Technical Reference Guide E6 PDF

Atoll
RF Planning and Optimisation Software
Version 2.7.1
Technical
Reference
Guide
AT271_TRG_E6
Technical Reference Guide
Contact Information
Forsk (Head Office) 7 rue des Briquetiers  www.forsk.com Web

31700 Blagnac  sales@forsk.com Sales and pricing information
France  helpdesk@forsk.com Technical support
 +33 (0) 562 74 72 10 General
+33 (0) 562 74 72 25 Technical support
 +33 (0) 562 74 72 11 Fax
Forsk (USA Office) 200 South Wacker Drive  sales_us@forsk.com Sales and pricing information
Suite 3100  support_us@forsk.com Technical support
Chicago, IL 60606  +1 312 674 4846 General
USA +1 888 GoAtoll (+1 888 462 8655) Technical support
 +1 312 674 4847 Fax
Forsk (China Office) Suite 302, 3/F, West Tower,  www.forsk.com.cn Web
Jiadu Commercial Building,  enquiries@forsk.com.cn Information and enquiries
No.66 Jianzhong Road,  +86 20 8553 8938 Telephone
Tianhe Hi-Tech Industrial Zone,  +86 20 8553 8285 Fax (Guangzhou)
Guangzhou, 510665, +86 10 6513 4559 Fax (Beijing)
People’s Republic of China
Atoll 2.7.1 Technical Reference Guide Release AT271_TRG_E6
© Copyright 1997 - 2009 by Forsk

The software described in this document is provided under a licence agreement. The software may only be used/copied
under the terms and conditions of the licence agreement. No part of this document may be copied, reproduced or
distributed in any form without prior authorisation from Forsk.
The product or brand names mentioned in this document are trademarks or registered trademarks of their respective
registering parties.
About Technical Reference Guide

This document is targeted at readers with a prior knowledge of Atoll, its operation and basic functioning. It is not the User
Manual for Atoll, and does not teach how to operate and use Atoll. It is a supplementary document containing detailed
descriptions of models, algorithms and concepts adopted in Atoll. Therefore, it concerns only the appropriate personnel.
Atoll Technical Reference Guide is divided into three parts with each part comprising similar topics. The first part contains
descriptions of general terms, entities, ideas and concepts in Atoll that are encountered throughout its use. It is followed
by the second part that consists of descriptions of entities common to all types of networks and the algorithms that are
technology independent and are available in any network type. Lastly, the guide provides detailed descriptions of each
basic type of network that can be modelled and studied in Atoll.
© Forsk 2009 AT271_TRG_E6 iii

iv AT271_TRG_E6 © Forsk 2009

Table of Contents
Table of Contents
1 Coordinate Systems and Units ....................................................... 25

1.1 Coordinate Systems............................................................................................................................... 25
1.1.1 Description of Coordinate Systems .................................................................................................. 25
1.1.1.1 Geographic Coordinate System.................................................................................................. 25
1.1.1.2 Datum ......................................................................................................................................... 25
1.1.1.3 Meridian ...................................................................................................................................... 25
1.1.1.4 Ellipsoid ...................................................................................................................................... 25
1.1.1.5 Projection.................................................................................................................................... 26
1.1.1.6 Projection Coordinate System .................................................................................................... 26
1.1.2 Coordinate Systems in Atoll ............................................................................................................. 26
1.1.2.1 Projection Coordinate System .................................................................................................... 26
1.1.2.2 Display Coordinate System ........................................................................................................ 26
1.1.2.3 Internal Coordinate Systems ...................................................................................................... 26
1.1.3 File Formats ..................................................................................................................................... 27
1.1.3.1 Unit Codes .................................................................................................................................. 27
1.1.3.2 Datum Codes.............................................................................................................................. 28
1.1.3.3 Projection Method Codes ........................................................................................................... 29
1.1.3.4 Ellipsoid Codes ........................................................................................................................... 29
1.1.3.5 Projection Parameter Indices...................................................................................................... 30
1.1.4 Creating a Coordinate System ......................................................................................................... 30
1.2 Units ....................................................................................................................................................... 30
1.2.1 Power Units ...................................................................................................................................... 30
1.2.2 Length Units ..................................................................................................................................... 31
1.3 BSIC Format .......................................................................................................................................... 31
2 Geographic and Radio Data ........................................................... 35

2.1 Geographic Data .................................................................................................................................... 35
2.1.1 Data Type......................................................................................................................................... 35
2.1.1.1 Digital Terrain Model (DTM) ....................................................................................................... 35
2.1.1.2 Clutter (Land Use) ...................................................................................................................... 36
2.1.1.2.1 Clutter Classes...................................................................................................................... 36
2.1.1.2.2 Clutter Heights ...................................................................................................................... 36
2.1.1.3 Traffic Data ................................................................................................................................. 36
2.1.1.3.1 Environment Traffic Maps ..................................................................................................... 36
2.1.1.3.2 User Profile Traffic Maps....................................................................................................... 36
2.1.1.3.3 Live Traffic Maps................................................................................................................... 36
2.1.1.3.4 User Density Traffic Maps..................................................................................................... 37
2.1.1.4 Vector Data................................................................................................................................. 37
2.1.1.5 Scanned Images......................................................................................................................... 37
2.1.1.6 Population................................................................................................................................... 37
2.1.1.7 Other Geographic Data............................................................................................................... 37
2.1.2 Supported Geographic Data Formats .............................................................................................. 37
2.2 Radio Data ............................................................................................................................................. 38
2.2.1 Site ................................................................................................................................................... 38
2.2.2 Antenna ............................................................................................................................................ 38
2.2.3 Transmitter ....................................................................................................................................... 38
2.2.4 Repeater........................................................................................................................................... 38
2.2.5 Remote Antenna .............................................................................................................................. 39
2.2.6 Station .............................................................................................................................................. 39
2.2.7 Hexagonal Design ............................................................................................................................ 39
2.2.8 GSM GPRS EGPRS Documents ..................................................................................................... 39
2.2.8.1 TRX............................................................................................................................................. 39
2.2.8.2 Subcell ........................................................................................................................................ 39
2.2.8.3 Cell Type..................................................................................................................................... 39
2.2.9 All CDMA and WiMAX BWA Documents ......................................................................................... 39
2.2.9.1 Cell.............................................................................................................................................. 39
3 File Formats .................................................................................... 43

3.1 BIL Format ............................................................................................................................................. 43
© Forsk 2009 AT271_TRG_E6 v

3.1.1 HDR Header File...............................................................................................................................43

