coverage consists of several layers with a layer per possible throughput. • Colour per effective quality level (Effective Eb/Nt) Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction DL properties). the obtained throughput ( TP obtained throughputs on each carrier. A services). i. A-capable terminals and EV-DO Rev.e. Nt 1xEV-DO Rev. DL ) in the downlink corresponds to the sum of the DL The obtained throughput ( TP  ic  ) on a carrier depends on the downlink transmission format. it Nt I req determines the downlink obtained throughput as follows: TP DL DL TP P – RLC  Index DL – Bearer  = -------------------------------------------------------------.e. The Early Termination Probability graph shows the probability of early termination ( p ) as a DL DL function of the number of retransmissions ( n Rtx ). 0 services). n TS The obtained throughput corresponds to the guaranteed throughput after a certain number of retransmissions (i. When HARQ (Hybrid Automatic Repeat Request) is used. i. ic BS b pilot    --- . B Users Single-carrier EV-DO Rev. the obtained throughput ( TP DL ) on downlink depends on the downlink transmission format.e the radio bearer index ( Index DL – Bearer ) with the number of timeslots ( n TS ). 0-capable terminals and EV-DO Rev. the required average number of retransmissions is smaller and the DL throughput is an average throughput ( TP av ) calculated as follows: DL TP P – RLC  Index DL – Bearer  DL TP av = --------------------------------------------------------------------DL  n Rtx (Index DL – Bearer. B service users. A Users For 1xEV-DO Rev. Atoll selects the downlink transmission format where ----. Atoll selects the downlink transmission format where E ----c.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 1xEV-DO Rev. TP DL is the value read in the graph “Peak throughput=f(C/I) (Rev0)” for the E calculated effective pilot quality level. ic b pilot    C ---  I  req and whose modulation scheme is supported by the terminal. n TS ).Atoll 3. ----c. the obtained throughput ( TP DL ) on downlink is determined from the graph “Peak throughput=f(C/I) (Rev0)” defined for the mobility type selected in the Condition tab (Prediction properties). 0 users (users with EV-DO Rev. ic BS b pilot  . For multi-carrier EV-DO Rev. Atoll calculates the average number of retransmissions (  n Rtx av ) as follows:  n DL   Rtx max  DL n DL DL DL DL  p  n Rtx  – p  n Rtx – 1    n Rtx =1 Rtx  n Rtx  av = -------------------------------------------------------------------------------------------DL p   n Rtx  max  1xEV-DO Rev.3. n TS ). It is determined as follows: DL TP P – RLC  Index DL – Bearer  DL TP  ic  = -------------------------------------------------------------n TS When HARQ (Hybrid Automatic Repeat Request) is used. A users (users with EV-DO Rev. the required average number of retransmissions is smaller and the DL throughput on a carrier is an average throughput ( TP av  ic  ) calculated as follows: 416 . the number of timeslots. A service users.n TS)  av DL The average number of retransmissions (  n Rtx  av ) is determined from early termination probabilities defined for the selected downlink transmission format. Nt The downlink obtained throughput corresponds to the guaranteed throughput after a certain number of retransmissions (i. 0 Users For 1xEV-DO Rev. Then. B service users are managed as 1xEV-DO Rev. the number of timeslots.e the radio bearer index ( Index DL – Bearer ) with the Ec C number of timeslots ( n TS ). 3 Uplink Service Area Analysis The results displayed when calculating the uplink service area analysis depend on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO).3.2. For each layer. the total uplink traffic channel quality (  Q MAX  ic   FCH ) is evaluated with respect to the receiver handover status. B users only. Then. see "Uplink Sub-Menu" on page 402. can be obtained. Each layer is assigned a colour and displayed with intersections between layers. TP • DL DL ). It enables you to view the obtained downlink DL throughput when HARQ is used. area is covered if ----c. area is covered if the average throughput. 5.n TS)  av DL The average number of retransmissions (  n Rtx av ) is determined from early termination probabilities defined for the selected downlink transmission format. TP av . Colour per average throughput This display option is available for 1xEV-DO Rev. For further details on calculations. Each layer is assigned a colour and displayed with intersections Nt between layers. Display Options You may choose the following display options: • Colour per C/I Coverage consists of several layers with a layer per quality threshold defined in the Display tab (Prediction properties). Each layer is assigned a colour and displayed with intersections between layers. Atoll displays uplink quality on FCH at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the receiver. No power control simulation is performed. Atoll calculates uplink quality on FCH from receiver for each cell (l. 5. For further details on formulas.2. Atoll determines uplink quality on FCH at the transmitter for the UL maximum terminal power. area is covered if the .Atoll 3. Coverage consists of several layers with a layer per possible average throughput ( TP av ). can be obtained. Atoll calculates the average number of retransmissions (  n Rtx  av ) as follows:  n DL   Rtx max  DL n DL DL DL DL  p  n Rtx  – p  n Rtx – 1    n Rtx =1 Rtx  n Rtx av = -------------------------------------------------------------------------------------------DL p   n Rtx  max  DL The average throughput ( TP av ) provided on downlink corresponds to the sum of the average throughputs obtained on each carrier. ic b pilot   Threshold .ic) (with ic=icBS or icgiven) in receiver active set. • Single colour 417 .5. For DL each layer. • Colour per throughput Coverage consists of several layers with a layer per possible throughput ( TP throughput.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 DL TP P – RLC  Index DL – Bearer  DL TP av  ic  = --------------------------------------------------------------------DL  n Rtx (Index DL – Bearer. see "Definitions and Formulas" on page 340.1 CDMA2000 1xRTT As in point analysis. A and 1xEV-DO Rev. Best server and active set determination is performed as in point prediction (AS analysis). The Early Termination Probability graph shows the probability of early termination ( p ) as a DL DL function of the number of retransmissions ( n Rtx ).5. For E each layer. • Colour per quality margin (Eb/Nt margin) Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction UL UL properties). For each layer. area is covered if P term – P term with intersections between layers. receiver is not completely defined and no mobility is assigned. Each layer is assigned a colour and displayed with intersections between layers. Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction FCH –re q properties).Atoll 3. • 418 Colour per throughput  ic   M arg in . For each layer. area is covered if P term  ic   Threshold . receiver is not completely defined and no service is assigned. For each layer. • Colour per probability This display option is available only if analysis is based on all simulations in a group (i. Coverage consists of several layers with a layer per UL UL user-defined mobility defined in Mobility sub-folder. area is covered if  Q MAX  ic p   FCH   Q req  FCH . area is covered if  Q MAX  ic   FCH   Q req  FCH . For each layer. Coverage colour is unique. area is covered if  Q MAX  ic   FCH  Threshold . There is a layer per transmitter with no intersection between layers. area is covered if UL UL  Q MAX  ic   FCH   Q req  FCH in the required number of simulations. • Colour per cell edge coverage probability Coverage consists of several layers with a layer per user-defined cell edge coverage probability.3. Each layer is assigned a colour and displayed with intersections between layers. P term . • Colour per maximum quality level (Max Eb/Nt) Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction UL properties). Layer colour is the colour assigned to best server transmitter. area is covered if  Q MAX  ic   FCH   Q req  FCH . Coverage consists of several layers with a layer per userdefined probability level defined in the Display tab (Prediction properties). Each layer is assigned a colour and displayed . area is covered if  Q effective  ic   FCH  Threshold . p. For each layer.e. defined in the Display tab UL UL (Prediction properties). • Colour per required power margin Coverage consists of several layers with a layer per user-defined power margin defined in the Display tab (Prediction max FCH –re q properties). if you select a group of simulations and the “All” option in the Condition tab of prediction properties). This parameter is user-defined for a given service and a terminal in the Services subfolder. Each layer is assigned a colour and displayed with intersections between layers. • Colour per effective quality level (Effective Eb/Nt) Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction UL properties). • Colour per required power FCH – re q The required terminal power. For each layer. • Colour per mobility In this case. Each layer is assigned a colour and displayed with intersections between layers. For each layer. Coverage consists of several layers with a layer per UL UL user-defined service defined in Services sub-folder. For each layer. Coverage consists of several layers with associated colours. Each layer is assigned a colour and displayed with intersections between layers. Each layer is assigned a colour and displayed with intersections between layers. For each layer. area is covered if  Q MAX  ic   FCH –  Q req  FCH  M arg in .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 UL UL UL Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH . Each layer is assigned a colour and displayed with intersections between layers. Each layer is assigned a colour and displayed with intersections between layers. • Colour per service In this case. • Colour per transmitter UL UL Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH .  Q req  FCH is the uplink data traffic quality target on the fundamental channel (FCH). is calculated as described in the Point analysis – AS analysis tab – Uplink sub-menu part. Then.2. area is covered if Q MAX  ic   Q req .6 kbps channel throughput. Atoll calculates the uplink quality from receiver for each cell (l. Q req  TP  is the uplink traffic data quality target for the throughput. A and 1xEV-DO Rev. 0 users. Each layer is assigned a colour and displayed with intersections between layers.ic) (with ic=icBS or icgiven) in receiver active set. TP UL UL UL UL UL . a given terminal and throughput in the service properties. For multi-carrier EV-DO users.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 This display option is relevant for CDMA2000 1xRTT data services only. • Colour per service In this case. Atoll considers the best sub-active set. For each layer. the total uplink quality ( Q MAX  ic  ) is evaluated with respect to the receiver handover status. TP P – UL   AF FCH + 16  ). There is a layer per transmitter with no intersection between layers. see "Definitions and Formulas" on page 340. For 1xEV-DO Rev.8 kbps channel throughput. TP P – UL   AF FCH + 8  . Q req is the quality required UL on uplink for a 9.6 kbps. B users. For 1xEV-DO Rev.Atoll 3. For each possible throughput. B terminals. This parameter is user-defined for the service. Coverage consists of several layers with a layer per UL UL user-defined mobility defined in Mobility sub-folder.e.2 CDMA2000 1xEV-DO As in point analysis. 5. TP P – UL   AF FCH + 4  . Best server and active set determination is performed as in point prediction (AS analysis). Atoll displays the uplink quality at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the receiver. 0 users. TP FCH UL FCH UL FCH UL FCH UL FCH UL ( TP P – UL  AF FCH . • Single colour UL UL UL Atoll displays a coverage if Q MAX  ic   Q req . Each layer is assigned a colour and displayed with UL UL intersections between layers. We have: E UL UL UL Q req =  ----c-  G p   1 + G ACK + G DRC + G TCH  for 1xEV-DO Rev. Coverage consists of several layers with a layer per UL UL user-defined service defined in Services sub-folder. For each layer.5. TP UL . N t min • Colour per transmitter UL UL Atoll displays a coverage if Q MAX  ic   Q req . Coverage consists of several layers with associated colours. area is covered if Q MAX  ic   Q req . For each layer. For further details on calculations. This parameter is calculated from the minimum uplink pilot quality and gains on the different uplink channels.3. For further details on formulas. For 1xEV-DO Rev. • Colour per probability This display option is available only if analysis is based on all simulations in a group (i. Coverage consists of several layers with a layer per possible throughput. Each layer is assigned a colour and displayed with intersections between layers. A and 1xEV-DO Rev. receiver is not completely defined and no mobility is assigned. Atoll determines the uplink quality at the transmitter for the maximum terminal power allowed and an UL uplink channel throughput of 4.8 kbps. • Colour per mobility In this case. Atoll determines the uplink quality at the transmitter for the maximum terminal power allowed and an uplink channel throughput of 9. Coverage colour is unique. see "Uplink Sub-Menu" on page 402. Atoll calculates the total UL UL uplink traffic channel quality ( Q MAX  ic TP  ). No power control simulation is performed. area is covered if Q MAX  ic TP   Q req  TP  . B users. UL TP P – UL   AF FCH + 2  . For 1xEV-DO Rev. 0 terminals. A and 1xEV-DO Rev.  N t min And E UL UL UL Q req =  ----c-  G p   1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary – Pilot  for 1xEV-DO Rev.3. Coverage consists of several layers with a layer per user- 419 . Q req is the quality required on uplink for a 4. if you select a group of simulations and the “All” option in the Condition tab of prediction properties). Layer colour is the colour assigned to best server transmitter. receiver is not completely defined and no service is assigned.  P max  Q MAX  ic   TCH = ---------------------term req P term With UL E UL UL  Q req  TCH =  ----c-  G p  G TCH  N t min Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction UL properties). Each layer is assigned a colour and displayed with intersections between layers. For each layer. • Colour per effective quality level (Effective Eb/Nt) Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction UL properties). 1xEV-DO Rev. area is covered if Q MAX  ic   Q req in the required number of simulations. • 420 Colour per required power margin corresponds . G TCH for 1xEV-DO Rev. 0 terminals. • Colour per cell edge coverage probability Coverage consists of several layers with a layer per user-defined cell edge coverage probability.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 UL UL defined probability level defined in the Display tab (Prediction properties). area is covered if  Q effective  ic   TCH  Threshold . p. Each layer is assigned a colour and displayed with intersections between layers. For each layer. TCH – re q For the selected configuration (i. G TCH for 1xEV-DO Rev. Each layer is assigned a colour and displayed with intersections between layers.Atoll 3. P term to the sum of the terminal powers required on each carrier of the configuration. P term . For each layer. area is covered if  Q MAX  ic   TCH  Threshold . the combination of carriers which provides the highest throughput). UL UL UL  Q effective  ic   TCH = min   Q MAX  ic   TCH  Q req  TCH  • Colour per quality margin (Eb/Nt margin) Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction UL UL properties). area is covered if P term  Threshold . Each layer is assigned a colour and displayed with intersections between layers. • Colour per maximum quality level (Max Eb/Nt) UL Here. TCH – re q P term req P term = --------------------------------------------------------. B service users Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction TCH – re q properties). 0. B service users For multi-carrier EV-DO users. Each layer is assigned a colour and displayed with intersections between layers.e. TCH –re q The required terminal power on traffic data channel.A and single-carrier 1xEV-DO Rev. 1 + G ACK + G DRC + G TCH And TCH – re q P term req P term . UL  Q req  TCH UL . • Colour per required power 1xEV-DO Rev. area is covered if Q MAX  ic p   Q req . For each layer. For each layer..3. is calculated as described in the Point analysis – AS analysis tab – Uplink sub-menu part. For each layer. A terminals. the coverage consists of several layers with a layer per user-defined power threshold defined TCH –re q in the Display tab (Prediction properties). Each layer is assigned a colour and displayed with intersections between layers. Each layer is assigned a colour and displayed with intersections between layers. = ------------------------------------------------------------------------------------------------------------------1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary – Pilot Multi-carrier 1xEV-DO Rev. For each layer. defined in the Display tab UL UL (Prediction properties). Atoll calculates the total uplink traffic channel quality (  Q MAX  ic   TCH ). area is covered if P term  ic   Threshold . area is covered if Q MAX  ic  – Q req  M arg in . It is calculated as follows: UL TP P – RLC  Index UL – Bearer  UL TP req = -------------------------------------------------------------n SF E UL UL UL Q req =  ----c-  G p   1 + G ACK + G DRC + G TCH + G RRI + G Auxiliary – pilot   N t min Where Ec UL  --- N t min is the minimum pilot quality level required on uplink to obtain the throughput. UL UL Q req  TP  is the uplink quality required to obtain the throughput. UL G ACK . Atoll calculates the total uplink quality ( Q MAX  ic TP  ). 0 tab). TP UL UL . G TCH .6.Atoll 3. They are defined in the terminal properties (1xEV-DO Rev.. 38. 1xEV-DO radio bearer ( Index UL – Bearer ) with a certain number of subframes ( n SF )) is used. TP UL UL UL . data rate control. is obtained when a certain uplink transmission format (i. one when the service uplink mode is "Low Latency" and another one for high capacity services. reverse rate indicator and auxiliary pilot channel gains relative to the pilot. 19. Coverage consists of several layers with a layer per possible throughput. B service users For multi-carrier 1xEV-DO Rev. data rate control and traffic data gains relative to the pilot.e. TP The possible throughputs on uplink. Multi-carrier 1xEV-DO Rev. .e. mobility and number of subframe). TP . For each layer. G RRI and G Auxiliary – pilot are respectively acknowledgement. G DRC and G TCH are respectively acknowledgement. Coverage consists of several UL UL UL UL layers with a layer per possible throughput. it selects the best configuration among all combinations of carriers. area is covered if Q MAX  ic TP   Q req  R v  . They are defined in the terminal properties (1xEVDO Rev. UL UL Q req  TP  is the uplink quality required to obtain the throughput. Then.8 and 153. Each layer is assigned a colour and displayed with intersections between layers. 1xEV-DO Rev. For each layer. For each layer. 76. B users. A tab). Atoll allocates the available terminal power to carriers sequentially and determines the uplink 1xEV-DO radio bearer obtained on each carrier. • Colour per throughput 1xEV-DO Rev.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Coverage consists of several layers with a layer per user-defined power margin defined in the Display tab (Prediction max req properties). TP The throughput. area is covered if TP UL UL  TP req . G ACK . B service users For each possible throughput. area is covered if Q MAX  ic TP   Q req  TP  . area is covered if P term – P term  ic   M arg in . Two values of G TCH are available. i. Atoll models load balancing between carriers. Each layer is assigned a colour and displayed with intersections between layers.2. For each layer. The value is defined in the UL 1xEV-DO Radio Bearer Selection (Uplink) table for the uplink transmission format (radio bearer Index. 0 service users For each possible throughput. Each layer is assigned a colour and displayed with intersections between layers. TP UL UL UL . G DRC . Coverage consists of several UL UL UL UL layers with a layer per possible throughput.4. the combination which provides the highest throughput. Two values are available for this parameter. Each layer is assigned a colour and displayed with intersections between layers. 421 . Atoll calculates the total uplink quality ( Q MAX  ic TP  ). one when the service uplink mode is "Low Latency" and another one for high capacity services. are: 9. A and single-carrier 1xEV-DO Rev. TP UL UL . traffic data channel.6 kbps E UL UL UL Q req =  ----c-  G p   1 + G ACK + G DRC + G TCH   N t min Where Ec  --- N t min is the minimum pilot quality level on uplink. This parameter is available in the Mobility types table.3. UL For the selected configuration (i. we have:  Ptot  icadj bpilot  DL DL N tot  ic  = txj j . TP av corresponds to the sum of the average throughputs obtained on each carrier of the configuration. B service users For multi-carrier 1xEV-DO Rev. the coverage consists of several layers with a layer per possible throughput. For CDMA2000 1xRTT systems. Each layer is assigned a colour and displayed with intersections between layers. TP UL • corresponds to the throughput of the best configuration. Colour per average throughput This display option is available for 1xEV-DO Rev. For each layer. A and single-carrier 1xEV-DO Rev.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 UL TP req is the uplink throughput associated with the layer. TP av .Atoll 3.3. The Early Termination Probability graph shows the probability of early termination ( p ) as a function of the number of retransmissions UL UL ( n Rtx ). i. B users. Atoll calculates the total uplink quality ( Q MAX  ic TP av  ).+ N0  Ptot  ic bpilot  + --------------------------------------------------RF  ic ic adj  DL term txj j term  N0  DL - Downlink noise rise. Atoll calculates the average number of retransmissions (  n Rtx  av ) as follows:  n UL   Rtx max  n UL UL UL UL UL  p  n Rtx  – p  n Rtx – 1    n Rtx =1 Rtx  n Rtx  av = -------------------------------------------------------------------------------------------UL p   n Rtx  max  1xEV-DO Rev. B service users UL UL UL For each possible average throughput. the combination which provides the highest throughput. B users only..e. A and 1xEV-DO Rev. When HARQ (Hybrid Automatic Repeat Request) is used. Multi-carrier 1xEV-DO Rev. NR DL  ic  . TP av . the combination of carriers which provides the highest throughput). UL TP req is the uplink throughput associated with the layer.e. For each UL UL layer. area is covered if TP av  TP req .5. Coverage consists of UL UL UL UL several layers with a layer per possible average throughput. Each layer is assigned a colour and displayed with intersections between layers. is calculated from the downlink total noise. UL UL UL Q req  TP av  is the uplink quality required to obtain the average throughput.4 Downlink Total Noise Analysis Atoll determines downlink total noise generated by cells. N tot . the required average number of retransmissions is smaller and the throughput is an average throughput UL ( TP av ) calculated as follows: UL TP P – RLC  Index UL – Bearer  UL TP av = ----------------------------------------------------------------------UL  n Rtx  Index UL – Bearer n SF   av UL The average number of retransmissions (  n Rtx  av ) is determined from early termination probabilities defined for the selected uplink transmission format (i. as: NR DL  ic  = – 10 log  ----------- N DL tot  422 .. area is covered if Q MAX  ic TP av   Q req  TP av  . the radio bearer index ( Index UL – Bearer ) with the number of subframes ( n SF )). we have:  Ptot  icadj  DL DL N tot  ic  = txj j + N0  Ptot  ic  + -----------------------------------RF  ic ic adj  DL term txj j For CDMA2000 1xEV-DO systems.2. 5.e. 5. allows the user to choose different displays. minimum or maximum) or any display per noise rise (average.6 Automatic Neighbour Allocation Atoll permits the automatic allocation of intra-technology neighbours in the current network. Each area is assigned a colour with intersections between areas. area is covered if min NR tot  ic   Threshold . Coverage consists of several areas with an area per user-defined ic noise rise threshold defined in the Display tab. • Colour per average noise level Coverage consists of several layers with a layer per user-defined noise level defined in the Display tab (Prediction properties).2.5. The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. The cells to be allocated will be called TBA cells. DL For each layer. • Colour per noise rise Atoll displays bins where NRDL  ic   Threshold .4. Coverage consists of several areas with an area per user-defined noise ic rise threshold defined in the Display tab. area is covered if max NR tot  ic   Threshold . minimum or maximum). They are located inside the focus zone. area is covered if N tot  ic   Threshold . Atoll determines DL total noise or DL noise rise on this carrier.4.2. 5.5. It means that all the cells of TBC transmitters of your . • Colour per minimum noise rise Atoll displays bins where min NR DL  ic   Threshold . area is covered if average NRtot  ic   Threshold . Atoll determines DL total noise for the best carrier. Coverage consists of several areas with an area per user-defined noise rise threshold defined in the Display tab. one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours. They must fulfill the following conditions: • • • They are active. DL For each layer. 423 . DL For each layer. Each layer is assigned a colour and displayed with intersections between layers. Then. • Colour per noise level Coverage consists of several layers with a layer per user-defined noise level defined in the Display tab (Prediction properties). Each area is assigned a colour with intersections between areas. Each layer is assigned a colour and displayed with ic intersections between layers.1 Analysis on the Best Carrier If the best carrier is selected. They satisfy the filter criteria applied to the Transmitters folder.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 5. In this case. Each layer is assigned a colour and displayed with intersections ic between layers. • Colour per maximum noise level Coverage consists of several layers with a layer per user-defined noise level defined in the Display tab (Prediction properties).Atoll 3. the displayed coverage is the same for any selected display per noise level (average.3. Coverage consists of several areas with an area per user-defined noise ic rise threshold defined in the Display tab.2 Analysis on a Specific Carrier When only one carrier is analysed. Each area is assigned a colour with intersections between areas. Each layer is assigned a colour and displayed with ic intersections between layers. DL For each layer. • Colour per average noise rise Atoll displays bins where average NRDL  ic   Threshold . • Colour per maximum noise rise Atoll displays bins where max NR DL  ic   Threshold . Two allocation algorithms are available. • Colour per minimum noise level Coverage consists of several layers with a layer per user-defined noise level defined in the Display tab (Prediction properties). Each area is assigned a colour with intersections between areas.atl document are potential neighbours. 0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks • ©Forsk 2015 They belong to the folder on which allocation has been executed. Otherwise. If the neighbours list of a cell is full.1 Neighbour Allocation for all Transmitters We assume that we have a reference cell A and a candidate neighbour. "Neighbour Allocation for a Group of Transmitters or One Transmitter" on page 427. If the distance between the reference cell and the candidate neighbour is greater than this value.ini file: [Neighbours] DoNotDeleteSymmetrics = 1 • • Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. If no focus zone exists in the ATL document. "Importance Calculation" on page 428. Therefore. Adjacence is relative to the number of pixels satisfying the criterion. If not selected. CellB is considered adjacent to CellA if there exists at least one pixel in the CellA Best Server coverage area where CellB is Best Server (if several cells have the same best server value) or CellB is the second best server that enters the Active Set (respecting the T_Drop of the allocation). • The calculation options: • Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. 424 . Atoll calculates the effective distance. adjacent cells are sorted and listed from the most adjacent to the least. 5. if the reference cell is a candidate neighbour of another cell.Atoll 3. Atoll deletes all the current neighbours and carries out a new neighbour allocation. the existing neighbours are kept. cell B. • If the Use Coverage Conditions check box is selected. depending on the above criterion. • Force symmetry: This option enables user to force the reciprocity of a neighbourhood link. Adjacency criterion: Let CellB be a candidate neighbour cell of CellA. there must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability.3. this one will be considered as candidate neighbour of the reference cell. you may force/forbid a cell to be candidate neighbour of the reference cell. Atoll takes into account the computation zone. the reference cell will not be added as a neighbour of that cell and that cell will be removed from the reference cell’s neighbours list. Therefore. In this section.6. Atoll checks following conditions: • The distance between both cells must be less than the user-definable maximum inter-site distance. • Force co-site cells as neighbours: This option enables you to force cells located on the reference cell site in the candidate neighbour list. For information on the effective distance calculation. Atoll will allocate neighbours to cells using the selected carriers. then the candidate neighbour is discarded. When automatic allocation starts. Only TBA cells may be assigned neighbours.This constraints can be weighted among the others and ranks the neighbours through the importance field (see after). This constraints can be weighted among the others and ranks the neighbours through the importance field (see after). This folder can be either the Transmitters folder or a group of transmitters or a single transmitter. Delete existing neighbours: When selecting the Delete existing neighbours option. You may choose one or more carriers. the following are explained: • • • "Neighbour Allocation for all Transmitters" on page 424. When the Force adjacent cells as neighbours check box is selected. You can force Atoll to keep that cell in the reference cell’s neighbours list by adding the following option in the Atoll. • Force adjacent cells as neighbours (only for intra-carrier neighbours): This option enables you to force cells geographically adjacent to the reference cell in the candidate neighbour list. only the distance criterion is taken into account. which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. see "Calculation of the Inter-Transmitter Distance" on page 430. SB is the area where: • • The pilot signal received from the cell B is greater than the minimum pilot signal level.3. The pilot signal from A is the highest one The pilot signal from A is lower than the minimum pilot signal level plus the margin.Atoll 3. SA is the area where: • • The pilot signal received from the cell A is greater than the minimum pilot signal level. It means that the cell A is the first one in the active set. It is strictly lower than the best pilot signal received and higher than the best pilot signal minus the margin.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 The overlapping zone ( S A  S B ) is defined as follows: • Intra-carrier neighbours: intra-carrier handover is a soft handover. 1st case: the reference cell A is located inside a continuous layer of cells with carrier c1 (c1 is the selected carrier on which you run the allocation) and the candidate cell B belongs to a layer of cells with carrier c2. The reference cell A and the candidate cell B are located inside a continuous layer of cells with carrier c1 (c1 is the selected carrier on which you run the allocation). It is needed in a multi-carrier (1xRTT and 1xEVDO carriers) CDMA network: • • To balance loading between carriers and layers (1st case). The pilot signal from B is the highest one. Figure 5. SB is the area where the cell B can enter the active set. Ec/I0.1st Case 2nd case: the reference cell A is located on the border of a layer with carrier c1 (c1 is the selected carrier on which you run the allocation) and the candidate cell B belongs to a layer of cells with carrier c2. The pilot quality from B is greater than T_Drop. SA is the area where: • • • The pilot signal received from the cell A is greater than the minimum pilot signal level. The pilot signal from B is the highest one. • • • The pilot signal received from the cell A is greater than the minimum pilot signal level. • • • The pilot signal received from the cell B is greater than the minimum pilot signal level. SB is the area where: • • The pilot signal received from the cell B is greater than the minimum pilot signal level. To make a coverage reason handover from micro cell frequency to macro cells (2nd case). The pilot signal from A is not the highest one. SA is the area where the cell A is the best serving cell. The pilot quality from A is the best. 425 . The pilot quality from A exceeds Min. Inter-carrier neighbours: inter-frequency handover is a hard handover.4: Overlapping Zones . In addition. For information on the importance calculation. Atoll displays the percentage of area meeting the coverage conditions and the corresponding surface area (km2).e. SA  SB . Among these 15 candidate neighbours. Atoll takes into account the pilot power of the cell. the candidate neighbour B is discarded.2nd Case Two ways enable you to determine the I0 value: • • Global Value: A percentage of the cell maximum power is considered. i. Therefore. If the % of maximum power is too low. 100 ) and compares this value to the % minimum covered area. I0 represents the sum of values calculated for each cell. if %  Pmax  P pilot . coverage or symmetric. • The importance of neighbours. if cells have previous allocations in the list. the percentage of area meeting the adjacency conditions and the corresponding surface area (km2). Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be allocated to each transmitter is exceeded. 426 . Atoll provides the list of neighbours. If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8.3. I0 represents the sum of total transmitted powers. a neighbour may be marked as exceptional pair. neighbours are marked as existing. it indicates the importance (in %) of each neighbour and the allocation reason. In the Results part. Finally. see "Importance Calculation" on page 428. Then. Defined per Cell: Atoll takes into account the total downlink power defined per cell.5: Overlapping Zones .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Figure 5. this value is taken into account instead of the default one available in the Neighbour Allocation dialogue. If Atoll calculates the percentage of covered area ( ----------------SA this percentage is not exceeded. Importance values are used by the allocation algorithm to rank the neighbours. Note that specific maximum numbers of neighbours (maximum number of intra-carrier neighbours.Atoll 3. adjacency and coverage reasons. the number of neighbours and the maximum number of neighbours allowed for each cell. only 8 (having the highest importance values) will be allocated to the reference cell. For neighbours accepted for co-site. adjacent. The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see after). co-site. maximum number of inter-carrier neighbours) can be defined at the cell level (property dialogue or cell table). If defined there. A forbidden neighbour must not be listed as neighbour except if the neighbourhood relationship already exists and the Delete existing neighbours option is unchecked when you start the new allocation. As a consequence. To avoid that.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 • By default. It will be the last one. Automatic neighbour allocation parameters are described in "Neighbour Allocation for all Transmitters" on page 424. Thus. it will not appear in the Results table. 2nd case: The cell B neighbour list is full: Atoll will not include cell A in the list and will cancel the link by deleting cell B from the cell A neighbour list. Atoll automatically calculates the path loss matrices if not found. Atoll considers the constraints between exceptional pairs in both directions so as to respect symmetry condition. On the other hand. The neighbour lists may be optionally used in the power control simulations to determine the mobile's active set. You can force Atoll to compare the Max inter-site distance with the real inter-site distance by adding the following lines in Atoll. you can force Atoll to prioritise the individual distances between reference cells and their respective neighbour candidates by adding the following lines in Atoll. it is interesting to know that the algorithm works such as finding the maximum number of neighbours by selection the multi-service traffic data as follows: Service: selection of the one with the lowest body loss. In this case.Loss) value. In this case. Neighbours of TBA cells marked as exceptional pair. if the cell B is a neighbour of the cell A while the cell A is not a neighbour of the cell B. When the options “Force exceptional pairs” and “Force symmetry” are selected. Mobility: no impact on the allocation. Terminal: selection of the one with the greatest (Gain . In the Results. if a TBA cell has already reached its maximum number of neighbours before starting the new allocation. This reciprocity is allowed only if the neighbour list is not already full. Atoll allocates neighbours to: • • • TBA cells.ini: [Neighbours] CandidatesMaxDistanceInImportanceCalculation=1 • • • • • • • • • • • • No simulation or prediction study is needed to perform an automatic neighbour allocation.2 Neighbour Allocation for a Group of Transmitters or One Transmitter In this case. because the effective distance is smaller. When starting an automatic neighbour allocation. the automatic neighbour allocation compares the defined Max intersite distance with the effective inter-cell distance. there can be cases where the real distance between assigned neighbours is higher than the Max inter-site distance. adjacent and symmetric.ini: [Neighbours] RealInterSiteDistanceCondition=1 • By default. the neighbour importance calculated with respect to distance is based on the global Max inter-site distance setting for all neighbour candidates. Neighbours of TBA cells that satisfy coverage conditions. no specific selection.3. symmetry cannot be respected.Atoll 3.6. Atoll displays a warning in the Event viewer. Atoll displays only the cells for which it finds new neighbours. 5. and. As a consequence. mobility or service is selected in the automatic allocation. two cases are possible: 1st case: There is space in the cell B neighbour list: the cell A will be added to the list. if equal. the one with the lowest noise figure. The force neighbour symmetry option enables the users to consider the reciprocity of a neighbourhood link. if neighbourhood relationship is forced in one direction and forbidden in the other one. Therefore. 427 . there can be cases where the calculated importance is different when the global Max inter-site distance is modified. Atoll displays a warning in the Event viewer indicating that the constraint on the forbidden neighbour will be ignored by algorithm because the neighbour already exists. Even if no specific terminal. 6. priority assigned to each neighbourhood cause is determined using the Importance Function (IF). Neighbourhood cause When Importance value Existing neighbour Only if the Delete existing neighbours option is not selected and in case of a new allocation Existing importance Exceptional pair Only if the Force exceptional pairs option is selected 100 % Co-site cell Only if the Force co-site cells as neighbours option is selected Importance Function (IF) Adjacent cell Only if the Force adjacent cells as neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions Only if the % minimum covered area is exceeded Importance Function (IF) Symmetric neighbourhood relationship Only if the Force neighbour symmetry option is selected Importance Function (IF) Except the case of forced neighbours (importance = 100%).6. d Di  = 1 – ---------d max d is the effective distance (in m). d max is the maximum distance between the reference transmitter and a possible neighbour. For information on the effective distance calculation.3 Importance Calculation Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason. 5.1 Importance of Intra-carrier Neighbours The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf. It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. The overlapping factor (O): the percentage of overlapping.Atoll 3. The minimum and maximum importance assigned to each of the above factors can be defined. The adjacency factor (A): the percentage of adjacency. and to quantify the neighbour importance. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 30% Adjacency factor (A) Min(A) 30% Max(A) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The Importance Function is evaluated as follows: Neighbourhood cause Importance Function Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%Max(Di))(O)}+Min(Di)+Delta(Di)(Di) 10%+20%{10%(Di)+90%(O)}+1%+9%(Di) No Yes Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 30%+30%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Yes Yes Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 60%+40%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Co-site Adjacent No Where: Delta(X)=Max(X)-Min(X) 428 .0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 5.3. table below). The IF considers the following factors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter.3. see "Calculation of the Inter-Transmitter Distance" on page 430. • • • The co-site factor (C): a Boolean. this value varies between 0 and 100%. The IF considers the following factors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. For information on the effective distance calculation. 5. this value varies between 0 to 100%. the neighbour importance depends on the distance and on the neighbourhood cause. The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The IF evaluates importance as follows: Co-site Neighbourhood cause IF Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)} Yes Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+ Max(O)(O)/(Max(Di)+Max(O))} 60%+40%{1/7%(Di)+6/7%(O)} Where Delta(X)=Max(X)-Min(X) 429 .3. • • The co-site factor (C): a Boolean. d max is the maximum distance between the reference transmitter and a possible neighbour. Neighbourhood cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site cell If the Force co-site cells as neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions If the % minimum covered area is exceeded Importance Function (IF) Symmetric neighbourhood relationship If the Force neighbour symmetry option is selected Importance Function (IF) Except the case of forced neighbours (importance = 100%). If the Min and Max value ranges of the importance function factors do not overlap. There can be a mix of the neighbourhood causes. neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping. priority assigned to each neighbourhood cause is determined using the Importance Function (IF). neighbours will be ranked in this order: co-site neighbours.2 Importance of Inter-carrier Neighbours As indicated in the table below. the neighbours will be ranked by neighbour cause.Atoll 3. The overlapping factor (O): the percentage of overlapping.3. With a value of Min(O) = 0%. If the Min and Max value ranges of the importance function factors overlap.6. d Di  = 1 – ---------d max d is the effective distance (in m). The IF is user-definable using the Min importance and Max importance fields. adjacent neighbours. It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. the neighbours may be ranked differently. With the default values for minimum and maximum importance fields. and neighbours allocated based on coverage overlapping.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 • • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. see "Calculation of the Inter-Transmitter Distance" on page 430. the neighbours may be ranked differently.4. neighbours will be ranked in this order: co-site neighbours and neighbours allocated based on coverage overlapping. There are a maximum of 512 PN offsets numbered from 0 to 511. There can be a mix of the neighbourhood causes. 5. Atoll takes into account the computation zone. They belong to the folder on which allocation has been executed.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks • • • • ©Forsk 2015 Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. If no focus zone exists in the . the neighbours will be ranked by neighbour cause. They satisfy the filter criteria applied to the Transmitters folder. Mobile processes the strongest received PN sequence and reads its phase that identifies the cell. The cells to which Atoll allocates PN offsets are referred to as the TBA cells (cells to be allocated). Figure 5. TBA cells fulfil following conditions: • • • • They are active. This folder can be either the Transmitters folder or a group of transmitters or a single transmitter.1 Calculation of the Inter-Transmitter Distance Atoll takes into account the real distance ( D in m) and azimuths of antennas in order to calculate the effective intertransmitter distance ( d in m).4 Appendices 5.6: Inter-Transmitter Distance Computation The formula above implies that two cells facing each other will have a smaller effective distance than the real physical distance. If the Min and Max value ranges of the importance function factors overlap. d = D   1 + x  cos  – x  cos   where x = 0.atl document. It is this effective distance that will be taken into account rather than the real distance. It is a time offset used by a cell to shift a Pseudo Noise sequence.6.6. If the Min and Max value ranges of the importance function factors do not overlap. neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping. 430 . With a value of Min(O) = 0%.7 PN Offset Allocation PN offset is used to identify a cell. 5.3. With the default values for minimum and maximum importance fields.Atoll 3.3% so that the maximum D variation does not exceed 1%. The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. They are located inside the focus zone. 7. • Neighbourhood between cells. 431 . Second order neighbours: The neighbours of neighbours. • • • • In the context of the PN offset allocation. Cells fulfilling a criterion on Ec/I0 (option “Additional Ec/I0 conditions”). If this value is not defined. You may consider: • • • First order neighbours: The neighbours of TBA cells listed in the Intra-technology neighbours table.12. the term "PN-cluster" refers to a sub-group of PN offsets. A PN offset cannot be reused at a site that is not at least as far away as the reuse distance from the site allocated with the particular PN offset. [4. • Co-PN Reuse distance. In this case. Atoll considers all the cells “B” that can enter the active set on the area where the reference cell is the best server (area where (Ec/I0)A exceeds Min.3. In 3GPP2 multi-RAT documents.Atoll 3. Atoll considers symmetry relationship between a cell. Third order neighbours: The neighbour’s neighbour’s neighbours. all PN offsets from 4 to 508 with a separation interval of 4 can be allocated). Atoll also tries to allocate different PN offsets to CDMA cells that are neighbours of a common LTE cell.g. The first PN offset is PILOT_INC and other ones are multiples of this value. If PILOT_INC is set to 4. and to calculate their importance.8.. the pool of possible PN offsets consists of PN offsets from 4 to 508 with a separation interval of 4 (i. Ec/I0 and is the highest one and (Ec/I0)B exceeds T_Drop). the existing PN offsets are kept.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 5. the term "neighbours" refers to intra-carrier neighbours. When this parameter is not specified in the cell properties. In this case. PN offset reuse distance can be defined at cell level. Atoll uses this parameter to determine the pool of possible PN offsets (512 divided by PILOT_INC value). or it allocates the same PN offset to each carrier of a transmitter if the option "Allocate carriers identically" is selected. • • • PN-cluster size. Atoll deletes all the current PN offsets and carries out a new PN offset allocation. The "Delete All Codes" option: When selecting this option. Within the context of PN offset allocation. its second order neighbours and its third order neighbours. Atoll reuses the intra-carrier neighbour allocation algorithm to determine the list of cells which cannot be allocated the same scrambling code.7.. When no domain is assigned to cells. then Atoll will use the default reuse distance defined in the PN offset Automatic Allocation dialogue..1. Reuse distance is a constraint on the allocation of PN offsets.. For example: When PILOT_INC is set to 4.508]). If not selected. Atoll considers either a percentage of the cell maximum powers or the total downlink power used by the cells in order to evaluate I0. For a reference cell “A”.1 Automatic Allocation Description 5.e. Atoll uses 50% of the maximum power.16. either Atoll independently plans PN offsets for the different carriers. Exceptional pairs. Domains of PN offsets.1 Options and Constraints The PN offset allocation algorithm can take into account following constraints and options: • PILOT_INC parameter. The possibility to use a maximum of PN offsets (option "Use a Maximum of PN Offsets"): Atoll will try to spread the PN offset spectrum the most. I0 equals the sum of total transmitted powers. its first order neighbours. • • • The carrier on which the allocation is run: It can be a given carrier or all of them. Atoll considers the PILOT_INC parameter only to determine available PN offsets (e. Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 In addition, it depends on the selected allocation strategy. Allocation strategies can be: • • • PN offset per cell: The purpose of this strategy is to reduce the spectrum of allocated PN offsets the maximum possible. Atoll will allocate the first possible PN offsets in the domain. Adjacent PN-Clusters per site: This strategy consists of allocating one cluster of adjacent PN offsets to each site, then, one PN offset of the cluster to each cell of each transmitter according to its azimuth. When all the clusters have been allocated and there are still sites remaining to be allocated, Atoll reuses the clusters at another site. Distributed PN-clusters per site: This strategy consists of allocating one cluster of PN offsets to each site in the network, then, one PN offset of the cluster to each cell of each transmitter according to its azimuth. With this strategy, the cluster is made of PN offsets separated as much as possible. When all the clusters have been allocated and there are still sites remaining to be allocated, Atoll reuses the clusters at another site. In the Results table, Atoll only displays PN offsets allocated to TBA cells. 5.7.1.2 Allocation Process For each TBA cell, Atoll lists all cells which have constraints with the cell. They are referred to as near cells. The near cells of a TBA cell may be: • • • • • • Its neighbour cells: the neighbours listed in the Intra-technology neighbours table (options “Existing neighbours” and "First Order"), The neighbours of its neighbours (options “Existing neighbours” and “Second Order”), The third order neighbours (options “Existing neighbours” and “Third Order”), The cells that fulfil Ec/I0 condition (option “Additional Ec/I0 conditions”), The cells with distance from the TBA cell less than the reuse distance, The cells that make exceptional pairs with the TBA cell. One additional constraint is considered in 3GPP2 multi-RAT documents: • The cell and its near cells are neighbours of the same LTE cell. These constraints have a certain weight taken into account to determine the TBA cell priority during the allocation process and the cost of the PN offset plan. During the allocation, Atoll tries to assign different PN offsets to the TBA cell and its near cells. If it respects all the constraints, the cost of the PN offset plan is 0. When a cell has too many constraints and there are not anymore PN offsets available, Atoll breaks the constraint with the lowest cost so as to generate the PN offset plan with the lowest cost. For information on the cost generated by each constraint, see "Cell Priority" on page 433. 5.7.1.2.1 Single Carrier Network The allocation process depends on the selected strategy. Algorithm works as follows: Strategy: PN offset per cell Atoll processes TBA cells according to their priority. It allocates PN offsets starting with the highest priority cell and its near cells, and continuing with the lowest priority cells not allocated yet and their near cells. For information on calculating cell priority, see "Cell Priority" on page 433. Strategy: Adjacent PN-Clusters per site All sites which have constraints with the studied site are referred to as near sites. Atoll assigns a PN-cluster of adjacent PN offsets to each site, starting with the highest priority site and its near sites, and continuing with the lowest priority sites not allocated yet and their near sites. When all the clusters have been allocated and there are still sites remaining to be allocated, Atoll reuses the clusters at another site. When the Co-PN Reuse Distance option is selected, the algorithm reuses the clusters as soon as the Co-PN reuse distance is exceeded. Otherwise, when the option is not selected, the algorithm tries to assign reused clusters as spaced out as possible. Then, Atoll allocates a PN offset from the cluster to each cell of each transmitter located on the sites according to the transmitter azimuth. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet and their near cells. For information on calculating site priority, see "Site Priority" on page 436. For information on calculating cell priority, see "Cell Priority" on page 433. Strategy: Distributed PN-Clusters per site All sites which have constraints with the studied site are referred to as near sites. Atoll assigns one cluster to each site, starting with the highest priority site and its near sites, and continuing with the lowest priority sites not allocated yet and their near sites. When all the clusters have been allocated and there are still sites remaining to be allocated, Atoll reuses the clusters at another site. When the Co-PN Reuse Distance option is selected, the algorithm reuses the clusters as soon as the Co-PN reuse distance is exceeded. Otherwise, when the option is not selected, the algorithm tries to assign reused clusters as spaced out as possible. 432 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Then, Atoll assigns a PN offset from the cluster to each cell of each transmitter located on the sites according to the transmitter azimuth. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet and their near cells. For information on calculating site priority, see "Site Priority" on page 436. For information on calculating cell priority, see "Cell Priority" on page 433. 5.7.1.2.2 Multi-Carrier Network In case you have a multi-carrier network and you run the PN offset allocation on all the carriers, the allocation process depends on wether the option "Allocate Carriers Identically" is selected or not. When the option is not selected, algorithm works for each strategy, as explained above. On the other hand, when the option is selected, allocation order changes. It is no longer based on the cell priority but depends on the transmitter priority. All transmitters which have constraints with the studied transmitter will be referred to as near transmitters. In case of a "Per cell" strategy (PN offset per cell), Atoll starts PN offset allocation with the highest priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. The same PN offset is assigned to each cell of the transmitter. In case of a "Per site" strategy (Adjacent and Distributed PN-clusters per site strategies), Atoll assigns a cluster to each site and then, allocates a PN offset to each transmitter. It starts with the highest priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. The same PN offset is assigned to each cell of the transmitter. For information on calculating cell priority, see "Cell Priority" on page 433. For information on calculating transmitter priority, see "Transmitter Priority" on page 435. When cells, transmitters or sites have the same priority, processing is based on an alphanumeric order. 5.7.1.2.3 Difference between Adjacent and Distributed PN-Clusters The following example explains the difference between "Adjacent PN-clusters" and "Distributed PN-clusters". The PILOT_INC has been set to 4 and the PN-cluster size to 3. There are: • • 128 PN offsets that can be allocated: they are all PN offsets from 4 to 508 with a separation interval of 4. Each PN-cluster consists of three PN offsets. So, there are 42 PN-clusters available. If you select "Adjacent PN-cluster per site" as allocation strategy, Atoll will consider PN-clusters consisted of adjacent PN offsets (e.g. {4,8,12}, {16,20,24}, ...,{496,500,504}). If you select "Distributed PN-cluster per site" as allocation strategy, Atoll will consider PN-clusters consisted of PN offsets separated as much as possible (e.g. {4,172,340}, {8,176,344}, ...,{168,336,504}). 5.7.1.3 Priority Determination 5.7.1.3.1 Cell Priority PN offset allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities assigned to cells depend upon how much constrained each cell is and the cost defined for each constraint. A cell without any constraint has a default cost, C , equal to 0. The higher the cost on a cell, the higher the priority it has for the PN offset allocation process. There are five criteria employed to determine the cell priority: PN Offset Domain Criterion The cost due to the domain constraint, C i  Dom  , depends on the number of PN offsets available for the allocation. The domain constraint is mandatory and cannot be broken. When no domain is assigned to cells, 512 PN offsets are available and we have: C i  Dom  = 0 When domains of PN offsets are assigned to cells, each unavailable PN offset generates a cost. The higher the number of codes available in the domain, the less will be the cost due to this criterion. The cost is given as: C i  Dom  = 512 – Number of PN Offsets in the domain 433 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Distance Criterion The constraint level of any cell i depends on the number of cells (j) present within a radius of "reuse distance" from its centre. The total cost due to the distance constraint is given as:  Cj  Dist  i   C i  Dist  = j Each cell j within the reuse distance generates a cost given as: C j  Dist  i   = w  d ij   c dis tan ce Where w  d ij  is a weight depending on the distance between i and j. This weight is inversely proportional to the inter-cell distance. For a reuse distance of 2000m, the weight for an inter-cell distance of 1500m is 0.25, the weight for co-site cells is 1 and the weight for two cells spaced out 2100m apart is 0. c dis tan ce is the cost of the distance constraint. This value can be defined in the Constraint Cost dialogue. • Exceptional Pair Criterion The constraint level of any cell i depends on the number of exceptional pairs (j) for that cell. The total cost due to exceptional pair constraint is given as: C i  EP  =  cEP  i – j  j Where c EP is the cost of the exceptional pair constraint. This value can be defined in the Constraint Cost dialogue. Neighbourhood Criterion The constraint level of any cell i depends on the number of its neighbour cells j, the number of second order neighbours k and the number of third order neighbours l. Let’s consider the following neighbour schema: Figure 5.7: Neighbourhood Constraints The total cost due to the neighbour constraint is given as:  Ci  N  =         Cj  N1  i   +  Cj – j  N1  i   +   Ck  N2  i   +  Ck – k  N2  i   +   Cl  N3  i   +  Cl – l  N3  i   j j k k l l Each first order neighbour cell j generates a cost given as: C j  N1  i   = I j  c N1 Where I j is the importance of the neighbour cell j. c N1 is the cost of the first order neighbour constraint. This value can be defined in the Constraint Cost dialogue. Because two first order neighbours must not have the same PN offset, Atoll considers the cost created by two first order neighbours to be each other. C j  N1  i   + C j  N1  i   C j – j  N1  i   = ----------------------------------------------------2 Each second order neighbour cell k generates a cost given as: 434 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 C k  N2  i   = Max ( C j  N1  i    C k  N1  j   , C j  N1  i    C k  N1  j   )  c N2 Where c N2 is the cost of the second order neighbour constraint. This value can be defined in the Constraint Cost dialogue. Because two second order neighbours must not have the same PN offset, Atoll considers the cost created by two second order neighbours to be each other. C k  N2  i   + C k  N2  i   C k – k  N2  i   = ------------------------------------------------------2 Each third order neighbour cell l generates a cost given as:  C  N1  i    C k  N1  j    C l  N1  k   C j  N1  i    C k  N1  j    C l N1  k   C l  N3  i   = Max  j   c N3   C j  N1  i    C k  N1  j     C l N1  k  C j  N1  i    C k  N1  j    C l N1  k   Where c N3 is the cost of the third order neighbour constraint. This value can be defined in the Constraint Cost dialogue. Because two third order neighbours must not have the same PN offset, Atoll considers the cost created by two third order neighbours to be each other. C l  N3  i   + C l  N3  i   C l – l  N3  i   = ----------------------------------------------------2 Atoll considers the highest cost of both links when a neighbour relation is symmetric and the importance value is different. . In this case, we have: C j  N1  i   = Max  I i – j I j – i   c N1 And C k  N2  i   = Max (C j  N1  i    C k  N1  j  ,C j  N1  k    C i  N1  j  )  c N2 LTE Neighbour Criterion This criterion is considered in 3GPP2 multi-RAT documents. If the cell i is neighbour of an LTE cell, the cell constraint level depends on how many cells j are neighbours of the same LTE cell. The total cost due to LTE neighbour constraint is given as: C i  N LTE  =  cNLTE  j – TxLTE  j Where cN LTE is the cost of the LTE neighbour constraint. This value can be defined in the Constraint Cost dialogue. Therefore, the total cost due to constraints on any cell i is defined as: C i = C i  Dom  + C i  U  With C i  U  = C i  Dist  + C i  EP  + C i  N  + C i  N 2G  5.7.1.3.2 Transmitter Priority In case you have a multi-carrier network and you run PN offset allocation on "all" the carriers with the option "allocate carriers identically", algorithm in Atoll allots priorities to transmitters. Priorities assigned to transmitters depend on how much constrained each transmitter is and the cost defined for each constraint. The higher the cost on a transmitter, the higher the priority it has for the PN offset allocation process. Let us consider a transmitter Tx with two cells using carriers 0 and 1. The cost due to constraints on the transmitter is given as: C Tx = C Tx  Dom  + C Tx  U  435 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Max  C  U   and C  Dom  = 512 – Number of PN offsets in the domain Tx i  Tx i With C Tx  U  = Here, the domain available for the transmitter is the intersection of domains assigned to cells of the transmitter. The domain constraint is mandatory and cannot be broken. 5.7.1.3.3 Site Priority In case of "Per Site" allocation strategies (Adjacent PN-clusters per site and Distributed PN-clusters per site), algorithm in Atoll allots priorities to sites. Priorities assigned to sites depend on how much constrained each site is. The higher the constraint on a site, the higher the priority it has for the PN offset allocation process. Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. The site constraint is given as: C S = C S  U  + C S  Dom  With, C S  U  =  CTx  U  , and CS  Dom  = 512 – Number of PN offsets in the domain Tx Here, the domain considered for the site is the intersection of domains available for transmitters of the site. 5.7.2 Allocation Examples In order to understand the differences between the different allocation strategies and the behaviour of the algorithm when using a maximum of PN offsets or not, let us consider the following sample scenario: Figure 5.8: PN Offset Allocation Let Site0, Site1, Site2 and Site3 be four sites with 3 cells using carrier 0 whom PN offsets have to be allocated. The PILOT_INC parameter has been set to 4 and the PN Cluster Size is 3. Therefore, all PN offsets from 4 to 508 with a separation interval of 4 can be allocated. The reuse distance is supposed to be lower than the inter-site distance. Only co-site neighbours exist and all of them have the same importance. The following section lists the results of each combination of options with explanation where necessary. 5.7.2.1 Strategy: PN Offset per Cell Since the restrictions of neighbourhood only apply to co-sites with the same importance and inter-site distances are greater than reuse distances, every cell has the same priority. Then, the PN offset allocation to cells is performed in an alphanumeric order. 436 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Without ‘Use a Maximum of PN Offsets’ With ‘Use a Maximum of PN Offsets’ Atoll allocates the first three PN offsets in the domain (4, 8 and 12) to the Site0’s cells. Under given constraints of neighbourhood and reuse distance, same PN offsets can be allocated to each site’s cells. Atoll allocates the first three PN offsets in the domain (4, 8 and 12) to the Site0’s cells. As it is allowed to use a maximum of PN offsets, Atoll allocates different PN offsets to each site’s cells so that there is least repetition. 5.7.2.2 Strategy: Adjacent PN-Clusters Per Site Since the restrictions of neighbourhood only apply to co-sites with the same importance and inter-site distances are greater than reuse distances, every cell has the same priority. Then, the PN offset allocation to cells is performed in an alphanumeric order. Without ‘Use a Maximum of PN Offsets’ With ‘Use a Maximum of PN Offsets’ Atoll allocates a PN cluster of adjacent PN offsets to Site0 and As it is possible to use a maximum of PN offsets, Atoll then, one PN offset of the PN cluster to each cell. Under given allocates different PN clusters of adjacent PN offsets to sites constraints of neighbourhood and reuse distance, the same so that there is least repetition of PN offsets. PN cluster can be allocated to each site and same PN offsets to each site’s cells. 5.7.2.3 Strategy: ‘Distributed PN-Clusters Per Site Since the restrictions of neighbourhood only apply to co-sites with the same importance and inter-site distances are greater than reuse distances, every cell has the same priority. Then, the PN offset allocation to cells is performed in an alphanumeric order. 437 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks Without ‘Use a Maximum of PN Offsets’ ©Forsk 2015 With ‘Use a Maximum of PN Offsets’ Atoll allocates a PN cluster of distributed PN offsets to Site0 As it is possible to use a maximum of PN offsets, Atoll and then, one PN offset of the PN cluster to each cell. Under allocates different PN clusters of distributed PN offsets to given constraints of neighbourhood and reuse distance, the sites so that there is least repetition of PN offsets. same PN cluster can be allocated to each site and same PN offsets to each site’s cells. 5.8 Automatic GSM-CDMA Neighbour Allocation 5.8.1 Overview You can automatically calculate and allocate neighbours between GSM/TDMA and CDMA2000 networks. In Atoll, it is called inter-technology neighbour allocation. Inter-technology handover is used in two cases: • • When the CDMA coverage is not continuous. In this case, the CDMA coverage is extended by CDMA-GSM handover into the GSM network, And in order to balance traffic and service distribution between both networks. Note that the automatic inter-technology neighbour allocation algorithm takes into account both cases. In order to be able to use the inter-technology neighbour allocation algorithm, you must have: • • An .atl document containing the GSM/TDMA network, GSM.atl, and another one containing the CDMA2000 network, CDMA.atl, An existing link on the Transmitters folder of GSM.atl into CDMA.atl. The external neighbour allocation algorithm takes into account all the GSM TBC transmitters. It means that all the TBC transmitters of GSM.atl are potential neighbours. The cells to be allocated will be called TBA cells which, being cells of CDMA.atl, fulfill following conditions: • • • • They are active, They satisfy the filter criteria applied to Transmitters folder, They are located inside the focus zone, They belong to the folder for which allocation has been executed. This folder can be either the Transmitters folder or a group of transmitters subfolder. Only CDMA TBA cells may be assigned neighbours. 5.8.2 Automatic Allocation Description The allocation algorithm takes into account criteria listed below: • • • • The inter-transmitter distance, The maximum number of neighbours fixed, Allocation options, The selected allocation strategy, Two allocation strategies are available: the first one is based on distance and the second one on coverage overlapping. We assume we have a CDMA reference cell, A, and a GSM candidate neighbour, transmitter B. 438 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 5.8.2.1 Algorithm Based on Distance When the automatic allocation starts, Atoll checks the following conditions: • The distance between the CDMA reference cell and the GSM neighbour must be less than the user-definable maximum inter-site distance. If the distance between the CDMA reference cell and the GSM neighbour is greater than this value, then the candidate neighbour is discarded. Atoll calculates the effective distance, which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter Distance" on page 430. • The calculation options, Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or more carriers. Atoll will allocate neighbours to cells using the selected carriers. Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site as the reference CDMA cell in the candidate neighbour list. This option is automatically selected. Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you may force/forbid a GSM transmitter to be candidate neighbour of the reference CDMA cell. Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept. • The importance of neighbours. Importance values are used by the allocation algorithm to rank the neighbours. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be allocated to each cell is exceeded. If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8 (having the highest importance values) will be allocated to the reference cell. Note that the maximum number of inter-technology neighbours can be defined at the cell level (property dialogue or cell table). If defined there, this value is taken into account instead of the default one available in the Neighbour Allocation dialogue. As indicated in the table below, the neighbour importance depends on the distance and on the neighbourhood cause; this value varies between 0 to 100%. Neighbourhood cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter If the Force co-site cells as neighbours option is selected 100 % Neighbourhood relationship that fulfils distance conditions If the maximum distance is not exceeded d1 – ---------d max Where d is the effective distance between the CDMA reference cell and the GSM neighbour and d max is the maximum intersite distance. In the Results part, Atoll provides the list of neighbours, the number of neighbours and the maximum number of neighbours allowed for each cell. In addition, it indicates the importance (in %) of each neighbour and the allocation reason. Therefore, a neighbour may be marked as exceptional pair, co-site, or distance. For neighbours accepted for distance reasons, Atoll displays the distance from the reference cell (m). Finally, if cells have previous allocations in the list, neighbours are marked as existing. 5.8.2.2 Algorithm Based on Coverage Overlapping When automatic allocation starts, Atoll checks following conditions: • The distance between the CDMA reference cell and the GSM neighbour must be less than the user-definable maximum inter-site distance. If the distance between the CDMA reference cell and the GSM neighbour is greater than this value, then the candidate neighbour is discarded. Atoll calculates the effective distance, which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter Distance" on page 430. • The calculation options, Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or more carriers. Atoll will allocate neighbours to cells using the selected carriers. 439 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site as the reference CDMA cell in the candidate neighbour list. This option is automatically selected. Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you may force/forbid a GSM transmitter to be candidate neighbour of the reference CDMA cell. Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept. • There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability where: Four different cases may be considered for SA: • 1st case: SA is the area where the cell A is the best serving cell of the CDMA network. • The pilot signal received from A is greater than the minimum pilot signal level, • The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0) and is the highest one. In this case, the Ec/I0 margin must be equal to 0dB and the max Ec/I0 option disabled. • 2nd case: SA represents the area where the pilot quality from the cell A strats decreasing but the cell A is still the best serving cell of the CDMA network. The Ec/I0 margin must be equal to 0dB, the max Ec/I0 option selected and a maximum Ec/I0 user-defined. • • • • The pilot signal received from A is greater than the minimum pilot signal level, The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/I0. The pilot quality from A is the highest one. 3rd case: SA represents the area where the cell A is not the best serving cell but can enter the active set. Here, the Ec/I0 margin has to be different from 0dB and the max Ec/I0 option disabled. • • • The pilot signal received from A is greater than the minimum pilot signal level, The pilot quality from A is within a margin from the best Ec/I0, where the best Ec/I0 exceeds the minimum Ec/ I0. 4th case: SA represents the area where: • The pilot signal received from A is greater than the minimum pilot signal level, • The pilot quality from A is within a margin from the best Ec/I0 (where the best Ec/I0 exceeds the minimum Ec/ I0) and lower than the maximum Ec/I0. In this case, the margin must be different from 0dB, the max Ec/I0 option selected and a maximum Ec/I0 userdefined. Two different cases may be considered for SB: • 1st case: SB is the area where the cell B is the best serving cell of the GSM network. In this case, the margin must be set to 0dB. • • The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is the highest one. 2nd case: The margin is different from 0dB and SB is the area where: • The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is within a margin from the best BCCH signal level. SA  SB Atoll calculates the percentage of covered area ( ------------------  100 ) and compares this value to the % minimum covered area. If SA this percentage is not exceeded, the candidate neighbour B is discarded. Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of covered area. When the automatic allocation is based on coverage overlapping, we recommend you to perform two successive automatic allocations: • • • 440 A first allocation in order to find handovers due to non-continuous CDMA coverage. In this case, you have to select the max Ec/I0 option and define a high enough value. A second allocation in order to complete the previous list with handovers motivated for reasons of traffic and service distribution. Here, the max Ec/I0 option must be disabled. The importance of neighbours. Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks AT330_TRR_E1 Importance values are used by the allocation algorithm to rank the neighbours according to the distance and the allocation reason. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be allocated to each cell is exceeded. If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8 (having the highest importance values) will be allocated to the reference cell. Note that the maximum number of inter-technology neighbours can be defined at the cell level (property dialogue or cell table). If defined there, this value is taken into account instead of the default one available in the Neighbour Allocation dialogue. As indicated in the table below, the neighbour importance depends on the neighbourhood cause; this value varies between 0 to 100%. Neighbourhood reason When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter If the Force co-site cells as neighbours option is selected IF Neighbourhood relationship that fulfils coverage conditions If the % minimum covered area is exceeded IF Except the case of forced neighbours (importance = 100%), priority assigned to each neighbourhood cause is determined using the Importance Function (IF). The IF considers the following factors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. d Di  = 1 – ---------d max d is the effective distance (in m). It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter Distance" on page 430. d max is the maximum distance between the reference transmitter and a possible neighbour. • • The co-site factor (C): a Boolean, The overlapping factor (O): the percentage of overlapping. The IF is user-definable using the Min importance and Max importance fields. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The IF evaluates importance as follows: Co-site Neighbourhood cause IF Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)} Yes Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+ Max(O)(O)/(Max(Di)+Max(O))} 60%+40%{1/7%(Di)+6/7%(O)} Where Delta(X)=Max(X)-Min(X) • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. If the Min and Max value ranges of the importance function factors do not overlap, the neighbours will be ranked by neighbour cause. With the default values for minimum and maximum importance fields, neighbours will be ranked in this order: co-site neighbours and neighbours allocated based on coverage overlapping. If the Min and Max value ranges of the importance function factors overlap, the neighbours may be ranked differently. There can be a mix of the neighbourhood causes. 441 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 5: CDMA2000 Networks ©Forsk 2015 In the Results part, Atoll provides the list of neighbours, the number of neighbours and the maximum number of neighbours allowed for each cell. In addition, it indicates the importance (in %) of each neighbour and the allocation reason. Therefore, a neighbour may be marked as exceptional pair, co-site or coverage. For neighbours accepted for co-site and coverage reasons, Atoll displays the percentage of area meeting the coverage conditions and the corresponding surface area (km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing. • • • No prediction study is needed to perform an automatic neighbour allocation. When starting an automatic neighbour allocation, Atoll automatically calculates the path loss matrices if not found. A forbidden neighbour must not be listed as neighbour except if the neighbourhood relationship already exists and the Delete existing neighbours option is unchecked when you start the new allocation. In this case, Atoll displays a warning in the Event viewer indicating that the constraint on the forbidden neighbour will be ignored by algorithm because the neighbour already exists. In the Results, Atoll displays only the cells for which it finds new neighbours. Therefore, if a TBA cell has already reached its maximum number of neighbours before starting the new allocation, it will not appear in the Results table. 5.8.2.3 Delete Existing Neighbours Option As explained above, Atoll keeps the existing inter-technology neighbours when the Delete existing neighbours option is not checked. We assume that we have an existing allocation of inter-technology neighbours. A new TBA cell i is created in CDMA.atl. Therefore, if you start a new allocation without selecting the Delete existing neighbours option, Atoll determines the neighbour list of the cell i, If you change some allocation criteria (e.g. increase the maximum number of neighbours or create a new GSM TBC transmitter) and start a new allocation without selecting the Delete existing neighbours option, it examines the neighbour list of TBA cells and checks allocation criteria if there is space in their neighbour lists. A new GSM TBC transmitter can enter the TBA cell neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list. 442 Chapter 6 LTE Networks This chapter covers the following topics: • "Definitions" on page 445 • "Calculation Quick Reference" on page 450 • "Available Calculations" on page 470 • "Calculation Details" on page 485 • "Automatic Planning Algorithms" on page 563 Atoll 3.3.0 Technical Reference Guidefor Radio Networks © Forsk 2015 444 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 6 LTE Networks This chapter describes all the calculations performed in Atoll LTE documents. The first part of this chapter lists all the input parameters in the LTE documents, their significance, location in the Atoll GUI, and their usage. It also contains the lists of the formulas used for the calculations. The second part describes all the calculation processes, i.e., signal level coverage predictions, point analysis calculations, signal quality coverage predictions, calculations on subscriber lists, and Monte Carlo simulations. The calculation algorithms used by these calculation processes are available in the next part. The third part describes all the calculation algorithms used in all the calculations. These algorithms include the calculation of signal levels, noise, and interference for downlink and uplink considering power control, MIMO, smart antennas, and the radio resource management algorithms used by the different available schedulers. If you are new to LTE, you can also see the Glossary of LTE Terms in the User Manual for information on LTE terms and concepts, especially in the context of their user in Atoll. • • • All the calculations are performed on TBC (to be calculated) transmitters. For the definition of TBC transmitters please refer to "Path Loss Matrices" on page 103. A cell refers to a transmitter-carrier (TX-c) pair. The cell being studied during a calculation is referred to as TXi(ic) in this chapter. All the calculation algorithms in this section are described for two types of cells: • • A studied cell (represented by the subscript "i") comprising the studied transmitter TXi and its carrier ic. It is the cell which is currently the focus of the calculation. For example, a victim cell when calculating the interference it is receiving from other cells. • Other cells (represented by the subscript "j") comprising the other transmitter TXj and its carrier jc. The other cells in the network can be interfering cells (downlink) or the serving cells of interfering mobiles (uplink). All the calculation algorithms in this section are described for two types of receivers: • Mi: A pixel (coverage predictions), subscriber (calculations on subscriber lists), or mobile (Monte Carlo simulations) covered/served by the studied cell TXi(ic). • Mj: A mobile (Monte Carlo simulations) covered/served by any other cell TXj(jc). Logarithms used in this chapter (Log function) are base-10 unless stated otherwise. • 6.1 Definitions This table lists the input to calculations, coverage predictions, and simulations. Name Value Unit Description D Frame 3GPP parameter (Fixed to 10 ms in Atoll) ms Frame duration W FB 3GPP parameter (Fixed to 180 kHz in Atoll) kHz Width of a resource/frequency block F 3GPP parameter (Fixed to 15 kHz in Atoll) kHz Subcarrier width N FB – SS PBCH 3GPP parameter (Fixed to 6 in Atoll) None Number of frequency blocks for SS and PBCH transmission N SF  Frame 3GPP parameter (Fixed to 10 in Atoll) None Number of subframes per frame N Slots  SF 3GPP parameter (Fixed to 2 in Atoll) None Number of slots per subframe K 1.38 x 10-23 J/K Boltzmann’s constant T 290 K Ambient temperature n0 Calculation result ( 10  Log  K  T  1000  = – 174 dBm/Hz ) dBm/Hz Power spectral density of thermal noise D CP Frame configuration or, otherwise, global parameter None Cyclic prefix duration N SD – PDCCH Frame configuration or, otherwise, global parameter SD Number of PDCCH symbol durations per subframe N FB – PUCCH Frame configuration or, otherwise, global parameter RB Average number of PUCCH frequency blocks per frame 445 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Name Value Unit Description M PC Global parameter dB Uplink power control adjustment margin CNR Min Global parametera dB Minimum signal to thermal noise threshold (interferer cutoff) W Channel Frequency band parameter MHz Channel bandwidth First Frequency band parameter None First channel number of the frequency band N Channel Last Frequency band parameter None Last channel number of the frequency band F Start – TDD Frequency band parameter MHz Start frequency of the TDD frequency band F Start – FDD – DL Frequency band parameter MHz DL start frequency of the FDD frequency band F Start – FDD – UL Frequency band parameter MHz UL start frequency of the FDD frequency band F Sampling Frequency band parameter MHz Sampling frequency f ACS Frequency band parameter dB Adjacent channel suppression factor N FB Frequency band parameter None Number of frequency blocks per channel bandwidth ICS Band Frequency band parameter MHz Inter-channel spacing CN Band Frequency band parameter None Channel number step N SCa – Total F Sampling -) Calculation result ( N SCa – Total = -------------------F None Total number of subcarriers N SCa – Used N FB  W FB -) Calculation result ( N SCa – Used = -----------------------F None Number of used subcarriers N SCa – DC Hard-coded parameter ( N SCa – DC = 1 ) None Number of DC subcarriers N SCa – Guard Calculation result ( N SCa – Guard = N SCa – Total – N SCa – Used – N SCa – DC ) None Number of guard subcarriers i Layer Layer parameter None Layer index p Layer Layer parameter None Layer priority V Layer Max Layer parameter km/h Maximum mobile speed supported by a layer  CE Frame configuration parameter dB Cell-edge power boost N FB – CE0 Frame configuration parameter None Number of cell-edge frequency blocks for PSS ID 0 N FB – CE1 Frame configuration parameter None Number of cell-edge frequency blocks for PSS ID 1 N FB – CE2 Frame configuration parameter None Number of cell-edge frequency blocks for PSS ID 2 B Bearer parameter None Bearer index Mod B Bearer parameter None Modulation used by the bearer CR B Bearer parameter None Coding rate of the bearer B Bearer parameter bits/ symbol Bearer efficiency TB Bearer parameter dB Bearer selection threshold Site Site parameter kbps Maximum S1 interface site downlink throughput N Channel TP S1 – DL 446 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value Unit Description Site Site parameter kbps Maximum S1 interface site uplink throughput Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) dB Transmitter noise figure N Ant – TX Transmitter parameter None Number of antenna ports used for transmission N Ant – RX Transmitter parameter None Number of antenna ports used for reception TX Transmitter antenna parameter dB Antenna gain TX Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) dB Transmitter loss TX Smart antenna parameter None Number of smart antenna elements Array Smart antenna parameter dB Array gain offset Combining Smart antenna parameter dB Power combining gain offset G SA Smart antenna parameter dB Diversity gain (cross-polarisation) N Channel Cell parameter None Cell’s channel number ID  Cell parameter None Cell’s physical ID ID SSS ID  Cell parameter: Floor  --------  3  None Cell’s SSS ID (one of 168 pseudorandom sequences) ID PSS Cell parameter: ID  Mod 3 None Cell’s PSS ID (one of 3 cyclic shifts of the sequence given by the SSS ID)  Shift Cell parameter: ID  Mod 6 None Cell’s v shift (also known as the reference signal hopping index) P Max Cell parameter dBm Maximum cell transmission power EPRE DLRS Cell parameter dBm Energy per resource element for the downlink reference signals (User-defined or calculated) EPRE SS Cell parameter dB Energy per resource element offset for the SS with respect to the downlink reference signal EPRE EPRE PBCH Cell parameter dB Energy per resource element offset for the PBCH with respect to the downlink reference signal EPRE EPRE PDCCH Cell parameter dB Energy per resource element offset for the PDCCH with respect to the downlink reference signal EPRE EPRE PDSCH Cell parameter dB Energy per resource element offset for the PDSCH with respect to the downlink reference signal EPRE T RSRP Cell parameter dB Minimum Required RSRP TX i  ic  Cell parameter dB Cell selection threshold TX  ic  i Cell parameter dB Cell individual offset Cell parameter dB Handover margin TL DL Cell parameter % Downlink traffic load r DL – CE Cell parameter % Downlink cell-edge traffic ratio TP S1 – UL nf G L TX E SA G SA G SA Div T Selection O Individual TX i  ic  M HO 447 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Name Value Unit Description TL UL Cell parameter % Uplink traffic load TL DL – Max Cell parameter % Maximum downlink traffic load TL UL – Max Cell parameter % Maximum uplink traffic load NR UL Cell parameter dB Uplink noise rise NRUL – ICIC Cell parameter dB ICIC uplink noise rise NR UL – Max Cell parameter dB Maximum uplink noise rise N Users – Max Cell parameter None Maximum number of users per cell N Users – DL Cell parameter None Number of users connected to the cell in downlink N Users – UL Cell parameter None Number of users connected to the cell in uplink L Path Cell parameter dB Delta path loss threshold N SF – DL Cell parameter None Number of downlink subframes per frame N SF – UL Cell parameter None Number of uplink subframes per frame N TDD – SSF TX i  ic  Cell parameter None Number of TDD special subframes per frame D Reuse Cell parameter m Channel and physical cell ID reuse distance G MU – MIMO – DL Cell parameter None Average number of co-scheduled MU-MIMO users in downlink G MU – MIMO – UL Cell parameter None Average number of co-scheduled MU-MIMO users in uplink  FPC Cell parameter None Fractional power control factor CINR PUSCH – Max Cell parameter dB Maximum PUSCH C/(I+N) Inter – Tech Cell parameter dB Inter-technology downlink noise rise Inter – Tech Cell parameter dB Inter-technology uplink noise rise AU DL Cell parameter % Downlink AAS usage ratio TX i  ic  Proportional Fair scheduler parameter None Downlink multi-user diversity gain (MUG) TX i  ic  Proportional Fair scheduler parameter None Uplink multi-user diversity gain (MUG) CINR MUG Proportional Fair scheduler parameter dB Maximum C/(I+N) above which no MUG gain is applied T SU – MIMO – UL Cell reception equipment parameter dB Uplink SU-MIMO threshold T MU – MIMO – UL Cell reception equipment parameter dB Uplink MU-MIMO threshold G SU – MIMO – UL Cell reception equipment parameter None Maximum uplink SU-MIMO gain G Div – UL Cell reception equipment parameter dB Receive, SU-MIMO, or MU-MIMO diversity gain T SCell UL Cell reception equipment parameter dB Uplink secondary cell activation threshold QCI Service parameter None QoS class identifier (QCI) of the service p QCI Service parameter (automatically determined from the QCI) None Service’s QCI priority NR DL NR UL G MUG – DL G MUG – UL Max Max 448 GBR) TPD Max – UL Service parameter kbps Maximum throughput demand in the uplink (Maximum Bit Rate.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value Unit Description p Service Service parameter None User-defined service priority B DL – Highest Service parameter None Highest bearer used by a service in the downlink B UL – Highest Service parameter None Highest bearer used by a service in the uplink B DL – Lowest Service parameter None Lowest bearer used by a service in the downlink B UL – Lowest Service parameter None Lowest bearer used by a service in the uplink f Act UL Service parameter % Uplink activity factor f Act DL Service parameter % Downlink activity factor TPD Min – UL Service parameter kbps Minimum throughput demand in the uplink (Guaranteed Bit Rate.Atoll 3.3. MBR) TPD Max – DL Service parameter kbps Maximum throughput demand in the downlink (Maximum Bit Rate. MBR) UL Service parameter kbps Average requested throughput in the uplink TP Average DL Service parameter kbps Average requested throughput in the downlink TP Offset Service parameter kbps Throughput offset f TP – Scaling Service parameter % Scaling factor L Body Service parameter dB Body loss N FB – UL Min Service parameter None Minimum number of frequency blocks P Min Terminal parameter dBm Minimum terminal power P Max Terminal parameter dBm Maximum terminal power nf Terminal parameter dB Terminal noise figure G Terminal parameter dB Terminal antenna gain L Terminal parameter dB Terminal loss N Ant – TX Terminal parameter None Number of antenna ports for transmission N Ant – RX Terminal parameter None Number of antenna ports for reception Max – DL Terminal parameter None Maximum number of downlink secondary cells Max – UL Terminal parameter None Maximum number of uplink secondary cells N TBB  TTI Max – DL UE category parameter Bits Maximum number of transport block bits per TTI (subframe) in downlink Max – UL UE category parameter Bits Maximum number of transport block bits per TTI (subframe) in uplink UE category parameter None Highest modulation supported in uplink UE category parameter None Maximum number of reception antenna ports supported in downlink TP Average N SCell N SCell N TBB  TTI Max – UL Mod UE Max – DL N Ant – UE 449 . GBR) TPD Min – DL Service parameter kbps Minimum throughput demand in the downlink (Guaranteed Bit Rate. 2. SU-MIMO. or MU-MIMO diversity gain T SCell DL Terminal reception equipment parameter dB Downlink secondary cell activation threshold DL Terminal reception equipment parameter dB Downlink AAS threshold UL Clutter parameter dB Additional uplink diversity gain G Div DL Clutter parameter dB Additional downlink diversity gain f SU – MIMO Clutter parameter None SU-MIMO gain factor L Indoor Clutter parameter dB Indoor loss L Path Propagation model result dB Path loss Max T AAS G Div F ICPDL Network parameter None Inter-technology downlink channel protection ratio for a frequency offset F between the interfered and interfering frequency channels M Shadowing – Model Monte Carlo simulations: Random result calculated from model standard deviation Coverage Predictions: Result calculated from cell edge coverage probability and model standard deviation dB Model shadowing margin M Shadowing – C  I Coverage Predictions: Result calculated from cell edge coverage probability and C/I standard deviation dB C/I shadowing margin Any interfering cell whose signal to thermal noise ratio is less than CNR Min will be discarded.2 Calculation Quick Reference The following tables list the formulas used in calculations. a.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Name Value Unit Description T SU – MIMO – DL Terminal reception equipment parameter dB SU-MIMO threshold G SU – MIMO – DL Terminal reception equipment parameter None Maximum downlink SU-MIMO gain T MU – MIMO – DL Terminal reception equipment parameter dB Downlink MU-MIMO threshold G Div – PBCH Terminal reception equipment parameter dB PBCH diversity gain G Div – PDCCH Terminal reception equipment parameter dB PDCCH diversity gain G Div – DL Terminal reception equipment parameter dB Transmit.Atoll 3. 6. 6.3.1 Downlink Transmission Powers Calculation Name Value Unit Description N Sym  SRB N SCa – FB  N SD  Slot  N Slot  SF None Number of symbols per scheduler resource block N Sym  SSF DwPTS N SCa – FB  N SD  SSF None Number of DwPTS modulation symbols per scheduler resource block in the TDD special subframes N SCa – FB W FB ---------F None Number of subcarriers per frequency block None Total number of symbols in downlink TX  ic  i N Sym – DL 450 DwPTS TX  ic  i N FB TX  ic  i TX  ic  i  N Sym  SRB  N SF – DL + N FB TX  ic  i DwPTS  N TDD – SSF  N Sym  SSF . Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value Unit Description None Number of symbols reserved for downlink reference signals in one scheduler resource block See "Downlink Transmission Power Calculation" on page 485 None Number of symbols reserved for downlink reference signals in DwPTS of one TDD special subframe TX i  ic  None Number of symbols reserved for downlink reference signals in one frame None Number of symbols for downlink reference signals in one scheduler resource block None Number of symbols for downlink reference signals in DwPTS of one TDD special subframe TX i  ic  None Number of symbols for downlink reference signals in one frame Where N Sym – PSS = 2  N FB – SS PBCH  N SCa – FB = 144 None Number of symbols for the PSS and the SSS None Number of symbols for the PBCH None Number of symbols for the PDCCH    8    16     24  TX  ic  i N Res  SRB TX i  ic  N Res  DwPTS TX i  ic  N Sym – Res TX i  ic  N SF – DL  N FB N DLRS  SRB TX i  ic  TX i  ic  N Sym – DLRS TX  ic  i if  N Ant – TX = 2 TX  ic  i if  N Ant – TX = 4 or 8   TX i  ic  TX i  ic  TX i  ic   N Res  SRB + N TDD – SSF  N FB    8    8     6  TX i  ic  N DLRS  DwPTS TX  ic  i if  N Ant – TX = 1   TX i  ic   N Res  DwPTS TX  ic  i if  N Ant – TX = 1   TX  ic  i if  N Ant – TX = 2 TX  ic  i if  N Ant – TX = 4 or 8   See "Downlink Transmission Power Calculation" on page 485 TX i  ic  TX i  ic  N SF – DL  N FB TX i  ic  TX i  ic  TX i  ic   N DLRS  SRB + N TDD – SSF  N FB  N DLRS  DwPTS N Sym – PSS + N Sym – SSS = 288 N Sym – SS N Sym – SSS = 2  N FB – SS PBCH  N SCa – FB = 144 TX i  ic  Extended CP: 216 Normal CP: 240 N Sym – PBCH TX  ic  i if  N SD – PDCCH = 0 : 0   TX  ic  TX  ic  i i if  N SD – PDCCH = 1 AND  N Ant – TX  2 : TX  ic  i N Sym – PDCCH TX  ic  TX  ic  TX i  ic  TX  ic  TX  ic  i N i   SD – PDCCH  N SCa – FB – 4  N FB  N SF – DL i i +  NSD – PDCCH  N SCa – FB – 4  N FB   Otherwise: TX  ic  TX i  ic   N TDD – SSF TX  ic  TX  ic  i i N i  N SCa – FB – 2  Min  4 N Ant – TX   N FB  SD – PDCCH   TX  ic  TX  ic  TX  ic  i i i +  Min  2 N SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB      TX i  ic  N Sym – PDSCH TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic   N SF – DL N Sym – DL – N Sym – Res – N Sym – SS – N Sym – PBCH – N Sym – PDCCH TX i  ic   N TDD – SSF None Number of symbols for the PDSCH 451 . Atoll 3. 10  Energy per resource element for 1 modulation symbol (dBm/Sym) of the downlink reference signals With reference signal EPRE calculation method is set to Calculated (with boost) or Calculated (without boost)    EPRE SS EPRE DLRS + EPRE SS TX i  ic  TX i  ic  dBm/Sym Energy per resource element for 1 modulation symbol (dBm/Sym) of the SS TX  ic  i EPRE PBCH TX  ic  i EPRE DLRS TX  ic  i EPRE PBCH dBm/Sym Energy per resource element for 1 modulation symbol (dBm/Sym) of the PBCH TX i  ic  TX i  ic  TX i  ic  dBm/Sym Energy per resource element for 1 modulation symbol (dBm/Sym) of the PDCCH TX i  ic  TX i  ic  dBm/Sym Energy per resource element for 1 modulation symbol (dBm/Sym) of the PDSCH dbm/Sym "Boosted" energy per resource element for 1 modulation symbol (dBm/Sym) of downlink reference signals when the reference signal EPRE calculation method is set to Calculated (with boost) dBm Instantaneous transmission power of the downlink reference signals + 10  Log  N SCa – FB  N FB – SS PBCH  dBm Instantaneous transmission power of the SS EPRE PBCH + 10  Log  N SCa – FB  N FB – SS PBCH  dBm Instantaneous transmission power of the PBCH TX i  ic  EPRE DLRS + EPRE PDCCH EPRE PDCCH TX i  ic  EPRE DLRS + EPRE PDSCH EPRE PDSCH TX  ic  i EPRE DLRS TX i  ic  P DLRS TX i  ic  P SS TX  ic  i P PBCH 452 + TX i  ic  N  TX i  ic  Sym – Res  EPRE DLRS + 10  Log  ------------------------ TXi  ic    N Sym – DLRS TX  ic  TX  ic  i i EPRE DLRS + 10  Log  2  N FB TX i  ic  EPRE SS TX  ic  i   .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks Name ©Forsk 2015 Value Unit Description TX  ic   P i  Max  ------------------- TX  ic  TX  ic  i i DwPTS 10 10  Log  10   N SD  Slot  N Slot  SF  N SF – DL + N TDD – SSF  N SD  SSF  –     TX  ic  TX  ic  i EPRE DLRS TX  ic  i i  EPRE SS EPRE PBCH  TX  ic  --------------------------------------------------------------------i 10 10 dBm/Sym 10  L og  N Sym – DLRS + N Sym – SS  10 + N Sym – PBCH  10    + N Sym – PDCCH  10 TX i  ic  EPRE PDCCH -----------------------------------10 + N Sym – PDSCH  10 TX i  ic   EPRE PDSCH -----------------------------------. 10  Energy per resource element for 1 modulation symbol (dBm/Sym) of the downlink reference signals With reference signal EPRE calculation method is set to Calculated (equal distribution of unused EPRE)    TX  ic   P i  Max  ------------------- TX i  ic  TX i  ic  DwPTS 10 10  Log  10   N SD  Slot  N Slot  SF  N SF – DL + N TDD – SSF  N SD  SSF  –        TX  ic  TX  ic  i EPRE DLRS TX  ic  i i  EPRE EPRE SS PBCH  TX  ic  ------------------------------------------------------------------i 10 10 dBm/Sym + N Sym – PBCH  10 10  L og  N Sym – Res + N Sym – SS  10    + N Sym – PDCCH  10 TX i  ic  EPRE PDCCH -----------------------------------10 + N Sym – PDSCH  10 TX i  ic   EPRE PDSCH -----------------------------------.3. 2 Co.3. or mobile Mi as follows: 453 .Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name TX  ic  i P PDCCH TX i  ic  P PDSCH Value Unit TX  ic    i TX  ic    N Sym–PDCCH i -dBm EPRE PDCCH + 10  Log  ------------------------------------------------------------------------------------------------------------------------------------- TX  ic  TX  ic  TX  ic    TXi  ic  i i i  N SD – PDCCH  N SF – DL + Min  2 N SD – PDCCH  N TDD – SSF         TX i  ic    TX  ic  N i Sym–PDSCH - EPRE PDSCH + 10  Log  ------------------------------------------------------------------------------------------------------------------ TX i  ic  TX  ic    N i  N   N – N SD  Slot Slot  SF SD – PDCCH SF – DL       TX i  ic  TX i  ic     DwPTS    + – Min 2  N  N N   SD  SSF SD – PDCCH  TDD – SSF  dBm Description Average transmission power of the PDCCH Average transmission power of the PDSCH 6.3 Signal Level Calculation (DL) The received signal levels (dBm) from any cell TXi(ic) are calculated for a pixel. subscriber.2.and Adjacent Channel Overlaps Calculation Name TX i  ic  F Start Value Unit Description  N TXi  ic  – N First – TXi  ic  TX i  ic  TX i  ic  Channel Channel   - F Start – Band + W Channel + ICS Band   ------------------------------------------------------TX i  ic        CN Band MHz Start frequency for the channel number assigned to a cell MHz End frequency for the channel number assigned to a cell MHz Co-channel overlap bandwidth None Co-channel overlap ratio MHz Bandwidth of the lower-frequency adjacent channel overlap None Lower-frequency adjacent channel overlap ratio MHz Bandwidth of the higher-frequency adjacent channel overlap None Higher-frequency adjacent channel overlap ratio None Adjacent channel overlap ratio None Total overlap ratio TX i  ic  TX i  ic  TX i  ic  TX  ic  – TX  jc  i j W CCO TX  jc  TX  ic  TX i  ic  – TX j  jc  L TX  ic  – TX  jc  i j H TX i  ic  – TX j  jc  TX  jc  TX  ic  TX  jc  TX i  ic  – TX j  jc  TX j  jc  TX i  ic  Min  F End  F End  TX  ic  TX  jc  TX  ic  i j i + W Channel – Max  F Start  F End     TX i  ic  – TX j  jc  W ACO H ---------------------------------TX j  jc  W Channel TX  ic  – TX  jc  i j TX i  ic  – TX j  jc  r ACO r ACO rO TX  ic  W ACO L ---------------------------------TX j  jc  W Channel H TX i  ic  – TX j  jc  TX  ic  j i j i i Min  F End  F Start  – Max  F Start  F Start – W Channel     TX i  ic  – TX j  jc  r ACO TX  ic  W CCO ----------------------------------TX j  jc  W Channel TX  ic  – TX  jc  i j r ACO L W ACO TX  jc  j i j i Min  F End  F End  – Max  F Start  F Start      TX  ic  – TX  jc  i j r CCO W ACO TX i  ic  F Start + W Channel F End L TX i  ic  – TX j  jc  r CCO TX  ic  – TX  jc  i j + r ACO H TX i  ic  – TX j  jc  + r ACO  10 TX i  ic  – f ACS ----------------------10 6.2. 0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Name TX  ic  i C Max Value TX  ic  i EIRP Max – L Path – M Shadowing – Model – L Indoor + G –L M i M M i TX i  ic  TX i TX i TX i TX i + 10  Log  E SA  +   Combining G SA TX i  ic  –L Mi TX i  ic  C SS TX i + TX i  ic  EIRP1 SS Mi TX i G Ant –L TX i TX i + 10  Log  E SA  + Mi TX i  ic  C PBCH TX i + G Ant – L TX i + G Ant – L TX i TX i + 10  Log  E SA  +   Combining G SA TX i  ic  Mi TX i  ic  TX i TX i TX i TX i + 10  Log  E SA  + Combining G SA TX  ic  i –L Mi TX  ic  i TX i TX i TX i TX i + 10  Log  E SA  +   TX i  ic  Mi Mi 454 PDCCH EIRP dBm Received PDSCH signal level dBm PDSCH EIRP dBm/Sym Received downlink reference signal energy per resource element (RSRP) Mi Mi TX i Array P PDSCH + G SA    + G SA E DLRS dBm Div G SA TX i TX i With smart antennas: TX i  ic  TX i  ic  Received PDCCH signal level – L Ant – L Body + f CP TX i  ic  TX i  ic  + EIRP1 PDSCH – L Path – M Shadowing – Model – L Indoor + G –L dBm i TX i Combining G SA Without smart antennas: P PDSCH + G Ant – L EIRP1 PDSCH M With smart antennas: TX i  ic  P PDCCH + G Ant – L C PDSCH PBCH EIRP – L Ant – L Body + f CP TX i  ic  TX i  ic  dBm Div G SA Mi Without smart antennas: P PDCCH + G Ant – L EIRP1 PDCCH + EIRP1 PDCCH – L Path – M Shadowing – Model – L Indoor + G Mi Received PBCH signal level TX i With smart antennas: TX i  ic  P PBCH + G Ant – L C PDCCH dBm Mi Mi TX i  ic  TX i  ic  SS EIRP – L Ant – L Body + f CP Without smart antennas: P PBCH + G Ant – L EIRP1 PBCH dBm Div G SA + EIRP1 PBCH – L Path – M Shadowing – Model – L Indoor + G Mi Received SS signal level TX i With smart antennas: –L dBm Mi Mi TX i  ic  P SS RS EIRP – L Ant – L Body + f CP TX i  ic  TX i  ic  dBm Combining G SA – L Path – M Shadowing – Model – L Indoor + G Without smart antennas: P SS EIRP1 SS Received downlink reference signal level TX i With smart antennas: TX  ic  i P DLRS –L dBm – L Ant – L Body + f CP TX i  ic  TX i  ic  Downlink max EIRP Mi Mi Without smart antennas: P DLRS + G Ant – L EIRP1 DLRS dbm Div G SA + EIRP1 DLRS – L Path – M Shadowing – Model – L Indoor + G Mi Received max cell power TX i With smart antennas: TX i  ic  P Max + G Ant – L C DLRS dBm i i TX i  ic  TX i  ic  Description – L Ant – L Body + f CP Without smart antennas: P Max + G Ant – L EIRP Max M Unit Combining + G SA Div + G SA – L TX i TX i  ic  EIRP2 DLRS – L Path – M Shadowing – Model – L Indoor + G Mi –L Mi Mi Mi – L Ant – L Body + f CP .3.Atoll 3. 5  If D CP 0 TX i + L Ant + L Body 10  Log  7  7.4 Noise Calculation (DL) Name Value Unit Description TX i  ic  n 0 + 10  Log  F  dBm Thermal noise for one resource element n 0 – Sym 455 .5  If D CP f CP dBm/Sym TX i TX i TX  ic  i L Total Received PDCCH energy per resource element – L Ant – L Body + f CP With smart antennas: EPRE PDCCH + G Ant – L EIRP2 PDSCH dBm/Sym Div TX i  ic  TX i  ic  TX i  ic  PBCH EIRP + G SA EIRP2 PDCCH – L Path – M Shadowing – Model TX i  ic  E PDSCH dBm/Sym TX i Without smart antennas: EPRE PDCCH + G Ant – L TX i  ic  Received PBCH energy per resource element – L Ant – L Body + f CP With smart antennas: EPRE PBCH + G Ant – L EIRP2 PDCCH dBm/Sym TX  ic  i EIRP2 PBCH – L Path – M Shadowing – Model TX  ic  i E PDCCH SS EIRP Combining Div + 10  Log  E SA  + G SA + G SA TX i  ic  TX i  ic  dBm/Sym TX i Without smart antennas: EPRE PBCH + G Ant – L EIRP2 PBCH Received SS energy per resource element TX i With smart antennas: TX i dBm/Sym – L Ant – L Body + f CP TX i  ic  TX i  ic  RS EIRP Combining Mi Without smart antennas: EPRE SS TX i  ic  dBm/Sym + 10  Log  E SA  + G SA   – L Path – M Shadowing – Model Mi Description i With smart antennas: TX i  ic  EIRP2 SS TX Unit If = Normal = Extended TX i  ic  is an interferer 6.Atoll 3.2.e.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value TX  ic  i TX i Without smart antennas: EPRE DLRS + G Ant – L TX  ic  i EIRP2 DLRS TX  ic  i EPRE DLRS TX i + G Ant – L TX i  ic  EIRP2 SS E SS TX i TX – L Indoor + G i –L Mi Mi EPRE SS TX i  ic  E PBCH TX i + G Ant – L + G Ant – L TX i – L Indoor + G Mi –L Mi Mi Mi TX i TX i  ic  TX i TX i TX i + 10  Log  E SA  + G SA   Combining – L Indoor + G Mi –L Mi Mi Mi TX i  ic  TX i TX i TX i PDCCH EIRP dBm/Sym Received PDSCH energy per resource element dBm/Sym PDSCH EIRP dB Path loss dB Total losses dB Cyclic prefix factor. the ratio of the useful symbol energy to the total symbol energy Combining Div + 10  Log  E SA  + G SA + G SA TX  ic  i EIRP2 PDSCH – L Path – M Shadowing – Model – L Indoor + G Mi –L Mi Mi Mi – L Ant – L Body + f CP TX i Without smart antennas: EPRE PDSCH + G Ant – L TX i With smart antennas: TX i  ic  TX i Array EPRE PDSCH + G SA    + G SA Combining + G SA Div + G SA – L TX i L Path L Model + L Ant L Path + L +L Mi TX i –G + L Indoor + M Shadowing – Model – G Mi Mi TX i Mi TX i  ic  TX i  ic  10  Log  6  7. i.3.. 3. and PDSCH of the interfering cell With 4 or 8 antenna ports and TX  ic  i N SD – PDCCH = 1 Interfering energy per resource element (dBm/Sym) received over downlink reference signals (Method 1: synchronised transmission and reception) Case: RS of the interfered cell collide with RS.5 Interference Calculation (DL) Name Value Unit Description dBm/Sym Interfering energy per resource element (dBm/Sym) received over downlink reference signals (Method 1: synchronised transmission and reception) Case: RS of the interfered cell collide only with RS of the interfering cell j  TX  jc  E DLRS  N j  N TXi  ic  – N TX j  jc  --------------------- TX i  ic  – TX j  jc  10 Ant – TX Ant – TX Ant – TX   10  Log ------------------ 10 + fO + 10  Log  -------------------------------------------- TX i  ic   TXi  ic    N Ant – TX  N Ant – TX     dBm/Sym TX j  jc  TX i  ic  – TX j  jc  TX j  jc  TX i  ic  – TX j  jc   E PDCCH + f PDCCH E PDSCH + f PDSCH --------------------------------------------------------------------------------------------------------------------------------------------. PDCCH. and PDSCH of the interfering cell With 4 or 8 antenna ports and TX i  ic  N SD – PDCCH  1 Interfering energy per resource element (dBm/Sym) received over downlink reference signals (Method 1: synchronised dBm/Sym transmission and reception) Case: RS of the interfered cell collide with PDCCH and PDSCH of the interfering cell With 1 or 2 antenna ports . PDCCH. and PDSCH of the interfering cell With 1 or 2 antenna ports TX  jc  j  TX  ic   E DLRS N i --------------------- TX  ic  – TX j  jc  10 Ant – TX  +f i 10  Log  ------------------ 10 O  TXj  jc    N Ant – TX    TX  jc  j  DLRS TX  jc  TX j  jc   DLRS TX  jc  TX  jc  j  DLRS j  TX  jc  E DLRS   N TXi  ic  – N TX j  jc  N j --------------------- TX i  ic  – TX j  jc  10  Ant – TX Ant – TX Ant – TX  + 10  Log  -------------------------------------------- 10  Log  ------------------ 10 TX i  ic  TX i  ic   + fO  N Ant – TX  N Ant – TX    dBm/Sym TX j  jc  TX j  jc  TX  ic  – TX  jc  TX i  ic  – TX j  jc  i j  E E +f +f PDCCH PDCCH PDSCH PDSCH --------------------------------------------------------------------------------------------------------------------------------------------. 10 10  10 + 5  10 ----------------------------------------------------------------------------------------------------------------------------------- 6    TX  jc  TX j  jc   DLRS j  TX  jc  E DLRS  N j  N TXi  ic  – N TX j  jc  --------------------- TX i  ic  – TX j  jc  10 Ant – TX Ant – TX Ant – TX  +f 10  Log  ------------------ 10  + 10  Log  --------------------------------------------O TX i  ic   TXi  ic     N Ant – TX  N Ant – TX    dBm/Sym TX j  jc  TX i  ic  – TX j  jc  TX j  jc  TX i  ic  – TX j  jc   E PDCCH + f PDCCH E PDSCH + f PDSCH --------------------------------------------------------------------------------------------------------------------------------------------. PDCCH. 3    TX  jc  TX  jc  j  DLRS TX i  ic  – TX j  jc  + fO 456 TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  j i j j i j  EPDCCH  + f PDCCH E PDSCH + f PDSCH -----------------------------------------------------------------------  ----------------------------------------------------------------------10 10  10  + 3  10 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ 4       Interfering energy per resource element (dBm/Sym) received over downlink reference signals (Method 1: synchronised transmission and reception) Case: RS of the interfered cell collide with RS. 10 10  + 3  10 10 ----------------------------------------------------------------------------------------------------------------------------------- 4    Interfering energy per resource element (dBm/Sym) received over downlink reference signals (Method 1: synchronised transmission and reception) Case: RS of the interfered cell collide with RS. 10 10  + 2  10 10 -----------------------------------------------------------------------------------------------------------------------------------.Atoll 3.2.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks Name ©Forsk 2015 Value TX  ic  i TX  ic  i n 0 – Sym + nf n Sym M i Unit Description dBm Downlink noise for one resource element 6. 0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value Unit TX  jc  TX  ic  – TX  jc  j i j E +f PDCCH PDCCH ----------------------------------------------------------------------10 TX j  jc   DLRS TX  jc  TX  ic  – TX  jc  j i j  E +f PDSCH PDSCH ----------------------------------------------------------------------- 10     10 + 5  10 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ 6       TX i  ic  – TX j  jc  Interfering energy per resource element (dBm/Sym) received over downlink reference signals (Method 1: synchronised transmission and reception) Case: RS of the interfered cell dBm/Sym collide with PDCCH and PDSCH of the interfering cell With 4 or 8 antenna ports and + fO TX i  ic  N SD – PDCCH = 1 TX j  jc  TX j  jc   DLRS TX i  ic  – TX j  jc  TX j  jc  TX i  ic  – TX j  jc   EPDCCH + fPDCCH  E PDSCH + f PDSCH -----------------------------------------------------------------------  ----------------------------------------------------------------------10 10  10  + 2  10 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ 3       TX i  ic  – TX j  jc  Interfering energy per resource element (dBm/Sym) received over downlink reference signals (Method 1: synchronised transmission and reception) Case: RS of the interfered cell dBm/Sym collide with PDCCH and PDSCH of the interfering cell With 4 or 8 antenna ports and + fO TX i  ic  N SD – PDCCH  1 TX j  jc  TX j  jc   SS PBCH Description TX j  jc  E PBCH  ESS  --------------------TX j  jc   ------------------- 10 10  10  N Sym – SS + 10  N Sym – PBCH - 10  Log  ------------------------------------------------------------------------------------------------------------TX j  jc    N + N Sym – SS Sym – PBCH     TX i  ic  – TX j  jc  + fO dBm/Sym TX j  jc  Interfering energy per resource element (dBm/Sym) received over the SS and the PBCH (Method 1: synchronised transmission and reception) + f MIMO TX  jc  TX j  jc   PDCCH j  E DLRS  TX  jc  j  -------------------- TX  ic  – TX j  jc  N Sym – DLRS in PDCCH 1 .3. 10 10  + f O i 10  Log  ------------------ ----------------------------------------TX  jc  TX  ic    j i N Sym – PDCCH  N Ant – TX    TX  jc  TX  ic  – TX  jc  j i j  TX  ic   E +f TX  jc  PDCCH PDCCH j N i ----------------------------------------------------------------------- – N 10 Sym – PDCCH Sym – DLRS in PDCCH  . 10 10  10  Log  ------------------ -----------------------------------------------------------------------------------------TX  ic   TXj  jc   i N Sym – PDCCH  N Ant – TX    TX  jc  TX  ic  – TX  jc  j i j  TX  jc   E +f PDCCH PDCCH N j  ----------------------------------------------------------------------TX  ic  – TX j  jc  10 Sym – PDCCH  +f i + 10  L og  ---------------------------- 10 O TX  ic    i  N Sym – PDCCH    TX j  jc  TX i  ic  – TX j  jc  E PDCCH + f PDCCH TX i  ic  – TX j  jc  + fO Interfering energy per resource element (dBm/Sym) received over the PDCCH (Method 1: synchronised dBm/Sym transmission and reception) Case: PDCCH of the interfered cell collides with PDCCH and some RS of the interfering cell Interfering energy per resource element (dBm/Sym) received over the PDCCH (Method 1: synchronised dBm/Sym transmission and reception) Case: PDCCH of the interfered cell collides only with PDCCH of the interfering cell 457 .Atoll 3. 10 + 10  L og  ----------------------------------------------------------------------------TX  ic    i   N Sym – PDCCH   Interfering energy per resource element (dBm/Sym) received over the PDCCH (Method 1: synchronised dBm/Sym transmission and reception) Case: PDCCH of the interfered cell collides with PDCCH and all the RS of the interfering cell TX  jc  TX j  jc   PDCCH TX j  jc   PDCCH j  TX  ic  E DLRS  TX  jc  TX  ic  j i N i --------------------- N – N Ant – TX Sym – DLRS in PDCCH Sym – DLRS in PDCCH .  10 10  10  Log  ------------------ ----------------------------------------------------------------------------------------TX  ic   TXj  jc   i N Sym – PDSCH  N Ant – TX    TX  jc  TX j  jc  TX j  jc   PDSCH TX i  ic  – TX j  jc  E PDSCH + f PDSCH TX  jc  TX  jc  j  DLRS TX  jc  j TX  jc  TX  jc  j 458 Interfering energy per resource element (dBm/Sym) received over the PDSCH (Method 1: synchronised dBm/Sym transmission and reception) Case: PDSCH of the interfered cell collides only with PDSCH of the interfering cell TX  ic  – TX  jc  dBm/Sym Interfering energy per resource element (dBm/Sym) received over downlink reference signals (Method 2: non-synchronised transmission and reception) +f E PBCH + f MIMO  ESS  MIMO----------------------------------------------TX  jc   --------------------------------------------- 10 10 j TX i  ic  – TX j  jc   10  N Sym – SS + 10  N Sym – PBCH --------------------------------------------------------------------------------------------------------------------------------------------------10  Log   + fO TX j  jc    N Sym – SS + N Sym – PBCH dBm/Sym     Interfering energy per resource element (dBm/Sym) received over the SS and the PBCH (Method 2: non-synchronised transmission and reception) TX j  jc  TX i  ic  – TX j  jc   E PDSCH + f PDSCH TX j  jc   ----------------------------------------------------------------------TX  ic  – TX j  jc  N 10 Sym – PDSCH  .+ fO i  ---------------------------+ 10 TX j  jc   N Sym – DL   TX  jc  j TX  jc   PDCCH TX i  ic  – TX j  jc  + fO j i j  E j E PDCCH + f PDCCH TX j  jc  TX j  jc  DLRS  -----------------------------------------------------------------------------------------N N Sym – PDCCH 10 10 Sym – DLRS  .+ 10  ------------------------ ----------------------------10  Log  10 TX j  jc  TX j  jc   N Sym – DL N Sym – DL  TX  jc  j  SS PBCH TX  ic  – TX  jc  j i j  TX  jc   E +f PDSCH PDSCH N j ----------------------------------------------------------------------- TX  ic  – TX j  jc  10 Sym – PDSCH  +f i .3.+ 10 ----------------------------10  Log  10   TX j  jc  TX j  jc   N Sym – DL N Sym – DL   TX j  jc  TX i  ic  – TX j  jc   E PDSCH + f PDSCH TX j  jc   ----------------------------------------------------------------------TX  ic  – TX j  jc  N 10 Sym – PDSCH  . 10  ----------------------------------------10  Log  ------------------O TX  jc  TX  ic   j i  N Ant – TX  N Sym – PDSCH   TX  jc  TX  ic  – TX  jc  j i j  TX  ic   E PDSCH + f PDSCH TX  jc  j N i ----------------------------------------------------------------------- – N 10 Sym – PDSCH Sym – DLRS in PDSCH   10 + 10  L og  ----------------------------------------------------------------------------TX  ic    i N Sym – PDSCH     Interfering energy per resource element (dBm/Sym) received over the PDSCH (Method 1: synchronised dBm/Sym transmission and reception) Case: PDSCH of the interfered cell collides with PDSCH and all the RS of the interfering cell TX  jc  TX j  jc   PDSCH j  TX  ic  E DLRS  TX  jc  TX  ic  j i N i --------------------- N – N Ant – TX Sym – DLRS in PDSCH Sym – DLRS in PDSCH .Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks Name ©Forsk 2015 Value Unit Description TX  jc  TX j  jc   PDSCH j   E TX  jc  DLRS  j  -------------------TX  ic  – TX  jc  N 1 j 10 Sym – DLRS in PDSCH  +f i .+ fO i  ---------------------------+ 10 TX j  jc   N Sym – DL   dBm/Sym Interfering energy per resource element (dBm/Sym) received over the PDCCH (Method 2: non-synchronised transmission and reception) . 10 + 10  L og  ---------------------------O TX  ic    i  N Sym – PDSCH    Interfering energy per resource element (dBm/Sym) received over the PDSCH (Method 1: synchronised dBm/Sym transmission and reception) Case: PDSCH of the interfered cell collides with PDSCH and some RS of the interfering cell TX  jc  j TX  jc  TX  jc  j TX  ic  – TX  jc  j i j  E j E PDCCH + f PDCCH TX j  jc  TX j  jc  DLRS  -----------------------------------------------------------------------------------------N Sym – DLRS N Sym – PDCCH 10 10 -------------------------. N j TX  ic  – TX j  jc  10 Sym – PDSCH  .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value TX  jc  TX j  jc   PDSCH TX j  jc  TX j  jc  TX i  ic  – TX j  jc  f PDCCH TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  dBm/Sym Interfering energy per resource element (dBm/Sym) received over the PDSCH (Method 2: non-synchronised transmission and reception)  E PDSCH + f PDSCH TX  jc   ----------------------------------------------------------------------.6 C/N Calculation (DL) Name TX i  ic  CNR DLRS Value TX i  ic  TX i  ic  E DLRS – n Sym Unit Description dB Downlink reference signals C/N 459 .+ fO i + 10  ---------------------------TX  jc   j N Sym – DL   TX i  ic  – TX j  jc   EPDSCH + fPDSCH TX j  jc   ---------------------------------------------------------------------10  N Sym – PDSCH  10  10 10  Log  --------------------------------------------------------------------------------------------TX j  jc  TX j  jc   N + N Sym – PDSCH Sym – PDCCH   Interfering energy per frequency block (dBm/RB) received over 1 frequency block during an OFDM symbol carrying reference signals dBm/RB  E DLRS TX j  jc   For number of antenna ports > 1. -------------------TX j  jc   TX i  ic  – TX j  jc   N Sym – PDCCH 10 10   . 10 + 10 + -------------------------------------------------------------------------------------------- 2  Min 2 N Ant – TX  + f O 8 is used instead of encircled 10 TX j  jc  TX j  jc    N Sym – PDSCH + N Sym – PDCCH   TX j  jc  TX i  ic  – TX j  jc  E PDCCH + f PDCCH ----------------------------------------------------------------------10 TX j  jc  TX  jc  TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  j j i j i j   f MIMO + f TL + f ICIC – DL + f ABS – DL  ------------------------------------------------------------------------------------------------------------------------------------------------- TX  jc  10   1 – AU j   10  DL      TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc    f + f + f   TL ICIC – DL ABS – DL -----------------------------------------------------------------------------------------------------------------------TX  jc    j 10 +  10 AU   DL TX  jc  f PDSCH Description TX  ic  – TX  jc  j i j  E j E +f TX  jc  TX  jc  DLRS PDCCH PDCCH j j  -----------------------------------------------------------------------------------------N Sym – DLRS N Sym – PDCCH 10 10 .3.+ 10 ---------------------------- ------------------------ 10  Log  10 TX  jc  TX  jc  j j  N Sym – DL N Sym – DL  TX j  jc   RSSI TX  jc  Unit TX  jc  TX  ic  – TX  jc  TX i  ic  – TX j  jc  10  Log  r O  TX  ic  – TX  jc  i j f ICIC – DL 10  Log  p Collision  PDCCH interference weighting factor dB PDSCH interference weighting factor dB Interference reduction factor due to channel overlap dB Interference reduction factor due to static downlink ICIC using fractional frequency reuse TX  ic  – TX  jc  j j i j i j   f MIMO + f TL + f ICIC – DL + f ABS – DL -------------------------------------------------------------------------------------------------------------------------------------------------  TX j  jc  10    1 – AU DL    10     TX TX TX  ic  – TX  jc  TX  ic  – TX  jc     j j  i j i j G SA    – G SA    + f ICIC – DL + f ABS – DL      -------------------------------------------------------------------------------------------------------------------------------------------------------  TX  jc  10  + AU DLj  10  fO dB   TX i  ic  – TX j  jc    TX  jc  TX j  jc  j 10  Log  TLDL   dB Interference reduction factor due to the downlink traffic load f MIMO TX j  jc  j 10  Log  N Ant – TX dB Interference increment due to more than one transmission antenna port Inter – Tech I DL TX k   P DL – Rec  -------------------------------------- F  TX i  ic  TX k    TX k  ICP DL W Downlink inter-technology interference f TL TX  jc   6.Atoll 3.2. 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Name Value TX  ic  i TX  ic  i CNR SS TX  ic  i CNR PBCH E SS TX  ic  i TX  ic  i CNR PDSCH Description dB SS C/N dB PBCH C/N dB PDCCH C/N dB PDSCH C/N Unit Description TX  ic  i E PBCH – n Sym TX i  ic  TX i  ic  Mi DL With MIMO: CNR PBCH = CNR PBCH + G Div – PBCH + G Div TX i  ic  TX i  ic  E PDCCH – n Sym TX i  ic  CNR PDCCH TX  ic  i – n Sym Unit TX i  ic  TX i  ic  Mi DL With MIMO: CNR PDCCH = CNR PDCCH + G Div – PDCCH + G Div TX i  ic  TX i  ic  E PDSCH – n Sym TX i  ic  TX i  ic  Mi DL With MIMO: CNR PDSCH = CNR PDSCH + G Div – DL + G Div 6.2.7 C/(I+N) Calculation (DL) Name Value TX  jc  TX i  ic  CINR DLRS TX  ic  i     j   n Sym  DLRS   ------------------  --------------------- TX i  ic    10 10  + I Inter – Tech + 10 10  + NR Inter – Tech  dB E DLRS –  10  Log  DL     DL      All TXj  jc            TX  jc  TX  ic  i CINR SS TX  ic  i   j     n Sym  SS PBCH  ------------------------  --------------------- TX i  ic   Inter – Tech Inter – Tech 10 10  10  +I  + NR dB  E SS + 10 –  10  Log  DL DL          All TXj  jc            TX  jc  TX i  ic  CINR PBCH Downlink reference signals C/(I+N) SS C/(I+N) TX  ic  i   j     n Sym  SS PBCH  ------------------------  --------------------- TX i  ic   Inter – Tech Inter – Tech 10 10  10  +I  + NR  E PBCH –  10  Log  + 10 DL DL       dB     All TXj  jc           TX  ic  i TX  ic  i M PBCH C/(I+N) DL i With MIMO: CINR PBCH = CINR PBCH + G Div – PBCH + G Div TX  jc  TX i  ic  CINR PDCCH TX  ic  i   j     n Sym  PDCCH-  -------------------   --------------------- TX i  ic  Inter – Tech Inter – Tech 10 10  10  +I  + NR  E PDCCH –  10  Log  + 10 DL DL     dB    All TXj  jc            TX i  ic  TX i  ic  Mi PDCCH C/(I+N) DL With MIMO: CINR PDCCH = CINR PDCCH + G Div – PDCCH + G Div TX  jc  TX i  ic  CINR PDSCH TX  ic  i      j  n Sym  PDSCH-    ---------------------------------------- TX i  ic    10 10  + I Inter – Tech + 10 10  + NR Inter – Tech E PDSCH –  10  Log  DL     DL dB   All TXj  jc            TX i  ic  TX i  ic  Mi PDSCH C/(I+N) DL With MIMO: CINR PDSCH = CINR PDSCH + G Div – DL + G Div RSRQ 460 TX i  ic  TX i  ic  10  Log  N FB  TX  ic   + E i – RSSI DLRS  TX i  ic  dB Reference signal received quality (RSRQ) .Atoll 3. + NR Inter + 10  Log ------------------------------------------------------------------DL TX  ic    i  N SD  Slot  N Slot  SF  N SF – DL TX  jc  TX i  ic   I + N  PDSCH TX  ic  i    j   n PDSCH Sym   ---------------------------------------- Inter – Tech 10 10  10  +I  10  Log  + 10    DL   All TXj  jc          TX i  ic   N Sym – PDSCH  – Tech  ---------------------------------------------.Atoll 3.+ NR Inter + 10  Log  TX  ic  DL TX i  ic   i  N SF – DL + N TDD – SSF TX  jc  I + TX  ic  i N  PDCCH TX  ic  i    j  n Sym  PDCCH   ----------------------------------------  10 10  + I Inter – Tech + 10 10  10  Log     DL  All TXj  jc          TX  ic  i  N TXi  ic   – Tech Sym – PDSCH + N Sym – PDCCH   .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value TX  jc  Unit Description dBm Received signal strength indicator (RSSI) dBm Downlink reference signals total noise (I+N) dBm SS and PBCH total noise (I+N) dBm PDCCH total noise (I+N) (Method 1: synchronised transmission and reception) dBm PDCCH total noise (I+N) (Method 2: non-synchronised transmission and reception) dBm PDSCH total noise (I+N) (Method 1: synchronised transmission and reception) TX  ic  i     j  n RSSI  Sym  TX  ic    --------------------------------------i Inter – Tech 10 10  10  +I 10  Log   RSSI + + 10  12 + DL      All TX j  jc        RSSI TX i  ic  Inter – Tech NR DL TX  ic  i + 10  Log  N FB   TX  jc  TX  ic  i    j  n Sym  DLRS -   ---------------------------------------  10 10  + I Inter – Tech + 10 10  10  Log    DL   All TXj  jc          TX i  ic   I + N  DLRS Inter – Tech + NR DL TX  ic  i + 10  Log  2  N FB TX  jc    TX  ic  i   j    n SS PBCH Sym  ------------------------ --------------------- Inter – Tech 10 10  10  +I  10  Log  + 10    DL    All TXj  jc         TX i  ic   I + N  SS PBCH Inter – Tech + NR DL + 10  Log  N SCa – FB  N FB – SS PBCH  TX  jc  TX i  ic   I + N  PDCCH TX  ic  i    j  n Sym  PDCCH-   ----------------------------------------  10 10  + I Inter – Tech + 10 10  10  Log     DL  All TXj  jc          TX i  ic   N Sym – PDCCH  – Tech  ---------------------------------------------.+ NR Inter + 10  Log TX  ic  DL TX i  ic    i  N SF – DL + N TDD – SSF 461 .3. Atoll 3.2.2.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Name Value TX  jc  I + TX  ic  i N  PDSCH Unit Description dBm PDSCH total noise (I+N) (Method 2: non-synchronised transmission and reception) Unit Description dBm Nominal PUSCH power dBm Maximum allowed transmission power of a user equipment dBm Received PUSCH and PUCCH signal level dBm PUSCH and PUCCH EIRP of a user equipment dB Cyclic prefix factor.e.10 Interference Calculation (UL) Name Mj I PUSCH PUCCH TX i  ic  – TX j  jc  fO M j f TL – UL 462 Value Mj TX i  ic  – TX j  jc  C PUSCH PUCCH + f O TX i  ic  – TX j  jc  Mj + f TL – UL + f ICIC – UL TX i  ic  – TX j  jc  10  Log  r O    M j 10  Log  TL UL   .8 Signal Level Calculation (UL) Name TX i  ic  P O_PUSCH Mi P Allowed Value TX i  ic  TX i  ic  CINR PUSCH – Max + NR UL TX i  ic  TX i  ic  + n PUSCH PUCCH – 10  Log  N FB  TX i  ic  TX i  ic  TX i  ic   Mi  Min  P Max 10  Log  N FB  + P O_PUSCH +  FPC  L Total    Mi Mi C PUSCH PUCCH TX i EIRP PUSCH PUCCH – L Path – M Shadowing – Model – L Indoor + G Ant –L TX i Mi Mi – L Ant – L Body + f CP P Mi EIRP PUSCH PUCCH   With P Mi Mi +G Mi –L Mi Mi = P Allowed without power control adjustment and P Mi Mi = P Eff after power control adjustment TX i  ic  10  Log  7  7. the ratio of the useful symbol energy to the total symbol energy Unit Description dBm PUSCH and PUCCH thermal noise dBm PUSCH and PUCCH noise Unit Description dBm Received PUSCH and PUCCH interference dB Interference reduction factor due to the co.9 Noise Calculation (UL) Name Value TX i  ic  TX i  ic  n 0 + 10  Log  N FB n 0 – PUSCH PUCCH TX i  ic   W FB  1000 TX i  ic  n 0 – PUSCH PUCCH + nf n PUSCH PUCCH TX i  ic  6. i.+ NR Inter + 10  Log DL TX  ic    i  N SD  Slot  N Slot  SF  N SF – DL 6.3.2.5  If D CP = Normal = Extended If M i is an interferer 0 6.and adjacent channel overlap dB Interference reduction factor due to the interfering mobile’s uplink traffic load TX  ic  i     j  n PDSCH Sym   ---------------------------------------- Inter – Tech 10 10   10  +I 10  Log  + 10    DL  All TXj  jc          TX  ic  i  N TXi  ic   – Tech Sym – PDSCH + N Sym – PDCCH   ------------------------------------------------------------------.5  If D CP f CP TX i  ic  10  Log  6  7.. 2.3.11 Noise Rise Calculation (UL) Name Value M TX  ic  i NR UL j   TX i  ic   IPUSCH PUCCH    non-ICIC M i n PUSCH PUCCH   - -------------------------------------------  ----------------------------------------------------------------------------10 10  10  Log  10 + 10        All M j        All TX  jc    j  Inter – Tech + NR UL TX i  ic  – n PUSCH PUCCH M j   TX i  ic   IPUSCH PUCCH    n  ICIC M   PUSCH  PUCCH i  ------------------------------------------------------------------- -------------------------------------------  10 10  10  Log   10  + 10      All Mj        All TXj  jc     TX i  ic  NR UL – ICIC Inter – Tech + NR UL TX i  ic  – n PUSCH PUCCH For any mobile Mi in cell centre of the interfered cell TXi(ic): I + TX  ic  i N  PUSCH PUCCH TX i  ic  NRUL TX i  ic  + n PUSCH PUCCH For any mobile Mi in cell-edge of the interfered cell TXi(ic): TX i  ic  TX i  ic  NRUL – ICIC + n PUSCH PUCCH 6.Atoll 3.12 C/N Calculation (UL) Name Value TX i  ic  Mi C PUSCH PUCCH – n PUSCH PUCCH Mi CNR PUSCH PUCCH With MIMO: Mi TX i  ic  Mi UL CNR PUSCH PUCCH = CNR PUSCH PUCCH + G Div – UL + G Div 6.2.13 C/(I+N) Calculation (UL) Name Value For any mobile Mi in cell centre of the interfered cell TXi(ic): TX i  ic  Mi CNR PUSCH PUCCH – NR UL Mi For any mobile Mi in cell-edge of the interfered cell TXi(ic): CINR PUSCH PUCCH TX i  ic  Mi CNR PUSCH PUCCH – NR ICIC – UL With MIMO: Mi Mi TX i  ic  UL CINR PUSCH PUCCH = CINR PUSCH PUCCH + G Div – UL + G Div 463 .2.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value TX  ic  – TX  jc  j  TX  ic  – TX  jc  i j i 10  Log  p Collision f ICIC – UL  Unit Description dB Interference reduction factor due to static uplink ICIC using fractional frequency reuse Unit Description dB Uplink noise rise for any mobile Mi in cell centre of the interfered cell TXi(ic) dB Uplink noise rise for any mobile Mi in cell-edge of the interfered cell TXi(ic) dBm PUSCH and PUCCH total noise (I+N) Unit Description dB PUSCH and PUCCH C/N Unit Description dB PUSCH and PUCCH C/(I+N) 6. Atoll 3.3.2.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Name Value Unit Description M M  Mi   TXi  ic    Mi  i Max  PAllowed –  CINR PUSCH PUCCH –  T M + M PC   P Min    B i   dBm Effective transmission power of a user equipment after power control adjustment i P Eff UL 6.14 Calculation of Downlink Cell Resources Name Value Unit Description N Sym  SRB N SCa – FB  N SD  Slot  N Slot  SF None Number of modulation symbols per scheduler resource block N Sym  SSF DwPTS N SCa – FB  N SD  SSF None Number of DwPTS modulation symbols per scheduler resource block in the TDD special subframes N SCa – FB W FB --------F None Number of subcarriers per frequency block None Total number of modulation symbols in downlink None Number of modulation symbols in DwPTS None Number of PDSCH modulation symbols None Number of PDSCH modulation symbols in the DwPTS None Downlink reference signals overhead None Downlink reference signals overhead in the DwPTS None Number of symbols reserved for downlink reference signals in one scheduler resource block DwPTS TX  ic  i TX  ic  i N Sym – DL N FB TX i  ic  TX i  ic  N Sym – DwPTS TX i  ic  R DL TX  ic  i N FB TX i  ic  TX  ic  i  N Sym  SRB  N SF – DL + N Sym – DwPTS TX i  ic  DwPTS  N TDD – SSF  N Sym  SSF TX i  ic  TX i  ic  TX i  ic  TX i  ic  N Sym – DL – O DLRS – O PSS – O SSS – O PBCH – O PDCCH – O UERS TX i  ic  TX i  ic  TX i  ic  TX i  ic  N Sym – DwPTS – O DLRS  DwPTS – O PDCCH  DwPTS R DwPTS TX  ic  i TX  ic  i O DLRS N FB TX i  ic  TX  ic  i TX i  ic  O DLRS  DwPTS N FB    8    16     24  TX i  ic  N DLRS  SRB TX  ic  i TX  ic  i  N DLRS  SRB  N SF – DL + O DLRS  DwPTS TX i  ic  TX i  ic   N DLRS  DwPTS  N TDD – SSF TX  ic  i if  N Ant – TX = 1 TX  ic  i if  N Ant – TX = 2   TX  ic  i if  N Ant – TX = 4 or 8   N DLRS  DwPTS See "Calculation of Downlink Cell Resources" on page 538 None Number of symbols reserved for downlink reference signals in DwPTS of one TDD special subframe O PSS 2  NFB – SS PBCH  N SCa – FB = 144 None PSS overhead O SSS 2  NFB – SS PBCH  N SCa – FB = 144 None SSS overhead Extended CP: 216 Normal CP: 240 None PBCH overhead None PDCCH overhead TX i  ic  TX i  ic  O PBCH TX  ic  i if  N SD – PDCCH = 0 : 0   TX  ic  TX i  ic  O PDCCH TX  ic  i i if  N SD – PDCCH = 1 AND  N Ant – TX  2 :     TX  ic  TX  ic  TX  ic  TX  ic  i i i N i  N SCa – FB – 4  N FB  N SF – DL + O PDCCH  DwPTS  SD – PDCCH  Otherwise: TX  ic  TX  ic  TX  ic  i i N i    SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB 464 TX i  ic  TX i  ic   N SF – DL + O PDCCH  DwPTS . 2.2. N Sym – UL N Sym  SRB None Uplink sounding reference signal overhead TX  ic  i TX i  ic  N SCa – FB .Atoll 3.2.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value Unit Description TX  ic  i if  N SD – PDCCH = 0 : 0   TX  ic  TX i  ic  O PDCCH  DwPTS TX  ic  i i if  N SD – PDCCH = 1 AND  N Ant – TX  2 :     TX  ic  TX  ic  i N i  N SCa – FB – 4  N FB  SD – PDCCH  Otherwise: TX  ic  PDCCH overhead in the DwPTS TX i  ic   N TDD – SSF TX  ic  TX  ic  i i  Min  2 N i    SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB   TX  ic  i  N TDD – SSF Without smart antennas and MIMO: 0 TX i  ic  With smart antennas and without MIMO: 12  N FB TX  ic  i O DMRS TX  ic  i With smart antennas and with MIMO: 24  N FB TX i  ic   N SF – DL TX  ic  i  N SF – DL None UE-specific reference signals or demodulation reference signal overhead Without smart antennas and with SU-MIMO or MU-MIMO and TX i  ic  TX i  ic  N Ant – TX  4 : 24  N FB TX i  ic   N SF – DL 6.3.15 Calculation of Uplink Cell Resources Name Value Unit Description N Sym  SRB N SCa – FB  N SD  Slot  N Slot  SF None Number of modulation symbols per scheduler resource block N SCa – FB W FB --------F None Number of subcarriers per frequency block None Total number of modulation symbols in uplink N Sym – UL – O ULSRS – O ULDRS None Nnumber of PUSCH modulation symbols TX i  ic  TX i  ic  N SCa – FB --------------------. N Sym 2  --------------------– UL N Sym  SRB None Uplink demodulation reference signal overhead Unit Description bps Maximum downlink throughput capacity of a UE category Unit Description bps Maximum uplink throughput capacity of a UE category TX i  ic  N Sym – UL TX i  ic  R UL O ULSRS O ULDRS TX  ic  N i  FB TX  ic  TX  ic  i i – N FB – PUCCH  N Sym  SRB  N SF – UL  TX i  ic  TX i  ic  TX i  ic  6.16 Calculation of Downlink UE Capacity Name Value Max TP UE – DL i N i + N TDD – SSF  Max – DL  SF – DL N TBB  TTI  ---------------------------------------------------D Frame TX  ic  TX  ic  6.17 Calculation of Uplink UE Capacity Name Value Max Max – UL N SF – UL N TBB  TTI  ---------------D Frame TP UE – UL TX i  ic  465 . Cell Capacity.– TPOffset PUTP E – DL  -----------------------100 Mi PUTP A – DL TX  ic  i R UL  M i B UL --------------------------------D Frame TX  ic  i R UL  B Mi CTP P – UL Mi TX  ic  i UL For proportional fair schedulers: --------------------------------.– TP Offset Cap E – DL  -----------------------100 Mi Cap A – DL M Mi Mi f TP – Scaling .3.Atoll 3. G MUG – UL D Frame With SU-MIMO:  Mi B UL =  Max – TX  ic  Mi B UL i   1 + f SU – MIMO  G SU – MIMO – UL – 1  With MU-MIMO in throughput coverage predictions: Mi TX i  ic  CTP P – UL  G MU – MIMO – UL 466 .– TP Offset CTP E – DL  -----------------------100 Mi CTP A – DL M M i i CTP P – DL   1 – BLER  B DL     i CTP E – DL Mi Cap P – DL ----------------------TX i  ic  N Users – DL i PUTP P – DL Mi Cap E – DL ----------------------TX i  ic  N Users – DL Mi PUTP E – DL M Mi i Mi f TP – Scaling . G MUG – DL D Frame With SU-MIMO:  Mi =  B DL Max – M Mi B DL i   1 + f SU – MIMO  G SU – MIMO – DL – 1  With MU-MIMO in throughput coverage predictions: TX i  ic  Mi CTP P – DL  G MU – MIMO – DL M M Mi Mi TX i  ic  Mi CTP P – DL  TL DL – Max Cap P – DL M M i i Cap P – DL   1 – BLER  BDL     i Cap E – DL Mi Mi Mi f TP – Scaling .18 Channel Throughput.2. Allocated Bandwidth Throughput. and Per-user Throughput Calculation Name Value TX i  ic  R DL  Unit Description kbps Downlink peak RLC channel throughput kbps Downlink effective RLC channel throughput kbps Downlink application channel throughput kbps Downlink peak RLC cell capacity kbps Downlink effective RLC cell capacity kbps Downlink application cell capacity kbps Downlink peak RLC throughput per user kbps Downlink effective RLC throughput per user kbps Downlink application throughput per user kbps Uplink peak RLC channel throughput M i B DL --------------------------------D Frame TX  ic  i R DL Mi CTP P – DL  M i B DL TX  ic  i For proportional fair schedulers: --------------------------------.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 6. – TP Offset ABTP E – UL  -----------------------100 Mi M M f TP – Scaling i i .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value M i i CTP P – UL   1 – BLER  B UL     Description kbps Uplink effective RLC channel throughput kbps Uplink application channel throughput kbps Uplink peak RLC cell capacity kbps Uplink effective RLC cell capacity kbps Uplink application cell capacity kbps Uplink peak RLC allocated bandwidth throughput kbps Uplink effective RLC allocated bandwidth throughput kbps Uplink application allocated bandwidth throughput  Cap Mi  M P – UL - ABTP P –i UL Min  ---------------------- TX i  ic    N Users – UL  kbps Uplink peak RLC throughput per user  Cap Mi  M E – UL - ABTP E –i UL Min  ---------------------- TXi  ic    N Users – UL  kbps Uplink effective RLC throughput per user kbps Uplink application throughput per user M i CTP E – UL M Unit M M i CTP A – UL Mi M i CTP E – UL i M f TP – Scaling i .Atoll 3.– TP Offset  -----------------------100 TX i  ic  Mi Cap P – UL CTP P – UL  TL UL – Max M i i Cap P – UL   1 – BLER  B UL     M i Cap E – UL Mi Cap A – UL M i ABTP P – UL M i ABTP E – UL Mi ABTP A – UL Mi PUTP P – UL Mi PUTP E – UL Mi PUTP A – UL M Mi Mi f TP – Scaling .3.– TP Offset Cap E – UL  -----------------------100 Mi Mi N FB – UL CTP P – UL  ----------------TX i  ic  N FB Mi M M i i ABTP P – UL   1 – BLER  B UL     Mi Mi Mi f TP – Scaling .– TP Offset PUTP E – UL  -----------------------100 6.19 Scheduling and Radio Resource Management Name Value Unit Description Sel Mi R Min – DL TPD Min – DL --------------------------- None Resources allocated to a mobile to satisfy its minimum throughput demand in downlink None Resources allocated to a mobile to satisfy its minimum throughput demand in uplink R Min – DL None Remaining downlink cell resources after allocation for minimum throughput demands Sel i R Min – UL None Remaining uplink cell resources after allocation for minimum throughput demands kbps Remaining throughput demand for a mobile in downlink Sel Mi Sel Mi CTP P – DL Sel Mi TPD Min – UL --------------------------- Sel Mi R Min – UL Sel Mi CTP P – UL TX i  ic  R Rem – DL TX i  ic  R Rem – UL Sel i TPD Rem – DL M TX i  ic  TL DL – Max –  Sel Mi Sel Mi TX  ic  i TL UL – Max –  M Sel Mi Sel Sel Mi Mi  Max  Min  TPD Max – DL – TPD Min – DL TP UE – DL   467 .2. 0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Name Value Sel i TPD Rem – UL M M  Max  i i Min  TPD Max – UL – TPD Min – UL TP UE – UL   M Sel Mi Sel Sel Mi TX i  ic  CTP P – DL Without MUG  G MUG – DL Sel i CTP P – UL Sel i CTP P – UL Without MUG TX  ic  i G MUG – UL Sel Mi RD Rem – DL Description kbps Remaining throughput demand for a mobile in uplink kbps Downlink peak channel throughput with multi-user diversity gain (Proportional Fair) kbps Uplink peak channel throughput with multi-user diversity gain (Proportional Fair) None Remaining resource demand for a mobile in downlink None Remaining resource demand for a mobile in uplink None Resources allocated to a mobile to satisfy its maximum throughput demand in downlink None Resources allocated to a mobile to satisfy its maximum throughput demand in uplink None Effective remaining downlink resources in a cell (Proportional Demand) Sel CTP P – DL M Unit M  Sel Mi TPD Rem – DL ---------------------------Sel Mi CTP P – DL Sel Mi RD Rem – UL Sel Mi TPD Rem – UL ---------------------------Sel Mi CTP P – UL Sel TX i  ic   Mi R Rem – DL Proportional Fair: Min  RD Rem – DL --------------------- N   Sel TX i  ic   Mi R Rem – DL - Round Robin: Min  RD Rem – DL -------------------N   Sel Mi R Max – DL Sel Mi TX i  ic  RD Rem – DL Proportional Demand: R Eff – Rem – DL  ---------------------------------Sel Mi  RDRem – DL Sel Mi Sel i TPD Rem – DL ---------------------------Sel M i CTP P – DL M Max C/I: Sel TX i  ic   Mi R Rem – UL - Proportional Fair: Min  RD Rem – UL -------------------N   Sel TX  ic  i  Mi R Rem – DL - Round Robin: Min  RD Rem – DL -------------------N   Sel Mi R Max – UL Sel Mi TX i  ic  RD Rem – UL Proportional Demand: R Eff – Rem – UL  ---------------------------------Sel Mi  RDRem – UL Sel Mi Sel Mi TPD Rem – UL Max C/I: --------------------------Sel Mi CTP P – UL TX i  ic  R Eff – Rem – DL 468   Sel Mi  TXi  ic   Min  R Rem – DL RD Rem – DL   Sel   M i  .Atoll 3.3. Atoll 3.– TP Offset UTP E – DL  -----------------------100 Sel Mi R UL Sel Mi  CTP P – UL Sel Sel M  Mi    i UTP P – UL   1 – BLER  B UL      Sel Mi Sel Mi Sel Mi f TP – Scaling UTP E – UL  ------------------------.– TP Offset 100 469 .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Name Value Unit Description TX  ic  i R Eff – Rem – UL   Sel M  TXi  ic   i Min  R Rem – UL RD Rem – UL   Sel   M i None Effective remaining uplink resources in a cell (Proportional Demand) Sel Sel   Mi   Mi    R Max – DL  CTP E – DL     Sel   M  Site i - Max  1 ----------------------------------------------------------------------------------------------------Sel Sel  Mi    Mi  Site TP – R  CTP   S1 – DL Min – DL E – DL     Sel   M  Site i None Site backhaul overflow ratio in downlink Sel Sel   Mi   Mi    R Max – UL  CTP E – UL     Sel   M i  Site Max  1 ------------------------------------------------------------------------------------------------------  Sel Sel  Mi    Mi  Site  R Min – UL  CTP E – UL   TPS1 – UL –    Sel   M i  Site None Site backhaul overflow ratio in uplink None Total resources assigned to a mobile in downlink (Downlink traffic load of the mobile) None Total resources assigned to a mobile in uplink (Uplink traffic load of the mobile) Unit Description kbps Downlink peak RLC user throughput kbps Downlink effective RLC user throughput kbps Downlink application user throughput kbps Uplink peak RLC user throughput kbps Uplink effective RLC user throughput kbps Uplink application user throughput   Site BHOF DL   Site BHOF UL  Sel Sel Mi TL DL Sel Mi = R DL M Sel i M Sel i R  Mi   Mi Max – DL  CTP P – DL  R Min – DL  CTP P – DL +  -----------------------------------------------Site     BHOF DL  -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – DL Sel Sel i TL UL M = Sel i R UL M Sel Mi Sel Mi R  Mi   Mi Max – UL  CTP P – UL  R Min – UL  CTP P – UL +  -----------------------------------------------Site     BHOF UL  -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – UL 6.20 User Throughput Calculation Name Sel Mi UTP P – DL Sel Mi UTP E – DL Sel Mi UTP A – DL Sel Mi UTP P – UL Sel Mi UTP E – UL Sel i UTP A – UL M Value Sel Mi R DL Sel Mi  CTP P – DL Sel Sel Mi   Mi   UTP P – DL   1 – BLER  B DL      Sel Mi Sel Mi Sel Mi f TP – Scaling .2.3. 3 Interference View Analysis provided in the interference view is based on path loss matrices.G M i M i M i .3. see the Administrator Manual.3. As well. L Ant . For each cell.3 Available Calculations 6. The bar graph displays cells whose received RSRP are higher than their minimum RSRP thresholds and are within a 30 dB margin from the studied signal level of the best server. L M i TX  ic  i • Downlink reference signal level C DLRS • Path loss L Path • Total losses L Total . Reception level bar graphs show the RSRP or signal levels in decreasing order. For more information on defining a different value for this margin. see the Administrator Manual.1. The bar graph displays cells whose C/ N are higher than the minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the highest interference level on the studied channel.3.4 Details View Analysis provided in the details view is based on path loss matrices. 6. for example a smaller value for improving the calculation speed. or PDSCH signal levels. 6.1. You can use a value other than 30 dB for the margin from the highest interference level on the studied channel.2 Reception View Analysis provided in the reception view is based on path loss matrices. You can use a value other than 30 dB for the margin from the studied signal level of the best server. So.1. Interference level bar graphs show the interference levels on different channels in decreasing order. All the cells from which the received RSRP is higher than their minimum RSRP thresholds are listed in the table. interference values are listed for all the cells whose C/N are higher than the minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the highest interference level on RS. Atoll displays the best server RS. for example a smaller value for improving the calculation speed. or PDSCH signal level. and f CP are not used in the calculations performed for the profile view. for example a smaller value for improving the calculation speed.3. you can display received downlink reference signal levels from the cells for which calculated path loss matrices are available. you can display the received signal level from the best server and interfering signal levels from other cells for which calculated path loss matrices are available.3. Atoll displays the received RSRP or reference signal.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 6. Other cells are listed in the decreasing order of RSRP. PBCH. 6.1 Profile View The point analysis profile view displays the following calculation results for the selected transmitter based on the calculation algorithm described in "Signal Level Calculation (DL)" on page 501. as well as interference levels on these channels from interfering cells. see the Administrator Manual. L Body . 470 . SS. SS. So.Atoll 3. SS. For each cell. For more information on defining a different value for this margin. PDCCH.3. and PDSCH signal levels. For each cell. For more information on defining a different value for this margin.1. The maximum number of bars in the graph depends on the studied signal level of the best server. The results for the best server (first row) are displayed using bold italic characters. and interference from other cells. You can use a value other than 30 dB for the margin from the highest interference level on RS. Atoll displays the RSRP and RS. So. you can display the received signal level from the best server and interfering signal levels from other cells for which calculated path loss matrices are available. The maximum number of bars in the graph depends on the highest interference level on the studied channel.1 Point Analysis 6. 0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 6. L Ant . Atoll determines the selected display parameter on each pixel inside the cell’s calculation area. and L Body are not considered in the calculations performed for the downlink signal level based coverage predictions. "Coverage Display Types" on page 472.3. Then.3. The Best function considers the highest value from a list of values. TX  ic  TX  ic  TX  ic  i i i MinimumThreshold  C DLRS  or L Total or L Path   MaximumThreshold   AND TX i  ic  TX j  jc  C DLRS  Best  C DLRS  – M  ji  Where M is the specified margin (dB). • • • If M = 0 dB. Coverage Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine coverage areas to display. There are three possibilities.2 Coverage Predictions 6. If M = 2 dB. For more information on coverage area determination and available display options. TX  ic  TX  ic  TX  ic  i i i MinimumThreshold  C DLRS  or L Total or L Path   MaximumThreshold AND 471 . G Mi Mi Mi . Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the highest. • All Servers The coverage area of each cell TXi(ic) corresponds to the pixels where. L Mi . Therefore. see "Path Loss Calculation Prerequisites" on page 57 for more information). Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. Atoll calculates the received downlink reference signal level.2. see "Signal Level Calculation (DL)" on page 501. These coverage predictions do not depend on the traffic input. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. these calculations are of special interest before and during the deployment stage of the network to study the coverage footprint of the system. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the highest or within a 2 dB margin from the highest If M = -2 dB. For more information on downlink reference signal level calculations.1 Downlink Signal Level Coverage Predictions The following coverage predictions are based on the received downlink reference signal levels: • • • Coverage by Transmitter Coverage by Signal Level Overlapping Zones For these calculations. TX  ic  TX  ic  TX  ic  i i i MinimumThreshold  C DLRS  or L Total or L Path   MaximumThreshold   • Best Signal Level and a Margin The coverage area of each cell TXi(ic) corresponds to the pixels where.3. see: • • "Coverage Area Determination" on page 471.Atoll 3. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher than the received downlink reference signal levels from the cells which are 2nd best servers • Second Best Signal Level and a Margin The coverage area of each cell TXi(ic) corresponds to the pixels where. If M = -2 dB. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. For more information on coverage area determination and available display options. Number of Servers: Atoll evaluates the number of cells that cover a pixel (i. Each pixel within the calculation area of TXi(ic) is considered a noninterfering receiver.3.e. dBµV/m): Where cell coverage areas overlap. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the second highest. Atoll calculates the received signal level or C/N level at each pixel for the channel type being studied.e. Best Server Total Losses (dB): Where cell coverage areas overlap. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher than the received downlink reference signal levels from the cells which are 3rd best servers.e. a mobility type. PDSCH. dBµV. the cell with the highest downlink reference signal level) and evaluates the total losses from this cell.Atoll 3. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the second highest or within a 2 dB margin from the second highest. If M = 2 dB. Coverage consists of several independent layers that can be displayed and hidden on the map. and PUSCH and PUCCH signal levels and noise. and take into account the receiver characteristics ( L Mi .. 