3.1.1.1 Description ..................................................................................................................................43
3.1.1.2 Samples ......................................................................................................................................44
3.1.1.2.1 Digital Terrain Model..............................................................................................................44
3.1.1.2.2 Clutter Classes File................................................................................................................44
3.1.1.2.3 BIL File...................................................................................................................................44
3.2 TIF Format ..............................................................................................................................................45
3.2.1 TFW Header File...............................................................................................................................45
3.2.2 Sample ..............................................................................................................................................46
3.2.2.1 Clutter Classes File .....................................................................................................................46
3.3 BMP Format............................................................................................................................................46
3.3.1 BMP File Description.........................................................................................................................46
3.3.1.1 BMP File Structure ......................................................................................................................46
3.3.1.2 BMP Raster Data Encoding ........................................................................................................48
3.3.1.2.1 Raster Data Compression Descriptions.................................................................................48
3.3.2 BPW/BMW Header File Description..................................................................................................49
3.3.3 Sample ..............................................................................................................................................49
3.4 Generic Raster Header File (.wld) ..........................................................................................................49
3.4.1 WLD File Description ........................................................................................................................49
3.4.2 Sample ..............................................................................................................................................49
3.5 DXF Format ............................................................................................................................................50
3.6 SHP Format ............................................................................................................................................50
3.7 MIF Format .............................................................................................................................................50
3.8 TAB Format ............................................................................................................................................50
3.9 ECW Format ...........................................................................................................................................51
3.10 Erdas Imagine Format ............................................................................................................................51
3.11 Planet EV/Vertical Mapper Geographic Data Format .............................................................................51
3.12 ArcView Grid Format ..............................................................................................................................52
3.12.1 ArcView Grid File Description ...........................................................................................................52
3.12.2 Sample ..............................................................................................................................................52
3.13 Other Supported Geographic Data File Formats ....................................................................................52
3.14 Planet Format .........................................................................................................................................52
3.14.1 DTM File............................................................................................................................................53
3.14.1.1 Description ..................................................................................................................................53
3.14.1.2 Sample ........................................................................................................................................53
3.14.2 Clutter Class Files .............................................................................................................................53
3.14.2.1 Description ..................................................................................................................................53
3.14.2.2 Sample ........................................................................................................................................54
3.14.3 Vector Files .......................................................................................................................................54
3.14.3.1 Description ..................................................................................................................................54
3.14.3.2 Sample ........................................................................................................................................55
3.14.4 Image Files........................................................................................................................................55
3.14.5 Text Data Files ..................................................................................................................................55
3.15 MNU Format ...........................................................................................................................................56
3.15.1 Description ........................................................................................................................................56
3.15.2 Sample ..............................................................................................................................................56
3.16 XML Table Export/Import Format ...........................................................................................................56
3.16.1 Index.xml File ....................................................................................................................................56
3.16.2 XML File ............................................................................................................................................57
3.17 Externalised Propagation Results Format ..............................................................................................58
3.17.1 DBF File ............................................................................................................................................58
3.17.1.1 DBF File Format ..........................................................................................................................58
3.17.1.1.1 DBF Structure ........................................................................................................................58
3.17.1.1.2 DBF Header (Variable Size - Depends on Field Count) ........................................................58
3.17.1.1.3 Each DBF Record (Fixed Length)..........................................................................................60
3.17.1.2 DBF File Content.........................................................................................................................60
3.17.2 LOS File ............................................................................................................................................61
3.18 Externalised Tuning Files .......................................................................................................................61
3.18.1 DBF File ............................................................................................................................................61
3.18.1.1 DBF File Format ..........................................................................................................................61
3.18.1.1.1 DBF Structure ........................................................................................................................62
3.18.1.1.2 DBF Header (Variable Size - Depends on Field Count) ........................................................62
3.18.1.1.3 Each DBF Record (Fixed Length)..........................................................................................63
3.18.1.2 DBF File Content.........................................................................................................................63
3.18.2 PTS File ............................................................................................................................................64
3.19 Interference Histograms File Formats ....................................................................................................64
vi AT271_TRG_E6 © Forsk 2009

Table of Contents
3.19.1 One Histogram per Line (.im0) Format............................................................................................. 64

3.19.1.1 Sample........................................................................................................................................ 65
3.19.2 One Value per Line with Dictionary File (.clc) Format ...................................................................... 65
3.19.2.1 CLC File...................................................................................................................................... 66
3.19.2.1.1 Description ............................................................................................................................ 66
3.19.2.1.2 Sample .................................................................................................................................. 66
3.19.2.2 DCT File...................................................................................................................................... 67
3.19.2.2.1 Description ............................................................................................................................ 67
3.19.2.2.2 Sample .................................................................................................................................. 67
3.19.3 One Value per Line (Transmitter Name Repeated) (.im1) Format ................................................... 68
3.19.3.1 Sample........................................................................................................................................ 68
3.19.4 Only Co-Channel and Adjacent Values (.im2) Format ..................................................................... 69
3.19.4.1 Sample........................................................................................................................................ 69
4 Calculations .................................................................................... 73
4.1 Overview ................................................................................................................................................ 73
4.2 Path Loss Matrices................................................................................................................................. 74
4.2.1 Calculation Area Determination........................................................................................................ 75
4.2.1.1 Computation Zone ...................................................................................................................... 75
4.2.1.2 Use of Polygonal Zones in Coverage Prediction Reports........................................................... 75
4.2.2 Calculate / Force Calculation Comparison ....................................................................................... 76
4.2.2.1 Calculate..................................................................................................................................... 76
4.2.2.2 Force Calculation........................................................................................................................ 76
4.2.3 Matrix Validity ................................................................................................................................... 76
4.3 Path Loss Calculations........................................................................................................................... 77
4.3.1 Ground Altitude Determination ......................................................................................................... 77
4.3.2 Clutter Determination ....................................................................................................................... 78
4.3.2.1 Clutter Class ............................................................................................................................... 78
4.3.2.2 Clutter Height.............................................................................................................................. 78
4.3.3 Geographic Profile Extraction........................................................................................................... 78
4.3.3.1 Extraction Methods ..................................................................................................................... 78
4.3.3.1.1 Radial Extraction ................................................................................................................... 78
4.3.3.1.2 Systematic Extraction ........................................................................................................... 79
4.3.3.2 Profile Resolution: Multi-Resolution Management...................................................................... 80
4.4 Propagation Models ............................................................................................................................... 82
4.4.1 Okumura-Hata and Cost-Hata Propagation Models......................................................................... 83
4.4.1.1 Hata Path Loss Formula ............................................................................................................. 83
4.4.1.2 Corrections to the Hata Path Loss Formula................................................................................ 83
4.4.1.3 Calculations in Atoll .................................................................................................................... 83
4.4.2 ITU 529-3 Propagation Model .......................................................................................................... 84
4.4.2.1 ITU 529-3 Path Loss Formula..................................................................................................... 84
4.4.2.2 Corrections to the ITU 529-3 Path Loss Formula ....................................................................... 84
4.4.2.2.1 Environment Correction ........................................................................................................ 84
4.4.2.2.2 Area Size Correction ............................................................................................................. 84
4.4.2.2.3 Distance Correction .............................................................................................................. 85
4.4.3 Standard Propagation Model (SPM) ................................................................................................ 85
4.4.3.1 SPM Path Loss Formula............................................................................................................. 85
4.4.3.2.1 Visibility and Distance Between Transmitter and Receiver................................................... 86
4.4.3.2.2 Effective Transmitter Antenna Height ................................................................................... 86
4.4.3.2.3 Effective Receiver Antenna Height ....................................................................................... 89
4.4.3.2.4 Correction for Hilly Regions in Case of LOS ......................................................................... 89
4.4.3.2.5 Diffraction .............................................................................................................................. 90
4.4.3.2.6 Losses due to Clutter ............................................................................................................ 90
4.4.3.2.7 Recommendations ................................................................................................................ 91
4.4.3.3 Automatic SPM Calibration......................................................................................................... 91
4.4.3.3.1 General Algorithm ................................................................................................................. 92
4.4.3.3.2 Sample Values for SPM Path Loss Formula Parameters ..................................................... 92
4.4.3.4 Unmasked Path Loss Calculation............................................................................................... 93
4.4.4 WLL Propagation Model ................................................................................................................... 94
4.4.4.1 WLL Path Loss Formula ............................................................................................................. 94
4.4.4.2.1 Free Space Loss ................................................................................................................... 94
4.4.4.2.2 Diffraction .............................................................................................................................. 94
4.4.5 ITU-R P.526-5 Propagation Model ................................................................................................... 95
4.4.5.1 ITU 526-5 Path Loss Formula..................................................................................................... 95
© Forsk 2009 AT271_TRG_E6 vii

4.4.5.2.1 Free Space Loss....................................................................................................................95

4.4.5.2.2 Diffraction...............................................................................................................................95
4.4.6 ITU-R P.370-7 Propagation Model....................................................................................................95
4.4.6.1 ITU 370-7 Path Loss Formula .....................................................................................................95
4.4.6.2 Calculations in Atoll .....................................................................................................................95
4.4.6.2.1 Free Space Loss....................................................................................................................95
4.4.6.2.2 Corrected Standard Loss.......................................................................................................95
4.4.7 Erceg-Greenstein (SUI) Propagation Model .....................................................................................96
4.4.7.1 SUI Terrain Types .......................................................................................................................97
4.4.7.2 Erceg-Greenstein (SUI) Path Loss Formula................................................................................97
4.4.8 ITU-R P.1546-2 Propagation Model..................................................................................................98
4.4.8.1.1 Step 1: Determination of Graphs to be Used.........................................................................99
4.4.8.1.2 Step 2: Calculation of Maximum Field Strength.....................................................................99
4.4.8.1.3 Step 3: Determination of Transmitter Antenna Height ...........................................................99
4.4.8.1.4 Step 4: Interpolation/Extrapolation of Field Strength .............................................................99
4.4.8.1.5 Step 5: Calculation of Correction Factors ............................................................................101
4.4.8.1.6 Step 6: Calculation of Path Loss..........................................................................................102
4.4.9 Sakagami Extended Propagation Model.........................................................................................103
4.4.10 Appendices .....................................................................................................................................104
4.4.10.1 Free Space Loss .......................................................................................................................104
4.4.10.2 Diffraction Loss..........................................................................................................................104
4.4.10.2.1 Knife-Edge Diffraction..........................................................................................................104
4.4.10.2.2 3 Knife-Edge Deygout Method.............................................................................................105
4.4.10.2.3 Epstein-Peterson Method ....................................................................................................107
4.4.10.2.4 Deygout Method with Correction .........................................................................................107
4.4.10.2.5 Millington Method.................................................................................................................107
4.5 Path Loss Tuning..................................................................................................................................108
4.6 Antenna Attenuation Calculation ..........................................................................................................109
4.6.1 Calculation of Azimuth and Tilt Angles............................................................................................109
4.6.2 Antenna Pattern 3-D Interpolation...................................................................................................110
4.6.3 Additional Electrical Downtilt Modelling...........................................................................................111
4.6.4 Antenna Pattern Smoothing ............................................................................................................111
4.6.4.1 Smoothing Algorithm .................................................................................................................113
4.7 Shadowing Model .................................................................................................................................113
4.7.1 Shadowing Margin Calculation........................................................................................................116
4.7.1.1 Shadowing Margin Calculation in Predictions ...........................................................................116
4.7.1.2 Shadowing Margin Calculation in Monte-Carlo Simulations......................................................117
4.7.2 Macro-Diversity Gains Calculation ..................................................................................................118
4.7.2.1 Uplink Macro-Diversity Gain Evaluation ....................................................................................118
4.7.2.1.1 Shadowing Error PDF (n Signals)........................................................................................118
4.7.2.1.2 Uplink Macro-Diversity Gain ................................................................................................121
4.7.2.2 Downlink Macro-Diversity Gain Evaluation ...............................................................................121
4.7.2.2.1 Shadowing Error PDF (n Signals)........................................................................................121
4.7.2.2.2 Downlink Macro-Diversity Gain............................................................................................123
4.8 Appendices ...........................................................................................................................................124
4.8.1 Transmitter Radio Equipment .........................................................................................................124
4.8.1.1 UMTS HSPA, CDMA2000 1xRTT 1xEV-DO, IS-95 cdmaOne, and
TD-SCDMA Documents124
4.8.1.2 GSM GPRS EGPRS Documents ..............................................................................................125
4.8.1.3 WiMAX 802.16d and WiMAX 802.16e Documents ...................................................................126
4.8.2 Secondary Antennas.......................................................................................................................127
5 GSM GPRS EGPRS Networks......................................................131

5.1 General Prediction Studies ...................................................................................................................131
5.1.1 Calculation Criteria..........................................................................................................................131
5.1.2 Point Analysis..................................................................................................................................131
5.1.2.1 Profile Tab .................................................................................................................................131
5.1.2.2 Reception Tab ...........................................................................................................................131
5.1.3 Coverage Studies............................................................................................................................132
5.1.3.1 Service Area Determination ......................................................................................................132
5.1.3.1.1 All Servers ...........................................................................................................................132
5.1.3.1.2 Best Signal Level and a Margin ...........................................................................................132
5.1.3.1.3 Second Best Signal Level and a Margin..............................................................................132
5.1.3.1.4 Best Signal Level per HCS Layer and a Margin ..................................................................133
5.1.3.1.5 HCS Servers and a Margin..................................................................................................133
5.1.3.1.6 Highest Priority HCS Server and a Margin ..........................................................................133
viii AT271_TRG_E6 © Forsk 2009