6. L Ant . Atoll keeps the highest value of the signal level. see: • • "C/N Calculation (DL)" on page 516. SS. These coverage predictions do not depend on the traffic input. PDSCH.e. . Atoll determines the best cell (i. the pixel falls within the coverage areas of these cells). Atoll determines the best cell (i. see "Path Loss Calculation Prerequisites" on page 57 for more information). The 2nd Best function considers the second highest value from a list of values. the cell with the highest downlink reference signal level) and evaluates the path loss from this cell.2. "Coverage Display Types" on page 473.. SS. It is possible to display the coverage predictions with colours depending on any transmitter or cell attribute. these calculations are of special interest before and during the deployment stage of the network to study the coverage footprint of the system. The properties of the non-interfering probe receiver are set by selecting a terminal. RS. "C/N Calculation (UL)" on page 529. PBCH. and a service. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction.. "Signal Level Calculation (UL)" on page 523. and other criteria such as: • • • • • • • Signal Level (dBm. see: • • 472 "Coverage Area Determination" on page 473.. For more information on C/N level calculations. and L Body ) when calculating the required parameter: • • Effective Signal Analysis (DL) Effective Signal Analysis (UL) For these calculations.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks TX  ic  ©Forsk 2015 TX  jc  nd i j C DLRS  2 Best  C DLRS  – M   ji Where M is the specified margin (dB). Path Loss (dB) Total Losses (dB) Best Server Path Loss (dB): Where cell coverage areas overlap. PDCCH. Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display parameter is greater than or equal to the defined thresholds values.3. i. dBµV. • • • If M = 0 dB. PUSCH and PUCCH. see: • • "Signal Level Calculation (DL)" on page 501.G Mi Mi Mi . Therefore. For more information on signal level calculations. dBµV/m) Best Signal Level (dBm.2 Effective Signal Analysis Coverage Predictions The following coverage predictions are based on the received downlink reference signal. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. The prediction is based on the delta path loss thresholds defined per cell. Total losses are calculated as explained in "Signal Level Calculation (DL)" on page 453. and bearer calculations. The downlink coverage predictions are based on the downlink traffic loads of the cells. It is possible to display the Effective Signal Analysis (UL) coverage prediction with colours depending on the following display options: • • PUSCH & PUCCH Signal Level (UL) (dBm) PUSCH & PUCCH C/N Level (UL) (dB) 6. 473 . Coverage consists of several independent layers that can be displayed and hidden on the map. L Ant . see "Path Loss Calculation Prerequisites" on page 57 for more information). Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. (I+N). • • • • • • • • Coverage by C/(I+N) Level (DL) Service Area Analysis (DL) Coverage by Throughput (DL) Coverage by Quality Indicator (DL) Coverage by C/(I+N) Level (UL) Service Area Analysis (UL) Coverage by Throughput (UL) Coverage by Quality Indicator (UL) These coverage predictions take into account the receiver characteristics ( L Mi . or set manually by the user for all the cells. the coverage area of each cell comprises the pixels where the cell is the best server.2. Atoll calculates the received signal level.e. It is possible to display the Effective Signal Analysis (DL) coverage prediction with colours depending on the following display options: • • • • • • • • • • • RSRP (RS EPRE) Level (DL) (dBm) RS Signal Level (DL) (dBm) SS Signal Level (DL) (dBm) PBCH Signal Level (DL) (dBm) PDCCH Signal Level (DL) (dBm) PDSCH Signal Level (DL) (dBm) RS C/N Level (DL) (dB) SS C/N Level (DL) (dB) PBCH C/N Level (DL) (dB) PDCCH C/N Level (DL) (dB) PDSCH C/N Level (DL) (dB) • Delta Path Loss (dB): Atoll calculates the difference of the total losses from the second best serving cells ( L Total ) and TX  jc  j TX i  ic  TX j  jc  TX i  ic  the total losses from the best serving cells ( L Total ) on each pixel of their coverage areas ( L Total – L Total ).3. C/(I+N). For these calculations. i.G Mi Mi Mi . "C/(I+N) and Bearer Calculation (UL)" on page 532. Pixels are coloured according to the colours of the transmitter symbols on the map.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Coverage Area Determination These coverage predictions are best server coverage predictions. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. a mobility type. noise. RSSI. and interference. The properties of the non-interfering probe receiver are set by selecting a terminal. These parameters can either be calculated by Atoll during the Monte Carlo simulations. ICIC Cell-edge Areas: Based on the delta path loss calculation as above.3. Pixels are • coloured according to the thresholds defined in the coverage prediction.3 C/(I+N)-based Coverage Predictions The following coverage predictions are based on the received signal levels. and interference at each pixel. see: • • "C/(I+N) and Bearer Calculation (DL)" on page 518. and a service.Atoll 3. and the uplink coverage predictions are based on the uplink noise rise values. For more information on RSRQ. Best server for each pixel is calculated as explained in "Best Server Determination" on page 535. Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display type parameter is greater than or equal to the defined thresholds values. total noise. and L Body ) when calculating the required parameter.. Allocated Bandwidth Throughput. For more information on coverage area determination and available display options. the coverage area of each cell comprises the pixels where the cell is the best server. for the calculated C/(I+N) and bearer. Atoll determines the best bearer available on each pixel.Atoll 3. it determines the value of the selected quality indicator from the quality graphs defined in the reception equipment of the selected terminal. It is possible to display the Coverage by C/(I+N) Level (DL) coverage prediction with colours depending on the following display options: • • • • • • • • • RSRQ Level (DL) (dB) RSSI Level (DL) (dBm) RS C/(I+N) Level (DL) (dB) SS C/(I+N) Level (DL) (dB) PBCH C/(I+N) Level (DL) (dB) PDCCH C/(I+N) Level (DL) (dB) SS & PBCH Total Noise (I+N) (DL) (dBm) PDSCH C/(I+N) Level (DL) (dB) PDSCH & PDCCH Total Noise (I+N) (DL) (dBm) It is possible to display the Service Area Analysis (DL) coverage prediction with colours depending on the following display options: • • • Bearer (DL) Modulation (DL): Modulation used by the bearer Service It is possible to display the Coverage by Throughput (DL) coverage prediction with colours depending on the following display options: • • • • • • • • • Peak RLC Channel Throughput (DL) (kbps) Effective RLC Channel Throughput (DL) (kbps) Application Channel Throughput (DL) (kbps) Peak RLC Cell Capacity (DL) (kbps) Effective RLC Cell Capacity (DL) (kbps) Application Cell Capacity (DL) (kbps) Peak RLC Throughput per User (DL) (kbps) Effective RLC Throughput per User (DL) (kbps) Application Throughput per User (DL) (kbps) It is possible to display the Coverage by Quality Indicator (DL) coverage prediction with colours depending on the following display options: • Quality indicators available in the document (Quality Indicators table): Atoll calculates the PDSCH C/(I+N) levels received from the best serving cells at each pixel of their coverage areas. Coverage consists of several independent layers that can be displayed and hidden on the map. see: • "Channel Throughput. and Per-user Throughput Calculation" on page 547. Then. of Frequency Blocks) • PUSCH & PUCCH C/(I+N) Level for 1 Frequency Block (UL) (dB): PUSCH & PUCCH C/(I+N) level with N FB – UL = 1 • Transmission Power (UL) (dBm) M i . It is possible to display the Coverage by C/(I+N) Level (UL) coverage prediction with colours depending on the following display options: 474 • • • PUSCH & PUCCH C/(I+N) Level (UL) (dB) PUSCH & PUCCH Total Noise (I+N) (UL) (dBm) Allocated Bandwidth (UL) (No. Coverage Area Determination These coverage predictions are all best server coverage predictions.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 For more information on thoughput calculations. see: • • "Coverage Area Determination" on page 474. i. Cell Capacity.e.3. From the C/(I+N). Best server for each pixel is calculated as explained in "Best Server Determination" on page 535.. Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display type parameter is greater than or equal to the defined thresholds values. "Coverage Display Types" on page 474. L G Mi Mi Mi . it determines the value of the selected quality indicator from the quality graphs defined in the reception equipment of the best serving cell. "Coverage Display Types" on page 476. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the highest or within a 2 dB margin from the highest If M = -2 dB. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. see: • • "Coverage Area Determination" on page 475. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher than the received downlink reference signal levels from the cells which are 2nd best servers 475 . • • • If M = 0 dB. see "Path Loss Calculation Prerequisites" on page 57 for more information). 6.3. Then. Atoll calculates the received downlink reference signal level then Atoll determines the selected display parameter on each pixel inside the cell’s calculation area. If M = 2 dB.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 It is possible to display the Service Area Analysis (UL) coverage prediction with colours depending on the following display options: • • • Bearer (UL) Modulation (UL): Modulation used by the bearer Service It is possible to display the Coverage by Throughput (UL) coverage prediction with colours depending on the following display options: • • • • • • • • • • • • Peak RLC Channel Throughput (UL) (kbps) Effective RLC Channel Throughput (UL) (kbps) Application Channel Throughput (UL) (kbps) Peak RLC Cell Capacity (UL) (kbps) Effective RLC Cell Capacity (UL) (kbps) Application Cell Capacity (UL) (kbps) Peak RLC Allocated Bandwidth Throughput (UL) (kbps) Effective RLC Allocated Bandwidth Throughput (UL) (kbps) Application Allocated Bandwidth Throughput (UL) (kbps) Peak RLC Throughput per User (UL) (kbps) Effective RLC Throughput per User (UL) (kbps) Application Throughput per User (UL) (kbps) It is possible to display the Coverage by Quality Indicator (UL) coverage prediction with colours depending on the following display options: • Quality indicators available in the document (Quality Indicators table): Atoll calculates the PUSCH and PUCCH C/(I+N) levels received at the best serving cells from each pixel of their coverage areas. Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the highest. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. Coverage Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine coverage areas to display. The coverage area of each cell TXi(ic) corresponds to the pixels where: TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX j  jc  MinimumThreshold  C DLRS  or L Total or L Path   MaximumThreshold AND C DLRS  Best  C DLRS  – M ji Where M is the specified margin (dB).2. see "Signal Level Calculation (DL)" on page 501. For more information on coverage area determination and available display options. Atoll determines the best bearer available on each pixel. L Ant . From the C/(I+N).3. Mi . For more information on downlink reference signal level calculations. The Best function considers the highest value from a list of values. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. It is possible to determine the coverage area based on the best signal level. and L Body are not considered in the calculations.Atoll 3.4 Cell Identifier Collision Zones Coverage Prediction The Cell Identifier Collision Zones coverage prediction is based on the received downlink reference signal levels. for the calculated C/(I+N) and bearer. 1. For more information. • Serving Base Station and Reference Cell as described in "Best Server Determination" on page 535.4. To have the same total number of users in each simulation of a group. 6. Allocated Bandwidth Throughput. • Generating a realistic user distribution as explained in "User Distribution" on page 476. add the following lines in the Atoll. and whether they are indoor or outdoor according to the percentage of indoor users per clutter class. • • "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 476. a second random trial is performed to obtain their geographical locations weighted according to the clutter classes. Atoll calculates the remaining parameters for each subscriber in the list that has a serving base station assigned. 6.ini file: [Simulation] RandomTotalUsers=0 6. using the properties of the default terminal and service. "C/(I+N) and Bearer Calculation (UL)" on page 532.1 User Distribution During each simulation.e.4. Mechanical Downtilt (  ): Angle with respect to the horizontal for pointing the subscriber terminal antenna towards its serving base station. Atoll calculates the path loss again for the subscriber locations and heights because the subscriber heights can be different from the default receiver height used for calculating the path loss matrices. Atoll calculates the following parameters for each subscriber in the list that has a serving base station assigned and whose Lock Status is set to None or Server. Atoll performs two random trials. The resulting user distribution complies with the traffic database and maps selected when creating simulations. "Signal Level Calculation (UL)" on page 523. Atoll determines the total number of users attempting connection in each simulation based on the Poisson distribution. This may lead to slight variations in the total numbers of users in different simulations.3. • • Azimuth (  ): Angle with respect to the north for pointing the subscriber terminal antenna towards its serving base station. Atoll calculates the following parameters for each subscriber in the list whose Lock Status is set to None. "Simulations Based on Sector Traffic Maps" on page 478. Cell Capacity. It is possible to display the coverage predictions with colours per cell or: • • Number of interferers Number of interferers per cell 6. number of users of a user profile per km². Coverage consists of several independent layers that can be displayed and hidden on the map. "Channel Throughput. The first random trial generates the number of users and their activity status as explained in the following sections depending on the type of traffic input.3. Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data.3.3.3. • Scheduling and Radio Resource Management as explained under "Simulation Process" on page 479.3 Calculations on Subscriber Lists When calculations are performed on a list of subscribers by running the Automatic Server Allocation.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display parameter is greater than or equal to the defined thresholds values.. 476 .4 Monte Carlo Simulations The simulation process is divided into two steps. Once all the user characteristics have been determined. see: • • • • • "Signal Level Calculation (DL)" on page 501. "C/(I+N) and Bearer Calculation (DL)" on page 518. i. and Per-user Throughput Calculation" on page 547.1 Simulations Based on User Profile Traffic Maps and Subscriber Lists User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of user profiles with an associated mobility type and a given density. inactive on both links. f Act and f Act . or the volume of the data transfer in the uplink and the downlink in each data session. a connected user can be either active on both links. the number of voice calls or data sessions. The average number of calls per hour N Call . Data Service (d) User profile parameters for data type services are: • The user terminal equipment used for the service (from the Terminals table).e. 477 . The number of users of each user profile is calculated from the surface area (SEnv) of each environment class map (or each polygon) and the user profile density (DUP). If the map is composed of points. N Call  D Call Calculation of the service usage duration per hour ( p 0 : probability of an active call): p 0 = ---------------------------3600 Calculation of the number of users trying to access the service v ( n v ): n v = N Users  p 0 The activity status of each user depends on the activity periods during the call. • The average duration of a call (seconds) D Call .e. N Users = S Env  D UP • In case of user profile traffic maps composed of lines. the number of users of each user profile is calculated from the line length (L) and the user profile density (DUP) (users per km): N Users = L  D UP • The number of users is a direct input when a user profile traffic map is composed of points. Voice Service (v) User profile parameters for voice type services are: • • The user terminal equipment used for the service (from the Terminals table).Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 User profile traffic maps: Each polygon or line of the map is assigned a density of users with a given user profile and mobility type. Atoll calculates the probability for a user being active at a given instant in the uplink and in the downlink according to the service usage characteristics described in the user profiles. i.3. User profiles model the behaviour of the different user categories. or active on DL only. Fixed subscribers listed in subscriber lists have a user profile assigned to each of them. Each user profile contains a list of services and parameters describing how these services are accessed by the user. each point is assigned a number of users with given user profile and mobility type. active on UL only.. the average duration of each voice call.. i. the uplink and downlink activity UL DL factors defined for the voice type service v. Calculation of activity probabilities: UL DL Probability of being inactive: p Inactive =  1 – f Act    1 – f Act  UL UL DL Probability of being active in the uplink: p Active = f Act   1 – f Act  DL DL UL Probability of being active in the downlink: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active in the uplink and downlink both: p Active = f Act  f Act Calculation of number of users per activity status: Number of inactive users: n v – Inactive = n v  p Inactive UL UL Number of users active in the uplink: n v – Active = n v  p Active DL DL Number of users active in the downlink: n v – Active = n v  p Active UL + DL UL + DL Number of users active in the uplink and downlink both: n v – Active = n v  p Active Therefore. and TP Average is the average downlink requested throughput of the service s.3. will correspond to calculated distributions. But if you check each simulation. • Sector Traffic Maps (# Active Users) UL Atoll directly uses the defined N and N coverage area using the service s.and N = ---------------------UL DL TP Average TP Average UL Where TP Cell is the total uplink throughput demand defined in the map for any service s for the coverage area of the DL transmitter.2 Simulations Based on Sector Traffic Maps Sector traffic maps are also referred to as live traffic maps. active on DL and active on UL and DL users. for the service d. For each transmitter TXi and each service s.e. the user distribution between services as well as the activity status distribution between users can be different in each of them.3. • The average data volume (in kBytes) transferred in the downlink V • The average throughputs in the downlink Calculation of activity probabilities: f UL DL TP Average DL and the uplink and the uplink V UL TP Average UL during a session. Live traffic data from the O&M is spread over the best server coverage areas of the transmitters included in the traffic map. The service and the activity status of each user are randomly drawn in each simulation. the average number of users per service and average numbers of inactive.Atoll 3. Either throughput demands per service or the number of active users per service are assigned to the coverage areas of each transmitter. active on UL. i.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 • The average number of data sessions per hour N Session .and f = -----------------------------------------UL DL TP Average  3600 TP Average  3600 UL DL Probability of being inactive: p Inactive =  1 – f    1 – f  UL Probability of being active in the uplink: p Active = f DL UL DL  1 – f  Probability of being active in the downlink: p Active = f DL UL  1 – f  UL + DL Probability of being active in the uplink and downlink both: p Active = f UL f DL Calculation of number of users: Number of inactive users: n d – Inactive = N Users  p Inactive UL UL Number of users active in the uplink: n d – Active = N Users  p Active DL DL Number of users active in the downlink: n d – Active = N Users  p Active UL + DL UL + DL Number of users active in the uplink and downlink both: n d – Active = N Users  p Active Calculation of the number of active users trying to access the service d (nd): UL UL + DL DL n d = n d – Active + n d – Active + n d – Active The user distribution per service and the activity status distribution between the users are average distributions. the number of active users on UL and DL in the transmitter .4. if you calculate several simulations at once. • Sector Traffic Maps (Throughputs) Atoll calculates the number of active users of each service s on UL and DL in the coverage area of TXi as follows: N UL UL DL TP Cell TP Cell DL = ---------------------. 478 DL values. UL DL N Session  V  8 N Session  V  8 DL = -----------------------------------------..1. 6. respectively. TP Average is the average uplink requested throughput of the service s. TP Cell is the total downlink throughput demand defined in the map for any service s for the coverage UL DL area of the transmitter. Therefore. active on DL and active on UL and DL users correspond to the calculated distribution. the simulation process. it is necessary to UL + DL DL ).3. if you calculate several simulations at once. the activity status distribution between users can be different in each of them. The activity status of each user depends on the activity periods during the call. Atoll considers both active and inactive users. The steps of this algorithm are listed below. optimising spectral efficiency.e. and satisfying the QoS demands of the users. Therefore. But if you check each simulation. 479 . active on UL. As for the other types of traffic maps. Calculation of activity probabilities: UL DL Probability of being inactive: p Inactive =  1 – f Act    1 – f Act  UL UL DL Probability of being active in the uplink: p Active = f Act   1 – f Act  DL DL UL Probability of being active in the downlink: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active in the uplink and downlink both: p Active = f Act  f Act Calculation of the number of active users trying to access the service: We have: N UL UL + DL UL =  p Active + p Active   n and N DL UL + DL DL =  p Active + p Active   n Where. Sets initial values for the following parameters: • Cell transmission powers and EPREs are calculated from the maximum power and EPRE offset values defined by the user as explained in "Downlink Transmission Power Calculation" on page 485. Therefore. 2.Atoll 3. 6.. i.4. and both ( n Active ). The simulation process can be summed up into the following iterative steps. average numbers of inactive. Atoll calculates the probability for a user being active in the uplink and in the downlink as follows: Users active in the uplink and downlink both are included in the N UL UL accurately determine the number of active users in the uplink ( n Active and N DL values. f Act and f Act . For each simulation. in each simulation. the uplink and downlink activity UL DL factors defined for the service. p Inactive Number of inactive users: n Inactive = --------------------------1 – p Inactive The activity status distribution between users is an average distribution. n is the total number of active users in the transmitter coverage area using the service. Calculation of number of users per activity status: UL UL + DL DL UL + DL  N  p Active N  p Active  UL + DL Number of users active in the uplink and downlink both: n Active = Min  -------------------------------------- -------------------------------------- or UL UL + DL DL + DL  p Active + p Active p Active + p UL Active  UL + DL simply. the activity status of each user is randomly drawn. 1. in the downlink ( n Active ). Generates mobiles according to the input traffic data as explained in "User Distribution" on page 476.3. n = n Active + n Active + n Active Calculation of the number of inactive users attempting to access the service: nv . Each Monte Carlo simulation in the Atoll LTE module is a snap-shot of the network with resource allocation carried out over a duration of 1 second (100 frames).0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 At any given instant. In fact.2 Simulation Process LTE cells include intelligent schedulers and radio resource management features for regulating network traffic loads. n Active = Min  N UL DL  f Act N DL UL  f Act  UL Number of users active in the uplink: n Active = N DL UL Number of users active in the downlink: n Active = N UL DL UL + DL – n Active DL UL + DL – n Active UL + DL And. NRUL – ICIC . 8. TX i  ic  For all the mobiles Mi served by any cell TXi(ic) in the uplink. n PUSCH PUCCH is the uplink thermal noise. Performs radio resource management and scheduling to determine the amount of resources to allocate to each mobile according to the service priorities and throughput demands of each mobile using the selected scheduler as explained in "Scheduling and Radio Resource Management" on page 552. as explained in "Best Server Determination" on page 535. NR UL TX  ic  i TX  ic  i TX  ic  i . Mi Transmitting P Allowed . 4. TL UL TX  ic  i . 3. NRUL is the noise rise. Determines the best servers for all the mobiles generated for the simulation. the simulation process. From fractional power control (see "Signal Level Calculation (UL)" on page 523). 7. we get the CINR PUSCH – Max for each mobile Mi that ensures access to the highest bearer: M TX  ic  M i i i CINR PUSCH – Max = T B +  1 –  FPC   L Total   For each cell TXi(ic). and determines whether they are in the cell centre or cell-edge.3. Mi Combining equations (1) and (2). Calculates the user throughputs after allocating resources to each mobile as explained in "User Throughput Calculation" on page 561. and L Total are the total losses. Determines the downlink and uplink C/(I+N) and bearers for each of these mobiles as explained in "C/(I+N) and Bearer Calculation (DL)" on page 518 and "C/(I+N) and Bearer Calculation (UL)" on page 532 respectively. 5. we know that: Mi P Allowed = CINR PUSCH – Max + NRUL + n PUSCH PUCCH +  FPC  L Total (1) Where CINR PUSCH – Max is the maximum PUSCH C/(I+N). Determines the channel throughputs at the mobile as explained in "Throughput Calculation" on page 538. the highest value is kept: TX  ic  M i i CINR PUSCH – Max = Max  CINR PUSCH – Max   All M i For each iteration k. and AU DL ) are set to their current values in the Cells table.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 M i • Mobile transmission power is set to the maximum mobile power ( P Max ).Atoll 3. r DL – CE . 6. a mobile Mi can access the highest bearer if: Mi Mi P Allowed – NR UL – n PUSCH PUCCH – L Total = T B (2) Mi Where T B is the bearer selection thresholds of the highest bearer defined in the reception equipment used by the cell TXi(ic).  FPC is the fractional power control factor. 480 . Sets the maximum PUSCH C/(I+N) of each cell to a value high enough to ensure that it will not cause any power constraints for cell-edge mobiles. • Cell loads ( TL DL TX  ic  i TX  ic  i . Atoll calculates CINR PUSCH – Max as follows to ensure access to the highest bearer using all the frequency blocks. and noise rise values of all the cells according to the resources in use and the total resources as follows: Calculation of Traffic Loads: Atoll calculates the traffic loads for all the cells TXi(ic).Atoll 3.3.1: LTE Simulation Algorithm 9. Calculation of Downlink Cell-edge Traffic Ratio: Atoll calculates the downlink cell-edge traffic ratio for all the cells as follows:  TX i  ic  CE Mi R DL CE Mi r DL – CE = --------------------TX i  ic  TL DL 481 . the uplink noise rise is calculated and updated by considering each interfering mobile Mj as explained in "Interference Calculation (UL)" on page 525. TL DL  = MU – MIMO – DL Mi RC DL TX i  ic  and TL UL = MU – MIMO – DL M i  MU – MIMO – UL Mi RC UL MU – MIMO – UL M i Calculation of Uplink Noise Rise: For each victim cell TXi(ic). TX  ic  i TL DL = M TX  ic  i  RDL and TLUL i Mi = M  RUL i Mi TX i  ic  For MU-MIMO.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Figure 6. Updates the traffic loads. the cell TXi(ic) requests its neighbouring cells to increase the uplink transmission powers of the mobiles they serve (mobiles interfering TXi(ic)). The default method of uplink noise rise control is the best effort method. This value can be changed through Atoll. 10. the ratio of interference over thermal noise I/N which can be calculated from the noise rise: IoT = I/N = (I+N)/N . the cell TXi(ic) does not request any change. In other words. the cell TXi(ic) requests its neighbouring cells to decrease the uplink transmission powers of the mobiles they serve (mobiles interfering TXi(ic)). you should decrease the uplink noise rise convergence threshold defined for the simulation so that the simulation takes more iterations to converge and allows noise rise control to reach its goal. i.e. CE i Calculation of Downlink AAS Usage: Atoll calculates the downlink AAS usages for all the cells as follows: M  AAS Mi TX i  ic  AAS = ------------------------------TX i  ic  TL DL AU DL  Where i R DL Mi M i R DL AAS is the sum of the percentages of the downlink cell resources allocated to mobiles served by the AAS smart antennas. Performs uplink noise rise control as follows: For each cell TXi(ic).  M NRC .1): TX  ic  TX  ic   NR i   NR i   UL UL – Max  -----------------------  ------------------------------  10 10 – 1  – 10  Log  10 – 1 =  10  Log  10                TX i  ic  NR UL TX  ic  i Here NRUL is the uplink noise rise of the cell TXi(ic) calculated in step 9. Atoll calculates the difference between the current and the maximum noise rise values (in terms of IoT..3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks Where M  M CE i R DL ©Forsk 2015 is the sum of the percentages of the downlink cell resources allocated to mobiles in the cell-edge.ini file by adding the following lines and setting it to a value other than "1" (positive values are considered as negative margins): [LTE] NR_CONTROL_MARGIN_MIN = 1 The uplink transmission powers of the mobiles in neighbouring cells of the cell TXi(ic) are adjusted according to the request in the next iteration by updating the maximum PUSCH C/(I+N) for the neighbouring cells TXj(jc): TX j  jc  CINR PUSCH – Max 482 TX j  jc  k = Min  CINR PUSCH – Max  TX i  ic  k–1 – NRUL TX  jc  j  CINR PUSCH – Max CINR PUSCH – Limit  . If you wish to achieve optimum noise rise control. This means that uplink noise rise control is not part of the simulation convergence criteria. If the resulting noise rise values are higher than the maximum allowed. The resulting noise rise values may be higher than the maximum allowed values defined per cell. The best effort noise rise control works as follows: • TX i  ic  If NR UL  0 . Here M NRC is a noise rise control margin set to -1 dB by default. irrespective of whether or not the noise rise control has been successful. a simulation will converge once the downlink and uplink traffic loads and the uplink noise rise values are stable. • • TX i  ic  If 0  NR UL TX i  ic  If NR UL  M NRC .Atoll 3. this means that the noise rise control requires more iterations for stabilising the overall network’s noise rise than those needed by the simulation to converge. the cell TXi(ic) requests its neighbouring cells to increase the uplink transmission powers of the mobiles they serve (mobiles interfering TXi(ic)). These six neighbouring cells are those whose served mobiles generate the highest interference for the studied cell. Performs the convergence test to see whether the differences between the previous and current values are within the convergence thresholds.ini file: [LTE] ULNRControlMethod = 1 The strict uplink noise rise control method makes the uplink noise rise control a part of the simulation convergence criteria. k–1 TX  jc  j CINR PUSCH – Limit is an upper limit fixed at 50 dB.ini file by adding the following lines: [LTE] ULNRControlPrecision = 5 Setting this option to X means that the precision will be taken as 0. If you wish to include the uplink noise rise control in the simulation convergence criteria.X dB. and the uplink noise rise values of all the cells are less than or equal to the defined maximum uplink noise rise. This value can be changed through Atoll. and CINR PUSCH – Max is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) as calculated in step 4.3. and k–1 TX j  jc  CINR PUSCH – Max is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) as calculated in step 4. Here m NRC is a noise rise control precision level set to 0.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  jc  j Here CINR PUSCH – Max TX  jc  j CINR PUSCH – Max k is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the current iteration k. The strict noise rise control works as follows: • TX i  ic  If NR UL  0 . a simulation will converge once the downlink and uplink traffic loads and the uplink noise rise values are stable. and can be written as follows: TX i  ic  TL DL k TX i  ic  TL UL k TX  ic  i NR UL k = TX  ic  i Max  TL DL  All TX  ic  i = k k – TL UL TX  ic  i Max  TL UL All TX  ic  i = TX i  ic  – TL DL TX i  ic  TX  ic  i Max  NR UL All TX  ic  i  k – 1 TX  ic  i k – NR UL  k – 1  k – 1 483 .5 dB by default. The convergence criteria are evaluated at the end of each iteration k. the cell TXi(ic) requests its neighbouring cells to decrease the uplink transmission powers of the mobiles they serve (mobiles interfering TXi(ic)). is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the previous iteration k-1. you can change the uplink noise rise control method from best effort to strict by setting the following option in the Atoll. At most six neighbouring cells are considered in uplink noise rise control. is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the previous iteration k-1. • TX i  ic  If NR UL  m NRC . In other words. The uplink transmission powers of the mobiles in neighbouring cells of the cell TXi(ic) are adjusted according to the request in the next iteration by updating the maximum PUSCH C/(I+N) for the neighbouring cells TXj(jc): TX j  jc  CINR PUSCH – Max TX  jc  k j = Min  CINR PUSCH – Max TX  jc  j Here CINR PUSCH – Max TX j  jc  CINR PUSCH – Max k TX i  ic  k–1 – NR UL TX  jc  j  CINR PUSCH – Max CINR PUSCH – Limit is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the current iteration k.5 dB). The default value is 5 (= 0. 11. or DL+UL. . Convergence: Simulation has converged between iteration k . DL. Repeats the above steps (from step 3.3. if: TX i  ic  TL DL k TX i  ic  NRUL TX i  ic   TL DL Req OR TX i  ic  TL UL TX i  ic  k  TL UL Req OR TX i  ic  NR UL TX i  ic  k  NR UL Req OR TX i  ic  k  NR UL – Max 12. with the best effort uplink noise rise control. this applies to the mobiles selected for scheduling by their primary cells.e.. Backhaul Saturation: If allocating resources to a mobile makes the effective RLC aggregate site throughputs exceed the maximum S1 interface throughputs defined for the site. the minimum uplink throughput demand is higher than the uplink allocated bandwidth throughput (step 7. for a user active in uplink. i. In addition to the above parameters. or if the mobile’s minimum throughput demand is higher than the UE throughput capacity.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks TX  ic  i If TL DL ©Forsk 2015 TX  ic  i Req . the simulations also list the connection status of each mobile. For LTE-A mobiles. Simulation Results At the end of the simulation process.). Mobiles can be rejected due to: • • • • • No Coverage: If an LTE mobile does not have any best serving cell with cell type "LTE" and if an LTE-A mobile does not have any best serving primary cell with cell type "LTE-A PCell" (step 3.1 and k. UL. if: TX  ic  i TL DL TX  ic  i k  TL DL TX  ic  i Req AND TL UL TX  ic  i k  TL UL TX  ic  i Req AND NR UL TX  ic  i k  NR UL Req Simulation has converged between iteration k . with the best effort uplink noise rise control.) for the iteration k+1 using the new calculated loads as the current loads. TL UL TX  ic  i Req . with the strict uplink noise rise control. and NR UL Req are the simulation convergence thresholds defined when creating the simulation.) Connected mobiles (step 7. this applies to their primary cells. with the strict uplink noise rise control. Atoll stops the simulation in the following cases. Resource Saturation: If all the cell resources are used up before allocation to the mobile or if.) can be: • • • 484 Connected UL: If a mobile active in UL is allocated resources in UL. if: TX i  ic  TL DL k TX  ic  i NRUL TX i  ic   TL DL Req TX i  ic  TL UL AND TX i  ic  k  TL UL TX i  ic  NR UL AND Req TX i  ic  k  NR UL Req AND TX  ic  i k  NR UL – Max No convergence: Simulation has not converged even after the defined maximum number of iterations. This condition is only verified if the simulation was created with the Backhaul capacity check box selected (step 7.1 and k. Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 7. if: TX i  ic  TL DL TX i  ic  k  TL DL TX i  ic  Req OR TL UL TX i  ic  k  TL UL TX i  ic  Req OR NR UL TX i  ic  k  NR UL Req Simulation has not converged even after the defined maximum number of iterations. For LTE-A mobiles. Connected DL: If a mobile active in DL is allocated resources in DL.). Connected DL+UL: If a mobile active in DL+UL is allocated resources in DL+UL.Atoll 3.). the main results obtained are: • • • • • • • • • • • Downlink traffic loads Uplink traffic loads Uplink noise rise Downlink ICIC ratio Uplink ICIC noise rise Downlink AAS usage Number of co-scheduled MU-MIMO users (DL) Number of co-scheduled MU-MIMO users (UL) Maximum PUSCH C/(I+N) Number of connected users in downlink Number of connected users in uplink These results can be used as input for C/(I+N)-based coverage predictions.) No Service: If the mobile is not able to access a bearer in the direction of its activity (step 5. PBCH. PSS. • TX i  ic  TX j  jc  N FB – CE2 and N FB – CE2 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell and PSS ID 2. • D CP • N SD  Slot : Number of symbol durations per slot (7 is D CP • TX  ic  i N SD – PDCCH : Number of PDCCH symbol durations per subframe defined in the TXi(ic) frame configuration or.Atoll 3. • TX i  ic  N TDD – SSF : Number of TDD special subframes (containing DwPTS. GP. • W FB : Width of a frequency block (180 kHz).1 Downlink Transmission Power Calculation LTE eNode-Bs have a maximum transmission power which is shared by downlink channels. • N Slot  SF : Number of slots per subframe (2). calculation of coverage predictions. The energy per resource element (EPRE) of the downlink reference signals is considered to be the reference with respect to which the EPRE of other channels is determined. and Monte Carlo simulations. You can either define the reference signal EPRE for each cell. These channels include the downlink reference signals. It is equal to 10 for FDD frequency bands. • TX i  ic  N SF – DL : Number of downlink subframes in the frame for the cell TXi(ic). otherwise. SSS. calculations on subscriber lists. • N FB TX i  ic  : Cyclic prefix duration defined in the TXi(ic) frame configuration or.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 6. and is determined from the cell’s TDD frame configuration for TDD frequency bands. and is determined from the cell’s TDD frame configuration for TDD frequency bands.4 Calculation Details The following sections describe all the calculation algorithms used in point analysis. • TX i  ic  TX j  jc  N FB – CE0 and N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell and PSS ID 0. : Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell. The EPRE offsets of channels other than the downlink reference signals can be positive values meaning a relative boost with respect to the downlink reference signals EPRE. or negative values meaning a reduction with respect to the downlink reference signals EPRE.4. TX i  ic  TX i  ic  N SF – DL and N TDD – SSF are determined as follows: TX i  ic  TX i  ic  Configuration N SF – DL N TDD – SSF FDD 10 0 DSUUU-DSUUU 2 2 DSUUD-DSUUD 4 2 DSUDD-DSUDD 6 2 DSUUU-DSUUD 3 2 485 . and PDSCH. 6.3. TX i  ic  TX  ic  i TX  jc  j and N FB TX i  ic  is Normal. global network settings. Atoll first determines the EPRE for each channel in the downlink and then the transmission power corresponding to each channel from the EPRE values. and UpPTS) in the frame for the cell TXi(ic). Input • F : Subcarrier width (15 kHz). or let Atoll calculate it from the cell’s maximum power and the EPRE offsets of other channels. • N FB – SS PBCH : Number of frequency blocks that carry the SS and the PBCH (6). 6 if D CP is Extended). It is equal to 0 for FDD frequency bands. otherwise. PDCCH (which is considered to include the PHICH and PCFICH). The transmission powers of various channels are determined from the distribution of the total energy over a frame among the resource elements corresponding to these channels. • TX i  ic  TX j  jc  N FB – CE1 and N FB – CE1 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell and PSS ID 1. global network settings. The number of modulation symbols (resource elements) per scheduler resource block is calculated as follows: N Sym  SRB = N SCa – FB  N SD  Slot  N Slot  SF Where N SCa – FB is the number of subcarriers per frequency block calculated as follows: W FB N SCa – FB = --------F The number of modulation symbols (resource elements) corresponding to the DwPTS per scheduler resource block in the TDD special subframes is calculated as follows: DwPTS DwPTS N Sym  SSF = N SCa – FB  N SD  SSF DwPTS Where N SD  SSF is the number of DwPTS symbol durations (OFDM symbols) per special subframe.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 TX  ic  i TX  ic  i Configuration N SF – DL N TDD – SSF DSUUU-DDDDD 6 1 DSUUD-DDDDD 7 1 DSUDD-DDDDD 8 1 TX i  ic  • N Ant – TX : Number of transmission (downlink) antenna ports defined for the cell TXi(ic). Calculation of Downlink Reference Signal EPRE In LTE. 1 frequency block by 1 subframe (2 slots) is called a scheduler resource block (SRB) in the calculations below. • EPRE SS • EPRE PBCH : Energy per resource element offset for the PBCH with respect to the downlink reference signals EPRE. determined from the TDD special subframe configuration according to the 3GPP specifications as follows: 486 .3. schedulers are able to perform resource allocation every subframe (2 slots). TX i  ic  TX i  ic  TX i  ic  Calculations If you have directly entered the downlink reference signal EPRE for the cell. • EPRE PDCCH : Energy per resource element offset for the PDCCH with respect to the downlink reference signals EPRE. TX i  ic  TX  ic  i TX i  ic  You can either set the P Max TX  ic  i TX i  ic  or EPRE DLRS for a cell. • EPRE PDSCH : Energy per resource element offset for the PDSCH with respect to the downlink reference signals EPRE. : Energy per resource element offset for the SS with respect to the downlink reference signals EPRE. a resource block (RB) is defined as 1 frequency block by 1 slot. However. • P Max : Maximum transmission power of the cell TXi(ic).Atoll 3. you can skip the section "Calculation of Downlink Reference Signal EPRE" on page 486 and go directly to the section "Calculation of Other EPREs and Per-channel Powers" on page 491. • EPRE DLRS : Downlink reference signal EPRE of the cell TXi(ic). Atoll then determines the numbers of modulation symbols corresponding to each control channel as follows: The number of modulation symbols for the downlink reference signals The number of modulation symbols reserved for downlink reference signal transmission in one scheduler resource block depends on the number of transmission antenna ports: TX i  ic  For all subframes except the TDD special subframes: N Res  SRB    8   =  16     24  TX  ic  i if  N Ant – TX = 1   TX  ic  i if  N Ant – TX = 2 TX  ic  i if  N Ant – TX = 4 or 8   For TDD special subframes: Special Subframe Configuration 0 1 2 3 Cyclic Prefix = Normal DwPTS N SD  SSF 3 9 10 11 Cyclic Prefix = Extended TX  ic  i N Ant – TX TX  ic  i N Res  DwPTS 1 2 2 4 4 8 8 8 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 DwPTS N SD  SSF 3 8 9 10 TX  ic  i TX  ic  i N Ant – TX N Res  DwPTS 1 2 2 4 4 8 8 8 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 1 8 2 16 4 24 8 24 487 .3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Special Subframe Configuration Cyclic Prefix = Normal DwPTS GP N SD  SSF DwPTS N SD  SSF 0 3 1 Cyclic Prefix = Extended UpPTS DwPTS GP UpPTS N SD  SSF DwPTS N SD  SSF 10 3 8 9 4 8 3 2 10 3 9 2 3 11 2 10 1 4 12 1 3 7 5 3 9 8 2 6 9 3 9 1 7 10 2 8 11 1 GP N SD  SSF 1 2 GP UpPTS UpPTS N SD  SSF 1 2 The total number of modulation symbols (resource elements) in downlink is calculated as follows: TX i  ic  TX i  ic  N Sym – DL = N FB TX i  ic  TX i  ic   N Sym  SRB  N SF – DL + N FB TX i  ic  DwPTS  N TDD – SSF  N Sym  SSF Out of the total number of modulation symbols.Atoll 3. Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks Special Subframe Configuration ©Forsk 2015 Cyclic Prefix = Normal TX  ic  i DwPTS N SD  SSF 4 12 5 3 6 9 7 10 8 11 Cyclic Prefix = Extended TX  ic  i N Ant – TX N Res  DwPTS 1 8 2 16 4 24 8 24 1 2 2 4 4 8 8 8 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 TX  ic  i DwPTS N SD  SSF 3 8 9 TX  ic  i N Ant – TX N Res  DwPTS 1 2 2 4 4 8 8 8 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 This gives a number of reserved modulation symbols per frame: TX i  ic  TX i  ic  TX i  ic  N Sym – Res = N SF – DL  N FB TX i  ic  TX i  ic  TX i  ic   N Res  SRB + N TDD – SSF  N FB TX i  ic   N Res  DwPTS The number of modulation symbols used for downlink reference signal transmission in one scheduler resource block is: TX i  ic  For all subframes except the TDD special subframes: N DLRS  SRB    8   =  8     6  TX  ic  i if  N Ant – TX = 1 TX  ic  i if  N Ant – TX = 2 TX  ic  i if  N Ant – TX = 4 or 8   For TDD special subframes: Special Subframe Configuration 0 1 488 Cyclic Prefix = Normal DwPTS N SD  SSF 3 9 TX  ic  i Cyclic Prefix = Extended TX  ic  i N Ant – TX N DLRS  DwPTS 1 2 2 2 4 2 8 2 1 6 2 6 4 5 8 5 DwPTS N SD  SSF 3 8 TX  ic  i TX  ic  i N Ant – TX N DLRS  DwPTS 1 2 2 2 4 2 8 2 1 6 2 6 4 5 8 5 .3. some modulation symbols reserved for downlink reference signals are subtracted: 216 for extended cyclic prefix 240 for normal cyclic prefix The number of modulation symbols for the PDCCH 489 . N Sym – PSS = 2  N FB – SS PBCH  N SCa – FB = 144 N Sym – SSS = 2  N FB – SS PBCH  N SCa – FB = 144 And.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Special Subframe Configuration Cyclic Prefix = Normal DwPTS N SD  SSF 2 10 3 11 4 12 5 3 6 9 7 10 8 11 TX  ic  i Cyclic Prefix = Extended TX  ic  i N Ant – TX N DLRS  DwPTS 1 6 2 6 4 5 8 5 1 6 2 6 4 5 8 5 1 8 2 8 4 6 8 6 1 2 2 2 4 2 8 2 1 6 2 6 4 5 8 5 1 6 2 6 4 5 8 5 1 6 2 6 4 5 8 5 DwPTS N SD  SSF 9 10 3 8 9 TX  ic  i TX  ic  i N Ant – TX N DLRS  DwPTS 1 6 2 6 4 5 8 5 1 8 2 8 4 6 8 6 1 2 2 2 4 2 8 2 1 6 2 6 4 5 8 5 1 6 2 6 4 5 8 5 This gives a number of downlink reference signal modulation symbols per frame: TX i  ic  TX i  ic  TX i  ic  N Sym – DLRS = N SF – DL  N FB TX i  ic  TX i  ic  TX i  ic   N DLRS  SRB + N TDD – SSF  N FB TX i  ic   N DLRS  DwPTS The number of modulation symbols for the SS The primary and secondary synchonisation signals are transmitted on 1 symbol duration each in the 1st and the 6th downlink subframes. The physical broadcast channel overlaps with the downlink reference signals. Therefore. over the center 6 frequency blocks. N Sym – SS = N Sym – PSS + N Sym – SSS = 288 The number of modulation symbols for the PBCH The physical broadcast channel is transmitted on four symbol durations in the 1st downlink subframe over the center 6 frequency blocks.3. therefore.Atoll 3. i.. The number of symbol durations for the PDCCH is defined in the global network settings.e. some modulation symbols reserved for downlink reference signals are subtracted: TX  ic  i if  N SD – PDCCH = 0 : TX i  ic  N Sym – PDCCH = 0 TX  ic  TX  ic  i i if  N SD – PDCCH = 1 AND  N Ant – TX  2 : TX  ic  TX i  ic  N Sym – PDCCH = TX  ic  i N i   SD – PDCCH  N SCa – FB – 4  N FB TX i  ic  TX  ic  i  N SF – DL TX i  ic  +  N SD – PDCCH  N SCa – FB – 4  N FB   TX i  ic   N TDD – SSF Otherwise: TX  ic  TX i  ic  N Sym – PDCCH = TX  ic  TX  ic  i i N i    SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB TX  ic  TX  ic  TX i  ic   N SF – DL TX  ic  i i i +  Min  2 N SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB TX i  ic   N TDD – SSF The number of modulation symbols for the PDSCH The total number of modulation symbols in the frame excluding all the control channel modulation symbols gives the number of modulation symbols available for user data. The physical downlink control channel overlaps with the downlink reference signals.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 The physical downlink control channel can be transmitted over up to 4 symbol durations in each subframe. for the PDSCH: TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  N Sym – PDSCH = N Sym – DL – N Sym – Res – N Sym – SS – N Sym – PBCH – N Sym – PDCCH The energy per resource element for 1 modulation symbol (dBm/Sym) of the downlink reference signals is calculated as follows: • If the reference signal EPRE calculation method is set to Calculated (equal distribution of unused EPRE): TX  ic  TX i  ic  EPRE DLRS  P i  Max  ------------------- TX i  ic  TX i  ic  DwPTS   10    N SD  Slot  N Slot  SF  N SF – DL + N TDD – SSF  N SD  SSF – = 10  Log 10        TX  ic  TX  ic  i i  EPRE SS EPRE PBCH  TX  ic  --------------------------------------------------------------------i 10 10 + N Sym – PBCH  10 10  L og  N Sym – DLRS + N Sym – SS  10    + N Sym – PDCCH  10 • 490 TX i  ic  EPRE PDCCH -----------------------------------10 + N Sym – PDSCH  10 TX i  ic  EPRE PDSCH  -----------------------------------. 10     If the reference signal EPRE calculation method is set to Calculated (with boost) or Calculated (without boost): . therefore.3.Atoll 3. The instantaneous SS transmission power is calculated as follows: TX i  ic  P SS TX i  ic  = EPRE SS + 10  Log  N SCa – FB  N FB – SS PBCH  The instantaneous PBCH transmission power is calculated as follows: TX i  ic  TX i  ic  P PBCH = EPRE PBCH + 10  Log  N SCa – FB  N FB – SS PBCH  Where N SCa – FB  N FB – SS PBCH implies that at the instant when the SS and the PBCH are transmitted. they are transmitted using 2 subcarriers in each frequency block.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  TX  ic  i EPRE DLRS  P i  Max  ------------------- TX  ic  TX  ic  10 i i DwPTS = 10  Log  10   N SD  Slot  N Slot  SF  N SF – DL + N TDD – SSF  N SD  SSF  –     TX  ic  TX  ic  i i  EPRE SS EPRE PBCH  TX  ic  --------------------------------------------------------------------i 10 10 10  L og  N Sym – Res + N Sym – SS  10 + N Sym – PBCH  10   + N Sym – PDCCH  10 TX  ic  i EPRE PDCCH -----------------------------------10 + N Sym – PDSCH  10 TX  ic  i EPRE PDSCH  -----------------------------------. The average PDCCH transmission power is calculated as follows: TX i  ic  P PDCCH TX i  ic    N Sym–PDCCH   = EPRE PDCCH + 10  Log  --------------------------------------------------------------------------------------------------------------------------------------- TX  ic  TX  ic  TX  ic  TX  ic    i i i i  N SD – PDCCH  N SF – DL + Min  2 N SD – PDCCH  N TDD – SSF TX i  ic  The average PDSCH transmission power is calculated as follows: 491 .3. they are transmitted using all the subcarriers in the centre 6 consecutive frequency blocks.Atoll 3. 10     Calculation of Other EPREs and Per-channel Powers The energy per resource element for 1 modulation symbol (dBm/Sym) of the SS is calculated as follows: TX i  ic  EPRE SS TX i  ic  TX i  ic  = EPRE DLRS + EPRE SS The energy per resource element for 1 modulation symbol (dBm/Sym) of the PBCH is calculated as follows: TX i  ic  TX i  ic  TX i  ic  EPRE PBCH = EPRE DLRS + EPRE PBCH The energy per resource element for 1 modulation symbol (dBm/Sym) of the PDCCH is calculated as follows: TX i  ic  TX i  ic  TX i  ic  EPRE PDCCH = EPRE DLRS + EPRE PDCCH The energy per resource element for 1 modulation symbol (dBm/Sym) of the PDSCH is calculated as follows: TX i  ic  TX i  ic  TX i  ic  EPRE PDSCH = EPRE DLRS + EPRE PDSCH If the reference signal EPRE calculation method is set to Calculated (with boost). the "boosted" RS energy per resource element is calculated as follows: TX i  ic  EPRE DLRS TX i  ic  = EPRE DLRS  N TXi  ic   Sym – Res  + 10  Log  ------------------------ TXi  ic    N Sym – DLRS The instantaneous downlink reference signal transmission power is calculated as follows: TX i  ic  P DLRS TX i  ic  TX i  ic  = EPRE DLRS + 10  Log  2  N FB  TX  ic  i Where 2  NFB   implies that at the instant when downlink reference signals are transmitted. PDCCH. No ICIC. we have. the instantaneous powers of the PDCCH and the PDSCH also vary over time. and PDSCH EPREs for cells using downlink static ICIC. Soft and partial soft FFR Cell-edge and cell-centre frequency blocks are transmitted at the same time. TX i  ic  TX i  ic  EPRE DLRS CC = EPRE DLRS TX i  ic  TX i  ic  TX i  ic  TX i  ic  N FB  ---------------------------------------------------------------------. This is why average transmission powers are calculated and used in Atoll. Therefore.and P DLRS  ---------------------------- CE = P DLRS  ----------------------------TX i  ic  TX i  ic  EPRE DLRS EPRE DLRS TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  EPRE PDCCH CC EPRE PDCCH CE . time-switched FFR. and hard FFR Cell-edge and cell-centre frequency blocks are not transmitted at the same time. EPRE and Transmission Power adjustment for ICIC The following applies to RS.and EPRE DLRS CE = EPRE DLRS CC   CE TX  ic  TX  ic  TX i  ic   i  N i  FB – CE + N FB – CC  CE TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  N FB . i.e..0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks TX  ic  i P PDSCH ©Forsk 2015         TX  ic  i   TX  ic  N Sym–PDSCH i  = EPRE PDSCH + 10  Log --------------------------------------------------------------------------------------------------------------------  TX  ic  TX  ic   i  N i  N   N – N Slot  SF SD – PDCCH SF – DL    SD  Slot   TX i  ic  TX i  ic     DwPTS     +  N SD  SSF – Min  2 N SD – PDCCH   N TDD – SSF As the number of subcarriers used for the PDCCH and PDSCH transmission varies over time. Therefore. TX i  ic  TX i  ic  EPRE DLRS CC = EPRE DLRS TX i  ic  EPRE PDCCH CC TX i  ic  EPRE PDSCH CC TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  N FB N FB and EPRE DLRS CE = EPRE DLRS  ---------------- ----------------TX i  ic  TX i  ic  N FB – CC N FB – CE TX i  ic  TX i  ic  TX i  ic  TX  ic  i TX  ic  i TX i  ic  TX i  ic  N FB N FB = EPRE PDCCH  ----------------and EPRE PDCCH CE = EPRE PDCCH  ----------------TX i  ic  TX i  ic  N FB – CC N FB – CE TX i  ic  TX i  ic  TX i  ic  N FB N FB and EPRE PDSCH CE = EPRE PDSCH  ----------------= EPRE PDSCH  ----------------TX  ic  TX  ic  i i N FB – CC N FB – CE TX i  ic  TX i  ic  P DLRS CC = P DLRS CE = P DLRS TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  P PDCCH CC = P PDCCH CE = P PDCCH P PDSCH CC = P PDSCH CE = P PDSCH 2.and EPRE PDCCH EPRE PDCCH CC = EPRE PDCCH  -------------------------------------------------------------------- CE = EPRE PDCCH CC   CE TX  ic  TX  ic  TX i  ic   i  N i  FB – CE + N FB – CC  CE TX i  ic  EPRE PDSCH CC TX i  ic  TX i  ic  TX i  ic  TX i  ic  N FB = EPRE PDSCH  ---------------------------------------------------------------------. 1. from one symbol duration to the next.and EPRE PDSCH CE = EPRE PDSCH CC   CE TX i  ic  TX i  ic  TX i  ic    N FB – CE + N FB – CC  CE  TX i  ic  P DLRS CC = P DLRS TX i  ic  P PDCCH CC 492 TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  EPRE EPRE DLRS CC DLRS CE .3. power is divided among cellcentre and cell-edge frequency blocks. therefore.Atoll 3.and P PDCCH = P PDCCH  -------------------------------- CE = P PDCCH  --------------------------------TX i  ic  TX i  ic  EPRE PDCCH EPRE PDCCH . • EPRE PDCCH : Energy per resource element of the PDCCH for cell TXi(ic).3. • EPRE SS • EPRE PBCH : Energy per resource element of the PBCH for cell TXi(ic). TX i  ic  • P PBCH : Instantaneous transmission power of the PBCH for cell TXi(ic). or N FB – CE2 depending on the PSS ID of TXi(ic). it is automatically calculated as follows:  CE TX i  ic  TX i  ic  N FB – CC = ----------------TX i  ic  N FB – CE TX i  ic  N FB – CC and N FB – CE are respectively the numbers of frequency blocks in cell centre and cell-edge of TXi(ic).Atoll 3.and Adjacent Channel Overlaps Calculation An LTE network can consist of cells that use different channel bandwidths.  CE TX i  ic  If  CE TX  ic  i P PDSCH TX  ic  i TX  ic  i TX  ic  TX  ic  EPRE PDSCH CC EPRE PDSCH CE i i . • P PDCCH : Average transmission power of the PDCCH for cell TXi(ic). 493 . Output TX  ic  i • EPRE DLRS : Energy per resource element of the downlink reference signals for cell TXi(ic). N FB – CE1 .4. Therefore. • P DLRS : Instantaneous transmission power of the downlink reference signals for cell TXi(ic). TX i  ic  TX i  ic  TX i  ic  TX  ic  i TX i  ic  : Instantaneous transmission power of the SS for cell TXi(ic).2 Co. Number of frequency blocks in ICIC mode Cell centre Cell edge TX i  ic  No FFR N FB Time-switched FFR N FB TX i  ic  TX i  ic  N FB – CEx TX i  ic  Soft FFR TX i  ic  TX i  ic  N FB – CEx TX i  ic  Hard FFR Partial soft FFR TX i  ic  N FB N FB TX i  ic  N FB TX i  ic  TX  ic  N FB – CEx TX i  ic  TX i  ic  – N FB – CEx TX  ic  N FB – CEx TX  ic  i i i –  N FB – CE0 + N FB – CE1 + N FB – CE2   TX i  ic  TX i  ic  N FB – CEx TX i  ic  Where N FB – CEx can be N FB – CE0 . TX i  ic  TX i  ic  6. • EPRE PDSCH : Energy per resource element of the PDSCH for cell TXi(ic). the start and end frequencies of all the channels may not exactly coincide. Channel bandwidths of cells can overlap each other with different ratios. • P PDSCH : Average transmission power of the PDSCH for cell TXi(ic). • P SS TX i  ic  : Energy per resource element of the SS for cell TXi(ic). By definition:  CE = ----------------EPRE CC TX i  ic  is left empty.and P PDSCH  ------------------------------- CE = P PDSCH  -------------------------------TX  ic  TX  ic  i i EPRE PDSCH EPRE PDSCH EPRE CE is the cell-edge power boost for cell TXi(ic)’s frame configuration.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  i P PDSCH CC = TX  ic  i In the above. "Total Overlap Ratio Calculation" on page 496. TX  ic  i If the studied cell is assigned a channel number N Channel . In order to calculate the co.and adjacent overlaps and the total overlap ratio are calculated as respectively explained in: • • • "Co-Channel Overlap Calculation" on page 495.e. • CN Band and CN Band : Channel number step of the frequency bands assigned to cells TXi(ic) and TXj(jc).1 Conversion From Channel Numbers to Start and End Frequencies Input • TX i  ic  TX j  jc  F Start – Band and F Start – Band : Start frequencies of the frequency bands assigned to the cells TXi(ic) and TXj(jc). TX j  jc  For FDD networks. 6. 494 TX i  ic  TX j  jc  • W Channel and W Channel : Bandwidths of the channels assigned to cells TXi(ic) and TXj(jc)..and adjacent channel overlaps are calculated between the channels used by any studied cell TXi(ic) and any other cell TXj(jc) of the network. F Start – Band can be the start frequency of a TDD frequency band ( F Start – TDD ).0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Figure 6.3..Atoll 3.2: Co-Channel and Adjacent Channel Overlaps The following sections describe how the co. Atoll considers that the same channel number is assigned to a cell in the downlink and uplink. • ICS Band and ICS Band : Inter-channel spacing of the frequency bands assigned to cells TXi(ic) and TXj(jc). TX i  ic  TX i  ic  TX j  jc  TX j  jc  . Once the start and end frequencies are known for the studied and other cells. In terms of interference calculation. and adjacent channel interference on the adjacent channel bandwidths. or the uplink or the downlink start frequency of an FDD frequency band ( F Start – FDD – UL or F Start – FDD – DL ). First – TX i  ic  First – TX j  jc  • N Channel • N Channel and N Channel : Channel numbers assigned to cells TXi(ic) and TXj(jc). the channel number you assign to a cell is considered for uplink and downlink both. corresponding to N Channel – 1 and TX i  ic  N Channel + 1 . "Adjacent Channel Overlap Calculation" on page 496. i. i.4.e.2. it receives co-channel interference on the channel bandwidth of TX i  ic  TX i  ic  N Channel . the co.and adjacent channel overlaps between two channels. the studied cell can be considered a victim of interference received from the other cells that might be interfering the studied cell. TX i  ic  and N Channel : First channel numbers the frequency band assigned to the cells TXi(ic) and TXj(jc). it is necessary to calculate the start and end frequencies of both channels (explained in "Conversion From Channel Numbers to Start and End Frequencies" on page 494). 495 . • F End TX i  ic  TX j  jc  and F End : End frequencies for the cells TXi(ic) and TXj(jc). Co-channel overlap exists if: TX i  ic  TX j  jc  F Start  F End TX i  ic  AND F End TX j  jc   F Start Otherwise there is no co-channel overlap.3. Atoll calculates the bandwidth of the co-channel overlap as follows: TX i  ic  – TX j  jc  W CCO TX  jc  TX  ic  TX  jc  TX  ic  j i j i = Min  FEnd  F End  – Max  F Start  F Start      The co-channel overlap ratio is given by: TX i  ic  – TX j  jc  r CCO TX  ic  – TX  jc  i j W CCO = ---------------------------------TX j  jc  W Channel Output • TX i  ic  – TX j  jc  r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc). • TX  ic  i W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic).2 Co-Channel Overlap Calculation Input • TX i  ic  TX j  jc  F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 494.Atoll 3. Calculations Atoll first verifies that co-channel overlap exists between the cells TXi(ic) and TXj(jc). 6.2. • TX i  ic  TX j  jc  F End and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 494.4.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Calculations Channel numbers are converted into start and end frequencies as follows: For cell TXi(ic): TX  ic  i F Start TX i  ic  F End = TX  ic  i F Start – Band TX i  ic   N TXi  ic  – N First – TX i  ic  TX  ic  TX  ic  i i Channel Channel   - + W Channel + ICS Band   ------------------------------------------------------TX i  ic       CN Band TX i  ic  = F Start + W Channel For cell TXj(jc): TX j  jc  F Start TX j  jc  F End  N TXj  jc  – N First – TX j  jc  TX j  jc  TX j  jc  Channel Channel   - = F Start – Band + W Channel + ICS Band   ------------------------------------------------------TX  jc     j   CN Band TX j  jc  TX j  jc  TX j  jc  = F Start + W Channel Output TX  ic  i TX  jc  j • F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc). Atoll determines the adjacent channel overlap ratio as follows: Bandwidth of the lower-frequency adjacent channel overlap: TX i  ic  – TX j  jc  W ACO L TX  jc  TX  ic  TX  jc  TX  ic  TX  ic  j i j i i = Min  F End  F Start  – Max  F Start  F Start – W Channel The lower-frequency adjacent channel overlap ratio is given by: TX i  ic  – TX j  jc  W ACO L = ---------------------------------TX j  jc  W Channel TX i  ic  – TX j  jc  r ACO L Bandwidth of the higher-frequency adjacent channel overlap: TX i  ic  – TX j  jc  W ACO H TX j  jc  TX i  ic  = Min  F End  F End  TX  ic  TX  jc  TX  ic  i j i + W Channel – Max  F Start  F End     The higher-frequency adjacent channel overlap ratio is given by: TX i  ic  – TX j  jc  W ACO H = ---------------------------------TX  jc  j W Channel TX i  ic  – TX j  jc  r ACO H The adjacent channel overlap ratio is given by: TX i  ic  – TX j  jc  r ACO TX i  ic  – TX j  jc  = r ACO L TX i  ic  – TX j  jc  + r ACO H Output • TX  ic  – TX  jc  i j r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).3 Adjacent Channel Overlap Calculation Input • TX  ic  i TX  jc  j F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 494.4. 496 . • TX i  ic  TX j  jc  F End and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 494.2.4.2. • TX i  ic  W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic). Adjacent channel overlap exists on the lower-frequency adjacent channel if: TX i  ic  TX i  ic  TX j  jc  F Start – W Channel  F End TX i  ic  TX j  jc  AND F Start  F Start Adjacent channel overlap exists on the higher-frequency adjacent channel if: TX i  ic  F End TX j  jc   F End TX i  ic  AND F End TX i  ic  TX j  jc  + W Channel  F Start Otherwise there is no adjacent channel overlap. Calculations Atoll first verifies that adjacent channel overlaps exist between (the lower-frequency and the higher-frequency adjacent channels of) the cells TXi(ic) and TXj(jc).Atoll 3.3.4 Total Overlap Ratio Calculation Input • TX i  ic  – TX j  jc  r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co-Channel Overlap Calculation" on page 495. 6.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 6. from the cell’s TDD frame configuration. ABS patterns are used in conjunction with cell range expansion for eICIC (enhanced inter-cell interference coordination. In order to calculate the collision between normal and almost blank subframes.e. These calculations are respectively explained in: • • • "Subframe Pattern Normalisation" on page 497. 6. FDD example: “01000100000100010000” = “0100010000010001000000000000000000000000” • If the ABS pattern is longer than the standard ABS pattern length. Atoll normalises the different ABS pattern lengths in order to perform logical (bit by bit) AND and OR operations afterwards.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • TX  ic  – TX  jc  i j r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Adjacent Channel Overlap Calculation" on page 496. The standard length of the ABS pattern of a cell is determined from its frequency band’s duplexing method and. ABS patterns of different lengths are normalised to 80 bits by Atoll.and adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).4.. For example.4. 2. FDD example: “01000100000100010000010001000001000100000111110000” = “0100010000010001000001000100000100010000” 497 . Any 0’s and 1’s entered after an asterisk will be ignored. and the used downlink.: • If the ABS pattern contains an asterisk. 6. The ABS pattern is corrected to match the standard ABS pattern lengths determined in step 1.3. The ABS pattern is a bit map. Almost blank subframes do not carry any traffic. i. In an ABS pattern. also known as time-domain ICIC) in an effort to minimise cell-edge interference between macro and small cells in heterogeneous LTE networks (HetNets).Atoll 3. which covers all the standard lengths. ABS patterns are normalised as follows: 1. each 0 signifies a normal subframe and 1 implies an almost blank subframe. a series of 0’s and 1’s where each bit corresponds to one subframe. the cells’ ABS patterns must be normalised. FDD example: “0100010000*” = “0100010000010001000001000100000100010000” • If the ABS pattern is shorter than the standard ABS pattern length. the pattern of 0’s and 1’s leading the asterisk is cyclically repeated until it matches the standard ABS pattern length.3. and special subframe patterns determined from the ABS patterns. The standard lengths of the ABS pattern bit maps as defined by the 3GPP are as follows: • • • • FDD cells: 40 bits TDD cells using the frame configuration 0: 70 bits TDD cells using the frame configuration 1 through 5: 20 bits TDD cells using the frame configuration 6: 60 bits The normalised ABS pattern length used in Atoll is 80 bits. it is filled with 0’s to match the standard ABS pattern length. it is truncated to match the standard ABS pattern length. uplink. "Determination of Effective Subframe Patterns" on page 498.3 Subframe Pattern Collision Calculation Subframe transmission and reception patterns can be defined for each cell using the Almost Blank Subframe (ABS) Pattern field. Calculations The total overlap ratio is: TX  ic  – TX  jc  i j rO = TX  ic  – TX  jc  i j r CCO + TX  ic  – TX  jc  i j r ACO  10 TX i  ic  – f ACS ----------------------10 Output • TX i  ic  – TX j  jc  rO : Total co. • TX  ic  i f ACS : Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic). in case the cell’s frequency band is a TDD band. the ABS pattern "0100001000" means that subframes 1 and 6 are almost blank subframes whereas all the other subframes are normal subframes carrying traffic. Only reference signals are transmitted over an ABS.1 Subframe Pattern Normalisation Prior to the calculation of subframe collision probabilities. "Calculation of Subframe Collision Probabilities" on page 499. and special subframe patterns ( SFP DL .4.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks • ©Forsk 2015 If the ABS pattern is empty.Atoll 3.3. the ABS pattern is concatenated with itself. is resized to 80 bits. More precisely. Examples: • FDD: “0100010000010001000001000100000100010000” = 01000100000100010000010001000001000100000100010000010001000001000100000100010000 • TDD frame configuration 0: “0100010000010001000001000100000100010000010001000001000100000100010000” = 01000100000100010000010001000001000100000100010000010001000001000100000100010000 • TDD frame configurations 1 through 5: “01000100000100010000” = 01000100000100010000010001000001000100000100010000010001000001000100000100010000 • TDD frame configuration 6: “010001000001000100000100010000010001000001000100000100010000” = 01000100000100010000010001000001000100000100010000010001000001000100000100010000 Once the ABS pattern SFP ABS has been normalised. it means that there are no almost blank subframes defined and all the subframes can carry traffic. FDD example: NULL = “0” = “0*” = “0000000000000000000000000000000000000000” 3. uplink. and SFM SSF ) to the normalised used subframe patterns SFP Used determined as explained in "Subframe Pattern Normalisation" on page 497: SFP DL = SFP Used AND SFM DL SFP UL = SFP Used AND SFM UL SFP SSF = SFP Used AND SFM SSF SFM DL . SFM UL . it is inverted to determine the used subframe pattern SFP Used that is used in further calculations: SFP Used = !SFP ABS 6. uplink and special subframe masks listed below: FDD SFM DL SFM UL 11111111111111111111111111111111111111111111111111111111111111111111111111111111 SFM SSF TDD frame configuration 0-DSUUU DSUUU SFM DL 10000100001000010000100001000010000100001000010000100001000010000100001000010000 SFM UL 00111001110011100111001110011100111001110011100111001110011100111001110011100111 SFM SSF 01000010000100001000010000100001000010000100001000010000100001000010000100001000 TDD frame configuration 1-DSUUD DSUUD SFM DL 10001100011000110001100011000110001100011000110001100011000110001100011000110001 SFM UL 00110001100011000110001100011000110001100011000110001100011000110001100011000110 SFM SSF 01000010000100001000010000100001000010000100001000010000100001000010000100001000 TDD frame configuration 2-DSUDD DSUDD 498 SFM DL 10011100111001110011100111001110011100111001110011100111001110011100111001110011 SFM UL 00100001000010000100001000010000100001000010000100001000010000100001000010000100 . the downlink. SFM UL . respectively. and SFP SSF ) are determined as follows by applying masks ( SFM DL .3. The ABS pattern determined in step 2.2 Determination of Effective Subframe Patterns Effective downlink. and SFM SSF are. SFP UL . subscriber.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 SFM SSF 01000010000100001000010000100001000010000100001000010000100001000010000100001000 TDD frame configuration 3-DSUUU DDDDD SFM DL 10000111111000011111100001111110000111111000011111100001111110000111111000011111 SFM UL 00111000000011100000001110000000111000000011100000001110000000111000000011100000 SFM SSF 01000000000100000000010000000001000000000100000000010000000001000000000100000000 TDD frame configuration 4-DSUUD DDDDD SFM DL 10001111111000111111100011111110001111111000111111100011111110001111111000111111 SFM UL 00110000000011000000001100000000110000000011000000001100000000110000000011000000 SFM SSF 01000000000100000000010000000001000000000100000000010000000001000000000100000000 TDD frame configuration 5-DSUDD DDDDD SFM DL 10011111111001111111100111111110011111111001111111100111111110011111111001111111 SFM UL 00100000000010000000001000000000100000000010000000001000000000100000000010000000 SFM SSF 01000000000100000000010000000001000000000100000000010000000001000000000100000000 TDD frame configuration 6-DSUUU DSUUD SFM DL 10000100011000010001100001000110000100011000010001100001000110000100011000010001 SFM UL 00111001100011100110001110011000111001100011100110001110011000111001100011100110 SFM SSF 01000010000100001000010000100001000010000100001000010000100001000010000100001000 6. AND and OR are logical bit-by-bit operators. This is equivalent to setting the following Atoll. or mobile in cell TXi(ic): • Subframe collision between cell centre of TXi(ic) and cell centre of TXj(jc):  TX i  ic  – TX j  jc  p ABS – DL – CC TX  jc  TX  jc   j j  OR SFMSSF    AND  SFM DL  1 = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------TX  ic  TX i  ic   SFM i  OR SFM SSF DL   TX  ic  i    SFMDL TX  ic  i  OR SFM SSF  1  TX i  ic  – TX j  jc  p ABS – UL – CC  j  AND  SFM j OR SFMSSF   UL     1 = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------TX i  ic  TX i  ic   SFM  OR SFM UL SSF   TX i  ic     SFMUL TX i  ic  OR SFM SSF TX  jc  TX  jc   1 499 . X. This means that all subframes are considered non-ABS subframes in the cell centre. • In the following equations.4.Atoll 3. Method 1: ABS Patterns Used Only at Cell Edges By default.3 Calculation of Subframe Collision Probabilities The probabilities of collision of subframes between a studied cell TXi(ic) and any interfering cell TXj(jc) are calculated as follows.ini option: [LTE] UseABSonCellEdgeOnly = 1 Different collision probabilities are calculated depending on the location of the served pixel. the operator  X implies the sum of 1’s in a given 1 • series of bits. ABS patterns are considered only to be used for serving users at cell edges.3.3. In the following equations. irrespective of the cell-edge area and the cell-edge traffic ratio.3. or mobile in cell TXi(ic) is calculated as follows: TX i  ic  – TX j  jc  f ABS – DL TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  j i j j i j = 10  Log  r DL – CE  p ABS – DL – CE +  1 – r DL – CE  p ABS – DL – CC      The uplink interference reduction factor due to subframe collisions for any served pixel.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks • ©Forsk 2015 Subframe collision between cell edge of TXi(ic) and cell centre of TXj(jc):  TX  ic  – TX  jc  i j p ABS – DL – CC TX  jc  TX  jc   j j AND  SFM DL OR SFM SSF       1 = -------------------------------------------------------------------------------------------------------------------------------------------------------------------TX  ic  TX  ic  i  SFP i  OR SFPSSF  DL  TX  ic  i    SFPDL TX  ic  i  OR SFPSSF  1  TX i  ic  – TX j  jc  p ABS – UL – CC TX  jc  TX  jc   j j  OR SFM SSF    AND  SFM UL  1 = -------------------------------------------------------------------------------------------------------------------------------------------------------------------TX  ic  TX i  ic   SFP i  OR SFPSSF UL   TX  ic  i    SFPUL TX  ic  i  OR SFPSSF  1 • Subframe collision between cell centre of TXi(ic) and cell edge of TXj(jc):  TX i  ic  – TX j  jc  p ABS – DL – CE  j j  OR SFP SSF    AND  SFP DL  1 = -------------------------------------------------------------------------------------------------------------------------------------------------------------------TX i  ic  TX i  ic   SFM  OR SFMSSF DL   TX  ic  i    SFMDL TX  ic  i  TX  jc  OR SFM SSF TX  jc   1  TX i  ic  – TX j  jc  p ABS – UL – CE  j  AND  SFP j OR SFP SSF   UL     1 = -------------------------------------------------------------------------------------------------------------------------------------------------------------------TX  ic  TX  ic  i  SFM i  OR SFMSSF  UL  TX i  ic     SFMUL TX i  ic  TX  jc  OR SFM SSF TX  jc   1 • Subframe collision between cell edge of TXi(ic) and cell edge of TXj(jc): TX i  ic  – TX j  jc  p ABS – DL – CE TX i  ic  TX i  ic  TX j  jc  TX j  jc    OR SFPSSF  AND  SFP DL OR SFP SSF     SFP DL      1 = ---------------------------------------------------------------------------------------------------------------------------------------------------------------TX i  ic  TX i  ic   SFP  OR SFP DL SSF     1  TX i  ic  – TX j  jc  p ABS – UL – CE  j  AND  SFP j OR SFP SSF   UL    1 = ---------------------------------------------------------------------------------------------------------------------------------------------------------------TX  ic  TX i  ic   SFP i  OR SFP SSF UL   TX i  ic     SFPUL TX i  ic  TX  jc  OR SFPSSF TX  jc   1 This method enables you to include the cell-edge traffic ratio in the calculation of interference. The downlink interference reduction factor due to subframe collisions for any served pixel.Atoll 3. subscriber.ini file: [LTE] UseABSonCellEdgeOnly = 0 The following collision probabilities are calculated between cells TXi(ic) and TXj(jc): 500 . subscriber. you can do so by adding the following lines in the Atoll. or mobile in cell TXi(ic) is calculated as follows: TX i  ic  – TX j  jc  f ABS – UL TX  ic  – TX  jc  TX  ic  – TX j  jc  i j i = 10  Log  p ABS – UL – CE  or f ABS – UL   TX  ic  – TX  jc  i j = 10  Log  p ABS – UL – CC    Method 2: ABS Patterns Used Throughout the Cell If you wish to apply the ABS patterns throughout the cell. "Interference Calculation (DL)" on page 505. "Interference Calculation (UL)" on page 525.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1  TX  ic  – TX  jc  i j p ABS – DL TX  jc  TX  jc   j j  OR SFPSSF    AND  SFP DL  1 = ---------------------------------------------------------------------------------------------------------------------------------------------------------------TX  ic  TX  ic  i  SFP i  OR SFP SSF DL   TX  ic  i    SFPDL TX  ic  i  OR SFP SSF  1  TX i  ic  – TX j  jc  p ABS – UL  j  AND  SFP j OR SFPSSF   UL    1 = ---------------------------------------------------------------------------------------------------------------------------------------------------------------TX i  ic  TX i  ic   SFP  OR SFP SSF UL   TX i  ic     SFPUL TX i  ic  OR SFP SSF TX  jc  TX  jc   1 The downlink interference reduction factor due to subframe collisions for any pixel. C/N. "C/N Calculation (UL)" on page 529. "Noise Calculation (DL)" on page 505. 6. "C/(I+N) and Bearer Calculation (DL)" on page 518. • TX  ic  i P SS : Transmission power of the SS for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on page 485. • TX i  ic  P PDCCH : Transmission power of the PDCCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on page 485. • TX  ic  i P PBCH : Transmission power of the PBCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on page 485. or mobile is calculated as follows: TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  = 10  Log  p ABS – DL  f ABS – DL   The uplink interference reduction factor due to subframe collisions for any pixel. • TX i  ic  EPRE DLRS : Energy per resource element of the downlink reference signals for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on page 485. • P DLRS : Transmission power of the downlink reference signals for cell TXi(ic) as calculated in "Downlink Transmission TX  ic  i Power Calculation" on page 485. the noise and interference. and C/(I+N) ratios are calculated in Atoll: • • • • • • • • • • • "Signal Level Calculation (DL)" on page 453. "C/(I+N) and Bearer Calculation (UL)" on page 532. subscriber.4 Signal Level and Signal Quality Calculations These calculations include the calculation of the received signal levels.4. subscriber. 501 . "Noise Calculation (UL)" on page 525.4.1 Signal Level Calculation (DL) Input TX i  ic  • P Max : Max power of the cell TXi(ic). or mobile is calculated as follows: TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  = 10  Log  p ABS – UL  f ABS – UL   6.4. "Noise Rise Calculation (UL)" on page 528.Atoll 3. "C/N Calculation (DL)" on page 516.3. and noise and interference. • TX i  ic  P PDSCH : Transmission power of the PDSCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on page 485. The following sections describe how the received signal levels. "Signal Level Calculation (UL)" on page 523. or mobile Mi. subscriber. or mobile Mi.0) from Mi the antenna patterns of the antenna used by Mi. TX i • G Ant : Transmitter antenna gain for the antenna used by the transmitter TXi. and the downlink reference signal level based coverage predictions. L Ant is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi. • L • L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model. TX i Array The smart antenna gain ( G SA    ) and the smart antenna array gain offset ( G SA ) are applied only if the AAS criterion (RS C/N. • G SA • G SA • Div G SA Array : Smart antenna array gain offset defined per clutter class. RS C/(I+N). • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi. G Mi Mi Mi . subscriber.3.V) = (0. For more information on TX i TX i the calculation of G SA    . while the antenna is pointed towards Mi’s best serving cell. and L Body are not used in the calculations performed for the point analysis tool’s profile tab.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 TX  ic  i • EPRE SS : Energy per resource element of the SS for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on page 485. subscriber. see "Beamforming Smart Antenna Models" on page 43. or PDSCH C/(I+N)) is less than the DL AAS threshold ( T AAS ) defined in the properties of the reception equipment used by the pixel. L Ant is determined in the direction (H. TX i In coverage predictions. Combining : Smart power combining gain offset defined per clutter class. shadowing margins are taken into account when the option "Shadowing taken into account" is selected. L Ant .Atoll 3. subscriber. : Smart antenna diversity gain (for cross-polarised smart antennas) defined per clutter class. • TX i  ic  EPRE PDCCH : Energy per resource element of the PDCCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on page 485. • M Shadowing – Model : Shadowing margin based on the model standard deviation. For calculating the interfering signal level from any interferer. L Mi . or mobile Mi. i : Receiver terminal’s antenna gain for the pixel. • G SA    : Smart antenna gain in the direction  of the served pixel. subscriber. • L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected. • TX i  ic  EPRE PDSCH : Energy per resource element of the PDSCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on page 485. 502 . Mi For calculating the useful signal level from the best serving cell. • Mi L Body : Body loss defined for the service used by the pixel. or mobile Mi. • EPRE PBCH : Energy per resource element of the PBCH for cell TXi(ic) as calculated in "Downlink Transmission Power TX  ic  i Calculation" on page 485. or mobile Mi. • L • • M G i M M : Receiver terminal losses for the pixel. i L Ant : Receiver terminal’s antenna attenuation calculated for the pixel. subscriber. TX i • E SA : Number of antenna elements defined for the smart antenna equipment used by the transmitter TXi. delta path loss calculation. or mobile Mi. TX i : Total transmitter losses for the transmitter TXi ( L TX i = L Total – DL ). Calculations The received signal levels (dBm) from any cell TXi(ic) are calculated for a pixel. or mobile Mi as follows: TX i  ic  RSRP: E DLRS TX i  ic  = EIRP2 DLRS – L Path – M Shadowing – Model – L Indoor + G TX i  ic  Without smart antennas: EIRP2 DLRS TX  ic  i With smart antennas: EIRP2 DLRS TX i  ic  TX i Mi = EPRE DLRS + G Ant – L TX  ic  i TX i = EPRE DLRS + G Ant – L TX i –L Mi Mi Mi – L Ant – L Body + f CP TX i TX i Combining + 10  Log  E SA  + G SA 503 . or mobile Mi as follows: TX i  ic  C Max TX i  ic  = EIRP Max – L Path – M Shadowing – Model – L Indoor + G TX i  ic  Without smart antennas: EIRP Max TX i  ic  With smart antennas: EIRP Max TX i  ic  TX i  ic  Mi TX i = P Max + G Ant – L TX i  ic  TX i = P Max + G Ant – L TX i TX i  ic  C DLRS = EIRP1 DLRS – L Path – M Shadowing – Model – L Indoor + G TX i  ic  Without smart antennas: EIRP1 DLRS TX i  ic  With smart antennas: EIRP1 DLRS TX  ic  i C SS TX  ic  i = EIRP1 SS TX i  ic  TX i TX i  ic  TX i  ic  TX i  ic  = P SS TX i  ic  = P SS TX i M –L TX i  ic  TX i  ic  TX  ic  i TX  ic  i i TX i  ic  TX i  ic  C PDCCH = EIRP1 PDCCH – L Path – M Shadowing – Model – L Indoor + G TX i  ic  TX i  ic  i TX i  ic  TX i  ic  –L TX i  ic  TX i  ic  C PDSCH = EIRP1 PDSCH – L Path – M Shadowing – Model – L Indoor + G TX i  ic  TX i  ic  Mi TX i  ic  i Mi Mi Combining Div + G SA Mi – L Ant – L Body + f CP TX i Combining Div + 10  Log  E SA  + G SA + G SA –L Mi Mi Mi – L Ant – L Body + f CP TX i TX i + 10  Log  E SA  + G SA   –L TX i Without smart antennas: EIRP1 PDSCH = P PDSCH + G Ant – L TX i  ic  M i TX i Mi TX i M – L Ant – L Body + f CP TX i TX i TX i i TX i Without smart antennas: EIRP1 PDCCH = P PDCCH + G Ant – L With smart antennas: EIRP1 PDCCH = P PDCCH + G Ant – L M Combining + 10  Log  E SA  + G SA   Mi TX Mi TX i –L TX i TX Mi – L Ant – L Body + f CP TX i Without smart antennas: EIRP1 PBCH = P PBCH + G Ant – L With smart antennas: EIRP1 PBCH = P PBCH + G Ant – L Mi + 10  Log  E SA  + G SA   i TX i C PBCH = EIRP1 PBCH – L Path – M Shadowing – Model – L Indoor + G TX i  ic  Mi TX i TX i Mi TX + G Ant – L + G Ant – L Mi – L Ant – L Body + f CP i Combining Div + 10  Log  E SA  + G SA + G SA TX i = P DLRS + G Ant – L Without smart antennas: EIRP1 SS TX i  ic  TX i  ic  Mi TX i = P DLRS + G Ant – L – L Path – M Shadowing – Model – L Indoor + G With smart antennas: EIRP1 SS –L Mi Mi Combining Div + G SA Mi – L Ant – L Body + f CP TX i TX i Array With smart antennas: EIRP1 PDSCH = P PDSCH + G SA    + G SA Combining + G SA Div + G SA – L TX i The energy per resource element (dBm/Sym) received from any cell TXi(ic) are calculated for a pixel. subscriber. subscriber.Atoll 3. otherwise.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • TX  ic  i D CP : Cyclic prefix duration defined in the TXi(ic) frame configuration or. in the global network settings. L Path is the path loss (dB) calculated as follows: TX i L Path = L Model + L Ant Furthermore. or mobile Mi. subscriber. subscriber. f CP  TX  ic   10  Log  7  7. Therefore. i.. or mobile Mi.3. The total symbol duration of a modulation symbol comprises the useful symbol duration. the total losses between the cell and the pixel. • C DLRS : Received downlink reference signal level from the cell TXi(ic) at the pixel.e.Atoll 3. or mobile Mi.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks TX  ic  i E SS TX  ic  i = EIRP2 SS ©Forsk 2015 – L Path – M Shadowing – Model – L Indoor + G TX  ic  i Without smart antennas: EIRP2 SS TX  ic  i With smart antennas: EIRP2 SS TX i  ic  TX  ic  i = EPRE SS TX  ic  i –L TX M i i + G Ant – L TX i  ic  E PBCH = EIRP2 PBCH – L Path – M Shadowing – Model – L Indoor + G TX i  ic  i + G Ant – L TX = EPRE SS M Mi TX i  ic  TX –L TX  ic  i TX  ic  i i TX i  ic  TX i  ic  E PDCCH = EIRP2 PDCCH – L Path – M Shadowing – Model – L Indoor + G TX  ic  i TX  ic  i Mi TX i  ic  TX i  ic  TX i  ic  TX i  ic  E PDSCH = EIRP2 PDSCH – L Path – M Shadowing – Model – L Indoor + G TX i  ic  TX i  ic  Mi TX i TX i  ic  TX i Mi Mi Mi – L Ant – L Body + f CP TX i TX i + 10  Log  E SA  + G SA   Mi Without smart antennas: EIRP2 PDSCH = EPRE PDSCH + G Ant – L TX i  ic  Div + G SA – L Ant – L Body + f CP Mi TX i –L Combining TX i TX i Mi i TX i Without smart antennas: EIRP2 PDCCH = EPRE PDCCH + G Ant – L With smart antennas: EIRP2 PDCCH = EPRE PDCCH + G Ant – L i i Combining Div + 10  Log  E SA  + G SA + G SA i –L TX TX Mi TX i TX TX i TX M i – L Ant – L Body + f CP + 10  Log  E SA  + G SA   i Without smart antennas: EIRP2 PBCH = EPRE PBCH + G Ant – L With smart antennas: EIRP2 PBCH = EPRE PBCH + G Ant – L M i Mi Combining Div + G SA Mi – L Ant – L Body + f CP TX i Array With smart antennas: EIRP2 PDSCH = EPRE PDSCH + G SA    + G SA Combining + G SA Div + G SA – L TX i In the above. subscriber. added to the useful data bits as padding against multi-path to avoid inter-symbol interference. the ratio of the useful symbol energy to the total symbol energy. Once a modulation symbol is received. only the energy of the useful data bits can be used for extracting the data. and a cyclic prefix. f CP implies that the energy belonging to the cyclic prefix is excluded from the useful signal level. the total energy within a modulation symbol belongs in part to the useful data bits and in part to the cyclic prefix. .5  If D CP  0 If TX  ic  is an interferer i  The cyclic prefix energy and the useful data bits energy are both taken into account when calculating interfering signal levels. • C SS TX i  ic  TX i  ic  : Received SS signal level from the cell TXi(ic) at the pixel. Output 504 TX i  ic  • C Max : Received max signal level from the cell TXi(ic) at the pixel.5  If D CPi = Normal  TX i  ic  =  = Extended  10  Log  6  7. Hence. carrying the actual data bits. The energy belonging to the cyclic prefix is lost once it has served its purpose of combatting inter-symbol interference. or mobile Mi can be calculated as follows: L Total = L Path + L TX i + L Indoor + M Shadowing – Model – G TX i +L M i –G M i M i M i + L Ant + L Body f CP is the cyclic prefix factor. subscriber. • L Total : Total losses between the cell TXi(ic) and the pixel. subscriber. TX i  ic  • E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel. or mobile Mi. subscriber.3 Interference Calculation (DL) The interference received by any pixel. or mobile Mi. Input • • • K: Boltzmann’s constant. subscriber.4.Atoll 3. is calculated as follows: TX i  ic  n 0 – Sym = n 0 + 10  Log  F  The downlink noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel. TX i  ic  TX  ic  i 6. subscriber. i.. subscriber.3. or mobile. and whether the cells support ICIC or not.. Input • TX  jc  j E DLRS : Received downlink reference energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal Level Calculation (DL)" on page 501 at the pixel. i. or mobile Mi. However. 6. or mobile Mi. subscriber. or mobile Mi. or mobile Mi. i. subscriber. T: Temperature in Kelvin. The thermal noise density depends on the temperature.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  i • C PBCH : Received PBCH signal level from the cell TXi(ic) at the pixel. subscriber. 505 . • nf Mi : Noise figure of the terminal used for calculations by the pixel. subscriber. or mobile Mi.e. over one subcarrier. is calculated as follows: TX i  ic  n Sym TX i  ic  = n 0 – Sym + nf Mi Output • TX i  ic  n Sym : Downlink noise for one subcarrier. • E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel.4. • C PDSCH : Received PDSCH signal level from the cell TXi(ic) at the pixel. • TX i  ic  E SS : Received SS energy per resource element from the cell TXi(ic) at the pixel. subscriber.e. Calculations The power spectral density of thermal noise is calculated as follows: n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz The thermal noise for one resource element. or mobile Mi. or mobile Mi. over one subcarrier.4. Atoll calculates the downlink noise which comprises thermal noise and the noise figure of the equipment. subscriber. on the traffic loads of the interfering cells TXj(jc). The downlink noise for one resource element. F : Subcarrier width (15 kHz). • C PDCCH : Received PDCCH signal level from the cell TXi(ic) at the pixel. or mobile Mi covered by the cell TXi(ic). or mobile Mi. subscriber. or mobile Mi. • E DLRS : Received downlink reference signal energy per resource element from the cell TXi(ic) at the pixel. subscriber.. it remains constant for a given temperature.4.2 Noise Calculation (DL) For determining the C/N and C/(I+N). TX  ic  i TX  ic  i TX i  ic  or mobile Mi. • E PDSCH : Received PDSCH eneregy per resource element from the cell TXi(ic) at the pixel. subscriber.e. the value of the thermal noise varies with the used bandwidth. served by a cell TXi(ic) from other cells TXj(jc) can be defined as the signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the channels used by the cells TXi(ic) and TXj(jc). • L Path : Path loss between the cell TXi(ic) and the pixel. subscriber. In Monte Carlo simulations. Calculation of traffic loads is explained in "Simulation Process" on page 479. • W FB : Width of a frequency block in the frequency domain (180 kHz). • M Shadowing – Model : Shadowing margin based on the model standard deviation. N Sym – SS : Number of SS resource elements as calculated in "Downlink Transmission Power Calculation" on page 485. see "Beamforming Smart Antenna Models" on page 43. or mobile Mi covered by the cell TXi(ic).3. • N FB – SS PBCH : Number of frequency blocks that carry the SS and the PBCH (6). For more information. as explained in "Signal Level Calculation (DL)" on page 501. As the received energies per resource element from interferers already include M Shadowing – Model . the ratio M Shadowing – Model – M Shadowing – C  I is applied to the interfering signals (for more information.Atoll 3. • TX j  jc  N Sym – DL : Total number of downlink resource elements as calculated in "Downlink Transmission Power Calculation" on page 485. subscriber.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks • TX  jc  j E SS ©Forsk 2015 : Received SS energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal Level Calculation (DL)" on page 501 at the pixel. see "Shadow Fading Model" on page 90).and Adjacent Channel Overlaps Calculation" on page 493. • TX j  jc  TL DL : Downlink traffic load of the interfering cell TXj(jc). subscriber. • M Shadowing – C  I : Shadowing margin based on the C/I standard deviation. • TX j  jc  N Sym – PDCCH : Number of PDCCH resource elements as calculated in "Downlink Transmission Power Calculation" on page 485. the received energies per resource element from interferers already include M Shadowing – Model . • TX  jc  j E PBCH : Received PBCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal Level Calculation (DL)" on page 501 at the pixel. • TX j G SA    : Smart antenna gain in the direction  calculated from the average array correlation matrix: H G SA    = g n     S   R Avg  S  . • TX j  jc  E PDSCH : Received PDSCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal Level Calculation (DL)" on page 501 at the pixel. or entered manually for each cell. . or mobile Mi covered by the cell TXi(ic). • N Sym – PBCH : Number of PBCH resource elements as calculated in "Downlink Transmission Power Calculation" on TX j  jc  page 485. M Shadowing – C  I is added to the received energies per resource element from interferers in order to achieve the ratio M Shadowing – Model – M Shadowing – C  I : E TX  jc  j = E TX  jc  j + M Shadowing – C  I In coverage predictions. see "Beamforming Smart Antenna Models" on page 43. • TX j  jc  E PDCCH : Received PDCCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal Level Calculation (DL)" on page 501 at the pixel. subscriber. TX j  jc  • N Sym – DLRS : Number of downlink reference signal resource elements as calculated in "Downlink Transmission Power • Calculation" on page 485. 506 TX j  jc  • AU DL : Downlink AAS usage of the interfering cell TXj(jc). • TX j  jc  N Sym – PDSCH : Number of PDSCH resource elements as calculated in "Downlink Transmission Power Calculation" on page 485. or mobile Mi covered by the cell TXi(ic). or mobile Mi covered by the cell TXi(ic). Traffic loads can either be calculated using Monte Carlo simulations. In coverage predictions. shadowing margins are taken into account when the option "Shadowing taken into account" is selected. • TX j G SA    : Smart antenna gain in the direction  . For more information. • TX i  ic  – TX j  jc  rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co. f PDCCH TX i  ic  – TX j  jc  .Atoll 3. • TX i  ic  TX j  jc  N FB – CE2 and N FB – CE2 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell and PSS ID 2. or mobile Mi is calculated as follows: • RS of the interfered cell TXi(ic) collide only with RS of the interfering cell TXj(jc) TX i  ic  This occurs when ID PSS TX j  jc  = ID PSS TX j  jc  TX i  ic  and N Ant – TX  N Ant – TX For the calculation of the probability of collision. and The interfered and interfering cells both have an even number of frequency blocks or both have an odd number of frequency blocks.ini file does not contain the following option: [LTE] SameItf_PDSCH_RS_PDCCH = 1 Synchronised transmission and reception means that the OFDM symbols of the interfered and interfering frames overlap and match each other in time. The interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell TXj(jc) at a pixel.ini file: [LTE] eICIConRS = 1 Method 1: Synchronised Transmission and Reception Atoll calculates the interference between two cells using this method when: • • The frequency channels assigned to the interfered and interfering cells have the same centre frequency. TX i  ic  TX j  jc  • W Channel and W Channel : Bandwidths of the channels assigned to cells TXi(ic) and TXj(jc). • TX i  ic  TX j  jc  F Start and F Start : Start frequencies of the channels assigned to the cells TXi(ic) and TXj(jc) calculated as explained in "Conversion From Channel Numbers to Start and End Frequencies" on page 494.3. here N Ant – TX = Min  4 NAnt – TX  . 507 . TX j  jc  TX i  ic  TX  jc  j Calculations Two interference calculation methods exist in Atoll. subscriber. • TX i  ic  TX j  jc  N FB – CE1 and N FB – CE1 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell and PSS ID 1. • N Ant – TX : Number of transmission (downlink) antenna ports defined for the cell TXj(jc). • ID  • r DL – CE and r DL – CE : Downlink cell-edge traffic ratios of the cells TXi(ic) and TXj(jc). TX i  ic  TX i  ic  TX j  jc  and ID  : Physical cell IDs of the cells TXi(ic) and TXj(jc). • N Ant – TX : Number of transmission (downlink) antenna ports defined for the cell TXi(ic). f ICIC – DL TX i  ic  – TX j  jc  . f TL TX i  ic  – TX j  jc  . • TX j  jc  In the calculations below. and • The Atoll.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • TX  ic  i N FB TX  jc  j and N FB : Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell. • TX j  jc  TX j  jc  Calculations of f MIMO . E DLRS TX i  ic  – TX j  jc  probability f ABS – DL is weighted by the downlink subframe collision when the relevant option is set in the Atoll. and f PDSCH are explained at the end of this section. • TX  ic  i TX  jc  j N FB – CE0 and N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell and PSS ID 0. 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 TX  jc  TX  jc  j  DLRS • j  TX  ic   E DLRS N i --------------------- TX  ic  – TX  jc  10 j Ant – TX  +f i = 10  Log  ------------------ 10 O TX  jc   j  N Ant – TX    RS of the interfered cell TXi(ic) collide with RS. here N Ant – TX = Min  4 N Ant – TX  . PDCCH. With 1 or 2 antenna ports: TX  jc  TX j  jc   DLRS j  TX  jc  E DLRS  N j --------------------- TX i  ic  – TX j  jc  10  Ant – TX = 10  Log  ------------------ 10 TX i  ic   + fO  N Ant – TX    TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  j i j j i j   E PDCCH + f PDCCH E PDSCH + f PDSCH TX j  jc  ---------------------------------------------------------------------------------------------------------------------------------------------  TXi  ic  10 10  N Ant – TX – N Ant – TX 10  + 3  10  ------------------------------------------------------------------------------------------------------------------------------------ + 10  L og  --------------------------------------------TX i  ic  4   N Ant – TX     TX  ic  i With 4 or 8 antenna ports and N SD – PDCCH = 1 : TX  jc  TX j  jc   DLRS j  TX  jc   E DLRS N j --------------------- TX i  ic  – TX j  jc  10  Ant – TX = 10  Log  ------------------ 10 + fO  TXi  ic    N Ant – TX    TX  jc  TX  ic  – TX  jc  j i j j i j   E PDCCH + f PDCCH E PDSCH + f PDSCH TX j  jc  ---------------------------------------------------------------------------------------------------------------------------------------------  TXi  ic  10 10  N Ant – TX – N Ant – TX 10  + 5  10  ------------------------------------------------------------------------------------------------------------------------------------ + 10  L og  --------------------------------------------TX i  ic  6   N Ant – TX     TX i  ic  With 4 or 8 antenna ports and N SD – PDCCH  1 : TX  jc  TX j  jc   DLRS j  TX  jc  E DLRS  N j --------------------- TX i  ic  – TX j  jc  10  Ant – TX = 10  Log  ------------------ 10 TX i  ic   + fO  N Ant – TX    TX  jc  TX  ic  – TX  jc  i j i j j j   E PDCCH + f PDCCH E PDSCH + f PDSCH TX j  jc  ---------------------------------------------------------------------------------------------------------------------------------------------  TXi  ic  10 10  N Ant – TX – N Ant – TX 10  + 2  10  ------------------------------------------------------------------------------------------------------------------------------------ + 10  L og  --------------------------------------------TX i  ic  3   N Ant – TX     • RS of the interfered cell TXi(ic) collide only with PDCCH and PDSCH of the interfering cell TXj(jc) This occurs TX i  ic  ID PSS when TX j  jc   ID PSS With 1 or 2 antenna ports: 508 TX i  ic  ( ID PSS TX j  jc  = ID PSS and TX i  ic   Shift TX j  jc  =  Shift  3 and TX i  ic  TX j  jc  N Ant – TX = N Ant – TX = 1 ) OR .Atoll 3. and PDSCH of the interfering cell TXj(jc) TX i  ic  This occurs when ID PSS TX j  jc  = ID PSS TX j  jc  TX i  ic  and N Ant – TX  N Ant – TX For the calculation of the probability of collision. • PDCCH of the interfered cell TXi(ic) collides with PDCCH and some RS of the interfering cell TXj(jc) TX i  ic  This occurs when ID PSS TX j  jc  = ID PSS TX j  jc  TX i  ic  and N Ant – TX  N Ant – TX 509 . subscriber. 10 = 10  Log  ------------------ ----------------------------------------O TX  jc  TX  ic    j i  N Ant – TX  N Sym – PDCCH   TX  jc  TX  ic  – TX  jc  j i j  TX  ic   E PDCCH + f PDCCH TX j  jc  N i ----------------------------------------------------------------------- – N 10 Sym – PDCCH Sym – DLRS in PDCCH  . here N Ant – TX = Min  4 NAnt – TX  .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  jc  TX  jc  j  DLRS TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  j i j j i j  EPDCCH  +f E +f PDCCH PDSCH PDSCH -----------------------------------------------------------------------  ----------------------------------------------------------------------10 10 TX  ic  – TX  jc   10  + 3  10 i j = 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ + f O 4       TX i  ic  With 4 or 8 antenna ports and N SD – PDCCH = 1 : TX  jc  TX j  jc   DLRS TX  ic  – TX  jc  j i j j i j  EPDCCH  + f PDCCH E PDSCH + f PDSCH -----------------------------------------------------------------------  ----------------------------------------------------------------------10 10 TX i  ic  – TX j  jc   10  + 5  10 = 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ + f O 6       TX i  ic  With 4 or 8 antenna ports and N SD – PDCCH  1 : TX  jc  TX j  jc   DLRS TX  ic  – TX  jc  j i j j i j  EPDCCH  + f PDCCH E PDSCH + f PDSCH -----------------------------------------------------------------------  ----------------------------------------------------------------------10 10 TX i  ic  – TX j  jc   10  + 2  10 = 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ + f O 3       The interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc) at a pixel.Atoll 3. N Sym – DLRS in PDCCH is the number of downlink reference signal resource elements that fall within the PDCCH. and N Sym – PDCCH is the number of PDCCH resource elements per frame.3. or mobile Mi is calculated as follows: TX  jc  j TX j  jc   SS PBCH TX  jc  j E PBCH  ESS  --------------------TX j  jc   ------------------- 10 10 TX  ic  – TX j  jc  TX j  jc   10  N Sym – SS + 10  N Sym – PBCH - + f O i = 10  Log  ------------------------------------------------------------------------------------------------------------+ f MIMO TX j  jc    N Sym – SS + N Sym – PBCH     The interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel. or mobile Mi is calculated as follows: • PDCCH of the interfered cell TXi(ic) collides with PDCCH and all the RS of the interfering cell TXj(jc) This occurs TX  ic  i ID PSS  when TX  ic  i ( ID PSS TX  jc  j = ID PSS and TX  ic  i  Shift TX  jc  j =  Shift  3 and TX  ic  i TX  jc  j N Ant – TX = N Ant – TX = 1 ) OR TX  jc  j ID PSS For the calculation of the probability of collision. subscriber. 10 + 10  L og  ----------------------------------------------------------------------------TX i  ic      N Sym – PDCCH   Here. TX  jc  TX j  jc   PDCCH j   E TX j  jc  DLRS   -------------------TX  ic  – TX j  jc  N 1 10 Sym – DLRS in PDCCH  +f i .  10 = 10  Log ------------------+ fO  ----------------------------------------TX i  ic   TXj  jc    N Ant – TX  N Sym – PDSCH   TX  jc  TX  ic  – TX  jc  j i j  TX  ic   E +f TX j  jc  PDSCH PDSCH N i ----------------------------------------------------------------------- – N 10 Sym – PDSCH Sym – DLRS in PDSCH   10 + 10  L og  ----------------------------------------------------------------------------TX i  ic      N Sym – PDSCH   Here. and N Sym – PDSCH is the number of PDSCH resource elements per frame. here N Ant – TX = Min  4 N Ant – TX  .Atoll 3. TX  jc  TX j  jc   PDCCH j  TX  ic  E DLRS  TX j  jc  TX i  ic  N i -------------------- TX  ic  – TX j  jc  N Sym – DLRS in PDCCH – N Sym – DLRS in PDCCH Ant – TX . and N Sym – PDCCH is the number of PDCCH resource elements per frame. 10 10  + f O i = 10  Log  ------------------ -----------------------------------------------------------------------------------------TX  jc  TX  ic    j i N Sym – PDCCH  N Ant – TX    TX  jc  TX  ic  – TX  jc  j i j  TX  jc   E PDCCH + f PDCCH N j ----------------------------------------------------------------------- 10 Sym – PDCCH   10 + 10  L og  ---------------------------- TXi  ic    N Sym – PDCCH    Here. TX  jc  TX j  jc   PDSCH j   E TX  jc  DLRS - j  -------------------TX i  ic  – TX j  jc  N 1 10  Sym – DLRS in PDSCH  . N Sym – DLRS in PDCCH is the number of downlink reference signal resource elements that fall within the PDCCH. or mobile Mi is calculated as follows: • PDSCH of the interfered cell TXi(ic) collides with PDSCH and all the RS of the interfering cell TXj(jc) This occurs TX i  ic  ID PSS when TX  ic  i ( ID PSS TX  jc  j = ID PSS and TX  ic  i  Shift TX  jc  j =  Shift  3 and TX  ic  i TX  jc  j N Ant – TX = N Ant – TX = 1 ) OR TX j  jc   ID PSS For the calculation of the probability of collision. here N Ant – TX = Min  4 N Ant – TX  .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 For the calculation of the probability of collision.3. N Sym – DLRS in PDSCH is the number of downlink reference signal resource elements that fall within the PDSCH. subscriber. here N Ant – TX = Min  4 N Ant – TX  . • PDCCH of the interfered cell TXi(ic) collides only with PDCCH of the interfering cell TXj(jc) TX i  ic  This occurs when ID PSS TX j  jc  TX j  jc  TX j  jc  = ID PSS TX i  ic  – TX j  jc   PDCCH = E PDCCH + f PDCCH TX j  jc  TX i  ic  and N Ant – TX  N Ant – TX TX i  ic  – TX j  jc  + fO The interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel. 510 . • PDSCH of the interfered cell TXi(ic) collides with PDSCH and some RS of the interfering cell TXj(jc) TX i  ic  This occurs when ID PSS TX j  jc  = ID PSS TX j  jc  TX i  ic  and N Ant – TX  N Ant – TX For the calculation of the probability of collision. or mobile Mi is calculated as follows: TX j  jc  TX j  jc   SS PBCH TX j  jc  TX j  jc  TX j  jc  + f MIMO E PBCH + f MIMO  ESS  ----------------------------------------------TX j  jc   --------------------------------------------- 10 10 TX  ic  – TX j  jc   N Sym – SS + 10  N Sym – PBCH  10 - + f O i = 10  Log  --------------------------------------------------------------------------------------------------------------------------------------------------TX j  jc    N Sym – SS + N Sym – PBCH     The interfering energy per resource element (dBm/Sym) received over the PDSCH and the PDCCH from any cell TXj(jc) at a pixel. subscriber. subscriber. • PDSCH of the interfered cell TXi(ic) collides only with PDSCH of the interfering cell TXj(jc) TX i  ic  TX j  jc  = ID PSS This occurs when ID PSS TX j  jc  TX j  jc  TX i  ic  – TX j  jc   PDSCH = E PDSCH + f PDSCH TX j  jc  TX i  ic  and N Ant – TX  N Ant – TX TX i  ic  – TX j  jc  + fO Method 2: Non-synchronised Transmission and Reception Atoll calculates the interference between two cells using this method when: • • The frequency channels assigned to the interfered and interfering cells do not have the same centre frequency. i.N j N 10 10 Sym – DLRS Sym – PDCCH  . The interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell TXj(jc) at a pixel. N Sym – DLRS in PDSCH is the number of downlink reference signal resource elements that fall within the PDSCH. or mobile Mi is calculated as follows: TX  jc  TX j  jc   DLRS TX  jc  TX  ic  – TX  jc  j i j  E j E PDCCH + f PDCCH TX j  jc  TX  jc  DLRS  -----------------------------------------------------------------------------------------.3. or mobile Mi is calculated as follows: 511 . inter-technology interference received from LTE cells calculated using the inter-technology IRFs. and N Sym – PDSCH is the number of PDSCH resource elements per frame.+ fO i  ---------------------------TX j  jc   N Sym – DL   The interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc) at a pixel.+ 10  ------------------------ ----------------------------= 10  Log 10 TX j  jc  TX j  jc   N Sym – DL N Sym – DL   + 10 TX j  jc  TX i  ic  – TX j  jc  E PDSCH + f PDSCH ----------------------------------------------------------------------10  TX j  jc  TX  ic  – TX j  jc  N Sym – PDSCH  . subscriber. or • The Atoll..e. or The interfered and interfering cells do not both have an even number of frequency blocks or do not both have an odd number of frequency blocks. 10 + 10  L og  ---------------------------TX i  ic    N Sym – PDSCH    Here.ini file contains the following option: [LTE] SameItf_PDSCH_RS_PDCCH = 1 This method is also used for calculating the interference received from LTE cells of an external network in co-planning mode.Atoll 3. 10  ----------------------------------------------------------------------------------------= 10  Log  ------------------O TX  jc  TX  ic   j i  N Ant – TX N Sym – PDSCH    TX  jc  TX  ic  – TX  jc  j i j  TX  jc   E PDSCH + f PDSCH N j ----------------------------------------------------------------------- 10 Sym – PDSCH  .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  jc  TX  jc  j  PDSCH j  TX  ic   E TX  jc  TX  ic  DLRS j i N i --------------------- TX  ic  – TX  jc  N – N 10 j Ant – TX Sym – DLRS in PDSCH Sym – DLRS in PDSCH  +f i . 0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 TX  jc  TX  jc  TX  jc  j  PDSCH + 10 TX  jc  TX  ic  – TX  jc  j i j E PDSCH + f PDSCH ----------------------------------------------------------------------10  TX j  jc  TX  ic  – TX j  jc  N Sym – PDSCH  . subscriber.  TX j  jc  10  + AU DL   10 Calculation of MIMO/Antenna Diversity Interference Factors 512 TX i  ic  – TX j  jc  and f PDSCH ) are calculated as follows: .3. TX j  jc  TX j  jc   RSSI TX i  ic  – TX j  jc   EPDSCH + fPDSCH TX j  jc   ---------------------------------------------------------------------10  N Sym – PDSCH  10  10 = 10  Log  --------------------------------------------------------------------------------------------TX j  jc  TX j  jc   N + N Sym – PDSCH Sym – PDCCH   TX j  jc  TX i  ic  – TX j  jc  E +f PDCCH PDCCH ----------------------------------------------------------------------10 TX j  jc   N Sym – PDCCH 10 . 10 + 10 + --------------------------------------------------------------------------------------------TX j  jc  TX j  jc  N Sym – PDSCH + N Sym – PDCCH TX j  jc  E DLRS --------------------10   TX j  jc   TX i  ic  – TX j  jc     2  Min 2 N Ant – TX  + f O     Calculation of PDCCH and PDSCH Interference Weighting Factors TX i  ic  – TX j  jc  The PDCCH and PDSCH interference weighting factors ( f PDCCH TX i  ic  – TX j  jc  f PDCCH TX j  jc  TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc    f MIMO + f TL + f ICIC – DL + f ABS – DL  ------------------------------------------------------------------------------------------------------------------------------------------------- TX j  jc  10   1 – AU    10 DL    =   TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc    j i j i j f TL + f ICIC – DL + f ABS – DL   ----------------------------------------------------------------------------------------------------------------------TX j  jc    10  10  + AU DL  TX  jc  TX i  ic  – TX j  jc  f PDSCH TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  j j i j i j   f MIMO + f TL + f ICIC – DL + f ABS – DL -------------------------------------------------------------------------------------------------------------------------------------------------  TX j  jc  10    1 – AU DL    10   =   TX j  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc    TX j  G    – G    + f + f  SA  SA ICIC – DL ABS – DL     --------------------------------------------------------------------------------------------------------------------------------------------------------.Atoll 3. the encircled 10 in the formulas below is replaced with 8.+ fO i  ---------------------------TX j  jc   N Sym – DL   TX  jc  TX j  jc  TX  ic  – TX  jc  j i j  E j E +f TX  jc  TX  jc  DLRS PDCCH PDCCH j j  -----------------------------------------------------------------------------------------N N Sym – PDCCH 10 10 Sym – DLRS  . or mobile Mi over 1 frequency block during an OFDM symbol carrying reference signals. Therefore. is given as follows: For cells using more than 1 antenna port.+ 10 ----------------------------= 10  Log  10  ------------------------ TX  jc  TX  jc  j j  N Sym – DL N Sym – DL   PDCCH TX  jc  TX  ic  – TX  jc  j i j  E j E PDCCH + f PDCCH TX j  jc  TX j  jc  DLRS  -----------------------------------------------------------------------------------------N Sym – DLRS N Sym – PDCCH 10 10  ---------------------------------------------------- TX  jc  + 10  = 10  Log 10 TX j  jc   j N Sym – DL N Sym – DL   + 10 TX j  jc  TX i  ic  – TX j  jc  E PDSCH + f PDSCH ----------------------------------------------------------------------10  TX j  jc  TX  ic  – TX j  jc  N Sym – PDSCH  . the interfering energy per frequency block (dBm/RB) received from any cell TXj(jc) at a pixel.+ fO i  ---------------------------TX j  jc   N Sym – DL   E-UTRA carrier RSSI is measured on the OFDM symbols that contain reference signals. TXi(ic) and TXj(jc) respectively. Therefore. are calculated as follows: SP TX  ic  i TX i  ic  TX j  jc  TX  jc  r DL – CE r DL – CE j = ---------------------------------------------------------------------------. and r DL – CE is the downlink cell-edge traffic ratios of the cells.and SP = ---------------------------------------------------------------------------TX i  ic  TX j  jc  TX i  ic  TX i  ic  TX j  jc  TX j  jc  N N FB – CE FB – CE r DL – CE +  1 – r DL – CE  ----------------r DL – CE +  1 – r DL – CE  ----------------TX i  ic  TX j  jc      N FB N FB Where. subscriber. Calculation of Interference Reduction Factors TX  jc  j Calculations for the interference reduction factors due to traffic load f TL TX i  ic  – TX j  jc  downlink ICIC using fractional frequency reuse ( f ICIC – DL TX  ic  – TX  jc  i j . SP is the switching point between the ICIC and the non-ICIC parts of the frame. or mobile Mi is in cell centre or cell-edge is determined as explained in "Best Server Determination" on page 535.3. and static ) are explained below: Interference reduction due to the traffic loads of the interfering cells: Interference reduction due to the traffic loads of the interfering cells TXj(jc) is calculated as follows: TX j  jc  f TL TX j  jc  = 10  Log  TL DL    Interference reduction due to the co. The switching points between the ICIC and non-ICIC parts of the frame of the victim and interfering cells. subscriber. an interfered user may receive interference from the cell-edge and cell-centre parts of the frame depending on time-domain switching points between the cell-edge and cell-centre parts of the frames.ini file: [LTE] MultiAntennaInterference = 0 MultiAntennaInterference is set to 1 by default. channel overlapping ( f O ). TX i  ic  – TX j  jc  Depending on the ICIC mode defined for the frame configuration of the cells TXi(ic) and TXj(jc).and adjacent channel overlap between the cells TXi(ic) and TXj(jc) is calculated as follows: TX i  ic  – TX j  jc  fO TX i  ic  – TX j  jc  = 10  Log  r O   Interference reduction due to static downlink ICIC using fractional frequency reuse: The interference reduction factor due to static downlink ICIC using fractional frequency reuse for any pixel. or mobile Mi is calculated as follows: TX i  ic  – TX j  jc  f ICIC – DL TX  ic  – TX j  jc  i = 10  Log  p Collision   Whether a pixel.Atoll 3. add the following lines in the Atoll.and adjacent channel overlap between the studied and the interfering cells: Interference reduction due to the co. 513 . Atoll determines the switching point between the ICIC and the non-ICIC parts of the frame using the ICIC ratios.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  jc  TX  jc  TX  jc  j j j f MIMO is the interference increment due to more than one transmission antenna port: f MIMO = 10  Log  N Ant – TX TX  jc  j If you do not wish to apply f MIMO . f ICIC – DL calculated as follows: • is If both TXi(ic) and TXj(jc) use time-switched FFR The cell-edge and cell-centre resources are time-divided. N FB – CE TX i  ic  is the number of cell-edge frequency blocks common in TXi(ic) and TXj(jc). and N FB – CE is the number of cell-edge frequency blocks in the cell TXi(ic). iii. i. subscriber. the total collision probability for the pixel. For a pixel. subscriber. there can be four different interference scenarios. Between the ICIC part of the victim and the ICIC part of the interferer. are calculated for cell-centre and cell-edge cases as follows: . Atoll calculates the probabilities of collision for each scenario and weights the total interference according to the total collision probability. iv. the total collision probability for the pixel. subscriber. subscriber. Therefore. Between the non-ICIC part of the victim and the ICIC part of the interferer. Between the ICIC part of the victim and the non-ICIC part of the interferer.= ---------------------------------------------- 1 – SP   W Channel N FB – CE ----------------SP  WChannel  N FB With cells using static downlink ICIC. ii. the switching point formula is derived from the equation: r DL – CE  TL DL  1 – r DL – CE   TL DL --------------------------------------------------------. p Collision TX i  ic  – TX j  jc  Cell centre: p Collision 514 Common N FB – CC = -------------------TX i  ic  N FB – CC .If SP j  SP i  TX  ic   1 – SP i      Other combinations of ICIC modes TX i  ic  – TX j  jc  Separate probabilities of collisions.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 The ICIC ratio is used to partition the total downlink traffic load into ICIC and non-ICIC parts of the frame. Therefore. or mobile Mi in the cell centre of TXi(ic). The probability of collision p Coll for each scenario is: Case Interfered cell TX i  ic  Interfering cell TX j  jc  p Coll i ICIC ICIC N FB – CE --------------------TX i  ic  N FB – CE ii ICIC Non ICIC 1 Common Common iii Non ICIC ICIC N FB – CE --------------------TX i  ic  N FB iv Non ICIC Non ICIC 1 Common Where. Between the non-ICIC part of the victim and the non-ICIC part of the interferer. or mobile Mi is calculated as follows: TX i  ic  – TX j  jc  p Collision • TX j  jc  TX i  ic   iv  p Coll If SP  SP  TX j  jc  TX  jc  TX  ic     + p iii   SP j – SP i  =  p iv TX  jc  TX  ic  Coll   1 – SP Coll    ---------------------------------------------------------------------------------------------------------------------------.3.Atoll 3. or mobile Mi is calculated as follows: TX i  ic  – TX j  jc  p Collision  TX j  jc  TX i  ic  i  p Coll If SP  SP   TX  jc  TX  ic  TX  jc  = i ii j i j  + p Coll   SP – SP TX j  jc  TX i  ic   p Coll  SP  ------------------------------------------------------------------------------------------------------------ SP If SP  TX i  ic   SP  For a pixel. or mobile Mi in the cell-edge of TXi(ic). or mobile Mi covered by a cell TXi(ic). • TX  jc  j  SS PBCH : Interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc) at a pixel. or N FB – CE2 depending on the PSS ID of TXi(ic). Calculation of the Downlink Inter-technology Interference The downlink inter-technology interference is calculated as follows: Inter – Tech I DL TX TX k   P DL – Rec  --------------------------------------- = F  TX  ic  TX   i k  TX k  ICP DL  k Here P DL – Rec is the received downlink power from an interfering cell TXk belonging to another technology. and F  TX i  ic  TX k  ICPDL is the inter-technology downlink channel protection ratio for a frequency offset F between the interfered and interfering frequency channels of TXi(ic) and TXk. CDMA2000. • TX j  jc   PDCCH : Interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel. N FB – CC Common N FB – CE = -------------------TX  ic  i N FB – CE Common is the number of common frequency blocks in TXi(ic) and TXj(jc) in cell centre. N FB – CE TX i  ic  is the TX i  ic  number of common frequency blocks in TXi(ic) and TXj(jc) on cell-edge. and TD-SCDMA cells. 515 . subscriber.Atoll 3. total power from UMTS.3. Output • TX j  jc   DLRS : Interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell TXj(jc) at a pixel. TX k P DL – Rec is calculated based on the EIRP from GSM cells.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  – TX  jc  i j Cell-edge: p Collision Common Where. maximum power from LTE cells. N FB – CC and N FB – CE are respectively the numbers of frequency blocks in cell centre and cell-edge of TXi(ic). Number of frequency blocks in ICIC mode Cell centre Cell edge TX i  ic  No FFR N FB Time-switched FFR N FB TX i  ic  N FB TX i  ic  TX i  ic  N FB – CEx TX i  ic  Hard FFR TX i  ic  N FB – CEx TX i  ic  Soft FFR N FB TX i  ic  Partial soft FFR N FB TX  ic  N FB – CEx TX i  ic  TX i  ic  – N FB – CEx TX  ic  N FB – CEx TX  ic  i i i –  N FB – CE0 + N FB – CE1 + N FB – CE2   TX i  ic  TX i  ic  TX i  ic  TX i  ic  N FB – CEx TX i  ic  Where N FB – CEx can be N FB – CE0 . and downlink cell power from Wi-Fi cells. • TX j  jc   PDSCH : Interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel. or mobile Mi covered by a cell TXi(ic). N FB – CE1 . Interference reduction due to subframe collision probabilities: TX i  ic  – TX j  jc  The interference reduction factor due to downlink subframe collision probabilities f ABS – DL is calculated as explained in "Subframe Pattern Collision Calculation" on page 497. subscriber. subscriber. preamble power from WiMAX cells. or mobile Mi covered by a cell TXi(ic). or mobile Mi covered by a cell TXi(ic). subscriber. subscriber. or mobile Mi. • T B : Bearer selection thresholds of the bearers defined in the reception equipment used by Mi’s terminal. or Mi mobile Mi. • i BLER  BDL : Downlink block error rate read from the graphs available in the reception equipment assigned to the M terminal used by the pixel. subscriber. Mi Mi or mobile Mi.Atoll 3. subscriber. or mobile Mi is located. TX i  ic  CINR DLRS : Downlink reference signal C/(I+N) from cell TXi(ic) at pixel. subscriber. • • TX  ic  i n Sym : Downlink noise for one subcarrier for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 505.4.4. or mobile Mi: TX i  ic  CNR DLRS 516 TX i  ic  TX i  ic  = E DLRS – n Sym . • B DL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. subscriber. or mobile Mi. Calculations The C/N for cell TXi(ic) are calculated as follows for any pixel. subscriber. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. • N Ant – RX : Number of reception (downlink) antenna ports defined for the terminal used by the pixel. or mobile Mi. subscriber. subscriber. or mobile Mi over 1 frequency block during an OFDM symbol carrying reference signals. 6. subscriber. • TX i  ic  E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. subscriber. • TX i  ic  E SS : Received SS energy per resource element from the cell TXi(ic) at the pixel.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks • ©Forsk 2015 TX  jc  j  RSSI : Interfering energy per frequency block (dBm/RB) received from any cell TXj(jc) at a pixel. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. • Mi G Div – PDCCH : PDCCH diversity gain defined in the reception equipment of the terminal used by the pixel. subscriber. • Inter – Tech I DL : Downlink inter-technology interference. subscriber. • Mobility  M i  : Mobility used for the calculations. • TX i  ic  E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel.4 C/N Calculation (DL) Input • TX i  ic  E DLRS : Received downlink reference signal energy per resource element from the cell TXi(ic) at the pixel. • TX i  ic  E PDSCH : Received PDSCH energy per resource element from the cell TXi(ic) at the pixel. subscriber. subscriber. Mi • T SU – MIMO – DL : SU-MIMO threshold defined in the reception equipment used by Mi’s terminal. TX i  ic  • N Ant – TX : Number of transmission (downlink) antenna ports defined for the cell TXi(ic). or mobile Mi as calculated in "C/ (I+N) and Bearer Calculation (DL)" on page 518.3. • DL G Div : Additional downlink diversity gain defined for the clutter class where the pixel. or mobile Mi. • Mi B DL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. • Mi G Div – PBCH : PBCH diversity gain defined in the reception equipment of the terminal used by the pixel. subscriber.   DL The additional downlink diversity gain defined for the clutter class of the pixel. if the corresponding option has been set in the Atoll.Atoll 3. corresponding to the bearer is applied to its selection threshold. and according to the Mi diversity mode depending on the SU-MIMO and MU-MIMO Mi thresholds and criteria. the PDSCH C/N calculated above becomes: TX i  ic  TX i  ic  Mi DL CNR PDSCH = CNR PDSCH + G Div – DL + G Div 517 . and Per-user Throughput Calculation" on page 547.3. and Per-user Throughput Calculation" on page 547. the bearers available for selection are all the bearers defined in the reception equipment for which the following is true: M i M i TX  ic  i DL T B – G Div – DL – G Div  CNR PDSCH The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). transmit diversity. Allocated Bandwidth Throughput. subscriber. subscriber. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. G Div – DL . Therefore.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  i CNR SS TX  ic  i = E SS TX  ic  i – n Sym TX  ic  i TX  ic  i TX  ic  i TX  ic  i TX  ic  i TX  ic  i TX i  ic  TX i  ic  TX i  ic  CNR PBCH = E PBCH – n Sym CNR PDCCH = E PDCCH – n Sym CNR PDSCH = E PDSCH – n Sym Bearer Determination: The bearers available for selection in the pixel. SU-MIMO diversity. MIMO Diversity Gain: With MIMO. or mobile Mi’s reception equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). Allocated Bandwidth Throughput. • Whose selection thresholds are less than the PDSCH C/N at Mi: T B  CNR PDSCH Mi TX i  ic  If the cell supports MIMO. or MU-MIMO diversity gain. • Peak RLC Throughput From among the bearers available for selection. the PBCH and PDCCH C/N become: TX i  ic  TX i  ic  Mi TX i  ic  TX i  ic  Mi DL CNR PBCH = CNR PBCH + G Div – PBCH + G Div DL CNR PDCCH = CNR PDCCH + G Div – PDCCH + G Div The PBCH and PDCCH diversity gains are applied to the C/N when the cell and the terminal both support any form of MIMO in downlink. • Bearer Index From among the bearers available for selection. For more information. the selected bearer is the one with the highest downlink effective RLC channel throughput as calculated in "Channel Throughput. Once the bearer is known. see the Administrator Manual. Cell Capacity. or mobile Mi for N Ant – TX . subscriber. Mobility  M i  . The additional downlink diversity gain defined per clutter is also applied. The gain is read from the properties of the TX i  ic  Mi reception equipment assigned to the pixel. or mobile Mi G Div is also applied. M i BLER  B DL . the selected bearer is the one with the highest index.ini file. Cell Capacity. N Ant – RX . • Effective RLC Throughput From among the bearers available for selection. the selected bearer is the one with the highest downlink peak RLC channel throughput as calculated in "Channel Throughput. Atoll takes the ratio of the signal level and the sum of the total interference from other cells and the downlink noise (as calculated in "Noise Calculation (DL)" on page 505). subscriber. • CNR PDCCH : PDCCH C/N from cell TXi(ic) at pixel. and is determined from the cell’s TDD frame configuration for TDD frequency bands.3. TX i  ic  TX i  ic  N SF – DL and N TDD – SSF are determined as follows: 518 TX i  ic  TX i  ic  Configuration N SF – DL N TDD – SSF FDD 10 0 DSUUU-DSUUU 2 2 DSUUD-DSUUD 4 2 DSUDD-DSUDD 6 2 DSUUU-DSUUD 3 2 DSUUU-DDDDD 6 1 . or mobile from all the interfering cells (as explained in "Interference Calculation (DL)" on page 505). Finally. subscriber. • TX i  ic  N TDD – SSF : Number of TDD special subframes (containing DwPTS.and adjacent channel overlap between the studied and the interfering cells. It is equal to 0 for FDD frequency bands. except when the "Lock Status" is set to "Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited.Atoll 3. Next. the receiver is oriented towards the best server just as in the case of LOS. otherwise. In the case of NLOS between the receiver and the best server. or mobile Mi. subscriber.4. GP. • N Slot  SF : Number of slots per subframe (2). or mobile under study.5 C/(I+N) and Bearer Calculation (DL) The carrier signal to interference and noise ratio is calculated in three steps. or mobile Mi. • W FB : Width of a frequency block (180 kHz).4. TX i  ic  TX i  ic  6. and UpPTS) in the frame for the cell TXi(ic). • CNR PDSCH : PDSCH C/N from cell TXi(ic) at pixel. First Atoll calculates the received signal level from the studied cell (as explained in "Signal Level Calculation (DL)" on page 501) at the pixel. : Number of frequency blocks. or mobile Mi. Input • F : Subcarrier width (15 kHz). • TX i  ic  N SF – DL : Number of downlink subframes in the frame for the cell TXi(ic). Interference from each cell is weighted according to the co. It is equal to 10 for FDD frequency bands. subscriber.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks M ©Forsk 2015 i Where G Div – DL is the transmit diversity. in the global network settings. or mobile Mi. SU-MIMO diversity. Atoll does not try to find the direction of the strongest signal. defined in the frequency bands table. The receiver terminal is always considered to be oriented towards its best server. Atoll calculates the interference received at the same studied pixel. for the channel bandwidth used by the cell TXi(ic). subscriber. or MU-MIMO diversity gain applied if the cell supports MIMO and depending on the Mi diversity mode. • CNR SS TX  ic  i : SS C/N from cell TXi(ic) at pixel. and is determined from the cell’s TDD frame configuration for TDD frequency bands. subscriber. 6 if D CP is Extended). the traffic loads of the interfering cells. TX i  ic  TX i  ic  TX i  ic  is Normal. the SU-MIMO and MU-MIMO thresholds and criteria. or mobile Mi. Output TX i  ic  • CNR DLRS : Downlink reference signal C/N from cell TXi(ic) at pixel. subscriber. • N FB – SS PBCH : Number of frequency blocks that carry the SS and the PBCH (6). • D CP • N SD  Slot : Number of symbol durations per slot (7 is D CP • N FB TX i  ic  : Cyclic prefix duration defined in TXi(ic) frame configuration or. and the probability of collision in case ICIC is used by the cells. TX i  ic  • CNR PBCH : PBCH C/N from cell TXi(ic) at pixel. 3. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. • TX j  jc   PDSCH : Interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel. subscriber. subscriber. Mi M i or mobile Mi. 519 . • TX i  ic  E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel. subscriber. • T B : Bearer selection thresholds of the bearers defined in the reception equipment used by Mi’s terminal. TX  ic  i • n Sym : Downlink noise for one subcarrier for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 505. Inter – Tech • NR DL • CNR DLRS : Downlink reference signal C/N from cell TXi(ic) at pixel. • TX  ic  i N Ant – TX : Number of transmission (downlink) antenna ports defined for the cell TXi(ic). subscriber. or mobile Mi as TX i  ic  calculated in "Signal Level Calculation (DL)" on page 501.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • TX  ic  i TX  ic  i Configuration N SF – DL N TDD – SSF DSUUD-DDDDD 7 1 DSUDD-DDDDD 8 1 TX i  ic  E DLRS : Received downlink reference signal energy per resource element from the cell TXi(ic) at the pixel. or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 505. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. subscriber. or mobile Mi. subscriber. • TX j  jc   SS PBCH : Interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc) at a pixel. or mobile Mi as calculated in "C/N : Inter-technology downlink noise rise.Atoll 3. or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 505. subscriber. •  DLRS : Interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell TX j  jc  TXj(jc) at a pixel. subscriber. or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 505. subscriber. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. • Mi B DL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. • TX  ic  i N Sym – PDCCH : Number of PDCCH resource elements as calculated in "Downlink Transmission Power Calculation" on page 485. • TX  ic  i N Sym – PDSCH : Number of PDSCH resource elements as calculated in "Downlink Transmission Power Calculation" on page 485. TX i  ic  • E SS : Received SS energy per resource element from the cell TXi(ic) at the pixel. subscriber. or mobile Mi as calculated in "Interference Calculation (DL)" on page 505. subscriber. subscriber. • TX j  jc   PDCCH : Interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel. TX i  ic  Calculation (DL)" on page 516. • TX i  ic  E PDSCH : Received PDSCH energy per resource element from the cell TXi(ic) at the pixel. subscriber. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. • B DL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 505. Mi • T SU – MIMO – DL : SU-MIMO threshold defined in the reception equipment used by Mi’s terminal. • TX j  jc   RSSI : Interfering energy per frequency block (dBm/RB) received over 1 frequency block during an OFDM symbol carrying reference signals from any cell TXj(jc) at a pixel. • E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel. or mobile Mi: TX  jc  TX i  ic  CINR PDCCH TX  ic  i      j   n PDCCH  Sym    ------------------- --------------------- TX i  ic  Inter – Tech Inter – Tech 10  10     10 +I +10 + NR DL = E PDCCH – 10  Log     DL     All TXj  jc             The PDSCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel. subscriber.Atoll 3. subscriber. • Mobility  M i  : Mobility used for the calculations. or mobile Mi: TX  jc  TX i  ic  CINR PDSCH TX  ic  i      j   n PDSCH  Sym   ------------------- --------------------- TX i  ic   Inter – Tech Inter – Tech 10 10  10  +I  + NR  +10 = E PDSCH –  10  Log  DL DL         All TXj  jc             The downlink reference signal received quality (RSRQ) for cell TXi(ic) is calculated as follows for any pixel. or mobile Mi. DL G Div : Additional downlink diversity gain defined for the clutter class where the pixel. Inter – Tech • I DL : Downlink inter-technology interference as calculated in "Interference Calculation (DL)" on page 505. or mobile Mi: TX  jc  TX i  ic  CINR DLRS TX  ic  i      j  n Sym  DLRS -    --------------------------------------- TX i  ic    10 10  + I Inter – Tech + 10 10  + NR Inter – Tech = E DLRS –  10  Log  DL     DL   All TXj  jc             The SS C/(I+N) for cell TXi(ic) is calculated as follows for any pixel.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks M • ©Forsk 2015 i N Ant – RX : Number of reception (downlink) antenna ports defined for the terminal used by the pixel. or mobile Mi: TX  jc  TX i  ic  CINR PBCH TX  ic  i      j  n Sym  SS PBCH    --------------------------------------------- TX i  ic    10 10  + I Inter – Tech + 10 10  + NR Inter – Tech = E PBCH –  10  Log  DL DL       All TXj  jc             The PDCCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel. subscriber. or mobile Mi. or mobile Mi. subscriber. subscriber. subscriber. or mobile Mi is • located. subscriber. or mobile Mi: RSRQ 520 TX i  ic  TX i  ic  = 10  Log  N FB  TX  ic   + E i – RSSI DLRS  TX i  ic  . Mi • G Div – PBCH : PBCH diversity gain defined in the reception equipment of the terminal used by the pixel. • i BLER  BDL : Downlink block error rate read from the graphs available in the reception equipment assigned to the M terminal used by the pixel.3. subscriber. Mi G Div – PDCCH : PDCCH diversity gain defined in the reception equipment of the terminal used by the pixel. subscriber. subscriber. subscriber. Calculations The downlink reference signal C/(I+N) for cell TXi(ic) is calculated as follows for any pixel. • or mobile Mi. or mobile Mi: TX  jc  TX  ic  i CINR SS = TX  ic  i E SS TX  ic  i      j  n Sym  SS PBCH-     --------------------------------------------- Inter – Tech Inter – Tech 10 10  + NR   10  +I –  10  Log  + 10 DL DL       All TXj  jc             The PBCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel. + NR Inter = 10  Log  + 10 DL TX i  ic    10  + I DL  + 10  Log   All TXj  jc      N SD  Slot  N Slot  SF  N SF – DL      With N SCa – FB calculated as follows: W FB N SCa – FB = --------F Bearer Determination: 521 . see "Interference Calculation (DL)" on page 456.+ NR Inter = 10  Log 10 + I DL + 10 + 10  Log ---------------------------------------------DL TX  ic    TXi  ic      i  N SF – DL + N TDD – SSF  All TXj  jc          Method 2: Non-synchronised Transmission and Reception For details.e.+ NR Inter = 10  Log 10 + I DL + 10 + 10  Log ------------------------------------------------------------------DL TX  ic        i  N SD  Slot  N Slot  SF  N SF – DL  All TXj  jc         TX  ic  i    j  n Sym  TX i  ic  PDSCH-   N TXi  ic    ---------------------------------------- – Inter – Tech 10  10  Sym – PDSCH + N Sym – PDCCH     -------------------------------------------------------------------.+ NR Inter = 10  Log 10 +I + 10 + 10  Log ---------------------------------------------DL TX i  ic    TX i  ic     DL    N SF – DL + N TDD – SSF  All TXj  jc         i    j   n TX  ic  PDSCH- Sym - i    --------------------------------------N Sym – PDSCH  10  10  Inter – Tech – Tech    . the received signals from the server (TXi(ic)). i.. or mobile Mi: TX  jc  TX i  ic   I + N  SS PBCH TX  ic  i    j  n Sym  SS PBCH   ---------------------------------------------  10 10  + I Inter – Tech + 10 10  + NRInter – Tech + 10  Log  N = 10  Log  DL SCa – FB  N FB – SS PBCH     DL  All TX j  jc          The PDSCH and PDCCH total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  i Where E DLRS is the cell’s RSRP and RSSI TX  ic  i is the received signal strength indicator. subscriber.Atoll 3. the encircled 10 in the formulas below is replaced with 8. TX i  ic   I + N  PDCCH TX i  ic   I + N  PDSCH TX  jc  TX  ic  TX  jc  TX  ic  i   j   n Sym  TX i  ic  PDCCH  -------------------  --------------------- N Sym – PDCCH  Inter – Tech – Tech 10  10     . subscriber. see "Interference Calculation (DL)" on page 456. calculated as follows: For cells using more than 1 antenna port. or mobile Mi: TX  jc  TX i  ic   I + N  DLRS TX  ic  i    j  n Sym  DLRS    --------------------------------------- TX  ic  Inter – Tech 10 10  + NR Inter – Tech + 10  Log  2  N i   10  +I = 10  Log  + 10 DL FB      DL  All TXj  jc          The SS and PBCH total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel. and all the interfering cells (TXj(jc)). TX  jc  RSSI TX  ic  i TX  ic  i     j  n Sym RSSI -  TX  ic    --------------------------------------TX  ic  10  10 i Inter – Tech Inter – Tech i   = 10  Log   RSSI + 10 + + 10  12 + NR DL + 10  Log  N FB    I DL    All TX j  jc        The downlink reference signal total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel. or mobile Mi: Method 1: Synchronised Transmission and Reception For details. TX  jc  TX i  ic   I + N  PDCCH  TX  jc  TX i  ic   I + N  PDSCH TX  ic  i    j  n Sym  TX  ic  PDCCH i  N TXi  ic     ---------------------------------------- – Inter – Tech 10  10  Sym – PDSCH + N Sym – PDCCH     . subscriber. • Whose selection thresholds are less than the PDSCH C/(I+N) at Mi: T B  CINR PDSCH Mi TX i  ic  If the cell supports MIMO. or mobile Mi. For more information. • Effective RLC Throughput From among the bearers available for selection. or mobile Mi for N Ant – TX . TX i  ic  . the selected bearer is the one with the highest index. corresponding to the bearer is applied to its selection threshold. transmit diversity. Allocated Bandwidth Throughput. Cell Capacity. subscriber.ini file. the selected bearer is the one with the highest downlink effective RLC channel throughput as calculated in "Channel Throughput. the PDSCH C/(I+N) calculated above becomes: TX i  ic  TX i  ic  Mi DL CINR PDSCH = CINR PDSCH + G Div – DL + G Div Mi Where G Div – DL is the transmit diversity. or mobile Mi G Div is also applied. Once the bearer is known. Output 522 TX i  ic  • CINR DLRS : Downlink reference signal C/(I+N) from cell TXi(ic) at pixel. • CINR SS : SS C/(I+N) from cell TXi(ic) at pixel.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 The bearers available for selection in the pixel. M i BLER  B DL . Therefore. subscriber. subscriber. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. or mobile Mi. SU-MIMO diversity. or mobile Mi’s reception equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). the PBCH and PDCCH C/(I+N) become: TX i  ic  TX i  ic  Mi TX i  ic  TX i  ic  Mi DL CINR PBCH = CINR PBCH + G Div – PBCH + G Div DL CINR PDCCH = CINR PDCCH + G Div – PDCCH + G Div The PBCH and PDCCH diversity gains are applied to the C/(I+N) when the cell and the terminal both support any form of MIMO in downlink. or MU-MIMO diversity gain. and Per-user Throughput Calculation" on page 547. subscriber. the selected bearer is the one with the highest downlink peak RLC channel throughput as calculated in "Channel Throughput. the bearers available for selection are all the bearers defined in the reception equipment for which the following is true: M M i i TX  ic  i DL T B – G Div – DL – G Div  CINR PDSCH The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). • Bearer Index From among the bearers available for selection. or mobile Mi. Cell Capacity. Mobility  M i  . and Per-user Throughput Calculation" on page 547. if the corresponding option has been set in the Atoll. • Peak RLC Throughput From among the bearers available for selection. N Ant – RX . MIMO Diversity Gain: With MIMO. Allocated Bandwidth Throughput. The gain is read from the properties of the TX i  ic  Mi reception equipment assigned to the pixel. • TX  ic  i CINR PBCH : PBCH C/(I+N) from cell TXi(ic) at pixel. or MU-MIMO diversity gain applied if the cell supports MIMO and depending on the Mi diversity mode. G Div – DL . see the Administrator Manual. the SU-MIMO and MU-MIMO thresholds and criteria.Atoll 3. and according to the Mi diversity mode depending on the SU-MIMO and MU-MIMO Mi thresholds and criteria. SU-MIMO diversity. The additional downlink diversity gain defined per clutter is also applied.   DL The additional downlink diversity gain defined for the clutter class of the pixel. subscriber. subscriber.3. or mobile Mi covered by a cell TXi(ic). 523 . subscriber. or mobile Mi. or mobile Mi in the downlink. and all the interfering cells (TXj(jc)). subscriber. or mobile Mi after power control : Fractional uplink power control factor defined for the cell TXi(ic). • P Eff : Effective transmission power of the terminal used by the pixel. subscriber. • M Shadowing – Model : Shadowing margin based on the model standard deviation. or mobile Mi covered by a cell TXi(ic).e. subscriber. or mobile TX  ic  i Mi covered by a cell TXi(ic). subscriber. • L • L Path : Path loss ( L Path = L Model + L Ant ). subscriber. • n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic). subscriber. or mobile Mi covered by a cell TXi(ic). TX  ic  i • NR UL – ICIC : ICIC uplink noise rise of the cell TXi(ic). or mobile Mi. subscriber.. at pixel. • TX i  ic   I + N  SS PBCH : SS and PBCH total noise from the interfering cells TXj(jc) at the pixel.Atoll 3. Mi Mi adjustment as calculated in "C/(I+N) and Bearer Calculation (UL)" on page 532. TX i • G Ant : Transmitter antenna gain for the antenna used by the transmitter TXi. • RSRQ TX  ic  i TX  ic  i : Downlink reference signal received quality from cell TXi(ic) at pixel.4.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  i • CINR PDCCH : PDCCH C/(I+N) from cell TXi(ic) at pixel. • TX i  ic   I + N  PDSCH : PDSCH total noise from the interfering cells TXj(jc) at the pixel. • TX i  ic   I + N  PDCCH : PDCCH total noise from the interfering cells TXj(jc) at the pixel. for the channel bandwidth used by the cell TXi(ic). • L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model. defined in the frequency bands table. • CINR PDSCH : PDSCH C/(I+N) from cell TXi(ic) at pixel. This value can be user-defined or calculated as explained in "Interference Calculation (UL)" on page 525. subscriber.3. the received signals from the server (TXi(ic)). • Mi B DL : Bearer assigned to the pixel. • N FB TX i  ic  TX i  ic  : Number of frequency blocks. TX i TX i In coverage predictions. 6.4. TX i  ic  • RSSI : Received signal strength indicator. i. subscriber. or mobile Mi. This value can be user-defined or calculated as explained in "Interference Calculation (UL)" on page 525. or mobile Mi. • NR UL TX  ic  i : Uplink noise rise of the cell TXi(ic).6 Signal Level Calculation (UL) Input TX i  ic  • CINR PUSCH – Max : Maximum PUSCH C/(I+N) defined for the cell TXi(ic). • L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected. subscriber. shadowing margins are taken into account when the option "Shadowing taken into account" is selected. or mobile Mi. • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi. TX i : Total transmitter losses for the transmitter TXi ( L TX i = L Total – UL ). •  I + N  DLRS : Downlink reference signals total noise from the interfering cells TXj(jc) at the pixel. TX  ic  i •  FPC • P Max : Maximum transmission power of the terminal used by the pixel. • L Total : Total loss calculated as explained in "Signal Level Calculation (DL)" on page 501. The total symbol duration of a modulation symbol comprises the useful symbol duration. the ratio of the useful symbol energy to the total symbol energy. the maximum allowed transmission power for the terminal used by the pixel.V) = (0. This power is calculated by performing fractional power control. • D CP TX i  ic  : Cyclic prefix duration defined in TXi(ic) frame configuration or.Atoll 3.  Next. f CP is the cyclic prefix factor. Hence. L Ant is determined in the direction (H. subscribers. subscriber. L Ant is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi. subscriber. i • G • L Ant : Receiver terminal’s antenna attenuation calculated for the pixel. Mi • L Body : Body loss defined for the service used by the pixel. or mobile Mi is calculated as follows: Mi TX i  ic  TX i  ic  TX i  ic   Mi  P Allowed = Min  P Max 10  Log  NFB  + P O_PUSCH +  FPC  L Total      Once the maximum allowed power has been calculated. and is P Mi Mi = P Eff after power control adjustment. or mobile Mi at its serving cell TXi(ic) is calculated as follows: Mi Mi TX i C PUSCH PUCCH = EIRP PUSCH PUCCH – L Path – M Shadowing – Model – L Indoor + G Ant – L TX i Mi Mi – L Ant – L Body + f CP Where EIRP is the effective isotropic radiated power of the terminal calculated as follows: Mi EIRP PUSCH PUCCH = P With P Mi Mi Mi +G Mi –L Mi = P Allowed without power control adjustment at the start of the calculations. or mobile Mi. or mobile Mi. Fractional Power Control: Fractional power control imposes a limitation on the maximum transmission power of the terminal..  TX  ic  i i i i i P O_PUSCH = CINR PUSCH – Max + NRUL – ICIC + n PUSCH PUCCH – 10  Log  NFB TX i  ic   for cell-edge.3. i Mi For calculating the useful signal level from the best serving cell. subscriber. otherwise. subscriber. and a cyclic prefix. This nominal PUSCH power is calculated as follows: TX i  ic  TX i  ic  TX i  ic  TX  ic  TX  ic  TX  ic  P O_PUSCH = CINR PUSCH – Max + NRUL TX  ic  TX  ic  i i + n PUSCH PUCCH – 10  Log  N FB TX  ic   for cell centre. while the antenna is pointed towards Mi’s best serving cell. it is used as an upper limit for transmission power in all the remaining calculations. or mobile Mi.e. subscriber. subscriber. A nominal PUSCH power is indicated by the cell to all the pixels. or mobile Mi. subscriber. carrying the actual data bits. or mobile Mi. M : Receiver terminal’s antenna gain for the pixel. added to the useful data bits as padding against multi-path to avoid inter-symbol interference. The received PUSCH and PUCCH signal level (dBm) from a pixel. or mobiles.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks • L M i M ©Forsk 2015 : Receiver terminal losses for the pixel. Calculations Atoll first calculates the allowed maximum transmission power for the terminal used by the pixel.  TX i  ic  Where n PUSCH PUCCH – 10  Log  N FB   corresponds to the uplink noise over 1 frequency block. i. For calculating the interfering signal level from any interferer. Once a modulation 524 .0) from Mi the antenna patterns of the antenna used by Mi. the total energy within a modulation symbol belongs in part to the useful data bits and in part to the cyclic prefix. in the global network settings. The uplink noise comprises thermal noise and the noise figure of the equipment. Mi 6. Output Mi • C PUSCH PUCCH : Received PUSCH and PUCCH signal level from the pixel. • P Allowed : Maximum allowed transmission power for the terminal used by the pixel.8 Interference Calculation (UL) The PUSCH and PUCCH interference is only calculated during Monte Carlo simulations. defined in the frequency bands table.. W FB : Width of a frequency block in the frequency domain (180 kHz). or mobile Mi. or mobile Mi at a cell TXi(ic). Input • • • K: Boltzmann’s constant. it remains constant for a given temperature. In coverage predictions. The used bandwidth depends on the number of used subcarriers. subscriber.3.4. for the channel bandwidth used by the cell TXi(ic). The energy belonging to the cyclic prefix is lost once it has served its purpose of combatting inter-symbol interference. Therefore. only the energy of the useful data bits can be used for extracting the data.5  If D CPi = Normal  TX i  ic  =  = Extended  10  Log  6  7. Atoll calculates the uplink noise over the channel bandwidth used by the cell.7 Noise Calculation (UL) For determining the C/N and C/(I+N).Atoll 3. T: Temperature in Kelvin. 525 . The thermal noise density depends on the temperature. However. the uplink noise rise values already available in simulation results or in the Cells table are used. 6. Calculations The power spectral density of thermal noise is calculated as follows: n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz The thermal noise for the PUSCH and the PUCCH is calculated as: TX i  ic  TX i  ic  n 0 – PUSCH PUCCH = n 0 + 10  Log  N FB  W FB  1000 The uplink noise is the sum of the thermal noise and the noise figure of the cell TXi(ic). i. The interference received by a cell TXi(ic) from an interfering mobile covered by a cell TXj(jc) can be defined as the PUSCH and PUCCH signal level received from the interfering mobile Mj depending on the overlap that exists between the channels used by the cells TXi(ic) and TXj(jc) and on the traffic load of the interfering mobile Mj.e.5  If D CP  is an interferer 0 If M i  The cyclic prefix energy and the useful data bits energy are both taken into account when calculating interfering signal levels. • N FB TX i  ic  : Number of frequency blocks.4.4. f CP  TX  ic   10  Log  7  7. subscriber.4.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 symbol is received. TX i  ic  TX i  ic  n PUSCH PUCCH = n 0 – PUSCH PUCCH + nf TX i  ic  Output • TX i  ic  n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic). f CP implies that the energy belonging to the cyclic prefix is excluded from the useful signal level. • nf TX i  ic  : Noise figure of the cell TXi(ic). the value of the thermal noise varies with the used bandwidth. Atoll 3. • TL UL : Uplink traffic load of the interfering mobile Mj. Mj Traffic loads are calculated during Monte Carlo simulations as explained in "Scheduling and Radio Resource Allocation" on page 552. Calculations The uplink interference received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) is calculated as follows: Mj TX i  ic  – TX j  jc  Mj I PUSCH PUCCH = C PUSCH PUCCH + f O Mj Mj TX i  ic  – TX j  jc  + f TL – UL + f ICIC – UL TX i  ic  – TX j  jc  + f ABS – UL Where f TL – UL is an interference reduction factor due to the uplink traffic load of the interfering mobile Mj.8. TX i  ic  TX j  jc  N FB – CE0 and N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the • channel bandwidth used by the cell and PSS ID 0.4. as explained in "Signal Level Calculation (UL)" on page 523. calculated as follows: M M j j f TL – UL = 10  Log  TL UL   526 . Mj C PUSCH PUCCH : PUSCH and PUCCH signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell • TXj(jc) as calculated in "Signal Level Calculation (UL)" on page 523. In Monte Carlo simulations.and Adjacent Channel Overlaps Calculation" on page 493. TX i  ic  TX j  jc  N FB – CE2 and N FB – CE2 : Number of cell-edge frequency blocks defined in the frame configurations table for the • channel bandwidth used by the cell and PSS ID 2. In coverage predictions. • M Shadowing – C  I : Shadowing margin based on the C/I standard deviation. Calculation of the uplink noise rise which represents the total uplink interference from all interfering mobiles as explained in "Noise Rise Calculation (UL)" on page 528. Interfering Signal Level Calculation (UL) Input TX  ic  i N FB • TX  jc  j and N FB : Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell. the ratio M Shadowing – Model – M Shadowing – C  I is applied to the interfering signals (for more information. M Shadowing – C  I is added to the received interfering signal levels in order to achieve the ratio M Shadowing – Model – M Shadowing – C  I : M M j j C PUSCH PUCCH = C PUSCH PUCCH + M Shadowing – C  I In coverage predictions.