Table of Contents
5.1.3.1.7 Second Best Signal Level per HCS Layer and a Margin .................................................... 134
5.1.3.1.8 Best Idle Mode Reselection Criterion (C2).......................................................................... 134
5.1.3.2 Coverage Display ..................................................................................................................... 135
5.1.3.2.1 Plot Resolution .................................................................................................................... 135
5.1.3.2.2 Display Types ..................................................................................................................... 135
5.2 Traffic Analysis..................................................................................................................................... 136
5.2.1 Traffic Distribution .......................................................................................................................... 136
5.2.1.1 Normal Cells (Nonconcentric, No HCS Layer).......................................................................... 136
5.2.1.1.1 Circuit Switched Services.................................................................................................... 136
5.2.1.1.2 Packet Switched Services................................................................................................... 136
5.2.1.2 Concentric Cells........................................................................................................................ 136
5.2.1.3 HCS Layers .............................................................................................................................. 136
5.2.2 Calculation of the Traffic Demand per Subcell ............................................................................... 137
5.2.2.1 Traffic Maps Based on Environments and User Profiles .......................................................... 137
5.2.2.1.1 Normal Cells (Nonconcentric, No HCS Layer) .................................................................... 137
5.2.2.1.2 Concentric Cells .................................................................................................................. 137
5.2.2.1.3 HCS Layers......................................................................................................................... 138
5.2.2.2 Traffic Maps Based on Transmitters and Services ................................................................... 142
5.2.2.2.1 Normal Cells (Nonconcentric, No HCS Layer) .................................................................... 142
5.2.2.2.2 Concentric Cells .................................................................................................................. 142
5.2.2.2.3 HCS Layers......................................................................................................................... 143
5.3 Network Dimensioning ......................................................................................................................... 146
5.3.1 Dimensioning Models and Quality Graphs ..................................................................................... 146
5.3.1.1 Circuit Switched Traffic ............................................................................................................. 146
5.3.1.2 Packet Switched Traffic ............................................................................................................ 146
5.3.1.2.1 Throughput.......................................................................................................................... 146
5.3.1.2.2 Delay ................................................................................................................................... 149
5.3.1.2.3 Blocking Probability............................................................................................................. 149
5.3.2 Network Dimensioning Process ..................................................................................................... 150
5.3.2.1 Network Dimensioning Engine.................................................................................................. 150
5.3.2.1.1 Inputs .................................................................................................................................. 150
5.3.2.1.2 Outputs ............................................................................................................................... 151
5.3.2.2 Network Dimensioning Steps.................................................................................................... 151
5.3.2.2.1 Step 1: Timeslots Required for CS Traffic........................................................................... 151
5.3.2.2.2 Step 2: TRXs Required for CS Traffic and Dedicated PS Timeslots................................... 151
5.3.2.2.3 Step 3: Effective CS Blocking, Effective CS Traffic Overflow and Served CS Traffic ......... 152
5.3.2.2.4 Step 4: TRXs to Add for PS Traffic ..................................................................................... 152
5.3.2.2.5 Step 5: Served PS Traffic ................................................................................................... 154
5.3.2.2.6 Step 6: Total Traffic Load.................................................................................................... 154
5.4 Key Performance Indicators Calculation .............................................................................................. 154
5.4.1 Circuit Switched Traffic................................................................................................................... 155
5.4.1.1 Erlang B .................................................................................................................................... 155
5.4.1.2 Erlang C.................................................................................................................................... 155
5.4.1.3 Served Circuit Switched Traffic................................................................................................. 155
5.4.2 Packet Switched Traffic .................................................................................................................. 155
5.4.2.1 Case 1: Total Traffic Demand > Dedicated + Shared Timeslots .............................................. 155
5.4.2.1.1 Traffic Load ......................................................................................................................... 155
5.4.2.1.2 Packet Switched Traffic Overflow ....................................................................................... 156
5.4.2.1.3 Throughput Reduction Factor ............................................................................................. 156
5.4.2.1.4 Delay ................................................................................................................................... 156
5.4.2.1.6 Served Packet Switched Traffic .......................................................................................... 156
5.4.2.2 Case 2: Total Traffic Demand < Dedicated + Shared Timeslots .............................................. 156
5.4.2.2.1 Traffic Load ......................................................................................................................... 156
5.4.2.2.2 Packet Switched Traffic Overflow ....................................................................................... 156
5.4.2.2.3 Throughput Reduction Factor ............................................................................................. 156
5.4.2.2.4 Delay ................................................................................................................................... 156
5.4.2.2.6 Served Packet Switched Traffic .......................................................................................... 157
5.5 Neighbour Allocation ............................................................................................................................ 157
5.5.1 Global Allocation for All Transmitters ............................................................................................. 157
5.5.2 Allocation for a Group of Transmitters or One Transmitter ............................................................ 160
5.6 Interference Prediction Studies ............................................................................................................ 160
5.6.1 Coverage Studies ........................................................................................................................... 160
© Forsk 2009 AT271_TRG_E6 ix

5.6.1.1 Service Area Determination ......................................................................................................161

5.6.1.1.1 All Servers ...........................................................................................................................161
5.6.1.1.2 Best Signal Level per HCS Layer and a Margin ..................................................................161
5.6.1.1.3 Best Signal Level of the Highest Priority HCS Layer and a Margin .....................................161
5.6.1.1.4 Second Best Signal Level per HCS Layer and a Margin .....................................................161
5.6.1.2 Carrier to Interference Ratio Calculation ...................................................................................161
5.6.1.2.1 Carrier Power Level .............................................................................................................162
5.6.1.2.2 Interference Calculation.......................................................................................................163
5.6.1.2.3 Collision Probability for Non Hopping Mode ........................................................................164
5.6.1.2.4 Collision Probability for BBH and SFH Modes.....................................................................164
5.6.1.3 Coverage Area Determination...................................................................................................165
5.6.1.3.1 Interference Condition Satisfied by At Least One TRX........................................................165
5.6.1.3.2 Interference Condition Satisfied by The Worst TRX ............................................................165
5.6.1.4 Coverage Area Display .............................................................................................................165
5.6.1.4.1 C/I Level...............................................................................................................................165
5.6.1.4.2 Max C/I Level.......................................................................................................................165
5.6.1.4.3 Min C/I Level........................................................................................................................165
5.6.1.4.4 Transmitter...........................................................................................................................165
5.6.2 Point Analysis..................................................................................................................................166
5.7 GPRS EGPRS Coverage Studies ........................................................................................................166
5.7.1 Coverage Area Determination.........................................................................................................166
5.7.1.1 All Servers .................................................................................................................................166
5.7.1.2 Best Signal Level per HCS Layer and a Margin ........................................................................166
5.7.1.3 Second Best Signal Level per HCS Layer and a Margin...........................................................166
5.7.1.4 HCS Servers and a Margin .......................................................................................................167
5.7.1.5 Highest Priority HCS Server and a Margin ................................................................................167
5.7.2 Calculation Options .........................................................................................................................167
5.7.2.1 Calculations Based on C ...........................................................................................................167
5.7.2.2 Calculations Based on C/I Without Considering Thermal Noise ...............................................167
5.7.2.3 Calculations Based on C/I Considering Thermal Noise ............................................................168
5.7.2.4 Ideal Link Adaptation (ILA) ........................................................................................................168
5.7.3 Coverage Study Scenarios .............................................................................................................168
5.7.3.1 GPRS/EGPRS Studies Based on C Without ILA ......................................................................168
5.7.3.1.1 Coding Scheme Based on C Without ILA............................................................................168
5.7.3.1.2 Throughput Based on C Without ILA ...................................................................................168
5.7.3.2 GPRS/EGPRS Studies Based on C With ILA ...........................................................................169
5.7.3.2.1 Coding Scheme Based on C With ILA.................................................................................169
5.7.3.2.2 Throughput Based on C With ILA ........................................................................................169
5.7.3.3 GPRS/EGPRS Studies Based on C/I Without ILA and Thermal Noise .....................................169
5.7.3.3.1 Coding Scheme Based on C/I Without ILA and Thermal Noise ..........................................169
5.7.3.3.2 Throughput Based on Worst Case Between C and C/I Without ILA....................................170
5.7.3.4 GPRS/EGPRS Studies Based on C/I With ILA and Without Thermal Noise.............................170
5.7.3.4.1 Coding Scheme Based on C/I With ILA and Without Thermal Noise ..................................170
5.7.3.4.2 Throughput Based on Worst Case Between C and C/I With ILA.........................................170
5.7.3.5 GPRS/EGPRS Studies Based on C/I Without ILA and With Thermal Noise.............................170
5.7.3.5.1 Coding Scheme Based on C/I Without ILA and With Thermal Noise ..................................170
5.7.3.5.2 Throughput Based on Interpolation Between C/N and C/(I+N) Without ILA ........................171
5.7.3.6 GPRS/EGPRS Studies Based on C/I With ILA and Thermal Noise..........................................171
5.7.3.6.1 Coding Scheme Based on C/I With ILA and Thermal Noise ...............................................171
5.7.3.6.2 Throughput Based on Interpolation Between C/N and C/(I+N) With ILA .............................172
5.7.4 Coverage Display............................................................................................................................172
5.7.4.1 GPRS/EGPRS Coding Schemes Study Display Types ............................................................172
5.7.4.1.1 Coding Schemes .................................................................................................................172
5.7.4.1.2 Best Coding Schemes .........................................................................................................173
5.7.4.2 RLC/MAC and Application Throughput/Timeslot Studies Display Types ..................................173
5.7.4.2.1 Relation Between RLC/MAC and Application Throughputs.................................................173
5.7.4.2.2 Throughput/Timeslot............................................................................................................173
5.7.4.2.3 Best Throughput/Timeslot....................................................................................................173
5.7.4.2.4 Average Throughput/Timeslot .............................................................................................173
5.7.4.2.5 Block Error Rate Computation .............................................................................................173
5.7.4.2.6 BLER Percentage ................................................................................................................173
5.7.4.2.7 Maximum BLER Percentage ...............................................................................................174
5.8 Circuit Quality Indicators Studies..........................................................................................................174
5.8.1 Circuit Quality Indicators .................................................................................................................174
5.8.2 Coverage Area Determination.........................................................................................................175
5.8.2.1 All Servers .................................................................................................................................176
5.8.2.2 Best Signal Level per HCS Layer and a Margin ........................................................................176
5.8.2.3 Second Best Signal Level per HCS Layer and a Margin...........................................................176
x AT271_TRG_E6 © Forsk 2009