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 The calculation of uplink interference can be divided into two parts: • • 6. As the interfering signal levels already include M Shadowing – Model .3.4. see "Shadow Fading Model" on page 90). interfering signal levels already include M Shadowing – Model .1 Calculation of the uplink interference from each individual interfering mobile as explained in "Interfering Signal Level Calculation (UL)" on page 526. TX i  ic  – TX j  jc  • rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co. shadowing margins are taken into account when the option "Shadowing taken into account" is selected. • M Shadowing – Model : Shadowing margin based on the model standard deviation. TX i  ic  TX j  jc  N FB – CE1 and N FB – CE1 : Number of cell-edge frequency blocks defined in the frame configurations table for the • channel bandwidth used by the cell and PSS ID 1. N FB – CC Common N FB – CC = -------------------TX i  ic  N FB – CC Common N FB – CE = -------------------TX  ic  i N FB – CE Common is the number of common frequency blocks in TXi(ic) and TXj(jc) in cell centre. or N FB – CE2 depending on the PSS ID of TXi(ic). subscriber. N FB – CE TX i  ic  is the TX i  ic  number of common frequency blocks in TXi(ic) and TXj(jc) on cell-edge. 527 . Whether a pixel. In Monte Carlo simulations.and adjacent channel overlap between the cells TXi(ic) and TXj(jc) is calculated as follows: TX i  ic  – TX j  jc  fO TX i  ic  – TX j  jc  = 10  Log  r O   Interference reduction due to static uplink ICIC using fractional frequency reuse: If the cell supports static ICIC in uplink. and for the mobiles located in the cell centre of the interfered cell Atoll calculates the UL Noise Rise. The interference reduction factor due to static uplink ICIC using fractional frequency reuse is calculated as follows: TX i  ic  – TX j  jc  f ICIC – UL TX i  ic  – TX j  jc  = 10  Log  p Collision  TX i  ic  – TX j  jc  Where p Collision   is the collision probability between the subcarriers used by the interfered and interfering cells.Atoll 3.and adjacent channel overlap between the studied and the interfering cells: Interference reduction due to the co. it means that a part of the LTE frame may use a fraction of the channel bandwidth. Atoll calculates two separate noise rise values. for the mobiles located in the cell-edge of the interfered cell Atoll calculates the ICIC UL Noise Rise. N FB – CC and N FB – CE are respectively the numbers of frequency blocks in cell centre and cell-edge of TXi(ic). It is determined during Monte Carlo simulations as follows: TX i  ic  – TX j  jc  Cell centre: p Collision TX i  ic  – TX j  jc  Cell-edge: p Collision Common Where. Interference reduction due to subframe collision probabilities: TX i  ic  – TX j  jc  The interference reduction factor due to uplink subframe collision probabilities f ABS – UL is calculated as explained in "Subframe Pattern Collision Calculation" on page 497. Number of frequency blocks in ICIC mode Cell centre Cell edge TX i  ic  No FFR N FB Time-switched FFR N FB TX i  ic  N FB TX i  ic  TX i  ic  N FB – CEx TX i  ic  Hard FFR TX i  ic  N FB – CEx TX  ic  i Soft FFR N FB Partial soft FFR TX  ic  i TX i  ic  N FB N FB – CEx TX  ic  i TX  ic  i – N FB – CEx TX  ic  TX  ic  N FB – CEx TX  ic  i i i –  N FB – CE0 + N FB – CE1 + N FB – CE2 TX  ic  i TX  ic  i TX i  ic  N FB – CEx TX  ic  i Where N FB – CEx can be N FB – CE0 . or mobile is located in the cell-edge is determined as explained in "Best Server Determination" on page 535.3. N FB – CE1 .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  – TX  jc  i j Calculations for the interference reduction factors due to channel overlapping ( f O TX  ic  – TX  jc  i j fractional frequency reuse ( f ICIC – UL ) and static uplink ICIC using ) are explained below: Interference reduction due to the co. subscriber.2 Noise Rise Calculation (UL) The uplink noise rise is defined as the ratio of the total uplink interference received by any cell TXi(ic) from all interfering mobiles Mj present in the coverage areas of all other cells TXj(jc) to the uplink noise of the cell TXi(ic). Atoll calculates the PUSCH and PUCCH total noise (I+N) as follows: TX i  ic  TX i  ic   I + N  PUSCH PUCCH = NR UL TX i  ic  + n PUSCH PUCCH For any mobile Mi in the cell-edge of the interfered cell TXi(ic). Calculations For any mobile Mi in the cell centre of the interfered cell TXi(ic). cell centre or cell-edge. TX i  ic  . see "Noise Rise Calculation (UL)" on page 528. In other words. Atoll calculates the ICIC UL Noise Rise as follows: M TX i  ic  NR UL – ICIC j   TX i  ic   IPUSCH PUCCH    ICIC M i n PUSCH PUCCH   - ------------------------------------------- TX  ic   -----------------------------------------------------------------10 10  + NR Inter – Tech – n i = 10  Log  10 + 10   UL PUSCH PUCCH     All Mj        All TXj  jc     For any pixel. according to the zone. TX i  ic  • NRUL – ICIC : ICIC uplink noise rise for the cell TXi(ic). or mobile Mi in cell-edge of the interfered cell TXi(ic). it is the ratio (I+N)/N. Atoll calculates the UL Noise Rise as follows: M TX i  ic  NR UL j   TX  ic   I PUSCH PUCCH  i   non-ICIC M i n PUSCH PUCCH   - ------------------------------------------- TX  ic   ----------------------------------------------------------------------------10 10  + NR Inter – Tech – n i = 10  Log   10  + 10 UL PUSCH PUCCH    All M j        All TXj  jc     For any pixel. 6. receiver. •  I + N  PUSCH PUCCH : PUSCH and PUCCH total noise for a cell TXi(ic) calculated for any pixel. Input Mj I PUSCH PUCCH : PUSCH and PUCCH interference signal levels received at a cell TXi(ic) from interfering mobiles Mj • covered by other cells TXj(jc) as calculated in "Interfering Signal Level Calculation (UL)" on page 526.Atoll 3. or mobile Mi in the cell centre of the interfered cell TXi(ic). Inter – Tech • NRUL : Inter-technology uplink noise rise. subscriber. Output Mj I PUSCH PUCCH : PUSCH and PUCCH interference signal level received at a cell TXi(ic) from an interfering mobile Mj • covered by a cell TXj(jc).3. Atoll calculates the PUSCH and PUCCH total noise (I+N) as follows: TX  ic  i TX  ic  i TX  ic  i  I + N  PUSCH PUCCH = NR UL – ICIC + n PUSCH PUCCH Output • 528 TX i  ic  NRUL : Uplink noise rise for the cell TXi(ic).8. or subscriber is located. Atoll uses either the ICIC UL Noise Rise or the UL Noise Rise to calculate the PUSCH and PUCCH C/(I+N).4. and calculations on subscriber lists. TX i  ic  n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic) as calculated in "Noise Calculation (UL)" • on page 525.4. where the pixel.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 In coverage predictions. For more information on the calculation of the uplink noise rise. subscriber. point analysis. or mobile Mi. Mi • P Min : Minimum transmission power of the terminal used by the pixel. subscriber.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 6.4.9 C/N Calculation (UL) Input • M i C PUSCH PUCCH : Received PUSCH and PUCCH signal level from the pixel. subscriber. TX i  ic  • N Ant – RX : Number of reception (uplink) antenna ports defined for the cell TXi(ic). or mobile Mi.4. subscriber. if the corresponding option has been set in the Atoll. or mobile Mi as Mi calculated in "Signal Level Calculation (UL)" on page 523. or mobile Mi at its serving cell TXi(ic) as calculated in "Signal Level Calculation (UL)" on page 523. 529 .ini file. • M i BLER  B UL : Uplink block error rate read from the graphs available in the reception equipment assigned to the cell   TXi(ic). • T B : Bearer selection thresholds of the bearers defined in the reception equipment used bythe cell TXi(ic). subscriber. • TX i  ic  N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 0. • M i N Ant – TX : Number of transmission (uplink) antenna ports defined for the terminal used by the pixel.3. • TX i  ic  N FB – CE2 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 2. • TX  ic  i N FB – CE1 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 1. subscriber. or mobile Mi. • T SU – MIMO – UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(ic). TX i  ic  • T B – Lowest : Bearer selection threshold of the lowest bearer in the reception equipment assigned to the cell TXi(ic). For more information. see the Administrator Manual. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. Calculations The PUSCH and PUCCH C/N from a pixel. TX i  ic  • n PUSCH PUCCH : PUSCH and PUCCH noise for the cell TXi(ic) as calculated in "Noise Calculation (UL)" on page 525. subscriber.Atoll 3. • M PC : Power control adjustment margin defined in the global network settings. • B UL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. • Mobility  M i  : Mobility used for the calculations. • P Allowed : Maximum allowed transmission power of the terminal used by the pixel. subscriber. or mobile Mi. • N FB TX i  ic  TX  ic  i : Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell TXi(ic). Mi Mi or mobile Mi. • Mi B UL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. or mobile Mi at its serving cell TXi(ic) is calculated as follows: Mi Mi TX i  ic  CNR PUSCH PUCCH = C PUSCH PUCCH – n PUSCH PUCCH Bearer Determination: The bearers available for selection in the cell TXi(ic)’s reception equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). as a mobile moves from good to bad radio conditions. and according to the Mi diversity mode depending on the SU-MIMO and MU-MIMO TX  ic  i thresholds and criteria. N Ant – RX . SU-MIMO diversity. The PUSCH and PUCCH C/N calculated above is given for the total number of frequency blocks associated with the channel TX i  ic  bandwidth of the cell.3. or mobiles in the uplink. the uplink bandwidth is not reduced to 4 because it does not provide any gain in terms of the bearer. and may reduce the number of used frequency blocks in order to satisfy the selected target. For example. the number of frequency blocks used by it for transmission in uplink are reduced one by one in order to improve the PUSCH and PUCCH C/N.. the PUSCH and PUCCH C/N calculated above become: Mi TX i  ic  Mi UL CNR PUSCH PUCCH = CNR PUSCH PUCCH + G Div – UL + G Div TX i  ic  Where G Div – UL is the receive diversity. • Maintain Connection The bandwidth used for transmission by a mobile is reduced only if the PUSCH and PUCCH C/N is not enough to even access the lowest bearer. subscribers. UL The additional uplink diversity gain defined for the clutter class of the pixel. or MU-MIMO diversity gain. and Per-user Throughput Calculation" on page 547. • Peak RLC Throughput From among the bearers available for selection. SU-MIMO diversity. there is no gain in the PUSCH and PUCCH C/N. BLER  B UL . Allocated Bandwidth Throughput. and using 6 it would only get access to the second best. Therefore. i. receive diversity. the selected bearer is the one with the highest uplink peak RLC channel throughput as calculated in "Channel Throughput. Although using 4 frequency blocks.Atoll 3. As there is no reduction in the bandwidth used for transmission. • Full Bandwidth Full channel width is used by each mobile in the uplink. subscriber. the selected bearer is the one with the highest index. the bearers available for selection are all the bearers defined in the reception equipment for which the following is true: Mi TX i  ic  Mi UL T B – G Div – UL – G Div  CNR PUSCH PUCCH The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). G Div – UL . • Effective RLC Throughput From among the bearers available for selection. the selected bearer is the one with the highest uplink effective RLC channel throughput as calculated in "Channel Throughput. N FB . its PUSCH and PUCCH C/N will be better than when using 5. For example. of Used Frequency Blocks): The uplink bandwidth allocation depends on the target defined for the scheduler used by the cell TXi(ic). the SU-MIMO and MU-MIMO threshold and criteria. • Best Bearer The bandwidth used for transmission by a mobile is reduced in order to improve the PUSCH and PUCCH C/N enough to access the best bearer. Cell Capacity.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks • ©Forsk 2015 M i M i Whose selection thresholds are less than the PUSCH and PUCCH C/N at Mi: T B  CNR PUSCH PUCCH If the cell supports MIMO. if using 5 frequency blocks. it will be assigned 5 frequency blocks as the used uplink bandwidth. Cell Capacity. or mobile Mi G Div is also applied. Mobility  M i  . The calculation of the gain introduced by the bandwidth reduction is explained below. • Bearer Index From among the bearers available for selection.e. a mobile is able to access the best bearer. Bandwidth allocation is performed for all the pixels. Allocated Bandwidth Throughput. i. corresponding to the bearer is applied to its selection threshold. The gain is read from the properties of the M TX  ic  M i i i reception equipment assigned to the cell TXi(ic) for N Ant – TX . The calculation of the gain introduced by the bandwidth reduction is explained below. and Per-user Throughput Calculation" on page 547. or MU-MIMO diversity gain applied if the cell supports MIMO and depending on the Mi diversity mode. the mobile already has the best bearer using 5 frequency blocks. 530 .e. Uplink Bandwidth Allocation (No.. MIMO Diversity Gain: Once the bearer is known. subscriber. or mobile Mi in the cell-edge of the interfered cell TXi(ic). N FB – CE1 .Atoll 3. with the highest peak RLC throughput. or N FB – CE2 depending on the PSS ID of TXi(ic). or mobile Mi at it serving cell TXi(ic).e. subscriber. subscriber. subscriber. and N FB – UL  Service   N FB – UL  N FB – CE for any pixel. or mobile Mi.. for the bearer selected for the pixel. or mobile Mi in the cell centre of the interfered TX i  ic  Mi Min cell TXi(ic). subscriber.. Atoll continues to work with the C/N given by the bandwidth allocation. i. Final The pixel. The transmission power of Mi is reduced to determine the effective transmission power from the pixel. The gain related to this bandwidth reduction is applied to the PUSCH and PUCCH C/N: Mi CNR PUSCH PUCCH Final  N TX i  ic   FB - = CNR PUSCH PUCCH + 10  Log  ---------------- Mi  All FB  N FB – UL Mi TX i  ic  Mi Min Where N FB – UL  Service   N FB – UL  N FB – CC for any pixel. or mobile Mi reduces its transmission power so that the PUSCH and PUCCH C/N from it at its cell is just enough to get the selected bearer. The uplink bandwidth allocation may result in the use of a number of frequency blocks which is less than the number of frequency blocks associated with the channel bandwidth of the cell.3. Uplink Power Control Adjustment: Once the bandwidth allocation is performed. Number of frequency blocks in ICIC mode Cell centre Cell edge TX  ic  i N FB No FFR TX  ic  i N FB TX i  ic  Time-switched FFR TX i  ic  N FB N FB – CEx TX i  ic  Hard FFR TX i  ic  N FB – CEx TX i  ic  Soft FFR N FB TX i  ic  Partial soft FFR N FB TX i  ic  N FB – CEx TX i  ic  TX i  ic  – N FB – CEx TX  ic  N FB – CEx TX  ic  TX  ic  TX i  ic  i i i –  N FB – CE0 + N FB – CE1 + N FB – CE2 TX i  ic  TX i  ic  N FB – CEx TX i  ic  Where N FB – CEx can be N FB – CE0 . or with the highest effective RLC throughput.e. subscriber. Mi Mi i. 531 . where T B UL TX i  ic  Mi B UL is the bearer selection threshold. from the reception equipment assigned to the cell TXi(ic). bearer with the highest index. or mobile Mi as follows: Mi Mi  Mi   TXi  ic    Mi  P Eff = Max  P Allowed –  CNR PUSCH PUCCH –  T M + M PC   P Min i    B   UL Mi Mi CNR PUSCH PUCCH is calculated again using P Eff . CNR PUSCH PUCCH = CNR PUSCH PUCCH . Output • Mi CNR PUSCH PUCCH : PUSCH and PUCCH C/N from a pixel. If with P Mi Mi Mi = P Allowed AND CNR PUSCH PUCCH  T TX i  ic  Mi + M PC .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 The definition of the best bearer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). N Ant – RX : Number of reception (uplink) antenna ports defined for the cell TXi(ic). subscriber. Atoll calculates the received signal level from each pixel. • M PC : Power control adjustment margin defined in the global network settings. the receiver is oriented towards the best server just as in the case of LOS. . • TX i  ic  N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 0. or mobile at its serving cell using the effective power of the terminal used by the pixel. TX i  ic  • NRUL – ICIC : ICIC uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 528. subscriber. Atoll does not try to find the direction of the strongest signal. Finally. subscriber. Atoll calculates the uplink carrier to noise ratio as explained in "C/N Calculation (UL)" on page 529.3. • TX i  ic  N FB – CE2 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 2. or mobile Mi. • N FB TX i  ic  TX i  ic  : Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell TXi(ic). • Mobility  M i  : Mobility used for the calculations. • TX  ic  i N FB – CE1 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 1. determines the uplink C/(I+N) by dividing the previously calculated uplink C/N by the uplink noise rise value of the cell as calculated in "Noise Rise Calculation (UL)" on page 528.4. First. Mi Mi or mobile Mi. In the case of NLOS between the receiver and the best server. or mobile Mi as Mi calculated in "Signal Level Calculation (UL)" on page 523. or mobile as explained in "Signal Level Calculation (UL)" on page 523. subscriber.10 C/(I+N) and Bearer Calculation (UL) The carrier signal to interference and noise ratio is calculated in three steps. subscriber.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 6. • T B : Bearer selection thresholds of the bearers defined in the reception equipment used bythe cell TXi(ic). or mobile Mi. except when the "Lock Status" is set to "Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited. Next. subscriber. or mobile Mi at it serving cell TXi(ic) as calculated in "C/N Calculation (UL)" on page 529. • TX i  ic  NRUL : Uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 528. or mobile Mi. • Mi B UL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. • P Allowed : Maximum allowed transmission power of the terminal used by the pixel. • Mi N Ant – TX : Number of transmission (uplink) antenna ports defined for the terminal used by the pixel. subscriber. The receiver terminal is always considered to be oriented towards its best server. subscriber.4. • 532 TX i  ic  • M i BLER  BUL : Uplink block error rate read from the graphs available in the reception equipment assigned to the cell   TXi(ic). Mi • P Min : Minimum transmission power of the terminal used by the pixel. Input • Mi CNR PUSCH PUCCH : PUSCH and PUCCH C/N from a pixel. TX i  ic  • T B – Lowest : Bearer selection threshold of the lowest bearer in the reception equipment assigned to the cell TXi(ic).Atoll 3. • T SU – MIMO – UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(c). • B UL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. The uplink noise rise can be set by the user manually for each cell or calculated using Monte Carlo simulations. Cell Capacity. N Ant – RX . Allocated Bandwidth Throughput.ini file. subscriber. the PUSCH and PUCCH C/(I+N) calculated above become: Mi Mi TX i  ic  UL CINR PUSCH PUCCH = CINR PUSCH PUCCH + G Div – UL + G Div TX i  ic  Where G Div – UL is the receive diversity. subscriber. Uplink Bandwidth Allocation (No. of Used Frequency Blocks): The uplink bandwidth allocation depends on the target defined for the scheduler used by the cell TXi(ic). The gain is read from the properties of the M TX  ic  M i i i reception equipment assigned to the cell TXi(ic) for N Ant – TX . the bearers available for selection are all the bearers defined in the reception equipment for which the following is true: M i TX  ic  i M UL i T B – G Div – UL – G Div  CINR PUSCH PUCCH The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). MIMO Diversity Gain: Once the bearer is known. SU-MIMO diversity. • Peak RLC Throughput From among the bearers available for selection. see the Administrator Manual. The PUSCH and PUCCH C/(I+N) calculated above is given for the total number of frequency blocks associated with the channel 533 . G Div – UL . SU-MIMO diversity. the SU-MIMO and MU-MIMO threshold and criteria.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Calculations For any pixel. if the corresponding option has been set in the Atoll. and according to the Mi diversity mode depending on the SU-MIMO and MU-MIMO TX i  ic  thresholds and criteria. • Bearer Index From among the bearers available for selection. and Per-user Throughput Calculation" on page 547. Therefore.3. subscriber. or mobile Mi G Div is also applied. or MU-MIMO diversity gain. the selected bearer is the one with the highest uplink effective RLC channel throughput as calculated in "Channel Throughput. Atoll calculates the PUSCH and PUCCH C/ (I+N) as follows: Mi TX i  ic  Mi CINR PUSCH PUCCH = CNR PUSCH PUCCH – NR UL – ICIC Bearer Determination: The bearers available for selection in the cell TXi(ic)’s reception equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). Allocated Bandwidth Throughput. the selected bearer is the one with the highest index. • Effective RLC Throughput From among the bearers available for selection.Atoll 3. • Whose selection thresholds are less than the PUSCH and PUCCH C/(I+N) at Mi: T B  CINR PUSCH PUCCH Mi Mi If the cell supports MIMO. or MU-MIMO diversity gain applied if the cell supports MIMO and depending on the Mi diversity mode. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi.   UL The additional uplink diversity gain defined for the clutter class of the pixel. corresponding to the bearer is applied to its selection threshold. Atoll calculates the PUSCH and PUCCH C/ (I+N) as follows: M M i TX  ic  i i CINR PUSCH PUCCH = CNR PUSCH PUCCH – NR UL For any pixel. or mobile Mi in the cell centre of the interfered cell TXi(ic). For more information. and Per-user Throughput Calculation" on page 547. BLER  BUL . Mobility  M i  . the selected bearer is the one with the highest uplink peak RLC channel throughput as calculated in "Channel Throughput. receive diversity. Cell Capacity. or mobile Mi in the cell-edge of the interfered cell TXi(ic). Atoll continues to work with the C/(I+N) given by the bandwidth Mi Mi allocation. i. subscriber.. N FB – CE1 . or mobile Mi reduces its transmission power so that the PUSCH and PUCCH C/(I+N) from it at its cell is just enough to get the selected bearer. if using 5 frequency blocks. i. the uplink bandwidth is not reduced to 4 because it does not provide any gain in terms of the bearer.e. and N FB – UL  Service   N FB – UL  N FB – CE for any pixel.3. subscriber.. i. 534 . or mobiles in the uplink. it will be assigned 5 frequency blocks as the used uplink bandwidth..Atoll 3. or with the highest effective RLC throughput. • Full Bandwidth Full channel width is used by each mobile in the uplink. or N FB – CE2 depending on the PSS ID of TXi(ic). and using 6 it would only get access to the second best.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 TX  ic  i bandwidth of the cell. a mobile is able to access the best bearer. The calculation of the gain introduced by the bandwidth reduction is explained below.e. the mobile already has the best bearer using 5 frequency blocks. CINR PUSCH PUCCH = CINR PUSCH PUCCH . • Maintain Connection The bandwidth used for transmission by a mobile is reduced only if the PUSCH and PUCCH C/(I+N) is not enough to even access the lowest bearer. subscriber. as a mobile moves from good to bad radio conditions. Number of frequency blocks in ICIC mode Cell centre Cell edge TX i  ic  No FFR N FB Time-switched FFR N FB TX i  ic  N FB TX i  ic  TX i  ic  N FB – CEx TX  ic  i Hard FFR TX  ic  i N FB – CEx TX i  ic  Soft FFR N FB Partial soft FFR TX i  ic  N FB TX i  ic  N FB – CEx TX i  ic  TX i  ic  – N FB – CEx TX  ic  TX  ic  N FB – CEx TX  ic  i i i –  N FB – CE0 + N FB – CE1 + N FB – CE2   TX i  ic  TX i  ic  TX i  ic  N FB – CEx TX i  ic  Where N FB – CEx can be N FB – CE0 . subscribers. i. The calculation of the gain introduced by the bandwidth reduction is explained below.. the number of frequency blocks used by it for transmission in uplink are reduced one by one in order to improve the PUSCH and PUCCH C/(I+N). or mobile Mi in the cell centre of the interfered TX i  ic  Mi Min cell TXi(ic). The gain related to this bandwidth reduction is applied to the PUSCH and PUCCH C/(I+N):  N TX i  ic   Mi Mi FB - CINR PUSCH PUCCH = CINR PUSCH PUCCH + 10  Log  ----------------Mi  All FB Final  N FB – UL TX i  ic  Mi Min Where N FB – UL  Service   N FB – UL  N FB – CC for any pixel. Final The pixel. and may reduce the number of used frequency blocks in order to satisfy the selected target. For example. Although using 4 frequency blocks. N FB . or mobile Mi in the cell-edge of the interfered cell TXi(ic). • Best Bearer The bandwidth used for transmission by a mobile is reduced in order to improve the PUSCH and PUCCH C/(I+N) enough to access the best bearer. The uplink bandwidth allocation may result in the use of a number of frequency blocks which is less than the number of frequency blocks associated with the channel bandwidth of the cell. As there is no reduction in the bandwidth used for transmission. Uplink Power Control Adjustment: Once the bandwidth allocation is performed. with the highest peak RLC throughput. bearer with the highest index.e.e. Bandwidth allocation is performed for all the pixels. The definition of the best bearer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). its PUSCH and PUCCH C/(I+N) will be better than when using 5. there is no gain in the PUSCH and PUCCH C/(I+N). For example. subscriber. • M HO TX i  ic  TX  ic  i TX i  ic  TX i  ic  : Handover margin defined for the cell TXi(ic). best server refers to a cell ("serving transmitter"-"reference cell" pair) that best covers a pixel. or mobile Mi. The serving cell selected for Monte Carlo simulations can also be based on the Random method instead of the Standard method. or mobile Mi at it serving cell TXi(ic). or mobile Mi and provides the best service. Input • TX i  ic  C DLRS : Downlink reference signal level received from any cell TXi(ic) at a pixel. or mobile Mi as follows: Mi Mi  Mi   TX i  ic    Mi  P Eff = Max  P Allowed –  CINR PUSCH PUCCH –  T M + M PC   P Min    B i   UL M M i i CINR PUSCH PUCCH is calculated again using P Eff . where T TX  ic  i M i B UL is the bearer selection threshold. or mobile Mi. TX i  ic  • p Layer : Priority defined for the layer assigned to for any cell TXi(ic). • T Selection : Cell selection threshold defined for the cell TXi(ic).5 Best Server Determination In LTE. subscriber. for the bearer selected for the pixel. If no serving cell is found for a mobile Mi. subscriber.3. or mobile Mi. • P Eff : Effective transmission power of the terminal used by the pixel. TX  ic  i • E DLRS : Received downlink reference signal energy per resource element (RSRP) from any cell TXi(ic) at a pixel. • N SCell Max – DL : Maximum number of downlink secondary cells defined for the terminal used by the pixel. subscriber. • B UL : Bearer assigned to the pixel. or mobile Mi. or mobile Mi in the uplink. subscriber. or mobile Mi after uplink bandwidth allocation.G Mi Mi Mi .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 If with P M i M i M i = P Allowed AND CINR PUSCH PUCCH  T TX  ic  i M i B UL + M PC . or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501. Output Mi • CINR PUSCH PUCCH : PUSCH and PUCCH C/(I+N) from a pixel. subscriber. subscriber. • O Individual : Cell individual offset defined for the cell TXi(ic). from the reception equipment assigned to the cell TXi(ic). The transmission power of Mi is reduced to determine the effective transmission power from the pixel. it is rejected for “No Coverage”. • N FB – UL : Number of frequency blocks used by the pixel. subscriber. subscriber. subscriber.Atoll 3. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501 using the terminal and service parameters ( L Mi . • T RSRP : Minimum RSRP defined for the cell TXi(ic). • Max – UL N SCell : Maximum number of uplink secondary cells defined for the terminal used by the pixel. L Ant . Calculations The serving cell selected for coverage predictions is based on the Standard serving cell selection method. and L Body ) of Mi.4. Mi Mi Mi 6. 535 . subscriber. 00080 0. The pixel..00200 0. or mobile Mi must be located within the round-trip time distances corresponding to the cells’ PRACH preamble format. • For potential serving cells that belong to layers of higher priorities.Atoll 3. • Random cell selection: a.00180 0. or mobile Mi’s service and terminal. subscriber.00020 29511 3 21024 0.00010 14521 1 21024 0. the RSRP received at the pixel. other than the preselected server ( S 0 ). subscriber. or mobile Mi’s service and terminal. Cyclic prefix Preamble sequence Cyclic prefix + preamble sequence Window size Guard period RTT distance Tsa Sec.00072 107269 4 448 0.00100 0. Sec.00013 0.00200 0. cells must fulfill the following requirements: • 536 The cells’ layers must be supported by the pixel. subscriber. b.3. Qualification: To qualify as potential servers. and fulfill the following condition: TX i  ic  TX i  ic  S0 S0 S0 E DLRS + O Individual  E DLRS + O Individual + M HO Atoll selects as the best server the cell from which the pixel. The speed defined in the pixel.00001 4096 0. or mobile Mi BSM is performed as follows: i • Standard cell selection based on 3GPP specifications for connected mode mobility: a. TX i  ic  TX i  ic  subscriber.00002 2811 The basic unit of time in LTE: Ts = 1/(15000 x 2048) seconds. and • From which the pixel. Preselection: From the list of cells that qualify as potential servers in step a. subscriber. c. subscriber.00052 77290 2 6240 0. cells must fulfill the following requirements: • • • The cells’ layers must be supported by the pixel. Sec. the cell that fulfills the following conditions is preselected as the serving cell ( S 0 ): • The cell belonging to the highest priority layer. subscriber. according to the defined best server selection criterion. PRACH preamble format a.00080 0.00300 0. .00015 0.00010 24576 0.00068 24576 0. or mobile Mi’s mobility type must be less than or equal to the maximum speed supported by the cells’ layers. Qualification: To qualify as potential servers. or mobile Mi receives the highest TX i  ic  S0 TX i  ic  S0 reference signal level or RSRP plus the cell individual offset ( C DLRS + O Individual or E DLRS + O Individual ). Metres 0 3168 0. Ts Sec.00160 0. • For the potential serving cells that belong to the layer of the lowest priority. then the preselcted server ( S 0 ) is selected as the best server.00017 0. subscriber.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 The best server selection for any pixel. If no cell fulfills the above condition.00160 0. Final selection: Among the cells that qualify as potential servers. or mobile Mi receives the highest reference signal level or RSRP ( C DLRS or TX i  ic  TX i  ic  E DLRS ) according to the defined best server selection criterion. subscriber.00228 0.00148 0.00020 49152 0. Sec. the RSRP received at the pixel.00090 0. or mobile Mi must be higher than or equal to the cells’ Min RSRP: E DLRS  T RSRP . or mobile Mi must be higher than or equal to the cells’ Min RSRP plus the cell selection threshold: TX  ic  TX  ic  TX  ic  i i i E DLRS  T RSRP + Max  0 T Selection .00068 49152 0. or mobile Mi. A list of potential primary serving cells whose cell type includes “LTE” and “LTE-A PCell” c. per layer the cell from which the pixel. subscriber. or mobile Mi is calculated as follows: 537 . respectively.3. Atoll keeps only one potential server per layer. or the numbers of downlink or uplink secondary serving cells Max – DL assigned to the user become equal to the maximum numbers defined in the terminal properties ( N SCell Max – UL N SCell and ). or mobile Mi must be higher than or equal to the cells’ Min RSRP: TX  ic  i TX  ic  i E DLRS  T RSRP . according to the defined best server selection criterion. In group-based carrier aggregation. Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell DL activation threshold defined in the terminal reception equipment properties ( T SCell ) are activated for aggregation in downlink. Secondary cells are selected based on the reference signal level or RSRP. subscriber. are activated for aggregation in uplink.. A list of potential serving cells whose cell type includes “LTE” LTE-A users: b. subscriber... subscriber. at this stage Atoll also eliminates cells belonging to other eNode-Bs than that of the selected primary cell. Here. or mobile Mi receives the highest reference signal level or RSRP. subscriber. see the Administrator Manual. Once a primary serving cell has been selected. co-channel cells are cells whose channels overlap the channel being used the primary serving cell. subscriber. and comparing it with the delta path loss threshold defined for the best server of the pixel. Similarly. subscriber. or mobile Mi’s mobility type must be less than or equal to the maximum speed supported by the cells’ layers.e. You can switch between carrier aggregation modes. Final selection: From the list of cells that qualify as potential servers in step a.Atoll 3. Atoll selects multiple servers by processing lists of potential servers according to the Standard or Random cell selection method: LTE users: a. and it is considered to be in cell centre otherwise. or mobile Mi is in the cell-edge or cell centre of TXi(ic) by calculating the difference between the path loss from the second best server and the best server. or mobile Mi must be located within the round-trip time distances corresponding to the cells’ PRACH preamble format (see table above). Atoll determines whether the pixel. subscriber.ini file. only secondary cells whose PDSCH C/(I+N) and PUSCH C/(I+N) are both higher than or equal to the DL secondary cell activation threshold defined in the terminal and cell reception equipment properties ( T SCell and UL T SCell ). The RSRP received at the pixel. BS M – 2ndBS M i i Log  r O   or BS M i mobile BS M i Mi is considered to be a cell edge BS M if i – L Total  L Path . is empty. and a primary serving cell for LTE-A users from the remaining list b. b. at this stage Atoll also eliminates cells not belonging to the carrier aggregation groups to which the selected primary cell belongs. The primary and secondary serving cells once assigned to a mobile do not change during a Monte Carlo simulation. Therefore. For more information. The second best server for a pixel. Here. This step is carried out until either list c. subscriber. Atoll selects secondary serving cells from list c. using the Atoll. The speed defined in the pixel. i.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • • • • The cells’ frequency band must be supported by the pixel. L Total is the 2ndBS total loss from Mi’s best server and L Total Mi is the total loss from Mi’s second best server calculated as explained in "Signal Level Calculation (DL)" on page 453. 2ndBS M L Total i a + 10  pixel. For carrier aggregation. For LTE-A users with a primary serving cell of type “LTE-A PCell” selected from list b. In intra-eNode-B carrier aggregation. or mobile Mi’s terminal. A list of potential secondary serving cells whose may include “LTE-A SCell DL” and “LTE-A SCell UL” Atoll selects the serving cell for LTE users from the list a. and then selects among these cells one cell as the best server at random. The pixel. Atoll eliminates the selected cell as well as any other co-channel cell from list c. 0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks 2ndBS M = TX i  ic  i C BS rO M – 2ndBS i M i ©Forsk 2015 TX  ic   TX i  ic   i = 2ndBest  C  DLRS All TX  ic   DLRS  i is the total channel overlap ratio between the best server and the second best server as calculated in "Co. It is equal to 10 for FDD frequency bands.1 Calculation of Total Cell Resources The total amount of resources in a cell is the number of modulation symbols that can be used for data transfer in each frame.6 Throughput Calculation Throughputs are calculated in two steps. otherwise. Output • BS M : Best serving cell of the pixel. 6. TX i  ic  N SD – PDCCH : Number of PDCCH symbol durations per subframe defined in TXi(ic) frame configuration or.and BS Mi Adjacent Channel Overlaps Calculation" on page 493. in the global network settings. • N Slot  SF : Number of slots per subframe (2).Atoll 3.1. • TX i  ic  N FB : Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell TXi(ic). • D CP • N SD  Slot : Number of symbol durations per slot (7 is D CP • TX  ic  i : Cyclic prefix duration defined in TXi(ic) frame configuration or. Allocated Bandwidth Throughput. and Per-user Throughput Calculation" on page 547. • N FB – SS PBCH : Number of frequency blocks that carry the SS and the PBCH (6). in the global network settings. 6 if D CP is Extended). or mobile Mi.1 "Calculation of Downlink Cell Resources" on page 538. • TX  ic  i N FB – CE1 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 1. L Path is the delta path loss threshold defined for the best server of the pixel.4. Calculation of Downlink Cell Resources Input • F : Subcarrier width (15 kHz). Calculation of uplink and downlink UE capacities as explained in "Calculation UE Capacities" on page 545. • • • Calculation of uplink and downlink total resources in a cell as explained in "Calculation of Total Cell Resources" on page 538. Cell Capacity. "Calculation of Uplink Cell Resources" on page 543.4. subscriber.4. TX i  ic  TX i  ic  is Normal. or mobile Mi. • TX i  ic  N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 0.3. otherwise.6. • W FB : Width of a frequency block (180 kHz).6. subscriber. i 6. The total cell resources can be calculated separately for the downlink and uplink as described in: • • 6. and is determined from the cell’s TDD frame configuration for TDD frequency bands. • TX i  ic  N SF – DL : Number of downlink subframes in the frame for the cell TXi(ic). • TX i  ic  N FB – CE2 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 2. Calculation of throughputs as explained in "Channel Throughput. 538 . Calculations In LTE. TX  ic  i TX  ic  i N SF – DL and N TDD – SSF are determined as follows: • TX i  ic  TX i  ic  Configuration N SF – DL N TDD – SSF FDD 10 0 DSUUU-DSUUU 2 2 DSUUD-DSUUD 4 2 DSUDD-DSUDD 6 2 DSUUU-DSUUD 3 2 DSUUU-DDDDD 6 1 DSUUD-DDDDD 7 1 DSUDD-DDDDD 8 1 TX i  ic  N Ant – TX : Number of transmission (downlink) antenna ports defined for the cell TXi(ic).Atoll 3. determined from the TDD special subframe configuration according to the 3GPP specifications as follows: Special Subframe Configuration Cyclic Prefix = Normal DwPTS GP N SD  SSF DwPTS N SD  SSF 0 3 1 Cyclic Prefix = Extended UpPTS DwPTS GP UpPTS N SD  SSF DwPTS N SD  SSF 10 3 8 9 4 8 3 2 10 3 9 2 3 11 2 10 1 4 12 1 3 7 5 3 9 8 2 6 9 3 9 1 7 10 2 8 11 1 GP N SD  SSF 1 2 GP UpPTS UpPTS N SD  SSF 1 2 The total number of modulation symbols (resource elements) in downlink is calculated as follows: TX i  ic  TX i  ic  N Sym – DL = N FB TX i  ic  TX i  ic   N Sym  SRB  N SF – DL + N Sym – DwPTS 539 . schedulers are able to perform resource allocation every subframe (2 slots). and is determined from the cell’s TDD frame configuration for TDD frequency bands.3. However.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • TX  ic  i N TDD – SSF : Number of TDD special subframes (containing DwPTS. The number of modulation symbols (resource elements) per scheduler resource block is calculated as follows: N Sym  SRB = N SCa – FB  N SD  Slot  N Slot  SF Where N SCa – FB is the number of subcarriers per frequency block calculated as follows: W FB N SCa – FB = --------F The number of modulation symbols (resource elements) corresponding to the DwPTS per scheduler resource block in the TDD special subframes is calculated as follows: DwPTS DwPTS N Sym  SSF = N SCa – FB  N SD  SSF DwPTS Where N SD  SSF is the number of DwPTS symbol durations (OFDM symbols) per special subframe. 1 frequency block by 1 subframe (2 slots) is called a scheduler resource block (SRB) in the calculations below. GP. a resource block (RB) is defined as 1 frequency block by 1 slot. and UpPTS) in the frame for the cell TXi(ic). It is equal to 0 for FDD frequency bands. e. TX  ic  i if  N Ant – TX = 4 or 8 TX  ic  i And N DLRS  DwPTS is determined from the table below: Special Subframe Configuration 0 1 2 540 Cyclic Prefix = Normal DwPTS N SD  SSF 3 9 10 TX i  ic  Cyclic Prefix = Extended TX i  ic  N Ant – TX N DLRS  DwPTS 1 2 4 8 8 DwPTS N SD  SSF TX i  ic  TX i  ic  N Ant – TX N DLRS  DwPTS 2 1 2 4 2 4 4 8 8 8 8 1 6 1 6 2 12 2 12 4 20 4 20 8 20 8 20 1 6 1 6 2 12 2 12 4 20 4 20 8 20 8 20 3 8 9 . TX  ic  i if  N Ant – TX = 1   TX  ic  i if  N Ant – TX = 2   .Atoll 3. O DMRS is the overhead corresponding to the UE-specific reference signals transmitted on the logical antenna port 5 or the demodulation reference signals transmitted using antenna ports 7 and 8 or 7 through 14. are calculated as follows: TX i  ic  TX i  ic  TX i  ic  TX i  ic  R DwPTS = N Sym – DwPTS – O DLRS  DwPTS – O PDCCH  DwPTS TX i  ic  Where O DLRS is the overhead corresponding to the downlink reference signals. and O PDCCH is the overhead corresponding to the physical TX i  ic  downlink control channel. i. O PBCH is the TX i  ic  overhead corresponding to the physical broadcast channel. These overheads are calculated as follows: Downlink reference signal overhead The downlink reference signal overhead depends on the number of transmission antenna ports: TX  ic  i TX  ic  i O DLRS = N FB TX  ic  i TX i  ic  TX i  ic  Where O DLRS  DwPTS = N FB TX  ic  i N DLRS  SRB TX  ic  i TX  ic  i  N DLRS  SRB  N SF – DL + O DLRS  DwPTS    8   =  16     24  TX i  ic  TX i  ic   N DLRS  DwPTS  N TDD – SSF . are calculated as follows: TX  ic  i TX  ic  i TX  ic  i = N Sym – DL – O DLRS – O PSS – O SSS – O PBCH – O PDCCH – O DMRS TX i  ic  The downlink DwPTS resources.. O PSS is the overhead corresponding to the TX i  ic  primary synchronisation signals.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks TX  ic  i TX  ic  i Where N Sym – DwPTS = N FB ©Forsk 2015 TX  ic  i DwPTS  N TDD – SSF  N Sym  SSF TX  ic  i The total downlink cell resources.3..e. i. R DwPTS . O SSS is the overhead corresponding to the secondary synchronisation signals. R DL TX  ic  i R DL TX  ic  i TX  ic  i . These downlink reference signal symbols are subtracted from the PDCCH overhead: TX  ic  i if  N SD – PDCCH = 0 : TX i  ic  O PDCCH = 0 541 . Therefore. therefore.Atoll 3. O PSS = 2  N FB – SS PBCH  N SCa – FB = 144 symbols O SSS = 2  N FB – SS PBCH  N SCa – FB = 144 symbols PBCH overhead The physical broadcast channel is transmitted on four symbol durations in the 1st downlink subframe over the center 6 frequency blocks. The PDCCH overlaps some downlink reference signal symbols.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Special Subframe Configuration Cyclic Prefix = Normal DwPTS N SD  SSF 3 4 5 6 7 8 11 12 3 9 10 11 TX  ic  i Cyclic Prefix = Extended TX  ic  i N Ant – TX N DLRS  DwPTS 1 6 2 12 4 20 8 20 1 8 2 16 4 24 8 24 1 2 2 4 4 8 8 8 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 DwPTS N SD  SSF 10 3 8 9 TX  ic  i TX  ic  i N Ant – TX N DLRS  DwPTS 1 8 2 16 4 24 8 24 1 2 2 4 4 8 8 8 1 6 2 12 4 20 8 20 1 6 2 12 4 20 8 20 PSS and SSS overhead The primary and secondary synchonisation signals are transmitted on 1 symbol duration each in the 1st and the 6th downlink subframes. some downlink reference signal modulation symbols are subtracted: 216 for extended cyclic prefix 240 for normal cyclic prefix PDCCH overhead The physical downlink control channel can be transmitted over up to 4 symbol durations in each subframe. The number of symbol durations for the PDCCH is defined in the global network settings. The physical broadcast channel overlaps with the downlink reference signals.3. over the centre 6 frequency blocks. N FB – CE1 . DMRS are transmitted on antenna ports 7 and 8. it is scaled down according to the ICIC mode used by the cell TXi(ic) depending on whether the downlink cell resources are being calculated for a cell-centre or cell-edge pixel. TX i  ic  Without smart antennas and MIMO: O DMRS = 0 TX i  ic  TX i  ic  With smart antennas and without MIMO: O DMRS = 12  N FB TX i  ic  TX i  ic  With smart antennas and with MIMO: O DMRS = 24  N FB TX i  ic   N SF – DL TX i  ic   N SF – DL TX i  ic  TX i  ic  TX i  ic  Without smart antennas and with SU-MIMO or MU-MIMO and N Ant – TX  4 : O DMRS = 24  N FB TX  ic  i Once R DL is known. TX i  ic  R DL TX i  ic  = R DL ICIC ABS  f Scaling  f Scaling ICIC f Scaling is calculated as follows for the different ICIC modes: ICIC ICIC cell resource scaling factor f Scaling for ICIC mode Cell centre Cell edge 1 1 Time-switched FFR 1 N FB – CEx -------------------TX i  ic  N FB Hard FFR N FB – CEx -------------------TX i  ic  N FB Soft FFR N FB – N FB – CEx -----------------------------------------TX  ic  i N FB Partial soft FFR N FB –  N FB – CE0 + N FB – CE1 + N FB – CE2 -----------------------------------------------------------------------------------------------------TX i  ic  N FB No FFR TX i  ic  TX i  ic  TX  ic  i TX i  ic  TX  ic  i TX  ic  i TX i  ic  TX  ic  i TX i  ic  N FB – CEx -------------------TX i  ic  N FB TX  ic  i TX i  ic  TX  ic  i TX  ic  i N FB – CEx -------------------TX  ic  i N FB TX i  ic  Where N FB – CEx can be N FB – CE0 . or N FB – CE2 depending on the PSS ID of TXi(ic). 542 TX i  ic   N SF – DL TX i  ic  N FB – CEx -------------------TX i  ic  N FB . or mobile.Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks TX  ic  ©Forsk 2015 TX  ic  i i if  N SD – PDCCH = 1 AND  N Ant – TX  2 : TX  ic  TX  ic  TX  ic  i i i O PDCCH =  N SD – PDCCH  N SCa – FB – 4  N FB TX  ic  i Where O PDCCH  DwPTS = TX  ic  i TX  ic  i  N SF – DL + O PDCCH  DwPTS TX  ic  TX  ic  i N i   SD – PDCCH  N SCa – FB – 4  N FB TX  ic  i  N TDD – SSF Otherwise: TX  ic  TX  ic  TX  ic  TX  ic  i i i i O PDCCH =  N SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB    TX  ic  TX i  ic  TX i  ic   N SF – DL + O PDCCH  DwPTS TX  ic  TX  ic  TX  ic  i i i i Where O PDCCH  DwPTS =  Min  2 N SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB      TX i  ic   N TDD – SSF UE-specific and demodulation reference signal overhead UE-specific reference signals are transmitted on the logical antenna port 5. subscriber. or on 7 through 14. GP. and is determined from the cell’s TDD frame configuration for TDD frequency bands. otherwise.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 ABS f Scaling is calculated as follows: • Method 1: ABS Patterns Used Only at Cell Edges ABS f Scaling •  1 Cell centre   TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic     R DL – R DwPTS + SFP SSF   R DwPTS SFP DL =       ------------------------------------------------------------------------------------------------------------------------------------------------1 1 . and UpPTS) in the frame for the cell TXi(ic). otherwise. • N Slot  SF : Number of slots per subframe (2). in the global network settings. • D CP • N SD  Slot : Number of symbol durations per slot (7 is D CP TX i  ic  : Cyclic prefix duration defined in TXi(ic) frame configuration or. 6 if D CP is Extended).3. Output • 6.Atoll 3.4. • TX  ic  i N TDD – SSF : Number of TDD special subframes (containing DwPTS. and is determined from the cell’s TDD frame configuration for TDD frequency bands.6. in the global network settings. • TX i  ic  N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 0. see "Subframe Pattern Collision Calculation" on page 497. It is equal to 10 for FDD frequency bands. • TX  ic  i N SF – UL : Number of uplink subframes in the frame for the cell TXi(ic). Calculation of Uplink Cell Resources Input • F : Subcarrier width (15 kHz). • TX i  ic  N FB – CE2 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 2. TX i  ic  TX i  ic  N SF – UL and N TDD – SSF are determined as follows: 543 . TX i  ic  TX i  ic  is Normal. It is equal to 0 for FDD frequency bands. • TX i  ic  N FB – CE1 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth used by the cell TXi(ic) and PSS ID 1. • W FB : Width of a frequency block (180 kHz). TX i  ic  • N FB – PUCCH : Average number of PUCCH frequency blocks per frame defined in TXi(ic) frame configuration or.Cell edge  TX i  ic   80  R DL    Method 2: ABS Patterns Used Throughout the Cell TX i  ic  ABS  SFPDL TX i  ic    R DL TX  ic  i – R DwPTS + TX i  ic   SFPSSF TX  ic  i   R DwPTS 1 1 f Scaling = ------------------------------------------------------------------------------------------------------------------------------------------------TX i  ic  80  R DL TX i  ic  For more information on SFP DL TX i  ic  and SFP SSF .2 TX  ic  i R DL : Amount of downlink resources in the cell TXi(ic). • N FB TX i  ic  : Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell TXi(ic).1. are calculated as follows: TX i  ic  = N Sym – UL – O ULSRS – O ULDRS TX i  ic  TX i  ic  Where O ULSRS is the overhead corresponding to the uplink sounding reference signals. These control channel overheads are calculated as follows: Calculations of uplink control channel overheads The uplink sounding reference signals are transmitted on 1 symbol duration in each uplink subframe.Atoll 3. Therefore.. and for SRS and PRACH if the UpPTS duration is 2 OFDM symbols. it is scaled down according to the ICIC mode used by the cell TXi(ic) depending on whether the uplink cell resources are being calculated for a cell-centre or cell-edge pixel.e. subscriber. or mobile. Calculations In LTE. TX i  ic  TX i  ic  N SCa – FB . a resource block (RB) is defined as 1 frequency block by 1 slot. R UL TX i  ic  R UL TX i  ic  TX i  ic  . i. TX i  ic  R UL ICIC TX i  ic  = R UL ICIC ABS  f Scaling  f Scaling f Scaling is calculated as follows for the different ICIC modes: 544 . the uplink cell capacity can be determined without considering the UpPTS symbols. Therefore. and O ULDRS is the overhead corresponding to the uplink demodulation reference signals. However. N Sym O ULDRS = 2  --------------------– UL N Sym  SRB TX i  ic  Once R UL is known. 1 frequency block by 1 subframe (2 slots) is called a scheduler resource block (SRB) in the calculations below.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 TX  ic  i TX  ic  i Configuration N SF – UL N TDD – SSF FDD 10 0 DSUUU-DSUUU 6 2 DSUUD-DSUUD 4 2 DSUDD-DSUDD 2 2 DSUUU-DSUUD 5 2 DSUUU-DDDDD 3 1 DSUUD-DDDDD 2 1 DSUDD-DDDDD 1 1 UpPTS is used for SRS (sounding reference signals) if the UpPTS duration is 1 OFDM symbol. The number of modulation symbols (resource elements) per resource block is calculated as follows: N Sym  SRB = N SCa – FB  N SD  Slot  N Slot  SF Where N SCa – FB is the number of subcarriers per frequency block calculated as follows: W FB N SCa – FB = --------F The total number of modulation symbols (resource elements) in uplink is calculated as follows: TX  ic  TX  ic  i i N Sym – UL =  N FB TX  ic  TX  ic  i i – N FB – PUCCH  N Sym  SRB  N SF – UL TX i  ic  The total uplink cell resources. schedulers are able to perform resource allocation every subframe (2 slots).3. Therefore. N Sym O ULSRS = --------------------– UL N Sym  SRB The uplink demodulation reference signals are transmitted on two symbol durations in each uplink subframe. TX i  ic  TX i  ic  N SCa – FB . 2. 545 . or N FB – CE2 depending on the PSS ID of TXi(ic). It is equal to 10 for FDD frequency bands. ABS f Scaling is calculated as follows: • Method 1: ABS Patterns Used Only at Cell Edges ABS f Scaling •  1 Cell centre  TX i  ic   =  SFP UL  1  -----------------------------. Calculation of Downlink UE Capacity Input • D Frame : Frame duration. "Calculation of Uplink UE Capacity" on page 546. see "Subframe Pattern Collision Calculation" on page 497. • N SF – DL : Number of downlink subframes in the frame for the cell TXi(ic).Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 ICIC ICIC cell resource scaling factor f Scaling for ICIC mode Cell centre Cell edge 1 1 Time-switched FFR 1 N FB – CEx -------------------TX i  ic  N FB Hard FFR N FB – CEx -------------------TX i  ic  N FB Soft FFR N FB – N FB – CEx -----------------------------------------TX i  ic  N FB Partial soft FFR N FB –  N FB – CE0 + N FB – CE1 + N FB – CE2 -----------------------------------------------------------------------------------------------------TX i  ic  N FB No FFR TX  ic  i TX i  ic  TX i  ic  TX  ic  i TX i  ic  TX i  ic  TX  ic  i TX i  ic  TX i  ic  N FB – CEx -------------------TX i  ic  N FB TX i  ic  TX  ic  i TX i  ic  N FB – CEx -------------------TX i  ic  N FB TX  ic  i TX i  ic  N FB – CEx -------------------TX i  ic  N FB TX i  ic  Where N FB – CEx can be N FB – CE0 .4.1 "Calculation of Downlink UE Capacity" on page 545. N FB – CE1 .3. 6. The UE capacities are calculated for the downlink and uplink as described in: • • 6.6. Output • TX i  ic  R UL : Amount of uplink resources in the cell TXi(ic).Cell edge 80   Method 2: ABS Patterns Used Throughout the Cell TX  ic  i  SFPUL ABS 1 f Scaling = ----------------------------80 TX i  ic  For more information on SFP UL .6. and Max – DL TX i  ic  is determined from the cell’s TDD frame configuration for TDD frequency bands.2 Calculation UE Capacities The UE category parameters define the maximum throughput that can be supported by a UE in downlink and uplink.4. • N TBB  TTI : Maximum number of transport block bits per TTI (subframe) in downlink defined for a UE category. Calculation of Uplink UE Capacity Input • D Frame : Frame duration. the maximum throughput that can be supported by a user equipment is defined through its UE category parameter Transport Block Size. The downlink UE capacity in terms of the maximum throughput supported by a UE in downlink is calculated as follows: TX  ic  Max TP UE – DL = Max – DL N TBB  TTI TX  ic  i N i   SF – DL + N TDD – SSF -------------------------------------------------- D Frame The maximum transport block sizes defined by the 3GPP for different UE categories correspond to the following maximum throughput capacities in FDD: UE Category 1 2 3 4 5 6 7 8 Max – DL 10296 51024 102048 150752 299552 301504 301504 2998560 Max 10.048 150. It is equal to 0 for FDD frequency bands. This is the maximum number of transport block bits that the UE can carry per subframe.752 299.296 51. • N SF – UL : Number of uplink subframes in the frame for the cell TXi(ic). It is equal to 10 for FDD frequency bands.6. It is equal to 0 for FDD frequency bands.3. and is determined from the cell’s TDD frame configuration for TDD frequency bands. TX i  ic  TX i  ic  N SF – UL and N TDD – SSF are determined as follows: 546 TX i  ic  TX i  ic  Configuration N SF – UL N TDD – SSF FDD 10 0 DSUUU-DSUUU 6 2 .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 TX  ic  i • N TDD – SSF : Number of TDD special subframes (containing DwPTS.024 102.504 2998.Atoll 3.504 301. and UpPTS) in the frame for the cell TXi(ic).4. • TX  ic  i N TDD – SSF : Number of TDD special subframes (containing DwPTS.552 301. and UpPTS) in the frame for the cell TXi(ic). GP.560 N TBB  TTI (bits/TTI) TP UE – DL (Mbps) Output • 6.2. and Max – UL TX  ic  i is determined from the cell’s TDD frame configuration for TDD frequency bands. GP.2 Max TP UE – DL : Maximum downlink throughput capacity of a UE category. • N TBB  TTI : Maximum number of transport block bits per TTI (subframe) in uplink defined for a UE category. and is determined from the cell’s TDD frame configuration for TDD frequency bands. TX  ic  i TX  ic  i N SF – DL and N TDD – SSF are determined as follows: TX i  ic  TX i  ic  Configuration N SF – DL N TDD – SSF FDD 10 0 DSUUU-DSUUU 2 2 DSUUD-DSUUD 4 2 DSUDD-DSUDD 6 2 DSUUU-DSUUD 3 2 DSUUU-DDDDD 6 1 DSUUD-DDDDD 7 1 DSUDD-DDDDD 8 1 Calculations In LTE. Allocated bandwidth throughputs are calculated for the number of used frequency blocks in uplink allocated to the pixel.024 51. The uplink UE capacity in terms of the maximum throughput supported by a UE in uplink is calculated as follows: Max TP UE – UL = Max – UL N TBB  TTI TX  ic  i N SF – UL  ----------------D Frame The maximum transport block sizes defined by the 3GPP for different UE categories correspond to the following maximum throughput capacities in FDD: UE Category 1 2 3 4 5 6 7 8 Max – UL 5160 25456 51024 51024 75376 51024 102048 1497760 Max 5.6. • TL UL – Max : Maximum uplink traffic load for the cell TXi(ic). or mobile Mi. 6. respectively. Cell Capacity.376 51.456 51.048 1497. • T SU – MIMO – UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(ic). Cell capacities are similar to channel throughputs but upper-bound by the maximum downlink and uplink traffic loads. subscriber. R UL i B DL i B UL TX i  ic  547 .Atoll 3.  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel. • R DL TX  ic  i TX i  ic  : Amount of downlink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on page 538. or mobile Mi. the maximum throughput that can be supported by a user equipment is defined through its UE category parameter Transport Block Size.3. subscriber. and Per-user Throughput Calculation Channel throughputs are calculated for the entire channel resources allocated to the pixel. or mobile Mi in the downlink in • "C/(I+N) and Bearer Calculation (DL)" on page 518. This is the maximum number of transport block bits that the UE can carry per subframe. or mobile Mi in the uplink in "C/ • (I+N) and Bearer Calculation (UL)" on page 532. • TX i  ic  • : Amount of uplink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on page 538. D Frame : Frame duration.760 N TBB  TTI (bits/TTI) TP UE – UL (Mbps) Output • Max TP UE – UL : Maximum uplink throughput capacity of a UE category. subscriber.024 75.  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel. Allocated Bandwidth Throughput.024 102.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  i TX  ic  i Configuration N SF – UL N TDD – SSF DSUUD-DSUUD 4 2 DSUDD-DSUDD 2 2 DSUUU-DSUUD 5 2 DSUUU-DDDDD 3 1 DSUUD-DDDDD 2 1 DSUDD-DDDDD 1 1 Calculations In LTE.4. Per-user throughputs are calculated by dividing the downlink cell capacities or uplink allocated bandwidth throughputs by the average number of downlink or uplink users defined for the cell. Input TX i  ic  • TL DL – Max : Maximum downlink traffic load for the cell TXi(ic).16 25.3 Channel Throughput. subscriber. while D Frame = 1 sec for Monte Carlo simulations. TX i  ic  For proportional fair schedulers. • T MU – MIMO – UL : MU-MIMO threshold defined in the reception equipment of the cell TXi(ic). TX i  ic  R DL Mi  B Mi TX  ic  i DL CTP P – DL = -------------------------------- G MUG – DL D Frame TX i  ic  Mi Max G MUG – DL = 1 if CINR PDSCH  CINR MUG If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph. or mobile Mi. • T SCell : Secondary cell activation threshold of the reception equipment assigned to the cell TXi(ic). • Mi N FB – UL : Number of frequency blocks used by the pixel. Mobility  M i  . DL • T SCell : Secondary cell activation threshold of the reception equipment assigned to the pixel. subscriber. • M M i i BLER  BUL : Uplink block error rate read from the BLER vs.3. • G MU – MIMO – UL : Average number of co-scheduled MU-MIMO users in uplink for the cell TXi(ic). M i • TP Offset : Throughput offset defined in the properties of the service used by the pixel. TX  ic  i • N Users – DL : Number of users connected to the cell TXi(ic) in downlink.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks M ©Forsk 2015 i • T SU – MIMO – DL : SU-MIMO threshold defined in the reception equipment of the pixel. CINR PDSCH graph available in the reception equipment TX  ic  i TX  ic  i TX i  ic  TX i  ic  TX  ic  M assigned to the terminal used by the pixel. • G MU – MIMO – DL : Average number of co-scheduled MU-MIMO users in downlink for the cell TXi(ic). the channel throughput is increased by the multi-user diversity gain G MUG – DL read Mi from the scheduler properties for the bearer B DL . the actual value of D Frame is used to calculate the channel throughput for coverage predictions. subscriber. defined in the frequency bands table. CINR PUSCH PUCCH graph available in the reception   equipment assigned to the cell TXi(ic). and the number of users connected to the cell in downlink. subscriber.Atoll 3. or mobile Mi. • N FB TX i  ic  : Number of frequency blocks. or mobile UL Mi Mi. subscriber. • i i BLER  BDL : Downlink block error rate read from the BLER vs. for the channel bandwidth used by the cell TXi(ic). subscriber. • T MU – MIMO – DL : MU-MIMO threshold defined in the reception equipment of the pixel. or mobile Mi. subscriber. or mobile Mi. • f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the pixel. MIMO – SU-MIMO Gain: 548 . or mobile Mi. or mobile Mi after uplink bandwidth allocation as calculated in "C/(I+N) and Bearer Calculation (UL)" on page 532. it is interpolated from the gain values available for the numbers of users just less than and just greater than the actual number of users. • N Users – UL : Number of users connected to the cell TXi(ic) in uplink. TX i  ic  Calculations Downlink: TX  ic  i • Mi R DL  Mi B DL Peak RLC Channel Throughput: CTP P – DL = --------------------------------D Frame In the above formula. subscriber. or mobile Mi is located.– TP Offset = CTP E – DL  -----------------------100 Mi Mi • Application Channel Throughput: CTP A – DL • Peak RLC Cell Capacity: Cap P – DL = CTP P – DL  TL DL – Max • i i i Effective RLC Cell Capacity: Cap E – DL = Cap P – DL   1 – BLER  B DL     • Mi TX i  ic  Mi M M Mi Application RLC Capacity: Cap A – DL M Mi Mi f TP – Scaling . it is interpolated from the gain values available for the C/(I+N) just less than and just greater than the actual C/(I+N). The gain is read from the TX  ic  i M i properties of the reception equipment assigned to the pixel.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 If the cell supports MIMO. or mobile Mi for N Ant – TX . subscriber.  Mi =  B DL Max – M i Mi B DL   1 + f SU – MIMO  G SU – MIMO – DL – 1     If the Max SU-MIMO Gain for the exact value of the C/(I+N) is not available in the table.– TP Offset = Cap E – DL  -----------------------100 Mi Mi • • • Mi Cap P – DL Peak RLC Throughput per User: PUTP P – DL = ----------------------TX i  ic  N Users – DL Effective RLC Throughput per User: Application Throughput per User: M i PUTP E – DL M i PUTP A – DL Mi Cap E – DL = ----------------------TX i  ic  N Users – DL = M i PUTP E – DL Mi M f TP – Scaling i . M i Mobility  M i  . BLER  B DL . which is the average number of co-scheduled users. the MU-MIMO gain.3. Atoll also takes into account the SU-MIMO Gain Factor f SU – MIMO defined for the clutter class where the pixel. Mi Mi TX i  ic  CTP P – DL = CTP P – DL  G MU – MIMO – DL • M M M i i i Effective RLC Channel Throughput: CTP E – DL = CTP P – DL   1 – BLER  B DL     Mi Mi f TP – Scaling . subscriber. and according to the Mi diversity mode depending on the MU-MIMO threshold and TX i  ic  criterion. MIMO – MU-MIMO Gain (for throughput coverage predictions): If the cell supports MU-MIMO. G SU – MIMO – DL . and according to the Mi diversity mode depending on the SU-MIMO threshold and criterion. N Ant – RX . corresponding to the bearer is applied to its efficiency. G MU – MIMO – DL .Atoll 3.– TP Offset  -----------------------100 549 . Max – M i the SU-MIMO gain. is applied to the channel throughput. 3. Mobility  M i  . the actual value of D Frame is used to calculate the channel throughput for coverage predictions.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Carrier Aggregation: Aggregated throughputs are calculated by summing the throughputs from each serving cell taking part in carrier aggregation for any LTE-A pixel. it is interpolated from the gain values available for the C/(I+N) just less than and just greater than the actual C/(I+N).Atoll 3. and according to the Mi diversity mode depending on the SU-MIMO threshold and criterion. Atoll also takes into account the SU-MIMO Gain Factor f SU – MIMO defined for the clutter class where the pixel. Uplink: TX i  ic  • Peak RLC Channel Throughput: M i CTP P – UL R UL  Mi B UL = --------------------------------D Frame In the above formula. TX i  ic  For proportional fair schedulers. subscriber. If the sum of the throughputs exceeds the maximum throughput supported by the UE category. BLER  B UL . MIMO – SU-MIMO Gain: If the cell supports MIMO. N Ant – TX . G SU – MIMO – UL . The gain is read from the M TX  ic  M i i i properties of the reception equipment assigned to the TXi(ic) for N Ant – RX . corresponding to the bearer is applied to its efficiency. it is interpolated from the gain values available for the numbers of users just less than and just greater than the actual number of users. the channel throughput is increased by the multi-user diversity gain G MUG – UL read Mi from the scheduler properties for the bearer B UL . and the number of users connected to the cell in uplink. the aggregated throughput is scaled down by the following ratio:  Mi  Max CTP P – DL Min  TP UE – DL   TX i  ic  r = --------------------------------------------------------------------------Mi CTP P – DL   TX i  ic  Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell DL activation threshold ( T SCell ) defined in the terminal reception equipment properties are activated for aggregation.  Mi B UL =  Max – TX i  ic  Mi B UL   1 + f SU – MIMO  G SU – MIMO – UL – 1     If the Max SU-MIMO Gain for the exact value of the C/(I+N) is not available in the table. TX i  ic  R UL Mi  B Mi TX  ic  i UL CTP P – UL = -------------------------------- G MUG – UL D Frame TX i  ic  Mi Max G MUG – UL = 1 if CINR PUSCH PUCCH  CINR MUG If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph. subscriber. or mobile Mi is located. Mobility  M i  . Max – TX i  ic  the SU-MIMO gain. 550 . while D Frame = 1 sec for Monte Carlo simulations. or mobile. 3.– TPOffset = ABTP E – UL  -----------------------100 Mi • Application Allocated Bandwidth Throughput: ABTPA – UL •  Cap M i  M M i P – UL - ABTPP –i UL Peak RLC Throughput per User: PUTP P – UL = Min  ----------------------TX  ic    i  N Users – UL  •  Cap Mi  Mi M E – UL - ABTP E –i UL Effective RLC Throughput per User: PUTP E – UL = Min  ----------------------TX  ic    i  N Users – UL  • Mi Application Throughput per User: PUTP A – UL Mi Mi f TP – Scaling .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 MIMO – MU-MIMO Gain (for throughput coverage predictions): If the cell supports MU-MIMO.– TP Offset = PUTP E – UL  -----------------------100 Mi Carrier Aggregation: Aggregated throughputs are calculated by summing the throughputs from each serving cell taking part in carrier aggregation for any LTE-A pixel. If the sum of the throughputs exceeds the maximum throughput supported by the UE category. subscriber.– TP Offset = CTP E – UL  -----------------------100 Mi • Application Channel Throughput: CTP A – UL • Peak RLC Cell Capacity: Cap P – UL = CTP P – UL  TL UL – Max • i i i Effective RLC Cell Capacity: Cap E – UL = Cap P – UL   1 – BLER  B UL     • M M i M Mi Application Cell Capacity: Cap A – UL TX  ic  i i M M Mi Mi f TP – Scaling . which is the average number of co-scheduled users. 551 . the MU-MIMO gain. is applied to the channel throughput. M i M i TX  ic  i CTP P – UL = CTP P – UL  G MU – MIMO – UL • M M M i i i Effective RLC Channel Throughput: CTP E – UL = CTP P – UL   1 – BLER  B UL     Mi Mi Mi f TP – Scaling . the aggregated throughput is scaled down by the following ratio:  Mi  Max Min  TP UE – UL CTP P – UL     TX i  ic  r = --------------------------------------------------------------------------Mi CTP P – UL   TX i  ic  Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell DL activation threshold ( T SCell ) defined in the terminal reception equipment properties and UL PUSCH C/(I+N) is higher than or equal to the secondary cell activation threshold ( T SCell ) defined in the cell reception equipment properties are activated for aggregation. and according to the Mi diversity mode depending on the MU-MIMO threshold and TX  ic  i criterion. or mobile. G MU – MIMO – UL .– TP Offset = Cap E – UL  -----------------------100 Mi M • • i Mi Mi N FB – UL Peak RLC Allocated Bandwidth Throughput: ABTP P – UL = CTP P – UL  ----------------TX  ic  i N FB M M M i i i Effective RLC Allocated Bandwidth Throughput: ABTP E – UL = ABTP P – UL   1 – BLER  B UL     Mi Mi Mi f TP – Scaling .Atoll 3. • p QCI : QCI priority of the service accessed by a mobile Mi. subscriber. subscriber. subscriber. • ABTP A – UL : Uplink application allocated bandwidth throughput at the pixel. • Cap A – DL : Downlink application cell capacity at the pixel. or mobile Mi. • CTP A – DL : Downlink application channel throughput at the pixel. subscriber. or mobile Mi. subscriber. subscriber. • CTP A – UL : Uplink application channel throughput at the pixel. TX i  ic  TX i  ic  Mi Mi .3. • PUTP P – DL : Downlink peak RLC throughput per user at the pixel. subscriber. • PUTP E – UL : Uplink effective RLC throughput per user at the pixel. • CTP E – UL : Uplink effective RLC channel throughput at the pixel. These resource allocation algorithms are explained in "Scheduling and Radio Resource Allocation" on page 552 and the calculation of user throughputs is explained in "User Throughput Calculation" on page 561. subscriber. or mobile Mi.7 Scheduling and Radio Resource Management Atoll LTE module includes a number of scheduling methods which can be used for scheduling and radio resource allocation during Monte Carlo simulations. • Cap P – DL : Downlink peak RLC cell capacity at the pixel. • ABTP P – UL : Uplink peak RLC allocated bandwidth throughput at the pixel. • Cap A – UL : Uplink application cell capacity at the pixel. subscriber. or mobile Mi. or mobile Mi. • Cap P – UL : Uplink peak RLC cell capacity at the pixel. subscriber. subscriber. subscriber. subscriber. or mobile Mi. • PUTP A – DL : Downlink application throughput per user at the pixel. • PUTP A – UL : Uplink application throughput per user at the pixel. • CTP P – UL : Uplink peak RLC channel throughput at the pixel. M i Mi M i Mi Mi Mi Mi M i Mi Mi Mi Mi M i Mi Mi Mi Mi Mi Mi Mi 6. • TL UL – Max : Maximum uplink traffic load for the cell TXi(ic). • Cap E – UL : Uplink effective RLC cell capacity at the pixel. or mobile Mi. or mobile Mi.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Output M i • CTP P – DL : Downlink peak RLC channel throughput at the pixel. subscriber. subscriber. subscriber. or mobile Mi. subscriber. or mobile Mi. subscriber. • Cap E – DL : Downlink effective RLC cell capacity at the pixel. or mobile Mi. subscriber.1 Scheduling and Radio Resource Allocation Input 552 TX  ic  i • TL DL – Max : Maximum downlink traffic load for the cell TXi(ic). • PUTP E – DL : Downlink effective RLC throughput per user at the pixel. 6.7. • p Service : User-defined priority of the service accessed by a mobile Mi. or mobile Mi.4. • PUTP P – UL : Uplink peak RLC throughput per user at the pixel. • ABTP E – UL : Uplink effective RLC allocated bandwidth throughput at the pixel. • N Users – Max : Maximum number of users defined for the cell TXi(ic). or mobile Mi. or mobile Mi. or mobile Mi. • CTP E – DL : Downlink effective RLC channel throughput at the pixel.4. or mobile Mi. or mobile Mi. subscriber. or mobile Mi. or mobile Mi. or mobile Mi.Atoll 3. subscriber. or mobile Mi. • T SCell : Secondary cell activation threshold of the reception equipment assigned to the cell TXi(ic).0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 M i • TPD Min – DL : Downlink minimum throughput demand for the service accessed by a mobile Mi. • TPD Min – UL : Uplink minimum throughput demand for the service accessed by a mobile Mi.  M : Bearer efficiency (bits/symbol) of the bearer assigned to the mobile Mi in the uplink in "C/(I+N) and Bearer • Calculation (UL)" on page 532. CINR PUSCH PUCCH graph available in the reception   equipment assigned to the cell TXi(ic).3. • TP Offset : Throughput offset defined in the properties of the service used by the mobile Mi. • TPD Max – UL : Uplink maximum throughput demand for the service accessed by a mobile Mi. • CTP P – DL : Downlink peak RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on UL Mi M i Mi page 538. •  • Calculation (DL)" on page 518. • TPD Max – DL : Downlink maximum throughput demand for the service accessed by a mobile Mi. • Mi CTP E – DL : Downlink effective RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on page 538.Atoll 3.  M M M i i Mi TX i  ic  TX i  ic  Mi B DL : Bearer efficiency (bits/symbol) of the bearer assigned to the mobile Mi in the downlink in "C/(I+N) and Bearer i B UL i B DL – Highest i B UL – Highest : Bearer efficiency (bits/symbol) of the highest bearer of the service being used by the mobile Mi in the uplink. • f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile Mi. DL • T SCell : Secondary cell activation threshold of the reception equipment assigned to the mobile Mi. • G MU – MIMO – UL : Average number of co-scheduled MU-MIMO users in uplink for the cell TXi(ic). • Mi ABTP P – UL : Uplink peak RLC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput Calculation" on page 538. CINR PDSCH graph available in the reception equipment   assigned to the terminal used by the mobile Mi. 553 . : Bearer efficiency (bits/symbol) of the highest bearer of the service being used by the mobile Mi in the  M • downlink. • Mi CTP E – UL : Uplink effective RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on page 538. • Mi ABTP E – UL : Uplink effective RLC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput Calculation" on page 538. i i BLER  B UL : Uplink block error rate read from the BLER vs. • Max TP UE – DL : Maximum downlink throughput capacity of the UE category of the mobile Mi as calculated in "Calculation of Downlink UE Capacity" on page 545. • • M TX  ic  M M i i BLER  B DL : Downlink block error rate read from the BLER vs. • G MU – MIMO – DL : Average number of co-scheduled MU-MIMO users in downlink for the cell TXi(ic). • Mi CTP P – UL : Uplink peak RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on page 538. • Max TP UE – UL : Maximum uplink throughput capacity of the UE category of the mobile Mi as calculated in "Calculation of Uplink UE Capacity" on page 546. Mobile Selection: TX  ic  i The scheduler selects N Users mobiles for the scheduling and RRM process. TPD Max – DL = --------------------------------------------Sel Sel   Mi     Mi    1 – BLER  B DL    1 – BLER  BDL         Sel Sel Mi Sel i TPD Min – UL M Sel Mi Uplink: TPD Min – UL = ---------------------------------------------. Calculation of Actual Minimum and Maximum Throughput Demands: If the service maximum throughput demand downgrading is active (for more information. Sel i TPD Max – DL = M Sel Mi Mi TPD Max – DL + TP Offset ----------------------------------------------------------------------------Sel  Mi   Mi   1 – BLER  B DL    f TP – Scaling    . TX  ic  TX  ic  TX  ic  i i i N Users = Min  N Users – Max N Users – Generated   Sel For a cell.. the maximum throughput demand of each user will be downgraded as follows: Downlink: Sel Mi TPD Max – DL Sel Mi Uplink: TPD Max – UL  Sel   Mi Sel Sel   Mi Mi B DL  = Max TPD Min – DL TPD Max – DL  ------------------------------    Sel   Mi B DL – Highest   Sel   Mi Sel Sel   Mi Mi B UL  = Max TPD Min – UL TPD Max – UL  -------------------------------    Sel   Mi B UL – Highest  Then. the actual minimum and maximum throughput demands can be considered as the peak RLC. TPD Max – UL Sel   Mi    1 – BLER  B UL      • Target Throughput = Application Throughput Downlink: Sel Mi Mi Sel TPD i Min – DL + TP Offset TPD Min – DL = -----------------------------------------------------------------------------Sel  Mi   Mi  M  1 – BLER  B DL    f TP – Scaling    554 Mi Mi   Min  TPD Max – UL ABTP P – UL   = ------------------------------------------------------------------------Sel   Mi    1 – BLER  B UL      . the scheduler keeps all the mobiles generated for the cell TXi(ic). effective RLC. mobiles M i TX  ic  i  N Users are selected for RRM by the scheduler.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Calculations The following calculations are described for any cell TXi(ic) containing the users Mi for which it is the best server. or application throughput. Therefore: • Target Throughput = Peak RLC Throughput M Sel i M Sel i Downlink: TPD Min – DL .3. Min  TPD Max – UL ABTP P – UL   • Target Throughput = Effective RLC Throughput Sel Mi Downlink: TPD Min – DL Sel Mi Sel Mi Sel Mi TPD Min – DL TPD Max – DL = --------------------------------------------. see the Administrator Manual)..Atoll 3. depending on the selected target throughput of the scheduler assigned to the cell TXi(ic). If the Monte Carlo user distribution has generated TX i  ic  a number of users which is less than N Users – Max . TPD Max – DL Sel Sel Mi Mi Mi   Uplink: TPD Min – UL . The mobiles are sorted first in the order of decreasing QCI priority (as listed in the table below) and then in the order of decreasing user-defined service priority within a QCI. For example: QoS class identifier 1 2 3 4 5 6 7 8 9 QCI priority 2 4 3 5 1 6 7 8 9 Sel Mi 1 Sel Mi p QCI p Service 1 i 2 : 3 : Sel Mi 0 2 i : : : 0 : 3 i : : : 0 : 4 i : : : 0 : 5 : : : : i 0 6 i : : : 0 : 7 i : : : 0 : 8 i 555 . This calculation is performed in order to limit the maximum uplink throughput demand to the maximum throughput that a user can get in uplink using the allocated bandwidth (number of frequency blocks) calculated for it in "C/(I+N) and Bearer Calculation (UL)" on page 532.Atoll 3. LTE-A users are only scheduled on their primary serving cells. Resource Allocation for Minimum Throughput Demands: • Sel 1.3. Sel Mi TX i  ic  Sel Mi  N Users in order of decreasing effective service priority (combination of p QCI and p Service ).  Sel Mi Mi  Mi  Min  TPD Max – UL ABTP P – UL + TP Offset Sel Mi   TPD Max–UL = -------------------------------------------------------------------------------------------------Sel  Mi   Mi   1 – BLER  B UL    f TP – Scaling    The Min() function selects the lower of the two values.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Uplink: Sel M M Sel i i TPD i Min – UL + TP Offset TPD Min – UL = -----------------------------------------------------------------------------Sel   Mi   Mi  1 – BLER  BUL    f TP – Scaling M   . Atoll sorts the M i For their minimum throughput demands. . If an active DL+UL mobile is only able to get its minimum throughput demand in one direction. Starting with M i M p Service NULL i : : N 0 TX  ic  i = N Users . • When/If in uplink  M Sel Mi TX i  ic  R Min – UL = TL UL – Max . 7. i..or 5. Atoll allocates the downlink and uplink resources required to satisfy each user’s minimum throughput demands in downlink and uplink as follows: Sel Mi Sel Mi R Min – DL Sel Mi Sel Mi TPD Min – DL TPD Min – UL = -------------------------. If  Sel Mi TX i  ic  R Min – DL  TL DL – Max or Sel Mi  Sel Mi TX i  ic  R Min – UL  TL UL – Max . Mobiles which are active DL+UL must be able to get their minimum throughput demands in both UL and DL in order to be considered connected DL+UL. 4. R Min – DL  -------------------Sel Mi CTP P – DL Sel Max Mi TP UE – UL . and all the minimum throughput resources demanded by Sel Mi the mobiles have been allocated.. respectively. R Min – UL  -------------------Sel Mi CTP P – UL 6.3.Atoll 3. Sel Mi Max TP UE – DL . Resource Allocation for Maximum Throughput Demands: For each cell.. Atoll goes to the next step for allocating resources to satisfy the maximum throughput demands. are rejected due to No Service. starting from the mobile with the lowest priority service. i. it is rejected.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Sel M Mi Sel i p QCI : : 0 9 : 0 : Sel = 1 up to M i i : : Sel Sel i : : 2. Mobiles which are active UL and whose minimum throughput demand in UL is higher than the uplink allocated Sel Mi Sel Mi bandwidth throughput ( TPD Min – UL  ABTP P – UL ) are rejected due to Resource Saturation.and R Min – UL = -------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL 3. • When/If in downlink Sel Mi  TX i  ic  R Min – DL = TL DL – Max .e. the resources available in downlink have been used up for Sel Mi satisfying the minimum throughput demands of the mobiles. among all the cells of the site in order to reach a downlink or uplink effective RLC aggregate site throughput ≤ the site’s maximum downlink or uplink S1 interface throughput.e. mobiles are rejected one by one due to Backhaul Saturation.e. Mobiles with minimum throughput demands higher than their UE capacities. Atoll stops the resource allocation in downlink or uplink. the remaining cell resources available are: 556 . Backhaul Saturation: If at this stage. a site’s downlink or uplink effective RLC aggregate throughput exceeds its maximum downlink or uplink S1 interface throughput. and the resources that were allocated to it in the one direction in which it was able to get a throughput are allocated to other mobiles. i. the resources available in uplink have been used up for Sel i satisfying the minimum throughput demands of the mobiles. Atoll 3. the following resource allocation methods are available: • Proportional Fair: The goal of this scheduling method is to distribute resources among users fairly in such a way that. Within each serving cell. are activated for aggregation in uplink. primary and Sel Mi Sel Mi secondary. see the Administrator Manual. resource allocation for the maximum throughput demands is carried out according to the scheduler used by that particular cell. Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell activation threshold defined in the DL terminal reception equipment properties ( T SCell ) are activated for aggregation in downlink. For the remaining throughput demands of the mobiles.3. only secondary cells whose PDSCH C/(I+N) and PUSCH C/(I+N) are both higher than or equal to the secondary cell activation threshold defined in DL UL the terminal and cell reception equipment properties ( T SCell and T SCell ). 557 . on the average. each user gets the highest possible throughput that it can get under the radio conditions at its location. LTE-A users are scheduled separately on each of their serving cells. Similarly. the remaining throughput demands are either the maximum UE capacities or the difference between the maximum and the minimum throughput demands. whichever is smaller: Sel Sel Sel M M M  i i i Max  Downlink: TPD Rem – DL = Min  TPD Max – DL – TPD Min – DL TP UE – DL   Sel Sel Sel Mi Mi Mi  Max  Uplink: TPD Rem – UL = Min  TPDMax – UL – TPD Min – UL TP UE – UL   For their maximum throughput demands.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 TX  ic  i TX  ic  i  Downlink: R Rem – DL = TL DL – Max – M TX  ic  i TX  ic  i Uplink: R Rem – UL = TL UL – Max –  M M Sel i R Min – DL Sel i M Sel i R Min – UL Sel i For each mobile.ini file to have each user’s remaining throughput demand distributed over each of its serving cells proportionally only to the resources available on each serving cell: Sel Mi Downlink: TPD Rem – DL TX i  ic  Sel Mi Server n R Rem – DL Server n = TPD Rem – DL  ----------------------------------------------------5 TX i  ic    RRem – DL  Server n n=1 Sel Mi Uplink: TPD Rem – UL TX i  ic  Sel Mi Server n R Rem – UL Server n = TPD Rem – UL  ----------------------------------------------------5 TX  ic  i   RRem – UL  Server n n=1 For more information. respectively. Each user’s remaining throughput demand ( TPD Rem – DL and TPD Rem – UL ) is distributed over each of its serving cells proportionally to the resources available on each serving cell and to the user’s downlink effective RLC channel throughput or uplink effective RLC allocated bandwidth throughput on each of its serving cell: Downlink: Sel i TPD Rem – DL Server n M = TX i  ic  Mi R Rem – DL  CTP E – DL Server n Server n  -------------------------------------------------------------------------------------------------5 Sel i TPD Rem – DL M TX i  ic    RRem – DL Mi Server n  CTP E – DL  Server n n=1 Sel Mi Uplink: TPD Rem – UL TX i  ic  Server n Mi R Rem – UL  ABTP E – UL Server n Server n = TPD Rem – UL  -----------------------------------------------------------------------------------------------------5 Sel Mi TX i  ic    RRem – UL Mi Server n  ABTP E – UL  Server n n=1 You can add an option in the Atoll. Mobility  M i  assigned to mobile .and RD Rem – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the peak channel throughputs at the user’s location.. • When/If in downlink Sel Mi  TX i  ic  R Max – DL = R Rem – DL . Sel Mi Sel Mi CTP P – DL = CTP P – DL Sel Mi TX i  ic  Without MUG Sel Mi TX i  ic  G MUG – DL = 1 if CINR PDSCH  Sel Mi  G MUG – DL and CTP P – UL = CTP P – UL  G MUG – UL Sel Mi TX i  ic  Max CINR MUG TX i  ic  Without MUG Max and G MUG – UL = 1 if CINR PUSCH PUCCH  CINR MUG . Atoll recalculates the remaining resources as follows: TX  ic  i TX  ic  i R Rem – DL = TL DL – Max –  M Sel i R Min – DL – Sel Mi TX i  ic  TX i  ic  R Rem – UL = TL UL – Max –  M 558 Sel i  M Sel i R Max – DL and Sel Mi Sel Mi R Min – UL –  M Sel i Sel Mi R Max – UL . in the cell TXi(ic) in the iteration k-1. the resources available in uplink have been used up for Sel Mi satisfying the maximum throughput demands of the mobiles. f.and -------------------N N c. i. i. b. d.e. TX  ic  i TX  ic  i a. DL or UL. g. Atoll stops the resource allocation in downlink or uplink.Atoll 3. this user is removed from the list of remaining users. and the number of connected users. Atoll divides the remaining resources in the cell into equal parts for each user: TX i  ic  TX i  ic  R Rem – DL R Rem – UL --------------------. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands: Sel Mi RD Rem – DL Sel Mi Sel Mi Sel Mi TPD Rem – DL TPD Rem – UL = --------------------------.. e. The resources allocated to each user by the Proportional Fair scheduling method for satisfying its maximum throughput demands are: Sel TX i  ic  Sel Sel Sel TX i  ic  M M  Mi  Mi R Rem – DL R Rem – UL i i - and R Max R Max – DL = Min  RD Rem – DL -------------------– UL = Min  RD Rem – UL --------------------- N  N    Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from the remaining resources of the cell.e.3. whichever is smaller. it is interpolated from the gain values available for the numbers of users just less than and just greater than the actual number of users. Each user’s channel throughput is increased by the multi-user diversity gain G MUG – DL or G MUG – UL read from the M Sel i scheduler properties for the downlink or uplink bearer ( B DL Sel Mi M Sel i Sel or B UL ). If the resources allocated to a user satisfy its maximum throughput demands.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Sel Let the total number of users be N  M i . • When/If in uplink Sel Mi  TX i  ic  R Max – UL = R Rem – UL . the resources available in downlink have been used up Sel Mi for satisfying the maximum throughput demands of the mobiles. If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph. • When/If in uplink  Sel Mi TX i  ic  R Max – UL = R Rem – UL . Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied TX  ic  i TX  ic  i until either R Rem – DL = 0 and R Rem – UL = 0 . a. the resources available in downlink have been used up Sel i for satisfying the maximum throughput demands of the mobiles.e. whichever is smaller.and RD Rem – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the peak channel throughputs at the user’s location. Atoll stops the resource allocation in downlink or uplink. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands: Sel Mi M RD Rem – DL Sel i M Sel Sel i Mi TPD Rem – DL TPD Rem – UL = --------------------------. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied TX i  ic  TX i  ic  until either R Rem – DL = 0 and R Rem – UL = 0 . the user throughputs for users with high throughput demands will be higher than those with low 559 . Therefore. this user is removed from the list of remaining users. d. Atoll divides the remaining resources in the cell into equal parts for each user: TX i  ic  TX i  ic  R Rem – DL R Rem – UL --------------------. i. c.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 h. i.e. Atoll recalculates the remaining resources as follows: TX i  ic  TX i  ic  R Rem – DL = TL DL – Max –  Sel Mi R Min – DL – Sel Mi TX i  ic  TX i  ic  R Rem – UL = TL UL – Max –   Sel Mi R Max – DL and Sel Mi Sel Mi R Min – UL – Sel Mi  Sel Mi R Max – UL Sel Mi g. If the resources allocated to a user satisfy its maximum throughput demands. or all the maximum throughput demands are satisfied. • Round Robin: The goal of this scheduling method is to allocate equal resources to users fairly.Atoll 3. • When/If in downlink Sel Mi  M TX i  ic  R Max – DL = R Rem – DL .. e. or all the maximum throughput demands are satisfied.and -------------------N N b.. • Proportional Demand: The goal of this scheduling method is to allocate resources to users weighted according to their remaining throughput demands. f. The resources allocated to each user by the Round Robin scheduling method for satisfying its maximum throughput demands are: Sel Mi R Max – DL TX i  ic  Sel Sel Sel TX i  ic  Mi  Mi  Mi R Rem – DL R Rem – UL - and R Max = Min  RD Rem – DL -------------------– UL = Min  RD Rem – UL --------------------- N N     Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from the remaining resources of the cell. the resources available in uplink have been used up for Sel Mi satisfying the maximum throughput demands of the mobiles. Sel Let the total number of users be N  M i . The resources allocated to each user by the Proportional Demand scheduling method for satisfying its maximum throughput demands are: Sel Mi R Max – DL • Sel Mi TX i  ic  Sel Mi Sel Mi TX i  ic  RD Rem – DL RD Rem – UL . Atoll calculates the amount of effective remaining resources of the cell to distribute among the users as follows:     Sel Sel TX i  ic  Mi TX i  ic  Mi  TXi  ic    TX i  ic   R Eff – Rem – DL = Min  R Rem – DL RD Rem – DL and R Eff – Rem – UL = Min  R Rem – UL RD Rem – UL     Sel Sel     Mi Mi   c. Spatial Multiplexing with Multi-User MIMO: MU-MIMO lets the system/scheduler work with parallel LTE frames. if the calculated value for the MU-MIMO criterion is higher TX i  ic  TX i  ic  than the MU-MIMO threshold T MU – MIMO – DL or T MU – MIMO – UL . Atoll stops the resource allocation in downlink or uplink.e. Atoll sorts the M i TX i  ic   N Users in order of decreasing PDSCH.and RD Rem – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the peak channel throughputs at the user’s location. In both downlink and uplink. the actual . b. However. more mobiles can therefore be allocated resources in the same frame.. depending on whether the allocation is being performed for the downlink or for the uplink. this scheduler distributes channel throughput between users proportionally to their demands. Atoll allocates the downlink and uplink resources required to satisfy each user’s remaining throughput demands in downlink and uplink as follows: Sel i R Max – DL M M Sel i M Sel Sel i M TPD Rem – DL TPD Rem – UL i = --------------------------. In other words. i.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 throughput demands. each mobile resource consumptions of a mobile 560 MU – MIMO Mi MU – MIMO Mi has a corresponding traffic load TL are given by: MU – MIMO Mi . i. the resources available in downlink have been used up Sel Mi for satisfying the maximum throughput demands of the mobiles.Atoll 3. the resources available in uplink have been used up for Sel Mi satisfying the maximum throughput demands of the mobiles. and therefore require less amount of resources. Therefore. and the end-throughput for each cell will be the highest compared to other types of schedulers. This is done by allocating as much resources as needed to mobiles with high C/(I+N) conditions. • When/If in uplink  Sel Mi TX i  ic  R Max – UL = R Rem – UL . Sel a. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands: Sel i RD Rem – DL M M Sel i M Sel Sel i M TPD Rem – DL TPD Rem – UL i = --------------------------. •  When/If in downlink M Sel i TX  ic  i R Max – DL = R Rem – DL .and R Max – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL c. a. MU-MIMO can be used if the cell supports MU-MIMO. As mobiles with high C/(I+N) can get higher bearers. many users can be co-scheduled on the same resources.3. or PUSCH and PUCCH C/(I+N).. b.and R Max = R Eff – Rem – DL  ---------------------------------– UL = R Eff – Rem – UL  ---------------------------------Sel Sel Mi Mi  RDRem – DL  RDRem – UL Sel Mi Sel Mi Max C/I: The goal of this scheduling method is to achieve the maximum aggregate throughput for the cells. Starting with the mobile with the highest rank. and the number of antenna ports is equal to 2 or more.e. 2 User Throughput Calculation User throughputs are calculated for the percentage of resources allocated to each mobile selected by the scheduling for RRM Sel during the Monte Carlo simulations. Sel Mi Sel = R UL : Uplink traffic load or the amount of uplink resources allocated to the mobile M i . 6.7.and RC UL TX  ic  i G MU – MIMO – DL Saturation occurs when  M MU – MIMO – UL i RC UL M MU – MIMO i TL UL = -----------------------------------TX  ic  i G MU – MIMO – UL TX  ic  i  = TL UL – Max or M MU – MIMO – DL i RC DL TX  ic  i = TL DL – Max . Carrier Aggregation: Aggregated throughputs are calculated by summing the throughputs from each serving cell taking part in carrier aggregation for any LTE-A mobile.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 MU – MIMO i RC DL M M MU – MIMO i MU – MIMO M TL DL i = ----------------------------------. limited by the maximum throughput supported by the UE category. 561 . Backhaul Capacity Limitation: Backhaul overflow ratios are calculated for each site as follows: Sel Sel   Mi   Mi    R Max – DL  CTP E – DL     Sel   M i  Site = Max  1 ------------------------------------------------------------------------------------------------------ and Sel Sel Mi    Mi  Site TP – R  CTP   S1 – DL Min – DL E – DL     Sel   M i  Site  Site BHOFDL  Sel Sel   Mi   Mi    R Max – UL  CTP E – UL     Sel   M  Site i  = Max  1 ------------------------------------------------------------------------------------------------------ Sel Sel Mi    Mi  Site  R Min – UL  CTP E – UL   TP S1 – UL –    Sel   M i  Site  Site BHOFUL  Total Amount of Resources Assigned to Each Selected Mobile: Sel Atoll calculates the amounts of downlink and uplink resources allocated to each individual mobile M i (which can also be referred to as the traffic loads of the mobiles) as follows: Sel Sel Mi Downlink: TL DL Sel Mi = R DL M Sel i M Sel i R  Mi   Mi Max – DL  CTP P – DL  R Min – DL  CTP P – DL +  -----------------------------------------------Site     BHOF DL  = -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – DL Sel Sel Mi Uplink: TL UL Sel Mi = R UL Sel Mi Sel Mi R  Mi   Mi Max – UL  CTP P – UL  R Min – UL  CTP P – UL +  -----------------------------------------------Site     BHOF UL  = -----------------------------------------------------------------------------------------------------------------Sel M Sel i CTP P – UL Output Sel Mi • TL DL • TL UL Sel Mi Sel Mi = R DL Sel : Downlink traffic load or the amount of downlink resources allocated to the mobile M i .Atoll 3. M i .4. . Sel Mi Sel Mi . subscriber. or mobile M i . . or mobile M i • Sel Mi UTP E – DL • UTP A – DL : Downlink application user throughput at the pixel. subscriber. • Sel Mi Sel CTP P – UL : Uplink peak RLC channel throughput at the mobile M i as calculated in "Throughput Calculation" on page 538. CINR PDSCH graph available in the reception   Sel equipment assigned to the terminal used by the mobile M i . or mobile M i . • Sel Mi Sel CTP P – DL : Downlink peak RLC channel throughput at the mobile M i as calculated in "Throughput Calculation" on page 538. subscriber. • Sel i R UL M Sel : Amount of uplink resources allocated to the mobile M i as calculated in "Scheduling and Radio Resource Allocation" on page 552. or mobile M i .0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Input • M Sel i R DL Sel : Amount of downlink resources allocated to the mobile M i as calculated in "Scheduling and Radio Resource Allocation" on page 552. Sel Mi Sel • f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile M i • TP Offset : Throughput offset defined in the properties of the service used by the mobile M i Sel Mi Sel Calculations Downlink: Sel Mi Sel Mi Sel Mi  CTP P – DL • Peak RLC User Throughput: UTP P – DL = R DL • Mi Mi   Mi   Effective RLC User Throughput: UTP E – DL = UTP P – DL   1 – BLER  B DL      Sel • Sel Mi Application User Throughput: UTP A – DL Sel Sel Sel Mi Sel Mi Sel Mi f TP – Scaling . CINR PUSCH PUCCH graph available in the reception   equipment assigned to the cell TXi(ic). Sel Sel UTP P – UL : Uplink peak RLC user throughput at the pixel. . Sel • TX i  ic   Mi  BLER  BDL  : Downlink block error rate read from the BLER vs.– TP Offset = UTP E – UL  -----------------------100 Output Sel UTP P – DL : Downlink peak RLC user throughput at the pixel.Atoll 3. • 562 Sel Mi • Sel : Downlink effective RLC user throughput at the pixel.3.– TP Offset = UTP E – DL  -----------------------100 Uplink: Sel Mi Sel Mi Sel Mi  CTP P – UL • Peak RLC User Throughput: UTP P – UL = R UL • Mi Mi   Mi   Effective RLC User Throughput: UTP E – UL = UTP P – UL   1 – BLER  B UL      Sel • Sel Mi Application User Throughput: UTP A – UL Sel Sel Mi Sel Sel Mi Sel Mi f TP – Scaling . Sel • Mi  Mi  BLER  BUL  : Uplink block error rate read from the BLER vs. subscriber. 3: Inter-Transmitter Distance Calculation The formula above implies that two cells facing each other have a smaller effective distance than the actual distance.5. "Automatic Frequency Planning Using the AFP" on page 570.Atoll 3. "Automatic PRACH RSI Planning Using the AFP" on page 576.3% so that the maximum variation in D does not to exceed 1%. This can be the Transmitters folder or a group of transmitters (subfolder). subscriber. Atoll checks the following conditions: 1. 6. It means that the cells of all the TBC transmitters of your ATL document are potential neighbours. "Automatic Physical Cell ID Planning Using the AFP" on page 572.1 Automatic Neighbour Planning The intra-technology neighbour planning algorithm takes into account the cells of all the TBC transmitters. They must fulfil the following conditions: • • • • They are active.3. When automatic allocation starts. or mobile M i • UTP A – UL : Uplink application user throughput at the pixel. "Automatic Inter-technology Neighbour Planning" on page 567. 563 . . The calculation options. Candidate neighbours are ranked in the order of increasing effective distance from the reference cell. The distance between both cells must be less than the user-definable maximum inter-site distance. Figure 6. Atoll calculates the effective distance between the reference cell and its candidate neighbour from the real distance between them and the azimuths of their antennas: Dist  CellA CellB  = D   1 + x  cos  – x  cos   Where x = 0. If no focus zone exists in the ATL document. They belong to the folder on which allocation has been executed. Only TBA cells are assigned neighbours. 2. then the candidate neighbour is discarded. The cells to be allocated will be called TBA cells. subscriber.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 M Sel i M Sel i Sel • UTP E – UL : Uplink effective RLC user throughput at the pixel. Sel 6. They satisfy the filter criteria applied to the Transmitters folder. If the distance between the reference cell and the candidate neighbour is greater than this value.5 Automatic Planning Algorithms The following sections describe the algorithms for: • • • • • "Automatic Neighbour Planning" on page 563. or mobile M i . We assume a reference cell TXi(ic) and a candidate neighbour cell TXj(jc). Atoll takes into account the computation zone. They are located inside the focus zone. D is stated in m. You can force Atoll to keep that cell in the reference cell’s neighbours list by adding the following option in the Atoll. Exceptional pairs are pairs of cells which will always or never be neighbours of each other. In this case. If the Use Coverage Conditions check box is selected. the reference cell will not be added as a neighbour of that cell and that cell will be removed from the reference cell’s neighbours list. The weight of this constraint is always the average of the Min and Max values defined for the adjacency factor. Atoll deletes all the current neighbours and carries out a new neighbour allocation.3. The neighbour list of TXj(jc) is not full. and its importance. Atoll displays a warning in the Event viewer. Atoll adds all the cells geographically adjacent to the reference cell to the candidate neighbour list. there can be two possibilities: i. Force Symmetry: If selected.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks • • ©Forsk 2015 Force Co-site Cells as Neighbours: If selected. A candidate neighbour cell TXi(ic) is considered adjacent to the reference cell TXi(ic) if there exists at least one pixel of TXj(jc)’s best server coverage area where TXi(ic) is the second best server. The weight of this constraint can be defined. only the distance criterion is taken into account. The overlapping zone ( S TX  ic   S TX  jc  ) is defined as follows i • 564 j Here S TX  ic  is the surface area covered by the cell TXi(ic) that comprises all the pixels where: i . the existing neighbours are kept in the list.ini file: [Neighbours] DoNotDeleteSymmetrics = 1 • Force Exceptional Pairs: This option enables you to force/forbid some neighbour relations. The neighbour list of TXj(jc) is full. A symmetric neighbour relation is allowed only if the neighbour list of the reference cell is not already full. Force Adjacent Cells as Neighbours: If selected. Atoll will add TXi(ic) to the end of the list. On the other hand. Figure 6. If not selected. This weight is used to calculate the rank of each neighbour and its importance. If the neighbours list of a cell is full. and its importance. The ranking of adjacent neighbour cells increases with the number of such pixels. • Delete Existing Neighbours: If selected. if neighbourhood relationship is forced in one direction and forbidden in the other. If TXj(jc) is a neighbour of TXi(ic) but TXi(ic) is not a neighbour of TXj(jc). Atoll considers the constraints between exceptional pairs in both directions so as to respect symmetry condition. The weight of this constraint can be defined. 3. the coverage areas of TXi(ic) and TXj(jc) must have an overlap. Cells are considered adjacent across layers if they belong to different layers and have a coverage overlap of at least one pixel. Adjacent cells are sorted in the order of decreasing rank. symmetry cannot be respected. Atoll adds all the cells located on the same site as the reference cell to the candidate neighbour list. It is used to calculate the rank of each neighbour. It is used to calculate the rank of each neighbour. Atoll adds the reference cell to the candidate neighbour list of the its candidate neighbour. Atoll adds all the cells adjacent across network layers to the reference cell to the candidate neighbour list.Atoll 3. Determination of Adjacent Cells: Geographically adjacent cells are determined on the basis of their best server coverage areas. Atoll will not be able to add TXi(ic) to the list. ii. Otherwise. so it will also remove TXj(jc) from the neighbour list of TXi(ic). If you select "Force exceptional pairs" and "Force symmetry".4: Determination of Adjacent Cells • • Force Adjacent Layers as Neighbours: If selected. TXj(jc) is considered a neighbour of TXi(ic) if S TX  ic   S TX  jc  i j -------------------------------------. S TX  ic   S TX  jc  i j When the above conditions are met. • The received RSRP is within E DLRS + O Individual + M HO TX  ic  i TX  ic  i TX i  ic  M HO TX i  ic  M HO • TX  ic  i TX  ic  i TX  ic  i TX  ic  i TX  ic  i TX  ic  i and E DLRS + O Individual + M HO + M End . Atoll calculates the importance of the automatically allocated neighbours. • The received RSRP with offset ( E DLRS + O Individual ) is the highest. When a global handover start value is used. S TX  ic  i and compares this value with the % Min Covered Area.Atoll 3. Atoll calculates the percentage of the coverage area overlap ( --------------------------------------. 100  % Min Coverage Area . table below). The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf.. this value varies between 0 and 100%. For calculating the overlapping coverage areas. global value and the value defined for that cell. i. 565 . Atoll keeps the ones with high importance. The service and terminal are selected such that the selection gives the largest possible coverage areas for the cells. S TX  ic  i Figure 6. Atoll sorts the neighbours by decreasing importance in order to keep the ones with high importance.5: Overlapping Zones Next. If the maximum number of neighbours to be allocated to each cell is exceeded. M Start and M End are global handover start and handover end values. is the handover margin defined for the cell TXi(ic). Atoll uses the service with the lowest body loss. the terminal that has the highest difference between gain and losses. S TX  jc  is the surface area covered by the cell TXj(jc) that comprises all the pixels where: j • The distance to the cell TXj(jc) is less than or equal to the round-trip time distance corresponding to the cell’s PRACH preamble format. • The received RSRP is greater than or equal to the cell’s Min RSRP: E DLRS  T RSRP . 100 ).e. for each cell.3. Atoll uses the higher of the two values.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • The distance to the cell TXi(ic) is less than or equal to the round-trip time distance corresponding to the cell’s PRACH preamble format. = M Start . and the shadowing margin calculated using the defined cell edge coverage probability. • The received RSRP is greater than or equal to the cell’s Min RSRP: E DLRS  T RSRP . if the option is selected. TX  jc  j TX j  jc  • • TX  jc  j TX j  jc  TX i  ic  If a global value of the minimum RSRP threshold ( T RSRP ) is set in the coverage conditions dialogue. 0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Neighbourhood cause When Importance value Existing neighbour Only if the Delete Existing Neighbours option is not selected and in case of a new allocation Existing importance Exceptional pair Only if the Force Exceptional Pairs option is selected 100 % Co-site cell Only if the Force Co-site Cells as Neighbours option is selected Importance Function (IF) Adjacent layer Only if the Force Adjacent Layers as Neighbours option is selected Importance Function (IF) Adjacent cell Only if the Force Adjacent Cells as Neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions Only if the % Min Covered Area is exceeded Importance Function (IF) Symmetric neighbourhood relationship Only if the Force Neighbour Symmetry option is selected Importance Function (IF) The importance is evaluated using an Importance Function (IF). • • • The co-site factor (C): a Boolean. The adjacency factor (A): the percentage of adjacency.3. The minimum and maximum importance assigned to each of the above factors can be defined.Atoll 3. d Di  = 1 – ---------d max d is the effective distance (in m). Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 30% Adjacency factor (A) Min(A) 30% Max(A) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The Importance Function is evaluated as follows: Neighbourhood cause Importance Function Resulting IF using the default values from the table above Coverage Min(O)+Delta(O){Max(Di)(Di)+(100%Max(Di))(O)}+Min(Di)+Delta(Di)(Di) 10%+20%{10%(Di)+90%(O)}+1%+9%(Di) Adjacent layer (Min(A)+Max(A))/2 45% Adjacent cells Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 30%+30%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Co-site cells Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 60%+40%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Where: Delta(X)=Max(X)-Min(X) 566 . It corresponds to the real inter-transmitter distance ( D in m) weighted by the azimuths of antennas. which takes into account the following factors: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. d max is the maximum distance between the reference transmitter and a possible neighbour. The overlapping factor (O): the percentage of overlapping. 2 Automatic Inter-technology Neighbour Planning The inter-technology neighbour planning algorithm takes into account all the TBC transmitters (if the other technology is GSM) or the cells of all the TBC transmitters (for any other technology than GSM).0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. Atoll takes into account the computation zone. With the default values for minimum and maximum importance fields. the neighbour importance calculated with respect to distance is based on the global Max inter-site distance setting for all neighbour candidates. They satisfy the filter criteria applied to the Transmitters folder.ini: [Neighbours] CandidatesMaxDistanceInImportanceCalculation=1 6. there can be cases where the calculated importance is different when the global Max inter-site distance is modified. With a value of Min(O) = 0%. This means that all the TBC transmitters (GSM) or the cells of all the TBC transmitters (all other technologies) of the linked document are potential neighbours. Atoll checks following conditions: 1.3. there can be cases where the real distance between assigned neighbours is higher than the Max inter-site distance. They must fulfil the following conditions: • • • • They are active. We assume a reference cell A and a candidate neighbour B. Cells whose channels have the same centre frequency are listed as intra-carrier neighbours. the candidate neighbour is discarded. As a consequence. the neighbours will be ranked by neighbour cause. Atoll lists only the cells for which it finds new neighbours. and neighbours allocated based on coverage overlapping. The cells to be allocated in the main document will be called TBA cells. The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. Atoll calculates the effective distance between the reference cell and its candidate neighbour from the real distance between them and the azimuths of their antennas: Dist  CellA CellB  = D   1 + x  cos  – x  cos   567 . When automatic allocation starts. you can force Atoll to prioritise the individual distances between reference cells and their respective neighbour candidates by adding the following lines in Atoll.ini: [Neighbours] RealInterSiteDistanceCondition=1 • By default. As a consequence. They belong to the folder on which allocation has been executed.Atoll 3. The distance between reference cell and the candidate neighbour must be less than the user-definable maximum inter-site distance. Only TBA cells are assigned neighbours. There can be a mix of the neighbourhood causes. If the distance is greater than this value. because the effective distance is smaller. In the results. neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping. If the Min and Max value ranges of the importance function factors do not overlap. • By default. the neighbours may be ranked differently. To avoid that. If the Min and Max value ranges of the importance function factors overlap. the automatic neighbour allocation compares the defined Max intersite distance with the effective inter-cell distance. adjacent cells. You can force Atoll to compare the Max inter-site distance with the real inter-site distance by adding the following lines in Atoll.5. neighbour cells are listed as inter-carrier neighbours. They are located inside the focus zone. neighbours will be ranked in this order: co-site neighbours. If no focus zone exists in the ATL document. This can be the Transmitters folder or a group of transmitters (subfolder). Otherwise. adjacent layers. • 2nd case: The margin is other than 0 dB. Atoll adds all the transmitters/cells located on the same site as the reference cell in its candidate neighbour list. Exceptional pairs are pairs of cells which will always or never be neighbours of each other. Candidate neighbours are ranked in the order of increasing effective distance from the reference cell. If not selected. The signal level received from B exceeds the minimum required. It is used to calculate the rank of each neighbour and its importance. • 2nd case: The margin is other than 0dB. 100 ) and compares this value with the % Atoll calculates the percentage of the coverage area overlap ( ----------------SA SA  SB . 568 . or TDSCDMA network. Atoll deletes all the current neighbours and carries out a new neighbour allocation. The calculation options: • • • • CDMA carriers: This option is available when an LTE network is being co-planned with a UMTS. with a 0 dB margin. SA  SB . Atoll will allocate only the cells using the selected carriers as neighbours. . SA is the area where: The reference signal energy per resource element received from A exceeds the minimum required (Min RSRP) and is within a margin from the highest signal level.6: Inter-Transmitter Distance Calculation The formula above implies that two cells facing each other have a smaller effective distance than the actual distance.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Where x = 0. • Algorithm based on coverage overlapping: The coverage areas of the reference cell A and the candidate neighbour B must overlap ( S A  S B ). 100  % Min Covered Area . This means that the reference signal energy per resource element received from A is greater than the minimum required (Min RSRP).3% so that the maximum variation in D does not to exceed 1%. SB is the area where: The signal level received from B exceeds the minimum required and is within a margin from the best signal level. Two cases may exist for SB: • 1st case: SB is the area where the candidate neighbour is the best server. 2. Delete existing neighbours: If selected. the margin must be set to 0dB. Force exceptional pairs: This option enables you to force/forbid some neighbour relations. and is the highest one. D is stated in m. You may choose one or more carriers. Two cases may exist for SA: • 1st case: SA is the area where the cell A is the best serving cell. Figure 6. Min Covered Area.3. B is considered a neighbour of A if ----------------SA Candidate neighbours are ranked in the order of decreasing coverage area overlap percentages. Force co-site cells as neighbours: If selected. the existing neighbours are kept in the list.Atoll 3. The weight of this constraint can be defined. In this case. CDMA. Neighbour relation criterion: • Allocation based on distance: The allocation algorithm is based on the effective distance between the reference cell and its candidate neighbour. and is the highest one. This option enables you to select the CDMA carrier(s) that you want Atoll to consider as potential neighbours of LTE cells. 3. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The IF evaluates importance as follows: Co-site Neighbourhood cause IF Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)} Yes Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+ Max(O)(O)/(Max(Di)+Max(O))} 60%+40%{1/7%(Di)+6/7%(O)} Where Delta(X)=Max(X)-Min(X) 569 . The importance (%) of neighbours depends on the distance and on the reason of allocation: • For allocation based on distance: Neighbour cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter/cell If the Force co-site cells as neighbours option is selected 100 % Neighbour relation that fulfils distance conditions If the maximum distance is not exceeded d1 – ---------d max d is the effective distance between the reference cell and the neighbour and d max is the maximum inter-site distance. The overlapping factor (O): the percentage of overlapping. • • The co-site factor (C): a Boolean. which takes into account the following factors: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas.Atoll 3. d max is the maximum distance between the reference transmitter and a possible neighbour. • For allocation based on coverage overlapping: Neighbour cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter/cell If the Force co-site cells as neighbours option is selected IF Neighbourhood relationship that fulfils coverage conditions If the % minimum covered area is exceeded IF The importance is evaluated using an Importance Function (IF). The IF is user-definable using the Min importance and Max importance fields.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 Next. Atoll keeps the ones with high importance. Atoll sorts the neighbours by decreasing importance in order to keep the ones with high importance. d Di  = 1 – ---------d max d is the effective distance (in m). Atoll calculates the importance of the automatically allocated neighbours. If the maximum number of neighbours to be allocated to each cell is exceeded. or optimum. In order to improve network performance. Default weight  IM = 0.5. i. They are located inside the focus zone. In the results. Default weight  Neighbour = 0. The AFP is based on a cost function which represents the interference level in the network.Atoll 3. the interference within the network is reduced as much as possible. If the Min and Max value ranges of the importance function factors do not overlap.and adjacent channel interference as much as possible while respecting any constraints input to it. neighbours. They must fulfil the following conditions: • • • • They are active. Channel separation is studied between each TBA cell and its related cells.3.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks • • • ©Forsk 2015 Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. and then the overall cost for the TBA cell. Atoll calculates the cost between each individual TBA and related cell. You can modify these weights in your LTE document. the neighbours may be ranked differently. neighbour relations.5. The main constraints are the resources available for allocation. In other words. the number of frequencies with which the AFP can work. 6. The following describes the AFP’s automatic planning method for carriers in LTE networks. the LTE AFP tries to minimise co. 6.3 • Cells within the cell’s (or the default) minimum reuse distance. which takes into account interference matrices. They satisfy the filter criteria applied to the Transmitters folder. and distance between transmitters.and Adjacent Channel Overlaps Calculation" on page 493. Atoll displays only the cells for which it finds new neighbours. and the relationships to take into account. If no focus zone exists in the ATL document. The AFP takes into account the cells of all the TBC transmitters. The absolute values of the constraint weights are calculated from the relative weights (%) defined in the Constraint Weights dialogue as follows: 570 .e. The cells to be allocated will be called TBA cells..1 Constraint and Relationship Weights The AFP is based on a cost function which takes into account channel separation constraints based on the channel overlap ratio as calculated in "Co. if the check box "Existing neighbours" is selected.5 • Cells that are listed in the interference matrix of the TBA cell. and distance between transmitters.. if the check box "Reuse distance" is selected.3. There can be a mix of the neighbourhood causes. The best. i.3 Automatic Frequency Planning Using the AFP The role of an Automatic Frequency Planning (AFP) tool is to assign frequencies (channels) to cells of a network such that the overall network performance is optimised. Co-channel interference is the main reason for overall network quality degradation in LTE. Their channel allocation status is not set to locked. neighbours will be ranked in this order: co-site neighbours and neighbours allocated based on coverage overlapping. With the default values for minimum and maximum importance fields. Related cells of a TBA cell are: • Its neighbours.e. interference matrices. The aim of the AFP is to minimise the cost.2 The sum of the weights assigned to the above relations is 1. If the Min and Max value ranges of the importance function factors overlap. the neighbours will be ranked by neighbour cause. frequency plan is the one which corresponds to the lowest cost. Default weight  Dis tan ce = 0. Atoll takes into account the computation zone. e. TX i  ic  – TX j  jc   Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance between TX i  ic  – TX j  jc  them.and adjacent channel interference probabilities calculated as TX  ic  – TX  jc  i j explained in "Interference Matrix Calculation" on page 578. the total cost of the current frequency plan for the entire network is simply the sum of the total TBA cell costs calculated above. For manual neighbour planning.3.3 AFP Algorithm The AFP algorithm is an iterative algorithm which: • • • • Calculates the cost (as describe d above) of the initial frequency plan.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 % Neighbour  Neighbour = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce % IM  IM = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce % Dis tan ce  Dis tan ce = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce 6. the frequency plan which provides the lowest total cost. $ Total =  TX i  ic  $ Total TX  ic  i 6.Atoll 3. TX i  ic  – TX j  jc  TX i  ic  – TX j  jc   Neighbour is the importance of the relationship between the TBA cell and its related neighbour cell. Tries different frequency plans in order to reduce the cost.3.5. this value is equal to 1. TX i  ic  – TX j  jc   IM is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows: TX i  ic  – TX j  jc  TX i  ic  – TX j  jc   IM = r CCO TX i  ic  – TX j  jc  TX i  ic  – TX j  jc   IM – CC and  IM – CC TX i  ic  – TX j  jc    IM – CC TX i  ic  – TX j  jc  + r ACO TX i  ic  – TX j  jc    IM – AC are respectively the co. r CCO TX  ic  – TX  jc  i j and r ACO are the co. 571 . i. Atoll calculates the quality reduction factor for the TBA cell and its related cell from the cost calculated above as follows: QRF TX i  ic  – TX j  jc  = 1–$ TX i  ic  – TX j  jc  The quality reduction factor is a measure of the cost of an individual relation. The total cost of the current frequency plan for any TBA cell is given as follows. Memorises the different frequency plans in order to determine the best one.  Dis tan ce is calculated as explained in "Distance Importance Calculation" on page 579.and Adjacent Channel Overlaps Calculation" on page 493.and adjacent channel overlap ratios as calculated in "Co. i.and Adjacent Channel Overlaps Calculation" on page 493... considering all the cells with which the TBA cell has relations: TX i  ic  $ Total = 1 –  QRF TX i  ic  – TX j  jc  TX j  jc  And.  Neighbour is calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 563. Stops when it is unable to improve the cost of the network.3. and proposes the last known best frequency plan as the solution.e.5.2 Cost Calculation The cost of the relation between the TBA cell and its related cell is calculated as follows: $ TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  = rO TX i  ic  – TX j  jc  Where r O TX i  ic  – TX j  jc     Neighbour   Neighbour TX i  ic  – TX j  jc  +  Dis tan ce   Dis tan ce TX  ic  – TX j  jc   +  i IM IM  is the channel overlap ratio as calculated in "Co. which takes into account interference matrices. distance between transmitters. Once they know the PSS ID of the cell. Assigned weight  CRS = 0 5. first-order neighbours of a common GSM or UMTS cell in 3GPP multi-RAT documents and CDMA cell in 3GPP2 multi-RAT documents. Once the physical cell ID and the associated pseudo-random sequence is known to the mobile. They must fulfil the following conditions: • • • • They are active. 504 physical cell IDs are available. There are as many pseudo-random sequences defined in the 3GPP specifications. SSS ID. Assigned weight  PSS = 0. The combination of these two IDs gives the physical cell ID and the associated pseudo-random sequence that is transmitted over the downlink reference signals. for UL DMRS sequence groups. Physical cell ID. The SSS and PSS are transmitted over the centre six frequency blocks independent of the channel bandwidths used by cells. with each group containing 3 unique identities (called PSS IDs in Atoll). PCI Mod (number of frequency blocks / 2). The cells to be allocated will be called TBA cells. PCI Mod 6.5.5. Cell search and selection will be impossible.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 6. PCI Mod 30. Assigned weight  SSS = 0. for PCFICH resource element groups. They satisfy the filter criteria applied to the Transmitters folder. neighbour relations (first-order neighbours. Therefore. if all the cells in the network transmit the same physical cell ID. and the frequency plan of the network. numbered from 0 to 503. they listen to the SSS of the cell in order to know the SSS ID. Mobiles synchronise there transmission and reception frequency and time by listening first to the PSS. PSS ID. the cell is recognized by the mobile based on the received downlink reference signals. and a PSS ID is defined by a number in the range of 0 to 2.4.02 3. The AFP takes into account the cells of all the TBC transmitters. The following describes the AFP’s automatic planning method for physical cell IDs in an LTE network. Assigned weight  ULDMRS = 0 6. Each cell’s downlink reference signals transmit a pseudo-random sequence corresponding to the physical cell ID of the cell. If no focus zone exists in the ATL document. for single-antenna port DL CRS). Their PSS ID and SSS ID statuses are both not set to locked. 572 .Atoll 3. and optionally second-order neighbours). As can be understood from the above description.23 4.4 Automatic Physical Cell ID Planning Using the AFP In LTE. in the order of priority: 1.3. it will be impossible for a mobile to identify different cells. Downlink channel quality measurements are also made on the downlink reference signals. They are located inside the focus zone. first-order neighbours of a common LTE cell. Atoll takes into account the computation zone. 6. An SSS ID is thus uniquely defined by a number in the range of 0 to 167. it is important to intelligently allocate physical cell IDs to cells so as to allow easy recognition of cells by mobiles. Physical cell IDs are grouped into 168 unique cell ID groups (called SSS IDs in Atoll). Assigned weight  PCFICH = 0 The sum of the weights assigned to the above constraints is 1.75 2. Assigned weight  ID = 0.1 Constraint and Relationship Weights The AFP is based on a cost-based function which takes into account the following constraints. Assigned weight  Neighbour = 0. and  Inter – Neighbour = 0.15 . Atoll calculates the cost between each individual TBA and related cell. In this case. and then the overall cost for the TBA cell. second-order neighbours can also be taken into account. the assigned weights are:  Neighbour = 0. Figure 6.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 You can modify these weights in your LTE document.ini file (see the Administrator Manual).Atoll 3.10 . 573 . The absolute values of the constraint weights are calculated from the relative weights (%) defined in the Constraint Weights dialogue as follows: % ID  ID = --------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % SSS + % CRS + % ULDMRS + % PCFICH % PSS  PSS = --------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % SSS + % CRS + % ULDMRS + % PCFICH % SSS  SSS = --------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % SSS + % CRS + % ULDMRS + % PCFICH % CRS  CRS = --------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % SSS + % CRS + % ULDMRS + % PCFICH % ULDMRS  ULDMRS = --------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % SSS + % CRS + % ULDMRS + % PCFICH % PCFICH  PCFICH = --------------------------------------------------------------------------------------------------------------------------------------------% ID + % PSS + % SSS + % CRS + % ULDMRS + % PCFICH The above constraints are studied between each TBA cell and its related cells. Assigned weight  Inter – Neighbour = 0.25 .3. if the check box "Existing neighbours" is selected. If the collision between neighbours of a common cell is not taken into account.  Inter – Neighbour applies to the relation between neighbours of a common cell. a UMTS cell or a GSM transmitter in 3GPP multi-RAT documents or an LTE or CDMA cell in 3GPP2 multi-RAT documents.5 and that of the collision between neighbours of a common cell is of course  Inter – Neighbour = 0 .7: Neighbour Relations for Physical Cell ID Allocation • Cells that are listed in the interference matrix of the TBA cell.7 on page 573 depicts the different neighbour relations that may exist in LTE. Figure 6.ini file (see the Administrator Manual).35 TBA cells which are first-order neighbours of a common cell are also related to each other through that cell. which can be an LTE cell. By adding an option in the Atoll. Related cells of a TBA cell are: • Its neighbours. the weight assigned to the direct first-order neighbour relation alone is  Neighbour = 0. This relation is also taken into account.15 You can choose to not take into account the physical cell ID collision between neighbours of a common cell by adding an option in the Atoll.  2nd – Neighbour = 0. The absolute values of the constraint weights are calculated from the relative weights (%) defined in the Constraint Weights dialogue as follows: % Neighbour  Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % Inter – Neighbour  Inter – Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % 2nd – Neighbour  2nd – Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % IM  IM = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % Dis tan ce  Dis tan ce = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce 6.2 Cost Calculation Atoll calculates the constraint violation levels between the TBA cell TXi(ic) and its related cell TXj(jc) as follows: TX  ic  – TX  jc  i j VL 1 TX  ic  – TX  jc  i j VL 2 ID SSS CRS ULDMRS PCFICH =  ID  p Coll +  SSS  p Penalty +  CRS  p Coll +  ULDMRS  p Coll +  PCFICH  p Coll PSS =  PSS  p Coll Where  ID . PSS ID.5.  ID PSS TX i  ic  TX i  ic  if R Co-site  3 AND ID SSS TX  ic  i if R Co-site  3 AND TX  ic  i ID  TX j  jc   ID SSS TX  jc  j – ID  TX  ic  i  R Co-site if the SSS ID Otherwise SSS planning strategy is set to "Same per site".2 The sum of the weights assigned to the above relations is 1. The SSS penalty models the SSS ID allocation constraint.Atoll 3. Assigned weight  Dis tan ce = 0. and  SSS are the weights assigned to the physical cell ID. ID p Coll is the physical cell ID collision probability given by ID p Coll   1 =    0   1 PSS PSS p Coll is the PSS ID collision probability given by p Coll =    0 SSS SSS p Penalty is the SSS ID penalty given by p Penalty   1  =   1   0 TX  ic  i if ID  TX i  ic  if ID  TX i  ic  if ID PSS TX j  jc  .3 • Cells within the cell’s (or the default) reuse distance. and SSS ID constraints. reference signal collision probability given by .3. and by p Penalty = 0 if the SSS ID planning strategy is set to "Free". if the check box "Reuse distance" is selected.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 Assigned weight  IM = 0. CRS p Coll is   1 CRS p Coll =    0 574 the single TX i  ic  if ID  TX i  ic  if ID  antenna downlink TX j  jc  Mod6 = ID  TX j  jc  Mod6  ID Mod6 Mod6 cell-specific .  ID  TX j  jc  = ID PSS if ID PSS TX i  ic  TX  jc  j = ID  TX j  jc  .  PSS .4. You can modify these weights in your LTE document. p Coll is the collision probability of the physical control format indicator channel resource element groups given by PCFICH p Coll    1  =     0  TX i  ic  TX j  jc  TX i  ic  TX j  jc  TX i  ic  TX j  jc   N FB   N FB  Mod  ---------------- = ID  Mod  -----------------  2   2  TX i  ic  TX j  jc   N FB   N FB  - Mod  ---------------Mod  ---------------  ID   2   2  if ID  if ID  . Next. the importance of the physical cell ID collision is the multiple of the importance values of the first order neighbour relations between the TBA cell and its second order neighbour. are respectively the co. If the TBA cell has the same physical cell ID assigned as one of its second-order neighbours.Atoll 3.  Inter – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour planning. Next. If more than one pair of neighbours of the TBA cell has the same physical cell ID assigned.and Adjacent Channel Overlaps Calculation" on page 493.  2nd – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour planning. and adjacent channel overlap ratios as calculated in "Co. r CCO TX i  ic  – TX j  jc  . For manual neighbour planning. r O TX i  ic  – TX j  jc  . TX i  ic  – TX j  jc   Interference TX i  ic  – TX j  jc   IM TX i  ic  – TX j  jc   IM TX i  ic  – TX j  jc  =  IM   IM TX i  ic  – TX j  jc   IM – CC TX i  ic  – TX j  jc   f Overlap is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows: TX i  ic  – TX j  jc  = r CCO TX i  ic  – TX j  jc  and  IM TX i  ic  – TX j  jc  +  Dis tan ce   Dis tan ce TX i  ic  – TX j  jc    IM – CC TX i  ic  – TX j  jc  =  IM – CC TX i  ic  – TX j  jc  and  IM – CC TX i  ic  – TX j  jc  + r ACO TX i  ic  – TX j  jc    IM – AC if the frequency plan is taken into account otherwise. this value is equal to 1. 575 . then the importance is the highest value among all the averages: TX i  ic  – TX j1  j1c   Inter – Neighbour TX i  ic  – TX j2  j2c    Neighbour  +  Neighbour = Max  --------------------------------------------------------------------------------- 2  All Neighbour Pairs  with ID Collisions Where TX j1  j1c  and TX j2  j2c  are two neighbours of the TBA cell TX i  ic  that have the same physical cell ID assigned.and adjacent channel interference probabilities calculated as TX i  ic  – TX j  jc  explained in "Interference Matrix Calculation" on page 578. If the TBA cell is related to its second order neighbour through more than one first order neighbour. and r ACO are the total. If two cells are neighbours of a common cell and have the same physical cell ID assigned. the importance of the physical cell ID collision is the average of their neighbour importance values with the common neighbour cell. TX i  ic  – TX j  jc   Neighbours TX i  ic  – TX j  jc  =  Neighbour   Neighbour TX i  ic  – TX j  jc  +  Inter – Neighbour   Inter – Neighbour +  2nd – Neighbour   2nd – Neighbour TX i  ic  – TX j  jc  Where  Neighbour is the importance of the relationship between the TBA cell and its related neighbour cell. Atoll calculates the importance of the neighbour relations between the TBA cell and its related cell. co-channel. The above applies to intra-technology as well as inter-technology neighbours in 3GPP multi-RAT and 3GPP2 multi-RAT documents. the importance is the highest value among all the multiples:  2nd – Neighbour = TX  ic  – TX  jc  j  i  Neighbour All Neighbour Pairs Max TX j  jc  – TX k  kc      Neighbour with ID Collisions Where TX k  kc  is the second-order neighbour of TX i  ic  through TX j  jc  .  Neighbour is calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 563.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 ULDRMS p Coll PCFICH   1 =    0 ULDMRS is the UL DMRS collision probability given by p Coll TX  ic  i if ID  TX  jc  j Mod30 = ID  TX  ic  i ID  Mod30 if  Mod30 TX  jc  j ID  Mod30 . Atoll calculates the importance of the interference relations between the TBA cell and its related cell.3. 6. and proposes the last known best physical cell ID plan as the solution.4. Tries different physical cell IDs to cells in order to reduce the costs.5 Automatic PRACH RSI Planning Using the AFP The following describes the AFP’s automatic planning method for PRACH RSIs in an LTE network.1 Constraint and Relationship Weights The AFP is based on a cost-based function which takes into account various relations between each TBA cell and its related cells. The total cost of the current physical cell ID plan for any TBA cell is given as follows. i. page 579. and then the overall cost for the TBA cell. i. 576 .3. Their PRACH RSI status is not set to locked.10 . and distance between transmitters. Related cells of a TBA cell are: • Its neighbours.Atoll 3.25 and  2nd – Neighbour = 0. The AFP takes into account the cells of all the TBC transmitters. The cells to be allocated will be called TBA cells.. Atoll calculates the quality reduction factor for the pair as follows: QRF TX  ic  – TX  jc  i j TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc    i j i j j i j i j i j   i + VL 2 + VL    f = 1 –   VL 1  Interference 1 Neighbours Overlap    The quality reduction factor is a measure of the cost of an individual relation. if the check box "Existing neighbours" is selected.5.50 By adding an option in the Atoll.5. $ Total =  TX i  ic  $ Total TX i  ic  6.3 AFP Algorithm The AFP algorithm is an iterative algorithm which: • • • • Calculates the cost (as described above) of the current physical cell ID plan. From the constraint violation levels and the importance values of the relations between the TBA and its related cell. 6. Assigned weight  Neighbour = 0. Memorises the different plans in order to determine the best plan. In this case. Stops when it is unable to improve the cost of the network.e.e. They must fulfil the following conditions: • • • • They are active.5. If no focus zone exists in the ATL document. Atoll takes into account the computation zone. Atoll calculates the cost between each individual TBA and related cell.. second-order neighbours can also be taken into account. the assigned weights are:  Neighbour = 0. considering all the cells with which the TBA cell has relations: TX i  ic  $ Total = 1 –  QRF TX i  ic  – TX j  jc  TX j  jc  And. TX  ic  – TX  jc  i j f Overlap ©Forsk 2015 is calculated TX  ic  – TX  jc  i j rO = as explained in "Distance if the frequency plan is taken into account and Importance Calculation" TX  ic  – TX  jc  i j f Overlap on = 1 otherwise. the total cost of the current physical cell ID plan for the entire network is simply the sum of the total TBA cell costs calculated above.5. which takes into account interference matrices. neighbour relations (first-order neighbours and optionally second-order neighbours). They satisfy the filter criteria applied to the Transmitters folder.ini file (see the Administrator Manual). which provides the lowest total cost. They are located inside the focus zone.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks TX  ic  – TX  jc  i j  Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance between TX  ic  – TX  jc  i j  Dis tan ce them. Atoll 3. Atoll calculates the importance of the neighbour relations between the TBA cell and its related cell. Next. You can modify these weights in your LTE document. Assigned weight  IM = 0.5. Assigned weight  Dis tan ce = 0. TX i  ic  N Req PRACH RSIs TX  ic  – TX  jc  i j TX  ic  i Where N Common PRACH RSIs is the number of PRACH RSIs common between cells TXi(ic) and TXj(jc).2 Cost Calculation Atoll calculates the constraint violation levels between the TBA cell TXi(ic) and its related cell TXj(jc) as follows: VL TX i  ic  – TX j  jc  PRACH p Coll PRACH = Min  1 p Coll  TX i  ic  – TX j  jc  N Common PRACH RSIs ID is the PRACH RSI collision probability given by p Coll = ----------------------------------------.  2nd – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour planning.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 • Cells that are listed in the interference matrix of the TBA cell. If the TBA cell has the same PRACH RSI assigned as one of its second-order neighbours. The absolute values of the constraint weights are calculated from the relative weights (%) defined in the Weights dialogue as follows: % Neighbour  Neighbour = ---------------------------------------------------------------------------------------------------------------------------------% Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % 2nd – Neighbour  2nd – Neighbour = ---------------------------------------------------------------------------------------------------------------------------------% Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % IM  IM = ---------------------------------------------------------------------------------------------------------------------------------% Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % Dis tan ce  Dis tan ce = ---------------------------------------------------------------------------------------------------------------------------------% Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce 6.2 The sum of the weights assigned to the above relations is 1. If the TBA cell is related to its second order neighbour through more than one first order neighbour.5. this value is equal to 1.3. TX i  ic  – TX j  jc   Neighbours TX i  ic  – TX j  jc  =  Neighbour   Neighbour TX i  ic  – TX j  jc  Where  Neighbour +  2nd – Neighbour   2nd – Neighbour TX i  ic  – TX j  jc  is the importance of the relationship between the TBA cell and its related neighbour cell.3 • Cells within the cell’s (or the default) reuse distance. TX i  ic  – TX j  jc   Interference TX i  ic  – TX j  jc  =  IM   IM TX i  ic  – TX j  jc  +  Dis tan ce   Dis tan ce TX i  ic  – TX j  jc   f Overlap 577 . and N Req PRACH RSIs is the number of PRACH RSIs required by the cell TXi(ic). the importance of the PRACH RSI collision is the multiple of the importance values of the first order neighbour relations between the TBA cell and its second order neighbour.  Neighbour is calculated during automatic neighbour planning by Atoll as explained in the Technical Reference Guide. Next. the importance is the highest value among all the multiples:  2nd – Neighbour = TX  ic  – TX  jc  j  i Neighbour  All Neighbour Pairs Max TX j  jc  – TX k  kc    Neighbour   with Collisions Where TX k  kc  is the second-order neighbour of TX i  ic  through TX j  jc  . if the check box "Reuse distance" is selected. For manual neighbour planning. Atoll calculates the importance of the interference relations between the TBA cell and its related cell. the total cost of the current PRACH RSI plan for the entire network is simply the sum of the total TBA cell costs calculated above. Atoll calculates the quality reduction factor for the pair as follows: QRF TX i  ic  – TX j  jc  = 1 – VL TX i  ic  – TX j  jc  TX i  ic  – TX j  jc     Interference  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  +  Neighbours  f Overlap   The quality reduction factor is a measure of the cost of an individual relation.3.6. r CCO TX  ic  – TX  jc  i j .Atoll 3.and adjacent channel interference probabilities calculated as TX  ic  – TX  jc  i j explained in "Interference Matrix Calculation" on page 578. considering all the cells with which the TBA cell has relations: TX i  ic   QRF $ Total = 1 – TX i  ic  – TX j  jc  TX j  jc  And.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks TX  ic  – TX  jc  i j  IM TX  ic  – TX  jc  i j  IM ©Forsk 2015 is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows: TX  ic  – TX  jc  i j = r CCO TX  ic  – TX  jc  i j and  IM TX i  ic  – TX j  jc  TX  ic  – TX  jc  i j   IM – CC TX  ic  – TX  jc  i j =  IM – CC TX i  ic  – TX j  jc   IM – CC and  IM – CC TX  ic  – TX  jc  i j + r ACO TX  ic  – TX  jc  i j   IM – AC if the frequency plan is taken into account otherwise.6 Appendices 6. are respectively the co.e. and proposes the last known best PRACH RSI plan as the solution.5. Stops when it is unable to improve the cost of the network.1 Interference Matrix Calculation The co-channel interference probability is calculated as follows: S TX  ic  i TX i  ic  – TX j  jc   IM – CC TX j  jc  TX i  ic    n  C Max + M Quality  Sym -------------------- ----------------------------------------------------  TX  ic  TX i  ic  TX  ic  TX i  ic  10 10 i   T i C – 10  Log  10 + 10 2N –n DLRS  FB   RSRP Sym        = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX  ic  i The adjacent channel interference probability is calculated as follows: 578 .3 AFP Algorithm The AFP algorithm is an iterative algorithm which: • • • • Calculates the cost (as described above) of the current PRACH RSI plan. From the constraint violation levels and the importance values of the relations between the TBA and its related cell. TX i  ic  – TX j  jc   Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance between TX i  ic  – TX j  jc   Dis tan ce them. and r ACO are the total.. r O TX  ic  – TX  jc  i j . Memorises the different plans in order to determine the best plan. co-channel.5. i.5.5. and adjacent channel overlap ratios calculated as calculated in "Co.e. which provides the lowest total cost. $ Total =  TX i  ic  $ Total TX  ic  i 6. = 1 otherwise.. i. TX i  ic  – TX j  jc  f Overlap is calculated TX i  ic  – TX j  jc  = rO as explained in "Distance Importance TX i  ic  – TX j  jc  if the frequency plan is taken into account and f Overlap Calculation" on page 579. 6. Tries different PRACH RSIs to cells in order to reduce the costs.and Adjacent Channel Overlaps Calculation" on page 493. The total cost of the current PRACH RSI plan for any TBA cell is given as follows. C Max TX i  ic  received maximum signal level from the cell TXj(jc) calculated using the Max Power defined for this cell.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks AT330_TRR_E1 S TX  ic  i TX  ic  – TX  jc  i j  IM – AC TX  jc  TX  ic  TX  ic  j i i   n  C Max + M Quality + f ACS  Sym ------------------------------------------------------------------------------------------------ TX  ic  TX  ic   TX  ic  TX  ic  i 10 10 i i    T i C 10  Log 10 + 10 2N DLRS –  FB   RSRP – n Sym        = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX  ic  i For frequencies farther than the adjacent channel.3. 579 . either defined for each TBA cell individually or set for all the TBA cells in the AFP dialogue.5.8: Weighted Distance Between Cells The above formula implies that two cells facing each other will have a shorter effective distance between them than the real distance.8 on page 579.2 Distance Importance Calculation TX  ic  – TX  jc  i j The distance importance between two cells (  Dis tan ce TX i  ic  – TX j  jc   Dis tan ce   1   D Reuse   2 =  Log   --------------------------------  TX i  ic  – TX j  jc    D    --------------------------------------------------------2  Log  D Reuse   if D ) is calculated as follows: TX  ic  – TX  jc  i j 1 Otherwise Where D Reuse is the minimum reuse distance. TX i  ic  TX i  ic  Here S TX  ic  is the best server coverage area of the cell TXi(ic). x is set TX i  ic  – TX j  jc  to 10 % so that the maximum variation in D due to the azimuths does not exceed 40 %. d = d TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  is the weighted distance between the TBA cell TXi(ic) and its related cell TXj(jc) calculated as   1 + x   cos    – cos    – 2   is weighted according to the azimuths of the TBA cell and its related cell with respect to the straight line joining TX i  ic  – TX j  jc  is the distance between the two cells considering any offsets with respect to the site locations. that comprises all the pixels where E DLRS  T RSRP i calculated in "Best Server Determination" on page 535.  and  are calculated from the azimuths of the two cells as shown in Figure 6. M Quality is the quality TX i  ic  margin used for the interference matrices calculation.6. 6. and D follows: D D TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  them.Atoll 3. C DLRS is the received downlink reference signal level from the cell TXi(ic). And. n Sym TX  ic  i subcarrier noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 505 and N FB is the is the per- is the total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell. S TX  ic  i Condition as is the best server coverage area of the cell TXi(ic) TX i  ic  TX j  jc  where the given condition is true. and two cells pointing in opposite directions will have a greater effective distance. f ACS is the adjacent channel suppression factor defined for the frequency band of the cell TXi(ic). the interference probability is 0. Figure 6. Figure 6.3.9: Importance Based on Distance Relation 580 .Atoll 3. and cells that are located far have low importance. which is interpreted as a high cost. This figure shows that cells that are located near (based on the effective distance which is weighted by the orientations of the cells) have high importance.9 on page 580. Cells that are further than the reuse distance do not have any cost related to the distance relation.0 Technical Reference Guide for Radio Networks Chapter 6: LTE Networks ©Forsk 2015 The importance of the distance relation is explained in Figure 6. Chapter 7 3GPP Multi-RAT Networks This chapter covers the following topics: • "Definitions" on page 583 • "Multi-RAT Monte Carlo Simulations" on page 583 • "Multi-RAT Coverage Predictions" on page 585 . 0 Technical Reference Guidefor Radio Networks © Forsk 2015 582 .3.Atoll 3. • • "Simulations Based on User Profile Traffic Maps" on page 584.3. add the following lines in the Atoll. "Simulations Based on Sector Traffic Maps" on page 584. Atoll determines the total number of users attempting connection in each simulation based on the Poisson distribution. The first random trial generates the number of users and their activity status as explained in the following sections depending on the type of traffic input. and whether they are indoor or outdoor according to the percentage of indoor users per clutter class. Once all the user characteristics have been determined.2 Multi-RAT Monte Carlo Simulations The simulation process is divided into two steps. The resulting user distribution complies with the traffic database and maps selected when creating simulations. • Scheduling and Radio Resource Management as explained under "Simulation Process" on page 585.2. "UMTS HSPA Networks" on page 211. Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data.Atoll 3. which can be provided by one or more technology. 7. This may lead to slight variations in the total numbers of users in different simulations.1 User Distribution During each simulation. Name Value Unit Description f act UL Service parameter None Uplink activity factor f act DL Service parameter None Downlink activity factor TL DL – GSM Subcell parameter % Downlink traffic load (GSM) 7. 7. To have the same total number of users in each simulation of a group. These service categories comprise the following service types in different technologies: Constant Bit Rate Services Variable Bit Rate Services GSM GPRS EDGE Circuit Packet (Constant Bit Rate) Packet (Max Bit Rate) UMTS HSPA Circuit R99 Packet HSPA (Constant Bit Rate) Packet R99 Packet HSDPA (Best Effort) Packet HSPA (Best Effort) LTE Voice Data 583 . a second random trial is performed to obtain their geographical locations weighted according to the clutter classes. • Generating a realistic user distribution as explained in "User Distribution" on page 583. Multi-RAT calculations that are the same as those in single-RAT documents can be found in: • • • "GSM GPRS EDGE Networks" on page 125.0 Technical Reference Guide for Radio Networks Chapter 7: 3GPP Multi-RAT Networks AT330_TRR_E1 7 3GPP Multi-RAT Networks This chapter describes the calculations specific to 3GPP multi-RAT documents. and "LTE Networks" on page 445. Atoll performs two random trials.ini file: [Simulation] RandomTotalUsers=0 In 3GPP multi-RAT documents.1 Definitions This table lists the input to coverage prediction and simulation calculations. services can be classified under constant bit rate and variable bit rate services. 3. the number of service sessions. number of users of a user profile per km². or the volume of the data transfer in the uplink and the downlink in each variable bit rate service session as explained in: • • GSM and LTE: "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 476 UMTS: "Simulations Based on User Profile Traffic Maps" on page 225. the number of users of each user profile is calculated from the line length (L) and the user profile density (DUP) • The number of users is a direct input when a user profile traffic map is composed of points. each point is assigned a number of users with given user profile and mobility type. Either throughput demands per service or the number of active users per service are assigned to the coverage areas of each transmitter.e. Each user profile contains a list of services and parameters describing how these services are accessed by the user. Atoll calculates the number of active users of each service UL and DL as follows: • • GSM and LTE: "Simulations Based on Sector Traffic Maps" on page 478 UMTS: "Simulations Based on Sector Traffic Maps" on page 229. or LTE: Terminal A: 30 % Terminal B: 50 % 584 . Live traffic data from the O&M is spread over the best server coverage areas of the transmitters included in the traffic map. For example. (users per km): N Users = L  D UP Atoll calculates the probability for a user being active at a given instant in the uplink and in the downlink according to the service usage characteristics described in the user profiles. i. UMTS. User profiles model the behaviour of the different user categories. if the percentages of terminals are defined as follows: Terminal A (GSM): 30 % Terminal B (GSM+UMTS): 50 % Terminal C (GSM+UMTS+LTE): 20 % For users of services that can be provided by GSM. the average duration of each constant bit rate service session.0 Technical Reference Guide for Radio Networks Chapter 7: 3GPP Multi-RAT Networks ©Forsk 2015 Simulations Based on User Profile Traffic Maps User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of user profiles with an associated mobility type and a given density. i. the final numbers of users are obtains as follows: inactive Number of inactive users: n j inactive = Average  n j inactive  nj GSM Number of users active on UL: n j  UL  = Average  n j  UL  GSM Number of users active on DL: n j  DL  = Average  n j  DL  GSM inactive UMTS  nj  n j  UL  UMTS  n j  DL  UMTS Number of users active on UL+DL: n j  UL + DL  = Average  n j  UL + DL  LTE  n j  UL  LTE  n j  DL  LTE GSM     n j  UL + DL  UMTS  n j  UL + DL  LTE  Simulations Based on Sector Traffic Maps Sector traffic maps are also referred to as live traffic maps. If the map is composed of points.. N Users = S Env  D UP • In case of user profile traffic maps composed of lines.e.Atoll 3.. once several numbers of users with different activity statuses have been calculated for different technologies. For any variable bit rate service (j). Distribution of Terminals Terminals assigned to users depend on the percentages defined per traffic map and the technologies supported by each terminal. User profile traffic maps: Each polygon or line of the map is assigned a density of users with a given user profile and mobility type. The number of users of each user profile is calculated from the surface area (SEnv) of each environment class map (or each polygon) and the user profile density (DUP). In the case of HSPA services. • • UMTS Effective Service Area Analysis (Eb⁄Nt) (DL+UL) is based on a combination of downlink and uplink service area predictions. Atoll selects the highest priority as defined in the service assigned to each mobile. Multi-RAT coverage predictions include: • • Effective Service Area Analysis (DL+UL) Coverage by Throughput (DL) Effective Service Area Analysis (DL+UL) The 3GPP multi-RAT effective service area is the combination of single-RAT effective service areas: • GSM Service Area Analysis (DL) is based on a coverage by coding scheme. as explained in "Effective Signal Analysis Coverage Predictions" on page 472. • Serving cell/technology selection For each mobile. a mobility type. LTE: Atoll determines the best server based on RSRP or RS level and the serving cell selection method. on all the interfered subcells.3 Multi-RAT Coverage Predictions Coverage predictions are calculated by determining the best server for each technology on each pixel and then determining the selected display parameter within the best server’s calculation area. and a service. This implies that a frequency plan has to be defined in order to obtain this GSM/GPRS/EDGE coverage. the mobile will be considered rejected by UMTS.6 % For users of services that can be provided by LTE only. The properties of the non-interfering probe receiver are set by selecting a terminal. The resolutions of coverage predictions do not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. depending on the type of service. If no best server can be found. Each pixel within the calculation area is considered a non-interfering receiver. as explained in "Downlink Service Area Analysis" on page 300 and "Uplink Service Area Analysis" on page 302. 