Table of Contents
5.8.2.4 HCS Servers and a Margin....................................................................................................... 176

5.8.2.5 Highest Priority HCS Server and a Margin ............................................................................... 176
5.8.3 Calculation Options ........................................................................................................................ 176
5.8.3.1 Calculations Based on C/N....................................................................................................... 177
5.8.3.2 Calculations Based on C/(I+N) ................................................................................................. 177
5.8.3.3 Ideal Link Adaptation (ILA) ....................................................................................................... 178
5.8.4 Calculation Scenarios..................................................................................................................... 178
5.8.4.1 CQI Study Based on C/N Without ILA ...................................................................................... 178
5.8.4.2 CQI Study Based on C/N With ILA ........................................................................................... 178
5.8.4.3 CQI Study Based on C/(I+N) Without ILA................................................................................. 178
5.8.4.4 CQI Study Based on C/(I+N) With ILA...................................................................................... 179
5.8.5 Coverage Display ........................................................................................................................... 179
5.8.5.1 Circuit Quality Indicators Study Display Types ......................................................................... 179
5.8.5.1.1 FER/BER/MOS ................................................................................................................... 179
5.8.5.1.2 Max FER/Max BER/Max MOS ............................................................................................ 179
6 UMTS HSPA Networks ................................................................. 183

6.1 General Prediction Studies .................................................................................................................. 183
6.1.1 Calculation Criteria ......................................................................................................................... 183
6.1.2 Point Analysis ................................................................................................................................. 183
6.1.2.1 Profile Tab ................................................................................................................................ 183
6.1.2.2 Reception Tab .......................................................................................................................... 183
6.1.3.1 Service Area Determination...................................................................................................... 184
6.1.3.1.1 All Servers........................................................................................................................... 184
6.1.3.1.2 Best Signal Level and a Margin .......................................................................................... 184
6.1.3.1.3 Second Best Signal Level and a Margin ............................................................................. 184
6.1.3.2.1 Plot Resolution .................................................................................................................... 185
6.1.3.2.2 Display Types ..................................................................................................................... 185
6.2 Definitions and Formulas ..................................................................................................................... 186
6.2.1 Inputs.............................................................................................................................................. 186
6.2.2 Ec/Io Calculation ............................................................................................................................ 190
6.2.3 DL Eb/Nt Calculation ...................................................................................................................... 191
6.2.4 UL Eb/Nt Calculation ...................................................................................................................... 192
6.3 Active Set Management ....................................................................................................................... 192
6.4 Simulations........................................................................................................................................... 193
6.4.1 Generating a Realistic User Distribution ........................................................................................ 193
6.4.1.1 Simulations Based on Raster Traffic and Vector Traffic Maps ................................................. 193
6.4.1.1.1 Circuit Switched Service (i) ................................................................................................. 194
6.4.1.1.2 Packet Switched Service (j) ................................................................................................ 194
6.4.1.2 Simulations Based on Traffic Map per Service and per Transmitter ........................................ 197
6.4.2 Power Control Simulation............................................................................................................... 199
6.4.2.1 Algorithm Initialization............................................................................................................... 200
6.4.2.2 R99 Part of the Algorithm ......................................................................................................... 200
6.4.2.3 HSDPA Part of the Algorithm.................................................................................................... 207
6.4.2.3.1 HSDPA Power Allocation .................................................................................................... 207
6.4.2.3.2 Number of HS-SCCH Channels and Maximum Number of HSDPA Users......................... 207
6.4.2.3.3 HSDPA Bearer Allocation Process ..................................................................................... 208
6.4.2.3.4 Fast Link Adaptation Modelling ........................................................................................... 209
6.4.2.3.5 Scheduling Algorithms ........................................................................................................ 215
6.4.2.4 HSUPA Part of the Algorithm.................................................................................................... 216
6.4.2.4.1 HSUPA Users ..................................................................................................................... 217
6.4.2.4.2 Admission Control ............................................................................................................... 217
6.4.2.4.3 HSUPA Bearer Allocation Process ..................................................................................... 218
6.4.2.5 Convergence Criteria................................................................................................................ 220
6.4.3 Results ........................................................................................................................................... 220
6.4.3.1 R99 Related Results................................................................................................................. 220
6.4.3.2 HSPA Related Results.............................................................................................................. 222
6.4.3.2.1 Statistics Tab ...................................................................................................................... 222
6.4.3.2.2 Mobiles Tab ........................................................................................................................ 222
6.4.3.2.3 Cells Tab ............................................................................................................................. 224
6.4.3.2.4 Sites Tab ............................................................................................................................. 225
6.4.4 Appendices..................................................................................................................................... 226
6.4.4.1 Admission Control in the R99 Part............................................................................................ 226
6.4.4.2 Resources Management........................................................................................................... 226
© Forsk 2009 AT271_TRG_E6 xi

6.4.4.2.1 OVSF Codes Management..................................................................................................226

6.4.4.2.2 Channel Elements Management .........................................................................................227
6.4.4.3 Downlink Load Factor Calculation .............................................................................................228
6.4.4.3.1 Downlink Load Factor per Cell.............................................................................................228
6.4.4.3.2 Downlink Load Factor per Mobile ........................................................................................229
6.4.4.4 Uplink Load Factor Due to One User ........................................................................................229
6.4.4.5 Inter-carrier Power Sharing Modelling .......................................................................................231
6.5 UMTS HSPA Prediction Studies...........................................................................................................232
6.5.1 Point Analysis..................................................................................................................................232
6.5.1.1 AS Analysis Tab ........................................................................................................................232
6.5.1.1.1 Bar Graph and Pilot Sub-Menu............................................................................................232
6.5.1.1.2 Downlink Sub-Menu.............................................................................................................234
6.5.1.1.3 Uplink Sub-Menu .................................................................................................................238
6.5.2 Coverage Studies............................................................................................................................241
6.5.2.1 Pilot Reception Analysis ............................................................................................................242
6.5.2.1.1 Prediction Study Inputs........................................................................................................242
6.5.2.1.2 Study Display Options .........................................................................................................242
6.5.2.2 Downlink Service Area Analysis ................................................................................................243
6.5.2.3 Uplink Service Area Analysis ....................................................................................................245
6.5.2.4 Downlink Total Noise Analysis ..................................................................................................246
6.5.2.4.1 Study Inputs.........................................................................................................................247
6.5.2.4.2 Analysis on All Carriers........................................................................................................247
6.5.2.4.3 Analysis on a Specific Carrier..............................................................................................247
6.5.2.5 HSDPA Prediction Study ...........................................................................................................248
6.5.2.6 HSUPA Prediction Study ...........................................................................................................252
6.5.2.6.2 Calculation Options..............................................................................................................252
6.5.2.6.3 Display Options....................................................................................................................252
6.6 Automatic Neighbour Allocation............................................................................................................254
6.6.1 Global Allocation for All Transmitters ..............................................................................................254
6.6.2 Allocation for a Group of Transmitters or One Transmitter .............................................................258
6.7 Primary Scrambling Code Allocation ....................................................................................................258
6.7.1 Automatic Allocation Description.....................................................................................................259
6.7.1.1 Options and Constraints ............................................................................................................259
6.7.1.2 Allocation Process.....................................................................................................................260
6.7.1.2.1 Single Carrier Network.........................................................................................................260
6.7.1.2.2 Multi-Carrier Network...........................................................................................................261
6.7.1.3 Priority Determination ................................................................................................................261
6.7.1.3.1 Cell Priority ..........................................................................................................................261
6.7.1.3.2 Transmitter Priority ..............................................................................................................264
6.7.1.3.3 Site Priority ..........................................................................................................................264
6.7.2 Allocation Examples........................................................................................................................264
6.7.2.1 Allocation Strategies and Use a Maximum of Codes ................................................................264
6.7.2.1.1 Strategy: Clustered ..............................................................................................................265
6.7.2.1.2 Strategy: Distributed ............................................................................................................265
6.7.2.1.3 Strategy: ‘One Cluster per Site ............................................................................................266
6.7.2.1.4 Strategy: ‘Distributed per Site ..............................................................................................266
6.7.2.2 Allocate Carriers Identically .......................................................................................................266
6.8 Automatic GSM-UMTS Neighbour Allocation .......................................................................................267
6.8.1 Overview .........................................................................................................................................267
6.8.2.1 Algorithm Based on Distance ....................................................................................................268
6.8.2.2 Algorithm Based on Coverage Overlapping ..............................................................................268
6.8.2.3 Appendices................................................................................................................................270
6.8.2.3.1 Delete Existing Neighbours Option......................................................................................270
6.8.2.3.2 Calculation of Inter-Transmitter Distance ............................................................................270
7 IS-95 cdmaOne and CDMA2000 Networks ...................................273

7.1 General Prediction Studies ...................................................................................................................273
7.1.1 Calculation Criteria..........................................................................................................................273
7.1.2 Point Analysis..................................................................................................................................273
7.1.2.1 Profile Tab .................................................................................................................................273
xii AT271_TRG_E6 © Forsk 2009

Table of Contents
7.1.2.2 Reception Tab .......................................................................................................................... 274

7.1.3.1 Service Area Determination...................................................................................................... 274
7.1.3.1.1 All Servers........................................................................................................................... 274
7.1.3.1.2 Best Signal Level and a Margin .......................................................................................... 274
7.1.3.1.3 Second Best Signal Level and a Margin ............................................................................. 274
7.1.3.2.1 Plot Resolution .................................................................................................................... 275
7.1.3.2.2 Display Types ..................................................................................................................... 275
7.2 Definitions and Formulas ..................................................................................................................... 276
7.2.1 Parameters Used for IS-95 cdmaOne and CDMA2000 1xRTT Modelling ..................................... 276
7.2.1.1 Inputs ........................................................................................................................................ 276
7.2.1.2 Ec/I0 Calculation....................................................................................................................... 280
7.2.1.3 DL Eb/Nt Calculation ................................................................................................................ 280
7.2.1.4 UL Eb/Nt Calculation ................................................................................................................ 281
7.2.1.5 Simulation Results .................................................................................................................... 282
7.2.2 Parameters Used for CDMA2000 1xEV-DO Modelling .................................................................. 283
7.2.2.1 Inputs ........................................................................................................................................ 283
7.2.2.2 Ec/I0 and Ec/Nt Calculations .................................................................................................... 287
7.2.2.3 UL Eb/Nt Calculation ................................................................................................................ 287
7.2.2.4 Simulation Results .................................................................................................................... 288
7.3 Active Set Management ....................................................................................................................... 289
7.4 Simulations........................................................................................................................................... 290
7.4.1 Generating a Realistic User Distribution ........................................................................................ 290
7.4.1.1 Number of Users, User Activity Status and User Data Rate..................................................... 290
7.4.1.1.1 Simulations Based on Raster Traffic and Vector Traffic Maps............................................ 290
7.4.1.1.2 Simulations Based on Traffic Maps per Service and per Transmitter ................................. 293
7.4.1.2 Transition Flags for 1xEV-DO Rev.0 User Data Rates............................................................. 297
7.4.1.3 User Geographical Position ...................................................................................................... 297
7.4.2 Network Regulation Mechanism..................................................................................................... 297
7.4.2.1 IS-95 cdmaOne Power Control Simulation Algorithm............................................................... 297
7.4.2.1.1 Algorithm Initialization ......................................................................................................... 298
7.4.2.1.2 Presentation of the Algorithm.............................................................................................. 298
7.4.2.1.3 Convergence Criterion ........................................................................................................ 302
7.4.2.2 CDMA2000 1xRTT Power Control Simulation Algorithm.......................................................... 302
7.4.2.3 CDMA2000 1xEV-DO Power/Data Rate Control Simulation Algorithm .................................... 310
7.4.3 Appendices..................................................................................................................................... 316
7.4.3.1 Admission Control..................................................................................................................... 316
7.4.3.2 Resources Management........................................................................................................... 317
7.4.3.2.1 Walsh Code Management .................................................................................................. 317
7.4.3.2.2 Channel Element Management .......................................................................................... 317
7.4.3.3 Downlink Load Factor Calculation ............................................................................................ 318
7.4.3.3.1 Downlink Load Factor per Cell ............................................................................................ 318
7.4.3.3.2 Downlink Load Factor per Mobile........................................................................................ 319
7.5 IS-95 cdmaOne and CDMA2000 Prediction Studies ........................................................................... 319
7.5.1 Point Analysis: The AS Analysis Tab ............................................................................................. 319
7.5.1.1 Bar Graph and Pilot Sub-Menu................................................................................................. 320
7.5.1.2 Downlink Sub-Menu.................................................................................................................. 321
7.5.1.2.1 IS-95 cdmaOne ................................................................................................................... 321
7.5.1.2.2 CDMA2000 1xRTT.............................................................................................................. 324
7.5.1.2.3 CDMA2000 1xEV-DO ......................................................................................................... 327
7.5.1.3 Uplink Sub-Menu ...................................................................................................................... 328
7.5.1.3.1 IS-95 cdmaOne ................................................................................................................... 328
7.5.1.3.2 CDMA2000 1xRTT.............................................................................................................. 331
7.5.1.3.3 CDMA2000 1xEV-DO ......................................................................................................... 335
7.5.2.1 Pilot Reception Analysis ........................................................................................................... 338
7.5.2.1.1 1st Case: Analysis Based on all Carriers ............................................................................ 338
7.5.2.1.2 2nd Case: Analysis Based on a Specific Carrier................................................................. 338
7.5.2.2 Downlink Service Area Analysis ............................................................................................... 339
7.5.2.2.1 IS-95 cdmaOne and CDMA2000 1xRTT............................................................................. 339
7.5.2.2.2 CDMA2000 1xEV-DO ......................................................................................................... 340
© Forsk 2009 AT271_TRG_E6 xiii