3GPP multi-RAT coverage predictions are combinations of corresponding single-RAT coverage predictions with specific parameter settings.2 Simulation Process Each Monte Carlo simulation is a snap-shot of the network where resource allocation is carried out. LTE Effective Service Area Analysis (DL+UL) is based on a combination of downlink and uplink service area predictions. If no best server can be found. Atoll searches for a serving cell of each supported and available technology as follows: • • • GSM: Atoll determines a best server based on the HCS layer/server selection algorithm. see "Path Loss Calculation Prerequisites" on page 57 for more information).3. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. the mobile will be considered rejected by LTE. as explained in "Circuit Quality Indicators Coverage Predictions" on page 149.2. the coverage is based on a combination of HSDPA et HSUPA service areas as explained in "HSDPA Prediction Study" on page 304 and "HSUPA Prediction Study" on page 309.0 Technical Reference Guide for Radio Networks Chapter 7: 3GPP Multi-RAT Networks AT330_TRR_E1 Terminal C: 20 % For users of services that can be provided by UMTS or LTE: Terminal B: 50/70 = 71.Atoll 3. Terminal C will be assigned. Codec modes and coding schemes are obtained from these radio conditions based on C/I+N without ideal link adaptation (as explained in "Throughput Calculation Based on Interpolation Between C/N and C/(I+N)" on page 138). UMTS: Atoll determines a best server based on Ec/Io.4 % Terminal C: 20/70 = 28. the mobile will be considered rejected by GSM. Two display options are available for this prediction: 585 . Once the potential serving technologies have been identified. Radio conditions are evaluated over the HCS server area with a margin of 4 dB. The steps of this algorithm are listed below. • Technology-wise Monte Carlo simulations as explained in: • • • GSM: "Radio Resource Management in GSM" on page 184 UMTS: "Power Control Simulation" on page 231 LTE: "Scheduling and Radio Resource Management" on page 552 7. If no best server can be found. as explained in "GPRS/EDGE Coverage Predictions" on page 140 or on a coverage by codec modes. 7. The best server and technology assigned to each mobile remains unchanged for the rest of the simulation. Atoll 3. • R99 Service Area Analysis (Eb⁄Nt) (DL) explained in "Downlink Service Area Analysis" on page 300 and HSDPA Throughput Analysis (DL) explained in "HSDPA Prediction Study" on page 304 R99: The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the effective RLC and application throughputs of the R99 layer (see "Downlink Service Area Analysis" on page 300 for more information).3. Coverage by Throughput (DL) The 3GPP multi-RAT throughput prediction is the combination of single-RAT throughput predictions: • GSM Packet Throughput Analysis (DL) explained in "Application Throughput Calculation" on page 139 The 3GPP multi-RAT effective RLC throughput is obtained from the maximum effective RLC throughput of the GSM layer. . considering all available technologies. Available Technologies: Pixels display the colour representing the combined areas over which a multi-technology terminal can be served. Max Effective RLC Throughput: The maximum throughput on the RLC layer that a cell can provide to the selected terminal per pixel taking into account possible transmission errors (BLER).0 Technical Reference Guide for Radio Networks Chapter 7: 3GPP Multi-RAT Networks • • ©Forsk 2015 Technologies: Each pixel displays the colour representing the visible technology having the highest priority defined in the selected service. considering all available technologies. Four display options are available for this prediction: • • • • 586 Effective RLC Throughput: The throughput on the RLC layer that a cell can provide to the selected terminal per pixel taking into account possible transmission errors (BLER) for the highest priority technology. the GSM+UMTS area shows the union between the GSM and the UMTS service areas as explained above. • LTE Coverage by Throughput (DL) explained in "C/(I+N)-based Coverage Predictions" on page 473 The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the Effective RLC Channel Throughput (DL) and the Application Channel Throughput (DL) (see "C/(I+N)-based Coverage Predictions" on page 473 for more information). Application Throughput: The throughput on the application layer that a cell can provide to the selected terminal per pixel taking into account possible transmission errors (BLER) for the highest priority technology. For instance. HSDPA: The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the effective RLC and application throughputs of the HSDPA layer (see "HSDPA Prediction Study" on page 304 for more information). Max Application Throughput: the maximum throughput on the application layer that a cell can provide to the selected terminal per pixel taking into account possible transmission errors (BLER). The 3GPP multi-RAT application throughput from the maximum application throughput of the GSM layer. Chapter 8 3GPP2 Multi-RAT Networks This chapter covers the following topics: • "Definitions" on page 589 • "Multi-RAT Monte Carlo Simulations" on page 589 • "Multi-RAT Coverage Predictions" on page 591 . 0 Technical Reference Guidefor Radio Networks © Forsk 2015 588 .Atoll 3.3. 1 Definitions This table lists the input to coverage prediction and simulation calculations. add the following lines in the Atoll. Name Value Unit Description f act UL Service parameter None Uplink activity factor DL Service parameter None Downlink activity factor f act 8. 8. These service categories comprise the following service types in different technologies: Constant Bit Rate Services Variable Bit Rate Services CDMA Speech 1xRTT Data 1xEV-DO rev. Once all the user characteristics have been determined. This may lead to slight variations in the total numbers of users in different simulations.2 Multi-RAT Monte Carlo Simulations The simulation process is divided into two steps. 589 .e.2.Atoll 3. Multi-RAT calculations that are the same as those in single-RAT documents can be found in: • • "CDMA2000 Networks" on page 337.3. A (Best Effort) 1xEV-DO rev. B (Best Effort) LTE Voice Data Simulations Based on User Profile Traffic Maps User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of user profiles with an associated mobility type and a given density. 0 1xEV-DO rev. B (Guaranteed Bit Rate) 1xEV-DO rev. which can be provided by one or more technology. A (Guaranteed Bit Rate) 1xEV-DO rev. • • "Simulations Based on User Profile Traffic Maps" on page 589. and "LTE Networks" on page 445 8. a second random trial is performed to obtain their geographical locations weighted according to the clutter classes. The resulting user distribution complies with the traffic database and maps selected when creating simulations. Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data.. • Generating a realistic user distribution as explained in "User Distribution" on page 589.0 Technical Reference Guide for Radio Networks Chapter 8: 3GPP2 Multi-RAT Networks AT330_TRR_E1 8 3GPP2 Multi-RAT Networks This chapter describes the calculations specific to 3GPP2 multi-RAT documents. and whether they are indoor or outdoor according to the percentage of indoor users per clutter class.ini file: [Simulation] RandomTotalUsers=0 In 3GPP2 multi-RAT documents. i. number of users of a user profile per km². Atoll determines the total number of users attempting connection in each simulation based on the Poisson distribution. Atoll performs two random trials. • Scheduling and Radio Resource Management as explained under "Simulation Process" on page 590. "Simulations Based on Sector Traffic Maps" on page 590. To have the same total number of users in each simulation of a group. services can be classified under constant bit rate and variable bit rate services.1 User Distribution During each simulation. The first random trial generates the number of users and their activity status as explained in the following sections depending on the type of traffic input. 0 Technical Reference Guide for Radio Networks Chapter 8: 3GPP2 Multi-RAT Networks ©Forsk 2015 User profile traffic maps: Each polygon or line of the map is assigned a density of users with a given user profile and mobility type. If the map is composed of points. If no best server can be found. the mobile will be considered rejected by LTE. the final numbers of users are obtains as follows: inactive Number of inactive users: n j inactive = Average  n j inactive CDMA  nj Number of users active on UL: n j  UL  = Average  n j  UL  CDMA Number of users active on DL: n j  DL  = Average  n j  DL  CDMA LTE   n j  UL  LTE  n j  DL  LTE Number of users active on UL+DL: n j  UL + DL  = Average  n j  UL + DL    CDMA  n j  UL + DL  LTE  Simulations Based on Sector Traffic Maps Sector traffic maps are also referred to as live traffic maps. Atoll calculates the number of active users of each service UL and DL as follows: • • LTE: "Simulations Based on Sector Traffic Maps" on page 478 CDMA: "Simulations Based on Sector Traffic Maps" on page 362. or the volume of the data transfer in the uplink and the downlink in each variable bit rate service session as explained in: • • LTE: "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 476 CDMA: "Simulations Based on User Profile Traffic Maps" on page 359. If no best server can be found.e. For any variable bit rate service (j). The number of users of each user profile is calculated from the surface area (SEnv) of each environment class map (or each polygon) and the user profile density (DUP). the average duration of each constant bit rate service session. N Users = S Env  D UP • In case of user profile traffic maps composed of lines. Atoll selects the highest priority as defined in the service assigned to each mobile. The steps of this algorithm are listed below.Atoll 3. Either throughput demands per service or the number of active users per service are assigned to the coverage areas of each transmitter. the number of service sessions.3. Atoll calculates the probability for a user being active at a given instant in the uplink and in the downlink according to the service usage characteristics described in the user profiles. The best server and technology assigned to each mobile remains unchanged for the rest of the simulation. Once the potential serving technologies have been identified. the mobile will be considered rejected by CDMA. i. LTE: Atoll determines the best server based on RSRP or RS level and the serving cell selection method. Atoll searches for a serving cell of each supported and available technology as follows: • • CDMA: Atoll determines a best server based on Ec/Io.2. • Technology-wise Monte Carlo simulations as explained in: • • 590 CDMA: "Simulations" on page 358 LTE: "Scheduling and Radio Resource Management" on page 552 . the number of users of each user profile is calculated from the line length (L) and the user profile density (DUP) (users per km): N Users = L  D UP • The number of users is a direct input when a user profile traffic map is composed of points. each point is assigned a number of users with given user profile and mobility type. 8. Each user profile contains a list of services and parameters describing how these services are accessed by the user. • Serving cell/technology selection For each mobile. User profiles model the behaviour of the different user categories.2 Simulation Process Each Monte Carlo simulation is a snap-shot of the network where resource allocation is carried out. once several numbers of users with different activity statuses have been calculated for different technologies. Live traffic data from the O&M is spread over the best server coverage areas of the transmitters included in the traffic map.. as explained in "Downlink Service Area Analysis" on page 413 and "Uplink Service Area Analysis" on page 417. as explained in "Effective Signal Analysis Coverage Predictions" on page 472. Two display options are available for this prediction: • • Technologies: Each pixel displays the colour representing the visible technology having the highest priority defined in the selected service. Application Throughput: The throughput on the application layer that a cell can provide to the selected terminal per pixel taking into account possible transmission errors (BLER) for the highest priority technology.Atoll 3. a mobility type. Max Application Throughput: the maximum throughput on the application layer that a cell can provide to the selected terminal per pixel taking into account possible transmission errors (BLER). the CDMA+LTE area shows the union between the CDMA and the LTE service areas as explained above. The properties of the non-interfering probe receiver are set by selecting a terminal.3 Multi-RAT Coverage Predictions Coverage predictions are calculated by determining the best server for each technology on each pixel and then determining the selected display parameter within the best server’s calculation area. Multi-RAT coverage predictions include: • • Effective Service Area Analysis (DL+UL) Coverage by Throughput (DL) Effective Service Area Analysis (DL+UL) The 3GPP2 multi-RAT effective service area is the combination of single-RAT effective service areas: • • CDMA Effective Service Area Analysis (Eb⁄Nt) (DL+UL) is based on a combination of downlink and uplink service area predictions. Max Effective RLC Throughput: The maximum throughput on the RLC layer that a cell can provide to the selected terminal per pixel taking into account possible transmission errors (BLER). 3GPP2 multi-RAT coverage predictions are combinations of corresponding single-RAT coverage predictions with specific parameter settings. Coverage by Throughput (DL) The 3GPP2 multi-RAT throughput prediction is the combination of single-RAT throughput predictions: • Service Area Analysis (Eb⁄Nt) (DL) explained in "Downlink Service Area Analysis" on page 413 The 3GPP2 multi-RAT effective RLC and application throughputs are respectively obtained from the effective RLC and application throughputs (see "Downlink Service Area Analysis" on page 300 for more information). considering all available technologies.3.0 Technical Reference Guide for Radio Networks Chapter 8: 3GPP2 Multi-RAT Networks AT330_TRR_E1 8. and a service. Each pixel within the calculation area is considered a non-interfering receiver. Four display options are available for this prediction: • • • • Effective RLC Throughput: The throughput on the RLC layer that a cell can provide to the selected terminal per pixel taking into account possible transmission errors (BLER) for the highest priority technology. see "Path Loss Calculation Prerequisites" on page 57 for more information). considering all available technologies. 591 . Available Technologies: Pixels display the colour representing the combined areas over which a multi-technology terminal can be served. For instance. The resolutions of coverage predictions do not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. • LTE Coverage by Throughput (DL) explained in "C/(I+N)-based Coverage Predictions" on page 473 The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the Effective RLC Channel Throughput (DL) and the Application Channel Throughput (DL) (see "C/(I+N)-based Coverage Predictions" on page 473 for more information). Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. LTE Effective Service Area Analysis (DL+UL) is based on a combination of downlink and uplink service area predictions. 3.0 Technical Reference Guide for Radio Networks Chapter 8: 3GPP2 Multi-RAT Networks 592 ©Forsk 2015 .Atoll 3. Chapter 9 TD-SCDMA Networks This chapter covers the following topics: • "Definitions and Formulas" on page 595 • "Signal Level Based Calculations" on page 602 • "Monte Carlo Simulations" on page 608 • "TD-SCDMA Prediction Studies" on page 626 • "Smart Antenna Modelling" on page 638 • "N-Frequency Mode and Carrier Allocation" on page 650 • "Neighbour Allocation" on page 651 • "Scrambling Code Allocation" on page 656 • "Automatic GSM/TD-SCDMA Neighbour Allocation" on page 666 . 0 Technical Reference Guidefor Radio Networks © Forsk 2015 594 .3.Atoll 3. and processes of the coverage predictions and the simulations available in TD-SCDMA documents.3. master and slave carriers. calculation parameters. the allocation of neigbours.28) Spread Global parameter None Minimum spreading factor (1) F Max Spread Global parameter None Maximum spreading factor (16) Proc Global parameter None P-CCPCH processing gain (13.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 9 TD-SCDMA Networks This chapter describes in detail the algorithms. location in the Atoll GUI.1 Inputs This table lists the inputs to computations.1.. 9.8 dB) N TS SF Global parameter None Number of timeslots per subframe (7) SF Global parameter ms Subframe duration (5) Frame Global parameter ms Frame duration (10) N Ch  TS GP Global parameter None Number of guard period chips per timeslot (16) N Ch  TS Data Global parameter None Number of data chips per timeslot (704) Midamble Global parameter None Number of midamble chips per timeslot (144) N Ch  PTS Global parameter None Number of guard period chips per pilot timeslot (96) N Ch  DwPTS GP Global parameter None Number of guard period chips per DwPTS timeslot (32) SYNC_DL Global parameter None Number of SYNC_DL chips per DwPTS timeslot (64) None Total number of chips per DwPTS timeslot (96) F Min G P – CCPCH D D N Ch  TS GP N Ch  DwPTS Total N Ch  DwPTS Global parameter Total N Ch  DwPTS GP SYNC_DL = N Ch  DwPTS + N Ch  DwPTS N Ch  UpPTS GP Global parameter None Number of guard period chips per UpPTS timeslot (32) SYNC_UL Global parameter None Number of SYNC_UL chips per UpPTS timeslot (128) None Total number of chips per UpPTS timeslot (160) N Ch  UpPTS Total N Ch  UpPTS Global parameter Total N Ch  UpPTS GP SYNC_UL = N Ch  UpPTS + N Ch  UpPTS 595 . which do not require simulation results. The third part describes the traffic scenario generation and Montel Carlo simulation algorithms including smart antenna modelling and dynamic channel allocation.Atoll 3. Name Value Unit Description R Ch Global parameter Mcps Chip rate (or Spreading rate) (1. The next sections are dedicated to TD-SCDMA coverage predictions which can be based on results obtained from simulations. and simulations. 9.e. and the allocation of scrambling codes. their significance. The last three sections describe in detail the allocation of frequencies. i. coverage predictions. Detailed explanation of the basic coverage predictions. is provided in the second part. and their usage. The first part of this chapter lists all the input and output parameters in the TD-SCDMA documents.1 Definitions and Formulas The tables in the following subsections list the input and output parameters and formulas used in simulations and other computations. 3.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks Name ©Forsk 2015 Value Unit Description Calculated global parameter Data W N Ch  TS W = --------------SF D bps Chip rate (140800 bps) F Avg Frequency band parameter MHz Average frequency range of the frequency band (2010) BW Frequency band parameter MHz Channel bandwidth of the carriers of a frequency band (1.Atoll 3.6) F IRF Cell parameter None Interference reduction factor F JD Site equipment parameter None Joint Detection (JD) factor TX Site equipment parameter None Multi-Cell Joint Detection factor Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) None BTS Noise Figure Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) None Transmitter loss L Tx = L Total – UL on uplink TX BTS parameter None Percentage of BTS signal correctly transmitted P TCH Max Cell parameter W Maximum cell traffic timeslot power P P – CCPCH Cell parameter W P-CCPCH power on TS0 P DwPCH Cell parameter W DwPCH power on DwPTS P OCCH – TS0 Cell parameter W Other common channel power on TS0 TComp P – CCPCH Cell parameter None P-CCPCH RSCP comparative threshold for baton handover P Max Cell parameter None Maximum difference between two transmitted powers Req Cell parameter None Required resource units in uplink RU DL Req Cell parameter None Required resource units in downlink P HS – PDSCH Available Cell parameter W HS-PDSCH power available per downlink timeslot P HR Cell parameter None Power headroom P HS – SCCH Cell parameter W HS-SCCH power per downlink timeslot N HS – SCCH Cell parameter None Number of HS-SCCH channels N HS – SICH Cell parameter None Number of HS-SICH channels Max Cell parameter None Maximum number of HSDPA users N HS-PDSCH Codes Min Cell parameter None Minimum number of HS-PDSCH codes Max Cell parameter None Maximum number of HS-PDSCH codes Max Cell parameter None Maximum number of intratechnology neighbours Max Cell parameter None Maximum number of intertechnology neighbours TX F MCJD NF L TX TX  RU UL N HSDPA N HS-PDSCH Codes N Intra – Neigh N Intra – Neigh 596 L Tx = L Total – DL on downlink . 0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Name Value Unit Description RSCP P – CCPCH Min Cell parameter or Global parameter W The minimum P-CCPCH RSCP required for a user to be connected to the cell P OCCH Timeslot parameter W Other common channel power DL Timeslot parameter W Downlink traffic power Timeslot parameter (Simulation constraint) None Maximum percentage of downlink used power Timeslot parameter (Simulation result) None Uplink load factor Timeslot parameter (Simulation constraint) None Maximum uplink load factor P HS – PDSCH Timeslot parameter W HS-PDSCH power available Min Timeslot parameter None Minimum number of HS-PDSCH codes N HS-PDSCH Codes Max Timeslot parameter None Maximum number of HS-PDSCH codes RU Overhead Timeslot parameter P TCH Max %P DL X UL Max X UL Available N HS-PDSCH Codes Overhead resource units Body Service parameter None Body loss Act Service parameter None Downlink activity factor for circuitswitched services and the A-DPCH activity factor for HSDPA services f UL Act Service parameter None Uplink activity factor for circuitswitched services and the A-DPCH activity factor for HSDPA services f DL Eff Service parameter None Downlink efficiency factor for circuitswitched services f UL Eff Service parameter None Uplink efficiency factor for circuitswitched services F Scaling Service parameter None Application througput scaling factor O TP Service parameter kbps Application throughput offset UL Service parameter (packet session modelling) None Average number of packet calls on the uplink during a session DL Service parameter (packet session modelling) None Average number of packet calls on the downlink during a session UL Service parameter (packet session modelling) ms Average time between two packet calls on the uplink T PacketCall DL Service parameter (packet session modelling) ms Average time between two packet calls on the downlink UL Service parameter (packet session modelling) KBytes Minimum packet call size on the uplink DL Service parameter (packet session modelling) KBytes Minimum packet call size on the downlink UL Service parameter (packet session modelling) KBytes Maximum packet call size on the uplink S Max – PacketCall DL Service parameter (packet session modelling) KBytes Maximum packet call size on the downlink T Packet UL Service parameter (packet session modelling) ms Average time between two packets on the uplink DL Service parameter (packet session modelling) ms Average time between two packets on the downlink L f DL N PacketCall N PacketCall T PacketCall S Min – PacketCall S Min – PacketCall S Max – PacketCall T Packet 597 .3.Atoll 3. 3.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks Name ©Forsk 2015 Value Unit Description UL Service parameter (packet session modelling) Bytes Packet size on uplink DL Service parameter (packet session modelling) Bytes Packet size on downlink Nom R99 bearer parameter kbps Downlink peak throughput Nom R99 bearer parameter kbps Uplink peak throughput WR99 bearer parameter (Can be calculated as ----------) Nom R DL None Downlink processing gain WR99 bearer parameter (Can be calculated as ----------) Nom R UL None Uplink processing gain Min R99 bearer parameter W Allowed minimum downlink traffic channel power Max R99 bearer parameter W Allowed maximum downlink traffic channel power N DL TS R99 bearer parameter None Number of downlink timelots TS R99 bearer parameter None Number of uplink timelots E Req C Req R99 bearer parameter per mobility (  ----b- or  --- )  N t TCH – UL  I  TCH – UL None Eb/Nt or C/I target on uplink E Req C Req R99 bearer parameter per mobility (  ----b- or  --- )  N t TCH – DL  I  TCH – DL None Eb/Nt or C/I target on downlink Req R99 bearer parameter per mobility W Target RSCP on uplink TCH Req R99 bearer parameter per mobility W Target RSCP on downlink TCH Div R99 bearer parameter per mobility None Downlink diversity gain Div R99 bearer parameter per mobility None Uplink diversity gain Term Terminal parameter W Maximum terminal power P Min Term Terminal parameter W Minimum terminal power P UpPCH Terminal parameter W UpPCH power Term Terminal parameter None Terminal Noise Figure Term Terminal parameter None Joint Detection (JD) factor Term Terminal parameter None Percentage of terminal signal correctly transmitted Term Terminal parameter None Terminal gain Term Terminal parameter None Terminal loss TAdd P – CCPCH Mobility parameter W Required RSCP T_Add for P-CCPCH TDrop P – CCPCH Mobility parameter W Required RSCP T_Drop for P-CCPCH Req Mobility parameter W Required RSCP threshold for DwPCH Req Mobility parameter W Required RSCP threshold for UpPCH E Req C Req Mobility parameter (  ----b- or  --- )  N t P – CCPCH  I  P – CCPCH None Required quality threshold for PCCPCH S Packet S Packet R DL R UL Proc G DL Proc G UL P TCH – DL P TCH – DL N UL Req Q TCH – UL Req Q TCH – DL RSCP TCH – UL RSCP TCH – DL G DL G UL P Max NF F JD  G L RSCP DwPCH RSCP UpPCH Req Q P – CCPCH 598 . 38 x 10-23 J/K Boltzman constant T 293 K Ambient temperature TX NFTX  K  T  BW W Thermal noise at transmitter Term NF Term  K  T  BW W Thermal noise at terminal TX Antenna parameter None Transmitter antenna gain Propagation model result None Path loss Result calculated from cell edge coverage probability and model standard deviation None Model shadowing margin used in coverage predictions P – CCPCH Result calculated from cell edge coverage probability and P-CCPCH Eb/Nt standard deviation None P-CCPCH Eb/Nt shadowing margin used in coverage predictions  Eb  Nt  DL Result calculated from cell edge coverage probability and DL Eb/Nt standard deviation None DL Eb/Nt shadowing margin used in coverage predictions  Eb  Nt  UL Result calculated from cell edge coverage probability and UL Eb/Nt standard deviation None UL Eb/Nt shadowing margin used in coverage predictions Req Q HS – SCCH Req Q HS – SICH Req Q DwPCH  Model Eb/Nt  P – CCPCH or CI  P – CCPCH CI Eb/Nt or  DL Eb/Nt or  UL  DL CI  UL L Indoor F DL N0 N0 G L Path Model M Shadowing M Shadowing M Shadowing M Shadowing 599 .3. optionally.Atoll 3. frequency band) parameter None Indoor loss Ortho Clutter class parameter None Downlink orthogonality factor F UL Ortho Clutter class parameter None Uplink orthogonality factor  Spread Clutter class parameter ° Spreading angle K 1.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Name Value Unit Description E Req Mobility parameter (  ----c- )  N t HS – SCCH None Required quality threshold for HSSCCH E P – CCPCH Mobility parameter (  ----c- )  N t HS – SICH None Required quality threshold for PCCPCH C Req Mobility parameter (  --- )  I  DwPCH None Required quality threshold for DwPCH Clutter class parameter None Model standard deviation Clutter class parameter None P-CCPCH Eb/Nt or C/I standard deviation Clutter class parameter None Downlink Eb/Nt or C/I standard deviation Clutter class parameter None Uplink Eb/Nt or C/I standard deviation Clutter (and. only carrier power is  Eb  Nt  UL For P-CCPCH Eb/Nt calculation TX Term attenuated by M Shadowing .0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Name Value Unit For RSCP calculation Model LT TX Term Body Transmitter-terminal total loss in coverage predictions Model M Shadowing L Path  L  L L  L Indoor  = ---------------------------------------------------------------------------------------------------------------------TX Term G G In UL.Atoll 3. carrier power and intra-cell interference are attenuated by M Shadowing or M Shadowing while  Eb  Nt  DL Body L Path  L  L L  L Indoor  M Shadowing = ---------------------------------------------------------------------------------------------------------------------TX Term G G TX Description  Eb  Nt UL Body DL Therefore.3 DwPCH C/I Calculation Name Value TX i TX i  ic  TX i  ic  N Tot – DL 600 TX i  ic    RSCP DwPCH ------------------------------------------TX i  ic  N Tot – DL C ---  I  DwPCH TX i  ic  TX i  ic  Term I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0 . G Proc P – CCPCH TX i  ic  N Tot – DL TX i  ic    RSCP P – CCPCH -----------------------------------------------TX i  ic  N Tot – DL TX TX i  ic  N Tot – DL TX i  ic  Term I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0 TX  ic  i RSCP P – CCPCH     TX i  ic  I Intra – DL With  TX i =  TX i TX i  ic  I IC – DL  ic jc  TX i TX  ic  i + RSCP OCCH – TS0   Ortho   1 – F DL   =  0  1 I Extra – DL None P-CCPCH Eb/Nt for the cell TX i  ic  None P-CCPCH C/I for the cell TX i  ic  W Downlink total noise for the cell TX i  ic  W Downlink intra-cell interference for the cell TX i  ic  W Downlink extra-cell interference for the cell TX i  ic  W Inter-carrier interference Unit Description None DwPCH C/I for the cell TX i  ic  W Downlink total noise for the cell TX i  ic  TX i  ic  TX i TX i  ic  Description TX  ic  i i C ---  I  P – CCPCH Unit Term    1 – F JD TX i  and Without Useful Signal Total Noise TX j  ic  TX j  ic  TX j  jc  TX j  jc    RSCPP – CCPCH + RSCPOCCH – TS0 ji   RSCPP – CCPCH + RSCPOCCH – TS0 TX j --------------------------------------------------------------------------------------F IRF  ic jc  9.1. 9. P – CCPCH Body L Path  L  L L  L Indoor  M Shadowing = ---------------------------------------------------------------------------------------------------------------------TX Term G G  Eb  Nt  P LT LT None For DL Eb/Nt calculation  Eb  Nt  DL LT TX Term  Eb  Nt  UL LT  Eb  Nt  DL P – CCPCH extra-cell interference is not.2 P-CCPCH Eb/Nt and C/I Calculation Name Value E b TX i  ic   --- N t P – CCPCH   RSCP P – CCPCH -----------------------------------------------.  Eb  Nt  For UL Eb/Nt calculation Term In DL. M Shadowing or L Path  L  L L  L Indoor  M Shadowing = ---------------------------------------------------------------------------------------------------------------------TX Term G G P – CCPCH M Shadowing are set to 1 in DL extracell interference calculation.3.1. or P Term P Max  --------------------------Max  ------------------------TX i  ic  TX i  ic  E C b  ---  -----  I  TCH – UL  N t TCH – UL E b  --- N t TCH – UL Term Req TX i  ic  Req 601 .3.4 DL TCH Eb/Nt and C/I Calculation Name Value TX i  ic  TX i TX i  ic    RSCP TCH – DL Div ---------------------------------------------. G Div UL TX i  ic  N Tot – UL Term P Req Q TCH – UL Q TCH – UL Term . G Div DL TX i  ic  N Tot – DL E b  --- N t TCH – DL TX i  ic  N Tot – DL TX i  ic  I Intra – DL TX i  ic  I Extra – DL I IC – DL  ic jc  TX i  ic  TX i C ---  I  TCH – DL TX i  ic  TX i  ic  Term I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0 TX i   TXi Ortho Term       1 – F DL    1 – F JD  +  1 –      TX  ic  TX  ic  i i   RSCP TCH – DL + RSCP OCCH    TX j  ic  TX j  ic  TX j  jc  TX j  jc    RSCPTCH – DL + RSCPOCCH  ji   RSCPTCH – DL + RSCPOCCH  TX j --------------------------------------------------------------------------F IRF  ic jc  9. G Proc UL  G UL TX i  ic  N Tot – UL i C ---  I  TCH – UL TX  ic    RSCP TCH – UL ------------------------------------------------.1.Atoll 3. G Proc DL  G DL TX i  ic  N Tot – DL TX i  ic    RSCP TCH – DL ---------------------------------------------.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Name Value TX  ic  i RSCP DwPCH     TX  ic  i I Intra – DL With  TX i =  TX i Ortho   1 – F DL   =  0  1 TX Description W Intra-cell interference for the cell TX i  ic  W Extra-cell interference for the cell TX i  ic  W Inter-carrier interference Unit Description None Downlink TCH Eb/Nt for the cell TX i  ic  None Downlink TCH C/I for the cell TX i  ic  W Downlink total noise for the cell TX i  ic  W Downlink intra-cell interference for the cell TX i  ic  W Downlink extra-cell interference for the cell TX i  ic  W Inter-carrier interference Unit Description None Uplink TCH Eb/Nt for the cell TX i  ic  None Uplink TCH C/I for the cell TX i  ic  W Uplink required power for the terminal i Term    1 – F JD  and Without Useful Signal Total Noise TX j  ic    RSCPDwPCH TX i  ic  I Extra – DL Unit ji TX j  jc    RSCPDwPCH I IC – DL  ic jc  TX j ---------------------------------------F IRF  ic jc  9.1.5 UL TCH Eb/Nt and C/I Calculation Name Value TX i  ic  Term TX i  ic    RSCP TCH – UL Div ------------------------------------------------. 1.1 Point Analysis For the selected transmitted TXi and carrier (ic).2. RSCP is the received signal code power for the P-CCPCH.1.7 HSDPA Dynamic Power Calculations Name Value TX i  ic  TX  ic  TX  ic  Ec i i i  ---   N t HS – SCCH   N Tot – DL –     RSCP HS – SCCH ----------------------------------------------------------------------------------------------------------------------------. Point Analysis: Real-time calculations for profile and reception analysis using the mouse to move a probe mobile on the map. L Model T TX i  TX i  ic  P HS – SCCH TX  ic  i TX  ic  i TX TX  ic  i TX  ic  i P Max – DL – Eff – P R99 – DL – P HR P HS – PDSCH TX i  ic  TX  ic  TX  ic  i – P HS – SCCH Ec i i i  ---   N t HS – SICH   N Tot – UL –     RSCP HS – SICH -------------------------------------------------------------------------------------------------------------------------.Atoll 3.2 Signal Level Based Calculations Two types of signal level based calculations are available in Atoll: 1. 2. you can study three parameters in point analysis Profile tab: Study criteria Formulas Signal level received from a transmitter on a carrier (cell) Signal level ( RSCP ) in dBm RSCP TX  ic  i TX  ic  i Model – L Path – M Shadowing – L Indoor TX i Path loss ( L Path ) in dB Total losses ( L T ) in dB = EIRP L Path = L Model + L Ant L T = L Path + L TX i Model + L Indoor + M Shadowing – G TX i Where. L Model T Mi  Mi P HS – SICH M M 9. EIRP TX i  ic  TX i  ic  = P P – CCPCH + G TX i –L TX i ic is a carrier number L Model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model 602 . RSCP Based Coverage Predictions: Calculation of RSCP related parameters on each pixel and colouring according to the selected display. 9.3. EIRP is the effective isotropic radiated power of the transmitter.6 Interference Calculation Name Value TX  jc  j I C2C  TX i TX j  TX j  ic  TX j  ic  Description W Cell to cell interference W UpPCH interference Unit Description W HS-SCCH power W HS-PDSCH power W HS-SICH power TX  jc  j   RSCPTCH – DL + RSCPOCCH  j   RSCPTCH – DL + RSCPOCCH  +--------------------------------------------------------------------------F IRF  ic jc  TX Unit TX j TX i TX i  ic  N0 I TS1 – UL TX i  ic  X TS1 – UL  ---------------------------------TX  ic  1 – X i  TS1 – UL  9.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 9. . Only bars for transmitters whose signal level is within a 30 dB margin from the best server signal are displayed.2 RSCP Based Coverage Predictions For each TBC transmitter. 9. Each pixel within the TXi calculation area is considered a probe receiver.2. You can use a value other than 30 dB for the margin from the best server signal level. 603 . or the total losses.3. or the highest signal level received on the best carrier. 9. taken into account when the option "Indoor coverage" is selected G L TX i TX i is the transmitter antenna gain is the transmitter loss ( L TX i = L Total – DL ) It is possible to analyse the best carrier. Coverage predictions are calculated using bilinear interpolation of multi-resolution path loss matrices (similar to the evaluation of site altitudes).0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 TX i L Ant is the transmitter antenna attenuation (from antenna patterns) Model M Shadowing is the shadowing margin. In this case. TXi. or the highest signal level received from the selected transmitter on the best carrier. For a selected transmitter. Coverage study parameters to be set are: • • The study conditions to determine the service area of each TBC transmitter The display settings to for colouring the covered pixels Atoll uses the parameters entered in the Condition tab of the coverage study properties dialogue to determine pixels covered by the each transmitter.1 Calculation Criteria The RSCP from a transmitter TXi and a selected carrier (ic) is given by: RSCP TX i  ic  = EIRP TX i  ic  Model – L Path – M Shadowing – L Body – L Indoor + G Term –L Term Where. 9. Atoll determines the value of the selected parameter on each studied pixel inside the TXi calculation area.Atoll 3. L T . and can be different for each coverage prediction. Atoll takes the highest P-CCPCH power of cells to calculate the signal level received from a transmitter. for example a smaller value for improving the calculation speed.2.1 Profile Tab TX i  ic  Atoll displays either the signal level received from the selected transmitter on a carrier ( RSCP P – CCPCH ). Received signal level bar graphs are displayed in a decreasing signal level order. Coverage prediction display resolution is independent of the path loss matrix and geographic data resolutions.2 Reception Tab Analysis provided in the Reception tab is based on path loss matrices. This parameter is taken into account when the option “Shadowing taken into account” is selected L Indoor are the indoor losses. L Path .1. it is also possible to study the path loss.1.2. 9.2. For more information on defining a different value for this margin. ( RSCP P – CCPCH ). Atoll displays either the signal level received on a carrier. The number of bars in the graph depends on the signal level received from the best server. see the Administrator Manual. Path loss and total losses are the same on any carrier. You can study reception from TBC transmitters for which path loss matrices have been calculated on their calculation areas.2. TX i  ic  For each transmitter. Atoll 3. This parameter is taken into account when the option “Shadowing taken into account” is selected L Indoor are the indoor losses. a service. Atoll calculates the RSCP considering: • • • the preferred carrier of the selected service. ic is a carrier number TX i L Path = L Model + L Ant L Model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model TX i L Ant is the transmitter antenna attenuation (from antenna patterns) Model M Shadowing is the shadowing margin. 604 . There are as many layers as defined thresholds. Each layer corresponds to an area where the RSCP from the best server exceeds a defined minimum threshold. or radiated power TX  ic  i EIRP DL – TCH = of the TX  ic  i P DL – TCH +G transmitter. If you perform this coverage prediction for the best carrier. • RSCP Margin (dB) Coverage consists of several layers with a layer per user-defined RSCP margin defined in the Display tab (Prediction TX i  ic  RSCP properties). A pixel of a service area is coloured if the RSCP level is greater than or equal to the defined thresholds.1 Coverage Condition This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the P-CCPCH. TX i –L TX i TX  ic  i TX  ic  i EIRP P – CCPCH = P P – CCPCH + G TX i –L TX i . area is covered if RSCP P – CCPCH – TAdd P – CCPCH  Mobility   M P – CCPCH . You can select the display colours according to the RSCP. or the master carrier in case of N-frequency mode compatible transmitters.2 Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • Best Signal Level (dBm) TX i  ic  Atoll calculates the best RSCP P – CCPCH received from each transmitter TX i  ic  on each pixel.2. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest PCCPCH power. For each layer. The pixel colour depends on the RSCP level. or from the master carrier in case of N-frequency mode compatible transmitters. there will not be any pixels covered by this transmitter.2. if no preferred carrier is defined for the service. If the selected carrier does not exist on a transmitter. Afterwards.2.2. or the carrier with the highest P-CCPCH power. Atoll chooses the highest RSCP. EIRP is the TX  ic  i EIRP DwPCH = effective TX  ic  i P DwPCH +G isotropic TX i –L TX i . . The coverage prediction is calculated for a given set of a terminal type. a mobility type.2.2. Each layer is assigned a colour and displayed with intersections between layers. DwPCH. the coverage prediction is calculated for the selected carrier. and for TS0. taken into account when the option "Indoor coverage" is selected L Term is the terminal loss L Body is the body loss defined in the service G G L Term TX i TX i is the receiver total gain is the transmitter antenna gain is the transmitter loss ( L TX i = L Total – DL ) 9.2 P-CCPCH RSCP Coverage Prediction 9.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 RSCP is the received signal code power.2. Coverage consists of several independent layers whose visibility in the workspace can be managed. or on any best server parameter. a carrier. Where other service areas overlap the studied one. RSCP can be calculated for P-CCPCH. 9.2. or the downlink TCH.3. The coverage prediction is calculated for a given set of a terminal type. the coverage prediction is calculated for the selected carrier. there will not be any pixels covered by this transmitter.2. or from the master carrier in case of N-frequency mode compatible transmitters.2. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. a service. Afterwards. there will not be any pixels covered by this transmitter. The pixels in the TX i  ic  TX i  ic  RSCP P – CCPCH = TX i  ic  Min coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH. 9. If the selected carrier does not exist on a transmitter.1 Coverage Condition This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the DwPCH.2. Coverage consists of several independent layers whose visibility in the workspace can be managed. there will not be any pixels covered by this transmitter. with different cell edge coverage probabilities. or the carrier with the highest P-CCPCH power. Atoll calculates the RSCP considering: • • • the preferred carrier of the selected service. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. or from the master carrier in case of N-frequency mode compatible transmitters. If you perform this coverage prediction for the best carrier. If you perform this coverage prediction for the best carrier. a service. a carrier. Atoll calculates the RSCP considering: • • • the preferred carrier of the selected service.4 P-CCPCH Pollution Analysis Coverage Prediction This coverage prediction calculates and displays the number of P-CCPCH polluters. Afterwards. or the master carrier in case of N-frequency mode compatible transmitters.2. and for DwPTS. 9.2. Afterwards. a mobility type. The coverage prediction is calculated for a given set of a terminal type. or the carrier with the highest P-CCPCH power.3 Best Server P-CCPCH Coverage Prediction This coverage prediction calculates and displays the best server RSCP for the P-CCPCH. a mobility type.5 DwPCH RSCP Coverage Prediction 9. if no preferred carrier is defined for the service. If you perform this coverage prediction for the best carrier. or the master carrier in case of N-frequency mode compatible transmitters. Atoll calculates the Received Signal Code TX i  ic  Power (RSCP) for the P-CCPCH for each pixel in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and determines the polluting transmitters according to: TX i  ic  TX j  jc  RSCP P – CCPCH  Best  RSCP P – CCPCH – M   ji Where M is the specified pollution margin. the coverage prediction is calculated for the selected carrier.2. if no preferred carrier is defined for the service.5. a service. a carrier.RSCP P – CCPCH) and where Best  RSCP TXj  jc   will be covered and coloured according to the transmitter colour. if no preferred carrier is defined for the service. or from the master carrier in case of N-frequency mode compatible transmitters. There is one coverage area per transmitter in the explorer. Atoll calculates the RSCP considering: • • • the preferred carrier of the selected service. and for TS0. or the carrier with the highest P-CCPCH power. the coverage prediction is calculated for the selected carrier.2. a mobility type.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 • Cell Edge Coverage Probability (%) TX  ic  i On each pixel of each transmitter service area. If the selected carrier does not exist on a transmitter. or the master carrier in case of N-frequency mode compatible transmitters.Atoll 3.3. Each layer corresponds to an area where the number of servers is greater than or equal to a defined minimum threshold. and for TS0. If the selected carrier does not exist on a transmitter. Atoll determines the number of transmitters covering each pixel and colours the pixel according to the number of polluting transmitters. the coverage corresponds to the pixels where the RSCP P – CCPCH from the transmitter exceeds TAdd P – CCPCH defined in the mobility selected in the Conditions tab.2. P – CCPCH j = All  9. The coverage prediction is calculated for a given set of a terminal type. There are as many layers as defined thresholds. 605 . a carrier. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest PCCPCH power. 2. 9. Atoll chooses the highest RSCP. the coverage corresponds to the pixels where the RSCP DwPCH from TX i  ic  the transmitter TX i  ic  exceeds RSCP DwPCH defined in the mobility selected in the Conditions tab. A pixel of a service area is coloured if TX i  ic  Req RSCP DwPCH  RSCP DwPCH  Mobility  .2. a mobility type.1 Coverage Condition This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the UpPCH in the uplink.. or the master carrier in case of N-frequency mode compatible transmitters.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks The pixels in the TX i  ic  TX  ic  i ©Forsk 2015 TX  ic  i Min coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH. or from the master carrier in case of N-frequency mode compatible transmitters. 9. TX i  ic  Min Term Req The pixels where RSCP P – CCPCH  Max (TAdd P – CCPCH.2 Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter. Each layer corresponds to an area where the RSCP at the best server exceeds a defined minimum threshold. Where other service areas overlap the studied one. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. • RSCP Margin (dB) Coverage consists of several layers with a layer per user-defined RSCP margin defined in the Display tab (Prediction TX  ic  i Req RSCP properties).2.2.RSCP P – CCPCH) and where RSCP UpPCH  RSCP UpPCH  Mobility  are covered and coloured according to the selected display parameter. Each layer is assigned a colour and displayed with intersections between layers. Each layer corresponds to an area where the RSCP from the best server exceeds a defined minimum threshold. if no preferred carrier is defined for the service.2.2.6. If you perform this coverage prediction for the best carrier.2. 9. or the carrier with the highest P-CCPCH power. There is one coverage area per transmitter in the explorer. Where other service areas overlap the studied one.Atoll 3. Atoll calculates the RSCP considering: • • • the preferred carrier of the selected service. a service.6 UpPCH RSCP Coverage Prediction 9. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • DwPCH RSCP (dBm) TX i  ic  Atoll calculates the best RSCP DwPCH received from each transmitter TX i  ic  on each pixel. area is covered if RSCP DwPCH – RSCP DwPCH  Mobility   M DwPCH . The pixel colour depends on the RSCP level. Atoll chooses the highest RSCP.5. there will not be any pixels covered by this transmitter. • 606 RSCP Margin (dB) .RSCP P – CCPCH) and where Req RSCP DwPCH  RSCP DwPCH  Mobility  are covered and coloured according to the selected display parameter.2 Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter. the coverage prediction is calculated for the selected carrier. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • UpPCH RSCP (dBm) Term Atoll calculates the best RSCP UpPCH received from each pixel of each transmitter service area at the transmitter.6. • Cell edge coverage probability (%) TX  ic  i On each pixel of each transmitter service area. Atoll uses the UpPCH power of the selected terminal to calculate the RSCP from each pixel of each transmitter’s best server coverage area. The coverage prediction is calculated for a given set of a terminal type. a carrier. Coverage consists of several independent layers whose visibility in the workspace can be managed.2.3. If the selected carrier does not exist on a transmitter. A pixel of a service area is coloured Term Req if RSCP UpPCH  RSCP UpPCH  Mobility  . and for UpPTS. The pixel colour depends on the RSCP level. For each layer. There are as many layers as defined thresholds. There are as many layers as defined thresholds. with different cell edge coverage probabilities. Afterwards. Coverage consists of several independent layers whose visibility in the workspace can be managed. Atoll calculates the Received Signal Code Power (RSCP) for the P-CCPCH for each pixel in the TX i  ic  coverage area where TX  ic  i RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and determines the interfering transmitters according to: TX i  ic  TX j  jc  RSCP P – CCPCH  Best  RSCP P – CCPCH – M   ji Where M is the specified pollution margin.2. Coverage consists of several independent layers whose visibility in the workspace can be managed. and for TS0. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest PCCPCH power. a service. If the pixel is interfered.7 Baton Handover Coverage Prediction 9. with different cell edge coverage probabilities. there will not be any pixels covered by this transmitter. If you perform this coverage prediction for the best carrier. the coverage corresponds to the pixels from which the RSCP UpPCH at Term the transmitter exceeds RSCP UpPCH defined in the mobility selected in the Conditions tab. pixels TX  ic  i RSCP P – CCPCH 9. • Cell Edge Coverage Probability (%) Term On each pixel of each transmitter service area. • Handover Areas Atoll displays the pixels where there are transmitters other than the best server that satisfy the above criteria. area is covered if RSCP UpPCH – RSCP UpPCH  Mobility   M UpPCH . if no preferred carrier is defined for the service. a carrier.2 are covered and coloured  TAdd P – CCPCH  Mobility  and according TX  jc  j RSCP P – CCPCH to the selected  TDrop P – CCPCH  Mobility  – display parameters. 9. If the selected carrier does not exist on a transmitter.2. a mobility type. The coverage prediction is calculated for a given set of a terminal type.2. Atoll colours it according to the colour assigned to the scrambling code in the display parameters. and for TS0. Coverage Display It is possible to display the potential handover areas or the number of transmitters covering each pixel. or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards. Atoll calculates the RSCP considering: • • • The the preferred carrier of the selected service. If you perform this coverage prediction for the best carrier. Atoll determines whether the cells of two transmitters covering a pixel have the same scrambling code. a service. if no preferred carrier is defined for the service. Coverage consists of a single layer with a defined colour whose visibility in the workspace can be managed.7.2. a mobility type. For each layer.2. 9. or the carrier with the highest P-CCPCH power.2.1 Coverage Condition This coverage prediction determines the pixels which receive RSCP from cells other than the best server high enough to perform baton handovers.8 Scrambling Code Interference Analysis This coverage prediction calculates and displays the pixels covered by two cells using the same scrambling code. • Number of Potential Servers Atoll determines the number of transmitters covering each pixel and colours the pixel according to the number of transmitters. The coverage prediction is calculated for a given set of a terminal type. There are as many layers as defined thresholds.Atoll 3. There is one coverage area per transmitter in the explorer. Each layer is assigned a colour and displayed with intersections between layers. the coverage prediction is calculated for the selected carrier.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Coverage consists of several layers with a layer per user-defined RSCP margin defined in the Display tab (Prediction Term Req RSCP properties).2. TX  jc  j TComp P – CCPCH where . Atoll calculates the RSCP considering: • • • the preferred carrier of the selected service.7. or the master carrier in case of N-frequency mode compatible transmitters. a carrier. there will not be any pixels covered by this transmitter.2. or the carrier with the highest P-CCPCH power. Afterwards. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. or the master carrier in case of N-frequency mode compatible transmitters.3. Received Signal Code Power (RSCP) is calculated for the P-CCPCH. Coverage 607 . the coverage prediction is calculated for the selected carrier. If the selected carrier does not exist on a transmitter. or from the master carrier in case of N-frequency mode compatible transmitters. Each layer corresponds to an area where the number of servers is greater than or equal to a defined minimum threshold. The user profile models the behaviour of the different user categories.e. and a mobility type.1. The first random trial generates the number of users and their activity status as explained in the following sections depending on the type of traffic input.3. User profile traffic maps: Each polygon or line of the map is assigned a density of users with a given user profile and mobility type.3 Monte Carlo Simulations The simulation process is divided into two steps. number of users of a user profile per km². • Generating a realistic user distribution as explained in "Generating a Realistic User Distribution" on page 608. 9. Each user profile contains a list of services and their associated parameters describing how these services are accessed by the user. i. Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data. If one of these two mobiles is rejected for some reason. 9.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 consists of several independent layers whose visibility in the workspace can be managed.ini file: [Simulation] RandomTotalUsers=0 Each user is randomly assigned a service. If the map is composed of points. This may lead to slight variations in the total numbers of users in different simulations. The resulting user distribution complies with the traffic database and maps selected when creating simulations. a terminal. a second random trial is performed to obtain their geographical locations weighted according to the clutter classes. There are as many layers as scrambling codes. the interference in the network.e.Atoll 3. Users accessing circuit-switched services transmit on both uplink and downlink simultaneously.. the other is also rejected due to the same reason. Atoll determines the total number of users attempting connection in each simulation based on the Poisson distribution. i.. 9. mobiles connected in uplink and downlink both.1 Simulations Based on User Profile Traffic Maps User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of user profiles with an associated mobility type and a given density. In TD-SCDMA networks users accessing packet-switched services can transmit either on uplink or on downlink. To have the same total number of users in each simulation of a group. and whether they are indoor or outdoor according to the percentage of indoor users per clutter class defined for the traffic maps. The number of users of each user profile is calculated from the surface area (SEnv) of each environment class map (or each polygon) and the user profile density (DUP).3. N Users = S Env  D UP 608 . Atoll also calculates the shadowing margin for each user based on the standard deviations defined for the clutter class of each user. A layer corresponding to areas where more than one scrambling code interferes is also available. Both active and inactive users use radio resources and generate interference. The activity status is determined based on the calculations of activity probabilities using the traffic inputs. Circuitswitched service users. are modelled in Atoll by two mobiles generated at the same location with one connected on the uplink and the other on the downlink. each point is assigned a number of users with given user profile and mobility type. add the following lines in the Atoll.3. "Simulations Based on Sector Traffic Maps" on page 612. • • "Simulations Based on User Profile Traffic Maps" on page 608. Once all the user characteristics have been determined. Each layer corresponds to the area where the corresponding scrambling code has interference. Atoll performs two random trials.1 Generating a Realistic User Distribution During each simulation. but never on both simultaneously. The user activity status influences the next step of the simulation. • Dynamic channel allocation and power control as explained under "Power Control Simulation" on page 613. The number of users and their distribution per activity status is determined as follows: • Calculation of the service usage duration per hour ( p 0 : probability of a connection): N call  d p o = ------------------3600 • Calculation of the number of users trying to access the service i ( n i ): n i = N Users  p 0 The activity status of each user depends on the activity periods during the connection.1. i. the number of voice calls or data sessions. The average number of calls per hour N Call . or the volumes of the data exchanged in the uplink and the downlink in each data session. f Act and f Act .. Each packet call is defined by its size and may be divided in packets of fixed size (1500 Bytes) separated by an inter-packet arrival time.2 Packet Switched Service (j) User profile parameters for packet switched services are: • • The user terminal equipment used for the service (from the Terminals table).Atoll 3. Atoll calculates the probability for a user being active in the uplink and in the downlink according to the service usage characteristics described in the user profiles.1.3. The average number of packet sessions per hour N Sess . • The volume (in kBytes) which is transferred on the downlink V DL and the uplink V UL during a session.1 Circuit Switched Service (i) User profile parameters for circuit switched services are: • • The user terminal equipment used for the service (from the Terminals table). or active on DL only. the average duration of each voice call. • The average duration of a call (seconds) D Call .3. the uplink and downlink activity UL DL factors defined for the circuit switched service i.e.3. • Calculation of activity probabilities: UL DL Probability of being inactive: p Inactive =  1 – f Act    1 – f Act  UL UL DL DL DL UL Probability of being active on UL: p Active = f Act   1 – f Act  Probability of being active on DL: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active both on UL and DL: p Active = f Act  f Act • Calculation of number of users per activity status: Number of inactive users: n i – Inactive = n i  p Inactive UL UL Number of users active in the uplink: n i – Active = n i  p Active DL DL Number of users active in the downlink: n i – Active = n i  p Active UL + DL UL + DL Number of users active in the uplink and downlink both: n i – Active = n i  p Active Therefore. 609 . a connected user can be either active on both links. active on UL only. A packet session consists of several packet calls separated by a reading time.1.e. inactive on both links.. i. 9. At any given instant.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 • In case of user profile traffic maps composed of lines. 9. the number of users per user profile is calculated from the line length (L) and the user profile density (DUP) (users per km): N Users = L  D UP • The number of users is an input when a user profile traffic map is composed of points.1. 0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Figure 9. Calculation of the average duration of inactivity within a packet call (c): UL UL DL DL  N Packet – 1   T Packet  N Packet – 1   T Packet UL DL .3. f Eff and f Eff are the uplink and downlink A-DPCH activity factors. D Connection 3600 3600 Calculation of the probability of being connected: 610 . the average duration of a connection in the session s is: UL UL UL DL DL DL D Connection =  D Activity  Session +  D Inactivity  Session and D Connection =  D Activity  Session +  D Inactivity  Session Calculation of the service usage duration per hour (probability of a connection): N Sess N Sess UL UL DL DL p Connection = -----------.and  D Inactivity  PacketCall = -------------------------------------------------------- D Inactivity  PacketCall = --------------------------------------------------------1000 1000 Calculation of the average duration of inactivity in a session (s): UL UL UL DL DL DL  D Inactivity  Session = N PacketCall   D Inactivity  PacketCall and  D Inactivity  Session = N PacketCall   D Inactivity  PacketCall Calculation of the average duration of activity in a session (s): UL UL DL DL N Packet  S Packet  8 UL UL .and S PacketCall = --------------------------------------UL UL DL DL N PacketCall  f Eff N PacketCall  f Eff UL DL In case of HSDPA services. D Connection and p Connection = -----------. Calculation of the average number of packets per packet call: UL DL  S PacketCall   S PacketCall  UL - + 1 and N DL - + 1 N Packet = Int  ------------------------------Packet = Int  ------------------------------UL  S Packet  1024  S DL Packet  1024 1 kBytes = 1024 Bytes. respectively.and  D Activity  Session = N PacketCall  -----------------------------------------------UL R Nom  1000 N Packet  S Packet  8 DL DL  D Activity  Session = N PacketCall  -----------------------------------------------DL R Nom  1000 Therefore.Atoll 3.1: Description of a Packet Session Calculation of the average packet call size (kBytes): UL DL UL DL V V S PacketCall = -------------------------------------. there can be three possible cases when a user is connected: a. 1st case: At a given time. packet are uploaded only. we have: a.and f = --------------------------------------------------------------------------------------UL UL DL DL  D Inactivity  Session +  D Activity  Session  D Inactivity  Session +  D Activity  Session Therefore. inactive on both links. the activity periods during the connection are taken into account. 1st case: At a given time.3. packets are downloaded and uploaded.1 on page 610 shows. DL The probability of the user being active on DL and inactive on UL: p1 Active = f DL DL  p Connected DL DL The probability of the user being inactive on both UL and DL: p3 Inactive =  1 – f   p Connected Calculation of number of users per activity status: Number of inactive users on UL and DL: n j – Inactive = n j   p1 Inactive + p2 Inactive + p3 Inactive  UL UL UL DL DL DL Number of users active on UL and inactive on DL: n j – Active = n j   p1 Active + p2 Active  Number of users active on DL and inactive on UL: n j – Active = n j   p1 Active + p3 Active  UL + DL UL + DL Number of users active on UL and DL: n j – Active = n j   p1 Active  Therefore. packet are downloaded only. packets are downloaded and uploaded. packet are downloaded only. 611 . or active on DL only. packet are uploaded only.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 UL DL p Connected = 1 –  1 – p Connection    1 – p Connection  Therefore. UL DL p Connection  p Connection UL + DL The probability of being connected is: p Connected = --------------------------------------------------------p Connected b. 3rd case: At a given time. active on UL only. 2nd case: At a given time. a connected user can be active on both links. UL The probability of the user being active on UL and inactive on DL: p2 Active = f UL UL  p Connected UL UL The probability of the user being inactive on both UL and DL: p2 Inactive =  1 – f   p Connected c. DL UL p Connection   1 – p Connection  DL The probability of being connected is: p Connected = ----------------------------------------------------------------------p Connected Calculation of the probability of being active: To determine the activity status of each user. 3rd case: At a given time. UL DL p Connection   1 – p Connection  UL The probability of being connected is: p Connected = ----------------------------------------------------------------------p Connected c. the number of users trying to access the service j is: n j = N Users  p Connected As Figure 9.Atoll 3. DL UL + DL UL UL + DL UL UL   1 – f   p Connected DL DL   1 – f   p Connected DL  p Connected The probability of the user being active on UL and inactive on DL: p1 Active = f The probability of the user being active on DL and inactive on UL: p1 Active = f UL + DL The probability of the user being active on both UL and DL: p1 Active = f UL f UL + DL UL DL UL + DL The probability of the user being inactive on both UL and DL: p1 Inactive =  1 – f    1 – f   p Connected b. f UL UL DL  D Activity  Session  D Activity  Session DL = --------------------------------------------------------------------------------------. 2nd case: At a given time. active on DL. DL UL + DL Atoll takes into account activity periods during the connection in order to determine the activity status of each user. if you compare each simulation. and active on UL and DL users.2 Simulations Based on Sector Traffic Maps Sector traffic maps are also referred to as live traffic maps. The service and the activity status of each user are random in each simulation. if you compute several simulations at once.2.3.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 The user distribution per service. Live traffic data from the OMC is spread over the best server coverage areas of the transmitters included in the traffic map. • And users active in both links ( n i – Active ).1.1 Throughputs in Uplink and Downlink When selecting Throughputs in Uplink and Downlink.1. But.and N = ----------. 9. you can input the throughput demands in the uplink and downlink for each sector and for each listed service. the average number of users per service and average numbers of inactive.3. or Erlangs per service are assigned to the coverage areas of each transmitter. we have: UL + DL UL + DL  N UL  p Active N DL  p Active  UL + DL  --------------------------------------Number of users active in UL and DL both: n i – Active = min  -------------------------------------- UL UL + DL DL + DL  p Active + p Active p Active + p UL Active  UL UL + DL Number of users active in UL and inactive in DL: n i – Active = N UL – n i – Active 612 . and the activity status distribution between the users are average distributions. NUL and NDL values include: UL • Users active in uplink and inactive in downlink ( n i – Active ). will correspond to calculated distributions. • Users active in downlink and inactive in uplink ( n i – Active ). Then. f Act and f Act are respectively the UL and DL activity factors defined for the service i. Atoll calculates the number of users per activity status: We have: UL UL + DL UL DL UL + DL DL UL + DL UL DL UL + DL  p Active + p Active    n i – Active + n i – Active + n i – Active  = N UL  p Active + p Active    n i – Active + n i – Active + n i – Active  = N DL Therefore. Throughput demands per service. Activity probabilities are calculated as follows: UL DL Probability of being inactive in UL and DL: p Inactive =  1 – f Act    1 – f Act  UL UL DL DL DL UL Probability of being active in UL only: p Active = f Act   1 – f Act  Probability of being active in DL only: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active both in UL and DL: p Active = f Act  f Act UL DL Where. Therefore. 9. Atoll calculates the number of users active in uplink and in downlink in the Txi cell using the service (NUL and NDL) as follows: N UL N DL UL DL RS RS DL = ----------. active on UL.3.Atoll 3. you will observe that the user distribution between services as well as the activity status distribution between users is different in each simulation.for R99 circuit and packet switched services UL DL R Nom R Nom DL RS = ---------for HSDPA service DL R Avg UL DL R S and R S are the uplink and downlink throughputs for service S in the TXi cell from the traffic map. the numbers of active users per service.  p inactive 1 – p inactive Therefore.Atoll 3. a TD-SCDMA network automatically regulates itself by using uplink and downlink power control in order to minimise interference and maximise capacity. active on DL.2 Power Control Simulation Based on CDMA air interface.e. or active in UL only. Atoll calculates the number of users per activity status: Number of inactive users in UL and DL: n i – Inactive = n i  p Inactive UL UL DL DL Number of users active in UL and inactive in DL: n i – Active = n i  p Active Number of users active in DL and inactive in UL: n i – Active = n i  p Active UL + DL UL + DL Number of users active in UL and DL both: n i – Active = n i  p Active Therefore. for each sector and for each service. The activity status of users is based on an average distribution.3. 9. the UL DL UL + DL number of users active in the uplink ( n i – Active ). The process is repeated from iteration to iteration and ends when the network is balanced. the average numbers of inactive. 9. or inactive in both links. But. you can input the number of connected users for each sector and for each listed service ( n i ). a connected user can have four different activity status: either active in both links. or active in UL only. you can directly input the number of inactive users ( n i – Inactive ). Atoll takes into account activity periods during the connection in order to determine the activity status of each user. The simulation algorithm also models the impact of smart antennas in the power control loop. a connected user can have four different activity status: either active in both links. i. or active in DL only. in the downlink ( n i – Active ) and in the uplink and downlink ( n i – Active ). 613 . Smart antennas improve the signal quality of each served mobile. or active in DL only. mobile terminal power. f Act and f Act are respectively the UL and DL activity factors defined for the service i.2 Total Number of Users (All Activity Statuses) When selecting Total Number of Users (All Activity Statuses). The influence of smart antennas is taken into account in signal quality calculations. active on UL. or inactive in both links.1.1.3. Atoll simulates these network regulation mechanisms using an iterative algorithm and calculates network parameters such as traffic power per cell and per timeslot.3. 9.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 UL + DL DL Number of users active in DL and inactive in UL: n i – Active = N DL – n i – Active UL UL + DL DL  n i – Active + n i – Active + n i – Active  Number of inactive users in UL and DL: n i – Inactive = -------------------------------------------------------------------------------. when the convergence criteria on uplink and downlink are satisfied. In each iteration. The activity status of each user is random in each simulation. Then. and handoff status for each terminal. Activity probabilities are calculated as follows: UL DL Probability of being inactive in UL and DL: p Inactive =  1 – f Act    1 – f Act  UL UL DL DL DL UL Probability of being active in UL only: p Active = f Act   1 – f Act  Probability of being active in DL only: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active both in UL and DL: p Active = f Act  f Act UL DL Where. will correspond to calculated distributions. and active on UL and DL users. you will observe that the activity status distribution between users is different in each simulation.3 Number of Users per Activity Status When selecting Number of Users per Activity Status. if you compute several simulations at once.2. if you compare each simulation.2. For each user distribution.3. all the mobiles (R99 and HSDPA service users) selected during generation of the user distribution attempt to connect to the network one by one. Therefore.. Afterwards. Interference on the downlink and the uplink is calculated on a per user.2: TD-SCDMA Power Control Algorithm 9.3.e.3.1 Algorithm Initialisation At the start of each simulation. Figure 9. no connected mobiles) • Term Uplink required power for mobiles is set to P Min 9.Atoll 3.2 R99 Part of the Algorithm Req The algorithm is described for an iteration k.2. all Q UL Req and Q DL thresholds depend on the user mobility. Power control is simulated over a sub-frame. i.2.1 Determination of Mi’s Best Server (SBS(Mi)) This step is performed for TS0 for each station TXi containing Mi in its calculation area.e.2. Here.3.. All the variables used in the description below are listed in "Definitions and Formulas" on page 595. the system loads for each carrier and timeslot are reset to initial values: • • Downlink traffic powers of cells P TCH – DL are initialised to 0 Watts Uplink interference powers received on all the carriers and timeslots I Intra – UL and I Extra – UL are initialised to 0 Watts (i. The best server for Mi is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power.2. 7 timeslots. 9. the P-CCPCH RSCP is calculated for: • 614 the preferred carrier of the service used by Mi.3. and are defined in the Service and Mobility parameter tables. or .. Xk is the value of the variable X at the iteration k. For HSDPA users. The following calculations are made for all R99 and HSDPA mobiles (Mi) using R99 bearers. The steps of this algorithm are detailed below.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 decrease the required powers and the loads of all the surrounding cells. In the algorithm. uplink and downlink power control is performed on the associated A-DCH bearer before fast link adaptation on downlink. or from the master carrier in case of N-frequency mode compatible transmitters. e. i. P – CCPCH j = All  The best server is determined once for the whole simulation during the first iteration. i.3. In this case. taken into account when the option "Indoor coverage" is selected L Mi is the los of the terminal used by Mi Mi L Body is the body loss defined in the service used by Mi G G L Mi TX i TX i is the receiver gain of the terminal user by Mi is the transmitter antenna gain is the transmitter loss ( L TX i = L Total – DL ) A cell TX i  ic  is considered the best server of a mobile Mi if it satisfies the following conditions: TX i  ic  Min RSCP P – CCPCH  RSCP P – CCPCH . if no preferred carrier is defined for the service. The aim of Dynamic Channel Allocation (DCA) is to reduce interference in order to maximise the usage of the radio resources. which when allocated to the mobiles will optimise the load balance between carriers. Mi is considered unable to connect to the network if no carrier or not enough timeslots have been selected. In this case. TX i  ic  And RSCP P – CCPCH = Best  RSCP TXj  jc   .. Mi can be allocated timeslots over more than one slave carrier.Atoll 3. If a preferred carrier is defined for the service requested by Mi and if it is available at TX i . it remains admitted for the all the iterations unless there are other reasons to reject it (following steps).2. 9. the DCA tries to find the "best carrier" and the "best timeslots". If Mi has no best server. BestCarrier  TX i M i  = the carrier preferred for the service. it is not taken into account in the next steps. Once a mobile has been admitted for a simulation. because the best server does not change during the simulation and smart antennas do not influence this step.e. k = 0. TX i  ic  RSCP P – CCPCH  TAdd P – CCPCH  Mobility  .. TX i L Path = L Model + L Ant L Model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model TX i L Ant is the transmitter antenna attenuation (from antenna patterns) Model M Shadowing is the shadowing margin. In other words. k = 0. Mi is rejected for the reason P-CCPCH RSCP < Min P-CCPCH RSCP. In the case of N-frequency compatible transmitters. This parameter is taken into account when the option “Shadowing taken into account” is selected L Indoor are the indoor losses. then the mobile will be rejected for the reason "DL Load Saturation" or "Admission Rejection" respectively. 615 .2.2 Dynamic Channel Allocation The dynamic channel allocation is performed once for the whole simulation during the first iteration. Mi is considered unable to connect to the network if no best server has been selected.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 • • the carrier with the highest P-CCPCH power. the mobile Mi will be rejected for the reason "RU Saturation". If the carrier and timeslot(s) selected by the DCA do not satisfy the control of radio resource limits for DL power or UL load. or the master carrier in case of N-frequency mode compatible transmitters.3. The RSCP from a transmitter TXi and a selected carrier ic is given by: TX  ic  i TX  ic  i RSCP P – CCPCH = P P – CCPCH + G TX i –L TX i Model M i – L Path – M Shadowing – L Body – L Indoor + G M i –L M i in dBm Where. The DCA controls the mobile admission. If a smart antenna is used by the transmitter in the uplink. 2.is the uplink required signal quality. 1. TX i  ic TS  M i   is described in "Uplink Power Control" on page 617. X X DCA X DCA DCA = DCA X UL N Tot – UL . if you enter an angular step of 15 degrees. and allocates a different carrier than the ones used by any interfering mobiles found. The best carrier for a mobile is the one that has the highest number of resource units: BestCarrier  TX i M i  = Carrier Max  RUs  Timeslot selection by Available RUs: From the selected carrier. Atoll searches for interfering mobiles within 15 degrees to the right and to the left of the served user. TX i  ic TS  M i   And. Timeslot selection by Load: From the selected carrier. Available RUs Carrier selection by Available RUs: The DCA determines the carrier which has the highest number of available resource units with enough timeslots to accomodate the service being used by each mobile Mi. The best carrier for a mobile is the one that is least loaded: BestCarrier  TX i M i  = Carrier Where.Atoll 3. For example. Direction of Arrival Carrier selection by Direction of Arrival: The DCA determines the direction of arrival of the signal from the served user Mi and checks whether there is an interfering mobile in the same direction as Mi. the direction of arrival for the served user Mi should not be the direction of arrival of an interfering mobile. Atoll searches for interfering mobiles within the angle defined by the Angular Step. = ----------------------------------------------TX i  ic TS  M i   TX i N Tot – UL + N0 is the load increment given by: Mi TX i    1 – f UL    1 – f JD    = ---------------------------------------------------------------------1 1 + ----------Req Q UL Ortho E b  --- N t UL Proc = -----------------. TX i  ic TS  M i   • N Tot – UL • N Tot – DL • The carrier is the same in the uplink and in the downlink for mobiles accessing circuitswitched services. The uplink processing gain G UL calculated Proc G UL Req Req C Req Where Q UL =  ---  I  UL from the service parameters. 616 . the smart antenna gain is taken into account in calculating Req Q UL . 3.3. if no smart antenna is used by the transmitter in the uplink. Load Carrier Selection by Load: The DCA determines the least loaded carrier with enough timeslots to accomodate the service being used by each mobile Mi. BestCarrier  TX i M i  = Carrier DoA  Mi   DoA  Mj  In other words. X DCA DCA = X DL Min  X DCA TX i  ic TS  M i   = N Tot – DL  if the mobile is connected in the downlink. is described in "Downlink Power Control" on page 619. X DCA if the mobile is connected in the uplink. Atoll selects the timeslots which are the least loaded and have enough resource units for the service being accessed by Mi. The best carrier for a mobile is the one which is not interfered by another mobile in the direction of the mobile Mi. These strategies are described below one by one. Atoll selects the timeslots which have the highest numbers of available resource units.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 There are four strategies for the DCA available in Atoll.  G Proc . TX i  ic TS  M i   N Tot – UL TX i  ic TS  M i   = I Tot – UL TX i + N0 Where TX i  ic TS  M i   I Tot – UL Mi = RSCP TCH – UL  TX i  ic TS  M i         Mj Mi + RSCP TCH – UL  TX i  ic TS  M i      Mi + M j  TX i  ic TS  M i   Mj  Mi  M M  1 –  j  RSCP j TCH – UL  TX i  ic TS  M i    +   M j  TX i  ic TS  M i   Mj  Mi  M TX j i RSCP TCH – UL  TX i  ic TS  M i      1 – F MCJD   M  TX  ic TS  M   j i i  M i =  M i Ortho   1 – F UL TX  i    1 – F JD  and  =  0  1 Without Useful Signal Total Noise The above formula gives the value of I Tot – UL for the uplink connection between Mi and TX i  ic  . 9. If the mobile Mi is connected (active or inactive) in the uplink and has a best server TX i  ic  assigned to it. Atoll calculates the signal quality on the uplink timeslots allocated to Mi by the DCA: TX i  ic TS  M i   E b  --- N t TCH – UL TX i  ic TS  M i   Mi Mi TX i  ic TS  M i   TX i  ic TS  M i     RSCP TCH – UL   RSCP TCH – UL Div  C . Atoll allocates timeslots to served users Mi in a sequential order.3 Uplink Power Control For each mobile Mi. 4.2. Mj. At the end of the DCA. an admitted mobile can have associated timeslots over more than one slave carrier. Atoll selects the timeslots which are not being used by any other mobile Mj located in the same direction as the served user Mi. the mobile that is listened to by the transmitter TX i  ic  .3. which is defined for the service being accessed by Mi. Sequential Sequential carrier selection: The DCA allocates carriers to served users Mi in a sequential order.Atoll 3. G Div = ------------------------------------------------------= ------------------------------------------------------UL  G UL or  --- UL TX i  ic TS  M i   TX i  ic TS  M i   I TCH – UL N Tot – UL N Tot – UL Calculation of Uplink Total Noise ( N Tot – UL ): The uplink total noise is calculated for the uplink connection between each mobile Mi and its best server TX i  ic  . Sequential timeslot selection: From the selected carrier.e. The mobile Mi is the focus. as well as located in the coverage areas of other cells. taking into account the interference received from other mobiles. 617 . which are located in the Mi best server coverage area.2..3. each admitted mobile has an associated carrier and timeslots. In case of N-frequency mode compatible transmitters.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Timeslot selection by Direction of Arrival: From the selected carrier. the uplink power control step calculates the uplink power required to satisfy the required quality level on the traffic channel. i. Interference is updated only for active mobiles on the uplink for circuit. TX i  ic  .Atoll 3. If Mi is an HSDPA user. Model LT 618 k–1 is the uplink required mobile . G TX i and L TX i are read from the main antenna model. P Req  TX i  ic TS  M i    = 0. and Mi P Req  TX i  ic TS  M i    is set to 0. • The extra-cell interference for which the best-server for the received mobile Mj is not TX i  ic  . for Mi P Req  TX i  ic TS  M i    .0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 The four terms comprising I Tot – UL are: • The useful signal for which the received mobile is the focus (Mi). G TX i SA = G UL and L TX i SA = L UL are calculated according to the smart antenna modelling method used.3. they are considered in the number of rejected mobiles. However. Otherwise. • The intra-cell interference for which the best-server is the same for the received mobile Mj and the focus Mi. if these mobiles are rejected. TX i  ic TS  M i   Where RSCP TCH – UL TX i Mi P Req  TX i  ic TS  M i    k–1 = --------------------------------------------------------------Model LT Mi Mi Model Mi L Path  L  L  L Body  L Indoor  M Shadowing = -----------------------------------------------------------------------------------------------------------------.and packetswitched services. • The intra-cell interference due to distortion in the terminal transmission. Mi P Req  TX i  ic TS  M i    k–1 The uplink received signal code power is: RSCP TCH – UL  TX i  ic TS  M i    = --------------------------------------------------------------Model LT Mi TX i Mi Mi Model Mi L Path  L  L  L Body  L Indoor  M Shadowing = -----------------------------------------------------------------------------------------------------------------. if a smart antenna is available in the uplink. Mi Mi P Min and P Max are set in the properties of the terminal used by the mobile Mi.1  PReq  TX i  ic TS  M i    Model In L T . Mi Calculation of Uplink Required Power ( P Req ): Then Atoll determines the required uplink power by: Req Mi P Req  TX i  ic TS  M i    E b  --- N t TCH – UL = P Req  TX i  ic TS  M i     -------------------------------------k–1 E b TXi  ic TS  M i    --- N t TCH – UL Mi k Req Mi or P Req  TX i  ic TS  M i    M C ---  I  TCH – UL -----------------------------------= P Req  TX i  ic TS  M i     TX i  ic TS  M i   k–1 C  ---  I  TCH – UL Mi k M i i M M i i And if P Req  TX i  ic TS  M i     P Min then P Req  TX i  ic TS  M i    = P Min M M i i If P Req  TX i  ic TS  M i     P Max then the mobile Mi is rejected for the reason "Pmob > PmobMax".and P Req  TX i  ic TS  M i    TX i Mi G G power for iteration k .1 transmitted on the timeslot allocated to Mi.and P Req  TX i  ic TS  M i    is the uplink required mobile power TX i Mi G G Model LT Mi Mi calculated for the timeslot allocated to Mi. TX i  ic TS  M i   N Tot – DL TX i  ic TS  M i   = I Tot – DL Mi + I IC – DL  ic jc  + I MM  M i M j  + N 0 Where TX i  ic TS  M i   I Tot – DL TX i  ic TS  M i   = RSCP Tot – DL  Mi      TX i + TX i  ic TS  M i    RSCP Tot – DL   1 –  i  RSCP i Tot – DL    Mj    TX i + M  TX  ic TS  M   j i i Mj  Mi TX TX  ic TS  M i    Mj  + M j  TX i  ic TS  M i   Mj  Mi  TX j  ic TS  M i   RSCP Tot – DL  Mj  M j  TX i  ic TS  M i   The four terms comprising I Tot – DL are: • The useful signal for which the received mobile is the focus (Mi).3. Otherwise. G TX i SA = G UL and L TX i SA = L UL are calculated according to the smart antenna modelling method used.4 Downlink Power Control For each mobile Mi. which is defined for the service being accessed by Mi. the downlink power control step calculates the downlink power for the best server TX i  ic  required to satisfy the required quality level on the traffic channel. • The extra-cell interference for which the best-server for the received mobile Mj is not TX i  ic  . they are considered in the number of rejected mobiles.or packetswitched services are calculated for information only. Atoll calculates the signal quality on the uplink timeslots allocated to Mi by the DCA: TX i  ic TS  M i   E b  --- N t TCH – DL TX i  ic TS  M i   TX i TX i TX i  ic TS  M i   TX i  ic TS  M i     RSCP TCH – DL   RSCP TCH – DL Div C  .2. • The intra-cell interference for which the best-server is the same for the received mobile Mj and the focus Mi. G Proc -.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Model In L T M . TX j  jc TS  M i    RSCPTot – DL  Mi  All TX j I IC – DL  ic jc  = ---------------------------------------------------------------F IRF  ic jc  619 . 9.2.Atoll 3. However.3. If the mobile Mi is connected (active or inactive) in the downlink and has a best server TX i  ic  assigned to it. if a smart antenna is available in the uplink. The uplink required powers for mobiles inactive in the uplink accessing circuit. G Div = ------------------------------------------------------- G or = -------------------------------------------------------DL DL DL TX i  ic TS  M i   TX i  ic TS  M i    I  TCH – DL N Tot – DL N Tot – DL Calculation of Downlink Total Noise ( N Tot – DL ): The downlink total noise is calculated for the downlink connection between each mobile Mi and its best server TX i  ic  . G TX i and L TX i are read from the main antenna model. TX i  ic  . for i P Req  TX i  ic TS  M i    . if these mobiles are rejected. • The intra-cell interference due to distortion in the transmitter. jc). if a smart antenna is available in the downlink.1  P TCH – DL SA = L DL are calculated according to the smart antenna modelling method used. P TCH – DL Model In L T .0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks  TX i =  TX i Ortho   1 – F DL ©Forsk 2015 M  i    1 – F JD  and  =  0    1 Without Useful Signal Total Noise I IC – DL  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink.and RSCP OCCH = -----------------------------Model Model LT LT TX i  ic TS  M i   LT TX i  ic TS  M i   + RSCP OCCH TX i Mi Mi Model TX i  ic TS  M i   L Path  L  L  L Body  L Indoor  M Shadowing = -----------------------------------------------------------------------------------------------------------------. and d is the distance between the mobiles Mi and Mj in km. Mj Mj P TCH – UL RSCP TCH – UL  M i  = ------------------L MM L MM   32. G TX i and L TX i are read from the main antenna model.and P TCH – DL TX i Mi G G is the downlink traffic power transmitted k–1 TX  ic TS  M   i i on the timeslot allocated to Mi during the iteration k . TX i  ic TS  M i   TX i  ic TS  M i   = RSCP TCH – DL RSCP Tot – DL TX i  ic TS  M i   With RSCP TCH – DL Model TX  ic TS  M   i i P TCH – DL TX i  ic TS  M i   P OCCH k–1 = ----------------------------------------.4 + 20  Log  F Avg  + 20  Log  d  =    49 + 30  Log  F Avg  + 40  Log  d  If d If d Mi – Mj Mi – Mj 3m with F Avg being the average frequency in MHz of the 3m frequency band used by the best server of the mobile Mi. for TX i  ic TS  M i   only and not for P OCCH . Otherwise.Atoll 3.3. If Mi is an HSDPA user. (TSMj). TX i  ic TS  M i   Calculation of Downlink Required Power ( P Req ): Then Atoll determines the required downlink power by: Req TX i  ic TS  M i   P Req TX i  ic TS  M i   k = P Req E b  --- N t TCH – DL  -------------------------------------k–1 E b TXi  ic TS  Mi    --- N t TCH – DL Req TX i  ic TS  M i   or P Req TX i  ic TS  M i   k = P Req TX i  ic TS  M i   And if P Req 620 Min C ---  I  TCH – DL  -----------------------------------TX i  ic TS  M i   k–1 C ---  I  TCH – DL TX i  ic TS  M i    P TCH – DL  Service  then P Req Min = P TCH – DL  Service  . G TX i  ic TS  M i   P TCH – DL TX i SA = G DL and L TX i TX  ic TS  M   i i = 0. • The interference received from the mobile Mj at the mobile Mi is calculated using either the free-space propagation model or the Xia model. which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic. Mj  RSCPTCH – UL  Mi  M j I MM  M i M j  = ------------------------------------------------is the interference from each mobile Mj transmitting in the uplink on the same F IRF  ic jc  timeslots as those on which the mobile Mi is receiving in the downlink. and The downlink timeslot of Mi (TSMi) is the same as the uplink timeslot of Mj.1. Mj can interfere Mi directly if and only if: Mi – Mj • The distance between Mi and Mj ( d ) is less than the Max Distance between interfering mobiles defined by the user when starting the simulation. e. i. k = 0. TX i  ic TS  M i   Otherwise. for the following iterations since the DCA has been performed.6 . Min Max P TCH – DL  Service  and P TCH – DL  Service  are set in the properties of the R99 bearer associated with the service used by the mobile Mi. Therefore. for the first iteration. This step updates the received signals for all the mobiles in the TX i  ic  coverage area which are interfered in the downlink by the connection between TX i  ic  and Mi..5 Uplink Signals Update This step uses the uplink terminal powers calculated for each timeslot allocated to the mobiles.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 TX  ic TS  M   i i If P Req TX  ic TS  M   i i Max  P TCH – DL  Service  then the mobile Mi is rejected for the reason "Ptch > PtchMax". Otherwise. this step is performed for any downlink timeslot for each mobile Mi that is connected and active.2. if a smart antenna is available in the downlink. Atoll also calculates the downlink traffic power for the different values of the Angular Step  Step . the mobile that is listened to by the transmitter TX i  ic  .Atoll 3. for . during the first iteration. TX i  ic  is the interfered receiver and Mi is the focus. only if TX j  ic  contain Mi in their coverage areas. TX i  ic TS  M i   For each mobile Mi interfered by Mj in the uplink by the connection between Mj and TX j  ic  .3. Downlink Signals Update For the first iteration. and P Req is set to 0. the downlink signals update step uses the actual downlink traffic powers calculated for each timeslot and the actual timeslots allocated to the mobiles. G TX i and L TX i are read from the main antenna model. The Dynamic Channel Allocation allocates timeslots and carriers to all the connected and active mobiles.2. The Dynamic Channel Allocation allocates timeslots and carriers to all the connected and active mobiles. for the following iterations. and the timeslot and carrier allocation remains the same for all the following iterations of a simulation. The Dynamic Channel Allocation is performed once only..and P Req TX i Mi G G .3. Atoll updates RSCP TCH – DL 621 .e.2. TX i  ic TS  M i   For each mobile interfered by Mi. TX i  ic TS  M i   RSCP TCH – DL TX i  ic TS  M i    Step = RSCP TCH – DL TX i  ic TS  M i   Where RSCP TCH – DL TX i SA G DL  -----------------------SA L DL   Step  TX i  ic TS  M i   P Req k–1 = ----------------------------------------Model LT Mi Mi Model TX i  ic TS  M i   L Path  L  L  L Body  L Indoor  M Shadowing = -----------------------------------------------------------------------------------------------------------------. the downlink traffic powers for all the downlink timeslots are set to 0 Watts. and this step is performed for all the downlink timeslots allocated to the mobile Mi on which it is connected and active. The Dynamic Channel Allocation is performed once only during the first iteration and the timeslot and carrier allocation remains the same for all the following iterations of a simulation.1 transmitted on the timeslot allocated to Mi. the downlink traffic power is incremented P TCH – DL TX i  ic TS  M i   = P TCH – DL TX i  ic TS  M i   + P Req For each mobile. G TX i  ic TS  M i   P Req TX i SA = G DL and L TX i k–1 is the downlink traffic power for iteration k SA = L DL are calculated according to the smart antenna modelling method used. However. This step updates the received signals for all the mobiles Mi interfered in the uplink by the uplink connection between interfering mobiles Mj and their best servers TX j  ic  .2. Model LT Model In L T . i. The downlink power for mobiles inactive in the downlink accessing circuit.or packetswitched services are calculated for information only. Therefore. this step is performed for any downlink timeslot for each mobile Mi that is connected and active for the first iteration.3. 9. Atoll updates RSCP TCH – UL 9. e.. The maximum number of supported HS-SCCH channels is defined per cell. 9.2. i TX i  ic TS  M i   P R99 – DL TX i  ic TS  M i   = P TCH – DL TX i  ic TS  M i   If P R99 – DL TX i  ic TS  M i   + P OCCH TX i  ic TS  M i    P Max – DL – Eff the mobile with the lowest service priority is rejected for the reason "DL Load Saturation". TX i  ic TS  M i   The maximum allowed uplink cell load. Uplink Load Control: Atoll verifies that the uplink load of any cell on any timeslot does not exceed the maximum uplink cell load allowed per timeslot.7 Control of Radio Resource Limits (Downlink Traffic Power and Uplink Load) This step checks whether the downlink traffic powers of the downlink timeslots and the uplink loads of the uplink timeslots of all the cells satisfy the conditions defined globally or per cell and timeslot.3. Downlink Power Control: Atoll verifies that the total R99 power transmitted by any cell on any timeslot does not exceed the effective maximum cell power per timeslot. is either taken from the properties of each cell or from the simulation properties if a global value is defined. TX i  ic TS  M i   TX i N Tot – UL + N0 If a smart antenna is used by the transmitter in the uplink. 9.2. The uplink load is given by: TX i  ic TS  M i   X UL TX i  ic TS  M i   N Tot – UL = ----------------------------------------------. the smart antenna gain is taken into account in the calculation of uplink load. i. • 622 Dynamic Allocation .3.2.e.Atoll 3..3. TX i  ic  TX i  ic  TX i  ic  P Tot – DL = P R99 – DL + P HR TX i  ic  TX i  ic  + P HS – SCCH + P HS – PDSCH The HSDPA powers. 9.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Where TX i  ic  is the transmitter considered and Mi is the focus.1 HSDPA Power Allocation TX i  ic  The total transmitted power of the cell ( P Tot – DL ) is the sum of the R99 transmitted power and the HSDPA powers. For each transmitter TXi. The effective maximum cell traffic power per timeslot is calculated as: TX i  ic TS  M i   TX i  ic TS  M i   P Max – DL – Eff = P Max – DL TX i  ic TS  M i   Where P Max – DL  %P Max – DL is the maximum cell power per timeslot defined per cell. i TX i  ic TS  M i   If X UL TX i  ic TS  M i    X Max – UL the mobile with the lowest service priority is rejected for the reason "UL Load Saturation". X Max – UL . carrier ic. Power can be allocated to HS-SCCH statically or dynamically: • Static Allocation The static HS-SCCH power is defined in the properties of the HSDPA cell.2. and %P Max – DL is the maximum allowed downlink load either taken from the properties of each cell or from the simulation properties if a global value is defined. For each transmitter TXi. carrier ic.if no smart antenna is used by the transmitter in the uplink. the HS-SCCH and HS-PDSCH powers are calculated as follows: • HS-SCCH Power: HS-SCCH channels are transmitted on DL traffic timeslots. and uplink timeslot TS M .3. i. the mobile that is the target for TX i  ic  . and downlink timeslot TS M .3.3 HSDPA Part of the Algorithm The following calculations are made for all HSDPA mobiles (Mi). 3. Model LT Model In L T . The effective maximum cell traffic power per timeslot is calculated as: TX i  ic  TX i  ic  TX i  ic  P Max – DL – Eff = P Max – DL  %P Max – DL . Otherwise. P Max – DL is the maximum power defined per cell.  N t HS – SCCH TX  ic  i TX  ic  i TX  ic  i TX  ic  i Where P Available – HS – SCCH = P Max – DL – Eff – P R99 – DL – P HR TX i  ic  . TX  ic  i P Max – DL is the maximum power defined per cell.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 TX  ic  TX  ic  E TXi  ic  Req i i HS-SCCH power is calculated for  ----c- = Q HS – SCCH  Mobility  so that P HS – SCCH  P Available – HS – SCCH . Power can be allocated to HS-SICH statically or dynamically: • Static Allocation The static HS-SICH power is defined in the properties of the terminal used by the HSDPA mobile Mi. The maximum number of supported HS-SICH channels is defined per cell. if a smart antenna is available in the downlink. • Dynamic Allocation HS-PDSCH power is calculated as follows: TX  ic  i TX  ic  i TX  ic  i TX  ic  i P HS – PDSCH = P Max – DL – Eff – P R99 – DL – P HR TX i  ic  TX i  ic  TX  ic  i – P HS – SCCH TX i  ic  Where P R99 – DL = P TCH – DL + P OCCH . and %P Max – DL is the maximum allowed downlink load either taken from the properties of each cell or from the simulation properties if a global value is defined.G TX i SA = G DL and L TX i SA = L DL are calculated according to the smart antenna modelling method used. TX i  ic  TX i  ic  The effective maximum cell traffic power per timeslot is calculated as: P Max – DL – Eff = P Max – DL  %P Max – DL .  TX i =  TX i Ortho   1 – F DL TX i Mi     1 – F JD  and  =  0  1 Mi Mi Without Useful Signal Total Noise Model TX i  ic  L Path  L  L  L Body  L Indoor  M Shadowing = -----------------------------------------------------------------------------------------------------------------. G TX i and L TX i are read from the main antenna model. TX  ic  i TX i  ic  P HS – SCCH TX  ic  TX  ic  Ec i i i  --- N –     RSCP HS – SCCH  N t HS – SCCH  Tot – DL  Model = ---------------------------------------------------------------------------------------------------------------------------- LT TX i  TX TX i  ic  Where N Tot – DL is the downlink total noise calculated in "Downlink Power Control" on page 619.Atoll 3. • Dynamic Allocation 623 .and P HS – SCCH is the HS-SCCH power calculated for the TX i Mi G G timeslots allocated to Mi. and %P Max – DL is the maximum allowed downlink load either taken from the properties of each cell or from the simulation properties if a global value is defined. Power can be allocated to HS-PDSCH statically or dynamically: • Static Allocation The static HS-PDSCH power is defined in the properties of the HSDPA cell. for TX i  ic  P HS – SCCH . • HS-PDSCH Power: HS-PDSCH channels are transmitted on DL traffic timeslots. The HS-SICH power is calculated as follows: • HS-SICH Power: HS-SICH channels can be transmitted on any UL traffic timeslot. and TX  ic  i P R99 – DL = TX  ic  i P TCH – DL + TX  ic  i P OCCH is the power available for HS-SCCH in the cell . G TX i SA = G UL and L TX i SA = L UL are calculated according to the smart antenna modelling method used. If no cell with such a resource unit is available.and P HS – SICH is the HS-SICH power calculated for the TX i Mi G G timeslots allocated to Mi. HS-SICH HS-SICH admission control is performed for active HSDPA users connected to A-DCH bearers on the uplink and having an HSSICH sub-connection status. TX i  ic  TX i  ic  TX i  ic  P Tot – DL = P R99 – DL + P HR 9. Each cell is able to manage a maximum number of HS-SCCH channels. if a smart antenna is available in the uplink.2 TX i  ic  TX i  ic  + P HS – SCCH + P HS – PDSCH Connection Status and Number of HSDPA Users HSDPA users cannot receive HS-SCCH and HS-PDSCH powers simultaneously.3. Therefore. L Model = ------------------------------------------------------------------------------------------------------------------------T Mi  N Tot – UL M i M M is the uplink total noise calculated in "Uplink Power Control" on page 617. Mi for P HS – SICH . the DCA provides a DL timeslot with one SF16 resource unit that has the downlink Ec/Nt higher than the required quality. the number of connected HSDPA users cannot exceed the number of HS-SCCH and HS-SICH channels per cell. Therefore.3.3.3. If no cell with such a resource unit is available. The sub-connection status can be: • • • HS-SCCH: HSDPA mobile that is receiving HS-SCCH power HS-PDSCH: HSDPA mobile that is receiving traffic power HS-SICH: HSDPA mobile that is transmitting HS-SICH power The number of active HSDPA users belonging to each sub-connection status is 1/3rd of the total number of active HSDPA users. Model LT Model In L T . the user is rejected. and transmitting HS-SICH because their occurrence is equally likely. G TX i and L TX i are read from the main antenna model. n HS – SCCH is the maximum number of HS-SCCH channels and n HS – SICH is the maximum number of HS-SICH channels that the cell can manage. Otherwise. 624 . at a given instance. HS-PDSCH arrives 3 timeslots after the HS-SCCH. the DCA provides an UL timeslot with one SF16 resource unit that has the uplink Ec/Nt higher than the required quality. Each HSDPA user consumes one HS-SCCH and HS-SICH channels. During the R99 part. TX  ic  i Mi P HS – SICH TX  ic  i Where  M i =  TX  ic  Ec  i i i  --- N –     RSCP HS – SICH  N t HS – SICH  Tot – UL  . Ortho   1 – F UL TX i TX  i    1 – F JD  and  =  0  1 Mi Mi Without Useful Signal Total Noise Model M L Path  L  L  L Body  L Indoor  M Shadowing i = -----------------------------------------------------------------------------------------------------------------.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 M TX  ic  E TXi  ic  Req i i HS-SICH power is calculated for  ----c- = Q HS – SICH  Mobility  so that P HS – SICH  P Max – HS – SICH and  N t HS – SICH M M i i P HS – SICH  P Max – HS – SICH . the user is rejected. TX i  ic  The total transmitted power of the cell ( P Tot – DL ) is the sum of the R99 transmitted power and the HSDPA powers. HS-SICH is 9 timeslots after the HS-PDSCH. The maximum number of HSDPA users ( n Max ) corresponds to the maximum number of HSDPA users that the cell can support. 9.Atoll 3. Each cell is able to manage a maximum number of HS-SICH channels.2.3 HSDPA Admission Control HS-SCCH HS-SCCH admission control is performed for active HSDPA users connected to A-DCH bearers on the downlink and having an HS-SCCH sub-connection status. During the R99 part. n HS – SCCH . n HS – SICH . each HSDPA user is assigned a sub-connection status randomly.2.3. Atoll assumes that an active HSDPA user has the same probability of receiving HSSCCH and HS-PDSCH. The scheduler allocates the best available HSDPA bearer to each user. 2. c..4 Convergence Criteria The convergence criteria are evaluated for each iteration and can be written as follows:  Max  TXi  ic TS  M i     DL = Int  P Err  100    All TX i   UL TX i  ic TS  M i   TX i  ic TS  M i     – N Tot – UL  Max N Tot – UL  k k – 1 = Int  ------------------------------------------------------------------------------------. b. HS-PDSCH Ec/Nt is calculated by taking into account all intra and extra cells interferences.4 HSDPA Dynamic Channel Allocation For each mobile connected to the A-DPCH bearer: 1. 4. If two bearers have the same RLC peak throughput. the best one is the one with the highest Ec/Nt. 9.3.3. The scheduler considers the group of HS-PDSCH users to whom bearers. If no bearer can be allocated due to low Ec/Nt. 3. The Ec/Nt value associated with the mobile-bearer pair is the worst one of all selected timeslots. and whose Ec/Nt is better than the minimum required and enough to reach the bearer’s resource unit requirements.2. Atoll checks that the Ec/Nt reaches the quality target. the resource units available in the master and slave carriers can be shared. The best is determined by applying the R99 Dynamic Channel Allocation algorithm. P Err is given by: 625 . the users will be connected. The scheduler sorts the HS-PDSCH users to whom bearers have been assigned in the order of decreasing RLC peak throughputs. If two users have the same bearer.2.5 Ressource Unit Saturation For each time slot. Other HS-PDSCH users will be rejected for the reason "HSDPA Scheduler Saturation".2. The required HS-PDSCH Ec/Nt value is read from receiver equipment properties. the user will be delayed or rejected. If the scheduler is unable to find a satisfactory timeslot collection. If there are enough HSDPA power and resource units available in order to obtain a HSDPA bearer. i. a minimum and maximum number of resource units for HSDPA users are defined in the cell properties. For each bearer supported by a mobile: a. If not. the user is rejected for the reason "HSDPA Scheduler Saturation". 3. The mobile is connected to the supported bearer having the highest RLC peak throughput. The scheduler calculates the HS-PDSCH Ec/Nt for each timeslot of the best collection. a mobile can be connected to timeslots belonging more than one carrier. they will be delayed and their connection status will be “HSDPA Delayed”. Atoll dynamically allocates the required number of codes respecting these limitations. The best available HSDPA bearer is selected depending on the user’s Ec/Nt.3. the user with the higher Ec/Nt has the higher rank. The scheduling is performed as follows: 1. 9. and HS-SICH have been assigned. Otherwise.3. Atoll selects the HSDPA bearers that match to the mobile terminal and UE category parameters.3. For N-frequency mode compatible transmitters. 9. 100 TX  ic  TS  M   All TX   i i i N Tot – UL   k TX  ic TS  M   i i Where. the scheduler allocates a lower HSDPA bearer which needs fewer codes. If there are no more resource units available for the lowest HSDPA bearer. For each bearer.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 HS-PDSCH Scheduling is performed for active HSDPA users connected to A-DCH bearers on the downlink and having an HS-PDSCH subconnection status.3. The number of HS-PDSCH users cannot exceed the maximum number of HSDPA users ( n Max ) supported by the cell. The scheduler searches for the best collection of "n" ordered timeslots that can provide enough resource units to support the service. The scheduler checks if enough codes are available for the selected HSDPA bearer (taking into account the maximum number of HSDPA codes). The minimum number of HSDPA codes is excluded from the set of codes available for R99 users. 2. HS-SCCH. the bearer is removed from the list of supported bearers.Atoll 3. Each HS-PDSCH user is considered as the only served user.e. 0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 TX  ic TS  M   i i TX  ic TS  M   i i P Err TX  ic TS  M   i i – P Rec P Rec Max  k  k–1 Step Step = ------------------------------------------------------------------------------------------------------------. If you perform these coverage predictions for the best carrier. Atoll stops the algorithm at the 56th iteration without converging. If  DL  5 and  UL  5 between the 4th and the 5th iteration. The simulation has converged. Each pixel within the TXi calculation area is considered a probe receiver. TXi. After the 30th iteration. the coverage predictions are calculated for the selected carrier. the simulation has reached convergence. 9. a carrier. a mobility type. Coverage predictions are calculated using bilinear interpolation of multi-resolution path loss matrices (similar to the evaluation of site altitudes). Atoll stops the algorithm after the 5th iteration. If  DL and  UL are lower than their respective thresholds. • Divergence: After 30 iterations. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. or from the master carrier in case of N-frequency mode compatible transmitters. there will not be any pixels covered by this transmitter. b. Example: Let us assume that the maximum number of iterations is 100. or the carrier with the highest P-CCPCH power.Atoll 3. and for TS0. the simulation has not converged. or the master carrier in case of N-frequency mode compatible transmitters.1 P-CCPCH Reception Analysis (Eb/Nt) or (C/I) E C These coverage predictions calculate and display the Eb/Nt or C/I on the P-CCPCH. TX  ic TS  M   0   Step  360 i i P Rec  TX  ic TS  M   i i P Err TX  ic TS  M   i i Step k TX  ic TS  M   i i P Rec – P Rec k k – 1 without smart antennas.4. If the selected carrier does not exist on a transmitter.with smart antennas. if no preferred carrier is defined for the service. Examples: Let us assume that the maximum number of iterations is 100. Coverage study parameters to be set are: • • The study conditions to determine the service area of each TBC transmitter The display settings to for colouring the covered pixels Atoll uses the parameters entered in the Condition tab of the coverage study properties dialogue to determine pixels covered by the each transmitter.  DL and  UL are less than or equal to their respective thresholds (defined when creating a simulation). The  N t P – CCPCH  I  P – CCPCH coverage predictions are calculated for a given set of a terminal type.3. a.4 TD-SCDMA Prediction Studies For each TBC transmitter. Atoll stops the algorithm at the 46th iteration. The simulation has not converged.  ----b- or  --- . they start decreasing slowly until the 40th iteration (without going under the thresholds) and then. Atoll determines the value of the selected parameter on each studied pixel inside the TXi calculation area. a service. 9.  DL and/or  DL equal 100 and do not decrease during the next 15 successive iterations. . Coverage prediction display resolution is independent of the path loss matrix and geographic data resolutions. = ---------------------------------------------------------------------------------TX i  ic TS  M i   P Rec k Atoll stops the simulations in the following cases: • Convergence: Between two successive iterations. and the UL and DL convergence thresholds are set to 5 %. and the UL and DL convergence thresholds are set to 5 %.  DL and/or  UL do not decrease during the next 15 successive iterations. and can be different for each coverage prediction. • Last Iteration: If  DL and/or  UL are still much higher than their respective thresholds after the last iteration. Atoll calculates the Eb/Nt or C/I considering: • • • 626 the preferred carrier of the selected service. Afterwards.  DL and/or  UL are still higher than their respective thresholds and from the 30th iteration. do not change during 15 successive iterations.  DL and/or  UL equal 80. After the 30th iteration. 0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 The pixels in TX i  ic  the TX  ic  i coverage area where TX  ic  i Min RSCP P – CCPCH  Max (TAdd P – CCPCH. which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic. TX TX  ic  i TX TX  ic  i i TX i  ic  E TX i  ic    RSCP P – CCPCH   RSCP P – CCPCH Proc C --- G Where  ----b- = ----------------------------------------------- and = -----------------------------------------------P – CCPCH TX i  ic   I  P – CCPCH TX i  ic  N t P – CCPCH N Tot – DL N Tot – DL i TX i  ic  TX i  ic  P P – CCPCH RSCP P – CCPCH = ---------------------LT The downlink total noise is calculated as follows: TX i  ic  TX i  ic  TX i  ic  Term N Tot – DL = I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0 Where TX i  ic  TX i  ic  I Intra – DL = RSCP P – CCPCH     With  TX i =  TX  ic  i I Extra – DL = TX i Ortho   1 – F DL TX i TX i  ic  + RSCP OCCH – TS0   Term    1 – F JD TX  ic  j TX i   and  =  0  1 Without Useful Signal Total Noise TX  ic  j   RSCPP – CCPCH + RSCPOCCH – TS0 ji TX j  jc  TX j  jc    RSCPP – CCPCH + RSCPOCCH – TS0 TX j I IC – DL  ic jc  = --------------------------------------------------------------------------------------F IRF  ic jc  I IC – DL  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink.RSCP P – CCPCH) . Coverage consists of several independent layers whose visibility in the workspace can be managed.Atoll 3. Each layer is assigned a colour and pixel is covered if  -----  N t P – CCPCH  I  P – CCPCH displayed with intersections between layers. 627 .3. Coverage consists of several independent layers whose visibility in the workspace can be managed. • Eb/Nt Margin or C/I Margin (dB) Atoll calculates the Eb/Nt or C/I margin on each pixel of the TX i  ic  best server coverage area. There are as many layers as thresholds defined in the Display tab (Prediction properties). and TX  ic  i E b Req Req  ---C ---  N t P – CCPCH  Q P – CCPCH or  I  P – CCPCH  Q P – CCPCH are covered and coloured according to the selected display option. For each layer. a TX  ic  TX  ic  Eb i C i  Threshold or  ---  Threshold . The pixel colour depends on the Eb/Nt or C/I margin value. TX i  ic  TX i  ic  RSCP OCCH – TS0 P OCCH – TS0 = -----------------------LT TX i Eb  Nt Term L Path  L  L  L Body  L Indoor  M Shadowing L T = ----------------------------------------------------------------------------------------------------------------------TX i Term G G  TX i Term and N 0 are defined in "Definitions and Formulas" on page 595. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • Eb/Nt or C/I (dB) Atoll calculates the Eb/Nt or C/I on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on the Eb/Nt or C/I level. Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter. There are as many layers as thresholds defined in the Display tab (Prediction properties). jc). Afterwards. there will not be any pixels covered by this transmitter. The coverage prediction is calculated for a  I  DwPCH given set of a terminal type. a pixel is covered if TX i  ic  E b  --- N t P – CCPCH Req  Q P – CCPCH or CECP Req  Q P – CCPCH .2 DwPCH Reception Analysis (C/I) C This coverage prediction calculates and displays the C/I on the DwPCH. TX i  ic  C ---  I  P – CCPCH For each layer.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 TX  ic  E TXi  ic  C i Req Eb  Nt Req CI For each layer. a mobility type. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. If you perform this coverage prediction for the best carrier. a pixel is covered if  ----b- – Q P – CCPCH  M P – CCPCH or  --- – Q P – CCPCH  M P – CCPCH .RSCP P – CCPCH) and  ---  Q DwPCH I DwPCH are covered and coloured according to the selected display option. and for DwPTS. If the selected carrier does not exist on a transmitter. or from the master carrier in case of N-frequency mode compatible transmitters. if no preferred carrier is defined for the service.Atoll 3. a carrier.  --- . or the master carrier in case of N-frequency mode compatible transmitters. or the carrier with the highest P-CCPCH power. Atoll calculates the C/I considering: • • • the preferred carrier of the selected service. The pixel colour depends on the cell edge coverage probability value. Coverage consists of several independent layers whose visibility in the workspace can be managed.4. which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic. • Cell Edge Coverage Probability (%) Atoll calculates the cell edge coverage probability on each pixel of the TX i  ic  best server coverage area. TX TX  ic  TX  ic  i i   RSCP DwPCH C i = -----------------------------------------Where  ---  I  DwPCH TX  ic  i N Tot – DL TX i  ic  RSCP DwPCH TX  ic  i P DwPCH = ---------------LT The downlink total noise is calculated as follows: TX i  ic  TX i  ic  TX i  ic  Term N Tot – DL = I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0 Where TX i  ic  TX i  ic  I Intra – DL = RSCP DwPCH     With  TX i  ic  TX i =  I Extra – DL = TX i Ortho   1 – F DL TX i Term    1 – F JD   and  =  0  1 Without Useful Signal Total Noise TX j  ic    RSCPDwPCH ji TX  jc  j   RSCPDwPCH TX j I IC – DL  ic jc  = ---------------------------------------F IRF  ic jc  I IC – DL  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink. Each layer is assigned a colour and displayed with intersections between layers. the coverage prediction is calculated for the selected carrier. 628 . TX i  ic  TX  ic  C i Min Req The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH. There are as many layers as thresholds defined in the Display tab (Prediction properties).  N t P – CCPCH  I  P – CCPCH Each layer is assigned a colour and displayed with intersections between layers.3. CECP 9. a service. jc). a service.4. a pixel is covered if  --- I DwPCH Req  Q DwPCH . There are as many layers as thresholds defined in the Display tab (Prediction properties). 9. if no preferred carrier is defined for the service. Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as thresholds defined in the Display tab TX  ic  C i (Prediction properties). There are as many layers as thresholds defined in the Display tab (Prediction properties). Each layer is assigned a colour CECP and displayed with intersections between layers. or the carrier with the highest P-CCPCH power.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 TX Eb  Nt Term i L Path  L  L  L Body  L Indoor  M Shadowing L T = ----------------------------------------------------------------------------------------------------------------------TX i Term G G  TX i Term and N 0 are defined in "Definitions and Formulas" on page 595. Afterwards. If you perform this coverage prediction for the best carrier. Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter. The pixel colour depends on the C/I margin value. or the master carrier in case of N-frequency mode compatible transmitters. a carrier. RSCP TCH – DL . If the selected carrier does not exist on a transmitter. or from the master carrier in case of N-frequency mode compatible transmitters. The pixel colour depends on the C/I level. Coverage consists of several independent layers whose visibility in the workspace can be managed. a pixel is covered if TX i  ic  C ---  Threshold . the coverage prediction is calculated for the selected carrier. For each layer.  I  DwPCH • C/I Margin (dB) Atoll calculates the C/I margin on each pixel of the TX i  ic  best server coverage area.RSCP P – CCPCH) and Req RSCP TCH – DL  RSCP TCH – DL  Service Mobility  are covered and coloured according to the selected display option. there will not be any pixels covered by this transmitter. a mobility type. Atoll calculates the RSCP considering: • • • The the preferred carrier of the selected service.3. For each layer. and for a downlink timeslot. • Cell Edge Coverage Probability (%) Atoll calculates the cell edge coverage probability on each pixel of the TX i  ic  best server coverage area. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest PCCPCH power. Each layer is assigned a colour and displayed with intersections between covered if  ---  I  DwPCH layers. pixels in the TX  ic  i TX i  ic  coverage area where TX i  ic  Min RSCP P – CCPCH  Max (TAdd P – CCPCH. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • C/I (dB) Atoll calculates the C/I on each pixel of the TX i  ic  best server coverage area. a pixel is TX i  ic  C Req CI – Q DwPCH  M DwPCH . The coverage prediction is calculated for a given set of a terminal type. TX  ic  i Where RSCP TCH – DL is given by: Max TX  ic  P TCH – DL  Service  i RSCP TCH – DL = ------------------------------------------Model LT 629 . For each layer.3 Downlink TCH RSCP Coverage This coverage prediction calculates and displays the RSCP for the downlink traffic channel.Atoll 3. Coverage consists of several independent layers whose visibility in the workspace can be managed. Each layer is assigned a colour and displayed with intersections between layers. The pixel colour depends on the cell edge coverage probability value. The pixel colour depends on the RSCP margin value. Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter. Atoll calculates the RSCP considering: • • • The the preferred carrier of the selected service. Otherwise. if a smart antenna is available in the downlink. Coverage consists of several independent layers whose visibility in the workspace can be managed. a mobility type. a carrier. The coverage prediction is calculated for a given set of a terminal type. The pixel colour depends on the RSCP level.4. For each layer. RSCP TCH – UL . G TX i SA = G DL and L TX i SA = L DL are calculated according to the smart antenna modelling method used. a pixel is TX i  ic  covered if RSCP TCH – DL  Threshold . Afterwards. Coverage consists of several independent layers whose visibility in the workspace can be managed. a service. For each layer. Each layer is assigned a colour and displayed with intersections between layers.RSCP P – CCPCH) RSCP TCH – UL  RSCP TCH – UL  Service Mobility  are covered and coloured according to the selected display option. G TX i and L TX i are read from the main antenna model. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • DL TCH RSCP (dBm) Atoll calculates the DL TCH RSCP on each pixel of the TX i  ic  best server coverage area. a pixel is covered if RSCP TCH – DL Req CECP  RSCP TCH – DL  Service Mobility  . the coverage prediction is calculated for the selected carrier. There are as many layers as thresholds defined in the Display tab (Prediction properties).Atoll 3. or the carrier with the highest P-CCPCH power. • Cell Edge Coverage Probability (%) Atoll calculates the cell edge coverage probability on each pixel of the TX i  ic  best server coverage area.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks TX ©Forsk 2015 Term i Model L Path  L  L  L Body  L Indoor  M Shadowing Max = ----------------------------------------------------------------------------------------------------------------------. There are as many layers as thresholds defined in the Display tab TX i  ic  (Prediction properties). Coverage consists of several independent layers whose visibility in the workspace can be managed. Each layer is assigned a colour and displayed with intersections between layers. There are as many layers as thresholds defined in the Display tab (Prediction properties). The pixel colour depends on the cell edge coverage probability value. there will not be any pixels covered by this transmitter. For each layer. 9. or from the master carrier in case of N-frequency mode compatible transmitters. Each layer is assigned a colour and displayed with intersections between layers.3. or the master carrier in case of N-frequency mode compatible transmitters. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. pixels in TX i  ic  the TX i  ic  Req coverage area where TX i  ic  Min RSCP P – CCPCH  Max (TAdd P – CCPCH. TX i  ic  Where RSCP TCH – UL is given by: Term TX i  ic  P Max RSCP TCH – UL = -------------Model LT 630 and . for Max P TCH – DL  Service  . If the selected carrier does not exist on a transmitter. if no preferred carrier is defined for the service. and for an uplink timeslot. If you perform this coverage prediction for the best carrier.4 Uplink TCH RSCP Coverage This coverage prediction calculates and displays the RSCP for the uplink traffic channel.and P TCH – DL  Service  is the maximum downlink traffic power TX i Term G G defined for the selected service. • RSCP Margin (dB) Atoll calculates the RSCP margin on each pixel of the TX i  ic  best server coverage area. a TX i  ic  Req RSCP pixel is covered if RSCP TCH – DL – RSCP TCH – DL  Service Mobility   M TCH – DL . Model LT Model In L T . 9.and P TCH – DL and P TCH – DL are respectively the downlink traffic TX Term i G G power and the other common control channel power for the selected timeslot.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 TX i Term Model L Path  L  L  L Body  L Indoor  M Shadowing Term = ----------------------------------------------------------------------------------------------------------------------. You can choose to display the minimum. the maximum.4. there will not be any pixels covered by this transmitter. Pixels are covered and coloured according to the total downlink noise thresholds defined in the display options. If you perform this coverage prediction for the best carrier. the most interfered carrier. The pixel colour depends on the RSCP margin value. For each layer. a pixel is covered if RSCP TCH – UL Req CECP  RSCP TCH – UL  Service Mobility  .G TX i SA = G UL and L TX i SA Term = L UL are calculated according to the smart antenna modelling method used. a mobility type.5 Downlink Total Noise This coverage prediction calculates and displays the total noise on the downlink. c.and P Max is the maximum uplink traffic power defined for the TX i Term G G selected terminal. • Cell Edge Coverage Probability (%) Atoll calculates the cell edge coverage probability on each pixel of the TX i  ic  best server coverage area. N Tot – DL .Atoll 3. for P Max . Each layer is assigned a colour and displayed with intersections between layers. The coverage prediction is calculated for a given set of a terminal type.3. a service.e. Each layer is assigned a colour and displayed with intersections between layers. a carrier. • RSCP Margin (dB) Atoll calculates the RSCP margin on each pixel of the TX i  ic  best server coverage area. or from the master carrier in case of N-frequency mode compatible transmitters. Coverage consists of several independent layers whose visibility in the workspace can be managed. Coverage consists of several independent layers whose visibility in the workspace can be managed.and RSCP OCCH = -------------With RSCP TCH – DL = ------------------Model Model LT LT TX i Term Model L Path  L  L  L Body  L Indoor  M Shadowing = ----------------------------------------------------------------------------------------------------------------------. Afterwards.. or the average total noise values from among the values calculated for all the carriers. Model LT 631 . There are as many layers as thresholds defined in the Display tab TX i  ic  (Prediction properties). If the selected carrier does not exist on a transmitter. and TS P TCH – DL P OCCH . Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter. The pixel colour depends on the cell edge coverage probability value. a TX  ic  i Req RSCP pixel is covered if RSCP TCH – UL – RSCP TCH – UL  Service Mobility   M TCH – UL . The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. For each layer. a pixel is TX i  ic  covered if RSCP TCH – UL  Threshold . Otherwise. and for a downlink timeslot. Each layer is assigned a colour and displayed with intersections between layers. i. G TX i and L TX i are read from the main antenna model. For each layer. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • UL TCH RSCP (dBm) Atoll calculates the UL TCH RSCP on each pixel of the TX i  ic  best server coverage area. Total downlink noise is given by: N Tot – DL =  Term  RSCP TCH – DL + RSCP OCCH  + N 0 All TX. Model LT Model In L T . Atoll calculates the downlink noise for all the carriers but keeps the worst case value. Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction properties). There are as many layers as thresholds defined in the Display tab (Prediction properties). the coverage prediction is calculated for the selected carrier. if a smart antenna is available in the uplink. The pixel colour depends on the RSCP level. a carrier. G TX i and L TX i are read from the main antenna model.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks Model In L T . for Max P TCH – DL  Service  .Atoll 3. if no preferred carrier is defined for the service. 9. a service. the coverage predictions are calculated for the selected carrier. and  -----  Q TCH – DL or  ---  Q TCH – DL are covered and N t TCH – DL I TCH – DL coloured according to the selected display option. for Max P TCH – DL  Service  . G Div = --------------------------------------------= --------------------------------------------Where  ----- DL  G DL and  --- DL TX  ic  TX i  ic  N t TCH – DL I TCH – DL i N Tot – DL N Tot – DL Max TX i  ic  P TCH – DL  Service  With RSCP TCH – DL = ------------------------------------------ Eb  Nt  DL LT TX  Eb  Nt  Term i DL L Path  L  L  L Body  L Indoor  M Shadowing Max LT = ----------------------------------------------------------------------------------------------------------------------. If you perform these coverage predictions for the best carrier. G Proc . or the master carrier in case of N-frequency mode compatible transmitters. Otherwise.RSCP P – CCPCH) . Atoll calculates the Eb/Nt or C/I considering: • • • the preferred carrier of the selected service. there will not be any pixels covered by this transmitter.and P TCH – DL  Service  is the maximum downlink traffic TX Term i G G power defined for the selected service. G TX i SA = G DL and L ©Forsk 2015 TX i SA = L DL are calculated according to the smart antenna modelling method used. and for a downlink timeslot. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power.  Eb  Nt  DL  Eb  Nt  In L T DL .3. G TX i SA = G DL and L TX i SA = L DL are calculated according to the smart antenna modelling method used. The pixels in TX i  ic  the coverage area where TX i  ic  TX i  ic  TX  ic  Min RSCP P – CCPCH  Max (TAdd P – CCPCH. N t TCH – DL I TCH – DL The coverage predictions are calculated for a given set of a terminal type. If the selected carrier does not exist on a transmitter. Otherwise. or from the master carrier in case of N-frequency mode compatible transmitters. TX i  ic  TX i TX i TX i  ic  TX  ic  i E b TXi  ic    RSCP TCH – DL   RSCP TCH – DL Div  C . G TX i and L TX i model. if a smart antenna is available in the downlink.  ----b- or  --- . TX  ic  i TX  ic  i TX  ic  i Term N Tot – DL = I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0 Where TX i  TX i  ic  TX i  ic  TX i  ic   TXi Ortho Term I Intra – DL =     1 – F DL    1 – F JD  +  1 –      RSCP TCH – DL + RSCP OCCH        TX i  ic  With RSCP OCCH TX i  ic  I Extra – DL = TX  ic  i P OCCH = --------------------- Eb  Nt  DL LT TX j  ic  TX j  ic    RSCPTCH – DL + RSCPOCCH  ji TX j  jc  TX j  jc    RSCPTCH – DL + RSCPOCCH  TX j I IC – DL  ic jc  = --------------------------------------------------------------------------F IRF  ic jc  632 are read from the main antenna . or the carrier with the highest P-CCPCH power.6 Downlink Service Area Analysis (Eb/Nt) or (C/I) E C These coverage predictions calculate and display the Eb/Nt or C/I on the downlink traffic channel. Afterwards. a mobility type.4. if a smart antenna is available in the downlink. TX i  ic  Eb C i Req Req Req RSCP TCH – DL  RSCP TCH – DL  Service Mobility  . Each layer is assigned a colour and displayed pixel is covered if  ----- N t TCH – DL I TCH – DL with intersections between layers. For each layer. There are as many layers as thresholds defined in the Display tab (Prediction properties).3. a pixel is covered if  E TXi  ic   Req Min   ----b-  Q TCH – DL  Threshold  N t TCH – DL  or  C TXi  ic   Req Min   ---  Q TCH – DL  Threshold . Each layer is assigned a colour and displayed with intersections between  I TCH – DL  layers. jc). The pixel colour depends on the required power level. a TX  ic  E b TXi  ic  C i  Threshold or  ---  Threshold . Max a pixel is covered Req  P TCH – DL  Service  or P TCH – DL = ---- N t TCH – DL if Req Max P TCH – DL – P TCH – DL  Service   M arg in . The pixel colour depends on the Eb/Nt or C/I level. where P TCH – DL = --------------------------TCH – DL  Service  or TX  ic  i E b  --- N t TCH – DL Req Q TCH – DL Req Max P TCH – DL = ------------------------ P TCH – DL  Service  . Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction properties). Each layer is assigned a colour and displayed with intersections between layers. a pixel is covered if  -----  N t TCH – DL  I  TCH – DL layer is assigned a colour and displayed with intersections between layers. Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 I IC – DL  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink. Req P TCH – DL = For each Req Q TCH – DL --------------------------TX i  ic   E b layer. There are as many layers as thresholds defined in the Display tab (Prediction properties). • Eb/Nt Margin or C/I Margin (dB) Atoll calculates the Eb/Nt or C/I margin on each pixel of the TX i  ic  best server coverage area. Coverage consists of several independent layers whose visibility in the workspace can be managed. For each layer. There are as many layers as thresholds defined in the Display tab (Prediction properties). 633 . Coverage consists of several independent layers whose visibility in the workspace can be managed. Coverage consists of several independent layers whose visibility in the workspace can be managed. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • Max Eb/Nt or Max C/I (dB) Atoll calculates the Eb/Nt or C/I on each pixel of the TX i  ic  best server coverage area. Req Q TCH – DL ------------------------TX i  ic  C ---  I  TCH – DL where Max  P TCH – DL  Service  . • Required Power (dBm) Atoll calculates the downlink required power on each pixel of the TX i  ic  best server coverage area. • Effective Eb/Nt or Effective C/I (dB) Atoll calculates the effective Eb/Nt or C/I on each pixel of the TX i  ic  best server coverage area. Req Q TCH – DL Req Req . There are as many layers as thresholds defined in the Display tab (Prediction properties). Coverage consists of several independent layers whose visibility in the workspace can be managed. The pixel colour depends on the effective Eb/Nt or C/I level. a pixel is covered if P TCH – DL  Threshold . P Max For each layer. • Required Power Margin (dB) Atoll calculates the downlink required power margin on each pixel of the TX i  ic  best server coverage area.Atoll 3. TX  ic  E b TX i  ic  C i Req Eb  Nt Req CI – Q TCH – DL  M TCH – DL or  --- – Q TCH – DL  M TCH – DL . which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic. Each layer is assigned a colour and displayed with intersections TX i  ic  C ---  I  TCH – DL between layers. The pixel colour depends on the Eb/Nt or C/I margin value. The pixel colour depends on the required power margin value. Each For each layer. Coverage Display It is possible to colour the transmitter service areas using a unique colour per transmitter.7 Uplink Service Area Analysis (Eb/Nt) or (C/I) E C These coverage predictions calculate and display the Eb/Nt or C/I on the uplink traffic channel.and P Max is the maximum power defined for the selected TX i Term G G terminal. there will not be any pixels covered by this transmitter. For each layer.4. Coverage consists of several independent layers whose visibility in the workspace can be managed. or from the master carrier in case of N-frequency mode compatible transmitters. Otherwise.or P Term . for P Max . and for an uplink timeslot. TX  ic  TX i  ic  E b TXi  ic  C i Req Req Req RSCP TCH – UL  RSCP TCH – UL  Service Mobility  .3. If the selected carrier does not exist on a transmitter.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks • ©Forsk 2015 Cell Edge Coverage Probability (%) Atoll calculates the cell edge coverage probability on each pixel of the TX i  ic  best server coverage area. G TX i and L TX i are read from the main antenna model. or colour the pixels in the coverage areas by any transmitter attribute or other criteria such as: • Max Eb/Nt or Max C/I (dB) Atoll calculates the Eb/Nt or C/I on each pixel of the TX i  ic  best server coverage area. Each layer is assigned a colour and displayed pixel is covered if  -----  N t TCH – UL  I  TCH – UL with intersections between layers. TX i  ic  C ---  I  TCH – DL For each layer. if no preferred carrier is defined for the service.  Eb  Nt  UL In L T . • 634 Effective Eb/Nt or Effective C/I (dB) . the coverage predictions are calculated for the selected carrier. Each layer is assigned a colour and displayed with intersections between layers. a TX i  ic  TX i  ic  Eb C  Threshold or  ---  Threshold . Atoll calculates the Eb/Nt or C/I considering: • • • The the preferred carrier of the selected service.and P Req = P Max  --------------------------With RSCP TCH – UL = ---------------------Req = P Max  ------------------------TX i  ic  TX i  ic   Eb  Nt UL E C b  ---  ----- LT  I  TCH – UL  N t TCH – UL  Eb  Nt  UL LT TX i  Eb  Nt  UL Term L Path  L  L  L Body  L Indoor  M Shadowing Term = ----------------------------------------------------------------------------------------------------------------------. or the carrier with the highest P-CCPCH power. if a smart antenna is available in the uplink. The pixel colour depends on the Eb/Nt or C/I level. N t TCH – UL I TCH – UL The coverage predictions are calculated for a given set of a terminal type.G TX i SA = G UL and L TX i SA Term = L UL are calculated according to the smart antenna modelling method used. There are as many layers as thresholds defined in the Display tab (Prediction properties).  ----b- or  --- . If you perform these coverage predictions for the best carrier. Coverage consists of several independent layers whose visibility in the workspace can be managed. CECP 9. a pixel is covered if TX i  ic  E b  --- N t TCH – DL Req  Q TCH – DL or CECP Req  Q TCH – DL . TX i  ic  TX i  ic  TX  ic  i E b TXi  ic    RSCP TCH – UL   RSCP TCH – UL Div C . There are as many layers as thresholds defined in the Display tab (Prediction properties).RSCP P – CCPCH) . Afterwards. G Div = ------------------------------------------------ G Where  ----- and UL UL UL TX i  ic  TX i  ic   I  TCH – UL = N t TCH – UL N Tot – UL N Tot – UL Term Term Req Req Term TX i  ic  Q TCH – UL Q TCH – UL P Max Term Term Term . pixels in TX i  ic  the coverage area where TX i  ic  Min RSCP P – CCPCH  Max (TAdd P – CCPCH. G Proc --- ------------------------------------------------. a mobility type. or the master carrier in case of N-frequency mode compatible transmitters.Atoll 3. and  -----  Q TCH – UL or  ---  Q TCH – UL are covered and N t TCH – UL I TCH – UL coloured according to the selected display option. The pixel colour depends on the cell edge coverage probability value. a service. a carrier. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. There are as many layers as thresholds defined in the Display tab (Prediction properties).0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Atoll calculates the effective Eb/Nt or C/I on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on the required power level. The pixel colour depends on the Eb/Nt or C/I margin value. CECP 9. Each layer is assigned a colour and displayed with intersections between layers. Each layer is assigned a colour and displayed with intersections between  I TCH – UL  layers. where P Req = --------------------------Max or TX i  ic  E b  --- N t TCH – UL Req Q TCH – UL Term .  ----b- or  --- and  N t TCH – DL  I  TCH – DL 635 . Coverage consists of several independent layers whose visibility in the workspace can be managed. a pixel is covered if P Req – P Max  M arg in .Atoll 3. Each layer is assigned a colour and displayed with intersections between layers. a pixel is covered if  ----b- – Q TCH – UL  M TCH – UL or  --- – Q TCH – UL  M TCH – UL . TX  ic  i C  -- I  TCH – UL • Required Power Margin (dB) Atoll calculates the uplink required power margin on each pixel of the TX i  ic  best server coverage area. These coverage E C predictions calculate the Eb/Nt or C/I on the downlink and uplink traffic channels. Coverage consists of several independent layers whose visibility in the workspace can be managed. Coverage consists of several independent layers whose visibility in the workspace can be managed. For each layer.8 Effective Service Area Analysis (Eb/Nt) or (C/I) These coverage predictions consist of pixels covered by the both the uplink and the downlink service areas. where Q TCH – UL Term . TX i  ic  C ---  I  TCH – UL For each layer. For each layer.3. a pixel is covered  E b TXi  ic   Req Min   -----  Q TCH – UL  Threshold  N t TCH – UL  if or  C TXi  ic   Req Min   ---  Q TCH – UL  Threshold . P Term P Req = --------------------------Max TX i  ic  E  ----b-  N t TCH – UL or Req Q TCH – UL Term . There are as many layers as thresholds defined in the Display tab (Prediction properties). • Required Power (dBm) Atoll calculates the uplink required power on each pixel of the TX i  ic  best server coverage area. Req For each layer. a pixel is covered if Term P Req  Threshold . The pixel colour depends on the required power margin value. TX  ic  E TXi  ic  C i Req Eb  Nt Req CI For each layer. • Eb/Nt Margin or C/I Margin (dB) Atoll calculates the Eb/Nt or C/I margin on each pixel of the TX i  ic  best server coverage area. There are as many layers as thresholds defined in the Display tab (Prediction properties). Each layer is assigned a colour and displayed with intersections between layers. P Term properties). Coverage consists of several independent layers whose visibility in the workspace can be managed. TX  ic  i C  -- I  TCH – UL • Cell Edge Coverage Probability (%) Atoll calculates the cell edge coverage probability on each pixel of the TX i  ic  best server coverage area. Each  N t TCH – UL  I  TCH – UL layer is assigned a colour and displayed with intersections between layers. a pixel is covered if TX i  ic  E b  --- N t TCH – UL Req  Q TCH – UL or CECP Req  Q TCH – UL . There are as many layers as thresholds defined in the Display tab (Prediction properties). The pixel colour depends on the effective Eb/Nt or C/I level. The pixel colour depends on the cell edge coverage probability value. P Term P Req = ------------------------Max . Coverage consists of several independent layers whose visibility in the workspace can be managed. P Term P Req = ------------------------Max .4. There are as many layers as thresholds defined in the Display tab (Prediction Req Q TCH – UL Term Term Term . service. the coverage prediction is calculated for the selected carrier.Atoll 3. and . or the carrier with the highest P-CCPCH power. Afterwards.RSCP P – CCPCH) • RSCP TCH – DL  RSCP TCH – DL  Service Mobility  • RSCP TCH – UL  RSCP TCH – UL  Service Mobility  • E b i C i Req Req  --- Q TCH – DL or  ---  Q TCH – DL for any of the 6 timeslots  N t TCH – DL  I  TCH – DL • E b C Req Req  --- Q TCH – UL or  ---  Q TCH – UL for any of the 6 timeslots  N t TCH – UL  I  TCH – UL TX i  ic  Req TX i  ic  Req TX  ic  TX  ic  TX i  ic  TX i  ic  9.9 Cell to Cell Interference This coverage prediction calculates and displays the interference received by cells receiving in uplink from other cells which are transmitting in downlink. If the selected carrier does not exist on a transmitter. if no preferred carrier is defined for the service. If you perform these coverage predictions for the best carrier. Atoll calculates the Eb/Nt or C/I considering: • • • the preferred carrier of the selected service. and a timeslot. the direction of the link changes twice (downlink to uplink. and display the pixels where both downlink and uplink Eb/Nt or C/I are above the required quality  N t TCH – UL  I  TCH – UL thresholds. = ------------------TX TX j LT LT j L Ant L Ant using a smart antenna. the coverage predictions are calculated for the selected carrier. and then uplink to downlink). or the carrier with the highest P-CCPCH power. or the master carrier in case of N-frequency mode compatible transmitters. a mobility type. on the same carrier ic or on another carrier jc. the cell to cell interference is given by: TX  jc  j I C2C  TX i TX j  = TX  ic  j TX  ic  j TX  jc  j   RSCPTCH – DL + RSCPOCCH  j   RSCPTCH – DL + RSCPOCCH  + --------------------------------------------------------------------------F IRF  ic jc  TX TX j TX j  ic  Where TX  ic  P TCH – DL    j RSCP TCH – DL = --------------------------LT TX  ic  j RSCP TCH – DL 636 TX j  ic  TX j TX j  jc  and TX  jc  P TCH – DL    j RSCP TCH – DL = --------------------------LT TX j  jc  TX j TX  jc  P TCH – DL G Ant P TCH – DL G Ant . ---------.3. The coverage prediction is calculated for a given set of a terminal type. a mobility type. a service. The coverage predictions are calculated for a given set of a terminal type. or the master carrier in case of N-frequency mode compatible transmitters. These transitions are referred to as switching points. and terminal are used to calculate the best server coverage of the interfered cell. During each subframe. The pixels in the TX i  ic  coverage area are covered and coloured according to the selected display option if all the following conditions are satisfied: TX i  ic  Min • RSCP P – CCPCH  Max (TAdd P – CCPCH. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. and for all the 6 timeslots. Afterwards. there will not be any pixels covered by this transmitter. a service.4. a carrier. a carrier. Assuming that a transmitter TX j is interfering a studied transmitter TX i on a timeslot.and RSCP TCHj – DL = ------------------. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest PCCPCH power. If the selected carrier does not exist on a transmitter. The timeslot configuration of each cell defines the direction of the link at any given instance.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 E b C  ---or  --- . The mobility. if no preferred carrier is defined for the service.otherwise. If you perform this coverage prediction for the best carrier. ---------. or from the master carrier in case of N-frequency mode compatible transmitters. Atoll calculates the RSCP considering: • • • the preferred carrier of the selected service. or from the master carrier in case of N-frequency mode compatible transmitters. there will not be any pixels covered by this transmitter. Each layer is assigned a colour and displayed with intersections between layers. a pixel is covered if I C2C  TX i TX j   Threshold . You can choose to display the minimum. The pixels in the TX i  ic  coverage area are covered and coloured if: • TX  ic  i RSCP P – CCPCH  TAdd P – CCPCH  Mobility  . a carrier. However. If some cells in a network use UpPCH shifting. i. The coverage prediction is calculated for a given set of an HSDPA terminal type. This coverage prediction calculates and displays the uplink interference on the TS1. The coverage prediction is calculated for a given set of a terminal type. from unsynchronised DwPTS or TS0 timeslots of other cells. If you perform this coverage prediction for the best carrier. TX i  ic  TX i The uplink interference on TS1 is given by: I TS1 – UL = N 0 TX i  ic  X TS1 – UL  ---------------------------------TX  ic  1 – X i  TS1 – UL  9. For each layer. if no preferred carrier is defined for the service. the maximum. and for all downlink timeslots. 9. TX j G Ant is the main antenna gain. the coverage prediction is calculated for the selected carrier.otherwise. If the selected carrier does not exist on a transmitter or if it does not support HSDPA. Atoll calculates the interference for all the carriers but keeps the worst case value.4. a mobility type. The coverage prediction is calculated using the main antenna. it is possible to shift the UpPCH to TS1. or from the master carrier in case of N-frequency mode compatible transmitters.10 UpPCH Interference UpPCH is usually carried by the UpPTS timeslot. TX  ic  i TX  ic  i Pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and I TS1 – UL  Threshold are covered and coloured according to the selected display option. an HSDPA service.  is the angle for the smart antenna pattern. ---------. or the carrier with the highest P-CCPCH power. the most interfered carrier. the coverage predictions are calculated for the selected carrier. TX j L Ant is the main antenna attenuation. Afterwards. a carrier. if the interference on UpPTS is high. and RSCP OCCH = --------------= ---------------. or from the master carrier in case of N-frequency mode compatible transmitters. there will not be any pixels covered by this transmitter. ---------TX TX LT LT j j L Ant L Ant TX j TX i  L TX  L RX is the path loss calculated using the ITU526-5 propagation model without antenna loss. a service.. The pixel colour depends on the cell to cell interference level.4.e.Atoll 3. I TS1 – UL . there will not be any pixels covered by this transmitter.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 TX  ic  j RSCP OCCH TX  ic  j ITU526 – 5 L T = L Path ITU526 – 5 L Path TX TX  jc  j j TX j TX  jc  P OCCH G Ant P OCCH G Ant j .11 HSDPA Predictions This coverage prediction calculates and displays the peak RLC throughput or the Peak MAC throughput per pixel covered by HSDPA cells. Coverage consists of several independent layers whose visibility in the workspace can be managed.3. and for TS1. If you perform these coverage predictions for the best carrier. There are as many layers as thresholds defined in the Display tab (Prediction properties). a mobility type. you can use this coverage prediction to study the interference on the shifted UpPCH of these cells from other cells. If the selected carrier does not exist on a transmitter. This is called UpPCH shifting. 637 . The interference from other cells is in this case generated by the traffic on the TS1 of interfering cells. or the average total noise. Atoll calculates the RLC or Peak MAC throughput considering: • • • the preferred carrier of the selected service. Afterwards. The uplink interference on TS1 is calculated from the uplink load calculated in the simulations or manually defiend for the TS1. or the master carrier in case of N-frequency mode compatible transmitters. Atoll calculates the cell to cell interference on each pixel of the TX i  ic  best server coverage area. The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power. The best server for the coverage prediction is determined according to the PCCPCH RSCP from the carrier with the highest P-CCPCH power.  N t HS – PDSCH • Peak RLC Throughput: After selecting the bearer. • E b TXi  ic TS  Max DL A-DPCH Eb/Nt: Atoll displays the A-DPCH Eb/Nt at the receiver (  ----- ) for the best server and the  N t TCH – DL – Max selected timeslot. it is defined for each HSDPA bearer in the related table. Atoll reads the corresponding RLC peak throughput. 638 TX i  ic  • HS-SCCH Power: On each pixel. Each layer is assigned a colour and displayed with intersections between layers. Atoll calculates RSCP HS – SCCH for all timeslots and selects the lowest value.Atoll 3. Coverage consists of several independent layers whose visibility in the workspace can be managed. No power control is performed as in simulations. and  N t HS – PDSCH • E C  ---is enough to select a bearer for the pixels. S Block is the transport block size (in kbits) of the selected HSDPA bearer. The pixel colour depends on the peak RLC throughput. • E b TX i  ic TS  Max UL A-DPCH Eb/Nt: Atoll displays the A-DPCH Eb/Nt at the best server (  ----- ) and the selected  N t TCH – UL – Max timeslot. Atoll calculates P HS – SICH for the selected timeslot. Atoll calculates  ----c- for the selected timeslot. • HS-SCCH RSCP: On each pixel. Here. Atoll determines uplink traffic channel quality for the maximum terminal power allowed. There are as many layers as thresholds defined in the Display tab (Prediction properties).3. Atoll calculates RSCP HS – SCCH for all timeslots and calculates the average of TX i  ic  these values.e -----------. see "HSDPA Part of the Algorithm" on page 622. There are as many layers as thresholds defined in the Display tab (Prediction properties). TX i  ic  Mi .  N t HS – PDSCH • E Average HS-PDSCH Ec/Nt: On each pixel. No power control is performed as in simulations. • E TXi  ic  Max HS-PDSCH Ec/Nt: On each pixel. Atoll calculates  ----C- for all timeslots and selects the highest value. The Peak MAC MAC throughput is calculated as follows: MAC R DL = S Block  500 Where. This is the highest throughput that the bearer can provide on each pixel. Atoll calculates  ----C- for all timeslots and selects the lowest value.  