7.5.2.3 Uplink Service Area Analysis ....................................................................................................342

7.5.2.3.1 IS-95 cdmaOne and CDMA2000 1xRTT .............................................................................342
7.5.2.3.2 CDMA2000 1xEV-DO ..........................................................................................................343
7.5.2.4 Downlink Total Noise Analysis ..................................................................................................345
7.5.2.4.1 Analysis on all Carriers ........................................................................................................346
7.5.2.4.2 Analysis on a Specific Carrier..............................................................................................346
7.6 Automatic Neighbour Allocation............................................................................................................346
7.6.1 Global Allocation for all Transmitters ..............................................................................................347
7.6.2 Allocation for a Group of Transmitters or One Transmitter .............................................................350
7.7 PN Offset Allocation..............................................................................................................................350
7.7.1.1 Options and Constraints ............................................................................................................351
7.7.1.2 Allocation Process.....................................................................................................................352
7.7.1.2.1 Single Carrier Network.........................................................................................................352
7.7.1.2.2 Multi-Carrier Network...........................................................................................................353
7.7.1.2.3 Difference between Adjacent and Distributed PN-Clusters .................................................353
7.7.1.3 Priority Determination ................................................................................................................353
7.7.1.3.1 Cell Priority ..........................................................................................................................353
7.7.1.3.2 Transmitter Priority ..............................................................................................................355
7.7.1.3.3 Site Priority ..........................................................................................................................355
7.7.2 Allocation Examples........................................................................................................................356
7.7.2.1 Strategy: PN Offset per Cell ......................................................................................................356
7.7.2.2 Strategy: Adjacent PN-Clusters Per Site...................................................................................357
7.7.2.3 Strategy: Distributed PN-Clusters Per Site................................................................................357
7.8 Automatic GSM-CDMA Neighbour Allocation.......................................................................................357
7.8.1 Overview .........................................................................................................................................357
7.8.2.3 Delete Existing Neighbours Option ...........................................................................................360
7.9 Automatic cdmaOne-CDMA2000 Neighbour Allocation .......................................................................360
7.9.1 Overview .........................................................................................................................................360
7.9.2.1 Delete Existing Neighbours Option ...........................................................................................362
8 TD-SCDMA Networks....................................................................365
8.1 Definitions and Formulas ......................................................................................................................365
8.1.1 Inputs ..............................................................................................................................................365
8.1.2 P-CCPCH Eb/Nt and C/I Calculation ..............................................................................................369
8.1.3 DwPCH C/I Calculation ...................................................................................................................369
8.1.4 DL TCH Eb/Nt and C/I Calculation..................................................................................................370
8.1.5 UL TCH Eb/Nt and C/I Calculation..................................................................................................370
8.1.6 Interference Calculation ..................................................................................................................371
8.1.7 HSDPA Dynamic Power Calculations .............................................................................................371
8.1.8 Smart Antenna Modelling................................................................................................................371
8.2 Signal Level Based Calculations ..........................................................................................................372
8.2.1 Point Analysis..................................................................................................................................372
8.2.1.1 Profile Tab .................................................................................................................................373
8.2.1.2 Reception Tab ...........................................................................................................................373
8.2.2 RSCP Based Coverage Predictions................................................................................................373
8.2.2.1 Calculation Criteria ....................................................................................................................373
8.2.2.2 P-CCPCH RSCP Coverage Prediction .....................................................................................374
8.2.2.2.1 Coverage Condition .............................................................................................................374
8.2.2.2.2 Coverage Display ................................................................................................................374
8.2.2.3 Best Server P-CCPCH Coverage Prediction.............................................................................374
8.2.2.4 P-CCPCH Pollution Coverage Prediction..................................................................................375
8.2.2.5 DwPCH RSCP Coverage Prediction .........................................................................................375
8.2.2.6 UpPCH RSCP Coverage Prediction..........................................................................................376
8.2.2.7 Baton Handover Coverage Prediction .......................................................................................376
8.2.2.8 Scrambling Code Interference Analysis ....................................................................................377
8.3 Monte Carlo Simulations.......................................................................................................................377
8.3.1 Generating a Realistic User Distribution .........................................................................................377
xiv AT271_TRG_E6 © Forsk 2009

Table of Contents
8.3.1.1 Simulations Based on Raster and Vector Traffic Maps ............................................................ 378
8.3.1.2 Simulations Based on Traffic Map Based on Transmitters and Services ................................. 381
8.3.2 Power Control Simulation............................................................................................................... 383
8.3.2.1 Algorithm Initialisation............................................................................................................... 383
8.3.2.2 R99 Part of the Algorithm ......................................................................................................... 384
8.3.2.2.1 Determination of Mi’s Best Server (SBS(Mi))...................................................................... 384
8.3.2.2.2 Dynamic Channel Allocation ............................................................................................... 384
8.3.2.2.3 Uplink Power Control .......................................................................................................... 386
8.3.2.2.4 Downlink Power Control...................................................................................................... 387
8.3.2.2.5 Uplink Signals Update......................................................................................................... 389
8.3.2.2.6 Downlink Signals Update .................................................................................................... 390
8.3.2.2.7 Control of Radio Resource Limits (Downlink Traffic Power and Uplink Load) .................... 390
8.3.2.3 HSDPA Part of the Algorithm.................................................................................................... 390
8.3.2.3.1 HSDPA Power Allocation .................................................................................................... 390
8.3.2.3.2 Connection Status and Number of HSDPA Users .............................................................. 392
8.3.2.3.3 HSDPA Admission Control.................................................................................................. 392
8.3.2.3.4 HSDPA Dynamic Channel Allocation.................................................................................. 393
8.3.2.3.5 Ressource Unit Saturation .................................................................................................. 393
8.3.2.4 Convergence Criteria................................................................................................................ 393
8.4 TD-SCDMA Prediction Studies ............................................................................................................ 394
8.4.1 P-CCPCH Reception Analysis (Eb/Nt) or (C/I) ............................................................................... 394
8.4.2 DwPCH Reception Analysis (C/I) ................................................................................................... 395
8.4.3 Downlink TCH RSCP Coverage ..................................................................................................... 397
8.4.4 Uplink TCH RSCP Coverage ......................................................................................................... 398
8.4.5 Downlink Total Noise...................................................................................................................... 398
8.4.6 Downlink Service Area (Eb/Nt) or (C/I)........................................................................................... 399
8.4.7 Uplink Service Area (Eb/Nt) or (C/I) ............................................................................................... 401
8.4.8 Effective Service Area (Eb/Nt) or (C/I) ........................................................................................... 402
8.4.9 Cell to Cell Interference .................................................................................................................. 403
8.4.10 UpPCH Interference ....................................................................................................................... 404
8.4.11 HSDPA Coverage .......................................................................................................................... 404
8.5 Smart Antenna Modelling..................................................................................................................... 405
8.5.1 Modelling in Simulations................................................................................................................. 405
8.5.1.1 Grid of Beams Modelling .......................................................................................................... 405
8.5.1.2 Adaptive Beam Modelling ......................................................................................................... 406
8.5.1.3 Statistical Modelling .................................................................................................................. 407
8.5.1.4 Optimum Beamformer Model.................................................................................................... 408
8.5.1.4.1 Downlink ............................................................................................................................. 408
8.5.1.4.2 Uplink .................................................................................................................................. 409
8.5.1.5 3rd Party Smart Antenna Modelling.......................................................................................... 411
8.5.2 Construction of the Geographic Distributions ................................................................................. 411
8.5.3 Modelling in Coverage Predictions ................................................................................................. 412
8.6 N-Frequency Mode and Carrier-Type Allocation.................................................................................. 413
8.6.1 Automatic Carrier-Type Allocation.................................................................................................. 413
8.7 Neighbour Allocation ............................................................................................................................ 413
8.7.1 Global Allocation for All Transmitters ............................................................................................. 414
8.7.2 Allocation for a Group of Transmitters or One Transmitter ............................................................ 417
8.8 Scrambling Code Allocation ................................................................................................................. 417
8.8.1 Automatic Allocation Description .................................................................................................... 418
8.8.1.1 Allocation Constraints and Options........................................................................................... 418
8.8.1.2 Allocation Strategies ................................................................................................................. 418
8.8.1.3 Allocation Process .................................................................................................................... 419
8.8.1.3.1 Single Carrier Network ........................................................................................................ 419
8.8.1.3.2 Multi-Carrier Network .......................................................................................................... 420
8.8.1.4 Priority Determination ............................................................................................................... 420
8.8.1.4.1 Cell Priority.......................................................................................................................... 420
8.8.1.4.2 Transmitter Priority.............................................................................................................. 423
8.8.1.4.3 Site Priority.......................................................................................................................... 423
8.8.2 IScrambling Code Allocation Example ........................................................................................... 423
8.8.2.1 Single Carrier Network.............................................................................................................. 423
8.8.2.1.1 Strategy: Clustered ............................................................................................................. 424
8.8.2.1.2 Strategy: Distributed per Cell .............................................................................................. 424
8.8.2.1.3 Strategy: One SYNC_DL Code per Site ............................................................................. 425
8.8.2.1.4 Strategy: Distributed per Site .............................................................................................. 425
© Forsk 2009 AT271_TRG_E6 xv