N t HS – PDSCH TX i  ic  For more information on HSDPA bearer selection.blocks per second). Atoll calculates  ----C- for all timeslots and calculates the average N t HS – PDSCH TX i  ic  TX i  ic  of these values. i. Coverage consists of several independent layers whose visibility in the workspace can be managed. • Average HS-PDSCH RSCP: On each pixel. Atoll determines downlink traffic channel quality at the receiver for a maximum traffic channel power allowed for the best server. Each layer is assigned a colour and displayed with intersections between layers. • E TX i  ic  HS-SCCH Ec/Nt: On each pixel.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 TX  ic  i • E C Req  --- Q HS – PDSCH . • E Min HS-PDSCH Ec/Nt: On each pixel. Here. 4 The pixel colour depends on the Peak MAC throughput. • Peak MAC Throughput: Atoll displays the Peak MAC throughput ( R DL ) provided on each pixel. TX i  ic  • Max HS-PDSCH RSCP: On each pixel. Coverage Display It is possible to colour the pixels in the coverage areas by criteria such as: TX  ic  i • Min HS-PDSCH RSCP: On each pixel. Atoll calculates RSCP HS – SCCH for the selected timeslot. N t HS – SCCH • HS-SICH Power: On each pixel. The value 500 corresponds to the number of blocks per second (there are 4 blocks per TTI and 2000 2000 TTI in one second. Atoll calculates RSCP HS – SCCH for all timeslots and selects the highest value. Atoll calculates P HS – SCCH for the selected timeslot. Each beam of a GOB has a different azimuth so that the GOB as a whole covers an entire sector. If a smart antenna model is only downlink or only uplink. N t HS – SICH • HS-PDSCH RSCP: On each pixel. Atoll determines the most suitable beam from the GOB for each user served by the smart antenna. called GOB. Atoll determines the gains and losses from the angular distributions calculated during the simulations for each timeslot and stored in the Cell Parameters per Timeslot table.1 Grid of Beams Modelling A grid-of-beams smart antenna. Atoll calculates RSCP HS – SICH for the selected timeslot. Atoll calculates  ----c- for the selected timeslot. the best beam is the one among all the beams of a GOB that has the highest difference between gain. and L DL = L TX = L Total – DL 9. N t HS – PDSCH M i TX  ic  i 9. the other direction uses the main antenna gain and losses for calculations. In words. L DL .0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 M i • HS-SICH RSCP: On each pixel. L DL = L DL and G UL = G Ant . • If a smart antenna is available on the downlink and uplink: TX SA TX SA TX SA TX SA G UL = G UL .5. Atoll determines the gains and losses using the smart antenna models. and whose horizontal patterns are pointed towards different directions as shown in the figure below: 639 . Therefore. and L UL ) are determined from the selected best beam. • E c TXi  ic  HS-PDSCH Ec/Nt: On each pixel.1 Modelling in Simulations 9.5. During the simulations.5 Smart Antenna Modelling Atoll calculates the smart antenna gains and losses in the direction of a user during the simulations. Atoll calculates  ----- for the selected timeslot. and vertical attenuations of the beams of the GOB. consists of more than one directional antenna pattern (beam) in different directions. • E HS-SICH Ec/Nt: On each pixel. The most suitable beam (best beam) is the one which provides the highest gain towards the served user: BeamBest = Beam H H V Max  G Beam – L Beam – L Beam V Where G Beam . L UL = L UL and G DL = G DL . L DL = L • TX If no smart antenna equipment is defined: TX TX TX TX G DL = G UL = G Ant . and L Beam are the gains. L DL = L DL • If a smart antenna is available on the downlink only: TX SA TX SA TX TX TX G DL = G DL . and in the direction of each pixel in coverage predictions. L Beam .Atoll 3. and horizontal and vertical SA SA SA SA attenuations.1. horizontal. Atoll calculates RSCP HS – PDSCH for the selected timeslot. The gains and losses of the GOB ( G DL . L UL = L TX TX = L Total – UL .3. G UL . The following example shows how Atoll calculates the GOB gains and losses. Example: Let us assume a GOB with 5 beams that have the same vertical patterns. L UL = L UL and G DL = G Ant . During simulations. In coverage predictions. L UL = L • = L Total – UL TX = L Total – DL If a smart antenna is available on the uplink only: TX SA TX SA TX TX TX G UL = G UL . as shown in the figure below.21 15 18 . If the user is located at  = 70 azimuth. SA SA H V G UL = 18 dB and L UL = L Beam + L Beam = 17.3: Grid Of Beams Modelling Let us assume that all the beams and the main antenna have the same 18 dBi gain.79 -30° 18 60 15 18 . which has the highest gain towards  . SA SA H V G DL = 18 dB and L DL = L Beam + L Beam = 17.60 . the total gain of the beam at 60° is the highest.15 -57 30° 18 60 15 18 . Therefore this beam is selected as the best beam.21 dB 640 .60 . and the vertical attenuation at the user location is 15 dB.4: GOB Modelling .15 -57  Transmitter Centre of the pixel where the served user is located  Angle between the user and the transmitter azimuth Figure 9.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Figure 9.15 0.3. which is also the same for all the beams because we assume that the vertical patterns are the same.60 .15 -57 -60° 18 60 15 18 .21.60 . If this beam has been selected in the downlink. as follows: Beam Gain (dBi) Horizontal Vertical Attenuation (dB) Attenuation (dB) G Beam – L Beam – L Beam Total Gain (dB) H V 0° 18 60 15 18 .15 -57 60° 18 2.2. Atoll determines the best beam.21 dB If this beam has been selected in the uplink.Determination of the Best Beam In our example.Atoll 3. Urban and dense urban clutter types introduce more multipath and spread the signal at a wider angle than an open or rural clutte type.Atoll 3.5: Adaptive Beam Modelling . the gain and losses of the adaptive beam at  are: SA SA H V G DL = 18 dB and L DL = L Beam + L Beam = 15 dB If the adaptive beam smart antenna is selected in the uplink. Different clutter types have different spreading effects on the propagation of radio waves. L Beam = 0 .1.5.3.  Spread . These values are used in interference calculation to determine the downlink interfering signal due to transmission towards the served user. Example: Let us assume an adaptive beam smart antenna selected for a transmitter along with a main antenna. Let us assume that the adaptive beam and the main antenna have the same 18 dBi gain. 641 . as well as for calculating the uplink interfering signals received at transmitter when decoding signal received from the served user. it reads the spreading angle from the clutter class properties.1. Atoll determines the clutter class of the served user. in the smart antenna equipment based on the statistical model. and reads the smart antenna C/ I gain defined for the Probability = 1 – TProb SA corresponding to the spreading angle. The following example shows how Atoll calculates the statistical C/I gains and losses. During the simulations. If the user is located at  = 60 azimuth. 9.5. as shown in the figure below:  Transmitter Centre of the pixel where the served user is located  Angle between the user and the transmitter azimuth Figure 9. To find the smart antenna gain.2 Adaptive Beam Modelling An adaptive beam smart antenna is capable of steering a given antenna pattern towards the direction of the served signal.Determination of the Best Beam If the adaptive beam smart antenna is selected in the downlink. You can create smart antenna equipment in Atoll based on the statistical approach by providing C/I gains and their cumulative probabilities for different spreading angles. and the vertical attenuation at the user location is 15 dB. In Atoll.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 9. it reads the probability threshold from the smart antenna properties. and the adaptive beam losses ( L DL and SA H V L UL ) are the horizontal and vertical pattern attenuations L Beam + L Beam towards the user direction.3 Statistical Modelling A statistical modelling approach is also available in Atoll which can be used to model the effect of smart antennas through C/ I gains. as the ideal beam steering algorithm steers the beam towards the served user. You can assign a spreading angle to each clutter class in your document. called a beam because of its highly directional shape. you can enter the C/I gains and their cumulative probabilities for each spreading angle. For each smart antenna equipment based on statistical modelling. The following example shows how Atoll calculates the adaptive beam gains and losses. TProb SA . Atoll reads the clutter class in which the served user is located to determine the spreading angle. this adaptive beam is oriented in the direction of each served user in order to model the effect of the smart antenna. the gain and losses of the adaptive beam at  are: SA SA H V G UL = 18 dB and L UL = L Beam + L Beam = 15 dB H In fact. you can set a probability threshold. this is modelled using a single antenna pattern. Once you have assigned the spreading angles to clutter classes. SA SA SA The adaptive beam gains ( G DL and G UL ) are the antenna gains defined for the beam. . Atoll has a number of points. If a gain for the exact probability value of 20% is not defined. Atoll will read the smart antenna C/I gain G for Prob = 20 % . The downlink traffic powers. using the same adaptive beam pointed towards the served user.7196 dB . The following example explains how the geographic distribution of downlink traffic power is created. i.1.e.e.4% = 4. Once Atoll has calculated the downlink traffic power and the uplink load using the smart antenna gains and losses determined as explained in the previous section. Example: Let us assume a smart antenna equipment using adaptive beam modelling.5. 9. available with the values of these results.2 Construction of the Geographic Distributions During simulations. The angular step defined for the simulations is  Step = 30 .1. Atoll uses the Angular Step value that you set when creating and running simulations to construct the geographic distributions of these results. which can be used in coverage predictions. is constructed by connecting the resulting value points. downlink traffic power and uplink loads. i. Atoll uses the smart antenna model selected for each transmitter to calculate the smart antenna gains and losses.6941 dB The smart antenna gains are the same for uplink and downlink.3. at the location of a given user. Therefore. Negative values of C/I gains are considered as losses.4 Beamforming Smart Antenna Models See "Beamforming Smart Antenna Models" on page 43. The smart antenna equipment SA SA has TProb = 80 % . only. 12 points. The downlink traffic power at the served user (W) with the adaptive beam pointing in the user’s direction is P W . If G SA Prob = 19% = 4. At the end of the simulations.5 3rd Party Smart Antenna Modelling 3rd party smart antenna models can be used in Atoll to determine the gains and losses during the simulations for a given user distribution generated.e.. Atoll linearly interpolates the gain value from the two surrounding values. The geographic distribution of uplink loads is constructed in the same manner. it calculates the same for points located at the angle equal to that of the Angular Step of the simulations. 9. 9.. These values are calculated and stored for each user generated for the simulations. Therefore. i.5. points.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Example: Let us assume that the served user is located at a an urban clutter class with  Spread = 10 . Angular Step apart. at the 12 other points are also determined.Atoll 3. 642 . these values are calculated and are available for the given locations of the users. The geographic distribution of these results.5. the results are calculated for each point located at regular steps of 30 . then G SA Prob = 20% = 4. Their are no losses for this type of smart antenna equipment. The smart antenna gains and losses are used during the simulations and the results are stored in the Cell Parameters per Timeslot table.6298 dB and G SA Prob = 20. the latter will be computed 45 times faster than the first.7: Geographic Distribution of Downlink Traffic Power The accuracy of the geographic distribution depends upon the value of the angular step.5. But. 9. The main results of Monte Carlo simulations used in coverage predictions are: • If a smart antenna is used in both uplink and downlink: Geographic distribution of UL load X • UL –  DL –  and DL traffic power P Traffic If a smart antenna is used in downlink only: DL –  Geographic distribution of DL traffic power P Traffic • Without smart antenna: UL load X UL DL and DL traffic power P Traffic 643 .3.3 Modelling in Coverage Predictions The results of Monte Carlo simulations.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Figure 9. The value of the Angular Step should be the best possible compromise between calculation speed and accuracy. and can be used to carry out coverage predictions. can be stored in the Cells and in the Cell Parameters per Timeslot tables. Figure 9. for example. A radiation pattern created at a 1 step will be much more accurate than one created at 45 .Atoll 3. including the smart antenna results.6: Construction of the Geographic Distribution of Downlink Traffic Power The resulting geographic distribution is formed by linearly joining the obtained results. Figure 9. a mobility. we have: 644 . Therefore. Fast link adaptation is modelled in a dedicated HSDPA coverage prediction. this depends on the option selected in Global parameters (HSDPA part): CQI based on P-CCPCH quality or CQI based on HS-PDSCH quality (CQI means channel quality indicator). deduces the best HSDPA bearer that can be used and selects the suitable bearer so as to comply with cell and terminal user equipment capabilities. and reading the uplink load and downlink traffic power from the geographic distribution results. P Traffic  30 dBm . available in Global Parameters. For example. If an exact value for the angle is not available. For each pixel.4 HSDPA Quality and Throughput Analysis Fast link adaptation (or Adaptive Modulation and Coding) is used in HSDPA.5.5.3. the way of calculating the dedicated HSDPA study depends on if CQI is based on the P-CCPCH quality or on the HS-PDSCH quality. it determines the HS-PDSCH CQI. and the number of codes every 2 ms during a communication. Then. and the number of codes are changed to adapt to the radio conditions variations. Atoll proceeds as follows. This receiver may be using a specific carrier or all of them. Based on the reported channel quality indicator (CQI).4. For a pixel located at  = 315 . Once the bearer selected. Atoll finds the highest downlink throughput that can be carried at each bin and may deduce the application throughput.Atoll 3. The coverage prediction can be calculated for an HSDPA compatible terminal. and HSDPA service. the figure below shows the distribution of downlink traffic power and uplink traffic load results from a DL – 315 simulation. Atoll determines the downlink traffic powers and the uplink loads from all the transmitters.5. Atoll calculates on each bin either the best pilot quality (P-CCPCH Ec/Nt) or the best HS-PDSCH quality (HS-PDSCH Ec/Nt). The probe receiver does not create any interference. and a downlink timeslot.1 Fast Link Adaptation Modelling As explained above. The power on the HS-DSCH channel is transmitted at a constant power while the modulation. a carrier. 9. ic  corresponds to the P-CCPCH quality.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 The uplink load and the downlink traffic power at a given pixel are determined by calculating the angle  of that pixel with respect to the transmitter azimuth. 9.  Nt  P – CCPCH Two options.75 % . the downlink traffic power P Traffic DL – 315 from these results. the coding. an HSDPA service. the coding. mobility. The probe receiver on each bin is allocated the cell’s HSDPA. In this example.1. Smart antenna results are taken into account in the computation of this study.4. and X UL – 315 and the uplink load X UL – 315 are read = 2. the Node-B may change the modulation (QPSK and optionally 16QAM). P-CCPCH Quality Calculation Ec Let us assume the following notation:  -----. the load and power are determined using linear interpolation for the given angle between two available values.1 CQI Based on P-CCPCH Quality When the option “CQI based on CPICH quality” is selected. Coverage area is limited by the RSCP P-CCPCH threshold. may be used to calculate Nt: option Without useful signal or option Total noise. Let us assume each bin on the map corresponds to a probe receiver with HSDPA capable terminal.8: Geographic Distribution of downlink traffic power and uplink load 9. 645 .  Nt  ic  P – CCPCH = -------------------------------------------------------------------DL N Tot  ic  And TXi  BTS    RSCP P – CCPCH  ic  Ec ---- ic  = -----------------------------------------------------------------------------------------------------------.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 TXi  BTS    RSCP P – CCPCH  ic  Ec ---- . CQIP – CCPCH is deduced from the table Ec  . which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic. This table is defined for the terminal reception equipment and the specified mobility. P P – CCPCH  ic  TXi RSCP P – CCPCH  ic  = ------------------------------LT CI L Path  L TX  L Term  L Body  L Indoor  M Shadowing L T = ---------------------------------------------------------------------------------------------------------------------G TX  G Term Term  BTS . jc).  Nt  HS – SCCH DL DL Term TXi N Tot  ic  –  1 – F Ortho    1 – F JD    BTS  RSCP HS – SCCH  ic  With DL DL DL DL Term N Tot  ic  = I Intra  ic  + I Extra  ic  + I Inter – Carrier  ic jc  + N 0 DL I Inter – Carrier  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink. we have: P HS – PDSCH  ic  = P HSDPA  ic  – n HS – SCCH  P HS – SCCH  ic  n HS – SCCH is the number of HS-SCCH channels and P HS – SCCH  ic  is the HS-SCCH power on carrier ic. jc).  Nt  P – CCPCH DL TXi N Tot  ic  –  1 –     BTS  RSCP P – CCPCH  ic  With DL DL DL DL Term N Tot  ic  = I Intra  ic  + I Extra  ic  + I Inter – Carrier  ic jc  + N 0 DL I Inter – Carrier  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink. P-CCPCH CQI Determination Let us assume the following notation: CQI P – CCPCH corresponds to the P-CCPCH CQI.3. It is specified in mobility  Nt  HS – SCCH properties.for the total noise option.  and N 0 are defined in "Definitions and Formulas" on page 595. P HS – SCCH  ic  is controlled so as to reach the required HS-SCCH Ec/Nt (  -----. P HSDPA  ic  is the power available for HSDPA on the carrier ic. ic  ).Atoll 3. P HSDPA  ic  = P HS – PDSCH  ic  + n HS – SCCH  P HS – SCCH  ic  Therefore.for the without useful signal option.  Nt  HS – SCCH DL N Tot  ic  And TXi  BTS  RSCP HS – SCCH  ic  Ec ---- ic  = -------------------------------------------------------------------------------------------------------------------------------------------------------. which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic. We have: TXi  BTS  RSCP HS – SCCH  ic  Ec ---- ic  = ---------------------------------------------------------.for the without useful signal option. It is either fixed by the Req Ec user. CQIP – CCPCH = f   -----. ic    Nt  P – CCPCH HS-PDSCH Quality Calculation Atoll proceeds as follows: 1st step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ).for the total noise option. This parameter is a user-defined cell input.  Nt  ic  HS – PDSCH DL N Tot  ic  And TXi  BTS  RSCP HS – PDSCH  ic  Ec ----= ----------------------------------------------------------------------------------------------------------------------------------------------------------. With DL DL DL DL Term N Tot  ic  = I Intra  ic  + I Extra  ic  + I Inter – Carrier  ic jc  + N 0 DL I Inter – Carrier  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink.for the without useful signal option. Therefore. Let us assume the following notation: CQI HS – PDSCH corresponds to the HS-PDSCH CQI. jc). F JD Term and N 0 are defined in "Definitions and Formulas" on page 595. P HS – PDSCH  ic  TXi RSCP HS – PDSCH  ic  = ---------------------------------LT And CI L Path  L TX  L Term  L Body  L Indoor  M Shadowing L T = ---------------------------------------------------------------------------------------------------------------------G TX  G Term Term  BTS .  ic   Nt  HS – PDSCH TXi RSCP HS – PDSCH  ic  DL DL Term N Tot  ic  –  1 – F Ortho    1 – F JD    BTS  --------------------------------------------n Here. F JD Term and N 0 are defined in "Definitions and Formulas" on page 595. HS-PDSCH CQI Determination The best bearer that can be used depends on the HS-PDSCH CQI. Atoll calculates the HS-PDSCH quality Ec Let us assume the following notation:  -----.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 P HS – SCCH  ic  TXi RSCP HS – SCCH  ic  = ------------------------------LT and CI L Path  L TX  L Term  L Body  L Indoor  M Shadowing L T = ---------------------------------------------------------------------------------------------------------------------G TX  G Term Term  BTS .Atoll 3.3.  Nt  HS – PDSCH Therefore.for the total noise option. Atoll deduces CQI HS – PDSCH as follows: CQI HS – PDSCH = CQI P – CCPCH – P P – CCPCH + P HS – PDSCH 646 . Req EcDL   ---- ic   N Tot  ic   HS – SCCH   Nt  TXi RSCP HS – SCCH  ic  =  -------------------------------------------------------------------  L T for the total noise option. F Ortho . Atoll works on the assumption that five HS-PDSCH channels are used (n=5). Req Ec    1 +  1 – F DL    1 – F Term    ----   ic  Ortho JD  BTS   Nt  HS – SCCH  2nd step: Then. we have: TXi  BTS  RSCP HS – PDSCH  ic  Ec ---- = ------------------------------------------------------------. which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic. F Ortho .   BTS    And Req EcDL  ----   ic   N Tot  ic   Nt  HS – SCCH   TXi RSCP HS – SCCH  ic  =  -------------------------------------------------------------------------------------------------------------------------------------------  L T for the without useful signal option. ic  corresponds to the HS-PDSCH quality. 647 . Bearer characteristics are provided in the HSDPA Bearer table. Atoll finds the best bearer that can be used in the table Best Bearer=f(HS-PDSCH CQI). Then.3.10: UE Categories Table HSDPA cell capabilities are: • Maximum number of HS-PDSCH channels: 15.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Bearer Selection Knowing the HS-PDSCH CQI. Atoll checks if best bearer characteristics are compliant with cell and user equipment category capabilities. This table is defined for the terminal reception equipment and the specified mobility.48 Mb/s Figure 9. Characteristics of this bearer are: • • • • Transport block size: 9719 Bytes Number of HS-PDSCH channels used: 7 16QAM modulation used: Yes Peak Throughput: 4. Its capabilities are: • • • • Maximum transport block size: 7298 Bytes Maximum number of HS-PDSCH channels used: 5 16QAM modulation used: Yes Minimum number of TTI between two TTI used: 2 Figure 9. Assuming the best bearer = 23.9: Radio Bearers Table Assuming user equipment category = 3.Atoll 3. Atoll selects the bearer which is the best bearer compliant with the cell and UE category capabilities.  Nt  ic  HS – SCCH DL DL Term TXi N Tot  ic  –  1 – F Ortho    1 – F JD    BTS  RSCP HS – SCCH  ic  With DL DL DL DL Term N Tot  ic  = I Intra  ic  + I Extra  ic  + I Inter – Carrier  ic jc  + N 0 DL I Inter – Carrier  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink. F Ortho . when the method “Without useful signal” is used. 648 Term and N 0 are defined in "Definitions and Formulas" on page 595. • • • The number of HS-PDSCH channels (5) does not exceed the maximum number of HS-PDSCH channels the terminal can use (5) and the maximum number of HS-PDSCH channels available at the cell level (15). jc).5. HS-PDSCH Quality Update Once the bearer selected. And the transport block size (9719 Bytes) exceeds the maximum transport block size (7298 Bytes) the terminal can carried. ic  ) specified in mobility properties. This parameter is a user-defined cell input.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 The bearer 23 cannot be selected because: • • The number of HS-PDSCH channels (7) exceeds the maximum number of HS-PDSCH channels the terminal can use (5).for the total noise option. Atoll searches a suitable bearer and selects the bearer index 22. P HS – SCCH  ic  TXi RSCP HS – SCCH  ic  = ------------------------------LT i And CI L Path  L TX  L Term  L Body  L Indoor  M Shadowing L T = ---------------------------------------------------------------------------------------------------------------------G TX  G Term Term  BTS . 9. Nt HS – SCCH We have: TXi  BTS  RSCP HS – SCCH  ic  Ec ---- . HS-PDSCH Quality Calculation Atoll proceeds as follows: 1st step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ). P HSDPA  ic  = P HS – PDSCH  ic  + n HS – SCCH  P HS – SCCH  ic  Therefore. we have: P HS – PDSCH  ic  = P HSDPA  ic  – n HS – SCCH  P HS – SCCH  ic  n HS – SCCH is the number of HS-SCCH channels and P HS – SCCH  ic  is the HS-SCCH power on carrier ic fixed by the user. The HSReq Ec SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (  -----.  Nt  ic  HS – SCCH = ---------------------------------------------------------DL N Tot  ic  And TXi  BTS  RSCP HS – SCCH  ic  Ec ---- = -------------------------------------------------------------------------------------------------------------------------------------------------------. P HSDPA  ic  is the power available for HSDPA on the carrier ic. Therefore. Atoll proceeds as follows.3. 16QAM modulation is supported by the terminal.for the without useful signal option. .Atoll 3. F JD Therefore.1.2 CQI Based on HS-PDSCH Quality When the option “CQI based on HS-PDSCH quality” is selected. The transport block size (7168 Bytes) does not exceed the maximum transport block size (7298 Bytes) the terminal can carried. Atoll can recalculate the HS-PDSCH quality with the real number of HS-PDSCH channels (A default value of 5 was taken into account in the first HS-PDSCH quality calculation). Atoll knows the number of HS-PDSCH channels. which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic.4. In the Bearer table.  Nt  HS – PDSCH Two options.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Req EcDL    ----  ic   HS – SCCH  N Tot  ic    Nt P HS – SCCH  ic  =  -------------------------------------------------------------------  L T for the total noise option. ic    Nt  HS – PDSCH mobility. may be used to calculate Nt: option Without useful signal or option Total noise. This table is defined for the terminal reception equipment and the specified table CQI HS – PDSCH = f   -----.4. HS-PDSCH CQI Determination Let us assume the following notation: CQIHS – PDSCH corresponds to the HS-PDSCH CQI. Atoll works on the assumption that five HS-PDSCH channels are used (n=5).Atoll 3.for the total noise option. Req Ec    1 +  1 – F DL    1 – F Term    ----   ic  Ortho JD  BTS   Nt  HS – SCCH  2nd step: Then. ic  corresponds to the HS-PDSCH quality. CQI HS – PDSCH is deduced from the Ec  .  Nt  ic  HS – PDSCH DL N Tot  ic  And TXi  BTS  RSCP HS – PDSCH  ic  Ec ----= ----------------------------------------------------------------------------------------------------------------------------------------------------------. available in Global parameters. which is reduced by the interference reduction factor F IRF  ic jc  defined for the pair (ic. F JD Term and N 0 are defined in "Definitions and Formulas" on page 595.  BTS     And Req EcDL  ----    Nt  ic  HS – SCCH  N Tot  ic    P HS – SCCH  ic  =  -------------------------------------------------------------------------------------------------------------------------------------------  L T for the without useful signal option.3. jc). Once the bearer selected. F Ortho .  ic   Nt  HS – PDSCH TXi RSCP HS – PDSCH  ic  DL DL Term N Tot  ic  –  1 – F Ortho    1 – F JD    BTS  --------------------------------------------n Here. 9. Atoll evaluates the HS-PDSCH quality Ec Let us assume the following notation:  -----. 649 . Then. P HS – PDSCH  ic  TXi RSCP HS – PDSCH  ic  = ---------------------------------LT And CI L Path  L TX  L Term  L Body  L Indoor  M Shadowing L T = ---------------------------------------------------------------------------------------------------------------------G TX  G Term Term  BTS . Bearer Selection The bearer is selected as described in "Bearer Selection" on page 646.5. We have: TXi  BTS  RSCP HS – PDSCH  ic  Ec ---- = ------------------------------------------------------------. With DL DL DL DL Term N Tot  ic  = I Intra  ic  + I Extra  ic  + I Inter – Carrier  ic jc  + N 0 DL I Inter – Carrier  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink. it deduces the HS-PDSCH CQI and the bearer to be used. Atoll exactly knows the number of HS-PDSCH channels and recalculates the HSPDSCH quality with the real number of HS-PDSCH channels.2 Coverage Prediction Display Options Three display options are available in the study property dialogue.for the without useful signal option. 4. Atoll reads the corresponding RLC peak throughput.2. Each layer is assigned a colour and displayed with intersections between layers.1 Automatic Carrier Allocation For each transmitter. 9. or the HS-PDSCH CQI when considering the CQI based on HS-PDSCH quality option. area is covered if the peak throughput can be provided. For each layer. 9. d TX i – TX j is the real distance between between TXi and TXj considering any offsets with respect to the site locations. The calculation of distance between TXi and any other transmitter TXj is performed using the equation below: D TX i – TX j Where D = d TX i – TX j TX i – TX j   1 + x   cos    – cos    – 2   is the weighted distance between TXi and TXj. x is set to 15 % so that the maximum variation in D TX i – TX j due to the azimuths does not exceed 60 %. For each layer. DL Coverage consists of several layers with a layer per possible peak throughput ( R Peak ). For any transmitter TXi.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks 9.6 N-Frequency Mode and Carrier Allocation Transmitters that support N-frequency mode are multi carrier transmitters with a master and one or more slave carrier. For each layer. Each layer is assigned a colour and displayed with intersections between area is covered if  -----.11 on page 650.11: Weighted Distance Between Transmitters The above formula implies that two transmitters facing each other will have a shorter weighted distance between them than the real distance. Ec  Threshold . 9. Coverage consists of several layers with a layer per CQI threshold ( CQI Threshold ).2. and are sorted according to their distance from it.6.5. and two transmitters pointing in opposite directions will have a greater weighted distance.  and  are calculated from the azimuths of the two cells as shown in Figure 9.1 ©Forsk 2015 Colour per CQI Atoll displays either the P-CCPCH CQI when the selected option in Global Parameters (HSDPA part) is CQI based on P-CCPCH quality.4.3 Colour per HS-PDSCH Ec/Nt Atoll displays on each bin the HS-PDSCH quality.5. You can assign master and slave carriers to transmitters manually.Atoll 3. Atoll determines a list of "near" transmitters. or use the automatic frequency allocation in Atoll to assign carrier types automatically.2.3. area is covered if CQI  CQI Threshold . Figure 9.5. This is the highest throughput that the bearer can provide on each bin. 9. Each layer is assigned a colour and displayed with intersections between layers. Coverage consists of several layers with a layer per threshold. its "near" transmitters are geographically located close to the transmitter.4. 650 . ic  Nt HS – PDSCH layers.2 Colour per Peak Throughput After selecting the bearer. cell B. If no focus zone exists in the . The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. This folder can be either the Transmitters folder or a group of transmitters or a single transmitter.Atoll 3. If the distance between the reference cell and the candidate neighbour is greater than this value. "Neighbour Allocation for a Group of Transmitters or One Transmitter" on page 655. Atoll calculates the effective distance. For information on the effective distance calculation. the candidate neighbour is discarded. 9.7. In this section. where TXj belongs to the list of "near" transmitters. It means that all the cells of TBC transmitters of your .atl document. The distance between both cells must be less than the user-defined maximum inter-site distance. For transmitters that support the N-frequency mode and have master carriers properly assigned. 2. Atoll assigns different carriers to cells of each co-site transmitter. Allocation of Master Carriers Atoll assigns one master carrier to each transmitter TXi.atl document are potential neighbours. They must fulfill the following conditions: • • • • • They are active Their transmitters support the N-frequency mode. Atoll takes into account the computation zone.3. and the cells are master carriers of their transmitters (neighbours are not allocated to standalone carriers) They satisfy the filter criteria applied to the Transmitters folder They are located inside the focus zone They belong to the folder on which allocation has been executed. Only TBA cells may be assigned neighbours. All the other cells of the transmitter are assigned the carrier-type "slave".1 Neighbour Allocation for All Transmitters We assume that we have a reference cell A and a candidate neighbour. which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. The cells to be allocated will be called TBA cells. such that the master carrier of TXi is different from the master carrier of TXj.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Allocation of All Carriers Co-N-Frequency Allocation Diff-N-Frequency Allocation Atoll assigns the same carriers to cells of each co-site transmitter. The calculation options: 651 . The master carrier is one of the cells defined in the transmitter. see "Appendix: Calculation of the InterTransmitter Distance" on page 656. When automatic allocation starts.7 Neighbour Allocation Atoll permits the automatic allocation of intra-technology neighbours in a TD-SCDMA document. "Importance Calculation" on page 655. 9. Atoll performs the neighbour and scrambling code allocation for the master carrier only. Atoll checks following conditions: 1. the following are explained: • • • "Neighbour Allocation for All Transmitters" on page 651. adjacent cells are sorted and listed from the most adjacent to the least.3. Force symmetry: This option enables you to force the reciprocity of a neighbourhood link. only the distance criterion is taken into account. Force adjacent cells as neighbours: This option enables you to force cells geographically adjacent to the reference cell in the candidate neighbour list. if the reference cell is a candidate neighbour of another cell. This constraint can be weighted among the others and ranks the neighbours through the importance field. Delete existing neighbours: When selecting the Delete existing neighbours option. The overlapping zone ( S A  S B ) is defined as follows: N-frequency handover is a baton handover.e. the existing neighbours are kept.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks • • • ©Forsk 2015 Force co-site cells as neighbours: This option enables you to force cells located on the same site as reference cell in the candidate neighbour list.Atoll 3. You can force Atoll to keep that transmitter in the reference transmitter’s neighbours list by adding the following option in the Atoll. Therefore. If the neighbours list of a transmitter is full. Atoll deletes all the current neighbours and carries out a new neighbour allocation. you may force/forbid a cell to be candidate neighbour of the reference cell. If not selected. • • 652 The P-CCPCH RSCP from the cell A is greater than the P-CCPCH RSCP T_Add. Otherwise. • Adjacency criterion: Geographically adjacent cells are determined on the basis of their best server coverages in TD-SCDMA projects. the later will be considered as candidate neighbour of the reference cell. there must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability. This constraint can be weighted among the others and ranks the neighbours through the importance field.) • When this option is selected. If the Use Coverage Conditions check box is selected. Therefore. Adjacency is relative to the number of pixels satisfying the criterion.ini file: [Neighbours] DoNotDeleteSymmetrics = 1 • • Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. The P-CCPCH RSCP from the cell A is greater than the P-CCPCH RSCP from all other cells. . Assuming that the reference cell A and the candidate cell B are located inside a continuous layer of cells: SA is the area where the cell A is the best serving cell. CellA is considered adjacent to CellB if there exists at least one pixel in the CellB best server coverage area (and P-CCPCH RSCP of CellB > P-CCPCH RSCP T_Add) where CellA is best server (of several cells have the same best server value) or CellA is the second best server that enters the handover set (i. Let CellA be a candidate neighbour cell of CellB. depending on the above criterion.. P-CCPCH RSCP of CellA > P-CCPCH RSCP T_Drop and P-CCPCH RSCP of CellA > P-CCPCH RSCP of CellB T_Comp. 3. the reference transmitter will not be added as a neighbour of that transmitter and that transmitter will be removed from the reference transmitter’s neighbours list. Figure 9.12: N-frequency Neighbour Allocation SA  SB . The importance of neighbours. • • The P-CCPCH RSCP from the cell B is greater than the P-CCPCH RSCP T_Drop.3. 653 . which it compares with the % minimum covered Atoll calculates the percentage of covered area ( ----------------SA area. the candidate neighbour B is discarded.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 SB is the area where the cell B can enter the handover set. : Overlapping Coverages 4. The P-CCPCH RSCP from the cell B is greater than the P-CCPCH RSCP from the cell A minus the P-CCPCH RSCP T_Comp. The coverage condition can be weighted among the others and ranks the neighbours through the importance field. If this percentage is not exceeded. 100 ). For information on the importance calculation. see "Importance Calculation" on page 655.Atoll 3. there can be cases where the calculated importance is different when the global Max inter-site distance is modified. You can force Atoll to compare the Max inter-site distance with the real inter-site distance by adding the following lines in Atoll. In addition. it indicates the importance (in %) of each neighbour and the allocation reason. or symmetric.ini: [Neighbours] RealInterSiteDistanceCondition=1 • By default. If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8. there can be cases where the real distance between assigned neighbours is higher than the Max inter-site distance. a neighbour may be marked as exceptional pair. Note that maximum numbers of neighbours can be defined at the cell level (properties dialogue or Cells table). neighbours are marked as existing.Atoll 3. the neighbour importance calculated with respect to distance is based on the global Max inter-site distance setting for all neighbour candidates.ini: [Neighbours] CandidatesMaxDistanceInImportanceCalculation=1 654 .3. Among these 15 candidate neighbours. the number of neighbours. only 8 (having the highest importance values) will be allocated to the reference cell. this value is taken into account instead of the default one available in the dialogue.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason. adjacency. the automatic neighbour allocation compares the defined Max intersite distance with the effective inter-transmitter distance. • By default.. and coverage reasons.e. Finally. the percentage of area that satisfies the adjacency conditions and the corresponding surface area (km2). and the maximum number of neighbours allowed for each cell. Atoll provides the list of neighbours. If defined there. coverage. For neighbours accepted for co-site. if cells have previous allocations in the list. because the effective distance is smaller. In the Results part. i. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be allocated to each transmitter is exceeded. adjacent. Atoll displays the percentage of area that satisfies the coverage conditions and the corresponding surface area (km2). As a consequence. As a consequence. you can force Atoll to prioritise the individual distances between reference transmitters and their respective neighbour candidates by adding the following lines in Atoll. To avoid that. co-site. Atoll considers the constraints between exceptional pairs in both directions so as to respect the symmetric relation. mobility. Symmetric neighbour relations are only added to the neighbour lists if the neighbour lists are not already full. table below). Although no specific terminal. 9. if a neighbour relation is forced in one direction and forbidden in the other. There is space in the cell B neighbour list: cell A will be added to the list. or symmetric Neighbours of TBA cells that satisfy coverage conditions Automatic neighbour allocation parameters are described in "Neighbour Allocation for All Transmitters" on page 651. it will not appear in the Results table. For automatic neighbour allocation.2 Neighbour Allocation for a Group of Transmitters or One Transmitter In this case. In this case. Atoll automatically calculates the missing path loss matrices. this value varies between 0 and 100%. or service is selected for automatic neighbour allocation. The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf. adjacent. Atoll allocates neighbours to: • • • TBA cells Neighbours of TBA cells marked as exceptional pair. • Mobility does not impact the allocation A forbidden neighbour must not be listed as neighbour except if the neighbourhood relationship already exists and the Delete existing neighbours option is not selected. On the other hand.3.Atoll 3. Thus. but cell A is not a neighbour of the cell B. Atoll displays only the cells for which it finds new neighbours. Atoll displays a warning message in the Event Viewer. The cell B neighbour list is full: Atoll will not include cell A in the list and will remove the symmetric relation by deleting cell B from the cell A neighbour list.3 Importance Calculation Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason and the distance. and to quantify the neighbour importance.7. In this case. 2. In the results. Therefore. If this is the same for all terminals. • • If you select Force exceptional pairs and Force symmetry options. the algorithm tries to find the maximum number of neighbours by selecting: • • • • The service with the lowest body loss The terminal with the highest difference between Gain and Losses. Atoll displays a warning message in the Event Viewer indicating that the constraint on the forbidden neighbour will be ignored by the algorithm because the neighbour already exists. if a TBA cell has already reached its maximum number of neighbours before starting the new allocation. Neighbourhood cause When Importance value Existing neighbour Only if the Delete existing neighbours option is not selected and in case of a new allocation Existing importance Exceptional pair Only if the Force exceptional pairs option is selected 100 % Co-site cell Only if the Force co-site cells as neighbours option is selected Importance Function (IF) Adjacent cell Only if the Force adjacent cells as neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions Only if the % minimum covered area is exceeded Importance Function (IF) Symmetric neighbourhood relationship Only if the Force neighbour symmetry option is selected Importance Function (IF) 655 .7. there can be two possibilities: 1.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 • • You do not require simulations or coverage predictions for an automatic neighbour allocation. symmetry cannot be respected. Atoll uses the terminal with the lowest noise figure. It will be the last one. 9. if the cell B is a neighbour of the cell A. the neighbours may be ranked differently. For information on the effective distance calculation. neighbours will be ranked in this order: co-site neighbours. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 30% Adjacency factor (A) Min(A) 30% Max(A) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The Importance Function is evaluated as follows: Neighbourhood cause Importance Function Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%Max(Di))(O)}+Min(Di)+Delta(Di)(Di) 10%+20%{10%(Di)+90%(O)}+1%+9%(Di) No Yes Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 30%+30%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Yes Yes Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 60%+40%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Co-site Adjacent No Where: Delta(X)=Max(X)-Min(X) • • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation.7. If the Min and Max value ranges of the importance function factors do not overlap.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Except the case of forced neighbours (importance = 100%). 9. The minimum and maximum importance assigned to each of the above factors can be defined. d = D   1 + x  cos  – x  cos   where x = 0. neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping. If the Min and Max value ranges of the importance function factors overlap.4 Appendix: Calculation of the Inter-Transmitter Distance Atoll takes into account the real distance ( D in m) and azimuths of antennas in order to calculate the effective intertransmitter distance ( d in m). With the default values for minimum and maximum importance fields. see "Appendix: Calculation of the Inter-Transmitter Distance" on page 656. the neighbours will be ranked by neighbour cause.3% so that the maximum D variation does not exceed 1%. d Di  = 1 – ---------d max d is the effective distance (in m). With a value of Min(O) = 0%.3. There can be a mix of the neighbourhood causes. The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. and neighbours allocated based on coverage overlapping.Atoll 3. priority assigned to each neighbourhood cause is determined using the Importance Function (IF). The overlapping factor (O): the percentage of overlapping. It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. d max is the maximum distance between the reference transmitter and a possible neighbour. adjacent neighbours. 656 . • • • The co-site factor (C): a Boolean. The adjacency factor (A): the percentage of adjacency. The IF considers the following factors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. • • Atoll can take into account inter-technology neighbour relations as constraints to allocate different scrambling codes to the TD-SCDMA neighbours of a GSM transmitter.Atoll 3.atl document accessible in the TD-SCDMA. Atoll takes into account either all the cells of TBC transmitters. Atoll calculates a scrambling code and a SYNC_DL code to all these cells.1 Automatic Allocation Description 9. Atoll takes into account the computation zone. In order to consider inter-technology neighbour relations in the scrambling code allocation. its second order neighbours and its third order neighbours. TBA cells are the cells that fulfill the following conditions: • • • • They are active They satisfy the filter criteria applied to the Transmitters folder They are located inside the focus zone They belong to the folder on which allocation has been executed. This folder can be either the Transmitters folder or a group of transmitters or a single transmitter. Depending on the options you select for automatic allocation of scrambling and SYNC_DL codes. Scrambling codes are numbered from 0 to 127.8 Scrambling Code Allocation Downlink scrambling codes enable mobile to distinguish one cell from another.atl documents. Furthermore. and SYNC_DL codes from 0 to 31. the scrambling code allocation also considers the master and slave carrier allocations.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 Figure 9.atl document. or only cells of active and filtered transmitters located inside the computation zone. 2. A different DL synchronisation code. or SYNC_DL code. The scrambling code reuse distance 657 .3. In TD-SCDMA.atl document. there are 128 scrambling codes (or P-CCPCH midamble codes) distributed in 32 clusters of 4 codes each.8. Neighbour relations between cells You may consider: • • • First order neighbours: The neighbours of TBA cells listed in the Intra-technology neighbours table. Atoll considers symmetry relationship between a cell. 9. you must make the Transmitters folder of the GSM. 9.1. it allocates scrambling codes and SYNC_DL codes only to TBA cells (cells to be allocated). its first order neighbours.8. It is this effective distance that will be taken into account rather than the real distance. If no focus zone exists in the . see the User Manual. For information on making links between GSM and TD-SCDMA . is assigned to each cluster.13: Inter-Transmitter Distance Computation The formula above implies that two cells facing each other will have a smaller effective distance than the real physical distance. But. if there are transmitters that support the N-frequency mode among the TBC transmitters of your network.1 Allocation Constraints and Options The scrambling code and SYNC_DL code allocation algorithm can take into account following constraints: 1. Second order neighbours: The neighbours of neighbours. Third order neighbours: The neighbour’s neighbour’s neighbours. e. 5. cells of a transmitter (i. • Distributed per site: This strategy allocates a group of adjacent clusters.. Atoll allocates the same scrambling code to each carrier of a transmitter. and there are still sites remaining to be allocated. Atoll reuses the adjacent clusters as far as possible at another site. consecutive SYNC_DL codes.ini file: [PSC] ConstantStep=1 For more information about setting options in the atoll. i.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Reuse Distance: It is a constraint on the allocation of scrambling codes. 3GPP specifications define 32 SYNC_DL codes with 4 corresponding scrambling codes each (SYNC_DL codes are numbered from 0 to 31). . Scrambling code reuse distance can be defined for each cell in the cell properties. depends on the number of transmitters per site. When the allocation is based on a Distributed strategy (Distributed per Cell or Distributed per Site). However. i.2 Allocation Strategies You can choose from the following four allocation strategies: • • • Clustered: The purpose of this strategy is to choose for a group of mutually constrained cells. If this value is not defined. different carriers) are assigned the same scrambling code. then one scrambling code from the cluster to each cell of the site. They are referred to as near cells. One SYNC_DL code per site: This strategy allocates one cluster. it is possible to define a different value (e. If you select "All". Existing allocation Delete All Codes: If you select this option. Same carriers must be assigned different scrambling codes. Atoll only displays scrambling codes and SYNC_DL codes allocated to TBA cells. Atoll keeps the existing allocation.8.e. Atoll will preferentially allocate codes from different clusters. this parameter can also be used to define the interval between the scrambling codes assigned to cells on a same site.1. The same scarmbling code or SYNC_DL code cannot be allocated to two sites that are not farther apart than the reuse distance. or consecutive SYNC_DL codes. In this case. Atoll lists all cells which have constraints with the cell. Atoll uses the default reuse distance defined in the Automatic Scrambling Code and SYNC_DL code Allocation dialogue.3 Allocation Process For each TBA cell. and finally one scrambling code from each cluster to each cell of each transmitter. When all the sites have been allocated adjacent clusters.g. The near cells of a TBA cell may be: • • 658 Its neighbour cells: the neighbours listed in the Intra-technology neighbours table (options “Existing neighbours” and "First Order"). Otherwise.3. The reuse distance constraint is used for clustered and distributed per cell allocation strategies. Atoll can use a maximum of codes Use a Maximum of Codes: If you choose to use a maximum of codes. scrambling codes among a minimum number of clusters. scrambling codes will be distributed among 64 SYNC_DL codes). Distributed per Cell: This strategy consists in using as many clusters as possible. The defined interval is applied by adding the following lines in the Atoll. Different carriers of the same site can be assigned the same scrambling code. The number of adjacent clusters. Atoll reuses the clusters as far as possible at another site.. The neighbours of its neighbours (options “Existing neighbours” and “Second Order”). then one cluster.ini file. or SYNC_DL code. In the Results table. per site. The carrier for which you want to perform the automatic allocation Carrier: You can select "All" or a specific carrier. 6. 9. 4. to each site. one SYNC_DL code. if the scrambling code domains associated with the carriers have a common cluster or enough codes in one cluster. see the Administrator Manual. 3. if you set the number of scrambling codes per SYNC_DL codes to 2. to each transmitter on the site according to its azimuth. Atoll will preferentially allocate all the codes within the same cluster. 9.e. Therefore.8. The number of scrambling codes per SYNC_DL code Each SYNC_DL code corresponds to a group of scrambling codes as defined in 3GPP specifications. When all the clusters have been allocated but there are still sites remaining. Atoll will delete any existing scrambling code allocation and perform a fresh allocation.1. Atoll will try to spread the allocated spectrum of scrambling codes as much as possible. For information on calculating cell priority.. Algorithm works as follows: Strategies: Clustered and Distributed per Cell Atoll processes TBA cells according to their priority. During the allocation. Then. For information on calculating site priority. Then. Strategy: One SYNC_DL Code per Site All sites which have constraints with the studied site are referred to as near sites.3.atl document). Atoll assigns a cluster. Atoll assigns each cluster of the group to each transmitter of the site according to the transmitter azimuth and selected neighbourhood constraints (options "Neighbours in Other Clusters" and "Secondary Neighbours in Other Clusters"). Atoll reuses the clusters at the other sites. When all the clusters have been allocated but there are still sites remaining. the algorithm tries to assign reused clusters as spaced out as possible. Atoll: • • • • Defines theoretical groups of adjacent clusters. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet and their near cells. to each site. i. Additional constraints are considered when: • • The cell and its near cells are neighbours of a same GSM transmitter (only if the Transmitters folder of the GSM. to each site.1. Otherwise. It allocates scrambling codes starting with the highest priority cell and its near cells. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet and their near cells. Then.e. see "Cell Priority" on page 660. For information on calculating site priority. see "Cell Priority" on page 660.Atoll 3. Atoll allocates a scrambling code from the cluster to each cell located on the sites (codes belong to the assigned clusters). when the option is not selected. For information on the cost generated by each constraint. and continuing with the lowest priority cells not allocated yet and their near cells. starting with the highest priority site and its near sites.e.atl document is accessible in the TD-SCDMA. Atoll reuses the adjacent clusters at other sites. The cells with distance from the TBA cell less than the reuse distance. the algorithm tries to assign reused clusters as spaced out as possible. see "Cell Priority" on page 660. These constraints have a certain weight taken into account to determine the TBA cell priority during the allocation process and the cost of the scrambling code plan. Atoll allocates a scrambling code to each cell located on the transmitters (codes belong to the assigned clusters). independent of the defined domain. 9. Atoll assigns a group of adjacent clusters.1 Single Carrier Network The allocation process depends on the selected strategy. and there are still sites remaining to be allocated. considering the 128 scrambling codes available and 4 codes per cluster. If it respects all the constraints. and continuing with the lowest priority sites not allocated yet and their near sites. a SYNC_DL code. When a cell has too many constraints and there are not anymore scrambling codes available. see "Cell Priority" on page 660. and continuing with the lowest priority sites not allocated yet and their near sites.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 • • • The third order neighbours (options “Existing neighbours” and “Third Order”).8. SYNC_DL codes. see "Site Priority" on page 663. Starts the distribution of clusters to groups from the cluster 0 Takes into account the maximum number of transmitters per site in order to determine the number of clusters in each group Determines the total number of groups 659 . When the Reuse Distance option is selected. or information on calculating cell priority. starting with the highest priority site and its near sites. Atoll breaks the constraint with the lowest cost so as to generate the scrambling code plan with the lowest cost. the algorithm reuses the clusters as soon as the reuse distance is exceeded.3. Otherwise. For information on calculating cell priority. The cells that make exceptional pairs with the TBA cell. i. the cost of the scrambling code plan is 0.. The neighbour cells cannot share the same cluster (for the "Distributed per site" allocation strategy only). see "Site Priority" on page 663. When all the sites have been allocated adjacent clusters. When the Reuse Distance option is selected. the algorithm reuses the clusters as soon as the reuse distance is exceeded. Strategy: Distributed per Site All sites which have constraints with the studied site are referred to as near sites. Atoll tries to assign different scrambling codes to the TBA cell and its near cells. Determination of Groups of Adjacent Clusters In order to determine the groups of adjacent clusters to be used. when the option is not selected. Group 11 Cluster 30 . depends on the number of scrambling codes available for the allocation. transmitters or sites have the same priority.. It starts with the highest priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters.. The higher the cost on a cell. Atoll can use all these groups for the allocation.3. In case of the "Distributed per site" strategy. equal to 0.8. In case of a "Per cell" strategy (Clustered and Distributed per cell). to each site. There are seven criteria employed to determine the cell priority.8. On the other hand. see "Transmitter Priority" on page 662. Cluster 31 If no domain is assigned to cells. 4 and 5 Group 3 with cluster 6. C . Atoll compares adjacent clusters actually available in the assigned domain with the theoretical groups and only keeps adjacent clusters common with the theoretical groups. 9. 7 and 8 Group 6 with cluster 12. The same scrambling code is assigned to each cell of the transmitter. the theoretical groups of adjacent clusters will be: Group 1 Group 2 Group 3 Group 4 Cluster 0 Cluster 3 Cluster 6 Cluster 9 Cluster 1 Cluster 4 Cluster 7 Cluster 10 Cluster 2 Cluster 5 Cluster 8 Cluster 11 .. i. The same scrambling code is assigned to each cell of the transmitter. If we have a domain comprising 12 clusters: clusters 1 to 8 and clusters 12 to 15. For information on calculating transmitter priority. In this case. C i  Dom  . then a cluster to each transmitter and finally. 9. the allocation order changes. A cell without any constraint has a default cost.1. 660 . i. It is no longer based on the cell priority but depends on the transmitter priority. allocates a scrambling code to each transmitter. Atoll assigns a cluster. The total cost due to constraints on any cell i is defined as: C i = C i  Dom  + C i  U  With C i  U  = C i  Dist  + C i  EP  + C i  N  + C i  N 2G  + C i  Cluster  + C i  CN  All the cost components are described below: • Scrambling Code Domain Criterion The cost due to the domain constraint. allocates a scrambling code to each transmitter. a SYNC_DL code.Atoll 3.3. if a domain is used.1.1 Cell Priority Scrambling code allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities assigned to cells depend upon how much constrained each cell is and the cost defined for each constraint.2 Multi-Carrier Network In case you have a multi-carrier network and you run the scrambling code allocation on all the carriers. Atoll displays a warning message in the Event Viewer. 2 and 15 will not be used..1. It starts with the highest priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. The domain constraint is mandatory and cannot be broken. to each site and then.e. All transmitters which have constraints with the studied transmitter will be referred to as near transmitters.. processing is based on an alphanumeric order. In case of the "One SYNC_DL code per site" strategy.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 If the number of scrambling codes per cluster is set to 4 and the maximum number of transmitters per site in the network is 3.. 13 and 14 The clusters 1. Atoll assigns a group of adjacent clusters.4 Priority Determination 9.8.4. SYNC_DL codes. If a domain does not contain any adjacent clusters. Atoll starts scrambling code allocation with the highest priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. the higher the priority it has for the scrambling code allocation process. The same scrambling code is assigned to each cell of the transmitter. When cells. Atoll will use the following groups of adjacent clusters: • • • Group 2 with cluster 3.e. • Neighbourhood Criterion The constraint level of any cell i depends on the number of its neighbour cells j. 128 scrambling codes are available and we have: C i  Dom  = 0 When domains of scrambling codes are assigned to cells. • Exceptional Pair Criterion The constraint level of any cell i depends on the number of exceptional pairs (j) for that cell. the weight for an inter-cell distance of 1500m is 0. The total cost due to exceptional pair constraint is given as: C i  EP  =  cEP  i – j  j Where c EP is the cost of the exceptional pair constraint. the number of second order neighbours k and the number of third order neighbours l.Atoll 3. the less will be the cost due to this criterion. the weight for co-site cells is 1 and the weight for two cells spaced out 2100m apart is 0.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 When no domain is assigned to cells. The total cost due to the distance constraint is given as:  Cj  Dist  i   C i  Dist  = j Each cell j within the reuse distance generates a cost given as: C j  Dist  i   = w  d ij   c dis tan ce Where w  d ij  is a weight depending on the distance between i and j. The higher the number of codes available in the domain. c dis tan ce is the cost of the distance constraint. This value can be defined in the Constraint Cost dialogue. Let’s consider the following neighbour schema: Figure 9. This value can be defined in the Constraint Cost dialogue.25. This weight is inversely proportional to the inter-cell distance. The cost is given as: C i  Dom  = 128 – Number of scrambling codes in the domain • Distance Criterion The constraint level of any cell i depends on the number of cells (j) present within a radius of "reuse distance" from its centre.14: Neighbourhood Constraints The total cost due to the neighbour constraint is given as:  Ci  N  =         Cj  N1  i   +  Cj – j  N1  i   +   Ck  N2  i   +  Ck – k  N2  i   +   Cl  N3  i   +  Cl – l  N3  i   j j k k l l Each first order neighbour cell j generates a cost given as: C j  N1  i   = I j  c N1 Where I j is the importance of the neighbour cell j. c N1 is the cost of the first order neighbour constraint. 661 . This value can be defined in the Constraint Cost dialogue. For a reuse distance of 2000m. each unavailable scrambling code generates a cost.3. Atoll considers the cost created by two second order neighbours to be each other. C l  N3  i   + C l  N3  i   C l – l  N3  i   = ----------------------------------------------------2 Atoll considers the highest cost of both links when a neighbour relation is symmetric and the importance value is different. The total cost due to GSM neighbour constraint is given as: 662 . This value can be defined in the Constraint Cost dialogue. If the cell i is neighbour of a GSM transmitter.3. we have: C j  N1  i   = Max  I i – j I j – i   c N1 And C k  N2  i   = Max (C j  N1  i    C k  N1  j  . Because two third order neighbours must not have the same scrambling code. C j  N1  i    C k  N1  j   )  c N2 Where c N2 is the cost of the second order neighbour constraint. the cell constraint level depends on how many cells j are neighbours of the same GSM transmitter. This value can be defined in the Constraint Cost dialogue. In this case.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 Because two first order neighbours must not have the same scrambling code. C i  CN  =  j di – j   I +  1 – ----------   i–j   Max  d CN    ----------------------------------------.Atoll 3. Because two second order neighbours must not have the same scrambling code. Atoll considers the cost created by two first order neighbours to be each other. the Transmitters folder of the GSM . c CN 2       Where c CN is the cost of the close neighbour constraint.C j  N1  k    C i  N1  j  )  c N2 • Close Neighbour Criterion The constraint level of any cell i depends on the number of its close neighbour cells j. Atoll considers the cost created by two third order neighbours to be each other. The close neighbour cost ( C i  CN  ) depends on two components: the importance of the neighbour relation ( I i – j ) and the distance ( d i – j ) relative to maximum Max close neighbour distance ( d CN ). • GSM Neighbour Criterion This criterion is considered when the co-planning mode is activated (i. C j  N1  i   + C j  N1  i   C j – j  N1  i   = ----------------------------------------------------2 Each second order neighbour cell k generates a cost given as: C k  N2  i   = Max ( C j  N1  i    C k  N1  j   .atl document) and inter-technology neighbours have been allocated.e.atl document is made accessible in the TD-SCDMA. C k  N2  i   + C k  N2  i   C k – k  N2  i   = ------------------------------------------------------2 Each third order neighbour cell l generates a cost given as:  C  N1  i    C k  N1  j    C l  N1  k   C j  N1  i    C k  N1  j    C l N1  k   C l  N3  i   = Max  j   c N3   C j  N1  i    C k  N1  j     C l N1  k  C j  N1  i    C k  N1  j    C l N1  k   Where c N3 is the cost of the third order neighbour constraint. This value can be defined in the Constraint Cost dialogue. 0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 C i  N 2G  =  cN2G  j – Tx2G  j Where cN 2G is the cost of the GSM neighbour constraint. The higher the cost on a site. Atoll allots priorities to sites. This value can be defined in the Constraint Cost dialogue.8. the domain available for the transmitter is the intersection of domains assigned to cells of the transmitter. In this case. you can consider additional constraints on allocated clusters (one cell.1. the constraint level of any cell i depends on the number of first and second order neighbours.3. each of them has two cells using carriers 0 and 1. The higher the cost on a transmitter. Atoll allots priorities to transmitters. The cost due to constraints on the transmitter is given as: C Tx = C Tx  Dom  + C Tx  U  With C Tx  U  = Max  C  U   and C  Dom  = 128 – Number of scrambling codes in the domain Tx i  Tx i Here. 9.3 Site Priority In case of "Per Site" allocation strategies (One SYNC_DL code per Site and Distributed per Site). 663 . the higher the priority it has for the scrambling code allocation process. the domain considered for the site is the intersection of domains available for transmitters of the site.4. Let us consider a transmitter Tx with two cells using carriers 0 and 1. its first order neighbours and its second order neighbours must be assigned scrambling codes from different clusters). The cost due to constraints on the site is given as: C S = C S  U  + C S  Dom  With C S  U  = Max  C  U   and C  Dom  = 128 – Number of scrambling codes in the domain S Tx  S Tx Here. The domain constraint is mandatory and cannot be broken.8.2 Transmitter Priority In case you have a multi-carrier network and you run scrambling code allocation on "all" the carriers.4. 9. The total cost due to the cluster constraint is given as: C i  Cluster  =  Cj  N1  i    cCluster +  Ck  N2  i    cCluster j k Where c Cluster is the cost of the cluster constraint.1. • Cluster Criterion When the "Distributed per Site" allocation strategy is used.Atoll 3. Priorities assigned to transmitters depend on how much constrained each transmitter is and the cost defined for each constraint. This value can be defined in the Constraint Cost dialogue. j and k. Let us consider a site S with three transmitters. the higher the priority it has for the scrambling code allocation process. The domain constraint is mandatory and cannot be broken. Priorities assigned to sites depend on how much constrained each site is and the cost defined for each constraint. This implies that the domain of scrambling codes for the four sites is from 0 to 23 (cluster 0 to cluster 5).1 Single Carrier Network In order to understand the differences between the different allocation strategies and the behaviour of algorithm when using a maximum of codes or not.2. with 3 transmitters each using carrier 0. The following section shows the results of each combination of options with explanations where necessary. and Site3 be four sites. Only co-site neighbours exist. in our case. and there are non allocated codes each site. Atoll first allocates those codes before reusing the already used ones.3. Without "Use a Maximum of Code" With "Use a Maximum of Code" As it is possible to use a maximum of codes. 9. The reuse distance is supposed to be less than the inter-site distance.8.2 Scrambling Code Allocation Example 9. Atoll starts allocation at the start of a different cluster at each site.8. to whom scrambling codes have to be allocated out of 6 clusters of 4 scrambling codes.1 Strategy: Clustered Since the restrictions of neighbourhood only apply to co-sites and.8.2. Site1. the distances between sites are greater than the reuse distance. left in the cluster. Site2. every cell has the same priority. Allocation is performed in an alphanumeric order. 664 .1.15: Scrambling Code Allocation Example Let Site0. Atoll starts allocating the codes from the start of cluster 0 at When a cluster is reused. let us consider the following sample scenario: Figure 9.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 9. therefore.3 Without "Use a Maximum of Code" With "Use a Maximum of Code" Atoll allocates codes from different clusters to each cell of the same site. the distances between sites are greater than the reuse distance. 665 . Strategy: One SYNC_DL Code per Site Since the restrictions of neighbourhood only apply to co-sites. Atoll can allocate different codes from a reused cluster at another site.8. As it is possible to use a maximum of codes.2 Strategy: Distributed per Cell Since the restrictions of neighbourhood only apply to co-sites and.1. same codes can be allocated to each site’s cells. every cell has the same priority.1.2. Under given constraints of neighbourhood and reuse distance. Without "Use a Maximum of Code" With "Use a Maximum of Code" In this strategy. every site has the same priority. Atoll allocates the codes so that there is least repetition of codes. In this case.8. When it is possible to use a maximum of codes.Atoll 3. Atoll reuses the cluster as far as possible at another site.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 9.2. Atoll allocates codes from different clusters to each site’s cells. Cluster allocation to sites is performed in an alphanumeric order. in our case.3. 9. Allocation is performed in an alphanumeric order. a cluster of codes is limited to be used at just one site at a time unless all codes and clusters have been allocated and there are still sites remaining to be allocated. 2 Multi Carrier Network If you have a multi carrier network. Let Site0. the same code is given to each cell of the transmitter. Cluster allocation to sites is performed in an alphanumeric order. Scrambling codes have to be allocated out of 6 clusters consisted of 4 scrambling codes.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks 9.8. Every site has the same priority and the cluster allocation to sites is performed in an alphanumeric order. therefore. i. Atoll reuses the group as far as possible at another site..16: Scrambling Code Allocation to All Carriers 666 . Atoll allocates the same scrambling code to each carrier of a transmitter.1. Then. and Site3 be four sites with 3 cells using carrier 0 and 3 cells using carrier 1. 9. Atoll allocates one cluster at each site and then.Atoll 3. 1.2. Only co-site neighbours exist. Site1.e.8. one code to each transmitter.2. transmitters with more than one cells using different carriers. every site has the same priority. Atoll can allocate different codes from a reused groups of adjacent clusters at another site. This implies that the domain of scrambling codes for the four sites is from 0 to 23 (cluster 0 to cluster 5).4 ©Forsk 2015 Strategy: Distributed per Site Since the restrictions of neighbourhood only apply to co-sites. and 2 is available). When it is possible to use a maximum of codes. Without "Use a Maximum of Code" With "Use a Maximum of Code" A group of adjacent clusters is allocated to one site at a time. In this case (here only one group of adjacent clusters 0. and you run scrambling code allocation on "all" the carriers. The reuse distance is supposed to be less than the inter-site distance.3. unless all the codes and groups of adjacent clusters have been allocated but there are still sites remaining to be allocated. Figure 9. Site2. An existing link on the Transmitters folder of GSM. The importance of neighbours.9. it is called inter-technology neighbour allocation.1 Automatic Allocation Description The allocation algorithm takes into account criteria listed below: • • • • The inter-transmitter distance The maximum number of neighbours Allocation options The selected allocation strategy Two allocation strategies are available: the first one is based on distance and the second one on coverage overlapping.atl are potential neighbours. In order to be able to use the inter-technology neighbour allocation algorithm. Atoll will allocate neighbours to cells using the selected carriers. Atoll calculates the importance of the automatically allocated neighbours. 667 .1 Algorithm Based on Distance When automatic allocation starts. In Atoll.atl document containing the GSM network.atl. and another one containing the TD-SCDMA network. you must have: • • An . You may choose one or more carriers.3. This option is automatically selected. The external neighbour allocation algorithm takes into account all the GSM TBC transmitters. A. The TD-SCDMA cells. Atoll sorts the neighbours by decreasing importance in order to keep the ones with high importance. Delete existing neighbours: When selecting the Delete existing neighbours option. Atoll deletes all the current neighbours and carries out a new neighbour allocation. For information on the effective distance calculation. TDSCDMA. existing neighbours are kept. Only TD-SCDMA TBA cells can be assigned neighbours. If the maximum number of neighbours to be allocated to each cell is exceeded. and a GSM candidate neighbour transmitter. see "Appendix: Calculation of the InterTransmitter Distance" on page 656. you may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.Atoll 3. Atoll checks following conditions: 1. Atoll calculates the effective distance. The calculation options: Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. to be allocated neighbours are called TBA cells which fulfill following conditions: • • • • They are active They satisfy the filter criteria applied to Transmitters folder They are located inside the focus zone They belong to the folder for which allocation has been executed. If the distance between the TD-SCDMA reference cell and the GSM neighbour is greater than this value. then the candidate neighbour is discarded.atl.atl. Atoll keeps the ones with high importance. Therefore. In this case. Atoll’s automatic inter-technology neighbour allocation algorithm takes into account both cases. Inter-technology handover is used in two cases: • • When the TD-SCDMA coverage is not continuous. in TD-SCDMA. B.9 Automatic GSM/TD-SCDMA Neighbour Allocation It is possible to automatically calculate and allocate neighbours between GSM and TD-SCDMA networks.atl into TD-SCDMA.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 9.1. 9. This folder can be either the Transmitters folder or one of its subfolders.9. If not selected.atl. 9. Next. It means that all the TBC transmitters of GSM. GSM. We assume we have a TD-SCDMA reference cell. In order to balance traffic and service distribution between both networks. the TD-SCDMA coverage is extended by TD-SCDMA to GSM handovers. The distance between the TD-SCDMA reference cell and the GSM neighbour must be less than the user-defined maximum inter-site distance. Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. 3. Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site than the reference TD-SCDMA cell in the candidate neighbour list. 2. In addition. If the distance between the TD-SCDMA reference cell and the GSM neighbour is greater than this value. 2. Finally. 9. Atoll deletes all the current neighbours and carries out a new neighbour allocation. Therefore.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks ©Forsk 2015 As indicated in the table below. Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Atoll provides the list of neighbours. the margin must be set to 0 dB. In this case. • The pilot signal received from A is greater than the minimum pilot signal level and is the highest one. Delete existing neighbours: When selecting the Delete existing neighbours option.2 Algorithm Based on Coverage Overlapping When automatic allocation starts. The distance between the TD-SCDMA reference cell and the GSM neighbour must be less than the user-defined maximum inter-site distance. co-site. it indicates the importance (in %) of each neighbour and the allocation reason. • 2nd case: The margin is different from 0 dB and SA is the area where: • The pilot signal level received from A exceeds the user-defined minimum pilot signal level and is within a margin from the highest signal level. existing neighbours are kept. then the candidate neighbour is discarded. For neighbours accepted for distance reasons. If not selected. a neighbour may be marked as exceptional pair. • • 668 The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is the highest one. if cells have previous allocations in the list.Atoll 3. The calculation options: Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. you may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell. 3. You may choose one or more carriers. Two different cases may be considered for SB: • 1st case: SB is the area where the cell B is the best serving transmitter of the GSM network. Neighbourhood cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter If the Force co-site cells as neighbours option is selected 100 % Neighbourhood relationship that fulfils distance conditions If the maximum distance is not exceeded d1 – ---------d max Where d is the effective distance between the TD-SCDMA reference cell and the GSM neighbour and d max is the maximum inter-site distance.3. • The margin is set to 0 dB. Atoll will allocate neighbours to cells using the selected carriers. Two different cases may be considered for SA: • 1st case: SA is the area where the cell A is the best serving cell of the TD-SCDMA network. For information on the effective distance calculation. .9. Therefore. Atoll checks following conditions: 1. this value varies between 0 to 100%. which corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. In the Results part. or distance. neighbours are marked as existing. the neighbour importance depends on the distance and on the neighbourhood cause.1. see "Appendix: Calculation of the InterTransmitter Distance" on page 656. Atoll calculates the effective distance. This option is automatically selected. the number of neighbours and the maximum number of neighbours allowed for each cell. Atoll displays the distance from the reference cell (m). There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability. 2nd case: The margin is different from 0 dB and SB is the area where: • The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is within a margin from the best BCCH signal level. Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site than the reference TD-SCDMA cell in the candidate neighbour list. 3. 100 ) and compares this value to the % minimum covered Atoll calculates the percentage of covered area ( ----------------SA area. d Di  = 1 – ---------d max d is the effective distance (in m). For information on the effective distance calculation. If this percentage is less than the minimum.Atoll 3. • • The co-site factor (C): a Boolean. 4. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The IF evaluates importance as follows: Co-site Neighbourhood cause IF Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)} Yes Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+ Max(O)(O)/(Max(Di)+Max(O))} 60%+40%{1/7%(Di)+6/7%(O)} Where Delta(X)=Max(X)-Min(X) 669 . d max is the maximum distance between the reference transmitter and a possible neighbour. see "Appendix: Calculation of the Inter-Transmitter Distance" on page 656. Next. Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of covered area. The IF considers the following factors for calculating the importance: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. Atoll keeps the ones with high importance. this value varies between 0 to 100%. priority assigned to each neighbourhood cause is determined using the Importance Function (IF). Neighbourhood reason When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter If the Force co-site cells as neighbours option is selected IF Neighbourhood relationship that fulfils coverage conditions If the % minimum covered area is exceeded IF Except the case of forced neighbours (importance = 100%). If the maximum number of neighbours to be allocated to each cell is exceeded. The importance of neighbours. It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. Atoll calculates the importance of the automatically allocated neighbours. The overlapping factor (O): the percentage of overlapping. the candidate neighbour B is discarded. Atoll sorts the neighbours by decreasing importance in order to keep the ones with high importance. the neighbour importance depends on the distance and on the neighbourhood cause. As indicated in the table below.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks AT330_TRR_E1 SA  SB . The IF is user-definable using the Min importance and Max importance fields. 1 Delete Existing Neighbours Option As explained above. the neighbours may be ranked differently. it will not appear in the Results table. 670 . It will be the first one in the neighbour list. in order to allocate neighbours to the new cell i only. Atoll automatically calculates the path loss matrices if not found. For neighbours accepted for co-site and coverage reasons. Atoll displays a warning in the Event viewer indicating that the constraint on the forbidden neighbour will be ignored by algorithm because the neighbour already exists. neighbours are marked as existing. Therefore. 9. If you change some allocation criteria (e.atl. it indicates the importance (in %) of each neighbour and the allocation reason.9. If the Min and Max value ranges of the importance function factors overlap. Atoll provides the list of neighbours. If the Min and Max value ranges of the importance function factors do not overlap. co-site or coverage.3. In the Results. a neighbour may be marked as exceptional pair. you can run the automatic allocation with the Delete existing neighbours option not selected. neighbours will be ranked in this order: co-site neighbours and neighbours allocated based on coverage overlapping. • • • No prediction study is needed to perform an automatic neighbour allocation. There can be a mix of the neighbourhood causes. In addition. Atoll keeps the existing inter-technology neighbours when the Delete existing neighbours option is not selected.3.1. Atoll examines the neighbour list of the TBA cells and checks allocation criteria only if there is still space left in their neighbour lists.3 Appendices 9. Atoll displays the percentage of area meeting the coverage conditions and the corresponding surface area (km2). With the default values for minimum and maximum importance fields.Atoll 3. the number of neighbours and the maximum number of neighbours allowed for each cell. Therefore.g. If a new TBA cell i is created in TD-SCDMA. the neighbours will be ranked by neighbour cause. if a TBA cell has already reached its maximum number of neighbours before starting the new allocation. A forbidden neighbour must not be listed as neighbour except if the neighbourhood relationship already exists and the Delete existing neighbours option is unchecked when you start the new allocation. When starting an automatic neighbour allocation. if cells have previous allocations in the list. In this case. Atoll displays only the cells for which it finds new neighbours. Finally. increase the maximum number of neighbours or create a new GSM TBC transmitter) and start a new allocation without selecting the Delete existing neighbours option. In the Results part.0 Technical Reference Guide for Radio Networks Chapter 9: TD-SCDMA Networks • • • ©Forsk 2015 Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. A new GSM TBC transmitter can enter the TBA cell neighbour list if allocation criteria are satisfied.9.1.. Chapter 10 WiMAX BWA Networks This chapter covers the following topics: • "Definitions" on page 673 • "Calculation Quick Reference" on page 678 • "Available Calculations" on page 690 • "Calculation Details" on page 702 • "Automatic Planning Algorithms" on page 758 . 0 Technical Reference Guidefor Radio Networks © Forsk 2015 672 .3.Atoll 3. 8.3. location in the Atoll GUI. i. or mobile (Monte Carlo simulations) covered/served by the studied cell TXi(ic). 4. and the radio resource management algorithms used by the different available schedulers. The third part describes all the calculation algorithms used in all the calculations. and their usage. and interference for downlink and uplink considering the effects of smart antennas.. A cell refers to a transmitter-carrier (TX-c) pair. a victim cell when calculating the interference it is receiving from other cells. The second part describes all the calculation processes. 12. 10. • Other cells (represented by the subscript "j") comprising the other transmitter TXj and its carrier jc. • Mi: A pixel (coverage predictions). 1/16. The first part of this chapter lists all the input parameters in the WiMAX BWA documents. subchannelisation. All the calculation algorithms in this section are described for two types of cells. For the definition of TBC transmitters please refer to "Path Loss Matrices" on page 103. The cell being studied during a calculation is referred to as TXi(ic) in this chapter.e. 20 r CP Frame configuration or. • 10.1 Definitions This table lists the input to calculations. It is the cell which is currently the focus of the calculation. otherwise. Name Value Unit Description K 1..Atoll 3. The other cells in the network can be interfering cells (downlink) or the serving cells of interfering mobiles (uplink). global parameter None Cyclic Prefix Ratio Choice List: 1/4. subscriber (calculations on subscriber lists). power control.5. signal quality coverage predictions. 1/8. These algorithms include the calculation of signal levels. calculations on subscriber lists. their significance. and Monte Carlo simulations. For example.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 10 WiMAX BWA Networks This chapter describes all the calculations performed in Atoll WiMAX documents. MIMO etc. noise. you can also see the Glossary of WiMAX Terms in the User Manual for information on WiMAX terms and concepts. If you are new to WiMAX. Logarithms used in this chapter (Log function) are base-10 unless stated otherwise. and simulations. • • • All the calculations are performed on TBC (to be calculated) transmitters. point analysis calculations. 2. signal level coverage predictions.38 x 10-23 J/K Boltzmann’s constant T 290 K Ambient temperature n0 Calculation result ( 10  Log  K  T  1000  = – 174 dBm/Hz ) dBm/Hz Power spectral density of thermal noise D Frame Global parameter ms Frame Duration Choice List: 2. All the calculation algorithms in this section are described for two types of receivers. 1/32 O Fixed DL Global parameter SD Fixed time-domain overhead (DL) O Fixed UL Global parameter SD Fixed time-domain overhead (UL) DL Global parameter % Variable time-domain overhead (DL) UL Global parameter % Variable time-domain overhead (UL) TDD Global parameter % Ratio of the DL subframe to the entire frame (TDD only) O Variable O Variable r DL-Frame 673 .5. The calculation algorithms used by these calculation processes are available in the next part. • • A studied cell (represented by the subscript "i") comprising the studied transmitter TXi and its carrier ic. especially in the context of their user in Atoll. • Mj: A mobile (Monte Carlo simulations) covered/served by any other cell TXj(jc). 5. coverage predictions. It also contains the lists of the formulas used for the calculations. Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Name Value Unit Description N SD – DL TDD Global parameter None Number of symbol durations per frame that corresponds to the DL subframe (TDD only) TDD Global parameter None Number of symbol durations per frame that corresponds to the UL subframe (TDD only) D TTG TDD Global parameter ms Transmit Time Guard (TDD only) D RTG TDD Global parameter ms Receive Time Guard (TDD only) M PC Global parameter dB Uplink power control margin CNR Min Global parametera dB Minimum signal to thermal noise threshold (interferer cutoff) PZ Permutation zone parameter None Number of subchannels per channel in UL subframe N SC – DL PZ Permutation zone parameter None Number of subchannels per channel in DL subframe N SCa – Total Frame configuration parameter None Total number of subcarriers per channel (FFT size) N SCa – Preamble Frame configuration parameter None Number of subcarriers used by the preamble PZ Permutation zone parameter None Number of used subcarriers per channel N SCa – Data PZ Permutation zone parameter None Number of subcarriers per channel used for data transfer N SCa – DC Hard-coded parameter ( N SCa – DC = 1 ) None Number of DC subcarriers per channel None Number of pilot subcarriers per channel None Number of guard subcarriers per channel N SD – UL N SC – UL N SCa – Used PZ N SCa – Pilot PZ PZ PZ Calculation result ( N SCa – Pilot = N SCa – Used – N SCa – Data ) Calculation result PZ N SCa – Guard PZ ( N SCa – Guard PZ = N SCa – Total – N SCa – Used – N SCa – DC ) PZ UL Permutation zone parameter None Uplink permutation zone PZ DL Permutation zone parameter None Downlink permutation zone QT PZ Permutation zone parameter dB Quality threshold: Required preamble C/N or C/(I+N) for accessing a zone Speed Max – PZ Permutation zone parameter Km/hr Speed limit for mobiles trying to access a permutation zone d Max – PZ Permutation zone parameter m Maximum distance from the transmitter covered by a zone p PZ Permutation zone parameter None Permutation zone priority W Channel Frequency band parameter MHz Channel bandwidth First Frequency band parameter None First channel number of the frequency band N Channel Last Frequency band parameter None Last channel number of the frequency band F Start – FB – TDD Frequency band parameter MHz Start frequency of the TDD frequency band F Start – FB – FDD – DL Frequency band parameter MHz DL Start frequency of the FDD frequency band F Start – FB – FDD – UL Frequency band parameter MHz UL Start frequency of the FDD frequency band N Channel 674 .3. Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Name Value Unit Description f Sampling Frequency band parameter None Sampling factor f ACS – FB Frequency band parameter dB Adjacent Channel Suppression Factor ICS FB Frequency band parameter MHz Inter-channel spacing CN FB Frequency band parameter None Channel number step Inter – Tech Network parameter dB Inter-technology interference reduction factor B Bearer parameter None Bearer index Mod B Bearer parameter None Modulation used by the bearer CR B Bearer parameter None Coding rate of the bearer B Bearer parameter bits/ symbol Bearer Efficiency TB Bearer parameter dB Bearer selection threshold TP BH – DL Site Site parameter kbps Maximum backhaul site downlink throughput Site Site parameter kbps Maximum backhaul site uplink throughput Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) dB Transmitter noise figure N Ant – TX Transmitter parameter None Number of antennas used for MIMO in transmission N Ant – RX Transmitter parameter None Number of antennas used for MIMO in reception TX Antenna parameter dB Transmitter antenna gain TX Transmitter parameter (user-defined or calculated from transmitter equipment characteristics) dB Transmitter loss TX Smart antenna parameter None Number of smart antenna elements Array Smart antenna parameter dB Array gain offset Combining Smart antenna parameter dB Power combining gain offset G SA Smart antenna parameter dB Diversity gain (cross-polarisation) N Channel Cell parameter None Cell’s channel number P Preamble Cell parameter dBm Preamble power dB Traffic power reduction dB Pilot power reduction dB Idle pilot power reduction f IRF TP BH – UL nf G L TX E SA G SA G SA Div Cell parameter P Traffic = P Preamble – P Traffic in dB P Traffic Ratio P Traffic = 10 P Traffic ------------------------10 in % Cell parameter P Pilot = P Preamble – P Pilot in dB P Pilot Ratio P Pilot = 10 P Pilot -------------------10 in % Cell parameter P Idle – Pilot = P Preamble – P Idle – Pilot in dB P Idle – Pilot Ratio P Idle – Pilot = 10 P Idle – Pilot -----------------------------------10 in % TL DL Cell parameter % Downlink traffic load TL UL Cell parameter % Uplink traffic load 675 .3. SU-MIMO or MUMIMO diversity gain f Bias QoS Scheduler parameter % QoS class bias factor QoS Service parameter None QoS class of the service p Service parameter None Service priority B DL – Highest Service parameter None Highest bearer used by a service in the downlink B UL – Highest Service parameter None Highest bearer used by a service in the uplink B DL – Lowest Service parameter None Lowest bearer used by a service in the downlink B UL – Lowest Service parameter None Lowest bearer used by a service in the uplink UL Service parameter % Uplink activity factor NR DL NR UL G MUG – DL G MUG – UL Max CINR MUG f Act 676 ©Forsk 2015 .3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks Name Value Unit Description TL DL – Max Cell parameter % Maximum downlink traffic load TL UL – Max Cell parameter % Maximum uplink traffic load NR UL Cell parameter dB Uplink noise rise NRUL – Seg Cell parameter dB Segmented zone uplink noise rise N Users – Max Cell parameter None Maximum number of users per cell N Users – DL Cell parameter None Number of users connected to the cell in downlink N Users – UL Cell parameter None Number of users connected to the cell in uplink SU DL Cell parameter % Downlink segmentation usage ratio AU DL Cell parameter % Downlink AAS usage ratio T AMS Cell parameter dB Adaptive MIMO switch threshold T MU – MIMO Cell parameter dB Multi-user MIMO threshold PI Cell parameter None Preamble index T Preamble Cell parameter dB Preamble C/N threshold D Reuse Cell parameter m Channel and preamble index reuse distance G MU – MIMO Cell parameter None Uplink MU-MIMO gain Inter – Tech Cell parameter dB Inter-technology downlink noise rise Inter – Tech Cell parameter dB Inter-technology uplink noise rise ZPBDL Cell parameter None Downlink zone permbase ZPB UL Cell parameter None Uplink zone permbase TX i  ic  Proportional Fair scheduler parameter None Downlink multi-user diversity gain (MUG) TX i  ic  Proportional Fair scheduler parameter None Uplink multi-user diversity gain (MUG) Proportional Fair scheduler parameter dB Maximum C/(I+N) above which no MUG gain is applied G SU – MIMO Max Cell WiMAX equipment parameter None Maximum SU-MIMO gain G Div – UL Cell WiMAX equipment parameter dB Uplink STTD/MRC.Atoll 3. 3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Name Value Unit Description f Act DL Service parameter % Downlink activity factor TPD Min – UL Service parameter kbps Minimum throughput demand in the uplink TPD Min – DL Service parameter kbps Minimum throughput demand in the downlink TPD Max – UL Service parameter kbps Maximum throughput demand in the uplink TPD Max – DL Service parameter kbps Maximum throughput demand in the downlink UL Service parameter kbps Average requested throughput in the uplink TP Average DL Service parameter kbps Average requested throughput in the downlink TP Offset Service parameter kbps Throughput offset f TP – Scaling Service parameter % Scaling factor L Body Service parameter dB Body loss N SC – UL Min Service parameter None Minimum number of subchannels P Min Terminal parameter dBm Minimum terminal power allowed P Max Terminal parameter dBm Maximum terminal power allowed nf Terminal parameter dB Terminal noise figure G Terminal parameter dB Terminal antenna gain L Terminal parameter dB Terminal loss N Ant – TX Terminal parameter None Number of antennas used for MIMO in transmission N Ant – RX Terminal parameter None Number of antennas used for MIMO in reception G SU – MIMO Max Terminal WiMAX equipment parameter None Maximum SU-MIMO gain G Div – DL Terminal WiMAX equipment parameter dB Downlink STTD/MRC or SU-MIMO diversity gain G Div – Preamble Terminal WiMAX equipment parameter dB Preamble diversity gain UL Clutter parameter dB Additional uplink diversity gain G Div DL Clutter parameter dB Additional downlink diversity gain f SU – MIMO Clutter parameter None SU-MIMO gain factor L Indoor Clutter parameter dB Indoor loss L Path Propagation model result dB Path loss TP Average G Div ICP DL Network parameter None Inter-technology downlink channel protection ratio for a frequency offset F between the interfered and interfering frequency channels M Shadowing – Model Monte Carlo simulations: Random result calculated from model standard deviation Coverage Predictions: Result calculated from cell edge coverage probability and model standard deviation dB Model shadowing margin M Shadowing – C  I Coverage Predictions: Result calculated from cell edge coverage probability and C/I standard deviation dB C/I shadowing margin F 677 .Atoll 3. 10.2. 10. ©Forsk 2015 Any interfering cell whose signal to thermal noise ratio is less than CNR Min will be discarded.if interferer uses a TDD frequency band and victim uses 100 an FDD frequency band.3.and Adjacent Channel Overlaps Calculation Name TX i  ic  F Start Value TX  ic  i TX  ic  i TX i  ic  TX i  ic  TX  jc  TX i  ic  – TX j  jc  TX  ic  Description MHz Start frequency for the channel number assigned to a cell MHz End frequency for the channel number assigned to a cell MHz Co-channel overlap bandwidth None Co-channel overlap ratio MHz Bandwidth of the lower-frequency adjacent channel overlap None Lower-frequency adjacent channel overlap ratio MHz Bandwidth of the higher-frequency adjacent channel overlap None Higher-frequency adjacent channel overlap ratio None Adjacent channel overlap ratio None FDD – TDD overlap ratio None Total overlap ratio TX i  ic  – TX j  jc  r CCO TX i  ic  – TX j  jc  TX  jc  TX  ic  TX  jc  TX  ic  TX  ic  j i j i i Min  F End  F Start  – Max  F Start  F Start – W Channel L TX i  ic  – TX j  jc  W ACO L ---------------------------------TX i  ic  W Channel TX i  ic  – TX j  jc  L TX i  ic  – TX j  jc  TX j  jc  TX i  ic  Min  F End  F End H TX  ic  TX  jc  TX  ic  i j i + W Channel – Max  F Start  F End  TX i  ic  – TX j  jc  W ACO H ---------------------------------TX  ic  i W Channel TX i  ic  – TX j  jc  r ACO TX  ic  W CCO ----------------------------------TX i  ic  W Channel TX i  ic  – TX j  jc  W ACO TX  jc  j i j i Min  F End  F End  – Max  F Start  F Start  W CCO r ACO TX i  ic  F Start + W Channel F End W ACO  N TXi  ic  – N First – TXi  ic  Channel Channel -   ------------------------------------------------------TX i  ic      CN FB TX  ic  i  F Start – FB +  W Channel + ICS FB Unit H TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  r ACO r ACO L TX i  ic  – TX j  jc  + r ACO H TDD TX i  ic  – TX j  jc  r FDD – TDD r DL – Frame ----------------------. 1 otherwise TX  ic  i  – f ACS – FB  TX  ic  – TX  jc  TX  ic  – TX  jc  ---------------------------- TX  ic  – TX  jc  10 j i j j r i  r i + r ACO  10 FDD – TDD  CCO      TX i  ic  – TX j  jc  rO TX i  ic  TX j  jc  if W Channel  W Channel TX  ic  i  – f ACS – FB TX  ic   TX  ic  – TX  jc  TX  ic  – TX  jc  --------------------------- TX  ic  – TX  jc  W i 10 j i j i j Channel r i    --------------------+ r ACO 10 TX j  jc   CCO  r FDD – TDD   W Channel   TX i  ic  TX j  jc  if W Channel  W Channel 678 .0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks a.1 Co.2 Calculation Quick Reference The following tables list the formulas used in calculations.Atoll 3. 5 Preamble C/N Calculation Name TX i  ic  CNR Preamble Value TX i  ic  TX i  ic  Mi DL C Preamble – n Preamble + G Div – Preamble + G Div 679 .3 Preamble Noise Calculation Name TX i  ic  n 0 – Preamble Preamble f Segment TX i  ic  TX i  ic  n 0 – Preamble + nf n Preamble Mi 10.and adjacent channel overlap dB Interference reduction factor due to segmentation None Preamble subcarrier collision probability W Downlink inter-technology interference Unit Description dB Preamble C/N for a cell TX  ic  i EIRP Preamble – L Path – M Shadowing – Model – L Indoor + G –L Mi Mi – L Ant – L Body TX i  ic  TX i  ic  TX i –L TX i With smart antennas: TX i  ic  P Preamble + G TX i –L TX i TX i + 10  Log  E SA  + TX L Path L Total i Mi Without smart antennas: P Preamble + G EIRP Preamble M Combining G SA + Div G SA i L Model + L Ant L Path + L Mi TX i + L Indoor + M Shadowing – Model – G TX i +L Mi –G Mi Mi + L Ant + L Body 10.2.2.2 Preamble Signal Level Calculation Name TX  ic  i C Preamble Value Unit Description dBm Received preamble signal level dBm Preamble EIRP of a cell dB Path loss dB Total losses Value Unit Description TX i  ic   TX  ic  N SCa – Preamble Preamble i  .2.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 10.Atoll 3. f Segment n 0 + 10  Log  F Sampling  --------------------------------TX i  ic    N SCa – Total  dBm Preamble thermal noise for a cell 1 --3 None Preamble segmenting factor dBm Preamble noise for a cell Unit Description dBm Total interference generated by an interfering cell dB Interference reduction factor due to the co.3.2.4 Preamble Interference Calculation Name TX j  jc  I Preamble Value TX j  jc  TX i  ic  – TX j  jc  C Preamble + f O TX i  ic  – TX j  jc  TX  ic  – TX  jc  i j  TX i  ic  – TX j  jc  10  Log  r O TX  ic  – TX  jc  i j 10  Log  p Collision  fO TX i  ic  – TX j  jc   TX i  ic  – TX j  jc  f Seg – Preamble p Collision Inter – Tech + f Seg – Preamble + I DL TX i  ic  1 if N Seg TX j  jc  = N Seg   TX i  ic  and 0 if N Seg TX j  jc   N Seg TX k   P DL – Rec  -------------------------------------- F  TX  ic   TX   i k  TX k  ICP DL  Inter – Tech I DL 10. 2.2.3.with downlink segmentation 15 TX i  ic  n 0 – DL + nf Mi .2.6 Preamble C/(I+N) Calculation Name Value Unit Description TX  ic  TX i  ic  CINR Preamble i      TXj  jc   n Preamble    IPreamble  ----------------------------- TX i  ic  Inter – Tech 10 ---------------------------- Inter – Tech     C Preamble – 10  Log + 10 + NR DL 10 I    10  DL   dB   All TXj  jc      Preamble C/(I+N) for a cell        Mi DL + G Div – Preamble + G Div TX  ic  TX  ic  i  I + N  Preamble i    TXj  jc   n Preamble   IPreamble ----------------------------- Inter – Tech 10 --------------------------  +I  + NR Inter – Tech dBm 10  Log  + 10 10 DL   10  DL   All TXj  jc          Preamble Total Noise (I+N) for a cell 10.8 Traffic and Pilot Noise Calculation (DL) Name Value Unit Description dBm Thermal noise for a cell None Downlink segmenting factor dBm Downlink noise for a cell Mi TX i  ic  n 0 – DL PZ DL   N SCa – Used   TXi  ic  n 0 + 10  Log  F Sampling  ------------------------ TX i  ic   N SCa – Total  With Segmentation: Mi PZ DL    TXi  ic   N SCa – Used n 0 + 10  Log  F Sampling  ------------------------ f Segment – DL TX i  ic    N SCa – Total   f Segment – DL TX i  ic  n DL 680 3  PSG + 2  SSG 1 without and --------------------------------------------.7 Traffic and Pilot Signal Level Calculation (DL) Name TX i  ic  C Traffic TX i  ic  C Pilot TX i  ic  EIRP Traffic TX i  ic  EIRP Pilot TX i  ic  P Traffic TX i  ic  P Pilot Value TX i  ic  EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G Mi –L Mi dBm Received traffic signal level dBm Received pilot signal level Mi Mi TX i  ic  Mi Description – L Ant – L Body EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G –L Unit Mi Mi Mi – L Ant – L Body TX i  ic  TX i TX i  ic  TX i P Traffic + G P Pilot + G Array + G SA Array + G SA + G SA + G SA Div TX i dBm Traffic EIRP of a cell Div TX i dBm Pilot EIRP of a cell Combining + G SA – L Combining + G SA – L TX i  ic  TX i  ic  dBm Traffic transmission power of a cell TX i  ic  TX i  ic  dBm Pilot transmission power of a cell P Preamble – P Traffic P Preamble – P Pilot 10.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 10. 2.9 Traffic and Pilot Interference Calculation (DL) Name Value TX  jc  Unit Description dBm Total interference generated by an interfering cell TX  jc  j  I j  I Non – AAS Idle  ---------------------------------------------- 10 10  Monte Carlo Simulations: 10  Log  10 + 10       TX  jc   I j  AAS   -----------------10  without smart antennas.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 10.Atoll 3. or 10  Log  10 with smart       antennas TX j  jc  I Total TX  jc  TX  jc  TX  jc  j j  I j  I I Non – AAS Idle AAS  ---------------------------------------------------------------- 10 10 10  Coverage Predictions: 10  Log  10 + 10 + 10       Monte Carlo Simulations: TX j  jc  EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G TX j  jc  Mi –L Mi Mi Mi – L Ant – L Body Coverage Predictions: I Traffic dBm TX j  jc  Traffic interference power of an interfering cell EIRP Traffic – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor +G Mi –L Mi Mi Mi – L Ant – L Body Monte Carlo Simulations: TX  jc  j EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G TX j  jc  M i –L M i M i M i – L Ant – L Body dBm Pilot interfering power of an interfering cell TX j dBm Traffic EIRP of an interfering cell TX j dBm Pilot EIRP of an interfering cell dBm Interference from the loaded part of the frame transmitted using the transmitter antenna of an interfering cell Coverage Predictions: I Pilot TX  jc  j EIRP Pilot – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor +G Mi Mi –L Mi Mi – L Ant – L Body TX j  jc  TX j  jc  TX j TX j  jc  TX j P Traffic + G EIRP Traffic TX j  jc  P Pilot + G EIRP Pilot –L –L TX  jc  TX j  jc  I Non – AAS  I j  TX j  jc  Traffic  ------------------ TX  jc  TX  jc  N SCa – Data 10 j j     -+ 10  Log TL DL  ----------------------- 1 – AU DL  10 TX j  jc      N SCa – Used     10 TX j  jc  I Pilot ------------------10  TX j  jc    N SCa – Data   1 – -------------------------     TX j  jc   N SCa – Used    Monte Carlo Simulations: TX j  jc  – L Path – M Shadowing – Model – L Indoor + G EIRP AAS TX j  jc  TX j  jc  EIRP AAS +G TX  jc  j –L Mi Mi Mi –L Mi – L Ant – L Body Coverage Predictions: I AAS EIRP AAS Mi dBm Interference power of an interfering cell transmitted using smart antenna dBm Traffic EIRP of an interfering cell using smart antenna – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor Mi Mi Mi – L Ant – L Body TX  jc  j P Traffic + G TX j –L TX j 681 .3. 0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Name Value TX  jc  j I Idle – Pilot TX  jc  j EIRP Idle – Pilot TX j  jc  – L Path – L Indoor + G TX j  jc  P Idle – Pilot + G EIRP Idle – Pilot TX j M i –L –L M i – M i L Ant – M i L Body TX j Unit Description dBm Interference from empty part of the frame transmitted using the transmitter antenna of an interfering cell dBm Idle pilot EIRP of an interfering cell dBm Interference from the empty part of the frame transmitted using the transmitter antenna of an interfering cell dB Interference reduction factor due to the co.and adjacent channel overlap TX  jc  TX j  jc  I Idle   I j  TX j  jc  Idle – Pilot.Atoll 3.   ----------------------------TX  jc   N 10 j SCa – Data   10  Log  1 – TL DL   10  1 – ------------------------   TX j  jc     N SCa – Used      TX  ic  – TX  jc  i j  TX i  ic  – TX j  jc  10  Log  r O TX  ic  – TX  jc  i j i j 10  Log  p Collision – DL    dB Interference reduction factor due to downlink segmentation TX k   P DL – Rec  -------------------------------------- F  TX i  ic  TX k    TX k  ICP DL W Downlink inter-technology interference Unit Description dB Traffic C/N for a cell dB Pilot C/N for a cell fO  TX  ic  – TX  jc  f Seg – DL  Inter – Tech I DL 10.3. Inter – Tech Inter – Tech 10  +I  + NR  +10 DL DL    dB       M i Pilot C/(I+N) for a cell DL With MIMO: CINR Pilot + G Div – DL + G Div TX  jc  TX i  ic   I + N  DL 682 TX  ic  i    I j  n DL DL   ------------------ ---------------------  10 10  + I Inter – Tech + 10 10  + NR Inter – Tech 10  Log  DL     DL  All TXj  jc          dBm Traffic Total Noise (I+N) for a cell .10 Traffic and Pilot C/N Calculation (DL) Name Value TX i  ic  TX i  ic  C Traffic – n DL TX i  ic  CNR Traffic TX i  ic  Mi DL With MIMO: CNR Traffic + G Div – DL + G Div TX i  ic  TX i  ic  C Pilot – n DL TX i  ic  CNR Pilot TX i  ic  Mi With MIMO: CNR Pilot + G Div – DL + DL G Div 10.11 Traffic and Pilot C/(I+N) Calculation (DL) Name Value    TXj  jc     IDL    -----------------C Traffic – 10  Log  10 10   All TX j  jc      TX i  ic  TX i  ic  CINR Traffic  TX i  ic  Unit  TX i  ic   n DL  + I Inter – Tech + -------------------10  DL 10   Mi      + NR Inter – Tech DL   dB     Description Traffic C/(I+N) for a cell DL With MIMO: CINR Traffic + G Div – DL + G Div TX  jc  TX i  ic  TX i  ic  CINR Pilot C Pilot    I j DL    -----------------10    10 – 10  Log      All TXj  jc       TX  ic  i TX  ic  i    n DL   --------------------.2.2. 2.Atoll 3.2.13 Traffic Noise Calculation (UL) Name Value Mi TX i  ic  n 0 – UL PZ UL    TXi  ic  N SCa – Used  n 0 + 10  Log  F Sampling  ------------------------ TX i  ic    N SCa – Total  TX i  ic  TX i  ic  n 0 – UL + nf n UL TX i  ic  10.12 Traffic Signal Level Calculation (UL) Name Value M M i C UL Unit Description dBm Received uplink signal level dBm Uplink EIRP of a user equipment Unit Description dBm Thermal noise for a cell dBm Uplink noise for a cell Unit Description dBm Uplink interference received at a cell dB Interference reduction factor due to the co.14 Traffic Interference Calculation (UL) Name M j I UL Value M TX  ic  – TX  jc  i j j C UL + f O M TX  ic  – TX  jc  i j j + f TL – UL + f Seg – UL TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  10  Log  r O fO   M Mj j 10  Log  TL UL   f TL – UL TX  ic  – TX  jc  TX  ic  – TX  jc  i j f Seg – UL TX i  ic  – TX j  jc  p Collision – UL TX i  ic  NR UL TX i  ic  NRUL – Seg TX i  ic   I + N  UL  TX i  ic   IMj    n UL  UL  non-seg M   --------------------- TX i  ic   --------------------------------------------i Inter – Tech 10   10 10  Log  + NR UL – n UL dB  10  + 10    All M j         All TX j  jc     TX i  ic   IMj    n UL UL    seg M  --------------------- TX i  ic  i   Inter – Tech 10  --------------------------------10  Log  + NR UL – n UL 10  10  + 10     All M j        All TX  jc    j  TX i  ic  NR UL TX i  ic  + n UL TX i  ic  TX i  ic  or NR UL – Seg + n UL dB Segmented zone uplink noise at a cell without smart antennas dBm Total Noise (I+N) for a cell dB Uplink noise at a cell with smart antenna 2 NR UL    I UL    +  n  I --------------------------------2 n  I Non-segmented zone uplink noise at a cell without smart antennas 683 .0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 10.3.and adjacent channel overlap dB Interference reduction factor due to the interfering mobile’s uplink traffic load i j 10  Log  p Collision – UL    db Interference reduction factor due to uplink segmentation SC Com -----------------TX i  ic  SC None Uplink segmentation collision probability i EIRP UL – L Path – M Shadowing – Model – L Indoor + G –L TX i Mi Mi With P Mi i Mi – L Ant – L Body P EIRP UL TX Mi +G Mi –L Mi Mi = P Max without power control and P Mi Mi = P Eff after power control 10.2. 2.16 Traffic C/(I+N) Calculation (UL) Name Value TX i  ic  Mi Without smart antennas: CNR UL – NR UL Mi CINR UL Mi TX i  ic  Mi With smart antennas: CNR UL – NR UL Mi TX i  ic  or CNR UL – NR UL – Seg TX i  ic   UL With MIMO: CINR UL + G Div – UL + G Div 10.3.17 Calculation of Total Cell Resources Name Value TX  ic  i F Sampling  W Channel  10  -  8000 Floor  f Sampling  ----------------------------------8000   F TX i  ic  TX  ic  i TX i  ic  TX i  ic  D Sym – Useful TX i  ic  Used D Frame TX i  ic  N  SD – Used   Frame –3 TX i  ic  TX i  ic  r CP -------------F D CP D Symbol 6 TX i  ic  TX i  ic  D Sym – Useful + D CP TDD TDD If DL:UL ratio is defined in percentage: TX  ic  TX  ic  i N  SD – DL   Subframe i TDD DL RoundUp  N  SD – Used   Frame  r DL – Frame – O Fixed If DL:UL ratio is defined in fraction: TDD  TXi  ic   N SD – DL DL RoundUp  N SD – Used   Frame  ----------------------------------------- – O Fixed TDD TDD  N SD – DL + N SD – UL 684 .2.Atoll 3.2.15 Traffic C/N Calculation (UL) Name Value TX i  ic  Mi C UL – n UL Mi CNR UL Mi TX i  ic  With MIMO: CNR UL + G Div – UL + UL G Div 10.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Name TX  ic  i  I + N  UL Value Unit Description dBm Total Noise (I+N) for a cell in case of smart antennas Unit Description dB Uplink C/N at a cell Unit Description dB Uplink C/(I+N) at a cell Unit Description Hz Sampling frequency F Sampling  10 -------------------------------------TX i  ic  N SCa – Total kHz Inter-subcarrier distance 1 ------------------TX i  ic  F ms Useful symbol duration ms Cyclic prefix duration ms Symbol duration D Frame – D TTG – D RTG ms Used frame duration  D Used  Frame  Floor  ---------------- TXi  ic    D Symbol SD Frame duration in terms of symbol durations SD Downlink subframe duration in terms of symbol durations 2 I UL    +  n  I  10. Atoll 3. i.e. the number of symbols in the downlink subframe SD Uplink subframe duration in terms of symbol durations Symbols Total uplink cell resources. i.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Name TX  ic  i R DL = TX  ic  i N  Sym – DL   Subframe Value Unit Description M   i DL PZ O Variable   TXi  ic  DL -  Floor  N  SD – DL   Subframe  N SCa – Data   1 – -------------------100     Symbols Total downlink cell resources. G MUG – DL D Frame Mi CTP P – DL M i With downlink segmentation: CTP P – DL  f Segment – DL With MIMO (SU-MIMO):  Mi B DL Max =  Mi   1 + f SU – MIMO  G SU – MIMO – 1   B DL With MIMO (AMS):  Mi B DL TX i  ic  =  Max Mi B DL TX i  ic  if CNR Preamble  T AMS M i CTP E – DL Mi CTP A – DL Mi Cap P – DL M i Cap E – DL Mi Cap A – DL M i PUTP P – DL   1 + f SU – MIMO  G SU – MIMO – 1   TX i  ic  M M i i CTP P – DL   1 – BLER  B DL     Mi Mi Mi f TP – Scaling .– TP Offset Cap E – DL  -----------------------100 Mi TX i  ic  or CINR Preamble  T AMS Mi Cap P – DL ----------------------TX i  ic  N Users – DL 685 .18 Channel Throughput. Allocated Bandwidth Throughput.e..2. and Per-user Throughput Calculation Name Value TX i  ic  R DL  M B Unit Description kbps Downlink peak MAC channel throughput kbps Downlink effective MAC channel throughput kbps Downlink application channel throughput kbps Downlink peak MAC cell capacity kbps Downlink effective MAC cell capacity kbps Downlink application cell capacity kbps Downlink peak MAC throughput per user i DL --------------------------------D Frame TX  ic  i R DL  M i B DL TX  ic  i For proportional fair schedulers: --------------------------------.– TP Offset CTP E – DL  -----------------------100 TX i  ic  Mi CTP P – DL  TL DL – Max M M i i Cap P – DL   1 – BLER  BDL     Mi Mi f TP – Scaling . the number of symbols in the uplink subframe If DL:UL ratio is defined in percentage: TX  ic  i TDD UL RoundDown  N SD – Used   Frame   1 – r DL – Frame  – O Fixed TX  ic  i If DL:UL ratio is defined in fraction: N  SD – UL   Subframe TDD  TX i  ic   N SD – UL UL RoundDown  N SD – Used   Frame  ----------------------------------------- – O Fixed TDD TDD  N SD – DL + N SD – UL TX i  ic  R UL = TX i  ic  N  Sym – UL   Subframe M   i UL PZ UL O Variable   TXi  ic   Floor  N SD – UL   Subframe  N SCa – Data   1 – ---------------------  100     10. Cell Capacity.. – TPOffset PUTP E – DL  -----------------------100 Mi PUTP A – DL TX  ic  i R UL  Mi B UL --------------------------------D Frame TX i  ic  R UL  B Mi TX  ic  i UL For proportional fair schedulers: --------------------------------.– TP Offset 100 CTP E – UL M i CTP A – UL M i Cap P – UL Mi M i TX  ic  i i CTP P – UL  TL UL – Max M M Mi i i Cap P – UL   1 – BLER  B UL  M Mi f TP – Scaling Cap E – UL  ------------------------. G MUi – MIMO D Frame M M Mi i i CTP P – UL   1 – BLER  B UL  M Mi f TP – Scaling CTP E – UL  ------------------------. G MUG – UL D Frame With MIMO (SU-MIMO): Mi CTP P – UL  M i B UL =  Max   1 + f SU – MIMO  G SU – MIMO – 1   M i B UL With MIMO (AMS):  B Max =  Mi B UL TX i  ic  Mi   1 + f SU – MIMO  G SU – MIMO – 1   UL TX i  ic  if CNR Preamble  T AMS TX i  ic  TX i  ic  or CINR Preamble  T AMS With MIMO (MU-MIMO) in uplink throughput coverage predictions: TX i  ic  R UL  Mi B UL TX  ic  --------------------------------.3.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Name Value Unit Description kbps Downlink effective MAC throughput per user kbps Downlink application throughput per user kbps Uplink peak MAC channel throughput kbps Uplink effective MAC channel throughput kbps Uplink application channel throughput kbps Uplink peak MAC cell capacity kbps Uplink effective MAC cell capacity kbps Uplink application cell capacity kbps Uplink peak MAC allocated bandwidth throughput kbps Uplink effective MAC allocated bandwidth throughput kbps Uplink application allocated bandwidth throughput  Cap Mi  M P – UL - ABTP P –i UL Min  ---------------------- TXi  ic    N Users – UL  kbps Uplink peak MAC throughput per user  Cap Mi  M E – UL - ABTP E –i UL Min  ----------------------TX i  ic    N Users – UL  kbps Uplink effective MAC throughput per user M i Cap E – DL ----------------------TX  ic  i N Users – DL M i PUTP E – DL M Mi i Mi f TP – Scaling .– TP Offset 100 Cap E – UL M i Cap A – UL Mi i M M ABTP P – UL Mi ABTPE – UL Mi ABTP A – UL M i PUTP P – UL M i PUTP E – UL 686 i N SC – UL CTP P – UL  ----------------Mi Mi i PZ UL N SC M M i i ABTP P – UL   1 – BLER  B UL     M i ABTP E – UL Mi M f TP – Scaling i .– TP Offset  -----------------------100 . 0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Name M Value M i PUTP A – UL M i PUTP E – UL i M f TP – Scaling i  ------------------------.3.19 Scheduling and Radio Resource Management Name Value Unit Description Sel Mi R Min – DL TPD Min – DL --------------------------- None Resources allocated to a mobile to satisfy its minimum throughput demand in downlink None Resources allocated to a mobile to satisfy its minimum throughput demand in uplink None Remaining downlink cell resources after allocation for minimum throughput demands R Min – UL None Remaining uplink cell resources after allocation for minimum throughput demands Sel Mi Sel Mi kbps Remaining throughput demand for a mobile in downlink Sel Mi Sel Mi kbps Remaining throughput demand for a mobile in uplink TX i  ic  kbps Downlink peak channel throughput with multi-user diversity gain (Proportional Fair) TX  ic  i kbps Uplink peak channel throughput with multi-user diversity gain (Proportional Fair) None Remaining resource demand for a mobile in downlink None Remaining resource demand for a mobile in uplink Sel Mi Sel Mi CTP P – DL M Sel i TPD Min – UL --------------------------- Sel M i R Min – UL M Sel i CTP P – UL TX i  ic  R Rem – DL TX i  ic  R Rem – UL TX i  ic   TL DL – Max – Sel Mi TX i  ic   TL DL – Max – M Sel Mi TPD Rem – DL Sel Mi TPD Rem – UL Sel Mi CTP P – DL Sel Mi CTP P – UL Sel Mi RD Rem – DL Sel Mi R Min – DL Sel Mi Sel i TPD Max – DL – TPD Min – DL TPD Max – UL – TPD Min – UL Sel Mi CTP P – DL M Without MUG  G MUG – DL Without MUG  G MUG – UL Sel i CTP P – UL Sel Mi TPD Rem – DL ---------------------------Sel Mi CTP P – DL M Sel M i RD Rem – UL Sel i TPD Rem – UL ---------------------------Sel M i CTP P – UL 687 .2.Atoll 3.– TP Offset 100 Unit Description kbps Uplink application throughput per user 10. 3.= -------------------------1 + --------= -------------------------= -------------------------Sel Sel Sel 100 Mi Mi Mi R Max – rtPS R Max – nrtPS R Max – BE r TX i  ic  R QoS – DL 1 QoS N QoS   --- TX i  ic   R Rem – DL  ------------------------------------------------------r 1 QoS N QoS   ---   All QoS 688 .Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Name Value Unit Description None Resources allocated to a mobile to satisfy its maximum throughput demand in downlink None Resources allocated to a mobile to satisfy its maximum throughput demand in uplink   Sel Mi  TXi  ic   Min  R Rem – DL RD Rem – DL   Sel   M i None Effective remaining downlink resources in a cell (Proportional Demand)   Sel Mi  TX i  ic   Min  R Rem – UL RD Rem – UL   Sel   Mi None Effective remaining uplink resources in a cell (Proportional Demand) None QoS class bias (Biased (QoS Class)) None Remaining downlink cell resources after allocation for minimum throughput demands for a QoS class (Biased (QoS Class)) TX  ic  i Sel R Rem – DL  Mi - Proportional Fair: Min  RD Rem – DL -------------------N   Sel Mi TX i  ic  RD Rem – DL Proportional Demand: R Eff – Rem – DL  ---------------------------------Sel Mi  RDRem – DL M Sel i R Max – DL M Biased (QoS Class): Sel i TX i  ic  Sel  Mi Min  RD Rem – DL R QoS – DL -------------------- N QoS   Sel Mi TPD Rem – DL Max Aggregate Throughput: --------------------------Sel Mi CTP P – DL TX i  ic  Sel  Mi R Rem – DL - Round Robin: Min  RD Rem – DL -------------------N   TX i  ic  Sel R Rem – UL  Mi - Proportional Fair: Min  RD Rem – UL -------------------N   Sel Mi TX i  ic  RD Rem – UL Proportional Demand: R Eff – Rem – UL  ---------------------------------Sel Mi  RDRem – UL M Sel i R Max – UL M Sel i TX i  ic  Sel Mi R QoS – UL  - Biased (QoS Class): Min  RD Rem – UL ------------------N QoS   Sel Mi TPD Rem – UL Max Aggregate Throughput: --------------------------Sel Mi CTP P – UL TX i  ic  Sel  Mi R Rem – UL - Round Robin: Min  RD Rem – UL -------------------N    TX i  ic  R Eff – Rem – DL  TX  ic  i R Eff – Rem – UL QoS  Sel Mi Sel Mi Sel Mi R Max – ErtPS f Bias R Max – rtPS R Max – nrtPS . 2.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Name Value Unit Description TX  ic  i R QoS – UL 1 QoS N QoS   --- TX  ic   i R Rem – UL  ------------------------------------------------------r 1 QoS N QoS   ---   None Remaining downlink cell resources after allocation for minimum throughput demands for a QoS class (Biased (QoS Class)) Sel Sel   Mi   Mi    R Max – DL  CTP E – DL     Sel   M i  Site Max  1 -------------------------------------------------------------------------------------------------------  Sel Sel  Mi    Mi  Site  R Min – DL  CTP E – DL   TP BH – DL –    Sel   M i  Site None Site backhaul overflow ratio in downlink Sel Sel   Mi   Mi   R  CTP  Max – UL  E – UL     Sel   M  Site i  Max  1 ------------------------------------------------------------------------------------------------------Sel Sel  M M     Site i i  R Min – UL  CTP E – UL   TP BH – UL –    Sel   M  Site i None Site backhaul overflow ratio in uplink None Total resources assigned to a mobile in downlink (Downlink traffic load of the mobile) None Total resources assigned to a mobile in uplink (Uplink traffic load of the mobile) Unit Description kbps Downlink peak MAC user throughput kbps Downlink effective MAC user throughput kbps Downlink application user throughput kbps Uplink peak MAC user throughput kbps Uplink effective MAC user throughput kbps Uplink application user throughput r  All QoS  Site BHOF DL   Site BHOF DL  Sel Sel Mi TL DL Sel Mi = R DL Sel Mi Sel Mi R  Mi   Mi Max – DL  CTP P – DL  R Min – DL  CTP P – DL +  -----------------------------------------------Site     BHOF DL  -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – DL Sel Sel Mi TL UL Sel Mi = R UL Sel Mi Sel Mi R  Mi   Mi Max – UL  CTP P – UL  R Min – UL  CTP P – UL +  -----------------------------------------------Site     BHOF UL  -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – UL 10.– TP Offset UTP E – UL  -----------------------100 689 .Atoll 3.– TP Offset 100 Sel Mi R UL Sel Mi  CTP P – UL Sel Sel Mi   Mi   UTP P – UL   1 – BLER  B UL      Sel Mi Sel Mi Sel Mi f TP – Scaling .3.20 User Throughput Calculation Name Sel Mi UTP P – DL Sel Mi UTP E – DL Sel i UTP A – DL M Sel Mi UTP P – UL Sel Mi UTP E – UL Sel Mi UTP A – UL Value Sel Mi R DL Sel Mi  CTP P – DL Sel Sel M  Mi    i UTP P – DL   1 – BLER  B DL      Sel Mi Sel Mi Sel Mi f TP – Scaling UTP E – DL  ------------------------. L M i TX  ic  i • Preamble signal level C Preamble • Path loss L Path • Total losses L Total . The bar graph displays cells whose C/ N are higher than the minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the highest interference level on the studied channel. You can use a value other than 30 dB for the margin from the highest interference level in the studied channel. see the Administrator Manual.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 10.1.4 Details View Analysis provided in the details view is based on path loss matrices.3. pilot. For each cell.3. for example a smaller value for improving the calculation speed. Interference level bar graphs show the interference levels on different channels in decreasing order. and interference from other cells.1 Profile View The point analysis profile view displays the following calculation results for the selected transmitter based on the calculation algorithm described in "Preamble Signal Level Calculation" on page 707. You can use a value other than 30 dB for the margin from the highest interference level on the preamble.2 Reception View Analysis provided in the reception view is based on path loss matrices. Atoll displays the best server preamble. For more information on defining a different value for this margin. interference values are listed for all the cells whose C/N are higher than the minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the highest interference level on the preamble. The bar graph displays cells whose received preamble signal levels are higher than their preamble C/N thresholds and are within a 30 dB margin from the highest preamble signal level. Reception level bar graphs show the signal levels or C/N in decreasing order. 10. for example a smaller value for improving the calculation speed. So.3 Available Calculations 10. see the Administrator Manual. For more information on defining a different value for this margin. you can display received signal levels from the cells for which calculated path loss matrices are available. The maximum number of bars in the graph depends on the preamble signal level of the best server. So.3. You can use a value other than 30 dB for the margin from the highest preamble signal level. see the Administrator Manual.3 Interference View Analysis provided in the interference view is based on path loss matrices. 10. or traffic signal level or C/N. Atoll displays the best server preamble signal level and interference from other cells. you can display the received signal level from the best server and interfering signal levels from other cells for which calculated path loss matrices are available. or traffic signal level. L Ant . The maximum number of bars in the graph depends on the highest interference level on the studied channel. you can display the received signal level from the best server and interfering signal levels from other cells for which calculated path loss matrices are available. Other cells are listed in the decreasing order of preamble signal level.1. As well. For more information on defining a different value for this margin. The results for the best server (first row) are displayed using bold italic characters. for example a smaller value for improving the calculation speed.G M i M i M i . For each cell. For each cell. pilot.1.3.1 Point Analysis 10.1. Atoll displays the received preamble. and L Body are not used in the calculations performed for the profile view. 690 .Atoll 3.3. So. 10.3. All the cells from which the received preamble signal level is higher than their preamble C/N thresholds are listed in the table. TX  ic  TX  ic  TX  ic  i i i MinimumThreshold  C Preamble  or L Total or L Path   MaximumThreshold   AND TX i  ic  TX j  jc  C Preamble  Best  C Preamble – M  ji  Where M is the specified margin (dB). Coverage Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine coverage areas to display. For these calculations. see "Path Loss Calculation Prerequisites" on page 57 for more information). Each pixel within the calculation area of TXi(ic) is considered a noninterfering receiver.2. The Best function considers the highest value from a list of values.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 10. Then. and L Body are not considered in the calculations performed for the preamble signal level based coverage predictions. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. • • • If M = 0 dB. TX  ic  TX  ic  TX  ic  i i i MinimumThreshold  C Preamble  or L Total or L Path   MaximumThreshold AND 691 . see: • • "Coverage Area Determination" on page 691. These coverage predictions do not depend on the traffic input. TX  ic  TX  ic  TX  ic  i i i MinimumThreshold  C Preamble  or L Total or L Path   MaximumThreshold   • Best Signal Level and a Margin The coverage area of each cell TXi(ic) corresponds to the pixels where. There are three possibilities.1 Preamble Signal Level Coverage Predictions The following coverage predictions are based on the received preamble signal levels: • • • Coverage by Transmitter Coverage by Signal Level Overlapping Zones For these calculations. Atoll considers pixels where the received preamble signal level from TXi(ic) is 2 dB higher than the received preamble signal levels from the cells which are 2nd best servers. If M = -2 dB. Atoll considers pixels where the received preamble signal level from TXi(ic) is either the highest or within a 2 dB margin from the highest. see "Preamble Signal Level Calculation" on page 707 For more information on coverage area determination and available display options.Atoll 3. Atoll calculates the received preamble signal level. For more information on preamble signal level calculations.3. "Coverage Display Types" on page 692. • All Servers The coverage area of each cell TXi(ic) corresponds to the pixels where. these calculations are of special interest before and during the deployment stage of the network to study the coverage footprint of the system. Atoll determines the selected display parameter on each pixel inside the cell’s calculation area. the best server calculation is always based on preamble signal level.G Mi Mi Mi . The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction.3. Atoll considers pixels where the received preamble signal level from TXi(ic) is the highest. L Mi . • Second Best Signal Level and a Margin The coverage area of each cell TXi(ic) corresponds to the pixels where. L Ant .3. Therefore. If M = 2 dB.2 Coverage Predictions 10. dBµV/m) Best Signal Level (dBm.3. the pixel falls within the coverage areas of these cells). and other criteria such as: • • • • • • • Signal Level (dBm.Atoll 3. see "Permutation Zone Selection" on page 714..0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks TX  ic  ©Forsk 2015 TX  jc  nd i j C Preamble  2 Best  C Preamble – M   ji Where M is the specified margin (dB). the cell with the highest preamble signal level) and evaluates the path loss from this cell. Atoll considers pixels where the received preamble signal level from TXi(ic) is 2 dB higher than the received preamble signal levels from the cells which are 3rd best servers. dBµV. and a service. If M = 2 dB.. see "Path Loss Calculation Prerequisites" on page 57 for more information).3.. see: • • • "Preamble Signal Level Calculation" on page 707. Atoll considers pixels where the received preamble signal level from TXi(ic) is the second highest. 10.. Therefore. Path Loss (dB) Total Losses (dB) Best Server Path Loss (dB): Where cell coverage areas overlap. see: • • 692 "Coverage Area Determination" on page 693. and take into account the receiver characteristics ( L • • Mi . Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver.e.e. Atoll calculates the received signal level or C/N level at each pixel for the channel type being studied. "Traffic Signal Level Calculation (UL)" on page 729 For more information on permutation zone selection. • • • If M = 0 dB. Number of Servers: Atoll evaluates the number of cells that cover a pixel (i. The properties of the non-interfering probe receiver are set by selecting a terminal.G Mi Mi Mi . "Traffic and Pilot Signal Level Calculation (DL)" on page 715. or pilot. For more information on C/N level calculations. preamble. "Traffic and Pilot C/N Calculation (DL)" on page 726 "Traffic C/N Calculation (UL)" on page 734. For more information on coverage area determination and available display options. traffic. Best Server Total Losses (dB): Where cell coverage areas overlap. L Ant .2 Effective Signal Analysis Coverage Predictions The following coverage predictions are based on the received preamble. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. These coverage predictions do not depend on the traffic input. It is possible to display the coverage predictions with colours depending on any transmitter or cell attribute. and L Body ) when calculating the required parameter: Effective Signal Analysis (DL) Effective Signal Analysis (UL) For these calculations. see: • • • "Preamble C/N Calculation" on page 712.2. "Coverage Display Types" on page 693. Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display parameter is greater than or equal to the defined thresholds values. For more information on signal level calculations. dBµV.e. these calculations are of special interest before and during the deployment stage of the network to study the coverage footprint of the system. The 2nd Best function considers the second highest value from a list of values. Atoll keeps the highest value of the signal level. i. a mobility type. . Atoll determines the best cell (i. Atoll considers pixels where the received preamble signal level from TXi(ic) is either the second highest or within a 2 dB margin from the second highest. If M = -2 dB. dBµV/m): Where cell coverage areas overlap.e. Coverage consists of several independent layers that can be displayed and hidden on the map. or pilot signal levels and noise. the cell with the highest preamble signal level) and evaluates the total losses from this cell. traffic. Atoll determines the best cell (i. 3 C/(I+N)-based Coverage Predictions The following coverage predictions are based on the received signal levels. These parameters can either be calculated by Atoll during the Monte Carlo simulations. • • • • • • • • Coverage by C/(I+N) Level (DL) Service Area Analysis (DL) Coverage by Throughput (DL) Coverage by Quality Indicator (DL) Coverage by C/(I+N) Level (UL) Service Area Analysis (UL) Coverage by Throughput (UL) Coverage by Quality Indicator (UL) These coverage predictions take into account the receiver characteristics ( L Mi . L Ant .Atoll 3. and the uplink coverage predictions are based on the uplink noise rise values.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Coverage Area Determination These coverage predictions are all best server coverage predictions.3. noise. For these calculations. Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display type parameter is greater than or equal to the defined thresholds values. total noise. and interference. see: • • • • "Preamble C/(I+N) Calculation" on page 712. 693 . and a service. the coverage area of each cell comprises the pixels where the cell is the best server. and interference at each pixel. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. Best server for each pixel is calculated as explained in "Best Server Determination" on page 713.3. Cell Capacity.G Mi Mi Mi . "Noise Rise Calculation (UL)" on page 733 For more information on thoughput calculations. For more information on C/(I+N). "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737. and L Body ) when calculating the required parameter. It is possible to display the Effective Signal Analysis (DL) coverage prediction with colours depending on the following display options: • • • • • • • • Preamble Signal Level (DL) (dBm) Pilot Signal Level (DL) (dBm) Traffic Signal Level (DL) (dBm) Preamble C/N Level (DL) (dB) Pilot C/N Level (DL) (dB) Traffic C/N Level (DL) (dB) Permutation Zone (DL) Segment It is possible to display the Effective Signal Analysis (UL) coverage prediction with colours depending on the following display options: • • • Signal Level (UL) (dBm) C/N Level (UL) (dB) Permutation Zone (UL) 10. The properties of the non-interfering probe receiver are set by selecting a terminal. The downlink coverage predictions are based on the downlink traffic loads of the cells. For more information on coverage area determination and available display options. Coverage consists of several independent layers that can be displayed and hidden on the map. Atoll calculates the received signal level. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. see: • • "Coverage Area Determination" on page 694.e.. or set manually by the user for all the cells. i. see: • "Channel Throughput. and Per-User Throughput Calculation" on page 743. a mobility type. see "Path Loss Calculation Prerequisites" on page 57 for more information). Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes.2. Allocated Bandwidth Throughput. (I+N). "Coverage Display Types" on page 694. and bearer calculations. "Traffic and Pilot C/(I+N) and Bearer Calculation (DL)" on page 727. Then. From the C/(I+N). it determines the value of the selected quality indicator from the quality graphs defined in the WiMAX equipment of the selected terminal. Coverage consists of several independent layers that can be displayed and hidden on the map. It is possible to display the Coverage by C/(I+N) Level (UL) coverage prediction with colours depending on the following display options: • • • • • C/(I+N) Level (UL) (dB) Total Noise (I+N) (UL) (dBm) Allocated Bandwidth (UL) (No. Atoll determines the best bearer available on each pixel. the coverage area of each cell comprises the pixels where the cell is the best server..3. It is possible to display the Coverage by C/(I+N) Level (DL) coverage prediction with colours depending on the following display options: • • • • • Preamble C/(I+N) Level (DL) (dB) Preamble Total Noise (I+N) (DL) (dBm) Traffic C/(I+N) Level (DL) (dB) Traffic Total Noise (I+N) (DL) (dBm) Pilot C/(I+N) Level (DL) (dB) It is possible to display the Service Area Analysis (DL) coverage prediction with colours depending on the following display options: • • • Bearer (DL) Modulation (DL): Modulation used by the bearer Service It is possible to display the Coverage by Throughput (DL) coverage prediction with colours depending on the following display options: • • • • • • • • • Peak MAC Channel Throughput (DL) (kbps) Effective MAC Channel Throughput (DL) (kbps) Application Channel Throughput (DL) (kbps) Peak MAC Cell Capacity (DL) (kbps) Effective MAC Cell Capacity (DL) (kbps) Application Cell Capacity (DL) (kbps) Peak MAC Throughput per User (DL) (kbps) Effective MAC Throughput per User (DL) (kbps) Application Throughput per User (DL) (kbps) It is possible to display the Coverage by Quality Indicator (DL) coverage prediction with colours depending on the following display options: • Quality indicators available in the document (Quality Indicators table): Atoll calculates the downlink traffic C/(I+N) levels received from the best serving cells at each pixel of their coverage areas. of Subchannels) C/(I+N) Level for 1 Subchannel (UL) (dB) Transmission Power (UL) (dBm) It is possible to display the Service Area Analysis (UL) coverage prediction with colours depending on the following display options: • • • Bearer (UL) Modulation (UL): Modulation used by the bearer Service It is possible to display the Coverage by Throughput (UL) coverage prediction with colours depending on the following display options: • • • • • • • • 694 Peak MAC Channel Throughput (UL) (kbps) Effective MAC Channel Throughput (UL) (kbps) Application Channel Throughput (UL) (kbps) Peak MAC Cell Capacity (UL) (kbps) Effective MAC Cell Capacity (UL) (kbps) Application Cell Capacity (UL) (kbps) Peak MAC Allocated Bandwidth Throughput (UL) (kbps) Effective MAC Allocated Bandwidth Throughput (UL) (kbps) .e. Best server for each pixel is calculated as explained in "Best Server Determination" on page 713. Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display type parameter is greater than or equal to the defined thresholds values.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Coverage Area Determination These coverage predictions are all best server coverage predictions.Atoll 3. for the calculated C/(I+N) and bearer. i. It is possible to display the coverage predictions with colours per cell or: • • Number of interferers Number of interferers per cell 10. and L Body are not considered in the calculations.Atoll 3.3.4 Cell Identifier Collision Zones Coverage Prediction The Cell Identifier Collision Zones coverage prediction is based on the received preamble signal levels. for the calculated C/(I+N) and bearer. see "Preamble Signal Level Calculation" on page 707 For more information on coverage area determination and available display options. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. Atoll considers pixels where the received preamble signal level from TXi(ic) is either the highest or within a 2 dB margin from the highest.3. it determines the value of the selected quality indicator from the quality graphs defined in the WiMAX equipment of the best serving cell. The Best function considers the highest value from a list of values. Coverage Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine coverage areas to display.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 • • • • Application Allocated Bandwidth Throughput (UL) (kbps) Peak MAC Throughput per User (UL) (kbps) Effective MAC Throughput per User (UL) (kbps) Application Throughput per User (UL) (kbps) It is possible to display the Coverage by Quality Indicator (UL) coverage prediction with colours depending on the following display options: • Quality indicators available in the document (Quality Indicators table): Atoll calculates the uplink traffic C/(I+N) levels received at the best serving cells from each pixel of their coverage areas. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. 695 .2. Coverage consists of several independent layers that can be displayed and hidden on the map. Atoll considers pixels where the received preamble signal level from TXi(ic) is 2 dB higher than the received preamble signal levels from the cells which are 2nd best servers. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes.3. see: • • "Coverage Area Determination" on page 695. Atoll calculates the following parameters for each subscriber in the list whose Lock Status is set to None. Atoll calculates the received preamble signal level then Atoll determines the selected display parameter on each pixel inside the cell’s calculation area. From the C/(I+N). The coverage area of each cell TXi(ic) corresponds to the pixels where: TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX j  jc  MinimumThreshold  C Preamble  or L Total or L Path   MaximumThreshold AND C Preamble  Best  C Preamble – M     ji Where M is the specified margin (dB). • • • If M = 0 dB. L Ant . 10.G M i M i M i . Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display parameter is greater than or equal to the defined thresholds values. Atoll considers pixels where the received preamble signal level from TXi(ic) is the highest. If M = 2 dB.3 Calculations on Subscriber Lists When calculations are performed on a list of subscribers by running the Automatic Server Allocation. "Coverage Display Types" on page 695. see "Path Loss Calculation Prerequisites" on page 57 for more information). L M i . Atoll calculates the path loss again for the subscriber locations and heights because the subscriber heights can be different from the default receiver height used for calculating the path loss matrices. For more information on preamble signal level calculations. • Serving Base Station and Reference Cell as described in "Best Server Determination" on page 713. Then. If M = -2 dB. Atoll determines the best bearer available on each pixel. It is possible to determine the coverage area based on the best signal level. To have the same total number of users in each simulation of a group. Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data. Atoll performs two random trials. "Channel Throughput. • Scheduling and Radio Resource Management as explained under "Simulation Process" on page 699. Fixed subscribers listed in subscriber lists have a user profile assigned to each of them. 10.ini file: [Simulation] RandomTotalUsers=0 10. a second random trial is performed to obtain their geographical locations weighted according to the clutter classes. "Preamble C/(I+N) Calculation" on page 712. User profile traffic maps: Each polygon or line of the map is assigned a density of users with a given user profile and mobility type. "Traffic and Pilot C/(I+N) and Bearer Calculation (DL)" on page 727. Cell Capacity.4. The number of users of each user profile is calculated from the surface area (SEnv) of each environment class map (or each polygon) and the user profile density (DUP).4 Monte Carlo Simulations The simulation process is divided into two steps.Atoll 3. N Users = S Env  D UP 696 . number of users of a user profile per km². • Generating a realistic user distribution as explained in "User Distribution" on page 696. This may lead to slight variations in the total numbers of users in different simulations.3. and Per-User Throughput Calculation" on page 743.1 Simulations Based on User Profile Traffic Maps and Subscriber Lists User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of user profiles with an associated mobility type and a given density. • • Azimuth (  ): Angle with respect to the north for pointing the subscriber terminal antenna towards its serving base station. If the map is composed of points. Atoll calculates the remaining parameters for each subscriber in the list that has a serving base station assigned. add the following lines in the Atoll.1.e. i. "Simulations Based on Sector Traffic Maps" on page 698. "Traffic Signal Level Calculation (UL)" on page 729. 10. • • "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 696. "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737. using the properties of the default terminal and service. Mechanical Downtilt (  ): Angle with respect to the horizontal for pointing the subscriber terminal antenna towards its serving base station.4..3. The first random trial generates the number of users and their activity status as explained in the following sections depending on the type of traffic input. "Traffic and Pilot Signal Level Calculation (DL)" on page 715. Allocated Bandwidth Throughput. User profiles model the behaviour of the different user categories. "Permutation Zone Selection" on page 714. Once all the user characteristics have been determined. "Noise Rise Calculation (UL)" on page 733.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Atoll calculates the following parameters for each subscriber in the list that has a serving base station assigned and whose Lock Status is set to None or Server. The resulting user distribution complies with the traffic database and maps selected when creating simulations. Atoll determines the total number of users attempting connection in each simulation based on the Poisson distribution. each point is assigned a number of users with given user profile and mobility type.3. Each user profile contains a list of services and parameters describing how these services are accessed by the user.3. For more information. and whether they are indoor or outdoor according to the percentage of indoor users per clutter class.1 User Distribution During each simulation. see: • • • • • • • • • "Preamble Signal Level Calculation" on page 707. inactive on both links. i. or active on DL only. Calculation of activity probabilities: UL DL Probability of being inactive: p Inactive =  1 – f Act    1 – f Act  UL UL DL Probability of being active in the uplink: p Active = f Act   1 – f Act  DL DL UL Probability of being active in the downlink: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active in the uplink and downlink both: p Active = f Act  f Act Calculation of number of users per activity status: Number of inactive users: n v – Inactive = n v  p Inactive UL UL Number of users active in the uplink: n v – Active = n v  p Active DL DL Number of users active in the downlink: n v – Active = n v  p Active UL + DL UL + DL Number of users active in the uplink and downlink both: n v – Active = n v  p Active Therefore.Atoll 3.. DL N Session  V  8 N Session  V  8 DL = -----------------------------------------. The average number of data sessions per hour N Session . • The average duration of a call (seconds) D Call . i.. Atoll calculates the probability for a user being active at a given instant in the uplink and in the downlink according to the service usage characteristics described in the user profiles. The average number of calls per hour N Call .3. Voice Service (v) User profile parameters for voice type services are: • • The user terminal equipment used for the service (from the Terminals table). the average duration of each voice call. for the service d. a user can be either active on both links. the number of voice calls or data sessions.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 • In case of user profile traffic maps composed of lines. or the volume of the data transfer in the uplink and the downlink in each data session. • The average data volume (in kBytes) transferred in the downlink V • The average throughputs in the downlink Calculation of activity probabilities: f UL DL TP Average DL and the uplink UL and the uplink V UL TP Average UL during a session.e. f Act and f Act . N Call  D Call Calculation of the service usage duration per hour ( p 0 : probability of an active call): p 0 = ---------------------------3600 Calculation of the number of users trying to access the service v ( n v ): n v = N Users  p 0 The activity status of each user depends on the activity periods during the call.and f = -----------------------------------------UL DL TP Average  3600 TP Average  3600 UL DL Probability of being inactive: p Inactive =  1 – f    1 – f  697 . the number of users of each user profile is calculated from the line length (L) and the user profile density (DUP) (users per km): N Users = L  D UP • The number of users is a direct input when a user profile traffic map is composed of points. Data Service (d) User profile parameters for data type services are: • • The user terminal equipment used for the service (from the Terminals table).e. the uplink and downlink activity UL DL factors defined for the voice type service v. active on UL only. 3. DL values. For each transmitter TXi and each service s.2 Simulations Based on Sector Traffic Maps Sector traffic maps per sector are also referred to as live traffic maps. The activity status of each user depends on the activity periods during the call. Atoll considers both active and inactive users. • Sector Traffic Maps (Throughputs) Atoll calculates the number of active users of each service s on UL and DL in the coverage area of TXi as follows: N UL UL DL TP Cell TP Cell DL = ---------------------. i. active on UL.. i.4. But if you check each simulation. • Sector Traffic Maps (# Active Users) UL Atoll directly uses the defined N and N coverage area using the service s. Either throughput demands per service or the number of active users per service are assigned to the coverage areas of each transmitter. in the downlink ( n Active ). it is necessary to DL UL + DL ). Live traffic data from the OMC is spread over the best server coverage areas of the transmitters included in the traffic map. As for the other types of traffic maps. active on DL and active on UL and DL users.. and TP Average is the average downlink requested throughput of the service s. f Act and f Act . the average number of users per service and average numbers of inactive. TP Average is the average uplink requested throughput of the service s. the uplink and downlink activity UL DL factors defined for the service. will correspond to calculated distributions. The service and the activity status of each user are randomly drawn in each simulation.Atoll 3. respectively.3. Calculation of activity probabilities: 698 . Therefore. 10. TP Cell is the total downlink throughput demand defined in the map for any service s for the coverage UL DL area of the transmitter. Therefore. and both ( n Active ).e.and N = ---------------------UL DL TP Average TP Average UL Where TP Cell is the total uplink throughput demand defined in the map for any service s for the coverage area of the DL transmitter. if you calculate several simulations at once.e. the user distribution between services as well as the activity status distribution between users can be different in each of them.1.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 UL Probability of being active in the uplink: p Active = f DL UL DL  1 – f  Probability of being active in the downlink: p Active = f DL UL  1 – f  UL + DL Probability of being active in the uplink and downlink both: p Active = f UL f DL Calculation of number of users: Number of inactive users: n d – Inactive = N Users  p Inactive UL UL Number of users active in the uplink: n d – Active = N Users  p Active DL DL Number of users active in the downlink: n d – Active = N Users  p Active UL + DL UL + DL Number of users active in the uplink and downlink both: n d – Active = N Users  p Active Calculation of the number of active users trying to access the service d (nd): UL UL + DL DL n d = n d – Active + n d – Active + n d – Active The user distribution per service and the activity status distribution between the users are average distributions. Atoll calculates the probability for a user being active in the uplink and in the downlink as follows: Users active in the uplink and downlink both are included in the N UL UL accurately determine the number of active users in the uplink ( n Active and N DL values. the number of active users on UL and DL in the transmitter At any given instant. Generates mobiles according to the input traffic data as explained in "User Distribution" on page 696. For each simulation. The steps of this algorithm are listed below. Calculation of number of users per activity status: UL UL + DL DL UL + DL  N  p Active N  p Active  UL + DL Number of users active in the uplink and downlink both: n Active = Min  -------------------------------------- -------------------------------------- or UL UL + DL DL + DL  p Active + p Active p Active + p UL Active  UL + DL simply. active on DL and active on UL and DL users correspond to the calculated distribution. the simulation process. The number of WiMAX frames in 1 second depends on the selected frame duration. the activity status of each user is randomly drawn. But if you check each simulation.2 Simulation Process WiMAX cells include intelligent schedulers and radio resource management features for regulating network traffic loads. Therefore. 699 . and P Idle – Pilot ) are set to the values defined by the user. and AU DL ) are set to their current values in the Cells table. in each simulation. Determines the best servers for all the mobiles generated for the simulation as explained in "Best Server Determination" on page 713. 1. The simulation process can be summed up into the following iterative steps. P Pilot . 10. 2. the activity status distribution between users can be different in each of them. n = n Active + n Active + n Active Calculation of the number of inactive users attempting to access the service: nv .0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 UL DL Probability of being inactive: p Inactive =  1 – f Act    1 – f Act  UL UL DL Probability of being active in the uplink: p Active = f Act   1 – f Act  DL DL UL Probability of being active in the downlink: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active in the uplink and downlink both: p Active = f Act  f Act Calculation of the number of active users trying to access the service: We have: N UL UL + DL UL =  p Active + p Active   n and N DL UL + DL DL =  p Active + p Active   n Where.4. and satisfying the QoS demands of the users. Sets initial values for the following parameters: • TX i  ic  TX i  ic  TX i  ic  TX i  ic  Cell transmission powers and reductions ( P Preamble . SU DL TX i  ic  . NR UL – Seg .Atoll 3. P Traffic . average numbers of inactive. optimising spectral efficiency. active on UL. 3.3. n is the total number of active users in the transmitter coverage area using the service. p Inactive Number of inactive users: n Inactive = --------------------------1 – p Inactive The activity status distribution between users is an average distribution. Each Monte Carlo simulation in the Atoll WiMAX BWA module is a snap-shot of the network with resource allocation carried out over a duration of 1 second. D Frame . NRUL TX i  ic  TX i  ic  .3. n Active = Min  N UL DL  f Act N DL UL  f Act  UL Number of users active in the uplink: n Active = N UL DL Number of users active in the downlink: n Active = N UL UL + DL – n Active DL UL + DL – n Active UL + DL DL And. • Cell loads ( TL DL TX i  ic  TX i  ic  . Mi • Mobile transmission power is set to the maximum mobile power ( P Max ). In fact. if you calculate several simulations at once. TL UL TX i  ic  . Determines the channel throughputs at the mobile as explained in "Channel Throughput. 9. and noise rise values of all the cells according to the resources in use and the total resources as follows: Calculation of Traffic Loads: Atoll calculates the traffic loads for all the cells TXi(ic). Calculates the user throughputs after allocating resources to each mobile as explained in "User Throughput Calculation" on page 757. 6. the simulation process. Cell Capacity.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Figure 10. Allocated Bandwidth Throughput. 7. 10. Updates the traffic loads. TX i  ic  TL DL = Mi Mi 700 TX i  ic   RDL and TLUL = Mi  RUL Mi .3. and Per-User Throughput Calculation" on page 743.1: WiMAX Simulation Algorithm For each iteration k. Determines the downlink and uplink traffic C/(I+N) and bearers for each of these mobiles as explained in "Traffic and Pilot C/(I+N) and Bearer Calculation (DL)" on page 727 and "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737 respectively. 5. Performs radio resource management and scheduling to determine the amount of resources to allocate to each mobile according to the QoS and throughput demands of each mobile using the selected scheduler as explained in "Scheduling and Radio Resource Allocation" on page 748. Determines the permutation zone assigned to each mobile as explained in "Permutation Zone Selection" on page 714. Determines the mobiles which are within the service areas of their best serving cells as explained in "Service Area Calculation" on page 714. 4. 8.Atoll 3. TL UL M  = MU – MIMO i RC UL MU – MIMO M i Calculation of Uplink Noise Rise: For each victim cell TXi(ic). MU – MIMO Mi 11. the uplink noise rise is calculated and updated by considering each interfering mobile Mj as explained in "Noise Rise Calculation (UL)" on page 733.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 TX  ic  i For uplink MU-MIMO.3. Calculation of Downlink AAS Usage: Atoll calculates the downlink AAS usages for all the cells as follows:  AAS Mi TX i  ic  AAS = ------------------------------TX i  ic  TL DL AU DL Where Mi R DL Mi  M i R DL AAS is the sum of the percentages of the downlink cell resources allocated to mobiles served by the AAS smart antennas. and can be written as follows: TX i  ic  TL DL k = TX  ic  i Max  TL DL All TX  ic  i TX i  ic  k – TL DL  k – 1 701 . Calculation of Uplink MU-MIMO Gain: Atoll calculates the uplink MU-MIMO gain for all the cells as follows: MU – MIMO Mi  TX i  ic  G MU – MIMO = R UL MU – MIMO Mi ------------------------------------------------------------MU – MIMO Mi RC UL  M MU – MIMO i MU – MIMO Mi  Where R UL is the sum of the percentages of the uplink cell resources allocated to MU-MIMO MU – MIMO Mi  mobiles and MU – MIMO Mi RC UL is the sum of the real resource consumption of MU-MIMO mobiles. The convergence criteria are evaluated at the end of each iteration k. Performs the convergence test to see whether the differences between the current and the new loads are within the convergence thresholds.Atoll 3. Calculation of Downlink Segmentation Usage: Atoll calculates the segmentation usages for all the cells as follows: Mi  M TX i  ic  i PZ Mi R DL Mi PZ DL = Seg = Seg DL = -----------------------------------------------------------TX i  ic  TL DL SU DL M  Where Mi i R DL Mi PZ DL = Seg is the sum of the percentages of the downlink cell resources allocated to mobiles M i PZ DL = Seg served by the downlink segmented permutation zone. Therefore. UL. the start and end frequencies of all the channels may not exactly coincide. if: TX  ic  i TL DL TX  ic  i k  TL DL TX  ic  i Req OR TL UL TX  ic  i k  TL UL TX  ic  i Req OR NR UL TX  ic  i k  NR UL Req 12. Repeats the above steps (from step 3. calculation of coverage predictions. In addition to the above parameters.1 and k if: TX i  ic  TL DL TX i  ic  k  TL DL TX i  ic  Req AND TL UL TX i  ic  k  TL UL TX i  ic  Req AND NR UL TX i  ic  k  NR UL Req No convergence: Simulation has not converged even after the last iteration.).3. Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 8. Simulation Results At the end of the simulation process.4. Mobiles can be rejected due to: • • • • • No Coverage: If the mobile does not have any best serving cell (step 3. 10.4 Calculation Details The following sections describe all the calculation algorithms used in point analysis. Atoll stops the simulation in the following cases.e.).) Resource Saturation: If all the cell resources are used up before allocation to the mobile or if. i. Connected DL: If a mobile active in DL is allocated resources in DL. Convergence: Simulation has converged between iteration k . DL.) for the iteration k+1 using the new calculated loads as the current loads. k = Max Number of Iterations defined when creating the simulation. Connected DL+UL: If a mobile active in DL+UL is allocated resources in DL+UL. This condition is only verified if the simulation was created with the Backhaul capacity check box selected (step 8.and Adjacent Channel Overlaps Calculation A WiMAX network can consist of cells that use different channel bandwidths. the main results obtained are: • • • • • • • • • Downlink traffic loads Uplink traffic loads Uplink noise rise received at the main antenna Ssegmented zone uplink noise rise received at the main antenna Angular distributions of downlink traffic power density for cells with smart antennas Angular distributions of uplink noise rise for cells with smart antennas Downlink AAS usage Downlink segmentation usage Uplink MU-MIMO capacity gain These results can be used as input for C/(I+N)-based coverage predictions.) or if the mobile is not within the service area of its best server (step 4. and NR UL Req are the simulation convergence thresholds defined when creating the simulation. i. Channel bandwidths of cells can overlap each other with different ratios. or DL+UL. calculations on subscriber lists.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks TX  ic  i TL UL = k k = TX i  ic  Req TX  ic  i k – TL UL TX  ic  i Max  NR UL All TX  ic  i TX i  ic  If TL DL TX  ic  i Max  TL UL All TX  ic  i TX  ic  i NR UL ©Forsk 2015 ..Atoll 3.) can be: • • • Connected UL: If a mobile active in UL is allocated resources in UL.e.1 Co. the simulations also list the connection status of each mobile. and Monte Carlo simulations. the minimum uplink throughput demand is higher than the uplink allocated bandwidth throughput (step 8.) Backhaul Saturation: If allocating resources to a mobile makes the effective MAC aggregate site throughputs exceed the maximum backhaul throughputs defined for the site.. No Service: If the mobile is not able to access a bearer in the direction of its activity (step 6. 702 . for a user active in uplink.) Connected mobiles (step 8. 10. TL UL  k – 1 TX  ic  i k – NR UL  k – 1 TX i  ic  Req . 10. TX i  ic  TX j  jc  • W Channel and W Channel : Bandwidths of the channels assigned to cells TXi(ic) and TXj(jc). "Total Overlap Ratio Calculation" on page 706. the studied cell can be considered a victim of interference received from the other cells that might be interfering the studied cell. or the uplink or the downlink start frequency of an FDD frequency band ( F Start – FB – FDD – UL or F Start – FB – FDD – DL ).and adjacent channel overlaps are calculated between the channels used by any studied cell TXi(ic) and any other cell TXj(jc) of the network. Once the start and end frequencies are known for the studied and other cells. • ICS FB • CN FB TX i  ic  TX j  jc  and ICS FB TX i  ic  : Inter-channel spacing of the frequency bands assigned to cells TXi(ic) and TXj(jc). the channel number you assign to a cell is considered for uplink and downlink both. For FDD networks.4.e.3. Atoll considers that the same channel number is assigned to a cell in the downlink and uplink. F Start – FB can be the start frequency of a TDD frequency band ( F Start – FB – TDD ). i. 703 . and adjacent channel interference on the adjacent channel bandwidths. TX i  ic  If the studied cell is assigned a channel number N Channel .and adjacent overlaps and the total overlap ratio are calculated as respectively explained in: • • • "Co-Channel Overlap Calculation" on page 704..e.and adjacent channel overlaps between two channels. "Adjacent Channel Overlap Calculation" on page 705. corresponding to N Channel – 1 and TX i  ic  N Channel + 1 .2: Co-Channel and Adjacent Channel Overlaps The following sections describe how the co.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Figure 10.. TX j  jc  and CN FB : Channel number step of the frequency bands assigned to cells TXi(ic) and TXj(jc).1 Conversion From Channel Numbers to Start and End Frequencies Input • TX i  ic  TX j  jc  F Start – FB and F Start – FB : Start frequency of the frequency band assigned to the cells TXi(ic) and TXj(jc). it is necessary to calculate the start and end frequencies of both channels (explained in "Conversion From Channel Numbers to Start and End Frequencies" on page 703). the co. In order to calculate the co. i. • • First – TX  ic  i N Channel TX i  ic  First – TX  jc  j and N Channel : First channel numbers the frequency band assigned to the cells TXi(ic) and TXj(jc).Atoll 3.1. In terms of interference calculation. TX j  jc  N Channel and N Channel : Channel numbers assigned to cells TXi(ic) and TXj(jc). it receives co-channel interference on the channel bandwidth of TX i  ic  TX i  ic  N Channel . 4. • TX i  ic  TX j  jc  F End and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 703. Atoll calculates the bandwidth of the co-channel overlap as follows: TX i  ic  – TX j  jc  W CCO TX  jc  TX  ic  TX  jc  TX  ic  j i j i = Min  F End  F End  – Max  F Start  F Start      The co-channel overlap ratio is given by: TX i  ic  – TX j  jc  r CCO TX  ic  – TX  jc  i j W CCO = ---------------------------------TX i  ic  W Channel Output • 704 TX i  ic  – TX j  jc  r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc).Atoll 3. • F End TX i  ic  TX j  jc  and F End : End frequencies for the cells TXi(ic) and TXj(jc). • TX  ic  i W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic). Co-channel overlap exists if: TX i  ic  TX j  jc  F Start  F End TX i  ic  AND F End TX j  jc   F Start Otherwise there is no co-channel overlap. 10. Calculations Atoll first verifies that co-channel overlap exists between the cells TXi(ic) and TXj(jc).0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Calculations Channel numbers are converted into start and end frequencies as follows: For cell TXi(ic): TX  ic  i F Start TX i  ic  F End = TX  ic  i F Start – FB TX  ic  TX i  ic   TXi  ic  – N First – TX i  ic  Channel Channel   N ------------------------------------------------------- TX i  ic       CN FB TX  ic  i i +  W Channel + ICS FB  TX i  ic  = F Start + W Channel For cell TXj(jc): TX j  jc  F Start TX j  jc  F End TX j  jc  TX  jc  TX j  jc   N TXj  jc  – N First – TX j  jc  Channel Channel    ------------------------------------------------------- TX  ic     i   CN FB TX  jc  j j = F Start – FB +  W Channel + ICS FB  TX j  jc  = F Start + W Channel Output TX  ic  i TX  jc  j • F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc).2 Co-Channel Overlap Calculation Input • TX i  ic  TX j  jc  F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 703. .3.1. 4.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 10.1. • F End TX i  ic  TX j  jc  and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 703. Atoll determines the adjacent channel overlap ratio as follows: Bandwidth of the lower-frequency adjacent channel overlap: TX i  ic  – TX j  jc  W ACO L TX  jc  TX  ic  TX  jc  TX  ic  TX  ic  j i j i i = Min  F End  F Start  – Max  F Start  F Start – W Channel The lower-frequency adjacent channel overlap ratio is given by: TX i  ic  – TX j  jc  W ACO L = ---------------------------------TX i  ic  W Channel TX i  ic  – TX j  jc  r ACO L Bandwidth of the higher-frequency adjacent channel overlap: TX i  ic  – TX j  jc  W ACO H TX j  jc  TX i  ic  = Min  F End  F End  TX  ic  TX  jc  TX  ic  i j i + W Channel – Max  F Start  F End     The higher-frequency adjacent channel overlap ratio is given by: TX i  ic  – TX j  jc  W ACO H = ---------------------------------TX  ic  i W Channel TX i  ic  – TX j  jc  r ACO H The adjacent channel overlap ratio is given by: TX i  ic  – TX j  jc  r ACO TX i  ic  – TX j  jc  = r ACO L TX i  ic  – TX j  jc  + r ACO H Output • TX  ic  – TX  jc  i j r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).3. Adjacent channel overlap exists on the lower-frequency adjacent channel if: TX i  ic  TX i  ic  TX j  jc  F Start – W Channel  F End TX i  ic  TX j  jc  AND F Start  F Start Adjacent channel overlap exists on the higher-frequency adjacent channel if: TX i  ic  F End TX j  jc   F End TX i  ic  AND F End TX i  ic  TX j  jc  + W Channel  F Start Otherwise there is no adjacent channel overlap.Atoll 3. Calculations Atoll first verifies that adjacent channel overlaps exist between (the lower-frequency and the higher-frequency adjacent channels of) the cells TXi(ic) and TXj(jc).4. However. Atoll models the co-existence of FDD and TDD cells in a network by determining the FDD – TDD overlap ratio as follows: 705 . coexisting FDD and TDD cells may also exist and interfere each other.3 Adjacent Channel Overlap Calculation Input TX  ic  i TX  jc  j • F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 703.4 FDD – TDD Overlap Ratio Calculation There are many different interference scenarios possible in a WiMAX network depending on the type of duplexing used by the cells of the network. • TX i  ic  W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic).1. The most common interference scenarios are FDD-only and TDD-only interferences. 10. Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Input • TDD r DL – Frame : Downlink subframe ratio defined in the global network settings. Calculations The FDD – TDD overlap ratio is calculated as follows depending on the frequency bands assigned to the cells TXi(ic) and TXj(jc): Frequency Band TX i  ic  – TX j  jc  Overlap Ratio r FDD – TDD TXi(ic) TXj(jc) TDD TDD 1 TDD FDD 1 FDD TDD r DL – Frame ----------------------100 FDD FDD 1 TDD Output • TX  ic  – TX  jc  i j r FDD – TDD : FDD – TDD overlap ratio between the cells TXi(ic) and TXj(jc). 10.4.1.5 Total Overlap Ratio Calculation Input TX i  ic  – TX j  jc  • r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co-Channel Overlap Calculation" on page 704. • r ACO TX i  ic  – TX j  jc  : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Adjacent Channel Overlap Calculation" on page 705. • TX i  ic  – TX j  jc  r FDD – TDD : FDD – TDD overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "FDD – TDD Overlap Ratio Calculation" on page 705. TX i  ic  • f ACS – FB : Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic). • W Channel and W Channel : Bandwidths of the channels assigned to the cells TXi(ic) and TXj(jc). TX  ic  i TX  jc  j Calculations The total overlap ratio is: TX i  ic  – TX j  jc  rO        =         TX  ic  i   –f ACS – FB-  TX  ic  – TX  jc  TX  ic  – TX  jc  --------------------------TX  ic  – TX j  jc  j i j 10 r i  r i + r ACO  10 FDD – TDD  CCO      TX i  ic  TX j  jc  TX i  ic  TX j  jc  if W Channel  W Channel TX  ic  i   –f TX i  ic  ACS – FB  TX  ic  – TX  jc  TX  ic  – TX  jc  ---------------------------- TX  ic  – TX j  jc  W i j i j 10 Channel r  r i --------------------+ r  10  ACO FDD – TDD TX j  jc   CCO    W Channel   if W Channel  W Channel TX i  ic  W Channel The multiplicative factor --------------------is used to normalise the transmission power of the interfering cell TXj(jc). This means that TX j  jc  W Channel TX j  jc  TX j  jc  if the interfering cell transmits at X dBm over a bandwidth of W Channel , and it interferes over a bandwidth less than W Channel , 706 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 TX  ic  i W Channel the interference from this cell should not be considered at X dBm but less than that. The factor --------------------converts X dBm over TX  jc  j W Channel TX  jc  j TX  jc  j W Channel to Y dBm (which is less than X dBm) over less than W Channel . Output • TX i  ic  – TX j  jc  rO : Total co- and adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc). 10.4.2 Preamble Signal Level and Quality Calculations These calculations include the calculation of the received preamble signal level, and the noise and interference on the preamble. The following sections also describe how the received preamble signal level, the noise and interference, C/N, and C/(I+N) ratios are calculated in Atoll: • • • • • "Preamble Signal Level Calculation" on page 707. "Preamble Noise Calculation" on page 708. "Preamble C/N Calculation" on page 712. "Preamble Interference Calculation" on page 710. "Preamble C/(I+N) Calculation" on page 712. 10.4.2.1 Preamble Signal Level Calculation Input TX i  ic  • P Preamble : Preamble transmission power of the cell TXi(ic). • E SA : Number of antenna elements defined for the smart antenna equipment used by the transmitter TXi. • G SA • Div G SA • TX i Combining G TX i TX i : Smart power combining gain offset defined per clutter class. : Smart antenna diversity gain (for cross-polarised smart antennas) defined per clutter class. : Transmitter antenna gain for the antenna used by the transmitter TXi. i = L Total – DL ). L • L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model. • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi. • M Shadowing – Model : Shadowing margin based on the model standard deviation. TX : Total transmitter losses for the transmitter TXi ( L TX • i In coverage predictions, shadowing margins are taken into account when the option "Shadowing taken into account" is selected. • L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected. • L Mi Mi : Receiver terminal losses for the pixel, subscriber, or mobile Mi. • G : Receiver terminal’s antenna gain for the pixel, subscriber, or mobile Mi. • L Ant : Receiver terminal’s antenna attenuation calculated for the pixel, subscriber, or mobile Mi. Mi Mi For calculating the useful signal level from the best serving cell, L Ant is determined in the direction (H,V) = (0,0) from Mi the antenna patterns of the antenna used by Mi. For calculating the interfering signal level from any interferer, L Ant is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi, while the antenna is pointed towards Mi’s best serving cell. • Mi L Body : Body loss defined for the service used by the pixel, subscriber, or mobile Mi. 707 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks L M i , G ©Forsk 2015 M i M i M i , L Ant , and L Body are not used in the calculations performed for the point analysis tool’s profile tab and the preamble signal level based coverage predictions. Calculations The received preamble signal level (dBm) from any cell TXi(ic) is calculated for a pixel, subscriber, or mobile Mi as follows: TX i  ic  TX i  ic  C Preamble = EIRP Preamble – L Path – M Shadowing – Model – L Indoor + G Mi –L Mi Mi Mi – L Ant – L Body Where EIRP is the effective isotropic radiated power of the cell calculated as follows: TX  ic  i TX  ic  i • Without smart antennas: EIRP Preamble = P Preamble + G • With smart antennas: EIRP Preamble = P Preamble + G TX i  ic  TX i  ic  TX i TX i –L –L TX i TX i TX i + 10  Log  E SA  + G SA   Combining Div + G SA L Path is the path loss (dB) calculated as follows: TX i L Path = L Model + L Ant Furthermore, the total losses between the cell and the pixel, subscriber, or mobile Mi can be calculated as follows: L Total = L Path + L TX i + L Indoor + M Shadowing – Model – G TX i +L M i –G M i M i M i + L Ant + L Body If you wish to exclude the the energy corresponding to the cyclic prefix part of the total symbol duration from the useful signal level, you must add the following lines in the Atoll.ini file: [WiMAX] ExcludeCPFromUsefulPower = 1 TX i  ic  When this option is active, the cyclic prefix energy is excluded from C Preamble . In other TX  ic  i words, the factor 10  Log  1 – r CP TX  ic   is added to C i Preamble .  Independant of the option, interference levels are calculated for the total symbol durations, i.e., the energy of the useful symbol duration and the cyclic prefix energy. Output TX i  ic  • C Preamble : Received preamble signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi. • L Path : Path loss between the cell TXi(ic) and the pixel, subscriber, or mobile Mi. • L Total : Total losses between the cell TXi(ic) and the pixel, subscriber, or mobile Mi. 10.4.2.2 Preamble Noise Calculation For determining the preamble C/N and C/(I+N), Atoll calculates the preamble noise over the bandwidth used by the cell. The used bandwidth depends on the number of subcarriers used by the preamble.The number of subcarriers used by the preamble can be different from the number of subcarriers used by the permutation zones. The preamble noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on the temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the used bandwidth. Input 708 • • K: Boltzmann’s constant. T: Temperature in Kelvin. • N SCa – Preamble : Number of subcarriers used by the preamble defined for the frame configuration of the cell TXi(ic). • N SCa – Total : Total number of subcarriers defined for the frame configuration of the cell TXi(ic). TX i  ic  TX i  ic  Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 TX  ic  i • F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on page 740. • nf M i : Noise figure of the terminal used for calculations by the pixel, subscriber, or mobile Mi. Calculations The power spectral density of thermal noise is calculated as follows: n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz The thermal noise over the preamble for a cell is calculated as: TX i  ic  n 0 – Preamble TX i  ic   TX  ic  N SCa – Preamble Preamble i  -  f Segment = n 0 + 10  Log  F Sampling  --------------------------------TX i  ic    N SCa – Total  Effect of Segmentation: The preamble is segmented and one of the three preamble carrier sets is used for transmission. Each preamble carrier set uses 1/3rd of the total number of preamble subcarriers. The power transmitted over the preamble has higher spectral density than the power transmitted over the entire channel bandwidth. This power concentration due to segmentation on the C/N and C/(I+N) results in an increase in the coverage footprint of the preamble. Hence, the Preamble thermal noise at the pixel, subscriber, or mobile Mi covered by the preamble is reduced by a factor of f Segment = 1 --- . 3 The following table shows the different types of subcarriers and their numbers for preamble transmission in WiMAX. N SCa – Total 128 512 1024 2048 Guard Subcarriers DC Subcarrier N SCa – Preamble All 1 (54) 107 1 0 1 (54) 35 0.3271 None 36 0.3364 2 None 36 0.3364 All 1 (214) 428 1 0 None 143 0.3341 1 (214) 142 0.3318 2 None 143 0.3341 All 1 (426) 851 1 0 1 (426) 283 0.3325 None 284 0.3337 2 None 284 0.3337 All 1 (852) 1703 1 0 1 (852) 567 0.3329 None 568 0.3335 None 568 0.3335 Segment Left 10 1 42 1 86 1 172 1 Right 10 41 86 172 2 Total 20 83 172 344 Preamble f Segment The preamble noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel, subscriber, or mobile Mi. TX i  ic  TX i  ic  n Preamble = n 0 – Preamble + nf Mi Output • TX i  ic  n Preamble : Preamble noise for the cell TXi(ic). 709 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 10.4.2.3 Preamble Interference Calculation The interference received by any pixel, subscriber, or mobile, served by a cell TXi(ic) from other cells TXj(jc) can be defined as the preamble signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the channels used by the cells TXi(ic) and TXj(jc) and which preamble carrier sets are used by the two cells. Input TX j  jc  • C Preamble : Preamble signal level received from an interfering cell TXj(jc) as calculated in "Preamble Signal Level Calculation" on page 707 at the pixel, subscriber, or mobile Mi covered by the cell TXi(ic). • M Shadowing – Model : Shadowing margin based on the model standard deviation. • M Shadowing – C  I : Shadowing margin based on the C/I standard deviation. In Monte Carlo simulations, interfering signal levels already include M Shadowing – Model , as explained in "Preamble Signal Level Calculation" on page 679. In coverage predictions, the ratio M Shadowing – Model – M Shadowing – C  I is applied to the interfering signals (for more information, see "Shadow Fading Model" on page 90). As the received interfering signal levels already include M Shadowing – Model , M Shadowing – C  I is added to the received interfering signal levels in order to achieve the ratio M Shadowing – Model – M Shadowing – C  I : TX  jc  j TX  jc  j C Preamble = C Preamble + M Shadowing – C  I In coverage predictions, shadowing margins are taken into account when the option "Shadowing taken into account" is selected. TX i  ic  – TX j  jc  • rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent Channel Overlaps Calculation" on page 702. • N Seg TX i  ic  TX j  jc  and N Seg : Segment numbers assigned to the cells TXi(ic) and TXj(jc) calculated from their respective TX i  ic  TX j  jc  preamble indexes ( n Preamble and n Preamble ) as follows: • Inter – Tech f IRF n Preamble N Seg 0 to 31, 96, 99, 102, 105, 108, 111 0 32 to 63, 97, 100, 103, 106, 109, 112 1 64 to 95, 98, 101, 104, 107, 110, 113 2 : Inter-technology interference reduction factor. Calculations The received preamble interference (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows: TX j  jc  TX j  jc  TX i  ic  – TX j  jc  I Preamble = C Preamble + f O TX  ic  – TX  jc  i j Where f O TX i  ic  – TX j  jc  Inter – Tech + f Seg – Preamble + I DL is the interference reduction factor due to channel overlap between the cells TXi(ic) and TXj(jc), calculated as follows: TX i  ic  – TX j  jc  fO TX i  ic  – TX j  jc  = 10  Log  r O   TX i  ic  – TX j  jc  f Seg – Preamble is the interference reduction factor due to preamble segmentation, calculated as follows: TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  f Seg – Preamble = 10  Log  p Collision  710   Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 TX  ic  – TX  jc  i j The probability of preamble subcarrier collision p Collision TX  ic  i 1 if N Seg TX  jc  j = N Seg TX  ic  i between the cells TXi(ic) and TXj(jc) is 0 if N Seg TX  jc  j  N Seg and . TX  jc  j TX  jc  j In case of smart antennas, C Preamble in I Preamble already includes the effect of the TX j number of antenna elements ( E SA ). If you wish to include the effect of the number of antennas in case of MIMO, you must add the following lines in the Atoll.ini file: [WiMAX] MultiAntennaInterference When the multi-antenna interference option is active, and TXj(jc) does not have a smart antenna equipment assigned, the interference is incremented by TX  jc  j + 10  Log  N Ant – TX .   TX j  jc  Where N Ant – TX is the number of MIMO transmission (downlink) antennas defined for the cell TXj(jc). Inter – Tech I DL is the inter-technology downlink interference from transmitters of an external network (linked document of any technology) calculated as follows: Inter – Tech I DL  = TX – External EIRP DL – L Path – L Indoor + G Mi –L Mi Mi Mi Inter – Tech – L Ant – L Body – f IRF All External TXs TX – External Where EIRP DL is the downlink EIRP of the external transmitter, L Path is the path loss from the external transmitters to the pixel, subscriber, or mobile location, L Indoor is the indoor losses taken into account when the option "Indoor coverage" is selected, L Mi is the receiver terminal losses for the pixel, subscriber, or mobile Mi, G Mi is the receiver terminal’s antenna Mi gain for the pixel, subscriber, or mobile Mi, L Ant is the receiver terminal’s antenna attenuation calculated for the pixel, Mi subscriber, or mobile Mi, and L Body is the body loss defined for the service used by the pixel, subscriber, or mobile Mi. Calculation of the Downlink Inter-technology Interference The downlink inter-technology interference is calculated as follows: Inter – Tech I DL TX k   P DL – Rec  -------------------------------------- = F  TX  ic  TX   i k  TX k  ICP DL  TX k Here P DL – Rec is the received downlink power from an interfering cell TXk belonging to another technology, and F  TX i  ic  TX k  ICPDL is the inter-technology downlink channel protection ratio for a frequency offset F between the interfered and interfering frequency channels of TXi(ic) and TXk. TX k P DL – Rec is calculated based on the EIRP from GSM cells, total power from UMTS, CDMA2000, and TD-SCDMA cells, maximum power from LTE cells, preamble power from WiMAX cells, and downlink cell power from Wi-Fi cells. Output • TX j  jc  I Preamble : Preamble interference received from any interfering cell TXj(jc) at the pixel, subscriber, or mobile Mi covered by a cell TXi(ic). • Inter – Tech I DL : Downlink inter-technology interference. 711 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 10.4.2.4 Preamble C/N Calculation Input • TX  ic  i C Preamble : Received preamble signal level from the cell TXi(ic) as calculated in "Preamble Signal Level Calculation" on page 707. TX i  ic  • n Preamble : Preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on page 708. • G Div – Preamble : Preamble diversity gain defined in the WiMAX equipment of the terminal used by the pixel, subscriber, Mi or mobile Mi. • DL G Div : Additional downlink diversity gain defined for the clutter class where the pixel, subscriber, or mobile Mi is located. Calculations The preamble C/N for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi: TX i  ic  TX i  ic  TX i  ic  Mi DL CNR Preamble = C Preamble – n Preamble + G Div – Preamble + G Div The preamble diversity gain is applied to the preamble C/N when the cell and the terminal both support any form of MIMO in downlink. The additional downlink diversity gain defined per clutter is also applied. Output • TX i  ic  CNR Preamble : Preamble C/N from the cell TXi(ic) at any pixel, subscriber, or mobile Mi. 10.4.2.5 Preamble C/(I+N) Calculation The carrier signal to interference and noise ratio is calculated in three steps. First Atoll calculates the received preamble signal level from the studied cell (as explained in "Preamble Signal Level Calculation" on page 707) at the pixel, subscriber or mobile under study. Next, Atoll calculates the interference received at the same studied pixel, subscriber, or mobile from all the interfering cells (as explained in "Preamble Interference Calculation" on page 710). Interference from each cell is weighted according to the co- and adjacent channel overlap between the studied and the interfering cells, and the probabilities of subcarrier collision. Finally, Atoll takes the ratio of the preamble signal level, and the sum of the total interference from all interfering cells and the noise (as calculated in "Preamble Noise Calculation" on page 708). The receiver terminal is always considered to be oriented towards its best server, except when the "Lock Status" is set to "Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited. In the case of NLOS between the receiver and the best server, Atoll does not try to find the direction of the strongest signal, the receiver is oriented towards the best server just as in the case of LOS. Input • TX i  ic  C Preamble : Preamble signal level received from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Preamble Signal Level Calculation" on page 707. TX i  ic  • n Preamble : Preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on page 708. • I Preamble : Preamble interference received from any cell TXj(jc) at a pixel, subscriber, or mobile Mi covered by a cell TX j  jc  TXi(ic) as calculated in "Preamble Interference Calculation" on page 710. Inter – Tech • NRDL • G Div – Preamble : Preamble diversity gain defined in the WiMAX equipment of the terminal used by the pixel, subscriber, M : Inter-technology downlink noise rise. i or mobile Mi. • DL G Div : Additional downlink diversity gain defined for the clutter class where the pixel, subscriber, or mobile Mi is located. • 712 Inter – Tech I DL : Downlink inter-technology interference as calculated in "Preamble Interference Calculation" on page 710. Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Calculations The preamble C/(I+N) for a cell TXi(ic) is calculated as follows at any pixel, subscriber, or mobile Mi: TX  ic  TX  ic  i CINR Preamble i      TX j  jc   n Preamble    IPreamble  ----------------------------- TX  ic  M i Inter – Tech DL 10 -  -------------------------- + NR Inter – Tech + G i = C Preamble –  10  Log  + I DL + 10 10 DL Div – Preamble + G Div    10      All TX j  jc             The preamble diversity gain is applied to the preamble C/(I+N) when the cell and the terminal both support any form of MIMO. The additional downlink diversity gain defined per clutter is also applied. The preamble total noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi: TX  ic  TX i  ic   I + N  Preamble i   TXj  jc   n Preamble   IPreamble ----------------------------- Inter – Tech 10 -  + NRInter – Tech  --------------------------= 10  Log  +I + 10 10 DL    10  DL  All TXj  jc          Output TX i  ic  • CINR Preamble : Preamble C/(I+N) from the cell TXi(ic) at a pixel, subscriber, or mobile Mi. •  I + N  Preamble : Preamble total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered TX i  ic  by a cell TXi(ic). 10.4.3 Best Server Determination In WiMAX, best server refers to a cell ("serving transmitter"-"reference cell" pair) from which a pixel, subscriber, or mobile Mi gets the highest preamble signal level or preamble C/(I+N). This calculation also determines whether the pixel, subscriber, or mobile Mi is within the coverage area of any transmitter or not. Input • TX i  ic  C Preamble : Preamble signal level received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Preamble Signal Level Calculation" on page 707 using the terminal and service parameters ( L M Mi , G Mi Mi , L Ant , and i L Body ) of Mi. "Preamble C/(I+N) Calculation" on page 712 • TX i  ic  CINR Preamble : Preamble C/(I+N) received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Preamble C/(I+N) Calculation" on page 712. Calculations The best server of any pixel, subscriber, or mobile Mi, BS M , is the cell from which the received preamble signal level or C/ i (I+N) is the highest among all the cells. The best server is determined as follows: BSM = TX i  ic  i TX i  ic   TX i  ic   C Preamble = Best C  All TX i  ic   Preamble  or BS M = TX i  ic  i TX i  ic  TX i  ic    CINR Preamble = Best  CINR Preamble  All TX i  ic    Here ic is the cell of the transmitter TXi with the highest preamble power. However, if more than one cell of the same transmitter covers the pixel, subscriber, or mobile, the final reference cell ic might be different from the initial cell ic (the one with the highest power) depending on the serving cell selection method: • • Random: In coverage prediction calculations and in calculations on subsriber lists, the cell of the highest priority layer is selected as the serving (reference) cell. In Monte Carlo simulations, a random cell is selected as the serving (reference) cell. Distributive: In coverage prediction calculations and in calculations on subsriber lists, the cell of the highest priority layer is selected as the serving (reference) cell. In Monte Carlo simulations, mobiles are distributed among cell layers one by one, i.e., if more than one cell layer covers a set of mobiles, the first mobile is assigned to the highest priority layer, the 2nd mobile to the second highest priority layer, and so on. 713 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 When using either the Random or the Distributive cell selection method, the reference cell once assigned to a mobile does not change during Monte Carlo simulations. Output • BS M : Best serving cell of the pixel, subscriber, or mobile Mi. i 10.4.4 Service Area Calculation In WiMAX, a pixel, subscriber, or mobile Mi can be covered by a cell (as calculated in "Best Server Determination" on page 713) but can be outside the service area. A pixel, subscriber, or mobile Mi is said to be within the service area of its best serving cell TXi(ic) if the preamble C/N from the cell at the pixel, subscriber, or mobile is greater than or equal to the preamble C/N threshold defined for the cell. Input • TX i  ic  CNR Preamble : Preamble C/N from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Preamble C/N Calculation" on page 712. • TX i  ic  T Preamble : Preamble C/N threshold defined for the cell TXi(ic). Calculations A pixel, subscriber, or mobile Mi is within the service area of its best serving cell TXi(ic) if: TX  ic  i TX  ic  i CNR Preamble  T Preamble Output • • True: If the calculation criterion is satisfied. False: Otherwise. 10.4.5 Permutation Zone Selection In order to be able to calculate the traffic C/(I+N) and the throughputs, a permutation zone is assigned to each pixel, subscriber, or mobile Mi located within the service area (as calculated in "Service Area Calculation" on page 714) of its best serving cell. The permutation zone assigned to Mi is one which covers Mi in terms of distance and preamble C/N or C/(I+N), and accepts user speeds equal to or higher than Mi’s speed selected for the calculation. A pixel, subscriber, or mobile Mi which is unable to get a permutation zone is considered to be outside the service area. Input TX i  ic  • d Max – PZ : Maximum distance covered by a permutation zone of a cell TXi(ic). • QT PZ TX  ic  i : Minimum preamble C/N or C/(I+N) required at the pixel, subscriber, or mobile Mi to connect to a permutation zone of a cell TXi(ic). TX i  ic  • Speed Max – PZ : Maximum speed supported by a permutation zone of a cell TXi(ic). • d • TX  ic  i CNR Preamble M – TX  ic  i i : Distance between the pixel, subscriber, or mobile Mi and a cell TXi(ic). : Preamble C/N from the cell TXi(ic) as calculated in "Preamble C/N Calculation" on page 712. TX i  ic  • CINR Preamble : Preamble C/(I+N) from the cell TXi(ic) as calculated in "Preamble C/(I+N) Calculation" on page 712. • Mobility  M i  : Speed of the pixel, subscriber, or mobile Mi. Calculations Mi is assigned the permutation zone with the highest priority among the permutation zones whose selection criteria Mi satisfies. Mi satisfies the selection criteria of a permutation zone if: • 714 The distance between Mi and TXi(ic) is less than or equal to the maximum distance covered by the permutation zone: Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 d • M – TX  ic  i i TX  ic  i  d Max – PZ The preamble C/N or C/(I+N) at Mi is better than or equal to the quality threshold defined for the permutation zone: TX  ic  i TX  ic  i CNR Preamble  QT PZ • TX  ic  i TX  ic  i or CINR Preamble  QT PZ The mobility of Mi is less than or equal to the maximum mobile speed supported by the permutation zone: TX i  ic  Mobility  M i   Speed Max – PZ Therefore, the permutation zones assigned to a pixel, subscriber, or mobile Mi in the downlink and uplink are: Mi PZ DL Mi PZ UL      TX  ic  i = Highest Priority  PZ DL              TX i  ic  TX i  ic      CNR   QT Preamble PZ   TX i  ic   TX i  ic     M i – TX i  ic      d Max – PZ AND AND  Mobility  M i   Speed Max – PZ  OR d        TX i  ic  TX i  ic   CINR   QT         TX  ic  i = Highest Priority  PZ UL             TX i  ic  TX i  ic      CNR   QT Preamble PZ   TX i  ic   TX i  ic     M i – TX i  ic     d d  AND AND Mobility  M  Speed  OR   Max – PZ   i Max – PZ      TX  ic  TX  ic  i  CINR i    QT   Preamble Preamble PZ PZ If more than 1 permutation zone satisfies the distance, speed, and quality threshold criteria, and all have the same priority, the permutation zone assigned to the pixel, subscriber, or mobile will be the first in the list of permutation zones (frame configuration) among these zones. Output • Mi Mi PZ DL and PZ UL : Downlink and uplink permutation zones assigned to the pixel, subscriber, or mobile Mi. 10.4.6 Traffic and Pilot Signal Level and Quality Calculations Traffic and pilot subcarriers can be transmitted with different transmission powers than the preamble power of a cell, and do not suffer the same interference and noise as the preamble. The following sections describe how traffic and pilot signal levels, noise and interference, C/N, and C/(I+N) ratios are calculated on the downlink and uplink. • • • • • • • • • • "Traffic and Pilot Signal Level Calculation (DL)" on page 715. "Traffic and Pilot Noise Calculation (DL)" on page 717. "Traffic and Pilot Interference Calculation (DL)" on page 718. "Traffic and Pilot C/N Calculation (DL)" on page 726. "Traffic and Pilot C/(I+N) and Bearer Calculation (DL)" on page 727. "Traffic Signal Level Calculation (UL)" on page 729. "Traffic Noise Calculation (UL)" on page 730. "Traffic Interference Calculation (UL)" on page 731. "Traffic C/N Calculation (UL)" on page 734. "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737. 10.4.6.1 Traffic and Pilot Signal Level Calculation (DL) Input TX i  ic  • P Preamble : Preamble transmission power of the cell TXi(ic). • P Traffic : Traffic power reduction of the cell TXi(ic). TX i  ic  715 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 TX  ic  i • P Pilot : Pilot power reduction of the cell TXi(ic). • G TX i : Transmitter antenna gain for the antenna used by the transmitter TXi. • Without smart antennas: G • With smart antennas: G G TX i TX i TX i is the transmitter antenna gain, i.e., G TX i TX i = G Ant . is the smart antenna gain in the direction of the pixel, subscriber, or mobile Mi, i.e., = G SA    . Where  is the direction in which Mi is located. For more information on the calculation of G SA    , refer to section "Beamforming Smart Antenna Models" on page 43. • Array G SA • G SA • G SA : Smart antenna diversity gain (for cross-polarised smart antennas) defined per clutter class. • L • L Path : Path loss ( L Path = L Model + L Ant ). • L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model. • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi. • M Shadowing – Model : Shadowing margin based on the model standard deviation. : Smart antenna array gain offset defined per clutter class. Combining : Smart power combining gain offset defined per clutter class. Div TX i : Total transmitter losses for the transmitter TXi ( L TX TX i = L Total – DL ). i TX i In coverage predictions, shadowing margins are taken into account when the option "Shadowing taken into account" is selected. • L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected. • L Mi G • Mi : Receiver terminal losses for the pixel, subscriber, or mobile Mi. : Receiver terminal’s antenna gain for the pixel, subscriber, or mobile Mi. Mi L Ant : Receiver terminal’s antenna attenuation calculated for the pixel, subscriber, or mobile Mi. • Mi For calculating the useful signal level from the best serving cell, L Ant is determined in the direction (H,V) = (0,0) from Mi the antenna patterns of the antenna used by Mi. For calculating the interfering signal level from any interferer, L Ant is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi, while the antenna is pointed towards Mi’s best serving cell. Mi • L Body : Body loss defined for the service used by the pixel, subscriber, or mobile Mi. Calculations The received traffic and pilot signal levels (dBm) from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi as follows: TX i  ic  TX i  ic  Mi TX i  ic  Mi C Traffic = EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G TX i  ic  C Pilot = EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G –L –L Mi Mi Mi Mi Mi Mi – L Ant – L Body and – L Ant – L Body Where EIRP is the effective isotropic radiated power of the cell calculated as follows: TX i  ic  TX i  ic  TX i TX i  ic  TX i EIRP Traffic = P Traffic + G TX i  ic  EIRP Pilot = P Pilot + G TX  ic  i TX  ic  i With P Traffic and P Pilot TX  ic  i TX  ic  i Array + G SA Array + G SA + G SA + G SA TX i Div TX i + G SA – L Combining + G SA – L and being the traffic and pilot transmission powers of the cell TXi(ic) calculated as follows: TX  ic  i TX  ic  i P Traffic = P Preamble – P Traffic and P Pilot 716 Div Combining TX  ic  i TX  ic  i = P Preamble – P Pilot Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 If you wish to exclude the the energy corresponding to the cyclic prefix part of the total symbol duration from the useful signal level, you must add the following lines in the Atoll.ini file: [WiMAX] ExcludeCPFromUsefulPower = 1 TX i  ic  When this option is active, the cyclic prefix energy is excluded from C Preamble . In other TX i  ic  words, the factor 10  Log  1 – r CP TX  ic   is added to C i Preamble .  Independant of the option, interference levels are calculated for the total symbol durations, i.e., the energy of the useful symbol duration and the cyclic prefix energy. Output TX i  ic  • C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi. • C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi. TX i  ic  10.4.6.2 Traffic and Pilot Noise Calculation (DL) For determining the traffic and pilot C/N and C/(I+N), Atoll calculates the downlink noise over the channel bandwidth used by the cell. The used bandwidth depends on the number of used subcarriers. The numbers of subcarriers used by different permutation zones can be different. The downlink noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on the temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the used bandwidth. Input • • K: Boltzmann’s constant. T: Temperature in Kelvin. M i PZ DL • N SCa – Used : Number of subcarriers used by the downlink permutation zone of a cell TXi(ic) assigned to Mi. • N SCa – Total : Total number of subcarriers defined for the frame configuration of a cell TXi(ic). • F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on page 740. • nf TX i  ic  TX i  ic  M i : Noise figure of the terminal used for calculations by the pixel, subscriber, or mobile Mi. Calculations The power spectral density of thermal noise is calculated as follows: n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz The thermal noise for a cell is calculated as: Mi TX i  ic  n 0 – DL PZ DL   N SCa – Used   TXi  ic  = n 0 + 10  Log  F Sampling  ------------------------ TX i  ic    N SCa – Total   The downlink noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel, subscriber, or mobile Mi. TX i  ic  n DL TX i  ic  = n 0 – DL + nf Mi Effect of Segmentation: If you select downlink segmentation support for the frame configuration used by the cell, it means that the first downlink PUSC permutation zone is segmented. All other zones are pooled together to form a non-segmented zone. The downlink segmenting factor, f Segment – DL , is calculated from the number of secondary subchannel groups assigned to the permutation zone in the Permutation Zones table. 717 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015  PSG + 2  SSGf Segment – DL = 3 -------------------------------------------15 Where, PSG is the number of primary subchannel groups and SSG is the number of used secondary subchannel groups. The multiplicative coefficients of 3 and 2 are derived from the ratio of the numbers of subchannels that belong to the primary and to the secondary subchannel gourps. For example, for the FFT size of 1024 (or 2048), each primary subchannel group contains 6 (or 12) subchannels, and each secondary subchannel group contains 4 (or 8) subchannels, which gives the ratio of 3:2. And, the denominator of 15 = 3 x 3 + 2 x 3. f Segment – DL represents the fraction of the channel bandwidth used by a downlink segment. The power transmitted 1 over a segment has ---------------------------- times the spectral density of the power transmitted over the entire channel f Segment – DL 1 bandwidth. When calculating the downlink C/N and C/(I+N) ratios, the increase in power by ---------------------------- due to this f Segment – DL power concentration is equivalent to a reduction in the noise level by f Segment – DL . Hence, if downlink segmentation is used, the thermal noise power at the pixel, subscriber, or mobile Mi covered by the downlink segmented permutation zone is reduced by the factor f Segment – DL . Which means that the thermal noise for the a segment of the channel used by a cell is calculated as: Mi TX i  ic  n 0 – DL PZ DL    TX i  ic  N SCa – Used  = n 0 + 10  Log  FSampling  ------------------------ f Segment – DL TX i  ic    N SCa – Total   Output • TX i  ic  n DL : Downlink noise for the cell TXi(ic). 10.4.6.3 Traffic and Pilot Interference Calculation (DL) The interference received by any pixel, subscriber, or mobile, served by a cell TXi(ic) from other cells TXj(jc) can be defined as the traffic and pilot signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the channels used by the cells TXi(ic) and TXj(jc), on the traffic loads of the interfering cells TXj(jc), and whether the cells use downlink segmentation or not. Moreover, the interference can come from cells using simple as well as smart antennas. The calculation can be divided into the two parts. • • 10.4.6.3.1 "Traffic and Pilot Interference Signal Levels Calculation (DL)" on page 718. "Effective Traffic and Pilot Interference Calculation (DL)" on page 722. Traffic and Pilot Interference Signal Levels Calculation (DL) The traffic and pilot signal levels received from interfering cells TXj(jc) at a pixel, subscriber, or mobile Mi, covered by a cell TXi(ic), are calculated in a different manner than the traffic and pilot signal levels from the studied cell TXi(ic). This section explains how these interfering signals are calculated. Input 718 TX j  jc  • P Preamble : Preamble transmission power of the cell TXj(jc). • P Pilot : Pilot power reduction of the interfering cell TXj(jc). • P Traffic : Traffic power reduction of the interfering cell TXj(jc). • P Idle – Pilot : Idle pilot power reduction of the interfering cell TXj(jc). • L TX j  jc  TX j  jc  TX j  jc  TX j : Total transmitter losses for the transmitter TXj ( L TX j = L Total – DL ). TX j • L Path : Path loss ( L Path = L Model + L Ant ). • L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model. Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 TX j • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXj. • M Shadowing – Model : Shadowing margin based on the model standard deviation. • M Shadowing – C  I : Shadowing margin based on the C/I standard deviation. In coverage predictions, shadowing margins are taken into account when the option "Shadowing taken into account" is selected. • L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected. • L • • Mi G : Receiver terminal losses for the pixel, subscriber, or mobile Mi. Mi : Receiver terminal’s antenna gain for the pixel, subscriber, or mobile Mi. Mi L Ant : Receiver terminal’s antenna attenuation calculated for the pixel, subscriber, or mobile Mi. M i L Ant is determined in the direction of TXj(jc) from the antenna patterns of the antenna used by Mi while the antenna is pointed towards TXi(ic). Mi • L Body : Body loss defined for the service used by the pixel, subscriber, or mobile Mi. • TL DL TX  jc  j : Downlink traffic load of the interfering cell TXj(jc). Traffic loads can either be calculated using Monte Carlo simulations, or entered manually for each cell. Calculation of traffic loads is explained in "Simulation Process" on page 699. • TX  jc  j AU DL : Downlink AAS usage ratio of the interfering cell TXj(jc). Downlink AAS usage ratios are calculated using Monte Carlo simulations as explained in "Simulation Process" on page 699. • TX j  jc  N SCa – Used : Number of used subcarriers defined for the first downlink permutation zone in the frame configuration assigned to the interfering cell TXj(jc). • TX j  jc  N SCa – Data : Number of data subcarriers defined for the first downlink permutation zone in the frame configuration assigned to the interfering cell TXj(jc). Calculations WiMAX cells can transmit different powers on pilot (NUsed – NData) and data (NData) subcarriers for the part of the frame with traffic, and a different pilot power for the part of the frame that does not have traffic bursts. Data subcarriers are off during the empty part of the frame. Therefore, the interference received from a cell depends on the traffic load and the different powers of the cell, i.e., pilot, traffic, and idle pilot powers. Monte Carlo simulations and coverage prediction calculations present different scenarios for interference calculations in the case of smart antennas. • Monte Carlo Simulations: In the case of Monte Carlo simulations, the interferer is either using the transmitter antenna or the smart antenna at any given moment. So, for each interfered pixel, subscriber, or mobile, Atoll already knows the type of the interference source. Therefore, the interference received from any cell TXj(jc) can be given by: TX  jc  TX j  jc  Without smart antennas: I Total TX  jc  j  I j  I Non – AAS Idle  ---------------------------------------------- 10 10   + 10 = 10  Log 10       TX  jc  With smart antennas: • TX  jc  j I Total  I j  AAS -  -----------------10   = 10  Log  10      Coverage Predictions: 719 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 In the case of coverage prediction calculations, the interferer could either be transmitting using the transmitter antenna, or using the smart antenna, or it could be empty, or not transmitting.Therefore, the interference received from any cell TXj(jc) can be given by: TX  jc  TX  jc  TX  jc  j j  I j  I I Non – AAS Idle AAS  ---------------------------------------------------------------- 10 10 10  + 10 + 10 = 10  Log  10       TX  jc  j I Total Where, the three components of the interference are: TX j  jc  • I Non – AAS : Interference from the loaded part of the frame transmitted using the main antenna, • I AAS • I Idle TX  jc  j TX j  jc  : Interference from the loaded part of the frame transmitted using the smart antenna, : Interference from the empty, or idle, part of the frame. The above components of the interference are calculated as follows: The interference from the loaded part of the frame transmitted using the main antenna is calculated as follows: The received interfering traffic and pilot signal levels (dBm) from any cell TXj(jc) are calculated for a pixel, subscriber, or mobile Mi as follows: In Monte Carlo simulations: TX j  jc  TX j  jc  Mi TX j  jc  Mi I Traffic = EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G TX j  jc  = EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G I Pilot –L –L Mi Mi Mi Mi Mi Mi – L Ant – L Body – L Ant – L Body In coverage prediction: TX j  jc  TX j  jc  Mi TX j  jc  Mi I Traffic = EIRP Traffic – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor + G TX j  jc  = EIRP Pilot – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor + G I Pilot –L –L Mi Mi Mi Mi Mi Mi – L Ant – L Body – L Ant – L Body Where EIRP is the effective isotropic radiated power of the cell calculated as follows: TX j  jc  TX j  jc  EIRP Traffic = P Traffic + G TX j  jc  TX j  jc  With P Traffic and P Pilot TX  jc  j TX  jc  j TX j –L TX j TX j  jc  and EIRP Pilot TX j TX j –L TX j being the traffic and pilot transmission powers of the cell TXj(jc) calculated as follows: TX  jc  j TX  jc  j P Traffic = P Preamble – P Traffic and P Pilot And G TX j  jc  = P Pilot + G TX  jc  j TX  jc  j = P Preamble – P Pilot TX j = G Ant , i.e., the transmitter antenna gain for the antenna used by the transmitter TXj. The interference from the loaded part of the frame transmitted using the main antenna is given as: TX  jc  TX j  jc  I Non – AAS TX  jc  j  I j   I Pilot TX j  jc  TX j  jc  Traffic   ------------------ TX  jc  ------------------  TX  jc  N N j j 10 10 SCa – Data SCa – Data - + 10  -----------------------  1 – -------------------------     1 – AU DL    10 = 10  Log  TL DL TX j  jc        TX j  jc  N SCa – Used    N SCa – Used      If you wish to include the effect of the number of antennas in case of MIMO, you must add the following lines in the Atoll.ini file: [WiMAX] MultiAntennaInterference = 1 When the multi-antenna interference option is active, the interference is incremented by TX  jc  TX  jc  j j + 10  Log  N Ant – TX . Where N Ant – TX is the number of MIMO transmission   (downlink) antennas defined for the cell TXj(jc). 720 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 The interference from the loaded part of the frame transmitted using the smart antenna is calculated as follows: The received interfering traffic signal level (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows: In Monte Carlo simulations: TX j  jc  I AAS TX j  jc  = EIRPAAS – L Path – M Shadowing – Model – L Indoor + G Mi –L Mi Mi Mi – L Ant – L Body In coverage prediction: TX j  jc  I AAS TX j  jc  = EIRPAAS – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor + G Mi –L Mi Mi Mi – L Ant – L Body Where EIRP is the effective isotropic radiated power of the cell calculated as follows: TX j  jc  EIRP AAS TX j  jc  = P Traffic + G TX j –L TX j TX j  jc  With P Traffic being the traffic transmission power of the cell TXj(jc) calculated as follows: TX j  jc  TX j  jc  TX j  jc  P Traffic = P Preamble – P Traffic And, G TX j = G SA    is the smart antenna gain in the direction of the victim mobile Mi, calculated from the angular distributions of the downlink traffic power density of the interfering cells. The angular distribution of the downlink traffic power density is determined from the array correlation matrices calculated during Monte Carlo simulations.  is the direction in which the victim pixel, subscriber, or mobile Mi is located. For more information on the calculation of G SA    , see "Beamforming Smart Antenna Models" on page 43. The gain of the interfering signal, G SA    , transmitted in the direction of each pixel  is given by: H G SA    = g n     S   R Avg  S  Where S  is the steering vector in the direction  (probe mobile/pixel), H denotes the Hilbert transform, R Avg is the average array correlation matrix, and g n    is the gain of the nth antenna element in the direction  . The interference from the empty, or idle, part of the frame transmitted using the transmitter antenna is calculated as follows: The received interfering pilot signal level (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows: TX j  jc  TX j  jc  I Idle – Pilot = EIRP Idle – Pilot – L Path – L Indoor + G Mi –L Mi Mi Mi – L Ant – L Body Where EIRP is the effective isotropic radiated power of the cell calculated as follows: TX j  jc  TX j  jc  EIRP Idle – Pilot = P Idle – Pilot + G TX j –L TX j TX j  jc  With P Idle – Pilot being the idle pilot transmission power of the cell TXj(jc) calculated as follows: TX j  jc  TX j  jc  TX j  jc  P Idle – Pilot = P Preamble – P Idle – Pilot And, G TX j TX j = G Ant , i.e., the transmitter antenna gain for the antenna used by the transmitter TXj. The interference from the empty, or idle, part of the frame transmitted using the transmitter antenna is given as: TX  jc  TX j  jc  I Idle  I j   TX j  jc  Idle – Pilot   ----------------------------   TX j  jc  N 10 SCa – Data     1 – -----------------------= 10  Log   1 – TL DL    10       TX j  jc     N SCa – Used      721 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 If you wish to include the effect of the number of antennas in case of MIMO, you must add the following lines in the Atoll.ini file: [WiMAX] MultiAntennaInterference = 1 When the multi-antenna interference option is active, the interference is incremented by TX  jc  TX  jc  j j + 10  Log  N Ant – TX . Where N Ant – TX is the number of MIMO transmission   (downlink) antennas defined for the cell TXj(jc). Output • 10.4.6.3.2 TX j  jc  I Total : Interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc). Effective Traffic and Pilot Interference Calculation (DL) The effective downlink traffic and pilot interference received at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) from interfering cells TXj(jc) depends on the co- and adjacent channel overlap that exists between the channel used by the studied cell and the interfering cells, and the downlink segmentation parameters of the studied and interfering cells. The first downlink PUSC zone can be segmented at the studied and the interfering cells. The probability of subcarrier collision depends on the lengths of the segmented zones and on the subchannel groups used at both sides. Input • TX j  jc  I Total : Interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc) as calculated in "Traffic and Pilot Interference Signal Levels Calculation (DL)" on page 718. TX i  ic  – TX j  jc  • : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent Channel Overlaps Calculation" on page 702. • SU DL rO TX i  ic  TX j  jc  and SU DL : Downlink segmentation usage ratios defined for cells TXi(ic) and TXj(jc). Calculations The total traffic and pilot interference (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows: TX j  jc  I DL TX j  jc  TX i  ic  – TX j  jc  = I Total + f O TX i  ic  – TX j  jc  + f Seg – DL Inter – Tech + I DL Calculations for the interference reduction factors due to channel overlapping and downlink segmentation are explained below: Interference reduction due to the co- and adjacent channel overlap between the studied and the interfering cells: Interference reduction due to the co- and adjacent channel overlap between the cells TXi(ic) and TXj(jc) is calculated as follows: TX i  ic  – TX j  jc  fO TX i  ic  – TX j  jc  = 10  Log  r O    Interference reduction due to downlink segmentation: If you select downlink segmentation support for the frame configuration that you are using, it means that the first zone in the downlink, i.e., the DL PUSC zone, is segmented. All other zones are pooled together to form a group of non-segmented zones. There are two effects of segmentation: 1. Power concentration, which means that the spectral density of the power transmitted over one segment is higher than the spectral density of the same power transmitted over the entire channel bandwidth. The effect of power concentration is visible when calculating the downlink C/(I+N). The power transmitted over a segmented zone has 1 --------------------------- times the spectral density of the power transmitted over the entire channel bandwidth. When f Segment – DL 1 calculating the C/(I+N) ratio, the increase in power by ---------------------------- is equivalent to decreasing the noise and f Segment – DL 722 50 % of 80 %). segmented or not. then this means that the downlink traffic load of the segmented zone is 40 % (i. the denominator of 15 = 3 x 3 + 2 x 3. SU is the downlink segmentation usage ratios of the cells.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 interference by f Segment – DL . which gives the bandwidth used by a segment. The following paragraphs explain how the collision probability is calculated.3. and the downlink traffic load of the non-segmented zones is 40 %. and the downlink segmentation usage ratio is 50 %.3: Downlink Segmentation Atoll determines the switching point between the segmented and the non-segmented zones using the downlink segmentation usage ratio.. the interference received at the pixel. Figure 10. SP is the switching point between the segmented and the non-segmented zones. subscriber. In simulations. Atoll uses the downlink segmentation usage ratios stored in the cell properties for determining the interference. For example. The downlink segmenting factor. The multiplicative coefficients of 3 and 2 are derived from the ratio of the numbers of subchannels that belong to the primary and to the secondary subchannel gourps. and f Segment – DL is downlink segmenting factor. Hence. 723 . In coverage predictions. For example. or mobile Mi covered by the segmented zone is reduced by a factor of f Segment – DL .e. 2. and then calculates the downlink segmentation usage ratios according to the traffic loads of the mobiles allocated to the segmented zone and in the non-segmented zones. which gives the ratio of 3:2. if downlink segmentation is used. and each secondary subchannel group contains 4 (or 8) subchannels. The downlink segmentation usage (SU) ratio is the percentage of the total downlink traffic load present in the segmented downlink PUSC zone. for the FFT size of 1024 (or 2048). is calculated from the number of secondary subchannel groups assigned to the first downlink PUSC permutation zone in the Permutation Zones table. PSG is the number of primary subchannel groups and SSG is the number of secondary subchannel groups.  PSG + 2  SSGf Segment – DL = 3 -------------------------------------------15 Where. TXi(ic) and TXj(jc) respectively. f Segment – DL . And. if the downlink traffic load is 80 %.Atoll 3. Collision probability between the subcarriers used by the subchannels belonging to the segment of the studied cell and the subcarriers used by other sectors. are calculated as follows: SP SP TX i  ic  TX i  ic  SU DL = ----------------------------------------------------------------------------------------------and TX i  ic  TX i  ic  TX i  ic    SU DL + f Segment – DL  1 – SU DL   TX j  jc  SU DL = ----------------------------------------------------------------------------------------------TX j  jc  TX j  jc  TX j  jc  SU DL + f Segment – DL   1 – SU DL    TX  jc  j Where. The switching points between the segmented and non-segmented zones of the victim and interfering cells. Atoll resets the downlink segmentation usage ratios for all the cells to 0. each primary subchannel group contains 6 (or 12) subchannels. PSG in the cell TXi(ic). Between the non-segmented zone of the victim and the segmented zone of the interferer. the switching point formula is derived from the equation: SU DL  TL DL  1 – SU DL   TL DL -------------------------------------------------------------------.= ----------------------------------------------SP  fSegment – DL  W Channel  1 – SP   W Channel With cells using downlink segmentation.4: Downlink Segmentation Interference Scenarios Therefore. it means that the segmented zone does not exist. The mapping between the preamble index. • • • • Between the segmented zone of the victim and the segmented zone of the interferer. The probability of collision p Coll for each scenario is given by the following formula: 3  PSG Com + 2  SSG Com p Coll = ---------------------------------------------------------------------TX i  ic  TX i  ic  3  PSG + 2  SSG Where. Setting SU to 0 gives SP = 0. and SSG TX i  ic  TX i  ic  is the number of primary subchannel groups is the number of secondary subchannel groups in the cell TXi(ic).0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 If the downlink segmentation usage ratio is set to 0. 2 PI  96 96  PI  114 PI Modulo 32 PI – 96 PI Floor  ------ 32  PI – 96  Modulo 3 There can be 2 cases for calculating the total probability of collision. Between the segmented zone of the victim and the non-segmented zone of the interferer. and Cell PermBase is available in the IEEE specifications. Therefore.3. Between the non-segmented zone of the victim and the non-segmented zone of the interferer. This mapping is performed in Atoll as follows: Preamble Index ( PI ) Range: 0 to 113 Cell PermBase ( PB ) Range: 0 to 31 Segment Number ( N Seg ) Range: 0. SSGCom is the number of secondary subchannel groups common in TXi(ic) and TXj(jc). the segment number. which shows how the switching point varies with the downlink segmentation usage ratio. and setting SU to 1 gives SP = 1 (or 100%). Figure 10.Atoll 3. Derivation of the switching point formula: The downlink segmentation usage ratio is used to partition the total downlink traffic load into segmented and non-segmented zones. 1. there can be four different interference scenarios. Atoll calculates the probabilities of collision for each scenario and weights the total interference according to the total collision probability. 724 . PSGCom is the number of primary subchannel groups common in TXi(ic) and TXj(jc). The segment numbers and the cell permutation base numbers (Cell PermBase) are determined from the cell’s preamble index. or mobile Mi from any interfering cell TXj(jc). subscriber. maximum power from LTE cells. or mobile location. preamble power from WiMAX cells.If SP j  SP i  TX  ic   1 – SP i      The interference reduction factor due to downlink segmentation for the pixel. L Indoor is the indoor losses taken into account when the option "Indoor coverage" is selected. the total collision probability for the pixel. or mobile Mi is covered by the non-segmented zone of TXi(ic). TX k P DL – Rec is calculated based on the EIRP from GSM cells. or mobile Mi is calculated as follows: TX  ic  – TX  jc  i j p Collision – DL  TX  jc  TX  ic  j i SS  p Coll If SP  SP   TX  jc  TX  ic  TX  jc  =  SS j i j SN  + p Coll   SP – SP TX j  jc  TX i  ic   p Coll  SP   ------------------------------------------------------------------------------------------------------------If SP  SP  TX i  ic   SP  Case 2: If the pixel. and TD-SCDMA cells. Mi subscriber. or mobile Mi is covered by the segmented zone of TXi(ic). subscriber.3. L Mi is the receiver terminal losses for the pixel. subscriber. or mobile Mi. subscriber. or mobile Mi is calculated as follows: • TX i  ic  – TX j  jc  p Collision – DL TX j  jc  TX i  ic   NN p Coll If SP  SP   TX j  jc  TX  jc  TX  ic     + p NS   SP j – SP i  =  p NN TX  jc  TX  ic  Coll   1 – SP Coll    ---------------------------------------------------------------------------------------------------------------------------.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 • Case 1: If the pixel. subscriber. subscriber. and F  TX i  ic  TX k  ICPDL is the inter-technology downlink channel protection ratio for a frequency offset F between the interfered and interfering frequency channels of TXi(ic) and TXk. and L Body is the body loss defined for the service used by the pixel. L Ant is the receiver terminal’s antenna attenuation calculated for the pixel. CDMA2000. subscriber. and downlink cell power from Wi-Fi cells. subscriber. 725 . or mobile Mi is calculated as follows: TX i  ic  – TX j  jc  f Seg – DL Inter – Tech I DL TX  ic  – TX  jc  i j = 10  Log  p Collision – DL    is the inter-technology downlink interference from transmitters of an external network (linked document of any technology) calculated as follows: Inter – Tech I DL  = TX – External EIRP DL – L Path – L Indoor + G M i –L M i M i M i Inter – Tech – L Ant – L Body – f IRF All External TXs TX – External Where EIRP DL is the downlink EIRP of the external transmitter.Atoll 3. total power from UMTS. the total collision probability for the pixel. Calculation of the Downlink Inter-technology Interference The downlink inter-technology interference is calculated as follows: Inter – Tech I DL TX TX k   P DL – Rec  --------------------------------------- = F  TX i  ic  TX k    TX  ICP DL k  k Here P DL – Rec is the received downlink power from an interfering cell TXk belonging to another technology. subscriber. G Mi is the receiver terminal’s antenna Mi gain for the pixel. • Inter – Tech I DL : Downlink inter-technology interference. Output • TX j  jc  I DL : Effective downlink traffic and pilot interference received at the pixel. subscriber. or mobile Mi. or mobile Mi. L Path is the path loss from the external transmitters to the pixel. or mobile Mi. the STTD/MRC or SU-MIMO diversity gain. the subchannel allocation mode of PZ DL . subscriber. • M i B DL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. subscriber. • M i BLER  BDL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the   terminal used by the pixel. corresponding to the bearer is applied to its selection threshold. • TX i  ic  C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel. 726 . • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. The gain is read from the properties of the WiMAX equipment assigned to the TX i  ic  Mi Mi pixel. • Whose selection thresholds are less than the traffic or pilot C/N at Mi: T B  CNR Traffic or T B  CNR Pilot Mi TX i  ic  Mi TX i  ic  Mi If the cell supports MIMO. Calculations The traffic and pilot C/N for a cell TXi(ic) are calculated as follows for any pixel. or mobile Mi. • N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel.Atoll 3. subscriber. or mobile Mi for N Ant – TX .0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 10. if the corresponding option has been set in the Atoll. or mobile Mi’s WiMAX equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). or mobile Mi as calculated in "Traffic and Pilot Signal Level Calculation (DL)" on page 715. or mobile Mi Mi as calculated in "Permutation Zone Selection" on page 714. or mobile Mi as calculated in "Traffic and Pilot Signal Level Calculation (DL)" on page 715. For more information. M i Mi or mobile Mi.4 Traffic and Pilot C/N Calculation (DL) Input • TX  ic  i C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel. or Mi mobile Mi. • Mobility  M i  : Mobility used for the calculations.4. M i BLER  B DL . Mobility  M i  . subscriber.ini file. N Ant – RX . TX i  ic  • T AMS : AMS threshold defined for the cell TXi(ic). subscriber.6. subscriber. or mobile Mi: TX  ic  i TX  ic  i TX  ic  i TX i  ic  TX i  ic  CNR Traffic = C Traffic – n DL TX i  ic  CNR Pilot = C Pilot – n DL Bearer Determination: The bearers available for selection in the pixel. TX i  ic  • N Ant – TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic). G Div – DL . see the Administrator Manual. subscriber. • Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel. • TX i  ic  n DL : Downlink noise for the cell TXi(ic) as calculated in "Traffic and Pilot Noise Calculation (DL)" on page 717.3. subscriber. or mobile Mi. • T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used by Mi’s terminal. • B DL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. subscriber. subscriber. Finally. subscriber. MIMO – STTD/MRC and SU-MIMO Diversity Gain: Once the bearer is known.4. Output TX  ic  i • CNR Traffic : Traffic C/N from the cell TXi(ic) at the pixel. the selected bearer is the one with the highest downlink effective MAC channel throughput as calculated in "Channel Throughput. Atoll takes the ratio of the signal level and the sum of the total interference from other cells and the downlink noise (as calculated in "Traffic and Pilot Noise Calculation (DL)" on page 717). the selected bearer is the one with the highest index. Therefore.Atoll 3. or mobile Mi as calculated in "Traffic and Pilot Signal Level Calculation (DL)" on page 715. or mobile under study. Input • TX  ic  i C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel. subscriber. or mobile Mi. the receiver is oriented towards the best server just as in the case of LOS. and Per-User Throughput Calculation" on page 743. Atoll does not try to find the direction of the strongest signal. the traffic and pilot C/N calculated above become: TX i  ic  TX i  ic  Mi DL TX i  ic  Mi DL CNR Traffic = CNR Traffic + G Div – DL + G Div TX i  ic  CNR Pilot = CNR Pilot + G Div – DL + G Div Mi Where G Div – DL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer. • Bearer Index From among the bearers available for selection. 727 .and adjacent channel overlap between the studied and the interfering cells. Cell Capacity. In the case of NLOS between the receiver and the best server. TX i  ic  10.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 DL The additional diversity gain defined for the clutter class of the pixel. and the probabilities of subcarrier collision if downlink segmentation is used. Interference from each cell is weighted according to the co.6. and Per-User Throughput Calculation" on page 743. • CNR Pilot : Pilot C/N from the cell TXi(ic) at the pixel. • TX  ic  i C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel.3. Allocated Bandwidth Throughput. subscriber. except when the "Lock Status" is set to "Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited. subscriber. • Peak MAC Throughput From among the bearers available for selection. or mobile Mi. Atoll calculates the interference received at the same studied pixel. the traffic loads of the interfering cells. or mobile Mi as calculated in "Traffic and Pilot Signal Level Calculation (DL)" on page 715. subscriber. First Atoll calculates the received signal level from the studied cell (as explained in "Traffic and Pilot Signal Level Calculation (DL)" on page 715) at the pixel. subscriber. The receiver terminal is always considered to be oriented towards its best server.5 Traffic and Pilot C/(I+N) and Bearer Calculation (DL) The carrier signal to interference and noise ratio is calculated in three steps. subscriber. the selected bearer is the one with the highest downlink peak MAC channel throughput as calculated in "Channel Throughput. Next. or mobile from all the interfering cells (as explained in "Traffic and Pilot Interference Calculation (DL)" on page 718). • TX  ic  i n DL : Downlink noise for the cell TXi(ic) as calculated in "Traffic and Pilot Noise Calculation (DL)" on page 717. the bearers available for selection are all the bearers defined in the WiMAX equipment for which the following is true: M i DL TX  ic  i Mi DL TX i  ic  M i T B – G Div – DL – G Div  CNR Traffic Mi T B – G Div – DL – G Div  CNR Pilot The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). Cell Capacity. • Effective MAC Throughput From among the bearers available for selection. Allocated Bandwidth Throughput. or mobile Mi G Div is also applied. or mobile Mi: TX  ic  TX i  ic   I + N  DL i   TX j  jc   n DL   I DL  --------------------- Inter – Tech 10  ----------------- + NR Inter – Tech = 10  Log  + 10 10  + I DL DL   10    All TX j  jc          Bearer Determination: The bearers available for selection in the pixel. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. subscriber. • Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel. Calculations The traffic and pilot C/(I+N) for a cell TXi(ic) is calculated as follows for any pixel. subscriber. or mobile Mi Mi as calculated in "Permutation Zone Selection" on page 714. or mobile Mi covered by a cell TXi(ic) as explained in "Traffic and Pilot Interference Calculation (DL)" on page 718.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 TX  jc  j • I DL : Effective downlink traffic and pilot interference from any cell TXj(jc) calculated for a pixel. or mobile Mi: TX i  ic  CINR Traffic TX i  ic  CINR Pilot    TXj  jc   TX i  ic  n    IDL    DL Inter – Tech Inter – Tech ------------------- ---------------------    and + I DL + + NR DL = C Traffic – 10  Log 10 10      10  10   All TXj  jc            TX i  ic  TX i  ic  = C Pilot   TXj  jc     TX i  ic  n DL  IDL      Inter – Tech Inter – Tech ------------------- ---------------------    + I DL + + NR DL 10 – 10  Log 10  10      10    All TXj  jc            The Traffic Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel.Atoll 3. if the corresponding option has been set in the Atoll. subscriber. Inter – Tech : Inter-technology downlink noise rise. • T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used by Mi’s terminal. or Mi mobile Mi. For more information. subscriber. • NRDL • T AMS : AMS threshold defined for the cell TXi(ic). or mobile Mi. or mobile Mi’s WiMAX equipment are the ones: 728 • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). TX i  ic  • N Ant – TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic).ini file. • Whose selection thresholds are less than the traffic or pilot C/(I+N) at Mi: T B  CINR Traffic or T B  CINR Pilot Mi TX i  ic  Mi TX i  ic  . • N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel. subscriber. or mobile Mi. subscriber. • Mobility  M i  : Mobility used for the calculations. TX  ic  i Mi M i or mobile Mi. • M i BLER  BDL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the terminal used by the pixel. • B DL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. subscriber.3. • Inter – Tech I DL : Downlink inter-technology interference as calculated in "Traffic and Pilot Interference Calculation (DL)" on page 718. • Mi B DL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. subscriber. subscriber. see the Administrator Manual. the subchannel allocation mode of PZ DL . and Per-User Throughput Calculation" on page 743. or mobile Mi G Div is also applied. subscriber. the traffic and pilot C/(I+N) calculated above become: TX i  ic  TX i  ic  Mi DL TX i  ic  Mi DL CINR Traffic = CINR Traffic + G Div – DL + G Div TX i  ic  CINR Pilot M = CINR Pilot + G Div – DL + G Div i Where G Div – DL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer. Cell Capacity.   DL The additional diversity gain defined for the clutter class of the pixel. the selected bearer is the one with the highest index. 10. •  I + N  DL TX i  ic  TX i  ic  : Traffic Total noise from the interfering cells TXj(jc) at the pixel. or mobile Mi for N Ant – TX . 729 .6. N Ant – RX . the bearers available for selection are all the bearers defined in the WiMAX equipment for which the following is true: Mi Mi DL TX i  ic  Mi Mi DL TX i  ic  T B – G Div – DL – G Div  CINR Traffic T B – G Div – DL – G Div  CINR Pilot The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). Therefore. subscriber. The gain is read from the properties of the WiMAX equipment assigned to the TX  ic  i M i M i pixel. the selected bearer is the one with the highest downlink peak MAC channel throughput as calculated in "Channel Throughput. • Bearer Index From among the bearers available for selection. or mobile Mi. and Per-User Throughput Calculation" on page 743. the selected bearer is the one with the highest downlink effective MAC channel throughput as calculated in "Channel Throughput. Allocated Bandwidth Throughput.6 Traffic Signal Level Calculation (UL) Input M i • P Max : Maximum transmission power of the terminal used by the pixel. Allocated Bandwidth Throughput. or mobile Mi. or mobile Mi after power control as Mi calculated in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737. Cell Capacity. Mobility  M i  . or mobile Mi covered by a cell TXi(ic).4. or mobile Mi in the downlink.Atoll 3. or mobile Mi without power control. Output TX i  ic  • CINR Traffic : Traffic C/(I+N) from the cell TXi(ic) at the pixel. subscriber. • Mi B DL : Bearer assigned to the pixel. • Effective MAC Throughput From among the bearers available for selection. corresponding to the bearer is applied to its selection threshold. subscriber. the STTD/MRC or SU-MIMO diversity gain.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 M i If the cell supports MIMO.3. M i BLER  B DL . subscriber. subscriber. • P Eff : Effective transmission power of the terminal used by the pixel. subscriber. G Div – DL . • Peak MAC Throughput From among the bearers available for selection. MIMO – STTD/MRC and SU-MIMO Diversity Gain: Once the bearer is known. • CINR Pilot : Pilot C/(I+N) from the cell TXi(ic) at the pixel. subscriber. or mobile Mi at a cell TXi(ic). while the antenna is pointed towards Mi’s best serving cell. The numbers of subcarriers used by different permutation zones can be different. subscriber. 730 . or mobile Mi. subscriber. 10.3. refer to section "Beamforming Smart Antenna Models" on page 43. subscriber. Calculations The received traffic signal level (dBm) from a pixel.e.e.7 Traffic Noise Calculation (UL) For determining the uplink C/N and C/(I+N). i. Atoll calculates the uplink noise over the channel bandwidth used by the cell. • L • L Path : Path loss ( L Path = L Model + L Ant ). • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi. : Receiver terminal’s antenna gain for the pixel. However. • M i For calculating the useful signal level from the best serving cell. G TX i TX i = G SA = 10  Log  E SA  . or mobile Mi at its serving cell TXi(ic) is calculated as follows: Mi Mi C UL = EIRP UL – L Path – M Shadowing – Model – L Indoor + G TX i –L TX i Mi Mi – L Ant – L Body Where EIRP is the effective isotropic radiated power of the terminal calculated as follows: Mi EIRP UL = P With P Mi Mi +G Mi –L Mi Mi = P Max without power control at the start of the calculations.e. and is the P Mi Mi = P Eff after power control. i. The uplink noise comprises thermal noise and the noise figure of the equipment. • L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected. L Ant is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi. it remains constant for a given temperature. Output • Mi C UL : Received uplink signal level from the pixel. The used bandwidth depends on the number of used subcarriers. • L Mi G • Mi : Receiver terminal losses for the pixel. subscriber. Mi • L Body : Body loss defined for the service used by the pixel. The thermal noise density depends on the temperature. TX i : Total transmitter losses for the transmitter TXi ( L TX i = L Total – UL ).0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks TX ©Forsk 2015 i • E SA : Number of antenna elements defined for the smart antenna equipment used by the transmitter TXi.Atoll 3. subscriber. is the uplink smart antenna beamforming gain..6... i. or mobile Mi. subscriber. TX i TX i In coverage predictions. or mobile Mi. Mi L Ant : Receiver terminal’s antenna attenuation calculated for the pixel. • M Shadowing – Model : Shadowing margin based on the model standard deviation.0) from Mi the antenna patterns of the antenna used by Mi.   For more information on the calculation of G SA . shadowing margins are taken into account when the option "Shadowing taken into account" is selected.4. • Without smart antennas: G • With smart antennas: G TX i TX i is the transmitter antenna gain. L Ant is determined in the direction (H. • L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model. G TX i TX i = G Ant . • G TX i : Transmitter antenna gain for the antenna used by the transmitter TXi. For calculating the interfering signal level from any interferer. the value of the thermal noise varies with the used bandwidth. or mobile Mi.V) = (0. 4. The interference received by a cell TXi(ic) from an interfering mobile covered by a cell TXj(jc) can be defined as the uplink signal level received from interfering mobiles Mj depending on the overlap that exists between the channels used by the cells TXi(ic) and TXj(jc). T: Temperature in Kelvin. M Shadowing – Model : Shadowing margin based on the model standard deviation. In coverage predictions. TX i  ic  • N SCa – Total : Total number of subcarriers defined for the frame configuration of a cell TXi(ic). In Monte Carlo simulations. on the traffic loads of the interfering mobile Mj. interfering signal levels already include M Shadowing – Model . • nf TX i  ic  TX  ic  i : Noise figure of the cell TXi(ic).6. In coverage predictions. The calculation of uplink interference can be divided into two parts: • • 10. as explained in "Traffic Signal Level Calculation (UL)" on page 729. Traffic Interference Signal Levels Calculation (UL) Input Mj • C UL : Uplink signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) as calculated in • "Traffic Signal Level Calculation (UL)" on page 729. Calculation of the uplink noise rise which represents the total uplink interference from all the interfering mobiles as explained in "Noise Rise Calculation (UL)" on page 733.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Input • • K: Boltzmann’s constant. the uplink noise rise values already available in simulation results or in the Cells table are used.Atoll 3.4. see "Shadow Fading Model" on page 90). • i UL N SCa – Used M PZ : Number of subcarriers used by the uplink permutation zone of a cell TXi(ic) assigned to Mi. Calculations The power spectral density of thermal noise is calculated as follows: n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz The thermal noise for a cell is calculated as: Mi TX i  ic  n 0 – UL PZ UL   N SCa – Used   TXi  ic  = n 0 + 10  Log  F Sampling  ------------------------ TX i  ic    N SCa – Total   The uplink noise is the sum of the thermal noise and the noise figure of the cell TXi(ic).8 Traffic Interference Calculation (UL) The uplink traffic interference is only calculated during Monte Carlo simulations. the ratio M Shadowing – Model – M Shadowing – C  I is applied to the interfering signals (for more information.1 Calculation of the uplink interference from each individual interfering mobile as explained in "Traffic Interference Signal Levels Calculation (UL)" on page 731. As the interfering signal levels already include 731 .8. TX i  ic  n UL TX i  ic  = n 0 – UL + nf TX i  ic  Output • TX i  ic  n UL : Uplink noise for the cell TXi(ic). • F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on page 740.6. 10.3. • M Shadowing – C  I : Shadowing margin based on the C/I standard deviation. • TX i  ic  – TX j  jc  rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co. and Cell PermBase is available in the IEEE specifications. This mapping is performed in Atoll as follows: Preamble Index ( PI ) Range: 0 to 113 Cell PermBase ( PB ) Range: 0 to 31 732 PI  96 96  PI  114 PI Modulo 32 PI – 96 . The interference reduction factor due to uplink segmentation is calculated as follows: TX  ic  – TX  jc  i j f Seg – UL TX  ic  – TX  jc  i j = 10  Log  p Collision – UL  TX i  ic  – TX j  jc  Where p Collision – UL is the collision probability between the subcarriers of the uplink segments being used by the interfered and interfering cells. and uplink segmentation are explained below: Interference reduction due to the co.and Adjacent Channel Overlaps Calculation" on page 702. uplink traffic load. All other zones are pooled together to form a group of nonsegmented zones. M Shadowing – C  I is added to the received interfering signal levels in order to achieve the ratio M Shadowing – Model – M Shadowing – C  I : M M j j C UL = C UL + M Shadowing – C  I In coverage predictions..e.3. Calculations The uplink interference received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) is calculated as follows: Mj Mj TX i  ic  – TX j  jc  I UL = C UL + f O TX i  ic  – TX j  jc  Mj + f TL – UL + f Seg – UL Calculations for the interference reduction factors due to channel overlapping. SC the cell TXi(ic). Traffic loads are calculated during Monte Carlo simulations as explained in "Scheduling and Radio Resource Allocation" on page 748. shadowing margins are taken into account when the option "Shadowing taken into account" is selected. TX i  ic  is the number of subchannels in The segment numbers and the cell permutation base numbers (Cell PermBase) are determined from the cell’s preamble index. it means that the first zone in the uplink.and adjacent channel overlap between the studied and the interfering cells: Interference reduction due to the co. the segment number. It is determined during Monte Carlo simulations as follows: TX i  ic  – TX j  jc  p Collision – UL SC Com = -----------------TX i  ic  SC Where. i. • Mj TL UL : Uplink traffic load of the interfering mobile Mj. SCCom is the number of subchannels common in TXi(ic) and TXj(jc).0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 M Shadowing – Model . The mapping between the preamble index. the UL PUSC zone.and adjacent channel overlap between the cells TXi(ic) and TXj(jc) is calculated as follows: TX i  ic  – TX j  jc  fO TX i  ic  – TX j  jc  = 10  Log  r O   Interference reduction due to interfering mobile’s traffic load: The interference reduction factor due to the interfering mobile’s uplink traffic load is calculated as follows: M M j j f TL – UL = 10  Log  TL UL   Interference reduction due to uplink segmentation: If you select uplink segmentation support for the frame configuration that you are using. is segmented.Atoll 3. 3. point analysis.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Segment Number ( N Seg ) Range: 0. In other words. 1.8. Atoll uses either the uplink segmented noise rise or the uplink noise rise to calculate the C/(I+N). Inter – Tech : Inter-technology uplink noise rise. and calculations on subscriber lists. TX i  ic  • n UL • NR UL : Uplink noise for the cell TXi(ic) as calculated in "Traffic Noise Calculation (UL)" on page 730. Atoll calculates the uplink total noise (I+N) as follows: TX i  ic   I + N  UL TX i  ic  TX i  ic  = NR UL – Seg + n UL 733 . or subscriber. Noise Rise Calculation (UL) The uplink noise rise is defined as the ratio of the total uplink interference received by any cell TXi(ic) from interfering mobiles Mj present in the coverage areas of other cells TXj(jc) to the uplink noise of the cell TXi(ic).4. or mobile Mi covered by the segmented zone in the interfered cell TXi(ic). it is the ratio (I+N)/N.Atoll 3. subscriber. subscriber.2 Mj I UL : Uplink interference signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc). see "Noise Rise Calculation (UL)" on page 733. that covers the pixel. Atoll calculates the uplink total noise (I+N) as follows: TX i  ic   I + N  UL TX i  ic  = NR UL TX i  ic  + n UL For any mobile Mi covered by the segmented zone in the interfered cell TXi(ic). Input • Mj I UL : Uplink interference signal levels received at a cell TXi(ic) from interfering mobiles Mj covered by other cells TXj(jc) as calculated in "Traffic Interference Signal Levels Calculation (UL)" on page 731. receiver. for the mobiles served by the segmented zone of the interfered cell Atoll calculates the uplink segmented noise rise. and for the mobiles served by the nonsegmented zones of the interfered cell Atoll calculates the uplink noise rise. segmented or nonsegmented. Atoll calculates two separate noise rise values. according to the zone.6. Calculations The uplink noise rise and total noise (I+N) for the cell TXi(ic) are calculated as follows: • Without smart antennas: For any mobile Mi covered by a non-segmented zone in the interfered cell TXi(ic). or mobile Mi covered by the non-segmented zone in the interfered cell TXi(ic). For more information on the calculation of the uplink noise rise. In coverage predictions. Atoll calculates the UL noise rise as follows: TX i  ic  NR UL  TX i  ic   IMj    n UL   UL -  non-seg M i  -------------------TX i  ic    Inter – Tech 10  -------------------------------------------= 10  Log  + NRUL – n UL 10  10  + 10     All Mj        All TX  jc    j  For any pixel. Atoll calculates the segmented zone UL noise rise as follows: TX i  ic  NR UL – Seg  TX i  ic   IMj    n UL UL    seg M i  --------------------- TX i  ic    10  Inter – Tech --------------------------------= 10  Log  + NR UL – n UL 10  10  + 10    All M j        All TX  jc    j  For any pixel. Output • 10. 2 PI Floor  ------  32  PI – 96  Modulo 3 In Monte Carlo simulations. • Mobility  M i  : Mobility used for the calculations. or mobile Mi. or mobile Mi. subscriber. subscriber.4. • : Uplink noise for the cell TXi(ic) as calculated in "Traffic Noise Calculation (UL)" on page 730. subscriber. • P Max : Maximum transmission power of the terminal used by the pixel.Atoll 3. or mobile M Mi as calculated in "Permutation Zone Selection" on page 714.6. TX i  ic  TX  ic  i Mi PZ UL N SC : Number of subchannels per channel defined for the uplink permutation zone assigned to the pixel. TX i  ic  • N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic). subscriber. or mobile Mi at its serving cell TXi(ic) as calculated in "Traffic Signal Level Calculation (UL)" on page 729. TX i  ic  TX i  ic  or  I + N  UL    : Total Noise for a cell TXi(ic) calculated for any pixel.9 Traffic C/N Calculation (UL) Input • M i C UL : Received uplink signal level from the pixel. subscriber. TX i  ic  • n UL • T AMS : AMS threshold defined for the cell TXi(ic). • NRUL •  I + N  UL TX i  ic  TX i  ic     : Angular distribution of the uplink noise rise for the cell TXi(ic). 10. subscriber. • Mi N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel.3. • M PC : Power control margin defined in the global network settings. • T B – Lowest : Bearer selection threshold of the lowest bearer in the WiMAX equipment assigned to the cell TXi(ic). • Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel. subscriber. • P Min : Minimum transmission power of the terminal used by the pixel.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks • ©Forsk 2015 With smart antennas: The angular distribution of the uplink noise rise is calculated during Monte Carlo simulations and can be stored in the Cells table in order to be used in coverage predictions. subscriber. or mobile Mi as calculated in "Permutation Zone Selection" on page 714. 734 i . or mobile Mi. or mobile Mi. or mobile Mi. subscriber. Mi Mi M i Mi or mobile Mi. Mi PZ UL = 8 • N SC  Seg : Number of subchannels per segment for the first uplink PUSC permutation zone. • T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used bythe cell TXi(ic). • B UL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. • M i B UL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. The angular distribution of the uplink noise rise is given by: 2 I UL    +  n  I NRUL    = --------------------------------2 n  I TX i  ic   I + N  UL 2    = I UL    +  n  I Output TX i  ic  • NRUL : Non-segmented uplink noise rise for the cell TXi(ic). • NRUL – Seg : Segmented uplink noise rise for the cell TXi(ic). SU-MIMO diversity or MU-MIMO diversity gain. The gain is read from the properties of the Mi TX i  ic  Mi WiMAX equipment assigned to the cell TXi(ic) for N Ant – RX . BLER  B UL . the STTD/MRC. M i Mobility  M i  . • Peak MAC Throughput From among the bearers available for selection. and Per-User Throughput Calculation" on page 743.3. Uplink Subchannelisation: The uplink subchannelisation depends on the uplink bandwidth allocation target defined for the scheduler used by the cell TXi(ic).0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 M i BLER  B UL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell • TXi(ic). subscriber. For more information. the selected bearer is the one with the highest index. the uplink C/N calculated above becomes: Mi Mi TX i  ic  UL CNR UL = CNR UL + G Div – UL + G Div TX i  ic  Where G Div – UL is the STTD/MRC. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. Therefore. • Bearer Index From among the bearers available for selection. UL The additional diversity gain defined for the clutter class of the pixel. see the Administrator Manual. and MU-MIMO Diversity Gain: Once the bearer is known. Cell Capacity. corresponding to the bearer is applied to its selection threshold. the selected bearer is the one with the highest uplink peak MAC channel throughput as calculated in "Channel Throughput. or MU-MIMO diversity gain corresponding to the selected bearer. G Div – UL . Allocated Bandwidth Throughput. N Ant – TX . Allocated Bandwidth Throughput. and Per-User Throughput Calculation" on page 743.ini file. the selected bearer is the one with the highest uplink effective MAC channel throughput as calculated in "Channel Throughput. • Effective MAC Throughput From among the bearers available for selection.Atoll 3. subscriber. if the corresponding option has been set in the Atoll. SU-MIMO Diversity. The uplink C/N calculated above is given for the total number of subchannels associated with the 735 . or mobile Mi G Div is also applied. SU-MIMO diversity. or mobile Mi at its serving cell TXi(ic) is calculated as follows: M M i i TX  ic  i CNR UL = C UL – n UL Bearer Determination: The bearers available for selection in the cell TXi(ic)’s WiMAX equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). Calculations The uplink C/N from a pixel. Cell Capacity. MIMO – STTD/MRC. the bearers available for selection are all the bearers defined in the WiMAX equipment for which the following is true: Mi TX i  ic  UL Mi T B – G Div – UL – G Div  CNR UL The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). the subchannel allocation mode of PZ UL . • Whose selection thresholds are less than the uplink C/N at Mi: T B  CNR UL Mi Mi TX i  ic  If the cell supports MIMO. or mobile Mi reduces its transmission power so that the uplink C/N from it at its cell is just enough to get the selected bearer. it will be assigned 5 subchannels as the used uplink bandwidth.Atoll 3. If with P Mi Mi Mi = P Max AND CNR UL  T TX i  ic  Mi + M PC . M M i i CNR UL = CNR UL .. The calculation of the gain introduced by the subchannelisation is explained below. for the bearer selected for the pixel. or with the highest effective MAC throughput.. i. Subchannelisation is performed for all the pixels. where T B UL TX i  ic  Mi B UL is the bearer selection threshold.e. subscriber. if using 5 subchannels.. and may reduce the number of used subchannels in order to satisfy the selected target. For example. subscriber.e. 736 . For example. The transmission power of Mi is reduced to determine the effective transmission power from the pixel. its uplink C/N will be better than when using 5. the mobile already has the best bearer using 5 subchannels. As there is no reduction in the bandwidth used for transmission. subscribers.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 M PZ permutation zone. • Full Bandwidth Full channel width is used by each mobile in the uplink. Final The pixel. as a mobile moves from good to bad radio conditions. and Min N SC – UL  Service  Mi Mi PZ UL = 8  N SC – UL  N SC  Seg for any pixel. there is no gain in the uplink C/N. The uplink subchannelisation may result in the use of a number of subchannels which is less than the total number of subchannels associated with the permutation zone. i.3. bearer with the highest index. Uplink Power Control: Once the subchannelisation is performed. or mobile Mi as follows: Mi M i  TX i  ic   Mi    Mi  P Eff = Max  PMax –  CNR UL –  T M + M PC   P Min    B i   UL Mi Mi CNR UL is calculated again using P Eff . and using 6 it would only get access to the second best. The definition of the best bearer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). Although using 4 subchannels.. subscriber. the uplink bandwidth is not reduced to 4 because it does not provide any gain in terms of the bearer. subscriber. Atoll continues to work with the C/N given by the subchannelisation. i. i. subscriber. The calculation of the gain introduced by the bandwidth reduction is explained below. The gain related to this bandwidth reduction is applied to the uplink C/N: Mi Mi CNR UL Final Where  PZUL  Mi  N SC  = CNR UL+ 10  Log  ----------------  N Mi  All SC  SC – UL Min N SC – UL  Service   M i N SC – UL PZ  N SC Mi UL for any pixel. or mobile Mi.e. • Maintain Connection The bandwidth used for transmission by a mobile is reduced only if the uplink C/N is not enough to even access the lowest bearer. with the highest peak MAC throughput. or mobile Mi covered by a non-segmented permutation zone in the interfered cell TXi(ic). a mobile is able to access the best bearer. • Best Bearer The bandwidth used for transmission by a mobile is reduced in order to improve the uplink C/N enough to access the best bearer. or mobile Mi covered by the segmented uplink PUSC zone in the interfered cell TXi(ic). the number of subchannels used by it for transmission in uplink are reduced one by one in order to improve the uplink C/N.e. or mobiles in the uplink. N SC i UL . from the WiMAX equipment assigned to the cell TXi(ic). Atoll calculates the received signal level from each pixel. • NR UL TX i  ic     : Angular distribution of the uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 733. The receiver terminal is always considered to be oriented towards its best server. or mobile Mi. TX i  ic  • NR UL – Seg : Segmented uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 733. the receiver is oriented towards the best server just as in the case of LOS. subscriber. or mobile Mi Mi as calculated in "Permutation Zone Selection" on page 714. In the case of NLOS between the receiver and the best server. Finally.4. • M PC : Power control margin defined in the global network settings.Atoll 3. • T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used bythe cell TXi(ic). subscriber. Atoll calculates the uplink carrier to noise ratio as explained in "Traffic C/N Calculation (UL)" on page 734. • P Min : Minimum transmission power of the terminal used by the pixel. subscriber. • B UL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. Next.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Output • M i CNR UL : Uplink C/N from a pixel. The uplink noise rise can be set by the user manually for each cell or calculated using Monte Carlo simulations. • Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel. First. • Mobility  M i  : Mobility used for the calculations. subscriber. • Mi B UL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. or mobile Mi. or mobile Mi at it serving cell TXi(ic). subscriber. or mobile Mi at it serving cell TXi(ic) as calculated in "Traffic C/N Calculation (UL)" on page 734. or mobile as explained in "Traffic Signal Level Calculation (UL)" on page 729. 737 . except when the "Lock Status" is set to "Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited. • P Max : Maximum transmission power of the terminal used by the pixel. Atoll does not try to find the direction of the strongest signal. Input Mi • CNR UL : Uplink C/N from a pixel. • T B – Lowest : Bearer selection threshold of the lowest bearer in the WiMAX equipment assigned to the cell TXi(ic). • N SC TX i  ic  Mi PZ UL : Number of subchannels per channel defined for the uplink permutation zone assigned to the pixel. or mobile Mi. 10. TX  ic  i • T AMS : AMS threshold defined for the cell TXi(ic). or mobile Mi. subscriber.10 Traffic C/(I+N) and Bearer Calculation (UL) The carrier signal to interference and noise ratio is calculated in three steps. subscriber. TX i  ic  • N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic).3. • NR UL TX i  ic  : Non-segmented uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 733. subscriber. subscriber.6. subscriber. subscriber. or mobile at its serving cell using the effective power of the terminal used by the pixel. or mobile Mi as calculated in "Permutation Zone Selection" on page 714. determines the uplink C/(I+N) by dividing the previously calculated uplink C/N by the uplink noise rise value of the cell as calculated in "Noise Rise Calculation (UL)" on page 733. Mi Mi Mi Mi or mobile Mi. M i PZ UL = 8 • N SC  Seg : Number of subchannels per segment for the first uplink PUSC permutation zone. • M i N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel. the selected bearer is the one with the highest uplink peak MAC channel throughput as calculated in "Channel Throughput. Victim and interfering mobiles are generated by a time-slot scenario as explained in "Simulation Process" on page 699. G Div – UL .ini file. see the Administrator Manual. Allocated Bandwidth Throughput. • Peak MAC Throughput From among the bearers available for selection. For more information. • Coverage predictions: CINR UL    = CNR UL – NR UL M i M i TX  ic  i  Bearer Determination: The bearers available for selection in the cell TXi(ic)’s WiMAX equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment).3. or mobile Mi covered by the segmented zone in the interfered cell TXi(ic): Mi Mi TX i  ic  CINR UL = CNR UL – NRUL – Seg • With smart antennas: • Monte Carlo simulations: The uplink C/(I+N) is calculated as described in the section "Beamforming Smart Antenna Models" on page 43. Therefore. the bearers available for selection are all the bearers defined in the WiMAX equipment for which the following is true: Mi TX i  ic  UL Mi Mi TX i  ic  UL Mi T B – G Div – UL – G Div  CINR UL and T B – G Div – UL – G Div  CINR UL    The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic). M i Mobility  M i  .   UL The additional diversity gain defined for the clutter class of the pixel. SU-MIMO diversity or MU-MIMO diversity gain. corresponding to the bearer is applied to its selection threshold. The gain is read from the properties of the M i TX  ic  i M i WiMAX equipment assigned to the cell TXi(ic) for N Ant – RX . the selected bearer is the one with the highest index. or mobile Mi covered by the non-segmented zone in the interfered cell TXi(ic): Mi Mi TX i  ic  CINR UL = CNR UL – NRUL For any pixel. subscriber. if the corresponding option has been set in the Atoll. the STTD/MRC. • Whose selection thresholds are less than the uplink C/(I+N) at Mi: T B  CINR UL and T B  CINR UL    Mi Mi Mi Mi TX  ic  i If the cell supports MIMO.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks • ©Forsk 2015 M i BLER  BUL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell TXi(ic). N Ant – TX . • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. subscriber. subscriber. and Per-User Throughput Calculation" on page 743. the subchannel allocation mode of PZ UL . or mobile Mi G Div is also applied.Atoll 3. Cell Capacity. • 738 Effective MAC Throughput . or mobile Mi at a cell TXi(ic) is calculated as follows: • Without smart antennas: For any pixel. subscriber. BLER  B UL . Calculations The uplink C/(I+N) for any pixel. • Bearer Index From among the bearers available for selection. and MU-MIMO Diversity Gain: Once the bearer is known. and using 6 it would only get access to the second best.e. Although using 4 subchannels. subscribers. it will be assigned 5 subchannels as the used uplink bandwidth. as a mobile moves from good to bad radio conditions.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 From among the bearers available for selection. • Best Bearer The bandwidth used for transmission by a mobile is reduced in order to improve the uplink C/(I+N) enough to access the best bearer.. or mobiles in the uplink. N SC . As there is no reduction in the bandwidth used for transmission. or mobile Mi covered by a non-segmented permutation zone in the interfered cell TXi(ic). the uplink bandwidth is not reduced to 4 because it does not provide any gain in terms of the bearer. and Per-User Throughput Calculation" on page 743. SU-MIMO diversity.. MIMO – STTD/MRC. Uplink Subchannelisation: The uplink subchannelisation depends on the uplink bandwidth allocation target defined for the scheduler used by the cell TXi(ic). its uplink C/(I+N) will be better than when using 5. For example. or MU-MIMO diversity gain corresponding to the selected bearer. Subchannelisation is performed for all the pixels. the selected bearer is the one with the highest uplink effective MAC channel throughput as calculated in "Channel Throughput. i. • Full Bandwidth Full channel width is used by each mobile in the uplink.Atoll 3.e. subscriber. subscriber. and Min N SC – UL  Service  Mi Mi PZ UL = 8  N SC – UL  N SC  Seg for any pixel. the mobile already has the best bearer using 5 subchannels. The calculation of the gain introduced by the bandwidth reduction is explained below. i. For example. The gain related to this bandwidth reduction is applied to the uplink C/(I+N): Mi Mi CINR UL Final Where  PZUL   N SC  = CINR UL+ 10  Log  ----------------  NMi  All SC SC – UL   Mi Min N SC – UL  Service   M i N SC – UL PZ  N SC Mi UL for any pixel. if using 5 subchannels. or with the highest effective MAC throughput. Uplink Power Control: 739 . and may reduce the number of used subchannels in order to satisfy the selected target.e. Allocated Bandwidth Throughput. The uplink subchannelisation may result in the use of a number of subchannels which is less than the total number of subchannels associated with the permutation zone. with the highest peak MAC throughput. i. • Maintain Connection The bandwidth used for transmission by a mobile is reduced only if the uplink C/(I+N) is not enough to even access the lowest bearer. the uplink C/(I+N) calculated above becomes: Mi Mi TX i  ic  UL CNR UL = CNR UL + G Div – UL + G Div Mi TX i  ic  Mi UL CINR UL = CINR UL + G Div – UL + G Div and Mi TX i  ic  Mi UL CINR UL    = CINR UL    + G Div – UL + G Div TX i  ic  Where G Div – UL is the STTD/MRC. The uplink C/(I+N) calculated above is given for the total number of subchannels associated with the Mi PZ UL permutation zone. SU-MIMO Diversity. Cell Capacity. or mobile Mi covered by the segmented uplink PUSC zone in the interfered cell TXi(ic). bearer with the highest index. The calculation of the gain introduced by the subchannelisation is explained below. The definition of the best bearer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic).3.. there is no gain in the uplink C/(I+N). a mobile is able to access the best bearer. the number of subchannels used by it for transmission in uplink are reduced one by one in order to improve the uplink C/(I+N). Allocated Bandwidth Throughput. from the WiMAX equipment assigned to the cell TXi(ic).3. Mi Mi M i 10.7. Final The pixel.Atoll 3.4. • N SC – UL : Number of subchannels used by the pixel.1 Calculation of Sampling Frequency Input TX  ic  i • f Sampling : Sampling factor defined for the frequency band of the cell TXi(ic).0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Once the subchannelisation is performed. subscriber. subscriber. Output Mi Mi • CINR UL or CINR UL    : Uplink C/(I+N) from a pixel. or mobile Mi in the uplink. or mobile Mi reduces its transmission power so that the uplink C/(I+N) from it at its cell is just enough to get the selected bearer. subscriber.e. • B UL : Bearer assigned to the pixel. Atoll continues to work with the C/(I+N) given by the subchannelisation. . subscriber.4. or mobile Mi as follows: Mi M i  TX i  ic   Mi    Mi  P Eff = Max  PMax –  CINR UL –  T M + M PC   P Min i    B   UL Mi Mi CINR UL is calculated again using P Eff . The following sections describe how the cell capacities are calculated for TDD and FDD networks.7. subscriber. where T B UL TX i  ic  Mi B UL is the bearer selection threshold. Cell Capacity. or mobile Mi. CINR UL = CINR UL .. subscriber. or mobile Mi. • P Eff : Effective transmission power of the terminal used by the pixel. • W Channel : Channel bandwidth of the cell TXi(ic). The total cell resources can be calculated separately for the downlink and the uplink subframes. for the bearer selected for the pixel. TX i  ic  Calculations Atoll determines the sampling frequency as follows: TX i  ic  TX  ic  i F Sampling 6  W Channel  10  -  8000 = Floor  f Sampling  ----------------------------------8000   Output • 740 TX i  ic  F Sampling : Sampling frequency for the cell TXi(ic). The transmission power of Mi is reduced to determine the effective transmission power from the pixel. and Per-User Throughput Calculation" on page 743.4.7 Throughput Calculation Throughputs are calculated in two steps. • • Calculation of uplink and downlink total resources in a cell as explained in "Calculation of Total Cell Resources" on page 740. or mobile Mi at it serving cell TXi(ic).1. M M i i i. or mobile Mi in the uplink after subchannelisation. 10.1 Calculation of Total Cell Resources The total amount of resources in a cell is the number of modulation symbols that can be used for data transfer in each frame. 10. Calculation of throughputs as explained in "Channel Throughput. If with P Mi Mi Mi = P Max AND CINR UL  T TX i  ic  Mi + M PC . subscriber. Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 10.4.7.1.2 Calculation of Symbol Duration Input TX  ic  i • F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on page 740. • N SCa – Total : Total number of subcarriers defined for the frame configuration of a cell TXi(ic). • r CP TX i  ic  TX i  ic  : Cyclic prefix ratio defined for the frame configuration of TXi(ic) or, otherwise, in the global network settings. Calculations From the sampling frequency, Atoll determines the inter-subcarrier spacing. F TX  ic  i TX i  ic  –3 F Sampling  10 = ------------------------------------TX i  ic  N SCa – Total Atoll calculates the useful symbol duration. TX i  ic  1 D Sym – Useful = ------------------TX  ic  i F And, the duration of the cyclic prefix. TX i  ic  D CP TX i  ic  r CP = -------------F Adding the Cyclic prefix ratio to the useful symbol duration, Atoll determines the total symbol duration. TX i  ic  TX i  ic  TX i  ic  D Symbol = D Sym – Useful + D CP Output • 10.4.7.1.3 TX i  ic  D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic). Calculation of Total Cell Resources - TDD Networks Input • D Frame : Frame duration. • D TTG : TTG duration. • D RTG : RTG duration. • D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic) as calculated in "Calculation of Symbol Duration" on page 741. • r DL – Frame : DL ratio. • N SD – DL : Number of symbol durations that correspond to the downlink subframe. • N SD – UL : Number of symbol durations that correspond to the uplink subframe. • O Fixed : Downlink fixed overhead. • O Variable : Downlink variable overhead. • O Fixed : Uplink fixed overhead. • O Variable : Uplink variable overhead. • • TDD TDD TX i  ic  TDD TDD TDD DL DL UL UL Mi PZ DL N SCa – Data : Number of data subcarriers of the downlink permutation zone of a cell TXi(ic) assigned to Mi. Mi PZ UL N SCa – Data : Number of data subcarriers of the uplink permutation zone of a cell TXi(ic) assigned to Mi. 741 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Calculations The downlink and the uplink subframes of a TDD frame are separated in time by the TTG and the RTG time guards. First of all, Atoll calculates the useful frame duration by removing the TTG and RTG from the frame duration: Used TDD TDD D Frame = D Frame – D TTG – D RTG Then, Atoll calculates the frame duration in terms of number of symbol durations:  D Used  TX  ic  i Frame  N  SD – Used   Frame = Floor  ----------------TX i  ic    D Symbol Next, Atoll calculates the downlink and uplink cell capacities as follows: Downlink Subframe: Atoll calculates the number of symbol durations in the downlink subframe excluding the fixed overhead defined in the global network settings: TX  ic  TX  ic  i i TDD DL N  SD – DL   Subframe = RoundUp  N SD – Used   Frame  r DL – Frame – O Fixed if DL:UL ratio is defined in percentage. TDD TX i  ic  N SD – DL  TXi  ic   DL Or N  SD – DL   Subframe = RoundUp  N  SD – Used   Frame  ----------------------------------------- – O Fixed if DL:UL ratio is defined in TDD TDD  N SD – DL + N SD – UL fraction. The RoundUp function rounds a float value up to the nearest integer value. The total number of symbols in the downlink subframe after removing the variable overhead is: TX i  ic  R DL TX i  ic  = N  Sym – DL   Subframe Mi   DL PZ DL O Variable   TXi  ic  = Floor  N  SD – DL   Subframe  N SCa – Data   1 – --------------------   100      Uplink Subframe: Atoll calculates the number of symbol durations in the uplink subframe excluding the fixed overhead defined in the global network settings: TX  ic  TX  ic  i i TDD UL N  SD – UL   Subframe = RoundDown  N SD – Used   Frame   1 – r DL – Frame  – O Fixed percentage. if DL:UL ratio is defined in TDD TX i  ic   TX i  ic  N SD – UL  UL Or N  SD – UL   Subframe = RoundDown  N SD – Used   Frame  ----------------------------------------- – O Fixed if DL:UL ratio is defined in TDD TDD  N SD – DL + N SD – UL fraction. The RoundDown function rounds a float value down to the nearest integer value. The total number of symbols in the uplink subframe after removing the variable overhead is: TX  ic  i R UL = TX  ic  i N  Sym – UL   Subframe Mi   UL PZ O Variable   TX i  ic  UL  = Floor  N  SD – UL   Subframe  N SCa – Data   1 – ---------------------  100     Output 10.4.7.1.4 TX i  ic  • R DL • R UL TX i  ic  TX i  ic  = N  Sym – DL   Subframe : Amount of downlink resources in the cell TXi(ic). TX i  ic  = N  Sym – UL   Subframe : Amount of uplink resources in the cell TXi(ic). Calculation of Total Cell Resources - FDD Networks The total cell resources calculation is the same for downlink and uplink subframes in FDD networks. Therefore, the symbol X is used to represent DL or UL in the expressions below. 742 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Input • D Frame : Frame duration. • D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic) as calculated in "Calculation of Symbol Duration" on page 741. • O Fixed : Downlink or uplink fixed overhead. • O Variable : Downlink or uplink variable overhead. • TX  ic  i X X Mi PZ X N SCa – Data : Number of data subcarriers of the downlink or uplink permutation zone of a cell TXi(ic) assigned to Mi. Calculations There are no transmit and receive time guards in FDD systems. Therefore, the downlink and the uplink subframe durations are the same as the frame duration. X D Subframe = D Frame The subframe durations in terms of the number of symbol durations excluding the fixed overheads are:  DX  TX  ic  i Subframe - – O XFixed N  SD – X   Subframe = Floor  ---------------------TX i  ic    D Symbol  The total numbers of symbols in the downlink or uplink subframes after removing the variable overheads are: TX i  ic  RX TX i  ic  = N  Sym – X   Subframe Mi   X PZ X O Variable   TXi  ic  = Floor  N  SD – X   Subframe  N SCa – Data   1 – --------------------   100      Output • TX i  ic  TX i  ic  = N  Sym – X   Subframe : Amount of downlink or uplink resources in the cell TXi(ic). RX 10.4.7.2 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-User Throughput Calculation Channel throughputs are calculated for the entire channel resources allocated to the pixel, subscriber, or mobile Mi. Cell capacities are similar to channel throughputs but upper-bound by the maximum downlink and uplink traffic loads. Allocated bandwidth throughputs are calculated for the number of used subchannels in uplink allocated to the pixel, subscriber, or mobile Mi. Per-user throughputs are calculated by dividing the downlink cell capacities or uplink allocated bandwidth throughputs by the average number of downlink or uplink users defined for the cell, respectively. Input TX i  ic  • TL DL – Max : Maximum downlink traffic load for the cell TXi(ic). • TL UL – Max : Maximum uplink traffic load for the cell TXi(ic). • R DL TX i  ic  TX i  ic  : Amount of downlink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on page 740. TX i  ic  • R UL • page 740.  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the downlink in : Amount of uplink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on • "Traffic and Pilot C/(I+N) and Bearer Calculation (DL)" on page 727.  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the uplink in i B DL i B UL "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737. 743 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks • • ©Forsk 2015 D Frame : Frame duration. f Segment – DL : Downlink segmenting factor for the first downlink PUSC zone as calculated in "Effective Traffic and Pilot Interference Calculation (DL)" on page 722. TX  ic  i • CNR Preamble : Preamble C/N the cell TXi(ic) as calculated in "Preamble C/N Calculation" on page 712. • T AMS : AMS threshold defined for the cell TXi(ic). • T MU – MIMO : MU-MIMO threshold defined for the cell TXi(ic). • G MU – MIMO : MU-MIMO gain defined for the cell TXi(ic). • • • TX  ic  i TX i  ic  TX i  ic  TX  ic  M i i BLER  BDL : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the WiMAX equipment   assigned to the terminal used by the pixel, subscriber, or mobile Mi. M M i i BLER  BUL : Uplink block error rate read from the BLER vs. CINR UL graph available in the WiMAX equipment assigned   to the cell TXi(ic). Mi f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the pixel, subscriber, or mobile Mi. • • M i TP Offset : Throughput offset defined in the properties of the service used by the pixel, subscriber, or mobile Mi. Mi PZ UL N SC : Number of subchannels per channel defined for the uplink permutation zone assigned to the pixel, subscriber, or mobile Mi as calculated in "Permutation Zone Selection" on page 714. • Mi N SC – UL : Number of uplink subchannels after subchannelisation with which the pixel, subscriber, or mobile Mi can get the highest available bearer, as calculated in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737. TX i  ic  • N Users – DL : Number of users connected to the cell TXi(ic) in downlink. • N Users – UL : Number of users connected to the cell TXi(ic) in uplink. TX i  ic  Calculations Downlink: TX i  ic  • Mi R DL  M i B DL Peak MAC Channel Throughput: CTP P – DL = --------------------------------D Frame In the above formula, the actual value of D Frame is used to calculate the channel throughput for coverage predictions, while D Frame = 1 sec for Monte Carlo simulations. TX  ic  i For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG – DL read from the scheduler properties for the Mobility  M i  and the number of users connected to the cell in downlink. TX i  ic  R DL Mi  M B i TX  ic  CTP P – DL i DL = -------------------------------- G MUG – DL D Frame TX  ic  i M i Max G MUG – DL = 1 if CINR Traffic  CINR MUG If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph, it is interpolated from the gain values available for the numbers of users just less than and just greater than the actual number of users. Downlink Segmentation: M i If the permutation zone assigned to the pixel, subscriber, or mobile Mi is the first downlink PUSC zone ( PZ DL = 0 ) and it is segmented, the channel throughput is calculated as: 744 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 M M i i CTP P – DL = CTP P – DL  f Segment – DL MIMO – SU-MIMO Gain: If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports SU-MIMO or AMS, SU-MIMO gain Max G SU – MIMO is applied to the bearer efficiency. The gain is read from the properties of the WiMAX equipment assigned to the pixel, subscriber, or mobile Mi for: TX  ic  i • N Ant – TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic). • N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel, subscriber, Mi or mobile Mi. • Mobility  M i  : Mobility used for the calculations. • Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel, subscriber, or Mi mobile Mi as calculated in "Permutation Zone Selection" on page 714. • M i B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink as explained in "Traffic and Pilot C/ (I+N) and Bearer Calculation (DL)" on page 727. • M i BLER  B DL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the TX  ic  i terminal used by the pixel, subscriber, or mobile Mi. BLER is determined for CINR Traffic . Atoll also takes into account the SU-MIMO Gain Factor f SU – MIMO defined for the clutter class where the pixel, subscriber, or mobile Mi is located. In case of SU-MIMO:  In case of AMS:  Mi Mi B DL =  B DL =  Max Mi B DL   1 + f SU – MIMO  G SU – MIMO – 1   TX i  ic  Max Mi TX i  ic    1 + f SU – MIMO  G SU – MIMO – 1   if CNR Preamble  T AMS TX i  ic  TX i  ic  or CINR Preamble  T AMS B DL If the Max SU-MIMO Gain for the exact value of the C/(I+N) is not available in the table, it is interpolated from the gain values available for the C/(I+N) just less than and just greater than the actual C/(I+N). • M M M i i i Effective MAC Channel Throughput: CTP E – DL = CTP P – DL   1 – BLER  B DL     M i CTP A – DL = M Mi i CTP E – DL M f TP – Scaling i - – TP Offset  -----------------------100 • Application Channel Throughput: • Peak MAC Cell Capacity: Cap P – DL = CTP P – DL  TL DL – Max • i i i Effective MAC Cell Capacity: Cap E – DL = Cap P – DL   1 – BLER  B DL     • • Mi TX i  ic  Mi M Application Cell Capacity: M i Cap A – DL M = Mi M i Cap E – DL Mi Peak MAC Throughput per User: PUTP P – DL M M f TP – Scaling i - – TP Offset  -----------------------100 Mi Cap P – DL = ----------------------TX i  ic  N Users – DL Mi • • Mi Cap E – DL Effective MAC Throughput per User: PUTP E – DL = ----------------------TX i  ic  N Users – DL Mi Application Throughput per User: PUTP A – DL Mi Mi f TP – Scaling - – TP Offset = PUTP E – DL  -----------------------100 Mi 745 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Uplink: TX  ic  i • Peak MAC Channel Throughput: R UL M i CTP P – UL  M B i UL = --------------------------------D Frame In the above formula, the actual value of D Frame is used to calculate the channel throughput for coverage predictions, while D Frame = 1 sec for Monte Carlo simulations. TX i  ic  For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG – UL read from the scheduler properties for the Mobility  M i  and the number of users connected to the cell in uplink. TX i  ic  R UL Mi  B Mi TX  ic  CTP P – UL i UL = -------------------------------- G MUG – UL D Frame TX i  ic  Mi Max G MUG – UL = 1 if CINR UL  CINR MUG If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph, it is interpolated from the gain values available for the numbers of users just less than and just greater than the actual number of users. MIMO – SU-MIMO Gain: If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports SU-MIMO or AMS, SU-MIMO gain Max G SU – MIMO is applied to the bearer efficiency. The gain is read from the properties of the WiMAX equipment assigned to the cell TXi(ic) for: • Mi N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel, subscriber, or mobile Mi. TX  ic  i • N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic). • Mobility  M i  : Mobility used for the calculations. • Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel, subscriber, or Mi mobile Mi as calculated in "Permutation Zone Selection" on page 714. • Mi B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink as explained in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737. • M i BLER  B UL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell   Mi TXi(ic). BLER is determined for CINR UL . Atoll also takes into account the SU-MIMO Gain Factor f SU – MIMO defined for the clutter class where the pixel, subscriber, or mobile Mi is located. In case of SU-MIMO:  Mi =  B UL In case of AMS:  Mi B UL =  Max Mi   1 + fSU – MIMO  G SU – MIMO – 1   B UL Max Mi TX i  ic  TX i  ic    1 + f SU – MIMO  G SU – MIMO – 1   if CNR Preamble  T AMS TX i  ic  TX i  ic  or CINR Preamble  T AMS B UL If the Max SU-MIMO Gain for the exact value of the C/(I+N) is not available in the table, it is interpolated from the gain values available for the C/(I+N) just less than and just greater than the actual C/(I+N). MIMO – MU-MIMO Gain (for uplink throughput coverage predictions only): If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports MU-MIMO and TX  ic  i TX  ic  i TX  ic  i TX  ic  i CNR Preamble  T MU – MIMO and N Ant – RX  2 , the MU-MIMO gain G MU – MIMO is applied to the channel throughput. The MU-MIMO gain is read from the properties of the cell TXi(ic). 746 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 M M i i TX  ic  i CTP P – UL = CTP P – UL  G MU – MIMO • M M M i i i Effective MAC Channel Throughput: CTP E – UL = CTP P – UL   1 – BLER  B UL     M i Mi f TP – Scaling - – TP Offset = CTP E – UL  -----------------------100 Mi Mi • Application Channel Throughput: CTP A – UL • Peak MAC Cell Capacity: Cap P – UL = CTP P – UL  TL UL – Max • i i i Effective MAC Cell Capacity: Cap E – UL = Cap P – UL   1 – BLER  B UL     • • Mi TX i  ic  Mi M Mi Application Cell Capacity: Cap A – UL M M Mi Mi f TP – Scaling - – TP Offset = Cap E – UL  -----------------------100 Mi Mi Peak MAC Allocated Bandwidth Throughput: ABTP P – UL Mi N SC – UL = CTP P – UL  ----------------M Mi PZ N SC • • • • • M i UL M M i i i Effective MAC Allocated Bandwidth Throughput: ABTP E – UL = ABTP P – UL   1 – BLER  B UL     Mi Application Allocated Bandwidth Throughput: ABTPA – UL Mi Peak MAC Throughput per User: PUTP P – UL  Cap M i  M P – UL - ABTPP –i UL = Min  ----------------------TX i  ic    N Users – UL  Mi Effective MAC Throughput per User: PUTP E – UL Mi Mi Mi f TP – Scaling - – TPOffset = ABTP E – UL  -----------------------100 Mi Application Throughput per User: PUTP A – UL  Cap Mi  M E – UL - ABTP E –i UL = Min  ---------------------- TXi  ic    N Users – UL  Mi Mi f TP – Scaling - – TP Offset = PUTP E – UL  -----------------------100 Mi Output Mi • CTP P – DL : Downlink peak MAC channel throughput at the pixel, subscriber, or mobile Mi. • CTP E – DL : Downlink effective MAC channel throughput at the pixel, subscriber, or mobile Mi. • CTP A – DL : Downlink application channel throughput at the pixel, subscriber, or mobile Mi. • Cap P – DL : Downlink peak MAC cell capacity at the pixel, subscriber, or mobile Mi. • Cap E – DL : Downlink effective MAC cell capacity at the pixel, subscriber, or mobile Mi. • Cap A – DL : Downlink application cell capacity at the pixel, subscriber, or mobile Mi. • PUTP P – DL : Downlink peak MAC throughput per user at the pixel, subscriber, or mobile Mi. • PUTP E – DL : Downlink effective MAC throughput per user at the pixel, subscriber, or mobile Mi. • PUTP A – DL : Downlink application throughput per user at the pixel, subscriber, or mobile Mi. • CTP P – UL : Uplink peak MAC channel throughput at the pixel, subscriber, or mobile Mi. • CTP E – UL : Uplink effective MAC channel throughput at the pixel, subscriber, or mobile Mi. • CTP A – UL : Uplink application channel throughput at the pixel, subscriber, or mobile Mi. • Cap P – UL : Uplink peak MAC cell capacity at the pixel, subscriber, or mobile Mi. Mi M i Mi Mi Mi Mi M i Mi Mi Mi Mi M i 747 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks M ©Forsk 2015 i • Cap E – UL : Uplink effective MAC cell capacity at the pixel, subscriber, or mobile Mi. • Cap A – UL : Uplink application cell capacity at the pixel, subscriber, or mobile Mi. • ABTP P – UL : Uplink peak MAC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi. • ABTP E – UL : Uplink effective MAC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi. • ABTP A – UL : Uplink application allocated bandwidth throughput at the pixel, subscriber, or mobile Mi. • PUTP P – UL : Uplink peak MAC throughput per user at the pixel, subscriber, or mobile Mi. • PUTP E – UL : Uplink effective MAC throughput per user at the pixel, subscriber, or mobile Mi. • PUTP A – UL : Uplink application throughput per user at the pixel, subscriber, or mobile Mi. M i M i Mi Mi Mi Mi Mi 10.4.8 Scheduling and Radio Resource Management Atoll WiMAX BWA module includes a number of scheduling methods which can be used for scheduling and radio resource allocation during Monte Carlo simulations. These resource allocation algorithms are explained in "Scheduling and Radio Resource Allocation" on page 748 and the calculation of user throughputs is explained in "User Throughput Calculation" on page 757. 10.4.8.1 Scheduling and Radio Resource Allocation Input TX i  ic  • TL DL – Max : Maximum downlink traffic load for the cell TXi(ic). • TL UL – Max : Maximum uplink traffic load for the cell TXi(ic). • N Users – Max : Maximum number of users defined for the cell TXi(ic). • QoS TX i  ic  TX i  ic  M i M i : QoS class of the service (UGS, ErtPS, rtPS, nrtPS, or Best Effort) accessed by a mobile Mi. • p • TPD Min – DL : Downlink minimum throughput demand for the service accessed by a mobile Mi. • TPD Min – UL : Uplink minimum throughput demand for the service accessed by a mobile Mi. • TPD Max – DL : Downlink maximum throughput demand for the service accessed by a mobile Mi. • TPD Max – UL : Uplink maximum throughput demand for the service accessed by a mobile Mi. • • : Priority of the service accessed by a mobile Mi. M M i i Mi Mi TX  ic  M i i BLER  BDL : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the WiMAX equipment   assigned to the terminal used by the mobile Mi. M M i i BLER  BUL : Uplink block error rate read from the BLER vs. CINR UL graph available in the WiMAX equipment assigned to the cell TXi(ic). Mi • f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile Mi. • TP Offset : Throughput offset defined in the properties of the service used by the mobile Mi. • CTP P – DL : Downlink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on Mi Mi page 740. • Mi CTP E – DL : Downlink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on page 740. 748 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 • M i CTP P – UL : Uplink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on page 740. • M i CTP E – UL : Uplink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on page 740. • Mi ABTP P – UL : Uplink peak MAC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput Calculation" on page 740. • QoS f Bias : Bias factor defined for the Biased (QoS Class) scheduling method. Calculations The following calculations are described for any cell TXi(ic) containing the users Mi for which it is the best server. Mobile Selection: TX  ic  i The scheduler selects N Users mobiles for the scheduling and RRM process. If the Monte Carlo user distribution has generated TX i  ic  a number of users which is less than N Users – Max , the scheduler keeps all the mobiles generated for the cell TXi(ic). TX  ic  TX  ic  TX  ic  i i i N Users = Min  N Users – Max N Users – Generated   TX  ic  i Sel  N Users are selected for RRM by the scheduler. For a cell, mobiles M i Calculation of Actual Minimum and Maximum Throughput Demands: Depending on the selected target throughput of the scheduler assigned to the cell TXi(ic), the actual minimum and maximum throughput demands can be considered as the peak MAC, effective MAC, or application throughput. Therefore: • Target Throughput = Peak MAC Throughput Sel Mi Sel Mi Downlink: TPD Min – DL , TPD Max – DL Sel Sel M M M   i i i Uplink: TPD Min – UL , Min  TPD Max – UL ABTP P – UL   • Target Throughput = Effective MAC Throughput Sel Mi Sel Mi Sel Mi Sel Mi TPD Min – DL TPD Max – DL Downlink: TPD Min – DL = --------------------------------------------- , TPD Max – DL = --------------------------------------------Sel Sel   Mi     Mi    1 – BLER  B DL    1 – BLER  BDL         Sel Sel Sel Mi Sel Mi Mi TPD Min – UL Uplink: TPD Min – UL = --------------------------------------------- , TPD Max – UL Sel   Mi    1 – BLER  BUL      • Mi Mi   Min  TPD Max – UL ABTP P – UL   = ------------------------------------------------------------------------Sel   Mi    1 – BLER  B UL      Target Throughput = Application Throughput Sel Mi Sel Mi Mi Sel Sel Mi Mi Mi TPD Min – DL + TP Offset TPD Max – DL + TP Offset - , TPD Max – DL = ----------------------------------------------------------------------------Downlink: TPD Min – DL = ----------------------------------------------------------------------------Sel Sel   Mi   Mi   Mi   Mi  1 – BLER  B DL    f TP – Scaling  1 – BLER  B DL    f TP – Scaling       Sel Mi Sel Mi Mi TPD Min – UL + TP Offset Uplink: TPD Min – UL = -----------------------------------------------------------------------------, Sel   Mi   Mi  1 – BLER  BUL    f TP – Scaling    749 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Sel Sel M i TPD Max – UL M M M  i i  i Min  TPD Max – UL ABTP P – UL + TP Offset   = -------------------------------------------------------------------------------------------------Sel   Mi   Mi  1 – BLER  BUL    f TP – Scaling    The Min() function selects the lower of the two values. This calculation is performed in order to limit the maximum uplink throughput demand to the maximum throughput that a user can get in uplink using the allocated bandwidth (number of used subchannels) calculated for it in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737. Resource Allocation for Minimum Throughput Demands: Sel 1. For the QoS classes UGS, ErtPS, rtPS, and nrtPS, Atoll sorts the M i p Sel Mi TX i  ic   N Users in order of decreasing service priority, : Sel Mi QoS 1 Sel Mi p UGS 2 p Sel Mi ... n > p : p : ErtPS : p p : rtPS : p Sel Mi Sel i ... n > p : p : nrtPS N–1 p p TX i  ic  =n Sel Mi =n > 0 ... =0 =n Sel Mi Sel Mi > 0 ... =0 Sel i Sel Mi ... n > p N M > 0 ... =0 Sel Mi Sel Mi M =n Sel Mi Sel Mi ... n > p : Sel Mi > 0 ... =0 TX i  ic  Where N  N Users , if there are some Best Effort users, or N = N Users if there are no Best Effort users selected. Sel 2. Starting with M i Sel = 1 up to M i = N , Atoll allocates the downlink and uplink resources required to satisfy each user’s minimum throughput demands in downlink and uplink as follows: Sel Mi Sel Mi Sel Mi Sel Mi TPD Min – DL TPD Min – UL R Min – DL = -------------------------- and R Min – UL = -------------------------Sel Mi CTP P – DL Sel Mi CTP P – UL 3. Atoll stops the resource allocation in downlink or uplink, 750 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 • When/If in downlink M  M Sel i TX  ic  i R Min – DL = TL DL – Max , i.e., the resources available in downlink have been used up for Sel i satisfying the minimum throughput demands of the mobiles. • When/If in uplink  M Sel i TX  ic  i R Min – UL = TL UL – Max , i.e., the resources available in uplink have been used up for Sel Mi satisfying the minimum throughput demands of the mobiles. 4. Mobiles which are active DL+UL must be able to get their minimum throughput demands in both UL and DL in order to be considered connected DL+UL. If an active DL+UL mobile is only able to get its minimum throughput demand in one direction, it is rejected, and the resources, that were allocated to it in the one direction in which it was able to get a throughput, are allocated to other mobiles. 5. Mobiles which are active UL and whose minimum throughput demand in UL is higher than the uplink allocated Sel Mi Sel Mi bandwidth throughput ( TPD Min – UL  ABTP P – UL ) are rejected due to Resource Saturation. 6. If Sel Mi TX i  ic   RMin – DL  TLDL – Max Sel Mi Sel Mi TX i  ic   RMin – UL  TLUL – Max , and all the minimum throughput resources demanded by or Sel Mi the mobiles have been allocated, Atoll goes to the next step for allocating resources to satisfy the maximum throughput demands. Backhaul Saturation: If at this stage, a site’s downlink or uplink effective MAC aggregate throughput exceeds its maximum downlink or uplink backhaul throughput, respectively, mobiles are rejected one by one due to Backhaul Saturation, starting from the mobile with the lowest priority service, among all the cells of the site in order to reach a downlink or uplink effective MAC aggregate site throughput ≤ the site’s maximum downlink or uplink backhaul throughput. Resource Allocation for Maximum Throughput Demands: For each cell, the remaining cell resources available are: TX  ic  i TX  ic  i  Downlink: R Rem – DL = TL DL – Max – M Sel i R Min – DL Sel Mi TX i  ic  TX i  ic  Uplink: R Rem – UL = TL UL – Max –  M Sel Mi R Min – UL Sel i For each mobile, the throughput demands remaining once the minimum throughput demands have been satisfied are the difference between the maximum and the minimum throughput demands: M Sel i M Sel i M Sel i Downlink: TPD Rem – DL = TPD Max – DL – TPD Min – DL Sel Mi Sel Mi Sel Mi Uplink: TPD Rem – UL = TPD Max – UL – TPD Min – UL For the remaining throughput demands of the mobiles belonging to the QoS classes ErtPS, rtPS, nrtPS, and Best Effort, the following resource allocation methods are available: • Proportional Fair: The goal of this scheduling method is to distribute resources among users fairly in such a way that, on the average, each user gets the highest possible throughput that it can get under the radio conditions at its location. 751 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Sel Let the total number of users belonging to the QoS classes ErtPS, rtPS, nrtPS, and Best Effort, be N  M i TX  ic  i . TX  ic  i a. Each user’s channel throughput is increased by the multi-user diversity gain G MUG – DL or G MUG – UL read from the Sel scheduler properties for the Mobility  M i  assigned to mobile M i and the number of connected users, DL or UL, in the cell TXi(ic) in the iteration k-1. Sel Mi Sel Mi CTP P – DL = CTP P – DL Sel Mi TX i  ic  Without MUG Sel Mi TX i  ic  Sel Mi  G MUG – DL and CTP P – UL = CTP P – UL Without MUG Sel Mi TX i  ic  Max TX i  ic  G MUG – DL = 1 if CINR Traffic  CINR MUG and G MUG – UL = 1 if CINR UL  G MUG – UL Max  CINR MUG . If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph, it is interpolated from the gain values available for the numbers of users just less than and just greater than the actual number of users. b. Atoll divides the remaining resources in the cell into equal parts for each user: TX i  ic  TX i  ic  R Rem – DL R Rem – UL --------------------- and -------------------N N c. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands: Sel Mi M RD Rem – DL Sel i M Sel Sel i Mi TPD Rem – DL TPD Rem – UL = --------------------------- and RD Rem – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the peak channel throughputs at the user’s location. d. The resources allocated to each user by the Proportional Fair scheduling method for satisfying its maximum throughput demands are: Sel Mi R Max – DL TX i  ic  Sel Sel Sel TX i  ic  Mi  Mi  Mi R Rem – DL R Rem – UL - and R Max = Min  RD Rem – DL -------------------– UL = Min  RD Rem – UL --------------------- N N     Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from the remaining resources of the cell, whichever is smaller. e. Atoll stops the resource allocation in downlink or uplink, • When/If in downlink Sel Mi  M TX i  ic  R Max – DL = R Rem – DL , i.e., the resources available in downlink have been used up Sel i for satisfying the maximum throughput demands of the mobiles. • When/If in uplink Sel Mi  RMax – UL = TX i  ic  R Rem – UL , i.e., the resources available in uplink have been used up for Sel Mi satisfying the maximum throughput demands of the mobiles. f. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of remaining users. g. Atoll recalculates the remaining resources as follows: TX i  ic  TX i  ic  R Rem – DL = TL DL – Max –  Sel Mi R Min – DL – Sel Mi TX i  ic  TX i  ic  R Rem – UL = TL UL – Max – Sel Mi Sel Mi Sel Mi  RMin – UL –  RMax – UL Sel Mi 752  Sel Mi R Max – DL and Sel Mi Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 h. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied TX  ic  i TX  ic  i until either R Rem – DL = 0 and R Rem – UL = 0 , or all the maximum throughput demands are satisfied. • Proportional Demand: The goal of this scheduling method is to allocate resources to users weighted according to their remaining throughput demands. Therefore, the user throughputs for users with high throughput demands will be higher than those with low throughput demands. In other words, this scheduler distributes channel throughput between users proportionally to their demands. a. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands: Sel Mi RD Rem – DL Sel Mi Sel Mi Sel Mi TPD Rem – DL TPD Rem – UL = --------------------------- and RD Rem – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the peak channel throughputs at the user’s location. b. Atoll calculates the amount effective remaining resources for the cell of each user to distribute among the users as follows: TX i  ic  R Eff – Rem – DL     Sel Sel Mi TX i  ic  Mi  TXi  ic    TXi  ic   RD Rem – DL and R Eff – Rem – UL = Min  R Rem – UL RD Rem – UL = Min  R Rem – DL     Sel Sel     Mi Mi   c. The resources allocated to each user by the Proportional Demand scheduling method for satisfying its maximum throughput demands are: Sel Mi Sel Mi TX i  ic  Sel Mi Sel Mi TX i  ic  RD Rem – DL RD Rem – UL - and R Max – UL = R Eff – Rem – UL  ---------------------------------R Max – DL = R Eff – Rem – DL  ---------------------------------Sel Sel Mi • Mi  RDRem – DL  RDRem – UL Sel Mi Sel Mi Biased (QoS Class): The goal of this scheduling method is to distribute resources among users of each QoS class fairly in such a way that, on the average, each user gets the highest possible throughput that it can get under the radio conditions at its location. The resources available for allocation to users of each QoS class depend on a bias factor. The QoS Class Bias Factor controls the amount of resources available for each QoS class. Calculation of the Remaining Resources per QoS Class: QoS The bias factor f Bias represents the bias in terms of resources allocated to 1 user of a QoS class with rank r to the resources allocated to 1 user of a QoS class with rank r–1: QoS Sel Mi Sel Mi Sel Mi f Bias R Max – ErtPS R Max – rtPS R Max – nrtPS  = 1 + --------- = -------------------------= -------------------------= -------------------------Sel Sel Sel 100 Mi Mi Mi R Max – rtPS R Max – nrtPS R Max – BE The ranks of QoS classes are: QoS Class QoS Class Rank r QoS ErtPS 1 rtPS 2 nrtPS 3 Best Effort 4 The resources available for the users of each QoS class from among the remaining resources is calculated as follows: 753 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 r TX  ic  i R QoS – DL r 1 QoS 1 QoS N QoS   --- N QoS   --- TX  ic  TX  ic  TX  ic    i i i = R Rem – DL  ------------------------------------------------------- and R QoS – UL = R Rem – UL  ------------------------------------------------------r r 1 QoS 1 QoS N QoS   --- N QoS   ---       All QoS All QoS Resource Allocation: Once the remaining resources available for the users of each QoS class have been determined, the allocation of resources within each QoS class is performed as for the proportional fair scheduler. Sel Let the number of users belonging to a QoS class N QoS  M i . a. Atoll divides the remaining resources of the QoS class into equal parts for each user: TX i  ic  TX i  ic  R QoS – DL R QoS – UL -------------------- and ------------------N QoS N QoS b. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands: Sel Mi Sel Mi RD Rem – DL Sel Mi Sel Mi TPD Rem – DL TPD Rem – UL = --------------------------- and RD Rem – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the peak channel throughputs at the user’s location. c. The resources allocated to each user by the Biased scheduling method for satisfying its maximum throughput demands are: Sel Mi R Max – DL TX i  ic  Sel Sel Sel TX i  ic  Mi  Mi  Mi R QoS – DL R QoS – UL - and R Max = Min  RD Rem – DL ------------------– UL = Min  RD Rem – UL -------------------- N N QoS    QoS  Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from the remaining resources of the QoS class, whichever is smaller. d. Atoll stops the resource allocation for a QoS class in downlink or uplink, • Sel Mi  When/If in downlink TX i  ic  R Max – DL = R QoS – DL , i.e., the resources available in downlink for the QoS class have Sel Mi been used up for satisfying the maximum throughput demands of the mobiles. • When/If in uplink  Sel Mi TX i  ic  R Max – UL = R QoS – UL , i.e., the resources available in uplink for the QoS class have been Sel Mi used up for satisfying the maximum throughput demands of the mobiles. e. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of remaining users. f. Atoll recalculates the remaining resources as follows: TX i  ic  TX i  ic   R QoS – DL = TL DL – Max – Sel Mi R Min – DL – Sel Mi TX i  ic  TX i  ic   R QoS – UL = TL UL – Max –  Sel Mi R Max – DL and Sel Mi Sel Mi R Min – UL – Sel Mi  Sel Mi R Max – UL Sel Mi g. Atoll repeats the all the above steps for the users of the QoS class whose maximum throughput demands have not TX i  ic  TX i  ic  been satisfied until either R QoS – DL = 0 and R QoS – UL = 0 , or all the maximum throughput demands are satisfied. • 754 Max Aggregate Throughput: Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 The goal of this scheduling method is to achieve the maximum aggregate throughput for the cells. This is done by allocating as much resources as needed to mobiles with high C/(I+N) conditions. As mobiles with high C/(I+N) can get higher bearers, and therefore require less amount of resources, more mobiles can therefore be allocated resources in the same frame, and the end-throughput for each cell will be the highest compared to other types of schedulers. Sel a. Atoll sorts the M i TX  ic  i  N Users in order of decreasing downlink or uplink traffic C/(I+N), depending on whether the allocation is being performed for the downlink or for the uplink. b. Starting with the mobile with the highest rank, Atoll allocates the downlink and uplink resources required to satisfy each user’s remaining throughput demands in downlink and uplink as follows: Sel Mi R Max – DL Sel Mi Sel Mi Sel Mi TPD Rem – DL TPD Rem – UL = --------------------------- and R Max – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL c. Atoll stops the resource allocation in downlink or uplink, • When/If in downlink  M Sel i TX  ic  i R Max – DL = R Rem – DL , i.e., the resources available in downlink have been used up Sel Mi for satisfying the maximum throughput demands of the mobiles. • When/If in uplink  Sel Mi TX i  ic  R Max – UL = R Rem – UL , i.e., the resources available in uplink have been used up for Sel Mi satisfying the maximum throughput demands of the mobiles. • Round Robin: The goal of this scheduling method is to allocate equal resources to users fairly. Sel Let the total number of users belonging to the QoS classes ErtPS, rtPS, nrtPS, and Best Effort, be N  M i . a. Atoll divides the remaining resources in the cell into equal parts for each user: TX i  ic  TX i  ic  R Rem – DL R Rem – UL --------------------- and -------------------N N b. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands: Sel Mi RD Rem – DL M Sel i M Sel Sel i Mi TPD Rem – DL TPD Rem – UL = --------------------------- and RD Rem – UL = --------------------------M Sel i M CTP P – DL Sel i CTP P – UL Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the peak channel throughputs at the user’s location. c. The resources allocated to each user by the Round Robin scheduling method for satisfying its maximum throughput demands are: Sel TX i  ic  Sel Sel Sel TX i  ic  Mi Mi  Mi  Mi R Rem – DL R Rem – UL - and R Max R Max – DL = Min  RD Rem – DL -------------------– UL = Min  RD Rem – UL --------------------- N N     Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from the remaining resources of the cell, whichever is smaller. d. Atoll stops the resource allocation in downlink or uplink, • When/If in downlink  M Sel Mi TX i  ic  R Max – DL = R Rem – DL , i.e., the resources available in downlink have been used up Sel i for satisfying the maximum throughput demands of the mobiles. 755 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks • ©Forsk 2015 When/If in uplink M  M Sel i TX  ic  i R Max – UL = R Rem – UL , i.e., the resources available in uplink have been used up for Sel i satisfying the maximum throughput demands of the mobiles. e. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of remaining users. f. Atoll recalculates the remaining resources as follows: TX i  ic  TX i  ic   R Rem – DL = TL DL – Max – Sel Mi R Min – DL – Sel Mi TX i  ic  TX i  ic  R Rem – UL = TL UL – Max –   Sel Mi R Max – DL and Sel Mi Sel Mi R Min – UL – Sel Mi  Sel Mi R Max – UL Sel Mi g. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied TX  ic  i TX  ic  i until either R Rem – DL = 0 and R Rem – UL = 0 , or all the maximum throughput demands are satisfied. Spatial Multiplexing with Uplink Multi-User MIMO: MU-MIMO lets the system/scheduler work with two parallel WiMAX frames (1 for each antenna). Therefore, a mobile connected to antenna 1 creates a corresponding resource availability on antenna 2. This resources made available on antenna 2 can then be assigned to another mobile without any effect on the overall load of the cell. When the second mobile is assigned to antenna 2, the resources allocated to it overlap with the resources made available by the first mobile on antenna 1. If the second mobile is allocated more resources than the first one made available, the second mobile will create resource availability on antenna 1. Each new mobile is either connected to antenna 1 or antenna 2. The part of the mobile’s resources which are not coupled with resources allocated to another mobile on the other antenna is called the real resource consumption. The part of the mobile’s resources which are coupled with the resources allocated to another mobile on the other antenna is called the virtual resource consumption. MU-MIMO can be used if the permutation zone assigned to the pixel, subscriber, or mobile Mi supports MU-MIMO, TX i  ic  TX i  ic  TX i  ic  CNR Preamble  T MU – MIMO , and N Ant – RX  2 . Let i be the index of connected MU-MIMO mobiles: i = 1 to N MU – MIMO Each mobile M i MU – MIMO Mi = 0 RR UL MU – MIMO Mi has a corresponding traffic load TL UL MU – MIMO Mi = 0 V UL = 100 % and available virtual resources . The scheduling starts with available real resources = 0 % . i = 0 means no MU-MIMO mobile has yet been scheduled. MU – MIMO The virtual resource consumption of a mobile M i MU – MIMO Mi is given by: VC UL MU – MIMO Mi MU – MIMO The real resource consumption of a mobile M i is given by: RC UL MU – MIMO The virtual resources made available by the mobile M i MU – MIMO Mi V UL MU – MIMO Mi – 1 = V UL Saturation occurs when  MU – MIMO Mi – VC UL MU – MIMO Mi RC UL MU – MIMO  Mi = Min  TL UL  MU – MIMO Mi = TL UL MU – MIMO Mi – 1   V UL   MU – MIMO Mi – VC UL are given by: MU – MIMO Mi + RC UL TX i  ic  = TL UL – Max . The following table gives an example: Mobile 756 MU – MIMO Mi TL UL (%) MU – MIMO Mi VC UL (%) MU – MIMO Mi RC UL (%) MU – MIMO Mi V UL M1 10 0 10 10 M2 5 5 0 5 (%) Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 M Mobile MU – MIMO i TL UL (%) M MU – MIMO i VC UL M (%) MU – MIMO i RC UL M MU – MIMO i V UL (%) M3 20 5 15 15 M4 40 15 25 25 … … … … … (%) Backhaul Capacity Limitation: Backhaul overflow ratios are calculated for each site as follows: Sel Sel   Mi   Mi    R Max – DL  CTP E – DL     Sel   M  Site i  = Max  1 ------------------------------------------------------------------------------------------------------- and Sel Sel Mi    Mi  Site  R Min – DL  CTP E – DL   TP BH – DL –    Sel   M  Site i  Site BHOFDL  Sel Sel   Mi   Mi    R Max – UL  CTP E – UL     Sel   M i  Site = Max  1 ------------------------------------------------------------------------------------------------------- Sel Sel Mi    Mi  Site  R Min – UL  CTP E – UL   TP BH – UL –    Sel   M  Site i  Site BHOFUL  Total Amount of Resources Assigned to Each Selected Mobile: Sel Atoll calculates the amounts of downlink and uplink resources allocated to each individual mobile M i (which can also be referred to as the traffic loads of the mobiles) as follows: Sel Sel Mi Downlink: TL DL Sel Mi = R DL Sel Mi Sel Mi R  Mi   Mi Max – DL  CTP P – DL  R Min – DL  CTP P – DL +  -----------------------------------------------Site     BHOF DL  = -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – DL Sel Sel Mi Uplink: TL UL Sel Mi = R UL M Sel i M Sel i R  Mi   Mi Max – UL  CTP P – UL  R Min – UL  CTP P – UL +  -----------------------------------------------Site     BHOF UL  = -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – UL Output • • Sel Mi TL DL Sel Mi TL UL Sel Mi = R DL Sel : Downlink traffic load or the amount of downlink resources allocated to the mobile M i . Sel Mi Sel = R UL : Uplink traffic load or the amount of uplink resources allocated to the mobile M i . 10.4.8.2 User Throughput Calculation User throughputs are calculated for the percentage of resources allocated to each mobile selected by the scheduling for RRM Sel during the Monte Carlo simulations, M i . Input • Sel Mi R DL Sel : Amount of downlink resources allocated to the mobile M i as calculated in "Scheduling and Radio Resource Allocation" on page 748. 757 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks • M ©Forsk 2015 Sel i Sel R UL : Amount of uplink resources allocated to the mobile M i as calculated in "Scheduling and Radio Resource Allocation" on page 748. • Sel i CTP P – DL M Sel : Downlink peak MAC channel throughput at the mobile M i as calculated in "Throughput Calculation" on page 740. • M Sel i Sel CTP P – UL : Uplink peak MAC channel throughput at the mobile M i as calculated in "Throughput Calculation" on page 740. Sel • TX i  ic   Mi  BLER  BDL  : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the WiMAX equipment   Sel assigned to the terminal used by the mobile M i . Sel • Mi  Mi  BLER  BUL  : Uplink block error rate read from the BLER vs. CINR UL graph available in the WiMAX equipment   assigned to the cell TXi(ic). Sel Mi Sel • f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile M i • TP Offset : Throughput offset defined in the properties of the service used by the mobile M i Sel Mi Sel Calculations Downlink: Sel Mi Sel Mi Sel Mi  CTP P – DL • Peak MAC User Throughput: UTP P – DL = R DL • Mi Mi   Mi   Effective MAC User Throughput: UTP E – DL = UTP P – DL   1 – BLER  B DL      Sel • Sel Mi Application User Throughput: UTP A – DL Sel Sel Sel Mi Sel Mi Sel Mi f TP – Scaling - – TP Offset = UTP E – DL  -----------------------100 Uplink: Sel Mi Sel Mi Sel Mi  CTP P – UL • Peak MAC User Throughput: UTP P – UL = R UL • M M   Mi   i i Effective MAC User Throughput: UTP E – UL = UTP P – UL   1 – BLER  B UL      Sel • Sel Mi Sel Sel Mi Sel Sel Mi Sel Mi f TP – Scaling Application User Throughput: UTP A – UL = UTP E – UL  ------------------------- – TP Offset 100 Output 758 Sel Mi Sel • UTP P – DL : Downlink peak MAC user throughput at the pixel, subscriber, or mobile M i • UTP E – DL : Downlink effective MAC user throughput at the pixel, subscriber, or mobile M i . . Sel Mi Sel Sel Mi Sel • UTP A – DL : Downlink application user throughput at the pixel, subscriber, or mobile M i . • UTP P – UL : Uplink peak MAC user throughput at the pixel, subscriber, or mobile M i • UTP E – UL : Uplink effective MAC user throughput at the pixel, subscriber, or mobile M i . • UTP A – UL : Uplink application user throughput at the pixel, subscriber, or mobile M i Sel Mi Sel . Sel Mi Sel Mi Sel Sel . . . Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 10.5 Automatic Planning Algorithms The following sections describe the algorithms for: • • • • • "Automatic Neighbour Planning" on page 759. "Automatic Inter-technology Neighbour Planning" on page 763. "Automatic Frequency Planning Using the AFP" on page 765. "Automatic Preamble Index Planning Using the AFP" on page 767. "Automatic Zone PermBase Planning Using the AFP" on page 771. 10.5.1 Automatic Neighbour Planning The intra-technology neighbour planning algorithm takes into account the cells of all the TBC transmitters. It means that the cells of all the TBC transmitters of your ATL document are potential neighbours. The cells to be allocated will be called TBA cells. They must fulfil the following conditions: • • • • They are active, They satisfy the filter criteria applied to the Transmitters folder, They are located inside the focus zone, They belong to the folder on which the allocation has been executed. This can be the Transmitters folder or a group of transmitters (subfolder). Only TBA cells are assigned neighbours. If no focus zone exists in the ATL document, Atoll takes into account the computation zone. We assume a reference cell TXi(ic) and a candidate neighbour cell TXj(jc). When automatic planning starts, Atoll checks the following conditions: 1. The distance between both cells must be less than the user-definable maximum inter-site distance. If the distance between the reference cell and the candidate neighbour is greater than this value, then the candidate neighbour is discarded. Atoll calculates the effective distance between the reference cell and its candidate neighbour from the real distance between them and the azimuths of their antennas: Dist  CellA CellB  = D   1 + x  cos  – x  cos   Where x = 0.3% so that the maximum variation in D does not to exceed 1%. D is stated in m. Figure 10.5: Inter-Transmitter Distance Calculation The formula above implies that two cells facing each other have a smaller effective distance than the actual distance. Candidate neighbours are ranked in the order of increasing effective distance from the reference cell. 2. The calculation options, • • Force Co-site Cells as Neighbours: If selected, Atoll adds all the cells located on the same site as the reference cell to the candidate neighbour list. The weight of this constraint can be defined. It is used to calculate the rank of each neighbour, and its importance. Force Adjacent Cells as Neighbours: If selected, Atoll adds all the cells geographically adjacent to the reference cell to the candidate neighbour list. The weight of this constraint can be defined. It is used to calculate the rank of each neighbour, and its importance. 759 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Figure 10.6: Determination of Adjacent Cells Determination of Adjacent Cells: Geographically adjacent cells are determined on the basis of their best server coverage areas. A candidate neighbour cell TXi(ic) is considered adjacent to the reference cell TXi(ic) if there exists at least one pixel of TXj(jc)’s best server coverage area where TXi(ic) is the second best server. The ranking of adjacent neighbour cells increases with the number of such pixels. Adjacent cells are sorted in the order of decreasing rank. • Force Symmetry: If selected, Atoll adds the reference cell to the candidate neighbour list of the its candidate neighbour. A symmetric neighbour relation is allowed only if the neighbour list of the reference cell is not already full. If TXj(jc) is a neighbour of TXi(ic) but TXi(ic) is not a neighbour of TXj(jc), there can be two possibilities: i. The neighbour list of TXj(jc) is not full, Atoll will add TXi(ic) to the end of the list. ii. The neighbour list of TXj(jc) is full, Atoll will not be able to add TXi(ic) to the list, so it will also remove TXj(jc) from the neighbour list of TXi(ic). If the neighbours list of a cell is full, the reference cell will not be added as a neighbour of that cell and that cell will be removed from the reference cell’s neighbours list. You can force Atoll to keep that cell in the reference cell’s neighbours list by adding the following option in the Atoll.ini file: [Neighbours] DoNotDeleteSymmetrics = 1 • Force Exceptional Pairs: This option enables you to force/forbid some neighbour relations. Exceptional pairs are pairs of cells which will always or never be neighbours of each other. If you select "Force exceptional pairs" and "Force symmetry", Atoll considers the constraints between exceptional pairs in both directions so as to respect symmetry condition. On the other hand, if neighbourhood relationship is forced in one direction and forbidden in the other, symmetry cannot be respected. In this case, Atoll displays a warning in the Event viewer. • Delete Existing Neighbours: If selected, Atoll deletes all the current neighbours and carries out a new neighbour allocation. If not selected, the existing neighbours are kept in the list. 3. If the Use Coverage Conditions check box is selected, the coverage areas of TXi(ic) and TXj(jc) must have an overlap ( S TX  ic   S TX  jc  ).Otherwise, only the distance criterion is taken into account. i j The overlapping zone ( S TX  ic   S TX  jc  ) is defined as follows: i • j Here S TX  ic  is the surface area covered by the cell TXi(ic) that comprises all the pixels where: i • The received preamble signal level is greater than or equal to the preamble signal level threshold. The received TX i  ic  TX i  ic  preamble signal level ( C Preamble ) and the preamble signal level threshold are calculated from CNR Preamble TX i  ic  TX i  ic  and T Preamble , respectively, by adding the value of the noise ( n Preamble ) to them. • TX  ic  i TX  ic  i S TX  ic  is the surface area covered by TXi(ic) within C Preamble + HO Start and C Preamble + HO End , or i TX i  ic  TX i  ic  CINR Preamble + HO Start and CINR Preamble + HO End . HOStart is the margin with respect to the best preamble 760 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 signal level or C/(I+N) at which the handover starts, and HO End is the margin with respect to the best • preamble signal level or C/(I+N) at which the handover ends. S TX  jc  is the coverage area where the candidate cell TXj(jc) is the best server. j • • • TX i  ic  If a global value of the preamble C/N threshold ( T Preamble ) is set in the coverage conditions dialogue, for each cell, Atoll uses the higher of the two values, i.e., global value and the value defined for that cell. For calculating the overlapping coverage areas, Atoll uses the service with the lowest body loss, the terminal that has the highest difference between gain and losses, and the shadowing margin calculated using the defined cell edge coverage probability, if the option is selected. The service and terminal are selected such that the selection gives the largest possible preamble C/N coverage areas for the cells. Atoll S TX  ic   S TX  jc  i j When the above conditions are met, Atoll calculates the percentage of the coverage area overlap ( ---------------------------------------  100 ), S TX  ic  i and compares this value with the % Min Covered Area. Figure 10.7: Overlapping Zones S TX  ic   S TX  jc  i j TXj(jc) is considered a neighbour of TXi(ic) if ---------------------------------------  100  % Min Coverage Area . S TX  ic  i Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by decreasing importance in order to keep the ones with high importance. If the maximum number of neighbours to be allocated to each cell is exceeded, Atoll keeps the ones with high importance. The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf. table below); this value varies between 0 and 100%. Neighbourhood cause When Importance value Existing neighbour Only if the Delete Existing Neighbours option is not selected and in case of a new allocation Existing importance Exceptional pair Only if the Force Exceptional Pairs option is selected 100 % Co-site cell Only if the Force Co-site Cells as Neighbours option is selected Importance Function (IF) Adjacent cell Only if the Force Adjacent Cells as Neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions Only if the % Min Covered Area is exceeded Importance Function (IF) Symmetric neighbourhood relationship Only if the Force Neighbour Symmetry option is selected Importance Function (IF) 761 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 The importance is evaluated using an Importance Function (IF), which takes into account the following factors: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. d Di  = 1 – ---------d max d is the effective distance (in m). It corresponds to the real inter-transmitter distance ( D in m) weighted by the azimuths of antennas. d max is the maximum distance between the reference transmitter and a possible neighbour. • • • The co-site factor (C): a Boolean, The adjacency factor (A): the percentage of adjacency, The overlapping factor (O): the percentage of overlapping. The minimum and maximum importance assigned to each of the above factors can be defined. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 30% Adjacency factor (A) Min(A) 30% Max(A) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The Importance Function is evaluated as follows: Neighbourhood cause Importance Function Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%Max(Di))(O)}+Min(Di)+Delta(Di)(Di) 10%+20%{10%(Di)+90%(O)}+1%+9%(Di) No Yes Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 30%+30%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Yes Yes Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 60%+40%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Co-site Adjacent No Where: Delta(X)=Max(X)-Min(X) • • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. If the Min and Max value ranges of the importance function factors do not overlap, the neighbours will be ranked by neighbour cause. With the default values for minimum and maximum importance fields, neighbours will be ranked in this order: co-site neighbours, adjacent neighbours, and neighbours allocated based on coverage overlapping. If the Min and Max value ranges of the importance function factors overlap, the neighbours may be ranked differently. There can be a mix of the neighbourhood causes. The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. With a value of Min(O) = 0%, neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping.If the Min and Max value ranges of the importance function factors overlap, the neighbours may be ranked differently. There can be a mix of the neighbourhood causes. In the results, Atoll lists only the cells for which it finds new neighbours. Cells whose channels have the same start frequency, the same channel width, and the same total number of subcarriers are listed as intra-carrier neighbours. Otherwise, neighbour cells are listed as inter-carrier neighbours. 762 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 • By default, the automatic neighbour allocation compares the defined Max intersite distance with the effective inter-cell distance. As a consequence, there can be cases where the real distance between assigned neighbours is higher than the Max inter-site distance, because the effective distance is smaller. You can force Atoll to compare the Max inter-site distance with the real inter-site distance by adding the following lines in Atoll.ini: [Neighbours] RealInterSiteDistanceCondition=1 • By default, the neighbour importance calculated with respect to distance is based on the global Max inter-site distance setting for all neighbour candidates. As a consequence, there can be cases where the calculated importance is different when the global Max inter-site distance is modified. To avoid that, you can force Atoll to prioritise the individual distances between reference cells and their respective neighbour candidates by adding the following lines in Atoll.ini: [Neighbours] CandidatesMaxDistanceInImportanceCalculation=1 10.5.2 Automatic Inter-technology Neighbour Planning The inter-technology neighbour planning algorithm takes into account all the TBC transmitters (if the other technology is GSM) or the cells of all the TBC transmitters (for any other technology than GSM). This means that all the TBC transmitters (GSM) or the cells of all the TBC transmitters (all other technologies) of the linked document are potential neighbours. The cells to be allocated in the main document will be called TBA cells. They must fulfil the following conditions: • • • • They are active, They satisfy the filter criteria applied to the Transmitters folder, They are located inside the focus zone, They belong to the folder on which allocation has been executed. This can be the Transmitters folder or a group of transmitters (subfolder). Only TBA cells are assigned neighbours. If no focus zone exists in the ATL document, Atoll takes into account the computation zone. We assume a reference cell A and a candidate neighbour B. When automatic planning starts, Atoll checks following conditions: 1. The distance between reference cell and the candidate neighbour must be less than the user-definable maximum inter-site distance. If the distance is greater than this value, the candidate neighbour is discarded. Atoll calculates the effective distance between the reference cell and its candidate neighbour from the real distance between them and the azimuths of their antennas: Dist  CellA CellB  = D   1 + x  cos  – x  cos   Where x = 0.3% so that the maximum variation in D does not to exceed 1%. D is stated in m. Figure 10.8: Inter-Transmitter Distance Calculation The formula above implies that two cells facing each other have a smaller effective distance than the actual distance. Candidate neighbours are ranked in the order of increasing effective distance from the reference cell. 2. The calculation options: 763 Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks • • • • ©Forsk 2015 CDMA Carriers: This option is available when an WiMAX network is being co-planned with a UMTS, CDMA, or TDSCDMA network. This option enables you to select the CDMA carrier(s) that you want Atoll to consider as potential neighbours of WiMAX cells. You may choose one or more carriers. Atoll will allocate only the cells using the selected carriers as neighbours. Force co-site cells as neighbours: If selected, Atoll adds all the transmitters/cells located on the same site as the reference cell in its candidate neighbour list. The weight of this constraint can be defined. It is used to calculate the rank of each neighbour and its importance. Force exceptional pairs: This option enables you to force/forbid some neighbour relations. Exceptional pairs are pairs of cells which will always or never be neighbours of each other. Delete existing neighbours: If selected, Atoll deletes all the current neighbours and carries out a new neighbour allocation. If not selected, the existing neighbours are kept in the list. 3. Neighbour relation criterion: • Allocation based on distance: The allocation algorithm is based on the effective distance between the reference cell and its candidate neighbour. • Algorithm based on coverage overlapping: The coverage areas of the reference cell A and the candidate neighbour B must overlap ( S A  S B ). Two cases may exist for SA: • 1st case: SA is the area where the cell A is the best serving cell, with a 0dB margin. This means that the preamble signal received from A is greater than the minimum required (calculated from the preamble C/N threshold), and is the highest one. . • 2nd case: The margin is other than 0dB. SA is the area where: The preamble signal level received from A exceeds the minimum required (calculated from the preamble C/N threshold) and is within a margin from the highest signal level. Two cases may exist for SB: • 1st case: SB is the area where the candidate neighbour is the best server. In this case, the margin must be set to 0dB. The signal level received from B exceeds the minimum required, and is the highest one. • 2nd case: The margin is other than 0dB. SB is the area where: The signal level received from B exceeds the minimum required and is within a margin from the best signal level. SA  SB Atoll calculates the percentage of the coverage area overlap ( ------------------  100 ) and compares this value with the % SA SA  SB Min Covered Area. B is considered a neighbour of A if ------------------  100  % Min Covered Area . SA Candidate neighbours are ranked in the order of decreasing coverage area overlap percentages. Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by decreasing importance in order to keep the ones with high importance. If the maximum number of neighbours to be allocated to each cell is exceeded, Atoll keeps the ones with high importance. The importance (%) of neighbours depends on the distance and on the reason of allocation: • 764 For allocation based on distance: Neighbour cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter/cell If the Force co-site cells as neighbours option is selected 100 % Neighbour relation that fulfils distance conditions If the maximum distance is not exceeded d1 – ---------d max Atoll 3.3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 d is the effective distance between the reference cell and the neighbour and d max is the maximum inter-site distance. • For allocation based on coverage overlapping: Neighbour cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter/cell If the Force co-site cells as neighbours option is selected IF Neighbourhood relationship that fulfils coverage conditions If the % minimum covered area is exceeded IF The importance is evaluated using an Importance Function (IF), which takes into account the following factors: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. d Di  = 1 – ---------d max d is the effective distance (in m). It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. d max is the maximum distance between the reference transmitter and a possible neighbour. • • The co-site factor (C): a Boolean, The overlapping factor (O): the percentage of overlapping. The IF is user-definable using the Min importance and Max importance fields. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The IF evaluates importance as follows: Co-site Neighbourhood cause IF Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)} Yes Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+ Max(O)(O)/(Max(Di)+Max(O))} 60%+40%{1/7%(Di)+6/7%(O)} Where Delta(X)=Max(X)-Min(X) • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. If the Min and Max value ranges of the importance function factors do not overlap, the neighbours will be ranked by neighbour cause. With the default values for minimum and maximum importance fields, neighbours will be ranked in this order: co-site neighbours and neighbours allocated based on coverage overlapping. If the Min and Max value ranges of the importance function factors overlap, the neighbours may be ranked differently. There can be a mix of the neighbourhood causes. In the results, Atoll displays only the cells for which it finds new neighbours. 10.5.3 Automatic Frequency Planning Using the AFP The role of an Automatic Frequency Planning (AFP) tool is to assign frequencies (channels) to cells of a network such that the overall network performance is optimised. In other words, the interference within the network is reduced as much as possible. Co-channel interference is the main reason for overall network quality degradation in WiMAX. In order to improve network performance, the WiMAX AFP tries to minimise co- and adjacent channel interference as much as possible while respecting 765 which takes into account interference matrices.3 • Cells within the cell’s (or the default) minimum reuse distance. The main constraints are the resources available for allocation.3. Assigned weight  Neighbour = 0. and distance between transmitters. Atoll takes into account the computation zone. The AFP is based on a cost function which represents the interference level in the network. The absolute values of the constraint weights are calculated from the relative weights (%) defined in the Constraint Weights dialogue as follows: % Neighbour  Neighbour = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce % IM  IM = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce % Dis tan ce  Dis tan ce = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce 10. They must fulfil the following conditions: • • • • They are active.and Adjacent Channel Overlaps Calculation" on .5 • Cells that are listed in the interference matrix of the TBA cell. You can modify these weights in your WiMAX document. frequency plan is the one which corresponds to the lowest cost. Atoll calculates the cost between each individual TBA and related cell.5.3. i.3. if the check box "Existing neighbours" is selected.Atoll 3. The aim of the AFP is to minimise the cost. i. Assigned weight  Dis tan ce = 0. neighbours.2 The sum of the weights assigned to the above relations is 1.. and then the overall cost for the TBA cell. Their channel allocation status is not set to locked. and distance between transmitters. Channel separation is studied between each TBA cell and its related cells. The following describes the AFP’s automatic planning method for frequencies in WiMAX networks. Assigned weight  IM = 0. 766 TX i  ic  – TX j  jc     Neighbour   Neighbour  TX i  ic  – TX j  jc  +  Dis tan ce   Dis tan ce TX  ic  – TX j  jc   +  i IM IM  is the channel overlap ratio as calculated in "Co. and the relationships to take into account.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 any constraints input to it.e. Related cells of a TBA cell are: • Its neighbours.. if the check box "Reuse distance" is selected. The AFP takes into account the cells of all the TBC transmitters. They are located inside the focus zone. the number of frequencies with which the AFP can work. interference matrices. or optimum.2 Cost Calculation The cost of the relation between the TBA cell and its related cell is calculated as follows: $ TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  = rO TX  ic  – TX  jc  i j Where r O page 702. 10. neighbour relations. They satisfy the filter criteria applied to the Transmitters folder.e. The best. If no focus zone exists in the ATL document.and Adjacent Channel Overlaps Calculation" on page 702.5. The cells to be allocated will be called TBA cells.1 Constraint and Relationship Weights The AFP is based on a cost function which takes into account channel separation constraints based on the channel overlap ratio as calculated in "Co. 3 AFP Algorithm The AFP algorithm is an iterative algorithm which: • • • • Calculates the cost (as described above) of the initial frequency plan. and proposes the last known best frequency plan as the solution.  Dis tan ce is calculated as explained in "Distance Importance Calculation" on page 774.5. r CCO TX  ic  – TX  jc  i j and r ACO are the co. the total cost of the current frequency plan for the entire network is simply the sum of the total TBA cell costs calculated above. k is a running index from 0 to 567 for FFT 2048.  Neighbour is calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 759. from 0 to 142 for FFT 512. and A pseudo-noise sequence transmitted using the subcarriers corresponding to the preamble carrier set.and Adjacent Channel Overlaps Calculation" on page 702. 1.Atoll 3..3. n is the number of the preamble carrier set indexed 0. $ Total =  TX i  ic  $ Total TX  ic  i 10. 10. Each preamble index. Atoll calculates the quality reduction factor for the TBA cell and its related cell from the cost calculated above as follows: QRF TX i  ic  – TX j  jc  = 1–$ TX i  ic  – TX j  jc  The quality reduction factor is a measure of the cost of an individual relation. The subcarriers are modulated using a BPSK modulation with a specific Pseudo-Noise (PN) sequence. from 0 to 283 for FFT 1024. Preamble carrier sets are defined using equation below: PreambleCarrierSet n = n + k  3 Where PreambleCarrierSetn gives the subcarriers used by the preamble. The downlink subframe can be divided into a 3-segment structure. or 2.and adjacent channel overlap ratios as calculated in "Co.5.3.16e defines 114 preamble indexes.. For manual neighbour planning.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 TX  ic  – TX  jc  i j TX  ic  – TX  jc  i j  Neighbour is the importance of the relationship between the TBA cell and its related neighbour cell.4 Automatic Preamble Index Planning Using the AFP IEEE 802.and adjacent channel interference probabilities calculated as TX  ic  – TX  jc  i j explained in "Interference Matrix Calculation" on page 774. Tries different frequency plans in order to reduce the cost. 767 . from 0 to 113. i. TX  ic  – TX  jc  i j  IM is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows: TX i  ic  – TX j  jc  TX i  ic  – TX j  jc   IM = r CCO TX i  ic  – TX j  jc  TX i  ic  – TX j  jc   IM – CC and  IM – CC TX i  ic  – TX j  jc    IM – CC TX i  ic  – TX j  jc  + r ACO TX i  ic  – TX j  jc    IM – AC are respectively the co. or 2). and includes a preamble which begins the transmission (the first symbol of the downlink transmission).e. Stops when it is unable to improve the cost of the network. DL PermBase (0 to 31) for the obligatory first DL PUSC zone. The preamble subcarriers are divided into 3 carrier sets. 1. and from 0 to 35 for FFT 128. this value is equal to 1. Memorises the different frequency plans in order to determine the best one. considering all the cells with which the TBA cell has relations: TX i  ic  $ Total = 1 –  QRF TX i  ic  – TX j  jc  TX j  jc  And. contains the following information: • • • Segment number (0. There are three possible groups consisting of a carrier set each which may be used by any segment. The total cost of the current frequency plan for any TBA cell is given as follows. These are defined by allocation of different subcarriers to each one of them. TX i  ic  – TX j  jc   Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance between TX i  ic  – TX j  jc  them.e. i. the frequency plan which provides the lowest total cost. and the frequency plan of the network. A mobile trying to connect to the network scans all the preamble subcarriers. Atoll takes into account the computation zone.4. Same preamble index. Once the best server is known.3. Therefore. 10. referred to as Cell PermBase in Atoll).02 3. As can be understood from the above description. in the order of priority: 1. first-order neighbours of a common WiMAX cell. and compares the PN sequences it is receiving with the 114 stored in its memory in order to detect the preamble index from the PN sequence. You can modify these weights in your WiMAX document. Related cells of a TBA cell are: • 768 Its neighbours. if the check box "Existing neighbours" is selected. out of the 114 available.23 The sum of the weights assigned to the above constraints is 1. UCD. If no focus zone exists in the ATL document. and then the overall cost for the TBA cell. Same cell permbase. Cell search and selection will be impossible. its PN sequence is used to identify its transmission. on the preamble carrier set.75 2. if all the cells in the network transmit the same preamble index. DCD. Their preamble index status or segment is not set to locked. . It selects the base station as its server whose preamble it receives with either the highest signal level or the highest C/(I+N). They satisfy the filter criteria applied to the Transmitters folder. Atoll calculates the cost between each individual TBA and related cell. the network will have 100% interference on downlink preambles. and allow easy recognition of cells by mobiles. They must fulfil the following conditions: • • • • They are active. and the IDCell (DL PermBase of the first DL PUSC zone. The following describes the AFP’s automatic planning method for preamble indexes in a WiMAX network. The absolute values of the constraint weights are calculated from the relative weights (%) defined in the Constraint Weights dialogue as follows: % PI  PI = -----------------------------------------------------% PI + % Seg + % PB % Seg  Seg = -----------------------------------------------------% PI + % Seg + % PB % PB  PB = -----------------------------------------------------% PI + % Seg + % PB The above constraints are studied between each TBA cell and its related cells. The PN sequence of the best server gives the preamble index.e. The AFP takes into account the cells of all the TBC transmitters.1 Constraint and Relationship Weights The AFP is based on a cost-based function which takes into account the following constraints. which takes into account interference matrices. which in turn gives the segment number.Atoll 3. Same segment number.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 In a WiMAX network. the mobile knows which subcarriers to listen to for the FCH.. distance between transmitters. Therefore. and optionally second-order neighbours). it is important to intelligently plan preamble indexes to cells so as to reduce preamble interference. Assigned weight  PI = 0. Assigned weight  Seg = 0. The cells to be allocated will be called TBA cells. and it will be impossible for a mobile to identify different cells. They are located inside the focus zone. Assigned weight  PB = 0. neighbour relations (first-order neighbours.5. and UL-MAP. DL-MAP. PN sequences) from all the base stations it can receive. listens to all the preambles (i. each base station transmits a different PN sequence. and N Seg TX i  jc   N Seg TX i  ic  – TX j  jc  .0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Assigned weight  Neighbour = 0. if the check box "Reuse distance" is selected.ini file (see the Administrator Manual). and  PB are the weights assigned to the preamble index. then VL 1 TX i  ic  – TX j  jc  + VL 2 = 1. and cell permbase constraints. Assigned weight  Inter – Neighbour = 0.4. second-order neighbours can also be taken into account.5 and that of the collision between neighbours of a common cell is of course  Inter – Neighbour = 0 .  Seg .ini file (see the Administrator Manual).5. • Cells that are listed in the interference matrix of the TBA cell.15 . the weight assigned to the direct first-order neighbour relation alone is  Neighbour = 0. and  Inter – Neighbour = 0. The absolute values of the constraint weights are calculated from the relative weights (%) defined in the Constraint Weights dialogue as follows: % Neighbour  Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % Inter – Neighbour  Inter – Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % 2nd – Neighbour  2nd – Neighbour = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % IM  IM = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce % Dis tan ce  Dis tan ce = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce 10. Assigned weight  Dis tan ce = 0.10 . In this case. If the collision between neighbours of a common cell is not taken into account.2 Cost Calculation Atoll calculates the constraint violation levels between the TBA cell TXi(ic) and its related cell TXj(jc) as follows: TX i  ic  – TX j  jc  VL 1 TX i  ic  – TX j  jc  VL 2 PI PB =  PI  p Coll +  PB  p Penalty Seg =  Seg  p Coll If TX i  ic  and TX i  jc  are co-transmitter cells. the assigned weights are:  Neighbour = 0. You can modify these weights in your WiMAX document.Atoll 3.3. Where  PI . 769 . This relation is also taken into account.   PI PI p Coll is the preamble index collision probability given by p Coll =  1   0 if PI if PI TX  ic  i TX  ic  i = PI  PI TX  jc  j TX  jc  j .  2nd – Neighbour = 0. segment number.3 • Cells within the cell’s (or the default) minimum reuse distance.35 TBA cells which are first-order neighbours of a common cell are also related to each other through that cell.2 The sum of the weights assigned to the above relations is 1.25 . By adding an option in the Atoll. Assigned weight  IM = 0.15 You can choose to not take into account the preamble index collision between neighbours of a common cell by adding an option in the Atoll. and the option Allocate Same Segment to Co-transmitter Cells has been TX i  ic  selected. 0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 Seg p Coll is the segment number collision probability. If the TBA cell is related to its second order neighbour through more than one first order neighbour. Next. If the TBA cell has the same preamble index assigned as one of its second-order neighbours. TX i  ic  – TX j  jc   Neighbours TX i  ic  – TX j  jc  =  Neighbour   Neighbour TX i  ic  – TX j  jc  Where  Neighbour +  Inter – Neighbour   Inter – Neighbour +  2nd – Neighbour   2nd – Neighbour TX i  ic  – TX j  jc  is the importance of the relationship between the TBA cell and its related neighbour cell. TX i  ic  – TX j  jc  + r ACO TX i  ic  – TX j  jc    IM – AC if the frequency plan is taken into account .001 if PB  0 Otherwise  PB cell permbase planning strategy is set to "Same per site". For manual neighbour planning.Atoll 3.  N Seg  TX  ic  TX  jc  TX i  ic  TX j  jc   1 if PB i  PB j AND Site = Site  PB is the cell permbase penalty given by p Penalty =  TX i  ic  TX j  jc  TX i  ic  TX j  jc  if the  PB AND Site  Site  0. the importance of the preamble index collision is the average of their neighbour importance values with the common neighbour cell. this value is equal to 1. Seg p Coll   1 =    0 PB p Penalty TX i  ic  if N Seg TX i  ic  if N Seg Seg p Coll is given by   0 =    1 Seg p Coll TX  ic  i if N Seg if TX  ic  i N Seg TX  jc  j = N Seg  TX  jc  j N Seg . Next. then the importance is the highest value among all the averages: TX i  ic  – TX j1  j1c   Inter – Neighbour TX i  ic  – TX j2  j2c    Neighbour  +  Neighbour = Max  --------------------------------------------------------------------------------- 2  All Neighbour Pairs  with PI Collisions Where TX j1  j1c  and TX j2  j2c  are two neighbours of the TBA cell TX i  ic  that have the same preamble index assigned. Otherwise. Atoll calculates the importance of the interference relations between the TBA cell and its related cell. Atoll calculates the importance of the neighbour relations between the TBA cell and its related cell.3. TX i  ic  – TX j  jc   Interference TX i  ic  – TX j  jc   IM TX i  ic  – TX j  jc   IM TX i  ic  – TX j  jc  =  IM   IM 770 TX i  ic  – TX j  jc   f Overlap is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows: TX i  ic  – TX j  jc  = r CCO TX i  ic  – TX j  jc  and  IM TX i  ic  – TX j  jc  +  Dis tan ce   Dis tan ce TX i  ic  – TX j  jc    IM – CC TX i  ic  – TX j  jc  =  IM – CC otherwise. If TX i  ic  and TX j  jc  are co-transmitter cells.  2nd – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour planning. and the option Allocate Same Segment to Co-transmitter Cells has been selected. The cell permbase penalty models the cell permbase constraint. the importance is the highest value among all the multiples:  2nd – Neighbour = TX  ic  – TX  jc  j  i Neighbour  All Neighbour Pairs Max TX j  jc  – TX k  kc    Neighbour   with PI Collisions Where TX k  kc  is the second-order neighbour of TX i  ic  through TX j  jc  . If two cells are neighbours of a common cell and have the same preamble index assigned. and by p Penalty = 0 if the cell permbase planning strategy is set to "Free". TX j  jc  = N Seg TX j  jc  .  Inter – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour planning. the importance of the preamble index collision is the multiple of the importance values of the first order neighbour relations between the TBA cell and its second order neighbour. If more than one pair of neighbours of the TBA cell has the same preamble index assigned.  Neighbour is calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 759. TX i  ic  – TX j  jc  f Overlap is TX i  ic  – TX j  jc  = rO calculated as explained in "Distance Importance TX i  ic  – TX j  jc  if the frequency plan is taken into account and f Overlap Calculation" on page 774.. TX i  ic  – TX j  jc   Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance between TX i  ic  – TX j  jc   Dis tan ce them. Subchannels in a channel contain different physical subcarriers when different permbases are used as seeds.5.Atoll 3. The following describes the AFP’s automatic planning method for zone permbases in a WiMAX network.and adjacent channel interference probabilities calculated as TX  ic  – TX  jc  i j explained in "Interference Matrix Calculation" on page 774. also a number from 0 to 31. The second DL PUSC permutation zone uses the zone permbase.. r O TX  ic  – TX  jc  i j . and r ACO are the total. and optionally second-order neighbours). The total cost of the current preamble index plan for any TBA cell is given as follows. Uplink permutation zones also use only zone permbases. It is a number from 0 to 31. Downlink PUSC permutation zones use 2 permbases: 1. 2.5 Automatic Zone PermBase Planning Using the AFP PermBases are numbers which are used as seeds in the permutation of subcarriers (mapping between physical and logical subcarrier numbers) and their allocation to subchannels. Other downlink permutation zones only use zone permbases.3 AFP Algorithm The AFP algorithm is an iterative algorithm which: • • • • Calculates the cost (as described above) of the initial preamble index plan. and the frequency plan of the network. The AFP takes into account the cells of all the TBC transmitters. co-channel.4. distance between transmitters. It is called IDCell in the IEEE specifications. $ Total =  TX  ic  i $ Total TX i  ic  10. and adjacent channel overlap ratios as calculated in "Co. neighbour relations (first-order neighbours.e. 10. Stops when it is unable to improve the cost of the network. first-order neighbours of a common WiMAX cell. They satisfy the filter criteria applied to the Transmitters folder. = 1 otherwise. Tries different preamble index plans in order to reduce the cost. From the constraint violation levels and the importance values of the relations between the TBA and its related cell. Atoll calculates the quality reduction factor for the pair as follows: QRF TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc      i = 1 –   VL1 + VL 2 + VL 1   Neighbours  f Overlap  Interference    The quality reduction factor is a measure of the cost of an individual relation. The cells to be allocated will be called TBA cells.and Adjacent Channel Overlaps Calculation" on page 702. i. They must fulfil the following conditions: • • • They are active. which takes into account interference matrices.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 TX  ic  – TX  jc  i j TX  ic  – TX  jc  i j  IM – CC and  IM – CC are respectively the co. considering all the cells with which the TBA cell has relations: TX i  ic  $ Total = 1 –  QRF TX i  ic  – TX j  jc  TX j  jc  And. Their zone permbase status is not set to locked. the uplink zone permbase is a number from 0 to 69.e. 771 . which provides the lowest total cost. However.3. the total cost of the current preamble index plan for the entire network is simply the sum of the total TBA cell costs calculated above. The first DL PUSC permutation zone uses the cell permbase (mapped to the preamble index of the cell). and proposes the last known best preamble index plan as the solution. r CCO TX  ic  – TX  jc  i j .5. i. Memorises the different plans in order to determine the best one. 0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks • ©Forsk 2015 They are located inside the focus zone. if the check box "Existing neighbours" is selected.3. By adding an option in the Atoll.5. • Cells that are listed in the interference matrix of the TBA cell.10 . If the collision between neighbours of a common cell is not taken into account. if the check box "Reuse distance" is selected.5 and that of the collision between neighbours of a common cell is of course  Inter – Neighbour = 0 . Assigned weight  Inter – Neighbour = 0. and then the overall cost for the TBA cell.15 You can choose to not take into account the zone permbase collision between neighbours of a common cell by adding an option in the Atoll.15 . Assigned weight  Dis tan ce = 0. Assigned weight  ZPB = 1 The above constraint is studied between each TBA cell and its related cells. ZPB is used for the downlink zone permbases ( ZPBDL ) and uplink zone permbases ( ZPBUL ) without distinction. TX i  ic  – TX j  jc   Neighbours 772 TX i  ic  – TX j  jc  =  Neighbour   Neighbour +  Inter – Neighbour   Inter – Neighbour +  2nd – Neighbour   2nd – Neighbour . Next.2 Cost Calculation Atoll calculates the constraint violation level between the TBA cell TXi(ic) and its related cell TXj(jc) as follows: VL TX  ic  – TX  jc  i j ZPB =  ZPB  p Coll Where  ZPB is the weight assigned to the zone permbase constraint. Assigned weight  IM = 0. • If no focus zone exists in the ATL document. This relation is also taken into account. Assigned weight  Neighbour = 0. and  Inter – Neighbour = 0.35 TBA cells which are first-order neighbours of a common cell are also related to each other through that cell. 10. In this case.25 .5. Atoll calculates the cost between each individual TBA and related cell.Atoll 3.5. Atoll calculates the importance of the neighbour relations between the TBA cell and its related cell. the assigned weights are:  Neighbour = 0. • In the following description.2 The sum of the weights assigned to the above relations is 1. Related cells of a TBA cell are: • Its neighbours.ini file (see the Administrator Manual).  2nd – Neighbour = 0.5. 10.ini file (see the Administrator Manual). the weight assigned to the direct first-order neighbour relation alone is  Neighbour = 0. second-order neighbours can also be taken into account.1 Constraint and Relationship Weights The AFP is based on a cost-based function which takes into account the following constraint: • Same zone permbase. Atoll takes into account the computation zone.   ZPB ZPB p Coll is the zone permbase collision probability given by p Coll =  1   0 if ZPB if ZPB TX i  ic  TX i  ic  = ZPB  ZPB TX j  jc  TX j  jc  .3 • Cells within the cell’s (or the default) minimum reuse distance. are respectively the co. is the importance of the relationship between the TBA and its related cell with respect to the distance between TX i  ic  – TX j  jc   Dis tan ce TX i  ic  – TX j  jc  f Overlap is TX i  ic  – TX j  jc  = rO calculated as explained in "Distance Importance TX i  ic  – TX j  jc  if the frequency plan is taken into account and f Overlap Calculation" on page 774. the importance is the highest value among all the multiples: TX  ic  – TX  jc  j  i  Neighbour All Neighbour Pairs  2nd – Neighbour = Max TX j  jc  – TX k  kc      Neighbour with ZPB Collisions Where TX k  kc  is the second-order neighbour of TX i  ic  through TX j  jc  . r CCO TX i  ic  – TX j  jc  . this value is equal to 1. If more than one pair of neighbours of the TBA cell has the same zone permbase assigned. Atoll calculates the importance of the interference relations between the TBA cell and its related cell.  Inter – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour planning.and Adjacent Channel Overlaps Calculation" on page 702. the importance of the zone permbase collision is the multiple of the importance values of the first order neighbour relations between the TBA cell and its second order neighbour. TX i  ic  – TX j  jc   Dis tan ce them.  Neighbour is calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 759. and r ACO are the total. If the TBA cell is related to its second order neighbour through more than one first order neighbour. co-channel.and adjacent channel interference probabilities calculated as TX i  ic  – TX j  jc  explained in "Interference Matrix Calculation" on page 774. The total cost of the current zone permbase plan for any TBA cell is given as follows. then the importance is the highest value among all the averages: TX i  ic  – TX j1  j1c  TX i  ic  – TX j2  j2c    Neighbour +  Neighbour   --------------------------------------------------------------------------------- 2  All Neighbour Pairs   Inter – Neighbour = Max with ZPB Collisions Where TX j1  j1c  and TX j2  j2c  are two neighbours of the TBA cell TX i  ic  that have the same zone permbase assigned.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 TX  ic  – TX  jc  i j TX  ic  – TX  jc  i j Where  Neighbour is the importance of the relationship between the TBA cell and its related neighbour cell. Atoll calculates the quality reduction factor for the pair as follows: QRF TX i  ic  – TX j  jc  = 1 – VL TX i  ic  – TX j  jc  TX i  ic  – TX j  jc     Interference  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  +  Neighbours  f Overlap   The quality reduction factor is a measure of the cost of an individual relation.  2nd – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour planning. considering all the cells with which the TBA cell has relations: TX i  ic  $ Total = 1 –  QRF TX i  ic  – TX j  jc  TX j  jc  773 .3. the importance of the zone permbase collision is the average of their neighbour importance values with the common neighbour cell. If two cells are neighbours of a common cell and have the same zone permbase assigned. = 1 otherwise. For manual neighbour planning. From the constraint violation level and the total importance of the relation between the TBA and its related cell. Next. If the TBA cell has the same zone permbase assigned as one of its second-order neighbours. and adjacent channel overlap ratios as calculated in "Co. r O TX i  ic  – TX j  jc  . TX i  ic  – TX j  jc   Interference TX i  ic  – TX j  jc   IM TX i  ic  – TX j  jc   IM TX i  ic  – TX j  jc  =  IM   IM TX i  ic  – TX j  jc  = r CCO TX i  ic  – TX j  jc    IM – CC TX i  ic  – TX j  jc  and  IM TX i  ic  – TX j  jc  TX i  ic  – TX j  jc   f Overlap is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows: TX i  ic  – TX j  jc   IM – CC TX i  ic  – TX j  jc  +  Dis tan ce   Dis tan ce =  IM – CC TX i  ic  – TX j  jc  and  IM – CC TX i  ic  – TX j  jc  + r ACO TX i  ic  – TX j  jc    IM – AC if the frequency plan is taken into account otherwise. i..e.6. Memorises the different plans in order to determine the best one.e.2 Distance Importance Calculation TX i  ic  – TX j  jc  The distance importance between two cells (  Dis tan ce TX i  ic  – TX j  jc   Dis tan ce 774   1     2 D Reuse =  Log   --------------------------------     D TXi  ic  – TXj  jc    --------------------------------------------------------2  Log  D Reuse   if D ) is calculated as follows: TX i  ic  – TX j  jc  Otherwise 1 . 10.3. and f ACS – FB is the adjacent channel suppression factor defined for the frequency band of the cell TXi(ic).5.5.. the total cost of the current zone permbase plan for the entire network is simply the sum of the total TBA cell costs calculated above. the interference probability is 0. that comprises all the pixels where CNR Preamble  T Preamble i as calculated in "Service Area Calculation" on page 714. n Preamble the preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on TX i  ic  page 708. M Quality is the quality margin used for the interference matrices calculation.1 Interference Matrix Calculation The co-channel interference probability is calculated as follows: S TX  ic  i TX  ic  – TX  jc  i j  IM – CC TX j  jc  TX i  ic    n  C Preamble + M Quality Preamble- --------------------------- ------------------------------------------------------------ TX i  ic  TX  ic  10 10  T i C Preamble – 10  Log  10 + 10  Preamble       = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX  ic  i The adjacent channel interference probability is calculated as follows: S TX  ic  i TX i  ic  – TX j  jc   IM – AC TX j  jc  TX i  ic  TX i  ic    n Preamble  C Preamble + M Quality + f ACS – FB -----------------------------  --------------------------------------------------------------------------------------------TX i  ic  TX  ic  10 10  T i C Preamble – 10  Log  10 + 10 Preamble         = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX  ic  i For frequencies farther than the adjacent channel. Tries different zone permbase plans in order to reduce the cost. S TX  ic  i TX i  ic  Condition is the best server coverage area of the cell TXi(ic) TX j  jc  where the given condition is true.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks ©Forsk 2015 And. Stops when it is unable to improve the cost of the network.6 Appendices 10. C Preamble and C Preamble are the received preamble signal levels from the cells TXi(ic) and TX  ic  i TXj(jc) respectively.  $ Total = TX  ic  i $ Total TX i  ic  10. and proposes the last known best zone permbase plan as the solution.6.Atoll 3. 10. which provides the lowest total cost. i.5. TX i  ic  TX i  ic  Here S TX  ic  is the best server coverage area of the cell TXi(ic).5.3 AFP Algorithm The AFP algorithm is an iterative algorithm which: • • • • Calculates the cost (as described above) of the initial zone permbase plan.5. Cells that are further than the reuse distance do not have any cost related to the distance relation.  and  are calculated from the azimuths of the two cells as shown in Figure 10. and cells that are located far have low importance. and two cells pointing in opposite directions will have a greater effective distance. x is set TX i  ic  – TX j  jc  to 10 % so that the maximum variation in D due to the azimuths does not exceed 40 %. which is interpreted as a high cost. and D follows: D D TX  ic  – TX  jc  i j TX i  ic  – TX j  jc  them. d = d TX  ic  – TX  jc  i j TX  ic  – TX  jc  i j is the weighted distance between the TBA cell TXi(ic) and its related cell TXj(jc) calculated as   1 + x   cos    – cos    – 2   is weighted according to the azimuths of the TBA cell and its related cell with respect to the straight line joining TX i  ic  – TX j  jc  is the distance between the two cells considering any offsets with respect to the site locations.10 on page 775.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks AT330_TRR_E1 Where D Reuse is the minimum reuse distance. This figure shows that cells that are located near (based on the effective distance which is weighted by the orientations of the cells) have high importance. either defined for each TBA cell individually or set for all the TBA cells in the AFP dialogue. The importance of the distance relation is explained in Figure 10.3.9 on page 775. Figure 10.Atoll 3.10: Importance Based on Distance Relation 775 .9: Weighted Distance Between Cells The above formula implies that two cells facing each other will have a shorter effective distance between them than the real distance. Figure 10. Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 10: WiMAX BWA Networks 776 ©Forsk 2015 .3. Chapter 11 Wi-Fi Networks This chapter covers the following topics: • "Definitions" on page 779 • "Calculation Quick Reference" on page 782 • "Available Calculations" on page 789 • "Calculation Details" on page 799 • "Automatic Planning Algorithms" on page 824 . Atoll 3.3.0 Technical Reference Guidefor Radio Networks © Forsk 2015 778 . e. The third part describes all the calculation algorithms used in all the calculations. subscriber (calculations on subscriber lists). otherwise.38 x 10-23 J/K Boltzmann’s constant T 290 K Ambient temperature n0 Calculation result ( 10  Log  K  T  1000  = – 174 dBm/Hz ) dBm/Hz Power spectral density of thermal noise r CP Frame configuration or. Logarithms used in this chapter (Log function) are base-10 unless stated otherwise.Atoll 3. and the radio resource management algorithms used in Monte Carlo simulations. calculations on subscriber lists. These algorithms include the calculation of signal levels.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 11 Wi-Fi Networks This chapter describes all the calculations performed in Atoll Wi-Fi documents. location in the Atoll GUI. • Other cells (represented by the subscript "j") comprising the other transmitter TXj and its carrier jc. Name Value Unit Description K 1. • Mj: A mobile (Monte Carlo simulations) covered/served by any other cell TXj(jc). The other cells in the network can be interfering cells (downlink) or the serving cells of interfering mobiles (uplink). The calculation algorithms used by these calculation processes are available in the next part. signal level coverage predictions. • Mi: A pixel (coverage predictions). It is the cell which is currently the focus of the calculation. For the definition of TBC transmitters please refer to "Path Loss Matrices" on page 103. For example. The second part describes all the calculation processes. and their usage. or mobile (Monte Carlo simulations) covered/served by the studied cell TXi(ic). All the calculation algorithms in this section are described for two types of cells. noise. coverage predictions. and Monte Carlo simulations. • 11. and interference for downlink and uplink. global parameter None Cyclic Prefix Ratio (guard interval) Choice List: 1/4 (long). • • A studied cell (represented by the subscript "i") comprising the studied transmitter TXi and its carrier ic.3. 1/8 (short) M PC Global parameter dB Uplink power control margin CNR Min Global parametera dB Minimum signal to thermal noise threshold (interferer cutoff) N SCa – Total Frame configuration parameter None Total number of subcarriers per channel (FFT size) N SCa – Used Frame configuration parameter None Number of used subcarriers per channel N SCa – Data Frame configuration zone parameter None Number of subcarriers per channel used for data transfer N SCa – DC Hard-coded parameter ( N SCa – DC = 1 ) None Number of DC subcarriers per channel N SCa – Pilot Calculation result ( N SCa – Pilot = N SCa – Used – N SCa – Data ) None Number of pilot subcarriers per channel 779 . and simulations. The cell being studied during a calculation is referred to as TXi(ic) in this chapter.. A cell refers to a transmitter-carrier (TX-c) pair. their significance. All the calculation algorithms in this section are described for two types of receivers. The first part of this chapter lists all the input parameters in the Wi-Fi documents. It also contains the lists of the formulas used for the calculations. signal quality coverage predictions. a victim cell when calculating the interference it is receiving from other cells. i. • • • All the calculations are performed on TBC (to be calculated) transmitters. point analysis calculations.1 Definitions This table lists the input to calculations. 3.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks Name Value Unit Description N SCa – Guard Calculation result ( N SCa – Guard = N SCa – Total – N SCa – Used – N SCa – DC ) None Number of guard subcarriers per channel W Channel Frequency band parameter MHz Channel bandwidth First Frequency band parameter None First channel number of the frequency band N Channel Last Frequency band parameter None Last channel number of the frequency band F Start – FB – DL Frequency band parameter MHz DL Start frequency of the frequency band F Start – FB – UL Frequency band parameter MHz UL Start frequency of the frequency band f ACS – FB Frequency band parameter dB Adjacent Channel Suppression Factor ICS FB Frequency band parameter MHz Inter-channel spacing CN FB Frequency band parameter None Channel number step Inter – Tech Network parameter dB Inter-technology interference reduction factor B Bearer parameter None Bearer index Mod B Bearer parameter None Modulation used by the bearer CR B Bearer parameter None Coding rate of the bearer B Bearer parameter bits/ symbol Bearer Efficiency TB Bearer parameter dB Bearer selection threshold TP BH – DL Site Site parameter kbps Maximum backhaul site downlink throughput Site Site parameter kbps Maximum backhaul site uplink throughput Transmitter parameter dB Transmitter noise figure N Ant – TX Transmitter parameter None Number of antennas used for MIMO in transmission N Ant – RX Transmitter parameter None Number of antennas used for MIMO in reception TX Antenna parameter dB Transmitter antenna gain TX Transmitter parameter dB Transmitter loss N Channel Cell parameter None Cell’s channel number P DL Cell parameter dBm Power TL DL Cell parameter % Downlink traffic load TL UL Cell parameter % Uplink traffic load TL DL – Max Cell parameter % Maximum downlink traffic load TL UL – Max Cell parameter % Maximum uplink traffic load NR UL Cell parameter dB Uplink noise rise N Users – Max Cell parameter None Maximum number of users per cell N Users – DL Cell parameter None Number of users connected to the cell in downlink N Channel f IRF TP BH – UL nf G L 780 ©Forsk 2015 TX . Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Name Value Unit Description N Users – UL Cell parameter None Number of users connected to the cell in uplink T AMS Cell parameter dB Adaptive MIMO switch threshold T Min Cell parameter dB Minimum C/N threshold Inter – Tech Cell parameter dB Inter-technology downlink noise rise Inter – Tech Cell parameter dB Inter-technology uplink noise rise G SU – MIMO Max Cell Wi-Fi equipment parameter None Maximum SU-MIMO gain G Div – UL Cell Wi-Fi equipment parameter dB Uplink STTD/MRC or SU-MIMO diversity gain p Service parameter None Service priority B DL – Highest Service parameter None Highest bearer used by a service in the downlink B UL – Highest Service parameter None Highest bearer used by a service in the uplink B DL – Lowest Service parameter None Lowest bearer used by a service in the downlink B UL – Lowest Service parameter None Lowest bearer used by a service in the uplink f Act UL Service parameter % Uplink activity factor f Act DL Service parameter % Downlink activity factor TPD Min – UL Service parameter kbps Minimum throughput demand in the uplink TPD Min – DL Service parameter kbps Minimum throughput demand in the downlink TPD Max – UL Service parameter kbps Maximum throughput demand in the uplink TPD Max – DL Service parameter kbps Maximum throughput demand in the downlink UL Service parameter kbps Average requested throughput in the uplink TP Average DL Service parameter kbps Average requested throughput in the downlink TP Offset Service parameter kbps Throughput offset f TP – Scaling Service parameter % Scaling factor L Body Service parameter dB Body loss P Min Terminal parameter dBm Minimum terminal power allowed P Max Terminal parameter dBm Maximum terminal power allowed nf Terminal parameter dB Terminal noise figure G Terminal parameter dB Terminal antenna gain L Terminal parameter dB Terminal loss N Ant – TX Terminal parameter None Number of antennas used for MIMO in transmission N Ant – RX Terminal parameter None Number of antennas used for MIMO in reception Terminal Wi-Fi equipment parameter None Maximum SU-MIMO gain NRDL NRUL TP Average Max G SU – MIMO 781 .3. 3. 11.2.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 Name Value Unit Description G Div – DL Terminal Wi-Fi equipment parameter dB Downlink STTD/MRC or SU-MIMO diversity gain UL Clutter parameter dB Additional uplink diversity gain G Div DL Clutter parameter dB Additional downlink diversity gain f SU – MIMO Clutter parameter None SU-MIMO gain factor L Indoor Clutter parameter dB Indoor loss L Path Propagation model result dB Path loss G Div F ICPDL Network parameter None Inter-technology downlink channel protection ratio for a frequency offset F between the interfered and interfering frequency channels M Shadowing – Model Monte Carlo simulations: Random result calculated from model standard deviation Coverage Predictions: Result calculated from cell edge coverage probability and model standard deviation dB Model shadowing margin M Shadowing – C  I Coverage Predictions: Result calculated from cell edge coverage probability and C/I standard deviation dB C/I shadowing margin Any interfering cell whose signal to thermal noise ratio is less than CNR Min will be discarded. 11.2 Calculation Quick Reference The following tables list the formulas used in calculations.1 Co.and Adjacent Channel Overlaps Calculation Name TX i  ic  F Start Value TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  – TX j  jc  TX  jc  L TX  jc  TX  ic  TX i  ic  – TX j  jc  H TX  ic  – TX  jc  i j r ACO H 782 TX  jc  TX  ic  TX  ic  j i j i i Min  F End  F Start  – Max  F Start  F Start – W Channel     MHz Start frequency for the channel number assigned to a cell MHz End frequency for the channel number assigned to a cell MHz Co-channel overlap bandwidth None Co-channel overlap ratio MHz Bandwidth of the lower-frequency adjacent channel overlap None Lower-frequency adjacent channel overlap ratio MHz Bandwidth of the higher-frequency adjacent channel overlap None Higher-frequency adjacent channel overlap ratio TX i  ic  – TX j  jc  W ACO L ---------------------------------TX i  ic  W Channel L W ACO TX  ic  Description TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  r ACO TX  jc  W CCO ----------------------------------TX i  ic  W Channel r CCO TX i  ic  – TX j  jc  TX  ic  j i j i Min  F End  F End  – Max  F Start  F Start      TX i  ic  – TX j  jc  W ACO TX i  ic  F Start + W Channel F End W CCO  TXi  ic  – N First – TXi  ic  Channel Channel   N ------------------------------------------------------- TX i  ic     CN FB   TX i  ic  F Start – FB +  W Channel + ICS FB  Unit TX j  jc  TX i  ic  Min  F End  F End  TX  ic  TX  jc  TX  ic  i j i + W Channel – Max  F Start  F End     TX i  ic  – TX j  jc  W ACO H ---------------------------------TX i  ic  W Channel .Atoll 3. a. 2.2.4 Interference Calculation (DL) Name TX j  jc  I DL TX i  ic  – TX j  jc  fO TX  jc  j f TL – DL Inter – Tech I DL Value TX j  jc  C DL TX i  ic  – TX j  jc  + fO TX j  jc  Inter – Tech + f TL – DL + I DL TX i  ic  – TX j  jc  10  Log  r O   TX j  jc  10  Log  TLDL    TX k   P DL – Rec  -------------------------------------- F  TX i  ic  TX k    TX  ICP DL k  783 .and adjacent channel overlap dB Interference reduction factor due to downlink traffic load W Downlink inter-technology interference TX  ic  – TX  jc  i j TX  ic  – TX  jc  i j r ACO r ACO TX  ic  – TX  jc  i j + r ACO L H TX  ic  i  – f ACS – FB  TX  ic  – TX  jc  TX  ic  – TX  jc  ---------------------------- j i j 10 r i  + r ACO  10  CCO      TX i  ic  rO TX j  jc  if W Channel  W Channel TX i  ic  – TX j  jc  TX  ic  i  – f ACS – FB TX  ic   TX  ic  – TX  jc  TX  ic  – TX  jc  --------------------------- W i j i j 10 Channel r i   --------------------+ r ACO 10 TX j  jc   CCO    W Channel   TX i  ic  TX j  jc  if W Channel  W Channel 11.2.3 Noise Calculation (DL) Name TX i  ic  n 0 – DL TX i  ic  TX i  ic  n 0 – DL + nf n DL Mi 11.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Name Value Unit Description None Adjacent channel overlap ratio None Total overlap ratio Unit Description dBm Received signal level dBm EIRP of a cell Value Unit Description  N TXi  ic   SCa – Used  n 0 + 10  Log  ------------------------ TXi  ic    N SCa – Total dBm Thermal noise for a cell dBm Downlink noise for a cell Unit Description dBm Interference generated by an interfering cell dB Interference reduction factor due to the co.3.Atoll 3.2 Signal Level Calculation (DL) Name TX  ic  i C DL EIRP Value TX i  ic  EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G –L Mi Mi Mi Mi – L Ant – L Body TX i  ic  TX i  ic  P DL +G TX i –L TX i 11. 7 Signal Level Calculation (UL) Name Value Mi EIRP UL – L Path – M Shadowing – Model – L Indoor + G M i C UL –L TX i M M i i – L Ant – L Body P M i EIRP UL With P M i TX i M Mi +G Mi –L Mi i = P Max without power control and P M i M i = P Eff after power control 11.5 C/N Calculation (DL) Name Value TX  ic  i Description dB Downlink C/N for a cell Unit Description TX  ic  i – n DL C DL TX  ic  i CNR DL Unit TX  ic  i With MIMO: CNR DL M i DL + G Div – DL + G Div 11.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 11.2.3.2.6 C/(I+N) Calculation (DL) Name Value TX i  ic  TX i  ic  CINR DL C DL    TXj  jc     I DL  -----------------10 – 10  Log  10   All TXj  jc       TX i  ic   n DL  + I Inter – Tech + -------------------10  DL 10    TX i  ic  With MIMO: CINR DL Mi TX i  ic  Downlink C/(I+N) for a cell DL + G Div – DL + G Div TX  jc   I + N  DL      + NRInter – Tech DL   dB     TX  ic  i    I j  n DL DL   ------------------ --------------------- 10 10 Inter – Tech  + NR Inter – Tech  10  +I 10  Log  + 10 DL    DL  All TXj  jc          dBm Total Noise (I+N) for a cell Unit Description dBm Received uplink signal level dBm Uplink EIRP of a user equipment Value Unit Description  N TXi  ic   SCa – Used  n 0 + 10  Log  ------------------------ TXi  ic    N SCa – Total dBm Thermal noise for a cell dBm Uplink noise for a cell Unit Description dBm Uplink interference received at a cell dB Interference reduction factor due to the co.9 Interference Calculation (UL) Name Mj I UL TX  ic  – TX  jc  i j fO 784 Value Mj TX i  ic  – TX j  jc  C UL + f O Mj + f TL – UL TX i  ic  – TX j  jc  10  Log  r O    .2.2.and adjacent channel overlap 11.8 Noise Calculation (UL) Name TX i  ic  n 0 – UL TX i  ic  TX i  ic  n 0 – UL + nf n UL TX i  ic  11.2. 12 Calculation of Total Cell Resources Name F TX i  ic  TX i  ic  D Sym – Useful Value TX i  ic  TX i  ic  TX i  ic  r CP -------------F D CP TX i  ic  D Symbol TX i  ic  R DL 6 TX i  ic  TX i  ic  D Sym – Useful + D CP TX  ic   1  -  N SCai – Data Floor  ----------------TX i  ic  D  Symbol 785 .2.2.3.10 C/N Calculation (UL) Name Value M C UL – n UL TX i  ic  Mi i CNR UL TX i  ic  Mi With MIMO: CNR UL + G Div – UL + UL G Div 11.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Name M Value M j 10  Log  TL UL j f TL – UL TX i  ic  NR UL TX i  ic   I + N  UL Unit Description dB Interference reduction factor due to the interfering mobile’s uplink traffic load  TX i  ic   IMj    n UL UL    non-seg M  --------------------- TX i  ic   --------------------------------------------i Inter – Tech 10   10  Log  + NR UL – n UL dB 10  10  + 10    All M j         All TX j  jc    TX i  ic  TX i  ic  Uplink noise at a cell dBm Total Noise (I+N) for a cell Unit Description dB Uplink C/N at a cell Unit Description dB Uplink C/(I+N) at a cell Unit Description W Channel  10 -----------------------------------TX i  ic  N SCa – Total kHz Inter-subcarrier distance 1 ------------------TX i  ic  F sec Useful symbol duration sec Cyclic prefix duration sec Symbol duration Symbols Total cell resources NRUL + n UL 11.11 C/(I+N) Calculation (UL) Name Value TX i  ic  Mi Mi CINR UL CNR UL – NR UL Mi TX i  ic  UL With MIMO: CINR UL + G Div – UL + G Div 11.2.Atoll 3. 3. Cell Capacity. and Per-user Throughput Calculation Name Value TX i  ic  R DL M i CTP P – DL With MIMO (AMS):  B Mi B Mi   1 + f SU – MIMO  G SU – MIMO – 1   if CNR DL M M Mi Mi Mi f TP – Scaling .– TP Offset CTP E – UL  -----------------------100 Mi Mi TX i  ic  Mi CTP P – UL  TL UL – Max M M Mi i i Cap P – UL   1 – BLER  B UL     M Mi f TP – Scaling .– TP Offset CTP E – DL  -----------------------100 Mi Mi TX i  ic  Mi CTP P – DL  TL DL – Max M M Mi i i Cap P – DL   1 – BLER  BDL  Mi Mi f TP – Scaling .0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 11.2.Atoll 3.– TP Offset Cap E – UL  -----------------------100 i Cap A – UL 786 M M Cap E – UL kbps Downlink effective MAC cell capacity kbps Downlink application cell capacity kbps Downlink peak MAC throughput per user kbps Downlink effective MAC throughput per user kbps Downlink application throughput per user kbps Uplink peak MAC channel throughput kbps Uplink effective MAC channel throughput kbps Uplink application channel throughput kbps Uplink peak MAC cell capacity kbps Uplink effective MAC cell capacity kbps Uplink application cell capacity TX i  ic  i i CTP P – UL   1 – BLER  B UL     Mi Downlink peak MAC cell capacity  T AMS Mi Cap P – UL kbps B UL TX i  ic  i CTP A – UL Downlink application channel throughput Mi B UL Max Mi if CNR DL CTP E – UL kbps Mi Cap E – DL ----------------------TX i  ic  N Users – DL Mi Mi Downlink effective MAC channel throughput Mi Cap P – DL ----------------------TX i  ic  N Users – DL Mi PUTP A – DL Mi Mi kbps TX i  ic  i i CTP P – DL   1 – BLER  B DL  Cap P – DL Downlink peak MAC channel throughput  T AMS Mi CTP A – DL kbps DL TX i  ic  CTP E – DL Description M i B DL Max =  DL  Unit Mi Mi .– TP Offset Cap E – DL  -----------------------100 Cap E – DL Cap A – DL PUTP P – DL PUTP E – DL Mi Mi Mi f TP – Scaling .13 Channel Throughput.– TPOffset PUTP E – DL  -----------------------100 TX i  ic  R UL Mi CTP P – UL With MIMO (AMS):  Mi =  B UL    1 + f SU – MIMO  G SU – MIMO – 1   M Mi f TP – Scaling . – TP Offset PUTP E – UL  -----------------------100 11.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Name Value M M i PUTP P – UL PUTP E – UL Mi PUTP A – UL Description kbps Uplink peak MAC throughput per user kbps Uplink effective MAC throughput per user kbps Uplink application throughput per user i Cap P – UL ----------------------TX  ic  i N Users – UL M Mi Unit i Cap E – UL ----------------------TX  ic  i N Users – UL Mi M M f TP – Scaling i i .14 Scheduling and Radio Resource Management Name Value Unit Description Sel Mi R Min – DL TPD Min – DL --------------------------- None Resources allocated to a mobile to satisfy its minimum throughput demand in downlink Sel Mi R Min – UL TPD Min – UL --------------------------- None Resources allocated to a mobile to satisfy its minimum throughput demand in uplink None Remaining downlink cell resources after allocation for minimum throughput demands R Min – UL None Remaining uplink cell resources after allocation for minimum throughput demands Sel Mi Sel Mi kbps Remaining throughput demand for a mobile in downlink Sel Mi Sel Mi kbps Remaining throughput demand for a mobile in uplink None Remaining resource demand for a mobile in downlink None Remaining resource demand for a mobile in uplink None Resources allocated to a mobile to satisfy its maximum throughput demand in downlink None Resources allocated to a mobile to satisfy its maximum throughput demand in uplink Sel Mi Sel Mi CTP P – DL Sel Mi Sel Mi CTP P – UL TX i  ic  R Rem – DL TX i  ic  R Rem – UL Sel Mi TPD Rem – DL Sel Mi TPD Rem – UL Sel Mi RD Rem – DL TX i  ic  TL DL – Max –  Sel Mi R Min – DL Sel Mi TX  ic  i TL DL – Max –  M Sel i Sel Mi TPD Max – DL – TPD Min – DL TPD Max – UL – TPD Min – UL Sel Mi TPD Rem – DL ---------------------------Sel Mi CTP P – DL Sel Mi Sel Mi RD Rem – UL TPD Rem – UL ---------------------------- Sel i R Max – DL  Mi R Rem – DL - Min  RD Rem – DL -------------------N   M Sel Mi R Max – UL Sel Mi CTP P – UL Sel Sel TX i  ic  TX i  ic  R Rem – UL  Mi - Min  RD Rem – UL -------------------N   787 .Atoll 3.2.3. – TP Offset 100 M Sel i R UL M Sel i  CTP P – UL Sel Sel Mi   Mi   UTPP – UL   1 – BLER  B UL      Sel Mi Sel Mi Sel Mi f TP – Scaling .0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 Name Value Unit Description Sel Sel   M  Mi i     R Max – DL  CTP E – DL     Sel   M  Site i - Max  1 ----------------------------------------------------------------------------------------------------- Sel Sel  M  Mi  Site i  TP – R  CTP  Min – DL  BH – DL E – DL     Sel   M i  Site None Site backhaul overflow ratio in downlink Sel Sel   Mi   Mi    R Max – UL  CTP E – UL     Sel   M i  Site  Max 1 -------------------------------------------------------------------------------------------------------  Sel Sel  Mi    Mi  Site  R Min – UL  CTP E – UL   TP BH – UL –    Sel   M i  Site None Site backhaul overflow ratio in uplink None Total resources assigned to a mobile in downlink (Downlink traffic load of the mobile) None Total resources assigned to a mobile in uplink (Uplink traffic load of the mobile) Unit Description kbps Downlink peak MAC user throughput kbps Downlink effective MAC user throughput kbps Downlink application user throughput kbps Uplink peak MAC user throughput kbps Uplink effective MAC user throughput kbps Uplink application user throughput  Site BHOF DL   Site BHOF DL  Sel Sel i TL DL M = Sel i R DL M Sel Mi Sel Mi R  Mi   Mi Max – DL  CTP P – DL  R Min – DL  CTP P – DL +  -----------------------------------------------Site     BHOF DL  -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – DL Sel Sel Mi TL UL Sel Mi = R UL M Sel i M Sel i Mi  Mi   R Max – UL  CTP P – UL  R Min – UL  CTP P – UL +  -----------------------------------------------Site     BHOF UL  -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – UL 11.3.15 User Throughput Calculation Name Sel Mi UTP P – DL Sel Mi UTPE – DL Sel Mi UTP A – DL M Sel i UTP P – UL Sel Mi UTP E – UL Sel Mi UTP A – UL 788 Value Sel Mi R DL Sel Mi  CTP P – DL Sel Sel Mi  Mi    UTP P – DL   1 – BLER  B DL      Sel i f TP – Scaling M Sel Mi Sel Mi UTP E – DL  ------------------------.– TP Offset UTP E – UL  -----------------------100 .2.Atoll 3. "Coverage Display Types" on page 790. The bar graph displays cells whose received signal levels are higher than their C/N thresholds and are within a 30 dB margin from the highest signal level.1 Profile View The point analysis profile view displays the following calculation results for the selected transmitter based on the calculation algorithm described in "Signal Level Calculation (DL)" on page 803.3. So.1. see the Administrator Manual. for example a smaller value for improving the calculation speed.1. You can use a value other than 30 dB for the margin from the highest interference level. and L Body are not used in the calculations performed for the profile view.2 Reception View Analysis provided in the reception view is based on path loss matrices. see "Path Loss Calculation Prerequisites" on page 57 for more information). L Ant . L Mi . You can use a value other than 30 dB for the margin from the highest signal level.1 Signal Level Coverage Predictions The following coverage predictions are based on the received signal levels: • • • Coverage by Transmitter Coverage by Signal Level Overlapping Zones For these calculations. Interference level bar graphs show the interference levels in decreasing order. 789 .3.Atoll 3. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. Reception level bar graphs show the signal levels or C/N in decreasing order.3. see: • • "Coverage Area Determination" on page 790. 11. The bar graph displays cells whose C/N are higher than the minimum interferer C/ N threshold and whose interference levels are within a 30 dB margin from the highest interference level. Atoll displays the best server signal level and interference from other cells. and L Body are not considered in the calculations performed for the signal level based coverage predictions. The maximum number of bars in the graph depends on the highest interference level. see the Administrator Manual. 11.1 Point Analysis 11. you can display received signal levels from the cells for which calculated path loss matrices are available.3. see "Signal Level Calculation (DL)" on page 803 For more information on coverage area determination and available display options. you can display the received signal level from the best server and interfering signal levels from other cells for which calculated path loss matrices are available.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 11.2. for example a smaller value for improving the calculation speed.3 Interference View Analysis provided in the interference view is based on path loss matrices. For more information on defining a different value for this margin. then determines the selected display parameter on each pixel inside the cell’s calculation area. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction.3. The maximum number of bars in the graph depends on the signal level of the best server. For each cell.G M i M i M i . 11. For more information on signal level calculations.3.1. For more information on defining a different value for this margin.3. So.2 Coverage Predictions 11. Atoll calculates the received signal level. L Ant .G Mi Mi Mi .3 Available Calculations 11. L M i TX  ic  i • Downlink signal level C DL • Path loss L Path • Total losses L Total . • • • If M = 0 dB.. dBµV.e. TX  ic  i  MinimumThreshold  C DL TX  ic  TX  ic  i i   or L Total or L Path   MaximumThreshold AND TX  ic  i C DL TX  jc  j  Best  C DL  – M ji Where M is the specified margin (dB). The Best function considers the highest value from a list of values. The 2nd Best function considers the second highest value from a list of values.Atoll 3.e.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 Coverage Area Determination Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine coverage areas to display. . dBµV/m) Best Signal Level (dBm. and other criteria such as: • • • • • • • 790 Signal Level (dBm. TX i  ic  MinimumThreshold  C DL TX  ic  TX  ic  i  or L i  Total or L Path   MaximumThreshold  AND TX i  ic  C DL TX  jc  nd j  2 Best  C DL  ji  –M  Where M is the specified margin (dB). Atoll determines the best cell (i. There are three possibilities.. Coverage consists of several independent layers that can be displayed and hidden on the map. If M = 0 dB. Number of Servers: Atoll evaluates the number of cells that cover a pixel (i.e. If M = -2 dB. the cell with the highest signal level) and evaluates the path loss from this cell. Atoll considers pixels where the received signal level from TXi(ic) is either the second highest or within a 2 dB margin from the second highest. If M = 2 dB. the cell with the highest signal level) and evaluates the total losses from this cell. Best Server Total Losses (dB): Where cell coverage areas overlap. Atoll considers pixels where the received signal level from TXi(ic) is the highest. dBµV/m): Where cell coverage areas overlap. Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display parameter is greater than or equal to the defined thresholds values. dBµV.. TX  ic  i  MinimumThreshold  C DL • TX  ic  TX  ic  i i   or L Total or L Path   MaximumThreshold Best Signal Level and a Margin The coverage area of each cell TXi(ic) corresponds to the pixels where.3. the pixel falls within the coverage areas of these cells). Atoll considers pixels where the received signal level from TXi(ic) is 2 dB higher than the received signal levels from the cells which are 3rd best servers. • All Servers The coverage area of each cell TXi(ic) corresponds to the pixels where. Atoll considers pixels where the received signal level from TXi(ic) is either the highest or within a 2 dB margin from the highest. Atoll considers pixels where the received signal level from TXi(ic) is 2 dB higher than the received • • • signal levels from the cells which are 2nd best servers. It is possible to display the coverage predictions with colours depending on any transmitter or cell attribute. • Second Best Signal Level and a Margin The coverage area of each cell TXi(ic) corresponds to the pixels where. Atoll considers pixels where the received signal level from TXi(ic) is the second highest. If M = 2 dB. Atoll keeps the highest value of the signal level. Path Loss (dB) Total Losses (dB) Best Server Path Loss (dB): Where cell coverage areas overlap. Atoll determines the best cell (i. If M = -2 dB. Atoll 3. 791 . see: • • "Signal Level Calculation (DL)" on page 803. and L Body ) when calculating the required parameter. Atoll calculates the received signal level or C/N level at each pixel. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. The properties of the non-interfering probe receiver are set by selecting a terminal.3 C/(I+N)-based Coverage Predictions The following coverage predictions are based on the received signal levels. For more information on C/N level calculations. "C/N Calculation (UL)" on page 812. Coverage Area Determination These coverage predictions are all best server coverage predictions. The properties of the non-interfering probe receiver are set by selecting a terminal. noise.3. It is possible to display the Effective Signal Analysis (DL) coverage prediction with colours depending on the following display options: • • Signal Level (DL) (dBm) C/N Level (DL) (dB) It is possible to display the Effective Signal Analysis (UL) coverage prediction with colours depending on the following display options: • • Signal Level (UL) (dBm) C/N Level (UL) (dB) 11.2. and interference.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 11. Coverage consists of several independent layers that can be displayed and hidden on the map.G Mi Mi Mi . • • • • • • • • Coverage by C/(I+N) Level (DL) Service Area Analysis (DL) Coverage by Throughput (DL) Coverage by Quality Indicator (DL) Coverage by C/(I+N) Level (UL) Service Area Analysis (UL) Coverage by Throughput (UL) Coverage by Quality Indicator (UL) These coverage predictions take into account the receiver characteristics ( L Mi . see: • • "C/N Calculation (DL)" on page 806. L Ant . Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. Atoll calculates the received signal level. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. For more information on coverage area determination and available display options.e. and a service. a mobility type. For these calculations. L Ant . and take into account the receiver characteristics ( L • • M i .. i. and interference at each pixel. For more information on signal level calculations. total noise.2. "Signal Level Calculation (UL)" on page 809. and L Body ) when calculating the required parameter: Effective Signal Analysis (DL) Effective Signal Analysis (UL) For these calculations. a mobility type. see: • • "Coverage Area Determination" on page 791.3. the coverage area of each cell comprises the pixels where the cell is the best server.G M i M i M i .2 Effective Signal Analysis Coverage Predictions The following coverage predictions are based on the received signal levels and noise. "Coverage Display Types" on page 791. Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display type parameter is greater than or equal to the defined thresholds values. Best server for each pixel is calculated as explained in "Best Server Determination" on page 815. The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. and a service.3. see "Path Loss Calculation Prerequisites" on page 57 for more information). The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data and can be defined separately for each coverage prediction. "Noise Rise Calculation (UL)" on page 811 For more information on throughput calculations. and the uplink coverage predictions are based on the uplink noise rise values. "Coverage Display Types" on page 792. see: • "Channel Throughput. For more information on C/(I+N).Atoll 3. It is possible to display the Coverage by C/(I+N) Level (UL) coverage prediction with colours depending on the following display options: • • • 792 C/(I+N) Level (UL) (dB) Total Noise (I+N) (UL) (dBm) Transmission Power (UL) (dBm) . and Per-user Throughput Calculation" on page 817. From the C/(I+N).e. For more information on coverage area determination and available display options. Coverage consists of several independent layers that can be displayed and hidden on the map. It is possible to display the Coverage by C/(I+N) Level (DL) coverage prediction with colours depending on the following display options: • • C/(I+N) Level (DL) (dB) Total Noise (I+N) (DL) (dBm) It is possible to display the Service Area Analysis (DL) coverage prediction with colours depending on the following display options: • • • Bearer (DL) Modulation (DL): Modulation used by the bearer Service It is possible to display the Coverage by Throughput (DL) coverage prediction with colours depending on the following display options: • • • • • • • • • Peak MAC Channel Throughput (DL) (kbps) Effective MAC Channel Throughput (DL) (kbps) Application Channel Throughput (DL) (kbps) Peak MAC Cell Capacity (DL) (kbps) Effective MAC Cell Capacity (DL) (kbps) Application Cell Capacity (DL) (kbps) Peak MAC Throughput per User (DL) (kbps) Effective MAC Throughput per User (DL) (kbps) Application Throughput per User (DL) (kbps) It is possible to display the Coverage by Quality Indicator (DL) coverage prediction with colours depending on the following display options: • Quality indicators available in the document (Quality Indicators table): Atoll calculates the downlink C/(I+N) levels received from the best serving cells at each pixel of their coverage areas. Best server for each pixel is calculated as explained in "Best Server Determination" on page 815. i. see "Path Loss Calculation Prerequisites" on page 57 for more information). Coverage Display Types A pixel of a coverage area is coloured if the calculated value of the selected display type parameter is greater than or equal to the defined thresholds values. These parameters can either be calculated by Atoll during the Monte Carlo simulations.3. it determines the value of the selected quality indicator from the quality graphs defined in the Wi-Fi equipment of the selected terminal. and bearer calculations. or set manually by the user for all the cells.. Cell Capacity. the coverage area of each cell comprises the pixels where the cell is the best server. see: • • • "C/(I+N) and Bearer Calculation (DL)" on page 807. Then. Atoll determines the best bearer available on each pixel. see: • • "Coverage Area Determination" on page 792. Coverage Area Determination These coverage predictions are all best server coverage predictions. "C/(I+N) and Bearer Calculation (UL)" on page 813. for the calculated C/(I+N) and bearer. Coverage predictions are generated using a bilinear interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes. (I+N).0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 The downlink coverage predictions are based on the downlink traffic loads of the cells. it determines the value of the selected quality indicator from the quality graphs defined in the Wi-Fi equipment of the best serving cell.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 It is possible to display the Service Area Analysis (UL) coverage prediction with colours depending on the following display options: • • • Bearer (UL) Modulation (UL): Modulation used by the bearer Service It is possible to display the Coverage by Throughput (UL) coverage prediction with colours depending on the following display options: • • • • • • • • • Peak MAC Channel Throughput (UL) (kbps) Effective MAC Channel Throughput (UL) (kbps) Application Channel Throughput (UL) (kbps) Peak MAC Cell Capacity (UL) (kbps) Effective MAC Cell Capacity (UL) (kbps) Application Cell Capacity (UL) (kbps) Peak MAC Throughput per User (UL) (kbps) Effective MAC Throughput per User (UL) (kbps) Application Throughput per User (UL) (kbps) It is possible to display the Coverage by Quality Indicator (UL) coverage prediction with colours depending on the following display options: • Quality indicators available in the document (Quality Indicators table): Atoll calculates the uplink C/(I+N) levels received at the best serving cells from each pixel of their coverage areas. 793 . From the C/(I+N). Atoll calculates the remaining parameters for each subscriber in the list that has a serving base station assigned. Atoll performs two random trials.3 Calculations on Subscriber Lists When calculations are performed on a list of subscribers by running the Automatic Server Allocation. for the calculated C/(I+N) and bearer. Atoll calculates the following parameters for each subscriber in the list whose Lock Status is set to None. "Signal Level Calculation (UL)" on page 809.Atoll 3. Atoll calculates the following parameters for each subscriber in the list that has a serving base station assigned and whose Lock Status is set to None or Server. Mechanical Downtilt (  ): Angle with respect to the horizontal for pointing the subscriber terminal antenna towards its serving base station.4.4 Monte Carlo Simulations The simulation process is divided into two steps. For more information. Then. 11. "Noise Rise Calculation (UL)" on page 811. Atoll determines the best bearer available on each pixel.1 User Distribution During each simulation. see: • • • • • • "Signal Level Calculation (DL)" on page 803. • Generating a realistic user distribution as explained in "User Distribution" on page 793. • • Azimuth (  ): Angle with respect to the north for pointing the subscriber terminal antenna towards its serving base station.3.3. 11. using the properties of the default terminal and service. The first random trial generates the number of users and their activity status as explained in the following sections depending on the type of traffic input. "Throughput Calculation" on page 816. 11. "C/(I+N) and Bearer Calculation (DL)" on page 807. The resulting user distribution complies with the traffic database and maps selected when creating simulations.3. • Scheduling and Radio Resource Management as explained under "Simulation Process" on page 797. • "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 794. • Serving Base Station and Reference Cell as described in "Best Server Determination" on page 815. Atoll calculates the path loss again for the subscriber locations and heights because the subscriber heights can be different from the default receiver height used for calculating the path loss matrices. "C/(I+N) and Bearer Calculation (UL)" on page 813.3. Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data. 0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks • ©Forsk 2015 "Simulations Based on Sector Traffic Maps" on page 795. i.e. and whether they are indoor or outdoor according to the percentage of indoor users per clutter class. • The average duration of a call (seconds) D Call . the average duration of each voice call. f Act and f Act . User profiles model the behaviour of the different user categories. the uplink and downlink activity UL DL factors defined for the voice type service v. the number of users of each user profile is calculated from the line length (L) and the user profile density (DUP) • The number of users is a direct input when a user profile traffic map is composed of points. The number of users of each user profile is calculated from the surface area (SEnv) of each environment class map (or each polygon) and the user profile density (DUP).. (users per km): N Users = L  D UP Atoll calculates the probability for a user being active at a given instant in the uplink and in the downlink according to the service usage characteristics described in the user profiles. Voice Service (v) User profile parameters for voice type services are: • • The user terminal equipment used for the service (from the Terminals table). add the following lines in the Atoll.3. number of users of a user profile per km². If the map is composed of points.1. i. The average number of calls per hour N Call .3.1 Simulations Based on User Profile Traffic Maps and Subscriber Lists User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of user profiles with an associated mobility type and a given density.. the number of voice calls or data sessions. a second random trial is performed to obtain their geographical locations weighted according to the clutter classes. Atoll determines the total number of users attempting connection in each simulation based on the Poisson distribution. User profile traffic maps: Each polygon or line of the map is assigned a density of users with a given user profile and mobility type.4. To have the same total number of users in each simulation of a group. Once all the user characteristics have been determined. This may lead to slight variations in the total numbers of users in different simulations. N Call  D Call Calculation of the service usage duration per hour ( p 0 : probability of an active call): p 0 = ---------------------------3600 Calculation of the number of users trying to access the service v ( n v ): n v = N Users  p 0 The activity status of each user depends on the activity periods during the call. Each user profile contains a list of services and parameters describing how these services are accessed by the user.ini file: [Simulation] RandomTotalUsers=0 11. each point is assigned a number of users with given user profile and mobility type. or the volume of the data transfer in the uplink and the downlink in each data session. i. N Users = S Env  D UP • In case of user profile traffic maps composed of lines. Calculation of activity probabilities: UL DL Probability of being inactive: p Inactive =  1 – f Act    1 – f Act  UL UL DL Probability of being active in the uplink: p Active = f Act   1 – f Act  DL DL UL Probability of being active in the downlink: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active in the uplink and downlink both: p Active = f Act  f Act 794 .e.Atoll 3..e. Fixed subscribers listed in subscriber lists have a user profile assigned to each of them. The service and the activity status of each user are randomly drawn in each simulation.1.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Calculation of number of users per activity status: Number of inactive users: n v – Inactive = n v  p Inactive UL UL Number of users active in the uplink: n v – Active = n v  p Active DL DL Number of users active in the downlink: n v – Active = n v  p Active UL + DL UL + DL Number of users active in the uplink and downlink both: n v – Active = n v  p Active Therefore. active on UL.3. 11. inactive on both links. Either throughput demands per service or the number of active users per service are assigned to the coverage areas of each transmitter. the user distribution between services as well as the activity status distribution between users can be different in each of them. The average number of data sessions per hour N Session . Therefore. respectively. the average number of users per service and average numbers of inactive. or active on DL only. Data Service (d) User profile parameters for data type services are: • • The user terminal equipment used for the service (from the Terminals table). For each transmitter TXi and each service s. will correspond to calculated distributions.Atoll 3. • The average data volume (in kBytes) transferred in the downlink V • The average throughputs in the downlink Calculation of activity probabilities: f UL DL TP Average DL and the uplink and the uplink V UL TP Average UL during a session. for the service d. a user can be either active on both links. Live traffic data from the OMC is spread over the best server coverage areas of the transmitters included in the traffic map. UL DL N Session  V  8 N Session  V  8 DL = -----------------------------------------.3. active on DL and active on UL and DL users.2 Simulations Based on Sector Traffic Maps Sector traffic maps per sector are also referred to as live traffic maps. if you calculate several simulations at once. But if you check each simulation.4. active on UL only. • Sector Traffic Maps (Throughputs) 795 .and f = -----------------------------------------UL DL TP Average  3600 TP Average  3600 UL DL Probability of being inactive: p Inactive =  1 – f    1 – f  UL Probability of being active in the uplink: p Active = f DL UL DL  1 – f  Probability of being active in the downlink: p Active = f DL UL  1 – f  UL + DL Probability of being active in the uplink and downlink both: p Active = f UL f DL Calculation of number of users: Number of inactive users: n d – Inactive = N Users  p Inactive UL UL Number of users active in the uplink: n d – Active = N Users  p Active DL DL Number of users active in the downlink: n d – Active = N Users  p Active UL + DL UL + DL Number of users active in the uplink and downlink both: n d – Active = N Users  p Active Calculation of the number of active users trying to access the service d (nd): UL DL UL + DL n d = n d – Active + n d – Active + n d – Active The user distribution per service and the activity status distribution between the users are average distributions. 0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 Atoll calculates the number of active users of each service s on UL and DL in the coverage area of TXi as follows: N UL UL DL TP Cell TP Cell DL = ---------------------. and TP Average is the average downlink requested throughput of the service s.and N = ---------------------UL DL TP Average TP Average UL Where TP Cell is the total uplink throughput demand defined in the map for any service s for the coverage area of the DL transmitter. TP Cell is the total downlink throughput demand defined in the map for any service s for the coverage UL DL area of the transmitter. DL values.e. p Inactive 1 – p Inactive 796 . Atoll calculates the probability for a user being active in the uplink and in the downlink as follows: Users active in the uplink and downlink both are included in the N UL UL accurately determine the number of active users in the uplink ( n Active and N DL values. the number of active users on UL and DL in the transmitter At any given instant. n Active = Min  N UL DL  f Act N DL UL  f Act  UL Number of users active in the uplink: n Active = N DL UL Number of users active in the downlink: n Active = N UL DL UL + DL – n Active DL UL + DL – n Active UL + DL And. As for the other types of traffic maps. • Sector Traffic Maps (# Active Users) UL Atoll directly uses the defined N and N coverage area using the service s.e. f Act and f Act . Therefore.. Atoll considers both active and inactive users. Calculation of activity probabilities: UL DL Probability of being inactive: p Inactive =  1 – f Act    1 – f Act  UL UL DL Probability of being active in the uplink: p Active = f Act   1 – f Act  DL DL UL Probability of being active in the downlink: p Active = f Act   1 – f Act  UL + DL UL DL Probability of being active in the uplink and downlink both: p Active = f Act  f Act Calculation of the number of active users trying to access the service: We have: N UL UL + DL UL =  p Active + p Active   n and N DL UL + DL DL =  p Active + p Active   n Where. and both ( n Active ).Atoll 3. i. TP Average is the average uplink requested throughput of the service s. Calculation of number of users per activity status: UL UL + DL DL UL + DL  N  p Active N  p Active  UL + DL Number of users active in the uplink and downlink both: n Active = Min  -------------------------------------- -------------------------------------- or UL UL + DL DL + DL  p Active + p Active p Active + p UL Active  UL + DL simply. it is necessary to UL + DL DL ). n = n Active + n Active + n Active Calculation of the number of inactive users attempting to access the service: nv Number of inactive users: n Inactive = ---------------------------. in the downlink ( n Active ). the uplink and downlink activity UL DL factors defined for the service. The activity status of each user depends on the activity periods during the call.. n is the total number of active users in the transmitter coverage area using the service.3. i. 11.2 Simulation Process Each Monte Carlo simulation in Atoll Wi-Fi is a snap-shot of the network with resource allocation carried out over a duration of 1 second. Therefore. in each simulation. 3.3. 1. the simulation process. Sets initial values for the following parameters: TX  ic  i • Cell transmission power ( P DL ) is set to the value defined by the user. TL UL TX i  ic  .1: Wi-Fi Simulation Algorithm For each iteration k.4. the activity status distribution between users can be different in each of them. Figure 11. average numbers of inactive. The steps of this algorithm are listed below. Determines the downlink and uplink C/(I+N) and bearers for each of these mobiles as explained in "C/(I+N) and Bearer Calculation (DL)" on page 807 and "C/(I+N) and Bearer Calculation (UL)" on page 813 respectively. 797 . the simulation process. In fact. Determines the best servers for all the mobiles generated for the simulation as explained in "Best Server Determination" on page 815. • Cell loads ( TL DL Mi TX i  ic  TX i  ic  . the activity status of each user is randomly drawn. 2.3. Generates mobiles according to the input traffic data as explained in "User Distribution" on page 793. active on DL and active on UL and DL users correspond to the calculated distribution. Determines the mobiles which are within the service areas of their best serving cells as explained in "Service Area Calculation" on page 815. For each simulation. The simulation process can be summed up into the following iterative steps. But if you check each simulation. 4. active on UL.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 The activity status distribution between users is an average distribution. if you calculate several simulations at once. 5. • Mobile transmission power is set to the maximum mobile power ( P Max ). and NR UL ) are set to their current values in the Cells table.Atoll 3. if: TX i  ic  TL DL TX i  ic  k  TL DL TX i  ic  Req OR TL UL TX i  ic  k  TL UL TX i  ic  Req OR NR UL TX i  ic  k  NR UL Req 11. Calculates the user throughputs after allocating resources to each mobile as explained in "User Throughput Calculation" on page 823. Atoll stops the simulation in the following cases. the simulations also list the connection status of each mobile. UL. Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 7.) for the iteration k+1 using the new calculated loads as the current loads. TL UL  k – 1  k – 1 TX i  ic  k – NR UL  k – 1 TX  ic  i Req . i.e. TX i  ic  TL DL TX i  ic  Mi  RDL and TLUL = Mi  RUL = Mi Mi Calculation of Uplink Noise Rise: For each victim cell TXi(ic). Performs the convergence test to see whether the differences between the current and the new loads are within the convergence thresholds. or DL+UL. the uplink noise rise is calculated and updated by considering each interfering mobile Mj as explained in "Noise Rise Calculation (UL)" on page 811.) or if the mobile is not within the service area of its best server (step 4. and Per-user Throughput Calculation" on page 817.).1 and k if: TX i  ic  TL DL TX i  ic  k  TL DL TX i  ic  Req AND TL UL TX i  ic  k  TL UL TX i  ic  Req AND NR UL TX i  ic  k  NR UL Req No convergence: Simulation has not converged even after the last iteration. Repeats the above steps (from step 3. Simulation Results At the end of the simulation process. Performs radio resource management and scheduling to determine the amount of resources to allocate to each mobile according to the throughput demands of each mobile using the selected scheduler as explained in "Scheduling and Radio Resource Allocation" on page 820. and NR UL Req are the simulation convergence thresholds defined when creating the simulation. and noise rise values of all the cells according to the resources in use and the total resources as follows: Calculation of Traffic Loads: Atoll calculates the traffic loads for all the cells TXi(ic). DL. Cell Capacity. 8. Mobiles can be rejected due to: • • • • 798 No Coverage: If the mobile does not have any best serving cell (step 3.3.e. the minimum uplink throughput demand is higher than the uplink allocated bandwidth throughput (step 7. No Service: If the mobile is not able to access a bearer in the direction of its activity (step 5. 10.. the main results obtained are: • • • Downlink traffic load Uplink traffic load Uplink noise rise These results can be used as input for C/(I+N)-based coverage predictions. and can be written as follows: TX i  ic  TL DL = k i TX i  ic  TL UL = k k = TX  ic  i Req – TL DL k – TL UL TX i  ic  TX  ic  i Max  NR UL All TX  ic  i TX  ic  i If TL DL TX i  ic  k TX  ic  i Max  TL UL All TX  ic  i TX i  ic  NR UL TX  ic  i Max  TL DL  All TX  ic  .Atoll 3. 7. 9.) .) Resource Saturation: If all the cell resources are used up before allocation to the mobile or if. Updates the traffic loads. In addition to the above parameters.. k = Max Number of Iterations defined when creating the simulation. for a user active in uplink. Determines the channel throughputs at the mobile as explained in "Channel Throughput.). The convergence criteria are evaluated at the end of each iteration k. Convergence: Simulation has converged between iteration k . i.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 6. ) Connected mobiles (step 7.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 • Backhaul Saturation: If allocating resources to a mobile makes the effective MAC aggregate site throughputs exceed the maximum backhaul throughputs defined for the site. the co. Figure 11. In order to calculate the co. Once the start and end frequencies are known for the studied and other cells.4. the start and end frequencies of all the channels may not exactly coincide.2: Co-Channel and Adjacent Channel Overlaps The following sections describe how the co. the studied cell can be considered a victim of interference received from the other cells that might be interfering the studied cell. Connected DL: If a mobile active in DL is allocated resources in DL. calculations on subscriber lists. Channel bandwidths of cells can overlap each other with different ratios. calculation of coverage predictions.e. 11. and adjacent channel interference on the adjacent channel bandwidths. Connected DL+UL: If a mobile active in DL+UL is allocated resources in DL+UL. "Total Overlap Ratio Calculation" on page 802. and Monte Carlo simulations.Atoll 3.and adjacent overlaps and the total overlap ratio are calculated as respectively explained in: • • • "Co-Channel Overlap Calculation" on page 800. it is necessary to calculate the start and end frequencies of both channels (explained in "Conversion From Channel Numbers to Start and End Frequencies" on page 800)..and adjacent channel overlaps are calculated between the channels used by any studied cell TXi(ic) and any other cell TXj(jc) of the network.1 Co. Therefore. it receives co-channel interference on the channel bandwidth of TX i  ic  TX i  ic  N Channel .and Adjacent Channel Overlaps Calculation A Wi-Fi network can consist of cells that use different channel bandwidths. TX  ic  i If the studied cell is assigned a channel number N Channel . i.) can be: • • • Connected UL: If a mobile active in UL is allocated resources in UL. "Adjacent Channel Overlap Calculation" on page 801. In terms of interference calculation. 799 . 11. This condition is only verified if the simulation was created with the Backhaul capacity check box selected (step 7. corresponding to N Channel – 1 and TX i  ic  N Channel + 1 .3.and adjacent channel overlaps between two channels.4 Calculation Details The following sections describe all the calculation algorithms used in point analysis. 2 Co-Channel Overlap Calculation Input • TX i  ic  F Start TX j  jc  and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 800.Atoll 3. i. TX  ic  i F End TX  jc  j • and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 800. TX i  ic  and N Channel : First channel numbers the frequency band assigned to the cells TXi(ic) and TXj(jc).0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 11. the channel number you assign to a cell is considered for uplink and downlink both. • W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic). TX  ic  i Calculations Atoll first verifies that co-channel overlap exists between the cells TXi(ic) and TXj(jc). TX j  jc  and CN FB : Channel number step of the frequency bands assigned to cells TXi(ic) and TXj(jc).1 Conversion From Channel Numbers to Start and End Frequencies Input • TX  ic  i TX  jc  j F Start – FB and F Start – FB : Start frequency of the frequency band assigned to the cells TXi(ic) and TXj(jc). TX i  ic  TX j  jc  • W Channel and W Channel : Bandwidths of the channels assigned to cells TXi(ic) and TXj(jc).1. Calculations Channel numbers are converted into start and end frequencies as follows: For cell TXi(ic): TX i  ic  F Start TX i  ic  F End TX i  ic  TX  ic  TX i  ic   TXi  ic  – N First – TX i  ic  Channel Channel   N ------------------------------------------------------- TX i  ic       CN FB TX  ic  i i = F Start – FB +  W Channel + ICS FB  TX i  ic  = F Start + W Channel For cell TXj(jc): TX j  jc  TX j  jc  TX j  jc  TX j  jc  F End TX j  jc   N TXj  jc  – N First – TX j  jc  Channel Channel    ------------------------------------------------------- TX  jc     j   CN FB TX j  jc  F Start = F Start – FB +  W Channel + ICS FB  TX j  jc  = F Start – FB + W Channel Output TX i  ic  TX j  jc  • F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc). F Start – FB can represent the uplink or the downlink start frequencies ( F Start – FB – UL or F Start – FB – DL ).3. • F End TX i  ic  TX j  jc  and F End : End frequencies for the cells TXi(ic) and TXj(jc). • ICS FB • TX  ic  i TX i  ic  CN FB TX  jc  j and ICS FB : Inter-channel spacing of the frequency bands assigned to cells TXi(ic) and TXj(jc). TX j  jc  Atoll considers that the same channel number is assigned to a cell in the downlink and uplink.1.. 11.e.4. First – TX i  ic  First – TX j  jc  • N Channel • N Channel and N Channel : Channel numbers assigned to cells TXi(ic) and TXj(jc).4. 800 . 3 Adjacent Channel Overlap Calculation Input TX i  ic  TX j  jc  • F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 800.Atoll 3.4.3.1.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Co-channel overlap exists if: TX  ic  i TX  jc  j TX  ic  i F Start  F End AND F End TX  jc  j  F Start Otherwise there is no co-channel overlap. TX i  ic  TX j  jc  and F End TX  ic  i Calculations Atoll first verifies that adjacent channel overlaps exist between (the lower-frequency and the higher-frequency adjacent channels of) the cells TXi(ic) and TXj(jc). • W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic). Atoll determines the adjacent channel overlap ratio as follows: Bandwidth of the lower-frequency adjacent channel overlap: TX  ic  – TX  jc  i j W ACO L TX  jc  TX  ic  TX  jc  TX  ic  TX  ic  j i j i i = Min  F End  F Start  – Max  F Start  F Start – W Channel The lower-frequency adjacent channel overlap ratio is given by: TX  ic  – TX  jc  i j r ACO L TX i  ic  – TX j  jc  W ACO L = ---------------------------------TX i  ic  W Channel Bandwidth of the higher-frequency adjacent channel overlap: TX i  ic  – TX j  jc  W ACO H TX j  jc  TX i  ic  = Min  F End  F End  TX  ic  TX  jc  TX  ic  i j i + W Channel – Max  F Start  F End     The higher-frequency adjacent channel overlap ratio is given by: 801 . • F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel Numbers to Start and End Frequencies" on page 800. 11. Atoll calculates the bandwidth of the co-channel overlap as follows: TX i  ic  – TX j  jc  W CCO TX  jc  TX  ic  TX  jc  TX  ic  j i j i = Min  FEnd  F End  – Max  F Start  F Start      The co-channel overlap ratio is given by: TX  ic  – TX  jc  i j TX i  ic  – TX j  jc  W CCO = ---------------------------------TX i  ic  W Channel r CCO Output • TX i  ic  – TX j  jc  r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc). Adjacent channel overlap exists on the lower-frequency adjacent channel if: TX i  ic  TX i  ic  TX j  jc  F Start – W Channel  F End TX i  ic  TX j  jc  AND F Start  F Start Adjacent channel overlap exists on the higher-frequency adjacent channel if: TX i  ic  F End TX j  jc   F End TX i  ic  AND F End TX i  ic  TX j  jc  + W Channel  F Start Otherwise there is no adjacent channel overlap. 1.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 TX  ic  – TX  jc  i j W ACO H = ---------------------------------TX  ic  i W Channel TX  ic  – TX  jc  i j r ACO H The adjacent channel overlap ratio is given by: TX i  ic  – TX j  jc  r ACO TX i  ic  – TX j  jc  = r ACO L TX i  ic  – TX j  jc  + r ACO H Output TX i  ic  – TX j  jc  r ACO • : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).Atoll 3.3.4.4 Total Overlap Ratio Calculation Input TX i  ic  – TX j  jc  r CCO • : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co-Channel Overlap Calculation" on page 800. This means that TX j  jc  W Channel TX j  jc  TX j  jc  if the interfering cell transmits at X dBm over a bandwidth of W Channel . TX i  ic  TX j  jc  Calculations The total overlap ratio is: TX i  ic  – TX j  jc  rO        =         TX  ic  i  – f ACS – FB  TX  ic  – TX  jc  TX  ic  – TX  jc  ---------------------------- 10 j i j r i  + r ACO  10  CCO      TX i  ic  TX j  jc  if W Channel  W Channel TX  ic  i  – f ACS – FB TX i  ic   TX  ic  – TX  jc  TX  ic  – TX  jc  ---------------------------- TX i  ic  TX j  jc  W Channel 10 i j i j r  -------------------- TX  jc  if W Channel  W Channel + r ACO  10  CCO  j   W Channel   TX i  ic  W Channel The multiplicative factor --------------------is used to normalise the transmission power of the interfering cell TXj(jc). 11. and it interferes over a bandwidth less than W Channel .and adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc). . TX  ic  i W Channel the interference from this cell should not be considered at X dBm but less than that. The factor --------------------converts X dBm over TX  jc  j W Channel TX j  jc  TX j  jc  W Channel to Y dBm (which is less than X dBm) over less than W Channel . TX i  ic  – TX j  jc  • r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Adjacent Channel Overlap Calculation" on page 801. TX i  ic  • f ACS – FB : Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic). Output • 802 TX i  ic  – TX j  jc  rO : Total co. • W Channel and W Channel : Bandwidths of the channels assigned to the cells TXi(ic) and TXj(jc). • Mi For calculating the useful signal level from the best serving cell. TX i In coverage predictions. or mobile Mi. Calculations The received signal levels (dBm) from any cell TXi(ic) are calculated for a pixel.4. "C/(I+N) and Bearer Calculation (DL)" on page 807.3. "Noise Calculation (DL)" on page 804. "Interference Calculation (UL)" on page 810. Mi L Ant : Receiver terminal’s antenna attenuation calculated for the pixel. • M Shadowing – Model : Shadowing margin based on the model standard deviation.0) from Mi the antenna patterns of the antenna used by Mi. L Ant is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi. "C/(I+N) and Bearer Calculation (UL)" on page 813. For calculating the interfering signal level from any interferer. "Signal Level Calculation (UL)" on page 809.1 Signal Level Calculation (DL) Input TX  ic  i P DL • • G • L TX i TX i : Transmission power of the cell TXi(ic). subscriber. • • • • • • • • • • "Signal Level Calculation (DL)" on page 803. subscriber. Mi • L Body : Body loss defined for the service used by the pixel. or mobile Mi as follows: TX i  ic  C DL = EIRP TX i  ic  – L Path – M Shadowing – Model – L Indoor + G Mi –L Mi Mi Mi – L Ant – L Body Where EIRP is the effective isotropic radiated power of the cell calculated as follows: EIRP TX i  ic  TX i  ic  = P DL +G TX i –L TX i 803 . subscriber. subscriber. or mobile Mi. TX i • L Path : Path loss ( L Path = L Model + L Ant ). or mobile Mi.V) = (0. ( G : Total transmitter losses for the transmitter TXi ( L TX i TX i TX i = G Ant ). L Ant is determined in the direction (H. "C/N Calculation (DL)" on page 806. and C/(I+N) ratios are calculated on the downlink and uplink.Atoll 3. • L • G Mi Mi : Receiver terminal losses for the pixel. noise and interference. while the antenna is pointed towards Mi’s best serving cell. C/N. • L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model. : Transmitter antenna gain for the antenna used by the transmitter TXi. = L Total – DL ).4. 11. : Receiver terminal’s antenna gain for the pixel. or mobile Mi. "C/N Calculation (UL)" on page 812. shadowing margins are taken into account when the option "Shadowing taken into account" is selected. • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi. "Noise Calculation (UL)" on page 810.2.2 Signal Level and Quality Calculations The following sections describe how signal levels. "Interference Calculation (DL)" on page 804. • L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected. subscriber.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 11. 3. subscriber. The thermal noise density depends on the temperature. the cyclic prefix energy is excluded from C DL TX i  ic  words. TX  ic  i n DL TX  ic  i = n 0 – DL + nf M i Output • TX i  ic  n DL : Downlink noise for the cell TXi(ic). i. it remains constant for a given temperature.2. 11. or mobile. the energy of the useful symbol duration and the cyclic prefix energy. i. 11.e. the value of the thermal noise varies with the used bandwidth. DL  Independant of the option. • nf TX i  ic  TX  ic  i Mi : Noise figure of the terminal used for calculations by the pixel. or mobile Mi. • N SCa – Used : Number of used subcarriers defined for the frame configuration of a cell TXi(ic).2 Noise Calculation (DL) For determining the C/N and C/(I+N). Calculations The power spectral density of thermal noise is calculated as follows: n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz The thermal noise for a cell is calculated as:  N TXi  ic   TX i  ic  SCa – Used  n 0 – DL = n 0 + 10  Log  ------------------------TX i  ic    N SCa – Total The downlink noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel. T: Temperature in Kelvin. Output • TX i  ic  C DL : Received signal level from the cell TXi(ic) at the pixel. subscriber. However. • N SCa – Total : Total number of subcarriers defined for the frame configuration of a cell TXi(ic).2. 804 . Atoll calculates the downlink noise over the channel bandwidth used by the cell. or mobile Mi. The downlink noise comprises thermal noise and the noise figure of the equipment.e.. served by a cell TXi(ic) from other cells TXj(jc) can be defined as the signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the channels used by the cells TXi(ic) and TXj(jc). the factor 10  Log  1 – r CP . or mobile Mi. In other TX  ic   is added to C i .4.Atoll 3. subscriber.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 If you wish to exclude the energy corresponding to the cyclic prefix (guard interval) in the total symbol duration from the useful signal level.. Input • • K: Boltzmann’s constant. you must add the following lines in the Atoll.3 Interference Calculation (DL) The interference received by any pixel. subscriber. and on the traffic loads of the interfering cells TXj(jc).ini file: [WiMAX] ExcludeCPFromUsefulPower = 1 TX i  ic  When this option is active. interference levels are calculated for the total symbol durations. The used bandwidth depends on the number of used subcarriers.4.   Independent of the option. subscriber.. In other TX i  ic  words.3. the factor 10  Log  1 – r CP  is added to C DL . the received signal levels from interferers already include M Shadowing – Model . • TX j  jc  TL DL : Downlink traffic load of the interfering cell TXj(jc). Calculation of traffic loads is explained in "Simulation Process" on page 797. you must add the following lines in the Atoll.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Input TX  jc  j • C DL : Received signal level from the cell TXi(ic) as explained in "Signal Level Calculation (DL)" on page 803. As the received signal levels from interferers already include M Shadowing – Model . i.and adjacent channel overlap between the cells TXi(ic) and TXj(jc) is calculated as follows: TX i  ic  – TX j  jc  fO TX i  ic  – TX j  jc  = 10  Log  r O    Interference reduction due to interfering cell’s traffic load: The interference reduction factor due to the interfering cell’s traffic load is calculated as follows: TX j  jc  TX j  jc  f TL – DL = 10  Log  TL DL  Inter – Tech I DL   is the inter-technology downlink interference from transmitters of an external network (linked document of any technology) calculated as follows: 805 . or entered manually for each cell. Traffic loads can either be calculated using Monte Carlo simulations. interference levels are calculated for the total symbol durations.and adjacent channel overlap between the studied and the interfering cells: Interference reduction due to the co. shadowing margins are taken into account when the option "Shadowing taken into account" is selected. see "Shadow Fading Model" on page 90). the energy of the useful symbol duration and the cyclic prefix energy. • TX i  ic  – TX j  jc  rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co. M Shadowing – C  I is added to the signal levels from interferers in order to achieve the ratio M Shadowing – Model – M Shadowing – C  I : TX j  jc  C DL TX j  jc  = C DL + M Shadowing – C  I In coverage predictions. or mobile Mi as follows: TX j  jc  I DL TX j  jc  = C DL TX i  ic  – TX j  jc  + fO TX j  jc  Inter – Tech + f TL – DL + I DL If you wish to exclude the energy corresponding to the cyclic prefix (guard interval) in the total symbol duration from the useful signal level.Atoll 3. the ratio M Shadowing – Model – M Shadowing – C  I is applied to the interfering signals (for more information. the cyclic prefix energy is excluded from C DL TX i  ic  .ini file: [WiMAX] ExcludeCPFromUsefulPower = 1 TX i  ic  When this option is active. Calculations Interference (dBm) from any cell TXj(jc) is calculated for a pixel. Calculations for the interference reduction factors due to channel overlapping and traffic load are explained below: Interference reduction due to the co. In Monte Carlo simulations. • M Shadowing – C  I : Shadowing margin based on the C/I standard deviation.and Adjacent Channel Overlaps Calculation" on page 799. In coverage predictions.e. as explained in "Signal Level Calculation (DL)" on page 803. TX i  ic  Mi Mi or mobile Mi. or mobile Mi. Output TX j  jc  • I DL • I DL : Downlink interference received at the pixel. subscriber.4 C/N Calculation (DL) Input • TX i  ic  C DL : Received signal level from the cell TXi(ic) at the pixel. or mobile Mi from any interfering cell TXj(jc). preamble power from WiMAX cells. and F  TX i  ic  TX k  ICP DL is the inter-technology downlink channel protection ratio for a frequency offset F between the interfered and interfering frequency channels of TXi(ic) and TXk. and L Body is the body loss defined for the service used by the pixel. subscriber. or Mi mobile Mi. L Path is the path loss from the external transmitters to the pixel. subscriber. subscriber. and TD-SCDMA cells. or mobile Mi. • Mi B DL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. or mobile Mi. • N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel. L Mi is the receiver terminal losses for the pixel.2.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks Inter – Tech I DL  = ©Forsk 2015 TX – External EIRP DL – L Path – L Indoor + G M i –L M i M i M i Inter – Tech – L Ant – L Body – f IRF All External TXs TX – External Where EIRP DL is the downlink EIRP of the external transmitter. or mobile Mi. • T AMS : AMS threshold defined for the cell TXi(ic). subscriber. • B DL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. subscriber. CDMA2000. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 803. Inter – Tech : Downlink inter-technology interference. • • 806 Mobility  M i  : Mobility used for the calculations. L Indoor is the indoor losses taken into account when the option "Indoor coverage" is selected. M i subscriber. or mobile location.3. 11. TX i  ic  • n DL : Downlink noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 804. or mobile Mi. subscriber. and downlink cell power from Wi-Fi cells. G Mi is the receiver terminal’s antenna Mi gain for the pixel. total power from UMTS. . maximum power from LTE cells. subscriber. subscriber. or mobile Mi. M i BLER  BDL : Downlink block error rate read from the graphs available in the Wi-Fi equipment assigned to the terminal   used by the pixel. L Ant is the receiver terminal’s antenna attenuation calculated for the pixel.Atoll 3.4. • T B : Bearer selection thresholds of the bearers defined in the Wi-Fi equipment used by Mi’s terminal. subscriber. TX k P DL – Rec is calculated based on the EIRP from GSM cells. Calculation of the Downlink Inter-technology Interference The downlink inter-technology interference is calculated as follows: Inter – Tech I DL TX k   P DL – Rec  --------------------------------------- = F  TX i  ic  TX k    TX k  ICP DL  TX k Here P DL – Rec is the received downlink power from an interfering cell TXk belonging to another technology. TX  ic  i • N Ant – TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic). 11. BLER  BDL . Output • TX i  ic  CNR DL : C/N from the cell TXi(ic) at the pixel. subscriber.2. First Atoll calculates the received signal level from the studied cell (as explained in "Signal Level Calculation (DL)" on page 803) at the pixel. Atoll does not try to find the direction of the strongest signal. or mobile under study.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Calculations The C/N for a cell TXi(ic) are calculated as follows for any pixel. subscriber. MIMO – STTD/MRC and SU-MIMO Diversity Gains: Once the bearer is known. The receiver terminal is always considered to be oriented towards its best server. Mobility  M i  .and adjacent channel overlap between the studied and the interfering cells. • TX i  ic  n DL : Downlink noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 804. In the case of NLOS between the receiver and the best server. Input • TX i  ic  C DL : Received signal level from the cell TXi(ic) at the pixel. the STTD/MRC or SU-MIMO diversity gain. 807 . corresponding to the bearer is applied to its selection threshold. or mobile Mi G Div is also applied. Finally. or mobile from all the interfering cells (as explained in "Interference Calculation (DL)" on page 804). or mobile Mi’s Wi-Fi equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). except when the "Lock Status" is set to "Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited.4. subscriber. the receiver is oriented towards the best server just as in the case of LOS. subscriber. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 803. or mobile Mi for N Ant – TX . Next. subscriber. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. The gain is read from the properties of the Wi-Fi equipment TX  ic  M M i i i assigned to the pixel.ini file. G Div – DL . Atoll takes the ratio of the signal level and the sum of the total interference from other cells and the downlink noise (as calculated in "Noise Calculation (DL)" on page 804). see the Administrator Manual. or mobile Mi: TX  ic  i CNR DL TX  ic  i = C DL TX  ic  i – n DL Bearer Determination: The bearers available for selection in the pixel. subscriber. subscriber. the bearers available for selection are all the bearers defined in the Wi-Fi equipment for which the following is true: Mi Mi TX i  ic  DL T B – G Div – DL – G Div  CNR DL The bearer selected for data transfer is the one with the highest index. Therefore. • Whose selection thresholds are less than the C/N at Mi: T B  CNR DL TX i  ic  Mi Mi If the cell’s frame configuration supports AMS.   DL The additional diversity gain defined for the clutter class of the pixel. N Ant – RX . or mobile Mi.Atoll 3. Interference from each cell is weighted according to the co. the C/N calculated above become: TX  ic  i TX  ic  i = CNR DL CNR DL M i DL + G Div – DL + G Div Mi Where G Div – DL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer. subscriber. For more information.3. and the traffic loads of the interfering cells. if the corresponding option has been set in the Atoll.5 C/(I+N) and Bearer Calculation (DL) The carrier signal to interference and noise ratio is calculated in three steps. Atoll calculates the interference received at the same studied pixel. • Mobility  M i  : Mobility used for the calculations. Inter – Tech • I DL : Downlink inter-technology interference as calculated in "Interference Calculation (DL)" on page 804. • N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel. the STTD/MRC or SU-MIMO diversity gain. • B DL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. : AMS threshold defined for the cell TXi(ic).ini file.3. or M i mobile Mi. or mobile Mi: TX i  ic  CINR DL TX i  ic  = C DL    TXj  jc   TX i  ic  n DL    IDL    Inter – Tech Inter – Tech ------------------- ---------------------    +I + + NR DL 10 – 10  Log 10      10  DL 10   All TXj  jc             The Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel. The gain is read from the properties of the Wi-Fi equipment TX  ic  M M i i i assigned to the pixel. N Ant – RX . subscriber. DL The additional diversity gain defined for the clutter class of the pixel. BLER  B DL . Mobility  M i  . the bearers available for selection are all the bearers defined in the Wi-Fi equipment for which the following is true: 808 . : Inter-technology downlink noise rise. see the Administrator Manual. if the corresponding option has been set in the Atoll. Mi Mi or mobile Mi. Inter – Tech • NRDL • TX  ic  i T AMS • T B : Bearer selection thresholds of the bearers defined in the Wi-Fi equipment used by Mi’s terminal. subscriber. or mobile Mi. or mobile Mi covered by a cell TXi(ic) as explained in "Interference Calculation (DL)" on page 804.Atoll 3. subscriber.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks TX  jc  j • I DL ©Forsk 2015 : Interference from any cell TXj(jc) calculated for a pixel. or mobile Mi: TX  ic  TX  ic  i  I + N  DL i  TX j  jc    n DL  I DL   --------------------- Inter – Tech Inter – Tech 10   -----------------= 10  Log  + 10 + NR DL 10  + I DL   10    All TX j  jc          Bearer Determination: The bearers available for selection in the pixel. subscriber. or mobile Mi’s Wi-Fi equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). Calculations The downlink C/(I+N) for a cell TXi(ic) is calculated as follows for any pixel. subscriber. Therefore. subscriber. For more information. G Div – DL . or mobile Mi for N Ant – TX . M i B DL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. • Whose selection thresholds are less than the downlink C/(I+N) at Mi: T B  CINR DL Mi TX i  ic  M i If the cell’s frame configuration supports AMS. corresponding to the bearer is applied to its selection threshold. or mobile Mi G Div is also applied. subscriber. subscriber. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. • or mobile Mi. subscriber. TX i  ic  • N Ant – TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic). subscriber. • i BLER  BDL : Downlink block error rate read from the graphs available in the Wi-Fi equipment assigned to the terminal M used by the pixel. subscriber.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 M M i i TX  ic  i DL T B – G Div – DL – G Div  CINR DL The bearer selected for data transfer is the one with the highest index. subscriber. subscriber. L Ant is determined in the direction (H. or mobile Mi. i L Ant : Receiver terminal’s antenna attenuation calculated for the pixel. TX i TX i In coverage predictions. • L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model.0) from M i the antenna patterns of the antenna used by Mi. subscriber. subscriber. or mobile Mi in the downlink. TX i TX i = G Ant ). subscriber. or mobile Mi. subscriber. TX i  ic  M : Total noise from the interfering cells TXj(jc) at the pixel. • L • • M G i M M : Receiver terminal losses for the pixel.V) = (0.6 Signal Level Calculation (UL) Input Mi • P Max : Maximum transmission power of the terminal used by the pixel. or mobile Mi without power control. or mobile Mi. subscriber. MIMO – STTD/MRC and SU-MIMO Diversity Gains: Once the bearer is known. G • L • L Path : Path loss ( L Path = L Model + L Ant ). or mobile Mi after power control as Mi calculated in "C/(I+N) and Bearer Calculation (UL)" on page 813. while the antenna is pointed towards Mi’s best serving cell.3. • L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi. Mi For calculating the useful signal level from the best serving cell. Output TX i  ic  • CINR DL •  I + N  DL • : Downlink C/(I+N) from the cell TXi(ic) at the pixel.Atoll 3. or mobile Mi. • L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected. 11. shadowing margins are taken into account when the option "Shadowing taken into account" is selected. subscriber. Calculations The received traffic signal level (dBm) from a pixel. L Ant is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi. For calculating the interfering signal level from any interferer. the C/(I+N) calculated above become: TX i  ic  CINR DL TX i  ic  = CINR DL Mi DL + G Div – DL + G Div Mi Where G Div – DL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer.4. or mobile Mi at its serving cell TXi(ic) is calculated as follows: 809 . subscriber. or mobile Mi covered by a cell TXi(ic). • M Shadowing – Model : Shadowing margin based on the model standard deviation. • P Eff : Effective transmission power of the terminal used by the pixel. i B DL : Bearer assigned to the pixel. • Mi L Body : Body loss defined for the service used by the pixel.2. i : Receiver terminal’s antenna gain for the pixel. TX i : Transmitter antenna gain for the antenna used by the transmitter TXi ( G TX i • : Total transmitter losses for the transmitter TXi ( L TX i = L Total – UL ). or mobile Mi. • nf TX i  ic  TX i  ic  TX i  ic  : Noise figure of the cell TXi(ic). on the traffic loads of the interfering mobile Mj. the value of the thermal noise varies with the used bandwidth. Output • M i C UL : Received uplink signal level from the pixel. In coverage predictions. 11. 11. However. • N SCa – Used : Number of used subcarriers defined for the frame configuration of a cell TXi(ic). The interference received by a cell TXi(ic) from an interfering mobile covered by a cell TXj(jc) can be defined as the uplink signal level received from interfering mobiles Mj depending on the overlap that exists between the channels used by the cells TXi(ic) and TXj(jc).4. Calculation of the uplink noise rise which represents the total uplink interference from all the interfering mobiles as explained in "Noise Rise Calculation (UL)" on page 811. The uplink noise comprises thermal noise and the noise figure of the equipment.4. or mobile Mi at a cell TXi(ic). and is the P Mi Mi = P Eff after power control. TX i  ic  n UL TX i  ic  = n 0 – UL + nf TX i  ic  Output • TX  ic  i n UL : Uplink noise for the cell TXi(ic).2. the uplink noise rise values already available in simulation results or in the Cells table are used. • N SCa – Total : Total number of subcarriers defined for the frame configuration of a cell TXi(ic).7 Noise Calculation (UL) For determining the uplink C/N and C/(I+N). Atoll calculates the uplink noise over the channel bandwidth used by the cell. The thermal noise density depends on the temperature. i. . The used bandwidth depends on the number of used subcarriers.. Calculations The power spectral density of thermal noise is calculated as follows: n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz The thermal noise for a cell is calculated as:  N TXi  ic   TX i  ic  SCa – Used  n 0 – UL = n 0 + 10  Log  ------------------------ TXi  ic    N SCa – Total The uplink noise is the sum of the thermal noise and the noise figure of the cell TXi(ic). The calculation of uplink interference can be divided into two parts: • • 810 Calculation of the uplink interference from each individual interfering mobile as explained in "Interference Signal Levels Calculation (UL)" on page 811.Atoll 3.e.2.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks M M i i C UL = EIRP UL – L Path – M Shadowing – Model – L Indoor + G ©Forsk 2015 TX i –L TX i M i M i – L Ant – L Body Where EIRP is the effective isotropic radiated power of the terminal calculated as follows: M i EIRP UL = P With P Mi M i +G M i –L M i Mi = P Max without power control at the start of the calculations.3. T: Temperature in Kelvin. Input • • K: Boltzmann’s constant. subscriber. it remains constant for a given temperature.8 Interference Calculation (UL) The uplink interference is only calculated during Monte Carlo simulations. it is the ratio (I+N)/N. • Mj TL UL : Uplink traffic load of the interfering mobile Mj. • TX i  ic  – TX j  jc  rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co.2 Mj I UL : Uplink interference signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc). Calculations The uplink noise rise and total noise (I+N) for the cell TXi(ic) are calculated as follows: TX i  ic  NRUL TX i  ic    Mj n  UL - I UL  -------------------TX i  ic   Inter – Tech 10  --------  = 10  Log  – n UL 10 + 10  + NRUL   10     All M j     All TX  jc    j For any pixel.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 11.3. subscriber.4.8. or mobile Mi in the interfered cell TXi(ic). TX i  ic  • n UL • NR UL : Uplink noise for the cell TXi(ic) as calculated in "Noise Calculation (UL)" on page 810.2. Inter – Tech : Inter-technology uplink noise rise.Atoll 3.8. Input • M j I UL : Uplink interference signal levels received at a cell TXi(ic) from interfering mobiles Mj covered by other cells TXj(jc) as calculated in "Interference Signal Levels Calculation (UL)" on page 811. Calculations The uplink interference received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) is calculated as follows: Mj TX i  ic  – TX j  jc  Mj I UL = C UL + f O Mj + f TL – UL Calculations for the interference reduction factors due to channel overlapping and uplink traffic load are explained below: Interference reduction due to the co.2. Noise Rise Calculation (UL) The uplink noise rise is defined as the ratio of the total uplink interference received by any cell TXi(ic) from interfering mobiles Mj present in the coverage areas of other cells TXj(jc) to the uplink noise of the cell TXi(ic).1 Interference Signal Levels Calculation (UL) Input • M j C UL : Uplink signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) as calculated in "Signal Level Calculation (UL)" on page 809.and adjacent channel overlap between the studied and the interfering cells: Interference reduction due to the co. Atoll calculates the uplink total noise (I+N) as follows: 811 .and adjacent channel overlap between the cells TXi(ic) and TXj(jc) is calculated as follows: TX i  ic  – TX j  jc  fO TX i  ic  – TX j  jc  = 10  Log  r O    Interference reduction due to interfering mobile’s traffic load: The interference reduction factor due to the interfering mobile’s uplink traffic load is calculated as follows: M M j j f TL – UL = 10  Log  TL UL Output • 11. In other words.and Adjacent Channel Overlaps Calculation" on page 799. Traffic loads are calculated during Monte Carlo simulations as explained in "Scheduling and Radio Resource Allocation" on page 820.4. 2.ini file. TX i  ic  • N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic). • Mobility  M i  : Mobility used for the calculations. subscriber. subscriber.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks TX  ic  i  I + N  UL TX  ic  i = NR UL ©Forsk 2015 TX  ic  i + n UL Output TX  ic  i • NRUL •  I + N  UL : Uplink noise rise for the cell TXi(ic). or mobile Mi at its serving cell TXi(ic) is calculated as follows: M M i i TX  ic  i CNR UL = C UL – n UL Bearer Determination: The bearers available for selection in the cell TXi(ic)’s Wi-Fi equipment are the ones: 812 • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). or mobile Mi. if the corresponding option has been set in the Atoll. • i BLER  BUL : Uplink block error rate read from the graphs available in the Wi-Fi equipment assigned to the cell TXi(ic). subscriber. TX i  ic  : Total noise for a cell TXi(ic) calculated for any pixel. 11. • M PC : Power control margin defined in the global network settings. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. TX i  ic  • n UL : Uplink noise for the cell TXi(ic) as calculated in "Noise Calculation (UL)" on page 810. Mi N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel. or mobile Mi. • P Max : Maximum transmission power of the terminal used by the pixel.   M Calculations The uplink C/N from a pixel. or • mobile Mi. TX i  ic  • T AMS : AMS threshold defined for the cell TXi(ic). subscriber. Mi B UL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel. subscriber. see the Administrator Manual. subscriber. subscriber. • Whose selection thresholds are less than the uplink C/N at Mi: T B  CNR UL Mi Mi . • B UL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. • T B : Bearer selection thresholds of the bearers defined in the Wi-Fi equipment used bythe cell TXi(ic).9 C/N Calculation (UL) Input Mi C UL : Received uplink signal level from the pixel. or mobile Mi. subscriber.4. • P Min : Minimum transmission power of the terminal used by the pixel. • or mobile Mi.3. Mi Mi Mi Mi or mobile Mi. For more information. or mobile Mi at its serving cell TXi(ic) as calculated in • "Signal Level Calculation (UL)" on page 809.Atoll 3. 2. where T B UL TX i  ic  Mi B UL is the bearer selection threshold. Atoll calculates the uplink carrier to noise ratio as explained in "C/N Calculation (UL)" on page 812. Uplink Power Control: The pixel. subscriber. Next.4. subscriber. for the bearer selected for the pixel.3. The receiver terminal is always considered to be oriented towards its best server. or mobile Mi reduces its transmission power so that the uplink C/N from it at its cell is just enough to get the selected bearer. or mobile Mi G Div is also applied.   UL The additional diversity gain defined for the clutter class of the pixel. or mobile Mi as follows: Mi M i  TX i  ic   Mi    Mi  P Eff = Max  P Max –  CNR UL –  T M + M PC   P Min    B i   UL Mi Mi CNR UL is calculated again using P Eff . from the Wi-Fi equipment assigned to the cell TXi(ic). the bearers available for selection are all the bearers defined in the Wi-Fi equipment for which the following is true: M TX  ic  i i M UL i T B – G Div – UL – G Div  CNR UL The bearer selected for data transfer is the one with the highest index. BLER  B UL . or mobile at its serving cell using the effective power of the terminal used by the pixel. In the case of NLOS between the receiver and the best server. Finally. subscriber.Atoll 3. Atoll calculates the received signal level from each pixel. subscriber.10 C/(I+N) and Bearer Calculation (UL) The carrier signal to interference and noise ratio is calculated in three steps. First. Output • Mi CNR UL : Uplink C/N from a pixel. 11. determines the uplink C/(I+N) by dividing the previously calculated uplink C/N by the uplink noise rise value of the cell as calculated in "Noise Rise Calculation (UL)" on page 811. Therefore. MIMO – STTD/MRC and SU-MIMO Diversity Gains: Once the bearer is known. N Ant – TX .0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 TX  ic  i If the cell’s frame configuration supports AMS. The uplink noise rise can be set by the user manually for each cell or calculated using Monte Carlo simulations. • TX i  ic  NR UL : Uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 811. corresponding to the bearer is applied to its selection threshold. G Div – UL . subscriber. subscriber. the STTD/MRC or SU-MIMO diversity gain. Mobility  M i  . 813 . or mobile as explained in "Signal Level Calculation (UL)" on page 809. the receiver is oriented towards the best server just as in the case of LOS. or mobile Mi. subscriber. the C/N calculated above become: Mi Mi TX i  ic  UL CNR UL = CNR UL + G Div – UL + G Div TX i  ic  Where G Div – UL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer. The transmission power of Mi is reduced to determine the effective transmission power from the pixel. or mobile Mi at it serving cell TXi(ic). or mobile Mi at it serving cell TXi(ic) as calculated in "C/N Calculation (UL)" on page 812. Atoll does not try to find the direction of the strongest signal. except when the "Lock Status" is set to "Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited. subscriber. Input • Mi CNR UL : Uplink C/N from a pixel. If with P Mi Mi Mi = P Max AND CNR UL  T TX i  ic  Mi + M PC . The gain is read from the properties of the Wi-Fi equipment TX  ic  M M i i i assigned to the cell TXi(ic) for N Ant – RX . • Whose selection thresholds are less than the uplink C/(I+N) at Mi: T B  CINR UL M M i i TX i  ic  If the cell’s frame configuration supports AMS. G Div – UL . The gain is read from the properties of the Wi-Fi equipment M TX  ic  M i i i assigned to the cell TXi(ic) for N Ant – RX . • P Max : Maximum transmission power of the terminal used by the pixel.   UL The additional diversity gain defined for the clutter class of the pixel. or mobile Mi G Div is also applied. Mi N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel. subscriber. subscriber. or mobile Mi. M M i i Mi Mi or mobile Mi. MIMO – STTD/MRC and SU-MIMO Diversity Gains: Once the bearer is known. corresponding to the bearer is applied to its selection threshold. • Mobility  M i  : Mobility used for the calculations. subscriber. • or mobile Mi. the C/(I+N) calculated above become: Mi Mi TX i  ic  UL CINR UL = CINR UL + G Div – UL + G Div TX i  ic  Where G Div – UL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer. subscriber. if the corresponding option has been set in the Atoll. subscriber. Therefore. TX i  ic  • N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic). • T B : Bearer selection thresholds of the bearers defined in the Wi-Fi equipment used bythe cell TXi(ic). BLER  BUL . Mobility  M i  . or mobile Mi at a cell TXi(ic) is calculated as follows: Mi TX i  ic  Mi CINR UL = CNR UL – NR UL Bearer Determination: The bearers available for selection in the cell TXi(ic)’s Wi-Fi equipment are the ones: • Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are defined in both equipment). • M PC : Power control margin defined in the global network settings.Atoll 3.ini file.   M Calculations The uplink C/(I+N) for any pixel. the bearers available for selection are all the bearers defined in the Wi-Fi equipment for which the following is true: Mi TX i  ic  UL Mi T B – G Div – UL – G Div  CINR UL The bearer selected for data transfer is the one with the highest index. • B UL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel. • Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service being accessed by Mi. Uplink Power Control: 814 . For more information. or mobile Mi. subscriber. N Ant – TX . • i BLER  BUL : Uplink block error rate read from the graphs available in the Wi-Fi equipment assigned to the cell TXi(ic). Mi B UL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 TX  ic  i • T AMS : AMS threshold defined for the cell TXi(ic). • P Min : Minimum transmission power of the terminal used by the pixel. or • mobile Mi. subscriber. the STTD/MRC or SU-MIMO diversity gain. see the Administrator Manual.3. Mi M i 11. or mobile Mi at it serving cell TXi(ic). • P Eff : Effective transmission power of the terminal used by the pixel.3 Best Server Determination In Wi-Fi. The best server is determined as follows: BSM = TX i  ic  i TX  ic   TX i  ic   i = Best C DL C  All TX i  ic   DL  Here ic is the cell of the transmitter TXi with the highest power. BS M . subscriber. or mobile Mi reduces its transmission power so that the uplink C/(I+N) from it at its cell is just enough to get the selected bearer. subscriber. subscriber. subscriber. L Ant . subscriber. subscriber. subscriber. or mobile Mi is said to be within the service area of its best serving cell TXi(ic) if the downlink C/N from the cell at the pixel. where T TX  ic  i M i B UL is the bearer selection threshold. However. or mobile Mi can be covered by a cell (as calculated in "Best Server Determination" on page 815) but can be outside the service area.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 The pixel. a pixel. or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 803 using the terminal and service parameters ( L Mi . subscriber. or mobile Mi gets the highest signal level. Output • BS M : Best serving cell of the pixel. or mobile Mi. or mobile Mi in the uplink. or mobile Mi.4. A pixel. subscriber. or mobile is greater than or equal to the minimum C/N threshold defined for the cell. The transmission power of Mi is reduced to determine the effective transmission power from the pixel. is the cell from which the received downlink signal level is the i highest among all the cells. Input • TX i  ic  C DL : Downlink signal level received from any cell TXi(ic) at a pixel. In coverage prediction calculations and in calculations on subsriber lists.3.4 Service Area Calculation In Wi-Fi.G Mi Mi Mi . If with P M i M i M i = P Max AND CINR UL  T TX  ic  i M i B UL + M PC . subscriber. subscriber. In Monte Carlo simulations. subscriber. or mobile Mi. i 11. best server refers to a cell ("serving transmitter"-"reference cell" pair) from which a pixel. This calculation also determines whether the pixel. or mobile. • B UL : Bearer assigned to the pixel. Output Mi • CINR UL : Uplink C/(I+N) from a pixel. a random cell is selected as the serving (reference) cell. the final reference cell ic might be different from the initial cell ic (the one with the highest power). subscriber. subscriber. or mobile Mi as follows: Mi M i  TX i  ic   Mi    Mi  P Eff = Max  P Max –  CINR UL –  T M + M PC   P Min i    B   UL Mi Mi CINR UL is calculated again using P Eff . 815 . for the bearer selected for the pixel. Calculations The best server of any pixel. from the Wi-Fi equipment assigned to the cell TXi(ic). and L Body ) of Mi. or mobile Mi is within the coverage area of any transmitter or not. if more than one cell of the same transmitter covers the pixel.Atoll 3. the cell of the highest priority layer is selected as the serving (reference) cell.4. subscriber. or mobile Mi. TX i  ic  D CP TX i  ic  r CP = -------------F Adding the cyclic prefix ratio to the useful symbol duration. • TX  ic  i T Min : Min C/N threshold defined for the cell TXi(ic). False: Otherwise. subscriber. TX i  ic  TX i  ic  TX i  ic  r CP Calculations Atoll determines the inter-subcarrier spacing. • • Calculation of uplink and downlink total resources in a cell as explained in "Calculation of Total Cell Resources" on page 816. Cell Capacity. • N SCa – Data : Number of data subcarriers defined for the frame configuration of a cell TXi(ic).4. otherwise.1 Calculation of Total Cell Resources The total amount of resources in a cell is the number of modulation symbols that can be used for data transfer per second. TX i  ic  1 D Sym – Useful = ------------------TX i  ic  F And. the duration of the cyclic prefix (guard interval). subscriber. TX i  ic  TX i  ic  TX i  ic  D Symbol = D Sym – Useful + D CP The total number of modulation symbols in the downlink and uplink are: 816 .5 Throughput Calculation Throughputs are calculated in two steps. Calculation of throughputs as explained in "Channel Throughput. in the global network settings.Atoll 3.5. • N SCa – Total : Total number of subcarriers defined for the frame configuration of a cell TXi(ic). Input TX  ic  i • W Channel : Channel bandwidth of the cell TXi(ic).0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 Input • TX  ic  i CNR DL : Downlink C/N from the cell TXi(ic) at a pixel.3. 11. Atoll determines the total symbol duration.4. or mobile Mi is within the service area of its best serving cell TXi(ic) if: TX  ic  i CNR DL TX  ic  i  T Min Output • • True: If the calculation criterion is satisfied. Calculations A pixel. • : Cyclic prefix ratio defined for the cell’s frame configuration of TXi(ic) or. 11. and Per-user Throughput Calculation" on page 817. F TX i  ic  TX i  ic  6 W Channel  10 = ----------------------------------TX i  ic  N SCa – Total Atoll calculates the useful symbol duration. or mobile Mi as calculated in "C/N Calculation (DL)" on page 806. CINR UL graph available in the Wi-Fi equipment assigned to the cell TXi(ic). Input TX i  ic  • TL DL – Max : Maximum downlink traffic load for the cell TXi(ic). • i i BLER  B DL : Downlink block error rate read from the BLER vs. 11.3. downlink or uplink. Cell Capacity. defined for the cell. and Per-user Throughput Calculation Channel throughputs are calculated for the entire channel resources allocated to the pixel. subscriber. • M M i i BLER  B UL : Uplink block error rate read from the BLER vs. • N Users – UL : Number of users connected to the cell TXi(ic) in uplink. CINR DL TX  ic  M graph available in the Wi-Fi equipment assigned to the terminal used by the pixel.4. Cell capacities are similar to channel throughputs but upper-bound by the maximum downlink and uplink traffic loads. • Mi f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the pixel.5.  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel. subscriber. or mobile Mi in the downlink in : Amount of uplink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on • "C/(I+N) and Bearer Calculation (DL)" on page 807. subscriber. TX i  ic  TX i  ic  Calculations Downlink: • M i TX  ic  i Peak MAC Channel Throughput: CTP P – DL = R DL  M i B DL MIMO – SU-MIMO Gain: 817 . • N Users – DL : Number of users connected to the cell TXi(ic) in downlink.Atoll 3. Mi • TP Offset : Throughput offset defined in the properties of the service used by the pixel. subscriber. or mobile Mi in the uplink in "C/ i B DL B i UL (I+N) and Bearer Calculation (UL)" on page 813. subscriber. TX i  ic  • R UL • page 816. Per-user throughputs are calculated by dividing the cell capacities by the average number of connects users. or mobile Mi.2 Channel Throughput. • TL UL – Max : Maximum uplink traffic load for the cell TXi(ic). or mobile Mi. subscriber.  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel. • R DL TX  ic  i TX i  ic  : Amount of downlink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on page 816. or mobile Mi.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 TX  ic  i R DL TX  ic  i = R UL TX  ic   1  -  N SCai – Data = Floor  ----------------TX  ic  D i  Symbol Output • TX i  ic  TX i  ic  R DL and R UL : Amount of downlink and uplink resources in the cell TXi(ic). or mobile Mi. TX i  ic  • T AMS : AMS threshold defined for the cell TXi(ic). • TX i  ic  CNR DL : Downlink C/N the cell TXi(ic) as calculated in "C/N Calculation (DL)" on page 806. The gain is read from the properties of the Wi-Fi equipment assigned to the cell TXi(ic) for: • Mi N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel. • N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel. or mobile Mi is located.3. M i or mobile Mi.– TP Offset  -----------------------100 Mi Cap P – DL = ----------------------TX i  ic  N Users – DL Mi • • Mi Cap E – DL Effective MAC Throughput per User: PUTP E – DL = ----------------------TX i  ic  N Users – DL Mi Application Throughput per User: PUTP A – DL Mi Mi f TP – Scaling . subscriber. • M i BLER  B DL : Downlink block error rate read from the graphs available in the Wi-Fi equipment assigned to the   TX i  ic  terminal used by the pixel. SU-MIMO gain G SU – MIMO is applied to the bearer efficiency. • B DL : Bearer assigned to the pixel.– TP Offset  -----------------------100 • Application Channel Throughput: • Peak MAC Cell Capacity: Cap P – DL = CTP P – DL  TL DL – Max • i i i Effective MAC Cell Capacity: Cap E – DL = Cap P – DL   1 – BLER  B DL     • • Mi TX i  ic  Mi M Application Cell Capacity: M i Cap A – DL M Mi M i Cap E – DL = Mi Peak MAC Throughput per User: PUTP P – DL M M f TP – Scaling i . subscriber. • 818 TX i  ic  N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic). it is interpolated from the gain values available for the C/(I+N) just less than and just greater than the actual C/(I+N). or mobile Mi. The gain is read from the properties of the Wi-Fi equipment assigned to the pixel.– TP Offset = PUTP E – DL  -----------------------100 Mi Uplink: • M i TX  ic  i Peak MAC Channel Throughput: CTP P – UL = R UL  M i B UL MIMO – SU-MIMO Gain: Max If the frame configuration supports AMS. or mobile Mi in the downlink as explained in "C/(I+N) and Bearer Mi Calculation (DL)" on page 807. subscriber. subscriber. subscriber. BLER is determined for CINR DL . or mobile Mi for: TX  ic  i • N Ant – TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic). • Mobility  M i  : Mobility used for the calculations.Atoll 3.  Mi B DL TX i  ic  Max =  Mi   1 + f SU – MIMO  G SU – MIMO – 1   if CNR DL TX i  ic   T AMS B DL If the Max SU-MIMO Gain for the exact value of the C/(I+N) is not available in the table. • M M M i i i Effective MAC Channel Throughput: CTP E – DL = CTP P – DL   1 – BLER  B DL     M i CTP A – DL = Mi M i CTP E – DL M f TP – Scaling i . . SU-MIMO gain G SU – MIMO is applied to the bearer efficiency. Atoll also takes into account the SU-MIMO Gain Factor f SU – MIMO defined for the clutter class where the pixel. subscriber. or mobile Mi.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 Max If the frame configuration supports AMS. • M i M i BLER  B UL : Uplink block error rate read from the graphs available in the Wi-Fi equipment assigned to the cell   Mi TXi(ic). or mobile Mi. subscriber. • PUTP E – DL : Downlink effective MAC throughput per user at the pixel. BLER is determined for CINR UL . • CTP E – DL : Downlink effective MAC channel throughput at the pixel. or mobile Mi in the uplink as explained in "C/(I+N) and Bearer Calculation (UL)" on page 813. Atoll also takes into account the SU-MIMO Gain Factor f SU – MIMO defined for the clutter class where the pixel. it is interpolated from the gain values available for the C/(I+N) just less than and just greater than the actual C/(I+N). • Cap A – DL : Downlink application cell capacity at the pixel. • CTP P – UL : Uplink peak MAC channel throughput at the pixel. subscriber. or mobile Mi. subscriber. or mobile Mi. subscriber. or mobile Mi.– TP Offset = PUTP E – UL  -----------------------100 Mi Output Mi • CTP P – DL : Downlink peak MAC channel throughput at the pixel. or mobile Mi is located. subscriber. • M M M i i i Effective MAC Channel Throughput: CTP E – UL = CTP P – UL   1 – BLER  B UL     Mi Mi f TP – Scaling .0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 • Mobility  M i  : Mobility used for the calculations. subscriber.– TP Offset = CTP E – UL  -----------------------100 Mi Mi • Application Channel Throughput: CTP A – UL • Peak MAC Cell Capacity: Cap P – UL = CTP P – UL  TL UL – Max • i i i Effective MAC Cell Capacity: Cap E – UL = Cap P – UL   1 – BLER  B UL     • Mi M Mi TX i  ic  Mi Application Cell Capacity: Cap A – UL M M Mi Mi f TP – Scaling .3. • Cap E – DL : Downlink effective MAC cell capacity at the pixel. or mobile Mi. subscriber. • CTP A – DL : Downlink application channel throughput at the pixel. • PUTP A – DL : Downlink application throughput per user at the pixel. or mobile Mi. or mobile Mi.Atoll 3. subscriber. • PUTP P – DL : Downlink peak MAC throughput per user at the pixel. or mobile Mi. Mi M i Mi Mi Mi Mi M i Mi Mi 819 . subscriber. • Cap P – DL : Downlink peak MAC cell capacity at the pixel. or mobile Mi. or mobile Mi.  Mi =  B UL TX i  ic  Max Mi   1 + f SU – MIMO  G SU – MIMO – 1   if CNR DL TX i  ic   T AMS B UL If the Max SU-MIMO Gain for the exact value of the C/(I+N) is not available in the table. subscriber. subscriber. subscriber.– TP Offset = Cap E – UL  -----------------------100 Mi Mi • • • Mi Cap P – UL Peak MAC Throughput per User: PUTP P – UL = ----------------------TX i  ic  N Users – UL Mi Mi Effective MAC Throughput per User: PUTP E – UL Mi Application Throughput per User: PUTP A – UL Cap E – UL = ----------------------TX i  ic  N Users – UL Mi Mi f TP – Scaling . • B UL : Bearer assigned to the pixel. 4. • Cap E – UL : Uplink effective MAC cell capacity at the pixel. • PUTP E – UL : Uplink effective MAC throughput per user at the pixel. Mi • CTP P – UL : Uplink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on page 816. subscriber. • • TX  ic  i TX i  ic  Mi : Priority of the service accessed by a mobile Mi. or mobile Mi. • PUTP P – UL : Uplink peak MAC throughput per user at the pixel.3. • TPD Max – UL : Uplink maximum throughput demand for the service accessed by a mobile Mi. • TPD Max – DL : Downlink maximum throughput demand for the service accessed by a mobile Mi. • CTP E – UL : Uplink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on Mi page 816. • TPD Min – UL : Uplink minimum throughput demand for the service accessed by a mobile Mi.6. subscriber. • PUTP A – UL : Uplink application throughput per user at the pixel.4. or mobile Mi. M graph available in the Wi-Fi equipment M i i BLER  BUL : Uplink block error rate read from the BLER vs. subscriber. or mobile Mi. • CTP P – DL : Downlink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on Mi Mi page 816. CINR DL   assigned to the terminal used by the mobile Mi. Mi • f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile Mi. or mobile Mi. • TL UL – Max : Maximum uplink traffic load for the cell TXi(ic). or mobile Mi.Atoll 3. • TP Offset : Throughput offset defined in the properties of the service used by the mobile Mi. • CTP A – UL : Uplink application channel throughput at the pixel. • Cap A – UL : Uplink application cell capacity at the pixel. subscriber.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks M ©Forsk 2015 i • CTP E – UL : Uplink effective MAC channel throughput at the pixel. subscriber. CINR UL graph available in the Wi-Fi equipment assigned   to the cell TXi(ic). subscriber. Mi Mi Mi Mi TX  ic  M i i BLER  BDL : Downlink block error rate read from the BLER vs. 11. • N Users – Max : Maximum number of users defined for the cell TXi(ic). or mobile Mi. or mobile Mi. subscriber.6 Scheduling and Radio Resource Management Wi-Fi scheduling and RRM algorithms are explained in "Scheduling and Radio Resource Allocation" on page 820 and the calculation of user throughputs is explained in "User Throughput Calculation" on page 823. • Cap P – UL : Uplink peak MAC cell capacity at the pixel. or mobile Mi.1 Scheduling and Radio Resource Allocation Input TX i  ic  • TL DL – Max : Maximum downlink traffic load for the cell TXi(ic). • Mi CTP E – DL : Downlink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on page 816. M M i i Mi Mi Mi Mi Mi 11. • p • TPD Min – DL : Downlink minimum throughput demand for the service accessed by a mobile Mi. subscriber. 820 . the scheduler keeps all the mobiles generated for the cell TXi(ic). starting from the mobile with the lowest priority service.e.and R Min – UL = -------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL 3. i. Atoll goes to the next step for allocating resources to satisfy the maximum throughput demands. respectively. Atoll sorts the M i Sel 2.. and all the minimum throughput resources demanded by Sel Mi the mobiles have been allocated. are allocated to other mobiles. If  Sel Mi TX i  ic   R Min – DL  TL DL – Max or Sel Mi Sel Mi TX i  ic  R Min – UL  TL UL – Max .e. mobiles M i TX  ic  i  N Users are selected for RRM by the scheduler. 5. Atoll allocates the downlink and uplink resources required to satisfy each user’s minimum throughput demands in downlink and uplink as follows: Sel Mi Sel Mi R Min – DL Sel Mi Sel Mi TPD Min – DL TPD Min – UL = -------------------------. Atoll stops the resource allocation in downlink or uplink. that were allocated to it in the one direction in which it was able to get a throughput. Starting with M i TX i  ic   N Users in order of decreasing service priority.Atoll 3. among all the cells of the site in order to reach a downlink or uplink effective MAC aggregate site throughput ≤ the site’s maximum downlink or uplink backhaul throughput.3. a site’s downlink or uplink effective MAC aggregate throughput exceeds its maximum downlink or uplink backhaul throughput. the resources available in downlink have been used up for Sel Mi satisfying the minimum throughput demands of the mobiles. Mobile Selection: TX  ic  i The scheduler selects N Users mobiles for the scheduling and RRM process.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Calculations The following calculations are described for any cell TXi(ic) containing the users Mi for which it is the best server. the resources available in uplink have been used up for Sel Mi satisfying the minimum throughput demands of the mobiles. and the resources. If the Monte Carlo user distribution has generated TX i  ic  a number of users which is less than N Users – Max . mobiles are rejected one by one due to Backhaul Saturation. If an active DL+UL mobile is only able to get its minimum throughput demand in one direction. p Sel = 1 up to M i Sel Mi : = N . Resource Allocation for Minimum Throughput Demands: Sel 1. Backhaul Saturation: If at this stage. • When/If in downlink Sel Mi  TX i  ic  R Min – DL = TL DL – Max .. Resource Allocation for Maximum Throughput Demands: For each cell. i. TX  ic  TX  ic  TX  ic  i i i N Users = Min  N Users – Max N Users – Generated   Sel For a cell. Mobiles which are active DL+UL must be able to get their minimum throughput demands in both UL and DL in order to be considered connected DL+UL. it is rejected. • When/If in uplink  M Sel i TX  ic  i R Min – UL = TL UL – Max . the remaining cell resources available are: TX  ic  i TX  ic  i Downlink: R Rem – DL = TL DL – Max –  M Sel i R Min – DL Sel Mi 821 . 4. i..e. 6. • Sel Mi  When/If in downlink TX i  ic  R Max – DL = R Rem – DL . Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied until TX i  ic  TX i  ic  either R Rem – DL = 0 and R Rem – UL = 0 .0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks TX  ic  i ©Forsk 2015 TX  ic  i M  Uplink: R Rem – UL = TL UL – Max – M Sel i R Min – UL Sel i For each mobile. the resources available in downlink have been used up for Sel Mi satisfying the maximum throughput demands of the mobiles. Atoll stops the resource allocation in downlink or uplink. this user is removed from the list of remaining users. 5.Atoll 3. 4. Atoll recalculates the remaining resources as follows: TX i  ic  TX i  ic  R Rem – DL = TL DL – Max –  Sel Mi R Min – DL – Sel Mi TX i  ic  TX i  ic  R Rem – UL = TL UL – Max –   Sel Mi R Max – DL and Sel Mi Sel Mi R Min – UL – Sel Mi  Sel Mi R Max – UL Sel Mi 7. 3. • When/If in uplink Sel Mi  TX i  ic  R Max – UL = R Rem – UL . If the resources allocated to a user satisfy its maximum throughput demands. Atoll converts the remaining throughput demands of all the users to their respective remaining resource demands: Sel Mi Sel Mi RD Rem – DL Sel Mi Sel Mi TPD Rem – DL TPD Rem – UL = --------------------------.and -------------------N N 2. the resources available in uplink have been used up for satisfying Sel Mi the maximum throughput demands of the mobiles. Atoll divides the remaining resources in the cell into equal parts for each user: TX i  ic  TX i  ic  R Rem – DL R Rem – UL --------------------. the throughput demands remaining once the minimum throughput demands have been satisfied are the difference between the maximum and the minimum throughput demands: Sel Mi Sel Mi Sel Mi Downlink: TPD Rem – DL = TPD Max – DL – TPD Min – DL M Sel i M Sel i M Sel i Uplink: TPD Rem – UL = TPD Max – UL – TPD Min – UL Sel Let the total number of users with remaining throughput demands greater than 0 be N  M i . i. or all the maximum throughput demands are satisfied. whichever is smaller.3. 1. The resources allocated to each user for satisfying its maximum throughput demands are: Sel i R Max – DL M TX i  ic  Sel Sel Sel TX i  ic  M  Mi  Mi R Rem – DL R Rem – UL i = Min  RD Rem – DL --------------------- and R Max – UL = Min  RD Rem – UL --------------------- N N     Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from the remaining resources of the cell.and RD Rem – UL = --------------------------Sel Mi Sel Mi CTP P – DL CTP P – UL Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the peak channel throughputs at the user’s location. Backhaul Capacity Limitation: 822 .e.. 3. • Sel Mi Sel CTP P – DL : Downlink peak MAC channel throughput at the mobile M i as calculated in "Throughput Calculation" on page 816.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 Backhaul overflow ratios are calculated for each site as follows: Sel Sel   M  Mi i     R Max – DL  CTP E – DL     Sel   M  Site i - and = Max  1 ----------------------------------------------------------------------------------------------------- Sel Sel  Mi    Mi  Site  R Min – DL  CTP E – DL   TP BH – DL –    Sel   M i  Site  Site BHOFDL  Sel Sel   Mi   Mi    R Max – UL  CTP E – UL     Sel   M i  Site = Max  1 -------------------------------------------------------------------------------------------------------  Sel Sel  Mi    Mi  Site TP – R  CTP  Min – UL  BH – UL E – UL     Sel   M i  Site  Site BHOFUL  Total Amount of Resources Assigned to Each Selected Mobile: Sel Atoll calculates the amounts of downlink and uplink resources allocated to each individual mobile M i (which can also be referred to as the traffic loads of the mobiles) as follows: Sel Sel Mi Downlink: TL DL Sel Mi = R DL Sel Mi Sel Mi R  Mi   Mi Max – DL  CTP P – DL  R Min – DL  CTP P – DL +  -----------------------------------------------Site     BHOF DL  = -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – DL Sel Sel Mi Uplink: TL UL Sel Mi = R UL Sel Mi Sel Mi R  M  Mi i  Max – UL  CTP P – UL  R Min – UL  CTP P – UL +  -----------------------------------------------Site     BHOF UL  = -----------------------------------------------------------------------------------------------------------------Sel Sel Mi CTP P – UL Output Sel Mi • TL DL • TL UL Sel Mi Sel Mi = R DL Sel : Downlink traffic load or the amount of downlink resources allocated to the mobile M i . M i . 823 . 11.2 User Throughput Calculation User throughputs are calculated for the percentage of resources allocated to each mobile selected by the scheduling for RRM Sel during the Monte Carlo simulations. • Sel Mi Sel R UL : Amount of uplink resources allocated to the mobile M i as calculated in "Scheduling and Radio Resource Allocation" on page 820. Sel Mi Sel = R UL : Uplink traffic load or the amount of uplink resources allocated to the mobile M i . Input • Sel Mi R DL Sel : Amount of downlink resources allocated to the mobile M i as calculated in "Scheduling and Radio Resource Allocation" on page 820.Atoll 3.6.4. • Sel Mi Sel CTP P – UL : Uplink peak MAC channel throughput at the mobile M i as calculated in "Throughput Calculation" on page 816. 5 Automatic Planning Algorithms 824 Sel • "Automatic Neighbour Planning" on page 825. subscriber. or mobile M i . • UTP P – UL : Uplink peak MAC user throughput at the pixel.Atoll 3. CINR Traffic graph available in the Wi-Fi equipment   Sel assigned to the terminal used by the mobile M i . or mobile M i . Sel Mi Sel Sel Mi Sel • UTP A – DL : Downlink application user throughput at the pixel. CINR UL graph available in the Wi-Fi equipment assigned   to the cell TXi(ic). . "Automatic Frequency Planning Using the AFP" on page 831.– TP Offset = UTP E – DL  -----------------------100 Uplink: Sel Mi Sel Mi Sel Mi  CTP P – UL • Peak MAC User Throughput: UTP P – UL = R UL • M M   Mi   i i Effective MAC User Throughput: UTP E – UL = UTP P – UL   1 – BLER  B UL      Sel • Sel Mi Application User Throughput: UTP A – UL Sel Sel Mi Sel Sel Mi Sel Mi f TP – Scaling . Sel Mi 11. . subscriber. Sel Mi Sel • f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile M i • TP Offset : Throughput offset defined in the properties of the service used by the mobile M i Sel Mi Sel Calculations Downlink: Sel Mi Sel Mi Sel Mi  CTP P – DL • Peak MAC User Throughput: UTP P – DL = R DL • Mi Mi   Mi   Effective MAC User Throughput: UTP E – DL = UTP P – DL   1 – BLER  B DL      Sel • Sel Mi Application User Throughput: UTP A – DL Sel Sel Sel Mi Sel Mi Sel Mi f TP – Scaling . "Automatic Inter-technology Neighbour Planning" on page 829.3. subscriber. subscriber. subscriber. Sel Sel . or mobile M i • UTP E – UL : Uplink effective MAC user throughput at the pixel. or mobile M i • UTP E – DL : Downlink effective MAC user throughput at the pixel. . or mobile M i Sel Mi Sel Mi The following sections describe the algorithms for: • • • Sel . Sel • M  Mi  i BLER  BUL  : Uplink block error rate read from the BLER vs.– TP Offset = UTP E – UL  -----------------------100 Output Sel Mi UTP P – DL : Downlink peak MAC user throughput at the pixel. or mobile M i .0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 Sel • TX  ic   Mi  i BLER  BDL  : Downlink block error rate read from the BLER vs. subscriber. • UTP A – UL : Uplink application user throughput at the pixel. . D is stated in m.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 11. Atoll adds all the cells located on the same site as the reference cell to the candidate neighbour list.5. Figure 11.3. • • Force Co-site Cells as Neighbours: If selected. Atoll takes into account the computation zone.3% so that the maximum variation in D does not to exceed 1%. When automatic planning starts. 825 . They are located inside the focus zone. They belong to the folder on which the allocation has been executed. Atoll checks the following conditions: 1. If the distance between the reference cell and the candidate neighbour is greater than this value. They satisfy the filter criteria applied to the Transmitters folder. 2. Force Adjacent Cells as Neighbours: If selected. Atoll adds all the cells geographically adjacent to the reference cell to the candidate neighbour list.3: Inter-Transmitter Distance Calculation The formula above implies that two cells facing each other have a smaller effective distance than the actual distance. The weight of this constraint can be defined. The weight of this constraint can be defined. It is used to calculate the rank of each neighbour. and its importance. Atoll calculates the effective distance between the reference cell and its candidate neighbour from the real distance between them and the azimuths of their antennas: Dist  CellA CellB  = D   1 + x  cos  – x  cos   Where x = 0. The calculation options. The distance between both cells must be less than the user-definable maximum inter-site distance. It means that the cells of all the TBC transmitters of your ATL document are potential neighbours. If no focus zone exists in the ATL document.1 Automatic Neighbour Planning The intra-technology neighbour planning algorithm takes into account the cells of all the TBC transmitters. This can be the Transmitters folder or a group of transmitters (subfolder). Only TBA cells are assigned neighbours. The cells to be allocated will be called TBA cells. then the candidate neighbour is discarded. We assume a reference cell TXi(ic) and a candidate neighbour cell TXj(jc). Candidate neighbours are ranked in the order of increasing effective distance from the reference cell.Atoll 3. It is used to calculate the rank of each neighbour. They must fulfil the following conditions: • • • • They are active. and its importance. symmetry cannot be respected. if neighbourhood relationship is forced in one direction and forbidden in the other. • Force Symmetry: If selected. The neighbour list of TXj(jc) is not full. If the neighbours list of a cell is full. only the distance criterion is taken into account.3. In this case. by adding the ) to them.Atoll 3. The ranking of adjacent neighbour cells increases with the number of such pixels. Adjacent cells are sorted in the order of decreasing rank. Atoll deletes all the current neighbours and carries out a new neighbour allocation. respectively. 826 . TX  ic  i S TX  ic  is the surface area covered by TXi(ic) within C DL i TX  ic  i + HO Start and C DL + HO End . Atoll adds the reference cell to the candidate neighbour list of the its candidate neighbour. On the other hand.ini file: [Neighbours] DoNotDeleteSymmetrics = 1 • Force Exceptional Pairs: This option enables you to force/forbid some neighbour relations. If not selected. • Delete Existing Neighbours: If selected. Atoll will add TXi(ic) to the end of the list. the existing neighbours are kept in the list. HOStart is the margin with respect to the best signal level at which the handover starts. so it will also remove TXj(jc) from the neighbour list of TXi(ic). A candidate neighbour cell TXi(ic) is considered adjacent to the reference cell TXi(ic) if there exists at least one pixel of TXj(jc)’s best server coverage area where TXi(ic) is the second best server. ii. A symmetric neighbour relation is allowed only if the neighbour list of the reference cell is not already full. Atoll displays a warning in the Event viewer.4: Determination of Adjacent Cells Determination of Adjacent Cells: Geographically adjacent cells are determined on the basis of their best server coverage areas.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 Figure 11. and HO End is the margin with respect to the best signal level at which the handover ends. Atoll considers the constraints between exceptional pairs in both directions so as to respect symmetry condition. You can force Atoll to keep that cell in the reference cell’s neighbours list by adding the following option in the Atoll. If TXj(jc) is a neighbour of TXi(ic) but TXi(ic) is not a neighbour of TXj(jc). i j The overlapping zone ( S TX  ic   S TX  jc  ) is defined as follows: i • j Here S TX  ic  is the surface area covered by the cell TXi(ic) that comprises all the pixels where: i • The received signal level is greater than or equal to the signal level threshold. If the Use Coverage Conditions check box is selected. there can be two possibilities: i. 3. the reference cell will not be added as a neighbour of that cell and that cell will be removed from the reference cell’s neighbours list. the coverage areas of TXi(ic) and TXj(jc) must have an overlap ( S TX  ic   S TX  jc  ).Otherwise. The received signal level TX i  ic  ( C DL TX i  ic  ) and the signal level threshold are calculated from CNR DL TX i  ic  value of the noise ( n DL • TX i  ic  and T Min . The neighbour list of TXj(jc) is full. Atoll will not be able to add TXi(ic) to the list. If you select "Force exceptional pairs" and "Force symmetry". Exceptional pairs are pairs of cells which will always or never be neighbours of each other. For calculating the overlapping coverage areas. Atoll uses the higher of the two values. 100 ). If the maximum number of neighbours to be allocated to each cell is exceeded. the terminal that has the highest difference between gain and losses. The service and terminal are selected such that the selection gives the largest possible C/N coverage areas for the cells. j • • • TX  ic  i If a global value of the C/N threshold ( T Min ) is set in the coverage conditions dialogue. table below). for each cell. Atoll calculates the percentage of the coverage area overlap ( -------------------------------------S TX  ic  i and compares this value with the % Min Covered Area. and the shadowing margin calculated using the defined cell edge coverage probability. Atoll sorts the neighbours by decreasing importance in order to keep the ones with high importance. global value and the value defined for that cell. S TX  ic  i Next.3. Atoll uses the service with the lowest body loss.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 • S TX  jc  is the coverage area where the candidate cell TXj(jc) is the best server. Atoll keeps the ones with high importance.Atoll 3. this value varies between 0 and 100%.. The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf. 100  % Min Coverage Area .5: Overlapping Zones S TX  ic   S TX  jc  i j TXj(jc) is considered a neighbour of TXi(ic) if --------------------------------------. Figure 11. Neighbourhood cause When Importance value Existing neighbour Only if the Delete Existing Neighbours option is not selected and in case of a new allocation Existing importance Exceptional pair Only if the Force Exceptional Pairs option is selected 100 % Co-site cell Only if the Force Co-site Cells as Neighbours option is selected Importance Function (IF) Adjacent cell Only if the Force Adjacent Cells as Neighbours option is selected Importance Function (IF) Neighbourhood relationship that fulfils coverage conditions Only if the % Min Covered Area is exceeded Importance Function (IF) Symmetric neighbourhood relationship Only if the Force Neighbour Symmetry option is selected Importance Function (IF) 827 .e. Atoll S TX  ic   S TX  jc  i j . if the option is selected. Atoll calculates the importance of the automatically allocated neighbours. i. When the above conditions are met. There can be a mix of the neighbourhood causes.3.If the Min and Max value ranges of the importance function factors overlap. d max is the maximum distance between the reference transmitter and a possible neighbour. adjacent neighbours. d Di  = 1 – ---------d max d is the effective distance (in m). The minimum and maximum importance assigned to each of the above factors can be defined. In the results. Cells whose channels have the same start frequency. Atoll lists only the cells for which it finds new neighbours. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 30% Adjacency factor (A) Min(A) 30% Max(A) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The Importance Function is evaluated as follows: Neighbourhood cause Importance Function Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%Max(Di))(O)}+Min(Di)+Delta(Di)(Di) 10%+20%{10%(Di)+90%(O)}+1%+9%(Di) No Yes Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 30%+30%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Yes Yes Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+ (100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) 60%+40%{10%(Di)+30%(O) +60%(A)}+1%+9%(Di) Co-site Adjacent No Where: Delta(X)=Max(X)-Min(X) • • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation. and neighbours allocated based on coverage overlapping. • • • The co-site factor (C): a Boolean. Otherwise. The overlapping factor (O): the percentage of overlapping. The default value of Min(O) = 1% ensures that neighbours selected for symmetry will have an importance greater than 0%. which takes into account the following factors: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. neighbours selected for symmetry will have an importance field greater than 0% only if there is some coverage overlapping. the neighbours may be ranked differently. neighbours will be ranked in this order: co-site neighbours. and the same total number of subcarriers are listed as intra-carrier neighbours.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 The importance is evaluated using an Importance Function (IF). There can be a mix of the neighbourhood causes. the neighbours may be ranked differently. It corresponds to the real inter-transmitter distance ( D in m) weighted by the azimuths of antennas. If the Min and Max value ranges of the importance function factors do not overlap. the same channel width. 828 .Atoll 3. If the Min and Max value ranges of the importance function factors overlap. neighbour cells are listed as inter-carrier neighbours. With a value of Min(O) = 0%. The adjacency factor (A): the percentage of adjacency. the neighbours will be ranked by neighbour cause. With the default values for minimum and maximum importance fields. The calculation options: 829 . If the distance is greater than this value. They must fulfil the following conditions: • • • • They are active.5. the candidate neighbour is discarded. Candidate neighbours are ranked in the order of increasing effective distance from the reference cell.Atoll 3. the neighbour importance calculated with respect to distance is based on the global Max inter-site distance setting for all neighbour candidates. This can be the Transmitters folder or a group of transmitters (subfolder). 2. They satisfy the filter criteria applied to the Transmitters folder.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 • By default. This means that all the TBC transmitters (GSM) or the cells of all the TBC transmitters (all other technologies) of the linked document are potential neighbours. the automatic neighbour allocation compares the defined Max intersite distance with the effective inter-cell distance.2 Automatic Inter-technology Neighbour Planning The inter-technology neighbour planning algorithm takes into account all the TBC transmitters (if the other technology is GSM) or the cells of all the TBC transmitters (for any other technology than GSM).ini: [Neighbours] RealInterSiteDistanceCondition=1 • By default.3. Figure 11. Only TBA cells are assigned neighbours. You can force Atoll to compare the Max inter-site distance with the real inter-site distance by adding the following lines in Atoll. D is stated in m. there can be cases where the real distance between assigned neighbours is higher than the Max inter-site distance. Atoll calculates the effective distance between the reference cell and its candidate neighbour from the real distance between them and the azimuths of their antennas: Dist  CellA CellB  = D   1 + x  cos  – x  cos   Where x = 0. When automatic planning starts.6: Inter-Transmitter Distance Calculation The formula above implies that two cells facing each other have a smaller effective distance than the actual distance. If no focus zone exists in the ATL document. Atoll takes into account the computation zone. The distance between reference cell and the candidate neighbour must be less than the user-definable maximum inter-site distance.ini: [Neighbours] CandidatesMaxDistanceInImportanceCalculation=1 11. As a consequence. We assume a reference cell A and a candidate neighbour B. The cells to be allocated in the main document will be called TBA cells. They are located inside the focus zone. They belong to the folder on which allocation has been executed. Atoll checks following conditions: 1. you can force Atoll to prioritise the individual distances between reference cells and their respective neighbour candidates by adding the following lines in Atoll. To avoid that.3% so that the maximum variation in D does not to exceed 1%. there can be cases where the calculated importance is different when the global Max inter-site distance is modified. because the effective distance is smaller. As a consequence. Force exceptional pairs: This option enables you to force/forbid some neighbour relations. Atoll calculates the importance of the automatically allocated neighbours. If not selected. The weight of this constraint can be defined. • 2nd case: The margin is other than 0 dB.Atoll 3. It is used to calculate the rank of each neighbour and its importance. • 2nd case: The margin is other than 0 dB. You may choose one or more carriers. 100  % Min Covered Area . Force co-site cells as neighbours: If selected. In this case. Delete existing neighbours: If selected. with a 0dB margin. If the maximum number of neighbours to be allocated to each cell is exceeded. CDMA. SA  SB Atoll calculates the percentage of the coverage area overlap ( -----------------. 3. This option enables you to select the CDMA carrier(s) that you want Atoll to consider as potential neighbours of Wi-Fi cells. The signal level received from B exceeds the minimum required. and is the highest one. Two cases may exist for SA: • 1st case: SA is the area where the cell A is the best serving cell. B is considered a neighbour of A if -----------------. • Algorithm based on coverage overlapping: The coverage areas of the reference cell A and the candidate neighbour B must overlap ( S A  S B ). or TDSCDMA network. SA is the area where: The signal level received from A exceeds the minimum required (calculated from the C/N threshold) and is within a margin from the highest signal level. Atoll deletes all the current neighbours and carries out a new neighbour allocation. The importance (%) of neighbours depends on the distance and on the reason of allocation: • 830 For allocation based on distance: Neighbour cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter/cell If the Force co-site cells as neighbours option is selected 100 % Neighbour relation that fulfils distance conditions If the maximum distance is not exceeded d1 – ---------d max . SA Candidate neighbours are ranked in the order of decreasing coverage area overlap percentages. SB is the area where: The signal level received from B exceeds the minimum required and is within a margin from the best signal level. Atoll sorts the neighbours by decreasing importance in order to keep the ones with high importance. Two cases may exist for SB: • 1st case: SB is the area where the candidate neighbour is the best server.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks • • • • ©Forsk 2015 CDMA Carriers: This option is available when a Wi-Fi network is being co-planned with a UMTS. Atoll adds all the transmitters/cells located on the same site as the reference cell in its candidate neighbour list. Neighbour relation criterion: • Allocation based on distance: The allocation algorithm is based on the effective distance between the reference cell and its candidate neighbour.3. Next. and is the highest one. This means that the signal received from A is greater than the minimum required (calculated from the C/N threshold). the existing neighbours are kept in the list. 100 ) and compares this value with the % SA SA  SB Min Covered Area. the margin must be set to 0 dB. Atoll keeps the ones with high importance. Atoll will allocate only the cells using the selected carriers as neighbours. Exceptional pairs are pairs of cells which will always or never be neighbours of each other. 3 Automatic Frequency Planning Using the AFP The role of an Automatic Frequency Planning (AFP) tool is to assign frequencies (channels) to cells of a network such that the overall network performance is optimised. Co-channel interference is the main reason for overall network quality degradation in Wi-Fi. the interference within the network is reduced as much as possible. the Wi-Fi AFP tries to minimise co.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 d is the effective distance between the reference cell and the neighbour and d max is the maximum inter-site distance. d Di  = 1 – ---------d max d is the effective distance (in m). • • The co-site factor (C): a Boolean.3. In order to improve network performance. Atoll displays only the cells for which it finds new neighbours. In the results. d max is the maximum distance between the reference transmitter and a possible neighbour. With the default values for minimum and maximum importance fields. There can be a mix of the neighbourhood causes. • For allocation based on coverage overlapping: Neighbour cause When Importance value Existing neighbour If the Delete existing neighbours option is not selected Existing importance Exceptional pair If the Force exceptional pairs option is selected 100 % Co-site transmitter/cell If the Force co-site cells as neighbours option is selected IF Neighbourhood relationship that fulfils coverage conditions If the % minimum covered area is exceeded IF The importance is evaluated using an Importance Function (IF). It corresponds to the real inter-transmitter distance weighted by the azimuths of antennas. The IF is user-definable using the Min importance and Max importance fields. The overlapping factor (O): the percentage of overlapping. If the Min and Max value ranges of the importance function factors do not overlap. which takes into account the following factors: • The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference transmitter. If the Min and Max value ranges of the importance function factors overlap.Atoll 3. 11. the neighbours will be ranked by neighbour cause.5. neighbours will be ranked in this order: co-site neighbours and neighbours allocated based on coverage overlapping. the neighbours may be ranked differently. In other words. Factor Min importance Default value Max importance Default value Distance factor (Di) Min(Di) 1% Max(Di) 10% Overlapping factor (O) Min(O) 10% Max(O) 60% Co-site factor (C) Min(C) 60% Max(C) 100% The IF evaluates importance as follows: Co-site Neighbourhood cause IF Resulting IF using the default values from the table above No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)} Yes Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+ Max(O)(O)/(Max(Di)+Max(O))} 60%+40%{1/7%(Di)+6/7%(O)} Where Delta(X)=Max(X)-Min(X) • • • Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance factor in the importance calculation.and adjacent channel interference as much as possible while respecting any 831 . They satisfy the filter criteria applied to the Transmitters folder.Atoll 3. Assigned weight  Dis tan ce = 0. and then the overall cost for the TBA cell. The absolute values of the constraint weights are calculated from the relative weights (%) defined in the Constraint Weights dialogue as follows: % Neighbour  Neighbour = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce % IM  IM = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce % Dis tan ce  Dis tan ce = ---------------------------------------------------------------------------------% Neighbour + % IM + % Dis tan ce 11. Related cells of a TBA cell are: • Its neighbours. The following describes the AFP’s automatic planning method for frequencies in Wi-Fi networks. Assigned weight  Neighbour = 0. The cells to be allocated will be called TBA cells. You can modify these weights in your Wi-Fi document. neighbours. Their channel allocation status is not set to locked.3. i.2 The sum of the weights assigned to the above relations is 1.and Adjacent Channel Overlaps Calculation" on .5.3.5.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks ©Forsk 2015 constraints input to it.e. the number of frequencies with which the AFP can work. and distance between transmitters. Atoll takes into account the computation zone. They are located inside the focus zone. i. Assigned weight  IM = 0.3 • Cells within the cell’s (or the default) minimum reuse distance.. if the check box "Existing neighbours" is selected. The best. if the check box "Reuse distance" is selected.1 Constraint and Relationship Weights The AFP is based on a cost function which takes into account channel separation constraints based on the channel overlap ratio as calculated in "Co. 11. interference matrices. If no focus zone exists in the ATL document.e. The AFP is based on a cost function which represents the interference level in the network. They must fulfil the following conditions: • • • • They are active. or optimum. 832 TX i  ic  – TX j  jc     Neighbour   Neighbour  TX i  ic  – TX j  jc  +  Dis tan ce   Dis tan ce TX  ic  – TX j  jc   +  i IM IM  is the channel overlap ratio as calculated in "Co.2 Cost Calculation The cost of the relation between the TBA cell and its related cell is calculated as follows: $ TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  = rO TX  ic  – TX  jc  i j Where r O page 799. The main constraints are the resources available for allocation. neighbour relations. and the relationships to take into account. frequency plan is the one which corresponds to the lowest cost. which takes into account interference matrices. The aim of the AFP is to minimise the cost. Channel separation is studied between each TBA cell and its related cells..and Adjacent Channel Overlaps Calculation" on page 799. Atoll calculates the cost between each individual TBA and related cell.5 • Cells that are listed in the interference matrix of the TBA cell. and distance between transmitters.3. The AFP takes into account the cells of all the TBC transmitters. the total cost of the current frequency plan for the entire network is simply the sum of the total TBA cell costs calculated above. Memorises the different frequency plans in order to determine the best one.5.  Neighbour is calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 825.and adjacent channel interference probabilities calculated as TX  ic  – TX  jc  i j explained in "Interference Matrix Calculation" on page 833.e. 11.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks AT330_TRR_E1 TX  ic  – TX  jc  i j TX  ic  – TX  jc  i j  Neighbour is the importance of the relationship between the TBA cell and its related neighbour cell.5. this value is equal to 1.1 Interference Matrix Calculation The co-channel interference probability is calculated as follows: S TX  ic  i TX  ic  – TX  jc  i j  IM – CC TX j  jc  TX i  ic    + M Quality n DL  C DL  -----------------------------------------------------------------------  TX i  ic  TX  ic  10 10  T i C DL – 10  Log  10 + 10  Min       = ------------------------------------------------------------------------------------------------------------------------------------------------S TX  ic  i The adjacent channel interference probability is calculated as follows: 833 .. TX i  ic  – TX j  jc   Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance between TX i  ic  – TX j  jc  them.  Dis tan ce is calculated as explained in "Distance Importance Calculation" on page 834. For manual neighbour planning.5..3 AFP Algorithm The AFP algorithm is an iterative algorithm which: • • • • Calculates the cost (as described above) of the initial frequency plan. $ Total =  TX i  ic  $ Total TX  ic  i 11. and proposes the last known best frequency plan as the solution.3.4 Appendices 11. the frequency plan which provides the lowest total cost. r CCO TX  ic  – TX  jc  i j and r ACO are the co.and adjacent channel overlap ratios as calculated in "Co. Stops when it is unable to improve the cost of the network.Atoll 3.and Adjacent Channel Overlaps Calculation" on page 799. The total cost of the current frequency plan for any TBA cell is given as follows.3. considering all the cells with which the TBA cell has relations: TX i  ic   QRF $ Total = 1 – TX i  ic  – TX j  jc  TX j  jc  And. Atoll calculates the quality reduction factor for the TBA cell and its related cell from the cost calculated above as follows: QRF TX i  ic  – TX j  jc  = 1–$ TX i  ic  – TX j  jc  The quality reduction factor is a measure of the cost of an individual relation. i.4.e. TX  ic  – TX  jc  i j  IM is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows: TX i  ic  – TX j  jc  TX i  ic  – TX j  jc   IM = r CCO TX i  ic  – TX j  jc  TX i  ic  – TX j  jc   IM – CC and  IM – CC TX i  ic  – TX j  jc    IM – CC TX i  ic  – TX j  jc  + r ACO TX i  ic  – TX j  jc    IM – AC are respectively the co. Tries different frequency plans in order to reduce the cost. i. either defined for each TBA cell individually or set for all the TBA cells in the AFP dialogue.7: Weighted Distance Between Cells The above formula implies that two cells facing each other will have a shorter effective distance between them than the real distance.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks S TX  ic  i TX  ic  – TX  jc  i j  IM – AC ©Forsk 2015 TX  jc  TX  ic  TX  ic  j i i   +M +f n  C DL  Quality ACS – FB DL ------------------------------------------------------------------------------------------------------  TX  ic  TX  ic  i 10 10  i  C + 10 – 10  Log  10 DL   T Min       = ----------------------------------------------------------------------------------------------------------------------------------------------------------------------S TX  ic  i For frequencies farther than the adjacent channel. and D follows: D D TX  ic  – TX  jc  i j TX i  ic  – TX j  jc  them. n DL TX j  jc  and C DL Condition TX i  ic   T Min as is the best server coverage area of the cell TXi(ic) where are the received downlink signal levels from the cells TXi(ic) and TXj(jc) the downlink noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 804. S TX  ic  i TX i  ic  the given condition is true. This figure shows that cells that are located near (based on the effective distance which is weighted by the orientations of the cells) have high importance. the interference probability is 0.5. which is 834 . The importance of the distance relation is explained in Figure 11. that comprises all the pixels where CNR DL i calculated in "Service Area Calculation" on page 815.7 on page 834. d = d TX i  ic  – TX j  jc  TX  ic  – TX  jc  i j is the weighted distance between the TBA cell TXi(ic) and its related cell TXj(jc) calculated as   1 + x   cos    – cos    – 2   is weighted according to the azimuths of the TBA cell and its related cell with respect to the straight line joining TX  ic  – TX  jc  i j is the distance between the two cells considering any offsets with respect to the site locations.4. and two cells pointing in opposite directions will have a greater effective distance. 11. M Quality is TX  ic  i the quality margin used for the interference matrices calculation. TX i  ic  Here S TX  ic  is the best server coverage area of the cell TXi(ic). x is set TX i  ic  – TX j  jc  to 10 % so that the maximum variation in D due to the azimuths does not exceed 40 %. C DL TX i  ic  respectively.2 Distance Importance Calculation TX i  ic  – TX j  jc  The distance importance between two cells (  Dis tan ce TX  ic  – TX  jc  i j  Dis tan ce   1     2 D Reuse =  Log   --------------------------------     D TXi  ic  – TXj  jc    --------------------------------------------------------2  Log  D Reuse   if D ) is calculated as follows: TX i  ic  – TX j  jc  1 Otherwise Where D Reuse is the minimum reuse distance.8 on page 835.  and  are calculated from the azimuths of the two cells as shown in Figure 11. and f ACS – FB is the adjacent channel suppression factor defined for the frequency band of the cell TXi(ic).Atoll 3. Figure 11.3. 3.AT330_TRR_E1 Atoll 3. Figure 11.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks interpreted as a high cost.8: Importance Based on Distance Relation 835 . and cells that are located far have low importance. Cells that are further than the reuse distance do not have any cost related to the distance relation. Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 11: Wi-Fi Networks 836 ©Forsk 2015 .3. Chapter 12 ACP Module This chapter covers the following topics: • "Objectives" on page 839 • "Quality Predictions and the Antenna Masking Method" on page 844 • "Configuration" on page 846 • "Multi-RAT and Co-planning Support" on page 848 • "Optimisation Methodology" on page 849 • "Load Balancing Objective" on page 855 • "EMF Exposure" on page 861 • "Shadowing Margin and Indoor Coverage" on page 865 • "Multi-Storey Optimisation" on page 865 • "ACP Software Data Flow" on page 868 . 0 Technical Reference Guidefor Radio Networks © Forsk 2015 838 .3.Atoll 3. the focus zone. a hot spot. the objective is said to be fulfilled if 90% of the pixels are covered by the objective rule. The objective is calculated only on the subset of pixels belonging to this zone. see "Optimisation Methodology" on page 849. CINR. CDMA2000. It uses an efficient global search algorithm to test many network configurations and propose the reconfigurations which best meet the objectives. An objective can combine several quality indicators from different technology layers. Both types of weighting can be used at the same time. overlap. Atoll ACP uses user-defined objectives to evaluate the quality and implementation cost of a network reconfiguration. The target zone can be either the computation zone. 12. LTE. A rule is a single quality indicator on a single technology layer fulfilling a defined threshold. Quality maps covering the computation zone are provided for the initial network (before reconfiguration) and final network (after reconfiguration). AND). The ACP can also select the best sites from a list of candidate sites. An objective can also be weighted according to traffic or weighted on a given zone.Atoll 3. UMTS.) on each pixel of the computation zone. ACP also supports 3GPP and 3GPP2 multi-RAT documents as well as co-planning.RSCP > -85dBm OR LTE 2010 . 12. a cost objective can be taken into account to reduce the expected implementation cost. or a zone defined as a group of clutter classes.1 Objectives Atoll ACP uses user-defined objectives to evaluate the quality and cost of the network reconfiguration or site selection. It calculates the basic quality indicators (RSCP. the pixel weights are taken from Atoll traffic maps. This is described by the following formula: Cov Obj =    i    1 Th1  Qual 1  i  OR 1 Th2  Qual 2  i    i  pixels of zone where.C/N > 20dB) The target for the objective defines the required percentage of pixels in the target zone (after applying any defined weight) which must fulfil the rule. When using traffic weighting. In addition. For example if the target is 90%. WiMAX.1.1 Definition and Evaluation The ACP calculates the quality objective using the user-defined resolution within the borders of the computation zone. Each quality indicator is technology-dependent. which are evaluated in a given zone and for a given traffic profile. and is consistent with the corresponding coverage predictions in Atoll.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12 ACP Module ACP optimises key network parameters in order to improve coverage and quality. etc. The ACP presents the changes ordered from the most to the least beneficial.1.1. EcIo. When using zone weighting. The defined weight enables you to assign a different importance to different pixels.1 Quality Objective Each quality objective is a logical combination of defined rules used to evaluate specific quality indicators. 12. and Wi-Fi.3. in which case the zone weight is taken as a supplementary factor to the traffic weight. An example of combined rules is: (UMTS 2100 . For more information on how weights are applied. the pixel weights are taken from a weight defined with each zone. An objective is defined by both a set of rules and a target. Each objective is measured on a defined target zone. Currently. 1 Th is the step function: 1 Th  x  = 1 if x  Th and 1 Th  x  = 0 if x  Th Qual k  i  is the basic quality measurement on pixel i   i  is the normalised weight for pixel i:  i = 1 i  pixels 839 . A pixel is said to be "covered" by the set of rules when it fulfils all the rules according to their logical relationship (OR. allowing phased implementation or implementation of just a subset of the suggested changes. ACP supports the following single-RAT radio access technologies: GSM. When this feature is supported for a quality parameter.Atoll 3.<Objective>.th.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module • • ©Forsk 2015 Thresholds on rules can be defined separately per zone or per clutter groups.ini file in order to modify maximum and minimum threshold values (see full option names and default values in the Atoll Administrator Manual): [ACPCore] <Technology>.ini file.<Objective>. • When the Progressive Threshold check box is selected. a progressive function is used instead with a weighting varying from 1 to 0. Progressive thresholds are proposed by default for some quality parameters.max • Threshold defined in the objective’s properties on the Objectives tab of the ACP setup. Main curve parameters can produce several shapes according to the user-defined values. The target threshold can be defined as absolute or relative compared to initial status: for example as 90% (absolute) or as increase of 5% over current coverage (relative). whether the objectives are met or not. The step function is described in "Definition and Evaluation" on page 839 and it is used by default in all objectives. the Progressive Threshold check box is enabled in the Thresholds Definition dialog box.1.1: Thresholds Definition dialog box • When the Progressive Threshold check box is cleared.1. You can disable it by setting the useProgressiveThreshold option to 0 in the [ACPCore] section of the ACP. leading to more intelligent decisions on improvements that may cause degradations elsewhere in the network.3.2 Progressive Thresholds Progressive thresholds allow the ACP to evaluate the amount of improvement or degradation of each objective. from a maximum to a minimum. a step function (1/0) is used and ACP objectives are evaluated on a "fixed-threshold basis". This value is used as the transition between two signal level ranges (below and above threshold). each range having its own hard coded modelling.min <Technology>. Below is a typical example with the signal level type objective (RSCP) and the default Min/Threshold/Max values (-120dBm/-90dBm/-60dBm). 12. Figure 12.2: Progressive Thresholds function 840 . Figure 12. on each pixel and in a logical manner. Below are the main curve parameters used by the progressive function: • The options below can be added in the ACP.th. RSSI (Total Noise Level Analysis (DL) + "Max Noise Level (dBm)") Overlap (Overlapping Zones (DL) + "Number of Servers") to measure pilot pollution as well as soft handover quality. This ensures that the measured values are the same in the ACP and in Atoll predictions.2. 12.1. but only on pixels of the target zone which are not filtered out.2 UMTS Quality Indicators The quality indicators used by ACP give the same maps and results as the reference prediction in Atoll. EcIo (Pilot Quality Analysis (DL) with "Ec/Io (dB)") to define the service area zone. CINR Co-channel (Coverage by C/I Level (DL) + "C/I Level (dB)") Overlap (Overlapping Zones (DL) + "Number of Servers") to define cell dominance and decrease the level of interference between cells while allowing a level of cell overlap.Atoll 3. For each quality indicator. In the following list. each quality indicator defined in UMTS is followed in italics by the reference prediction in Atoll (with the relevant "Field" setting). each quality indicator defined in CDMA2000 is followed in italics by the reference prediction in Atoll (with the relevant "Field" setting). When using a filter. For example. EcIo (Pilot Quality Analysis (DL) with "Ec/Io (dB)") to define the service area zone.1.1.2 Quality Indicators in the ACP Atoll ACP defines a set of basic quality indicators. formulas are similar to the ones in UMTS.2. In the following list. 841 . Best Server Distance 1st-2nd Difference 1st-Nth Difference 12. if any: • • • • • • • RSCP (Coverage by Signal Level (DL) + "Best Signal Level (dBm)") used as a measure of raw network coverage. These quality indicators are used when defining a rule to form complex objectives.1. Best Server Distance 1st-2nd Difference 1st-Nth Difference ACP handles CDMA2000 similarly to UMTS. each quality indicator defined in GSM is followed in italics by the reference prediction in Atoll (with the relevant "Field" setting).1. the target percentage coverage is not evaluated on all pixels of the target zone. 12.3.2.1. The main difference is that the formula for deriving signal level and Ec⁄Io differs between 1xRTT and 1xEv-DO: • In 1xRTT. Best Server Distance 1st-2nd Difference 1st-Nth Difference The ACP manages interference quality in the network by measuring signal pollution: a limited number of overlapping cells are allowed in order to allow for coverage continuity and handover capability. The number should be consistent with the frequency reuse ratio used for the network. if any: • • • • • • BCCH Signal Level (Coverage by Signal Level (DL) + "Best Signal Level (dBm)") used as a measure of raw network coverage. you could calculate an objective only on the pixels of a zone for which there is no coverage in a given technology: Target: 90% of pixel with UMTS RSCP > -95dB for which GSM Signal Level < -95dBm The 90% target will be applied only to the subset of pixels for which the GSM signal level is below -90dBm 12. Overlap (Overlapping Zones (DL) + "Number of Servers") to measure pilot pollution as well as soft handover quality. taking into account the pilot power as the basis for signal level computation.3 CDMA2000 Quality Indicators The quality indicators used by ACP give the same maps and results as the reference prediction in Atoll.1 GSM Quality Indicators The quality indicators used by ACP give the same maps and results as the reference prediction in Atoll. In the following list. Atoll ACP uses the same formulas as used elsewhere in Atoll.3 Target Filtering Atoll ACP allows you to filter pixels on which the target percentage will be evaluated according to defined filter conditions. if any: • • • • • • Signal Level (Coverage by Signal Level (DL) + "Best Signal Level (dBm)") used as a measure of raw network coverage.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12. RSRQ and RSSI. RS CINR (Coverage by C/(I+N) Level (DL) + "RS C/(I+N) Level (DL) (dB)") to measure and control interference. in which case it is recommended to perform one or several ACP -> AFP cycles. you can ensure that the quality indicator will be calculated the same as the reference prediction.1. RS C (Effective Signal Analysis (DL) + "RS Signal Level (DL) (dBm)") used as a measure of raw network coverage. In some cases this may lead to suboptimal reconfiguration. for example which service and terminal to use to define body loss and other losses (terminal antenna gain and loss).1. Preamble C/N (Effective Signal Analysis (DL) + "Preamble C/N Level (DL) (dB)") used as a measure of raw network coverage. as well as the Ec/Io computation. enabling comparison of the quality map with the Atoll coverage prediction. you can specify a reference prediction from among the predictions already calculated. RSRQ (Coverage by C/(I+N) Level (DL) + "RSRQ Level (DL) (dB)") to measure and control interference. RSRP (Effective Signal Analysis (DL) + "RSRP Level (DL) (dBm)") used as a measure of raw network coverage. In the following list. two methods are currently provided by the ACP: • • Using the current frequency plan: The existing frequency plan is taken into account when calculating RS CINR. you can consider shadowing in the calculation. In some cases this may lead to suboptimal reconfiguration. RS C/N (Effective Signal Analysis (DL) + "RS C/N Level (DL) (dB)") used as a measure of raw network coverage. Best Server Distance 1st-2nd Difference 1st-Nth Difference Because Preamble CINR depends strongly on the frequency plan. Ignoring the current frequency plan and ICIC: All the network cells are assumed to be on the same channel.2. Preamble CINR (Effective Signal Analysis (DL) + "Preamble C/(I+N) Level (DL) (dB)") to measure and control interference and signal quality. For most quality indicators.2. Preamble C (Effective Signal Analysis (DL) + "Preamble Signal Level (DL) (dBm)") used as a measure of raw network coverage. 12. Additionally. 12. the pilot is transmitted at full cell power. 12. each quality indicator defined in LTE is followed in italics by the reference prediction in Atoll (with the relevant "Field" setting). In the following list. in which case it is recommended to perform one or several ACP -> AFP cycles.6 Quality Indicator Parameters and Reference Maps The parameters that define how each quality indicator is calculated are under "Parameters" on the Objectives tab.2. if any: • • • • • • • • Signal Level (Coverage by Signal Level (DL) + "Best Signal Level (dBm)") used as a measure of raw network coverage.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module • ©Forsk 2015 In 1xEv-DO. RSRQ and RSSI depend strongly on the frequency plan. Overlap (Overlapping Zones (DL) + "Number of Servers") to better control cell dominance. Currently the frequency plan is not dynamically recalculated while changing network parameters. Overlap (Overlapping Zones (DL) + "Number of Servers") to better control cell dominance. By using a reference prediction. RSSI (Coverage by C/(I+N) Level (DL) + "RSSI Level (DL) (dBm)") to measure and control interference. each quality indicator defined in WiMAX is followed in italics by the reference prediction in Atoll (with the relevant "Field" setting). Ignoring the current frequency plan and segmentation: All the network cells are assumed to be on the same channel. if any: • • • • • • • • • • • Signal Level (Coverage by Signal Level (DL) + "Best Signal Level (dBm)") used as a measure of raw network coverage.5 WiMAX Quality Indicators The quality indicators used by ACP give the same maps and results as the reference prediction in Atoll.Atoll 3.3. two methods are currently provided by the ACP: • • Using the current frequency plan: The existing frequency and segmentation plan are taken into account when calculating the CINR.1. 842 . Currently the frequency plan and segmentation plans are not dynamically recalculated while changing network parameters. The cell max power is thus used as the basis of the signal level computation.4 LTE Quality Indicators The quality indicators used by ACP give the same maps and results as the reference prediction in Atoll. Best Server Distance 1st-2nd Difference 1st-Nth Difference Because RS CINR. When the resolution of the optimisation is different from the resolution of the path loss matrices. ACP’s "EcIo" prediction versus Atoll’s "Pilot Quality Analysis (DL)" prediction). At the scale of overall maps. the ACP and Atoll coverage predictions will match except in corner cases which are difficult to identify and manage.1. You can automatically create a set of site classes and their associated costs by defining some options in the [ACPGeneralPage] section of the ACP. • Setup Preferences > Calculation setting (on the Preferences tab of the ACP Properties dialog box): • In "High speed" mode. ACP reduces the number of cells it monitors for each pixel. see the Administrator Manual.g. For more information. 12. some of which may only create a bit of interference at first. it will inherit the site class of the existing site. Coverage predictions are therefore similar in most cases. The site class defined in the custom field in the Sites table will be assigned automatically to each site in the database when a new ACP setup is created. the path loss resolution.g.2. 843 .7 Advanced Objective Configuration By combining several rules to define one objective. etc). You can also assign different sets of costs for different site classes. and the raster resolution are identical. If a new candidate site is created in ACP and is co-located with an existing site. potential mismatches between ACP and Atoll core predictions may appear according the following parameter settings: • Resolution: • The best match between ACP and Atoll coverage predictions is obtained when the ACP resolution matches the path loss resolution. and later create significantly more interference after antenna parameters are changed during optimisation. Sites are assigned to a site class either manually or automatically.3 Atoll and ACP Prediction Matching ACP coverage predictions try as much as possible to match the Atoll coverage predictions (e. If it is not co-located with an existing site.Atoll 3. [ACPCustomFieldExtraction] site.3.costClass=SITECLASS # The name of the custom column in SITE table used to define the 'cost class'.costClass" option.2. Hence.ini file. Trade-off between quality and cost: The ACP will select the changes which have the most benefit for the least cost. Each site class can be assigned a different set of costs. You can assign them automatically by defining a custom field in the Sites table in Atoll and then defining the custom field in the ACP.ini using the "site. The "Automatic Candidate Positioning" functionality (New Candidate Setup dialog box > Action button) can be impacted in "High speed" mode. ACP performs a bilinear interpolation by using the four closest path loss values and interpolating. they still match pretty well despite small cosmetic mismatches in some very specific corner cases. the site class is set to Default and can be changed manually. they will be at pixel level and are negligible (e.1. Generally speaking. in spite of the variety of potentially conflictuous conditions such as varying resolutions. When there are differences.1. you can even more advanced objectives by applying the rules only to certain pixels. small map shifts. etc. # 'cost class' is used to define precisely the cost of changes applied to a site. For example: Example UMTS Overlap > 0 AND UMTS Overlap < 4 (UMTS RSCP > -90dBm AND UMTS EcIo > -12dB) (LTE RS C > -85dBm AND LTE RS CINR > 4dB) OR Description Pilot Pollution avoidance (UMTS) Coverage offered by at least one technology By defining a filter.8 Cost Objective Atoll ACP also takes cost objectives into consideration.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12. For example: Example GSM BCCH > -90dBm FOR Pixels where: (UMTS RSCP < -100 AND Overlap < 2 Description Possibility of Inter-technology handover UMTS->GSM 12. the Atoll prediction resolution. • The most acute match between ACP and Atoll coverage predictions is obtained when the ACP resolution. There are two modes of operation: • • Cost limit: The total cost of the reconfiguration will not exceed a given maximum cost. you can define more advanced objectives. 2 Quality Predictions and the Antenna Masking Method Atoll ACP needs to correctly assess how well a reconfigured network will meet quality objectives when performing an antenna reconfiguration such as changing the antenna model. The ACP performs an un-masking operation with the current antenna pattern.ini file. 12.2. ACP proposes three different modes: Basic. using one of two internal methods: • • Direct calculation: ACP calculates incidence angles by direct calculation using the raster data.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module • • ©Forsk 2015 In "High precision" mode. see the Administrator Manual. ACP proposes 4 modes: "Basic". Cost Hata. You can set the following parameters for the default method: 844 . Improved.3. "Antenna Correction" and "Full Path Loss".2. or azimuth. the coverage predictions calculated by the ACP may differ slightly from Multi-Storey predictions due to different methodologies used by the ACP and the Atoll platforms. All others native models: ACP calculates directly. but with a few additional parameters. In addition.2.1 Optimised Method The optimised method is used for propagation methods which are native to Atoll: the Standard Propagation Model (SPM). These types of reconfiguration are performed by direct path loss calculation. ACP assesses this change by calculating how the path loss matrices change when the antenna is modified. ACP selects by default the Optimised mode.2. 12. 12. CrossWave. etc. Atoll ACP distinguishes between two categories of propagation models: native and non-native. tilt.Atoll 3.2 Antenna Masking Modes for Non-Native Propagation Models For non-native propagation models. refer to the "Configuring Default Settings" section in the User Manual). The ACP uses a mix combining a radial method for lower storeys (based on Atoll's "CalculateGrid" API) and a systematic method for upper storeys where few evaluation points are present (using Atoll's "CalculateSubscribers" API). For native propagation models. ACP increases the number of cells it monitors for each pixel. 12. providing that the propagation model implements the appropriate methods of Atoll's API ACP automatically selects which internal method to use for each native propagation model: • • CrossWave: ACP delegates the calculation to model the propagation model. "Improved". You can define the internal method used by setting the appropriate option in the ACP. Delegating to the model: ACP calculates incidence angles by delegating the calculation to the propagation model. The antenna masking method is not used for site selection and antenna height optimisation. For non-native propagation models. When the Multi-Storey extension is enabled. The Optimised antenna masking method provides accurate prediction of the emission angles. This calculation depends strongly on the horizontal and vertical emission angles between a transmitter and the receiving pixel. thereby reducing potential inconsistencies with Atoll coverage predictions (for more information. and Full Path Loss.ini file.1 Basic Method The Basic method is similar to the Optimised method with direct calculation. This process is strongly dependent on the type of propagation model used originally to produce the path loss matrices. The optimised method ensures that the ACP prediction correlates strongly with the propagation model calculation. followed by remasking with the new antenna pattern. For information on modifying the ACP. reconfiguring power is performed by direct scaling of existing path loss matrices and therefore does not use either an antenna masking method or recalculation of the path loss matrices. the Antenna correction method is not available. The ACP only calculates the path loss matrices for the changes which need to be evaluated by the optimisation algorithm. Either: • • Native 3D Interpolation method: The method used by Atoll. the ACP recalculates all path loss matrices for all combinations of parameters which are tested. Atoll ACP first calculates new path loss matrices for every possible combination of antenna parameters which needs to be tested.2 Improved Method The Improved method performs antenna masking by delegating the calculation of the angles of incidence to the propagation model. If the propagation model does not implement the appropriate methods of Atoll’s API. When using the pre-calculated method.2. It usually produces the best results. By pre-calculating only this subset. 845 . for example.2. has no impact with this method as antenna pattern interpolation is calculated by the propagation model and the method used is the propagation model’s embedded method.2. the ACP recalculates the path loss matrices for that change only. If the propagation model does not implement the appropriate methods of Atoll's API. For more information on Atoll’s method for 3D interpolation.e. the ACP reduces the number of path loss matrices to be calculated and the calculation time. Antenna losses recovered by ACP may include antenna correction and 3D antenna extrapolation. 12.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 Parameter Description Antenna pattern interpolation The antenna gain calculation method for deriving the antenna gain from a set of angles of incidence.3. Either: • • Native 3D Interpolation method: The method used by Atoll. For more information on Atoll’s method for 3D interpolation. 12. If a change is tested on a transmitter that was not taken into consideration when the path loss matrices were calculated. The selection of antenna pattern interpolation.Atoll 3. see the Technical Guide Linear Interpolation method: A simple linear method with a smoothing parameter: G =  A hor  azi  + A ver  elev    smooth Direct view When selected. for example. The optimisation process then uses these pre-calculated path loss matrices to determine how attenuation changes when an antenna is modified. five possible changes in electrical tilt and five possible changes in azimuth.2. see the Technical Guide Linear Interpolation method: A simple linear method with a smoothing parameter: G =  A hor  azi  + A ver  elev    smooth The Improved method usually gives accurate results. the Improved method is not available. You can adjust the following parameter when using Improved method: Parameter Description Antenna pattern interpolation The antenna gain calculation method for deriving the antenna gain from a set of angles of incidence. This is a fall-back method for complex propagation models not accurately modelled by the Basic or Improved methods. or from default clutter heights based on the clutter class file. If available. 25 path loss matrices. Receiver on top of clutter Specify whether the receiver should be considered to be on top of the clutter or not.. i.2. Clutter heights are either extracted from the clutter height file. proposed in an additional parameters column.3 Antenna Correction Method The Antenna Correction method performs antenna masking by delegating both the calculation of the angles of incidence and antenna 3D interpolation to the propagation model. Operator-specific propagation models can often be modelled correctly using the Basic mode. the angle of incidence will be the direct Tx-Rx angle Use clutter height Specify whether clutter heights should be applied along the profiles between transmitter and receiver. the Antenna correction method is the recommended one.2. The ACP does not calculate all path loss matrices for all possible combinations. for complex ray tracing propagation models.4 Full Path Loss Method With the Full Path loss method. 12. In the case of the azimuth. The concepts on which Atoll ACP antenna modelling are based are the following: • • • Antenna Element: An antenna element groups all instances of an antenna.3. Remote Antennas. Use a path loss storage directory which is dedicated to your project region. Therefore. Physical Antenna: A physical antenna is a multi-band antenna. use one of these methods instead. belonging to the same frequency band. • • Power optimisation and site selection (without reconfiguration) do not require recalculation of the path loss matrices. Antenna height reconfiguration as well as new candidates always use a method similar to the Full Path loss method to calculate missing path loss matrices. all antennas will be all displaced to the same height. However. only enable two or three azimuth options. Atoll enables you to carry this out in several different ways: • 846 Using the Physical Antenna field of the Antenna table: You can assign the same name in the Physical Antenna field in the Antenna table to antennas belonging to the same physical antenna. Antenna Groups (Optional): An antenna group is a user-defined subset of the physical antenna enabling you to select antenna model reconfiguration to be done within this subset. and Secondary Antennas ACP fully supports repeaters. when a change to antenna height or azimuth is made to one transmitter on a site. height.3. power. In other words. the same change is made to all transmitters of a site. When using the pre-calculated method. Also carefully design your antenna groups. You have the option of locking height and azimuth optimisation per site. The ACP will then use this information to automatically create all antenna elements and physical antennas. the clutter height files and DTM must be accurate in Atoll so that the ACP can access the terrain profile (even when you have configured CrossWave to directly access building vectors). grouping all antenna elements from different frequency bands which are physically the same antenna. The physical antenna name is displayed in the "Model" column of the ACP Antenna Pattern Table. For each parameter change. 12. etc.Atoll 3.4 Antenna Masking and Repeaters.1 Configuring an Optimisation Setup Setting up the reconfiguration parameters is straightforward.3 Configuration 12. independently of the frequency band they use. If the Basic or Improved method gives accurate ACP predictions that are in line with Atoll. However the "donor side" of repeaters can not be reconfigured. while height locking is enabled by default for all co-localised transmitters on the same site. a range for the parameter can be specified. with different electrical tilts.3 CrossWave Propagation Model Atoll ACP supports the CrossWave propagation model as a native model using the Optimised method and delegating the calculation of the angles of incidence to the model. cascaded repeaters. The repeater or remote antenna can be reconfigured for the "coverage side".2. ACP correctly takes into account the path loss produced by transmitters using secondary antennas. 12. By default. and for antenna height.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 The main drawback of the pre-calculated method is the lengthy pre-calculation required and the disk space required to store the path loss matrices.3.1. .2. for example: • • Maximum variation for azimuth Minimum/maximum range for electrical tilt. For example. azimuth locking is disabled.1 Antenna Setup Electrical tilt and antenna model optimisation require correct antenna modelling. limit the number of parameters covered. mechanical tilt. Modelling the antennas normally only needs to be done once. 12. the following are recommended: • • • Use the pre-calculated method only when necessary. 12. refers to the rotation of the mast. This ensures that future optimisations in that region will have path loss matrices that have already been calculated. and remote antennas. new columns appear in the Antenna Pattern table on the Antenna > Patterns vertical tab to allow you to configure which antenna uses AEDT and the range of allowed electrical tilt.asRelative=1 The following option can be used to create an additional constraint on the Reconfiguration tab that will be applied to electrical tilt changes.max=ACP_ETILT_MAX For more information on the available options in the ACP.etilt.1. the hard-coded value (935. Band" field in the ACP "Antenna Pattern Table": • • If a FREQUENCY custom field exists in Atoll.2 Additional Electrical Tilt (AEDT) Atoll ACP supports additional electrical downtilt (AEDT) processing. The patterns are derived by Atoll ACP using geometric down-tilts of the original antenna pattern.use=ACP_AEDT_USE antenna. then the value is set to 0 and the cell remains empty.1. The ACP will use the data entered in these custom fields to set the default values in a new optimisation setup.etilt. The same default settings apply to mechanical tilt parameters.aedt.3. the first FREQBAND defining a frequency within this range is used.etilt. see the Administrator Manual. You must identify this column using the "antenna. 12. rather than absolute ones: [ACPReconfPage] tx. then a frequency will be determined in the following order: • If "antennaPattern" is referenced by a transmitter.share" option in the ACP.etilt.Atoll 3. • • Manually defining Antenna Elements and Physical Antennas: You can manually define antenna elements and physical or use a REGEX expression. see the User Manual. The following example forces the ACP to find an optimal electrical tilt 4 degrees higher than or 4 degrees below the current electrical tilt. You can enable AEDT support in the ACP by setting the following option in the ACP. If a FREQUENCY custom field does not exist in Atoll or exists and is less than or equal to 0. the ACP allows the reconfiguration of electrical tilt parameters based on absolute values. 12.etilt. This constraint enables the user to define a range of electrical tilt changes within a defined number of degrees above or below the current electrical tilt.ini file. see the Administrator Manual.ini file: [ACPAntennaPage] enableAedt=1 When you have activated AEDT support.3. for all transmitters. This is the preferred method. 2110.deltaLimitConstraint=4 847 . If all fails. You can use the following ACP.ini file.ini file.ini options to reference custom columns in the Antennas table. the frequency defined by transmitter’s FREQBAND is used. AEDT is used when antenna patterns are not available for changes in electrical tilts. Detecting automatically the "Freq. FREQBAND is the "Frequency" field (in GSM) or "Start Frequency" field (in UMTS and LTE) in the Frequency Bands table available from Parameters > Network Settings > Frequencies > Bands. [ACPCustomFieldExtraction] antenna.3 Relative Electrical Tilt Values By default. The following option allows you to display the electrical tilt values in the Transmitters table (on the Reconfiguration tab) as relative values.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 You can also create a custom column in the Antennas table to automatically link antenna elements of a multi-band physical antenna which have the same electrical tilt. [ACPReconfPage] tx. • • If the project contains a single FREQBAND. For more information on the ACP. or 1805) contained in the [FMIN-FMAX] range is used. Else.min=ACP_ETILT_MIN antenna. as ACP will then automatically create all antenna elements and physical antennas each time a setup is created. the value it contains will be extracted. For more information on manually defining antenna elements and physical antennas.3. then the frequency defined by this FREQBAND is used If "antennaPattern" defines a [FMIN-FMAX] range. and provided that co-located transmitters are within a user-definable inter-antenna distance (default = 1m).e. If at least one transmitter defines a "Shared Antenna". and same antenna height (within a user-definable inter-antenna distance [default = 1m] for position. etc. same azimuth. and co-planning documents.colocated = 2 Co-located Antenna (i. the ACP automatically detects colocated sites and antennas. The ACP setup then becomes a multi-RAT setup. the ACP will issue a nonblocking warning. The benefits of using the ACP in multi-RAT mode are: • • • You can define multi-frequency band/multi-RAT combined objectives You can automatically synchronise shared multi-band antennas.ini option below: [ACPTplReconfPage] tx. multi-RAT. • When two transmitters are linked. problems in the Atoll database can mean that the ACP does not recognise that sites or antennas are co-located.3.min. Transmitters) • User-definable inter-antenna distance (default=1m) using the ACP.4 Multi-RAT and Co-planning Support The ACP fully supports multi-frequency band. If this happens. All transmitters from co-located sites with the same shared antenna are linked if they have a different frequency band. or if the technology is different. you can upgrade existing sites with a new radio access technology.4. The ACP automatically detects sites supporting several technologies.distance.ini option below: [ACPTplReconfPage] site.). including secondary antennas (i. and take the upgrade cost into consideration. you can manually set the sites or antennas to be co-located. then the following logic is used: • • Detection of co-located sites as sites located within a user-definable inter-site distance (default = 2m). and within 1 metre for antenna height). same mechanical tilt. In a multi-RAT document.Atoll 3. • The normal way of detecting linked transmitters is to use the "Shared Antenna" field in the Transmitters table (SHAREDMAST). within 1 degree for tilt.min. In addition. then the ACP will use another mode where it automatically detects the linked transmitter using the same criteria as the one used in sanity check (within a user-definable interantenna distance [default = 1m] . 848 . If the "Shared Antenna" field is not used by at least one transmitter.e. when Shared Antenna is not or is only partially used) using the following algorithm: Parameter Description Co-located site • User-definable inter-site distance (default = 2m) using the ACP. • When two linked transmitters are not consistent. as well as shared multi-band antennas using the Shared Antenna field of the Atoll Transmitters table when present (SHAREDMAST in the database). although you should also review the database to correct any errors there. within 2 degrees for azimuth.1 Multi-RAT and Co-planning Modes When working in co-planning mode with several Atoll documents. This ensures that any antenna reconfiguration is properly taken into account in all impacted technologies.colocated = 1 • Antenna height within 1 metre • Antenna azimuth within 2 degrees • Mechanical tilt within 1 degree Same physical antenna when the antenna defines this field • Occasionally.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 12. Shared site location is automatically managed for site location and site candidates. 12.distance. their values in the Current columns on the Reconfiguration > Transmitters vertical tab are highlighted in red. within 2 degrees for azimuth. ACP enables you to import the other Atoll project into one ACP setup. Sanity check is performed to validate that antenna parameters are consistent: same position. e. Each quality indicator is evaluated for the technology layer to which it is assigned. A number of additional techniques are used to improve the basic process. consisting of enumerating all possible solutions. it optimises all the objectives (quality and cost) combined into a single global score function. This search algorithm uses the concept of iterations: each iteration consists of one parameter change on one of the sectors or sites of the network. you can put more emphasis on some technology layers by modifying the global weight of the objectives of each technology layer. you need to: • • Define a heavier weight on the objectives related to the "target" technology layer. the algorithm attempts to find the best parameter combination to minimise the global score function. Some iterations might also cancel each other. although this parameter still depends strongly on the size of the network and the quality of the initial network. 849 . however you can group quality indicators from different technology layers within a same objective. aspiration. any new change is forbidden during a certain number of iterations. Or use a coverage target for the objectives of the "constraint" technology layers which are relative to the current coverage (where a successful optimisation would be defined as "no coverage decrease"). there is not.. while allowing for the exploration of new locations in the search space.. When you are using the ACP with more than one technology layer (and.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12. and long-term search. When a transmitter has been allowed a parameter change.5 Optimisation Methodology When the Atoll ACP performs the optimisation. and should be high enough to ensure that the search space is properly covered. more than the number of atoms in the universe.Atoll 3. i..2 Technology Layer Definition ACP sees each radio access technology as one or several technology layers which are defined according to the following rules: Technology Technology Layer Definition Example GSM Each independent frequency band is seen as a separate technology layer • • GSM 900 GSM 1800 UMTS and CDMA Each carrier is seen as a separate technology layer • • • Carrier 10562 of 2110 MHz band Carrier 10587 of 2110 MHz band Carrier 2937of 925 MHz band LTE and WiMAX Each independent frequency band is seen as a separate technology layer • • 2010 MHz band 900 MHz band When defining objectives. this algorithm performs local greedy optimisation. very quickly becomes unmanageable. i. in multi-RAT projects as well)..5. therefore. such as randomisation. a sector is returned to its initial state at the second iteration.e.4. for example) is also used to improve the process and make it efficient. If one needs to be optimised without degrading others. All are considered the same. The number of iterations is a key parameter of the optimisation. i. a sorting algorithm provides an implementation plan where the most useful changes (in terms of minimizing the score) are done first and the least useful changes are done last. In the final step. In short.1 Search Algorithm The possible number of configurations grows exponentially with the number of sectors to optimise. It should be noted that a given sector might be modified in several steps. leads to a search space of 10100 possible solutions. one target technology layer and one or more constraint technology layers. selecting the best antenna among 10 possible antennas on a 100-sector network. for example. Atoll ACP uses a Tabu-based search algorithm with fast convergence. i. 12.e. For example. Usually a few times the number of entities to optimise is sufficient. ACP optimises the quality objectives for all technology layers. each rule is associated with a single technology layer. The naive search method. Knowledge of the particular nature of the network (cell neighbour relations.e.3. This global score function is used as the basis for the search algorithm. 12. the final change might be the result of several different iterations. diversification. 12. different points on the Pareto surface of multi-objective optimisation problems. It performs a local optimisation of the network while preventing useless changes from being done.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 The ACP recommends a number of iterations. the change which mostly improves the score function STOT(x). and better management of early manual stops by the user. i. The tuning phase is fully transparent to the end-user and it provides the following benefits: • • • • • Removes the changes with insignificant benefits. Figure 12. The tuning phase stops as soon as the score function (see "Global Score Function" on page 851).3. etc. about 2/3 of the iterations are used in the Tabu-based search phase.e. Provides a better automatic stop condition for the Tabu-based search phase. Due to the behaviour of the optimisation phase. The requested number of iterations is used in both phases. The recommended number of iterations is calculated by multiplying the number of entities to optimise by two. and to the possibility of early stopping. without having to also perform local optimisation around the best candidate solution.e. it can happen that the ACP finds the non-optimal local optima.2 Tuning Algorithm The ACP search algorithm includes a tuning phase between the search and sorting phases (corresponding to optimisation and finalisation phases in the ACP GUI). By default. and 1/3 in the tuning phase. Allows the Tabu-based search phase to concentrate more on spanning the solution phase. changes with less than 1% of the benefit provided by the best single change. i. you ensure that the search space is explored correctly.Atoll 3. On the Graph tab of the Optimisation window. Finds the local optima around the best solution of the optimisation phase. i. the ACP proceeds as in the Tabu-based search phase. By defining a number of iterations equal to or greater than the recommended number.5. can no longer be improved. The tuning phase consists in improving the best solution found during the search phase. These ratios can differ when an early stop (automatic or manual) is performed during the Tabu-based search phase.3: Graph tab of the Optimisation window 850 . It simply finds the best neighbour candidate to move after each iteration. randomisation. but without using a Tabu list.e. During the tuning phase. Will be useful in future releases for better management of multiple solution findings corresponding to different qualities or cost trade-offs. a vertical bar is displayed to show the switch point between optimisation and tuning phases. This feature is fully transparent. The sorting algorithm recursively finds the best change to apply among all remaining changes. The best change is the one which improves the total cost function the best: • • The first changes proposed have more benefits (in terms of the trade-off between quality and cost) than later changes. 1 Cov obj2 = --N    i   1 Th  E c I o  i   Thresh EcIo  i  pixels Where: •   i  is the normalised traffic density on pixel "i":   i = 1 i  pixels • • E c and E c I o are the basic quality measurements on one pixel as described earlier.5.5.1 Search Algorithm The global score function used as a basis for the search algorithm is created by a linear combination of the cost objective and every quality objective. The coverage costs are null if the target coverage is reached: f i  x  = 0 for x  T arg etCov • f c is a one-dimensional function mapping the network financial cost suitable to be used alongside the quality objective costs. The global score function is in the following form: C TOT  x  =  ai  fi  Qi  x   + k  fc  C  x   i  quality obj Where: • i is an index spanning all quality objectives defined • n is the network configuration to be tested • Q i  x  is the "ith" quality objective evaluation • a i is a weight factor associated with the "ith" quality objective. f obj1 and f obj2 are the one-dimensional mapping functions expressing the individual cost for a coverage figure or network quality: 851 . A change is typically a modification to an antenna parameter or (for candidate sites) deploying a site or (for existing sites) sector or removing a site or sector. and applies different importance on the different • quality objectives for the different technology layers.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12. Atoll ACP proposes a solution consisting of a number of changes to be applied to the initial network. Atoll ACP then uses a sorting algorithm to create an ordered implementation plan. Example C TOT  x  = a obj1  f obj1  Cov obj1  x   + a obj2  f obj2  Cov obj2  x   Where: Cov obj1  x  (or Cov obj2  x  ) is the percentage of coverage over a specified threshold for configuration "x". 12.3 Sorting Algorithm After applying the search and tuning algorithms. For example.5. It is then possible to select a subset of the total number of changes by selecting only the N first changes. in UMTS: 1 Cov obj1 = --N    i   1 Th  E c  i   Thresh Ec  i  pixels Where: 1 Th is the step function.4. C  x  is the (financial) cost associated with configuration "x" • f i is a one-dimensional function expressing the individual given cost for the "ith" quality objective measurement.4 Global Score Function 12.3. focus. but takes longer to run.2 Tuning Algorithm The global score function used as a basis for the tuning algorithm is the following: S TOT ( x )  CTOT ( x )  p  f n  N ( x )  Where: • x • CTOT (x) is the global score function described earlier • N (x) • p is the network configuration to be tested is the number of changes performed in the network from the initial configuration is a weight factor derived automatically to have an appropriate scaling of the new term with chosen such that the "cost" of one change is equal to q% of the score function improvement provided by the best individual change in all the proposed changes. tilt. the optimisation will likely find a better solution but will take longer to run. The following table gives typical values to be used for a good optimisation: Parameter 852 Typical Value Number of iterations Around 1 to 2 per item (antenna.3.Atoll 3. the sampling is more approximate but the speed is highly increased.5 Weighting Several types of weight are applied during the calculation of the global score function.5. It is q  1. in initial implementations. Similarly. These parameters affect the quality and speed of the optimisation. hot-spot zones or clutter-group zone) can have an additional weight which increases or decreases the importance of the zone.6 Controlling the Optimisation Although the Atoll ACP process is designed to be as automatic as possible. there are a couple of parameters that require some consideration: • • Number of iterations: This option defines the number of iterations in the search algorithm. azimuth. Zone weighting: Each pixel within a defined zone (computation. On a geographical level (used to calculate the weights   i  in the above formulas): • • Traffic weighting: Each pixel can have an importance proportional to the traffic supported on it. Atoll ACP provides information on the total number of pixels. if the number of iterations is high.5. On a global level (forming the weights a i in the above formulas): • Quality objective weighting: Within a technology layer. cell pilot power. Resolution: The resolution defines the size of the pixels used to measure the quality objectives. each quality objective can be given more or less importance as compared to other quality objectives of that technology layer.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 f obj1  cov  = 0 for cov  T arg etRSCP f obj2  cov  = 0 for cov  T arg etEcIo 12.) to be optimised Resolution The average number of positions per site is between 300 and 3000. . as well as the average number of pixels per site. By default • fn CTOT ( x) . 25% is a one-dimensional function expressing an individual cost for a given number of changes. etc. 12. it is the identity function: fn y  y 12. If the resolution is high. If the resolution is low.4. As a suggestion. By default. ACP does a better job of sampling the network zone.5. with the selection of Maximum Cost or Quality/Cost tradeoff options).3. if used. It takes into account the coverage and quality objectives (i.Atoll 3.4: Setup > Properties > Optimisation tab > Cost Control dialog box Only the data displayed on the Change Details tab is actually separated for quality objectives and cost of changes: • Quality Improvement Ratio %: this column shows the ratio of attained quality VS the maximum quality when all changes are made (the displayed values range from 0% and 100%).7 Implementation Plan The sorting process in the implementation plan is based on a global score function which includes quality objectives and the cost of changes (i.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 You can start with a low resolution first. then the first one in the list will be the one with the lowest associated cost. However. The data is separated as it makes more sense to display understandable values in 2 columns rather than display a "Score improvement ratio %" which would be difficult to understand. the financial cost. it takes into account the quality settings as well as the Cost Control settings if cost of changes is considered.e. 12. LTE RS Coverage and LTE RS CINR) and. • Total Cost: this column shows the associated cost of changes.e. 853 . When the ACP is running an optimisation. C TOT  x  =  ai  fi  Qi  x   + k  fc  C  x   i  quality obj Figure 12. before using a higher resolution for more accurate results. One consequence of this is that if you obtain 2 changes providing the same quality improvement ratio. the load balancing objective is not considered. the tabs of the Optimisation dialogue provide feedback which can help you to decide to stop the optimisation early if the overall network quality seems to have improved enough. This ratio allows you to know the relative gain of each change.5. For more information on ACP memory limits. a list of the neighbouring cells in order to be able to find new best servers and calculate interference levels.5. and the algorithm is optimised to improve speed. Normal: The normal mode with a balanced trade-off between speed. and accuracy.5. These internal data structures are created during the loading phase of an ACP run. each time on a different portion of the entire area. The ACP needs to store. You can change how the ACP manages the data it loads into memory by setting certain options in the ACP.5. If this happens. You can use the "threshLevelMonitorCell" option to define the best server signal threshold (dB) of the cell in order for the cell to be monitored.3. As well. Given the amount of data processed. If this is the case.5: Setup > Optimisation > Properties > Change Details tab 12. and the accuracy of the optimisation (especially on measures related to interference): • • • High speed: The cell list is shortened to reduce memory use. for each pixel. you can set the number of cells that the ACP monitors by setting the "maxMonitorCell" option in the ACP. 854 . you can rerun the optimisation by decreasing the resolution or decreasing the size of the computation zone.ini file. In practice. it is usually sufficient to change the mode of operation. it is recommended to use the High precision mode in order to ensure that all neighbour candidate sites are well monitored by the ACP. see the Administrator Guide. 12. as allocations to memory are very limited during the solution search. When doing site selection in Greenfield scenarios where a lot of candidate site are defined close to each other. for example) in memory in a format optimised for fast processing. For each pixel. High precision: The high precision mode results in higher memory usage and a lower speed. it would be prohibitive in terms of the amount of time necessary to read path loss matrices from the disk on each iteration.8 Memory Usage and Optimisation Resolution The administrator can set an option in the ACP. memory use. but offers the highest accuracy by monitoring a longer list of cells. For example.ini file. you can limit memory usage by performing the optimisation in several steps. you might reach the set limit when using a high resolution and Atoll will stop the ACP optimisation early. however. Each objective is the combination of one or several quality indicators evaluated on several hundred thousand or even millions of pixels depending on the resolution and network size. see the Administrator Guide.9 Internal Data Management and Performance 12.Atoll 3.1 Memory Usage For each tested network change.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 Figure 12. Any memory issue that could occur when optimising a large network should happen during this initial phase. thereby performing a trade-off between memory usage. the ACP keeps the list of neighbour cells and their attribute (path loss attenuation. the ACP recalculates how each objective improves or degrades.ini file to set a limit on the amount of memory that Atoll ACP can use.9. when performing an optimisation on a large area. For more information on these options. e.9. Other technologies use the same principles. Another approach is to consider that the actual cell loads are fixed. In particular this approach would still provide useful insights on load imbalance from only a partial traffic model (considering for example only 1 or 2 services typically representative of the traffic distribution). across technology layers. i. By detecting when the cell loads become imbalanced or in excess of given cell resources. To better use the ACP and avoid lengthy recalculations after rollback changes. etc.1 Traffic Capture for Load Balancing The ACP is designed to perform load balancing across multiple technology layers. and then make sure that the supported traffic is maximised and well balanced. the cell capacity load is related to transmitted power: • • A total transmitted power is computed over the whole Best Server area by adding transmitted power for each pixel. Derivation of the score function of the load balancing objective from the cell capacity loads. however with a different definition of cell capacity load. The ACP approach is basically to perform cell capacity load balancing. This approach avoids the complexity of cell load calculation through complex Monte Carlo simulation.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12. specify a Shared directory for path loss matrix storage in the predictions’s Properties dialogue. but still assuming that the cells have a fixed load for the purpose of interference calculation. For UMTS R99. This means that the requested traffic will be shared across the available technology layers within the allowed technologies for this service.6. This also holds for the path loss directory. Since the cell power load is fixed in ACP calculation.2 Disk Space Usage To reduce disk space usage. fixing the level of interference generated in the network corresponding to a target load.6. but is only a capacity indicator computed by assuming a fixed cell load and a traffic model which could be simplified model as long as it is representative of the actual traffic distribution.6.1 Principle Used in ACP One obvious approach for load control and balancing within ACP is to compute the new actual cell loads for each tested reconfiguration. and we believe it is more robust to imperfect inputs. avoiding that some cells use excessive power.. it is possible to avoid such reconfiguration. Moreover. and for which: 855 . Computation of cell capacity load from the assigned traffic.2 Optimisation Principle The calculation is performed in 3 steps: • • • Assignment of the requested traffic to the various cells on the network. The cell capacity load is not the actual cell load derived from a set of realistic traffic maps and services using a Monte Carlo simulation with power convergence loop. The ratio of this total cell power over the maximum cell power is the cell capacity load. by focusing on a what-if scenario where cell loads are set to an average target load (using for example a target network load of 70%). Any imbalance in the actual cell load is thus reflected by imbalance in the capacity cell load.5.6 Load Balancing Objective 12. service. the user can define in Windows a private storage directory for ACP with compression set to ON. The overall difficulty is the calculation of cell loads given the precise inputs required for traffic. When the cell capacity loads are being successfully balanced. This is the approach used in ACP. Each pixel power is scaled with traffic density distribution. another method is needed to insure that the real cell load does not increase beyond capacity and is correctly balanced between cells. 12. as well as the complex and lengthy calculation involved which would require full Monte Carlo simulations for each tested reconfiguration. 12. The key reason why cell capacity load balancing is a suitable approach is that cell capacity load is correlated to the actual cell load.3. then the computed capacity cell loads measure directly how much room is available or missing in actual cell capacity to support this target network load. (but which are often imperfect in practice). 12.Atoll 3. The requested traffic for each service is assigned to cells according to the following rules: • Candidate cells for assignment of a pixel traffic are selected among all best server cells in all different technology layers. and is expressed in % of the available resources.2. The ACP equalises the cell capacity load. Each pixel transmitted power is computed by using the load factor equation. they tend to converge loosely towards the actual cell load. 6. The traffic is assigned partially to each of these candidate cells. 12.3.3 Load Balancing Score Function The Load Balancing Score Function being minimised to drive ACP optimisation is the following: 856 . it will be assigned to the Cell with the minimum cell load. including antenna gain and losses • G proc is the service processing gain (Gproc = 3.84e6/Tputserv) In LTE Given the pixel SINR. the process still tends to equalise technology layer loads as much as possible. followed by a convergence loop to reach a minimum state. In other words. while computing the cells capacity loads. In UMTS R99 The load factor equation used is the following: Eb No  Io 1  = ----------. The end result of this process is to distribute the traffic across cells and technology layers in such a way that overlapping cells from different technology layers tend to be equally loaded. in UMTS the cells belonging to same transmitter usually have the same footprint. The procedure uses a simulation process where the pixel traffic is added gradually to the network.2. and the cell capacity loads are updated after each assignment. this process simulates a perfect call admission control procedure whose purpose is to perfectly equalise the requested traffic among all cells and technology layers in the network: a new call is always assigned to the technology layer and cell having the minimum load. if a pixel traffic requested can be assigned to Cell1 from TechnologyLayer1 and to Cell2 from TechnologyLayer2. etc).2 Cell Capacity Load Calculation The traffic assignment stage basically balance the traffic request across the different cells. and is given by: Tput serv  = ------------------Tput max Where Tput serv is the average requested throughput for the service 12. for cells which do not fully overlap.Atoll 3. -----------------------P max A tt  G proc Where: •  is the pixel load ratio • P max is the maximum cell power • E b N o is the target EbNo for the given service • I o is the total noise and interference • A tt is the attenuation towards the cell.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module • • • ©Forsk 2015 The technology layer is allowed for the service. The service traffic capture conditions are fulfilled. The cell capacity load is the sum of all pixel contributions.6. The unit of capacity load is a percentage (%). For example. and as such the computed load will tend to be equal across these cells. one derives first the best Bearer which can be assigned to a call originating from that pixel. The exact method for computing the contribution of a pixel to the cell capacity load is technology-dependent. For example. The pixel load ratio is then the ratio of resources used by the service on this pixel. scaled with assigned traffic to the cell’s technology layer:  Li = T k i   k Pixel  k  Cell i  Where:  k is the pixel load ratio and is T k i is the traffic assigned to pixel "k" and cell "i".2. This process basically simulates a network where the traffic is dynamically assigned to technology layers in order to equalise cell loads. Technology layers with high capacity (for example LTE vs GSM) tend to acquire more traffic automatically. such that the cells capacity load is minimal. then derives the maximum possible throughput Tputmax which could be provided to that pixel-originated call (given bandwith. Inactive cells are not considered in the calculation (in term of average/standard deviation values and number of cells). 12. due to traffic capture condition being fulfilled b is a scaling factor to give more or less weighting to the traffic captured ratio (default is 1) This score function will be considered in the ACP global score function when load balancing is activated.e. 12.  Wli   Li – l   Wli Celli  Layerl i Where W li is a weighting factor applied on each cell load.6. The overall Load Quality index is an average of the all the technology layers’ Load Quality indexes. to the load imbalance (default =1) 1  l = -------------.4 Load Quality Index The Load Quality Index is defined as: QI = Mean  QI l  Where: • QI l =   l + a   l  is the Quality Index computed for technology layer "l" and: •  l is the weighted average of the cell capacity load •  l is the weighted standard deviation of the cell capacity load • a is a scaling factor to give more or less weighting to the standard deviation. and is used to reduce the effect or completely deselect a cell in the calculation: • • Cells inside the target zone are considered as having a weight of 1.2. i.e. whose minimisation reduces both the average cell load and load imbalance (explained below with formulas)  is the Traffic Captured ratio which measures how much traffic is potentially served in the target zone. This is described in "Impact on the Global Score Function" on page 861.  Wli  Li W  li Celli  Layerl i 1 2  l = -------------. See definition of "b" in "Load Balancing Score Function" on page 856.5 Captured Traffic Ratio The captured traffic ratio is defined as: T ass  = -------T req Where: • T req is the total traffic requested in the target zone • T ass is the total traffic assigned in the target zone Increasing scaling factor "b" leads to increase the total traffic assigned when the Score function is being minimised. for technology layer "l".2. and cells outside the target zone are allocated a weight of 0. 857 .0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 Score = QI   1 + b   1 –    Where: • • • QI is the Load Quality Index. each technology layer gets identical importance.Atoll 3.6.3. i. A default Load Balance Target value is defined for the Load Balance in the user interface (e. when the Load Balance target is reached. the cell load distribution is not derived from a Normal distribution. we then decrease the load of the 85th percentile of the distribution. however the argument still holds: decreasing the   + a    will focus on the capacity network bottleneck. 12. Cell Load Distribution Percentiles represent the area under the normal curve. only the average load (  l ) is considered for minimisation in the score function. By using "a = 2". 80%). we directly increase the capacity by focusing on the high distribution percentile. Hence. As the bottleneck in network capacity for a given quality is often given by its most loaded cells. 858 . the load decrease is for the 98th percentile of the distribution. minimising the Load Quality Index can be seen as a way of decreasing both the Load Dispersion (thus improving the Load Balance) and the Average Load (thus improving the network capacity). the Load dispersion part of the Load Quality index is set to null.6.3 Quality Figures Used for Graphs and Statistics Results The ACP provides graphs for the Load Balance and the Average Load values. The goal is to provide a percentage value for an improvement and a graph which increases when the quality indicator increases.Atoll 3.6. Hence.6.3.2. • • By decreasing the Quality Index with "a = 1". In the most general cases. in terms of quality figure for an easy understanding by users. in other words a low level of imbalance. for example 10% or 20% without further optimisation.g. Let's make the assumption that the cell capacity loads are drawn from a Normal random distribution (Gaussian process).0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 12. This Load Balance target allows a margin. Figure 12.6 Introduction of Load Balancing as a Quality Indicator The Load Quality Index can be rewritten as follows: l QI l =  l  1 + a ---- =  l  1 + aD l  l Where: •  D l = ----l is the Load Dispersion measure for technology layer "l" l • the Load Balance (B) is defined as B = 1 – D l This load dispersion parameter directly measures the load imbalance. It is completely described by both mean and standard deviations. since the objective is above target. When this target is reached. increasing from left to right. The minimisation of the Quality index can also be seen as minimisation of the number of overloaded cells. 1 Load Balancing Tab The Average Load and Load Balance quality figures are shown on the Load Balancing tab for any specified zone.6. However. They are based on the cell capacity loads which are displayed on the right on the Load Balancing tab. a 100% improvement means a decrease by 2 of the average load. which is expected during optimisation. etc. with initial and final values available on the Statistics tab. it is possible to recompute these quality figures using the formulas described in the previous section. B = 1 –  --. This quality figure will tend to increase when the average load decreases.8 -- when 1 LB = --------------------------- 2 1 + 4   ---    when  It is displayed as a curve on the Graphs tab.6. 12. and to show 0% for total imbalance. Consequently.Atoll 3. must be adapted due to negative values obtained when    . 859 .2 Average Load The quality figure is given in terms of improvement (%) from the initial average load.3. 200% a decrease by 3.4. or  0 the value for step N and  the value for step N+1.6.1 Load Balance The figure is designed to show 100% for perfect load balance. Hence.3. the formulas used  for the Load Balance quality figure are:  LB = 1 – 0.6.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12. It is displayed as a curve on the Graphs tab and the final improvement can be found on the Statistics tab.4 Optimisation Results 12. In both cases. The graph also directly shows the cumulative distribution of cell capacity loads. 12. the formula used for calculation.– 1  Where  0 is the initial value and  is the final value. The formula used for the quality figure is 0 ----. Note that the Load balance value will tend to increase (and the graph to go up) when cell capacity loads are better equalised. -75% an increase by 4. from the list of cell capacity loads. -50% an increase by 2..3. thus providing the ratio of cell capacity load being smaller than some load value. 860 . 12. However the capacity load statistics displayed on the Load Balancing tab are calculated based on the technology layers or hetnet layers and the zones that are currently selected in the dialog box.3.4.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 Figure 12. the displayed initial/final load balance values can be different on each tab.2 Graphs The graph representing the Load Balance quality figure shows the progress of this quality figure for each iteration.7: Setup > Optimisation > Properties > Load Balancing tab The initial/final load balance values displayed on the Statistics tab and the Load Balancing tab will be identical if: • The same technology layers are considered: • in the Load Balancing page of the setup properties’ Objectives tab • next to For Technology Layer on the optimisation properties’ Load Balancing tab • The same zone is considered: • next to Evaluation Zone on the setup properties’ Optimisation tab • next to For Zone on the optimisation properties’ Load Balancing tab The values displayed on the Statistics tab are calculated for the cells based on the selected technology layers or hetnet layers and located within the target zone (more precisely for the cells that are actually considered for load balancing).Atoll 3. Therefore.6. ACP calculates the EMF exposure in two dimensions (for open areas such as parks or roads) or in three dimensions (for buildings). a "worst case" mode can be used for the EMF exposure calculation where EMF predictions are very pessimistic. 861 . with buildings.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 Figure 12. The internal propagation model evaluates the field strength in V/m using a model based on "free space" propagation. where the EMF exposure will be the greatest.6.8: Graphs for Quality Figures Figure 12. 12.9: Statistics results (extract) 12. For example. providing the highest EMF values which would ever likely be seen in the real world.3. For regulation requirements. hospitals. Additionally. It is calibrated in such a way that X% improvement of the Load quality index is equivalent to an identical improvement on other objectives. coverage) can be modified from the user interface (Objectives tab > Load Balancing > Weight parameter).Atoll 3. The ACP can analyse and optimise the EMF exposure in the network in order to reduce excessive electromagnetic radiation in populated areas. but it can take diffraction into account when required. Using an internal propagation model specific to EMF exposure. This mode is useful to ensure that an unacceptable level of EMF exposure is never reached in sensitive areas such as schools. The weighting of the load balance objective versus other quality objectives (e. it is typically a few V⁄m. you can choose to measure the exposure only at the front façade. 1% improvement of the Load quality index is equivalent to 1% improvement of the RSCP coverage. etc. Although the exact limit on the acceptable level of EMF exposure varies by jurisdiction.5 Impact on the Global Score Function The Load Balancing Score Function is added to the ACP global score function. which already includes quality objectives and the cost of changes.7 EMF Exposure EMF exposure is defined as the total electromagnetic field measured at a given location.g. They can be considered as transparent but with a certain degree of diffraction loss. each polygon is associated with a single propagation class and a height. you can distribute evaluation points on one geo vector entry for one subset of polygons. such as parks. The signal experiences some loss when going through and also suffer from diffraction loss. but instead it uses all the geo data available in the Atoll project: • • Geo raster data: Raster data give a grid-based representation of the terrain with a defined resolution. The raster files needed are DTM (Digital Terrain Model) representing the ground altitude. habitation versus monuments). For areas covered both by vector data and raster data. ACP uses the geo data to create a 3D representation of the terrain in the form of a fine raster of pixels with a default resolution of 2 meters. specified by the "Linear loss start distance (m)" parameter. the default clutter class height is used.1. with no diffraction loss.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 12. Geo vector data: Geo vector data model the buildings and their height. Currently. only the geo vector data are used. for a bridge) to better reflect where people will be.Atoll 3. the points can be placed either at the bottom (for example. Similarly you distribute evaluation points on only selected clutter classes.7.3 Distribution of Evaluation Points ACP uses the internal terrain representation to specify where to set evaluation points for EMF exposure evaluation. It is possible to define new propagation classes.7. The loss is applied only after a given number of meters. For example. The measurement points can be distributed in either a 3D pattern or in a 2D pattern. Building: The Building propagation class is used for opaque objects such as buildings.7. The propagation classes have the following parameters: • • • Penetration loss (dB): The loss occurring when the signal enters the object. For a two-dimensional distribution. clutter classes representing the type of terrain and clutter heights (also called a digital height model) representing individual heights (for example.1. Distribution of evaluation points: Field strength measurements are made on a set of points within an object. 862 . but not on another vector entry. wood constructions. both the height and propagation class information are encoded: • • For geo vectors. 12.) or for differentiating items on which EMF evaluation should be done (for example. 12. ACP uses the associated DBF file to map the polygons to propagation classes and heights. etc.1. building heights). for example to differentiate between similar items with different penetration loss characteristics (for example. Vegetation: The Vegetation propagation class is used for areas covered with vegetation.7. The following default propagation classes are provided: • • • Open: The Open propagation class is for areas without obstacles. meaning that the signal experiences diffraction losses at the edge of the object but does not go completely through. in a park) or at the top (for example.1 Concepts of ACP EMF Exposure 12. Each propagation class is either opaque.3. glass buildings. meaning that the signal passes through it (with perhaps some losses) and does not experience diffraction loss. An open area can also be an elevated area such as a bridge. Such areas are transparent. in the form of one or several ArcView SHP files defining numerous polygons. For geo rasters. each clutter class is associated with a single propagation class. user-defined classes are always of the type "opaque". Linear loss (dB/m): A linear loss applied for each meter within an object. It is recommended to always provide either geo vector data or clutter heights raster data to have the most accurate EMF exposure prediction. stone buildings. For each pixel in this raster representation. It is possible to distribute evaluation points separately on each propagation class and for each terrain entry. If no clutter height file is present. Geo raster data are only used for the areas not covered by geo vector data. The height is obtained from the clutter height raster file.1 Propagation Classes The internal propagation model calculates EMF exposure using propagation classes which are retrieved from input files. or transparent. the ACP does not need any specific terrain modelling. such as an open area or water. If a geo vector contains more than one polygon.2 Terrain Profile To measure EMF exposure. This mode gives you a pessimistic view of the actual exposure since.7. 12. The overall EMF exposure calculation is obtained by adding the electromagnetic signal level generated by each technology involved (GSM. parks. This mode is useful when optimising the networks while ensuring that regulatory limits are never exceeded. WiMAX) and considering all carriers and channels used. It then distributes a number of evaluation points in this representation according to the parameters set in the setup. In this case it is also possible to specify the maximum indoor distance on which to measure EMF exposure. The ACP also takes into consideration the antenna radiation pattern by creating a 3D interpolation from the 2D horizontal and vertical cross-sections of the antenna radiation pattern.3 EMF Exposure Calculation The calculation of EMF exposure is based on the following formula giving the electromagnetic E field (in Volt/meter) at distance d .2 General Workflow ACP creates a representation of the terrain in 3D. the following conditions must be met for the most reliable results: • • GSM: the number of TRXs must be correctly referenced in the Atoll database.4 The Contribution of Transmitter Power to EMF Exposure The ACP takes the maximum power transmission on all the carriers and channels used by transmitters in the network into consideration in its calculations.7. as well as some diffraction and indoor losses. LTE. forests. The method is similar to the one used elsewhere in Atoll. roads.g.7. Vegetation: For the Vegetation class (e. This ensures that all provided results are for a fully loaded network. bridges. The EMF exposure is therefore evaluated as if the area was completely free space. penetration loss through obstacles as well as diffraction and reflection around obstacles tends to strongly decrease the signal strength compared to a completely free space model.. 12. It takes into account the antenna gain and attenuation patterns towards each evaluation point. open spaces.3. ACP calculates the quadratic sum of all channels in all multi-RAT technologies: E = 2 2  E GSM  +  E UMTS  +  E LTE  2 The following parameters are used in the calculation: o  BcchPower • GSM: N • UMTS:  MaxPowerCelli TRX ACP takes into account the maximum transmission powers of cells or transmitters to simulate "worst case" scenarios.) evaluation points are distributed in 2D on the top of the class height.g. thereby giving a worst case calculation of EMF exposure. in free space far field: 863 .1. In UMTS. Building: For the Building class evaluation points are distributed in 3D. 12. To predict the overall EMF exposure.) evaluation points are distributed in 2D at the bottom of the class height.7. UMTS. Hence. etc. and or to restrict the prediction on the building front façade.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 Evaluation points are distributed either in a 2D or 3D pattern depending on the propagation class: • • • Open: For the Open class (e. in the real world. the other carriers will also be included in the calculation.Atoll 3. ACP evaluates EMF exposure on each of these evaluation points using an integrated propagation model specially tailored for the evaluation of strong signals in the vicinity of the antennas. 12. CDMA. It is highly recommended to activate all technology layers. where any opaque propagation class such as Building becomes fully transparent to electromagnetic waves. even if only one technology layer is activated in the ACP interface for a given frequency band.1. and UMTS and LTE: all the cells which are to be considered must be present and activated in the Atoll database.5 Worst-case Mode The ACP allows you to consider a worst case scenario. etc. This modelling is based on free space formulas which are applicable starting at a few meters from the antennas up to a few hundred meters.. the plane wave power density is given (in Watts per square meter) by: P WM 2 P tx G tx = -------------2 4d The power received by a theoretical ideal antenna with an effective aperture of A er and a gain of G rx is given by the Friis transmission formula for wavelength  : P rx = P Wm 2 A er 2 Where  A = -----. Propagation models designed for coverage analysis typically deal with signal levels usually lower than -40 dBm. In non-line of sight situations or far away from the base station (i.G rx 4 By combining the Friis formula with the expression of E previously defined...Atoll 3. This model is based on the free space far field formula since "line of sight" exposure can cause significant EMF exposure. each one having a far field starting at a distance of 2 (or less than 1 meter for a typical frequency of interest). the significant EMF exposure level (> 0. urban empirical models such as Cost-Hata models are typically an extension of the Friis formula where the 2 n distance denominator 1  d is replaced with 1  d . the far field formula usually leads to good field estimates starting at a distance of around 5 meters from the antenna. For example.3. The signal strength then becomes smaller than the range of interest (a few tenths of V/m). beyond a few hundred meters). distance. etc. resulting in a potential health hazard (when exposure is above a few tenths of V/m). In addition they only measure the signal level in a 2D horizontal plane. reflection. 864 . with n being a value from 3 to 5. However. This is in part because antennas are formed of several stacked dipoles (for example 8 to 10) with low coupling between them. Reason for Using the Free Space Far Field Model In the free space far field model. the total EM signal can be obtained by adding the signal generated by each dipole. In practice. The Far-field Restriction The far field area is usually defined by the area beyond a distance related to antenna size D : d far – field = 2 When the largest dimension D of the antenna is less than the wave length (  ). diffraction. and not in 3D horizontal and vertical planes. and reflection phenomena decrease the signal strength very rapidly. They model complex phenomena such as diffraction. ACP uses a simple propagation model dedicated to cope with the requirements of EMF exposure evaluation. 2 d far – field =  2D    When the largest dimension D of the antenna is greater than the wave length (  ).e. multipath transmission.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 30 EIRP E = ---------------------d Where EIRP = P tx G tx and P tx is the transmitted power in Watts and G tx is the antenna gain. either deterministically (ray tracing) or empirically. The far field starts at around 10 to 20 meters from the antenna. the relationship between the EMF exposure level (in dBV⁄m) and the received signal power level (in dBm) when considering a receiving antenna with 0 dB gain at frequency F (in MHz) is: P rx  dBm  = E  dB V  m  + 42.8 – 20log 10 F  MHz  For frequencies around 1 GHz.1 V⁄m) corresponds to a received signal level greater than -37 dBm. then the ACP uses a point-to-point calculation instead of full path loss matrices (i. If the number of points distributed for a given height is low. etc. The ACP then calculates the angles of incidence which are used for the antenna masking method. The ACP stores the path loss attenuation to the multi-storey evaluation points in the ACP storage directory. On further ACP runs. When indoor coverage is used. if a pixel presents a weight of 1 and a total of 5 points at that location (one point at ground level and four additional points. or multiplied by the number of vertical points at that pixel. when using a propagation model such as Crosswave). The ACP then proceeds with its optimisation algorithm as usual. values seen at a given position: When several points are present for a pixel (1 ground level point + one or several multi-storey points at different heights). and in that case reuses them if possible.9.9. but it is possible to specify which clutter classes should be considered as indoors. each point either takes a weight of 0.2 Pixel Weighting The total weight associated with an x/y pixel (derived from traffic and zone weighting) is either shared equally among all vertical evaluation points present at this pixel. all pixels are considered as being indoors.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12. Three options are provided: • • • Display at ground level (defined receiver height): Only the prediction values seen at the ground pixels are shown.. 12.2 (when vertical weight sharing is selected) or 1 (when vertical weight sharing is not selected).e. for example. This margin usually depends on: • • Cell coverage probability. However all predictions provided by the ACP relate only to the ground layer by default. in calculating the macrodiversity gain in UMTS. A new tab is available in the properties of the prediction to show results at different heights.8 Shadowing Margin and Indoor Coverage Atoll ACP supports both indoor coverage and a shadowing margin. 12.9. 12. The ACP first detects if path loss matrices created by the Atoll MultiStorey Add-in are present.3. it uses API CalculateSubscriber instead of CalculateGrid). Display values at given storey: Only the points at the given storey are displayed Example: 865 . the minimum value of those points is shown. If matrices are not present. even after modifying parameters such as resolution. it recalculates path loss matrices itself for different heights. All 3-D points participate in the objective optimisation. through the weighting process described earlier. The shadowing margin is applied in the same way as it is in Atoll coverage predictions. By default. When the shadowing margin has been enabled. This enables the ACP to take all floors of the building into account during optimisation.1 Path Loss Calculation and Data Caching The ACP multi-storey extension calculates the path losses from each transmitter to the points distributed in three dimensions by calling certain methods of Atoll's API on the propagation model (CalculateSubscriber and CalculateGrid). one every 10 meters). when using "delegation to the propagation model" (for example. a shadowing margin is added to the basic quality measurement.9 Multi-Storey Optimisation The ACP includes a multi-storey extension where evaluation points are distributed on all floors of buildings defined in a clutter heights map. For example. The model standard deviation which is clutter-dependent and defined separately for different quality measures. Display min. there is usually little or no need for path loss recalculation. The calculation method depends on the propagation model: • • Direct calculation at the required height when not using "delegation to the propagation model" Angle estimation from the original angle of incidence calculated at ground level and taking into account geometrical considerations. an additional indoor loss is added to all pixels marked as being indoors.3 Results All statistical results provided take into account both 2-D and 3-D points. This indoor loss is clutter-dependent. according to the log normal distribution function.Atoll 3. 12. Viewing detailed results for 3-D points is done by creating Quality Analysis and Objective Analysis predictions in the ACP. The following default propagation classes are provided: • • • Open: The Open propagation class is for areas without obstacles. or transparent. Building: The Building propagation class is used for opaque objects such as buildings. specified by the "Linear loss start distance (m)" parameter.onlyDHM=true • • The actual heights used for multi-storey evaluation depend on the receiver height defined in Atoll. meaning that the signal experiences diffraction losses at the edge of the object but does not go completely through.5m.3. Vegetation: The Vegetation propagation class is used for areas covered with vegetation. habitation versus monuments). 13. then the actual heights are 1. The loss is applied only after a given number of meters. and a systematic method (CalculateSubscribers) for the upper storey where few evaluation points are present. If a receiver height of 1. It is possible to define new propagation classes. For a two-dimensional distribution. An open area can also be an elevated area such as a bridge. This is the same as the process used by the Multi-storey Prediction add-in. stone buildings. the points can be placed either at the bottom (for example. Linear loss (dB/m): A linear loss applied for each meter within an object.4 Notes • ACP distributes multi-storey evaluation points only where clutter heights are present. 866 . no point is created using the default clutter class height when only a clutter class file is present. user-defined classes are always of the type "opaque". in a park) or at the top (for example.9. Currently. Such areas are transparent. 12.9. glass buildings.5 Concepts of ACP EMF Exposure 12. with no diffraction loss. meaning that the signal passes through it (with perhaps some losses) and does not experience diffraction loss. Atoll uses in general a "radial" method. The measurement points can be distributed in either a 3D pattern or in a 2D pattern. To enable taking the default clutter class height into account.5.) or for differentiating items on which EMF evaluation should be done (for example. wood constructions. you can define the following option in the ACP.5 m is used.5m. etc. for example to differentiate between similar items with different penetration loss characteristics (for example. such as parks.5m. and a vertical step of 2 storeys (with a storey defined in this example as being 3 m).9. Distribution of evaluation points: Field strength measurements are made on a set of points within an object.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module Signal Level at ground level ©Forsk 2015 Signal Level at 3rd storey 12. The propagation classes have the following parameters: • • • Penetration loss (dB): The loss occurring when the signal enters the object.INI file: [ACPCore] multistorey. etc.Atoll 3. Each propagation class is either opaque. They can be considered as transparent but with a certain degree of diffraction loss. 7.1 Propagation Classes The internal propagation model calculates EMF exposure using propagation classes which are retrieved from input files. for a bridge) to better reflect where people will be. Predictions calculated by the ACP might differ slightly from multi-storey predictions due to different methodologies used by the two tools: • • The ACP uses a mix of a radial method (CalculateGrid) for lower storeys. such as an open area or water. The signal experiences some loss when going through and also suffer from diffraction loss. By default. 5. The ACP also takes into consideration the antenna radiation pattern by creating a 3D interpolation from the 2D horizontal and vertical cross-sections of the antenna radiation pattern. open spaces.9.9.Atoll 3. 12.9. If a geo vector contains more than one polygon.5.6 General Workflow ACP creates a representation of the terrain in 3D. the ACP does not need any specific terrain modelling. roads.9..g. Geo raster data are only used for the areas not covered by geo vector data. clutter classes representing the type of terrain and clutter heights (also called a digital height model) representing individual heights (for example.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module AT330_TRR_E1 12. This ensures that all provided results are for a fully loaded network. where any opaque propagation class such as Building becomes fully transparent to electromagnetic waves. For areas covered both by vector data and raster data. 12. For each pixel in this raster representation. 12.g. If no clutter height file is present. The height is obtained from the clutter height raster file. Building: For the Building class evaluation points are distributed in 3D.5. This mode is useful when optimising the networks while ensuring that regulatory limits are never exceeded. in the real world. 12.) evaluation points are distributed in 2D on the top of the class height. The method is similar to the one used elsewhere in Atoll. It then distributes a number of evaluation points in this representation according to the parameters set in the setup. both the height and propagation class information are encoded: • • For geo vectors. etc. Similarly you distribute evaluation points on only selected clutter classes. Evaluation points are distributed either in a 2D or 3D pattern depending on the propagation class: • • • Open: For the Open class (e. The EMF exposure is therefore evaluated as if the area was completely free space.3 Distribution of Evaluation Points ACP uses the internal terrain representation to specify where to set evaluation points for EMF exposure evaluation. 867 . This mode gives you a pessimistic view of the actual exposure since. each clutter class is associated with a single propagation class.3.5 Worst-case Mode The ACP allows you to consider a worst case scenario. In this case it is also possible to specify the maximum indoor distance on which to measure EMF exposure. penetration loss through obstacles as well as diffraction and reflection around obstacles tends to strongly decrease the signal strength compared to a completely free space model. It is recommended to always provide either geo vector data or clutter heights raster data to have the most accurate EMF exposure prediction. building heights). parks. It is possible to distribute evaluation points separately on each propagation class and for each terrain entry. ACP uses the associated DBF file to map the polygons to propagation classes and heights. bridges.9.4 The Contribution of Transmitter Power to EMF Exposure The ACP takes the maximum power transmission on all the carriers and channels used by transmitters in the network into consideration in its calculations.5. the default clutter class height is used. etc. and or to restrict the prediction on the building front façade. but instead it uses all the geo data available in the Atoll project: • • Geo raster data: Raster data give a grid-based representation of the terrain with a defined resolution. each polygon is associated with a single propagation class and a height. For geo rasters. Geo vector data: Geo vector data model the buildings and their height. For example. Vegetation: For the Vegetation class (e. but not on another vector entry. ACP uses the geo data to create a 3D representation of the terrain in the form of a fine raster of pixels with a default resolution of 2 meters. you can distribute evaluation points on one geo vector entry for one subset of polygons.2 Terrain Profile To measure EMF exposure.) evaluation points are distributed in 2D at the bottom of the class height. thereby giving a worst case calculation of EMF exposure. forests. in the form of one or several ArcView SHP files defining numerous polygons. The raster files needed are DTM (Digital Terrain Model) representing the ground altitude. only the geo vector data are used.. The ACP manages cases of data mismatch by using the concept of a locked setup node. It will be unlocked if the Atoll project is rolled back to its initial state. A new setup needs to be created in order for the changes to be taken into account. as well as some diffraction and indoor losses. after modifying transmitters directly in Atoll). The data model also contains information identifying the version used to generate it. • 868 Atoll ACP enables you to duplicate an existing setup node while at the same time updating its internal data model to be consistent with the current state of the Atoll project. Setup: The ACP Setup dialogue allows you to view and modify the optimisation parameters. If changes are introduced into the Atoll database later (such as changes to the antennas. This data model is saved in a "Setup" node and enables each optimisation setup to be reviewed or replayed later. 12. but is instead accessed directly by the core optimisation engine. nor does it contain the path losses matrices.Automatic Cell Planning context menu). DHM maps. Because the path loss information is not stored in the setup node. Atoll ACP produces a result model which is found under the original setup in an Optimisation node. The data model is not accessible using the Setup dialogue. cells. . it can happen that there is a mismatch between stored path loss matrices and the data model in the setup node (for example. cells. It works using the extracted internal data model in the Setup node. you can view the results and generate analysis maps that can be displayed directly in the Atoll map window. meaning that the results produced by a previous release can in general be reloaded or replayed. etc. and service definitions. This behaviour is particularly true when new settings produced by an optimisation run are committed to Atoll. UMTS. The raster data and path loss matrices are accessed directly by the core optimisation engine during ACP calculations.10 ACP Software Data Flow Understanding the Atoll ACP software data flow will help understand how the module works and some of its internal constraints.3. the ACP extracts all relevant information from the current Atoll project and builds its internal data model. Using the Optimisation node. site. these changes are not taken into account in any existing setup node. transmitters. DEM. This modelling is based on free space formulas which are applicable starting at a few meters from the antennas up to a few hundred meters. it is important to understand that: • • • An optimisation runs on the data model stored in the setup node. You can also commit the set of recommended changes directly into the Atoll database. Optimisation Engine: The optimisation engine is the core algorithm that performs the optimisation on a defined setup. WiMAX) and considering all carriers and channels used. The data model does not include raster information such as clutter. The overall EMF exposure calculation is obtained by adding the electromagnetic signal level generated by each technology involved (GSM. The network configuration is essentially frozen in the setup node in the state it was in when the setup was created. LTE..e. No optimisation can be run on a locked setup node unless the path loss information is consistent with the internal data model of the setup. but also uses direct access to raster and path loss information. when the user selects New from the ACP . Setup nodes are automatically unlocked when the path loss information and the internal data model once again match. Results: After an optimisation run.). Here are some of the concepts related to the data flow: • • • • • Data Model Extraction: When first run (i.Atoll 3.0 Technical Reference Guide for Radio Networks Chapter 12: ACP Module ©Forsk 2015 ACP evaluates EMF exposure on each of these evaluation points using an integrated propagation model specially tailored for the evaluation of strong signals in the vicinity of the antennas. sites. CDMA. Data Model Content: The data model includes all necessary data from the Atoll database. It takes into account the antenna gain and attenuation patterns towards each evaluation point. Because Atoll ACP uses this internal data model. essentially all antennas. The setup node is locked after a commit. 3.0 Technical Reference Guide for Radio Networks 869 .AT330_TRR_E1 Atoll 3. • Head Office 7 rue des Briquetiers 31700 Blagnac. France Tel: +33 562 747 210 Fax: +33 562 747 211 AT330_TRR_E1 • US Office • China Office 200 South Wacker Drive – Suite 3100 Chicago. 66 Jianzhong Road. Tianhe Hi-Tech Industrial Zone. West Tower. IL 60606. R. 3/F. of China Tel: +86 20 8553 8938 Fax: +86 20 8553 8285 www. USA Tel: +1 312 674 4800 Fax: +1 312 674 4847 Suite 302. No. P.forsk.com March 2015 . Guangzhou. Jiadu Commercial Building. 510665.