8.8.2.2 Multi Carrier Network.................................................................................................................425

8.9 Automatic GSM/TD-SCDMA Neighbour Allocation ..............................................................................426
8.9.1.3 Appendices................................................................................................................................428
8.9.1.3.1 Delete Existing Neighbours Option......................................................................................428
8.9.1.3.2 Calculation of Inter-Transmitter Distance ............................................................................429
9 WiMAX BWA Networks..................................................................433

9.1 Definitions and Formulas ......................................................................................................................433
9.1.1 Input ................................................................................................................................................433
9.1.2 Co- and Adjacent Channel Overlaps Calculation............................................................................436
9.1.3 Preamble Signal Quality Calculations .............................................................................................437
9.1.3.1 Preamble Signal Level Calculation............................................................................................437
9.1.3.2 Preamble Noise Calculation ......................................................................................................438
9.1.3.3 Preamble Interference Calculation ............................................................................................438
9.1.3.4 Preamble C/N Calculation .........................................................................................................438
9.1.3.5 Preamble C/(I+N) Calculation....................................................................................................438
9.1.4 Traffic and Pilot Signal Quality Calculations ...................................................................................438
9.1.4.1 Traffic and Pilot Signal Level Calculation (DL) ..........................................................................438
9.1.4.2 Noise Calculation (DL) ..............................................................................................................439
9.1.4.3 Traffic and Pilot Interference Calculation (DL) ..........................................................................439
9.1.4.4 Traffic and Pilot C/N Calculation (DL) .......................................................................................440
9.1.4.5 Traffic and Pilot C/(I+N) Calculation (DL) ..................................................................................440
9.1.4.6 Traffic Signal Level Calculation (UL) .........................................................................................441
9.1.4.7 Noise Calculation (UL) ..............................................................................................................441
9.1.4.8 Traffic Interference Calculation (UL) .........................................................................................441
9.1.4.9 Traffic C/N Calculation (UL) ......................................................................................................441
9.1.4.10 Traffic C/(I+N) Calculation (UL) .................................................................................................442
9.1.5 Channel Throughput Calculation.....................................................................................................442
9.1.5.1 Calculation of Cell Capacity (Total Resources) .........................................................................442
9.1.5.2 Bearer Determination ................................................................................................................443
9.1.5.3 Channel Throughput Calculation...............................................................................................443
9.1.6 Scheduling and Radio Resource Management...............................................................................443
9.1.6.1 User Throughput Calculation.....................................................................................................445
9.1.7 Optimum Beamformer Model ..........................................................................................................445
9.2 Calculation Processes ..........................................................................................................................446
9.2.1 Point Analysis..................................................................................................................................446
9.2.1.1 Profile Tab .................................................................................................................................447
9.2.1.2 Reception Tab ...........................................................................................................................447
9.2.2 Preamble Signal Level Coverage Predictions.................................................................................447
9.2.2.1.1 All Servers ...........................................................................................................................447
9.2.2.1.2 Best Signal Level and a Margin ...........................................................................................447
9.2.2.1.3 Second Best Signal Level and a Margin..............................................................................448
9.2.2.2 Coverage Display ......................................................................................................................448
9.2.2.2.1 Coverage Resolution ...........................................................................................................448
9.2.2.2.2 Display Types ......................................................................................................................448
9.2.3 Signal Quality Analysis Coverage Predictions ................................................................................449
9.2.3.2 Coverage Parameter Calculation ..............................................................................................450
9.2.3.3.2 Signal Quality Analysis (DL) Display Types.........................................................................450
9.2.3.3.3 Signal Quality Analysis (UL) Display Types.........................................................................451
9.2.4 Calculations on Subscriber Lists .....................................................................................................451
9.2.5 Monte Carlo Simulations .................................................................................................................452
9.2.5.1 Generating a Realistic User Distribution ...................................................................................452
9.2.5.1.1 Simulations Based on Raster Traffic Maps, Vector Traffic Maps, and Subscriber Lists......452
9.2.5.1.2 Simulations Based on Traffic Map Based on Transmitters and Services ............................454
9.2.5.2 Simulation Process....................................................................................................................455
9.2.6 C/(I+N)-Based Coverage Predictions..............................................................................................459
9.2.6.2 Coverage Parameter Calculation ..............................................................................................460
9.2.6.3.2 Coverage by C/(I+N) Level (DL) Display Types ..................................................................460
xvi AT271_TRG_E6 © Forsk 2009

Table of Contents
9.2.6.3.3 Coverage by Best Bearer (DL) Display Types .................................................................... 461

9.2.6.3.4 Coverage by Channel Throughput (DL) Display Types ...................................................... 461
9.2.6.3.5 Coverage by C/(I+N) Level (UL) Display Types.................................................................. 461
9.2.6.3.6 Coverage by Best Bearer (UL) Display Types .................................................................... 462
9.2.6.3.7 Coverage by Channel Throughput (UL) Display Types ...................................................... 462
9.3 Calculation Algorithms ......................................................................................................................... 462
9.3.1 Co- and Adjacent Channel Overlaps Calculation ........................................................................... 462
9.3.1.1 Conversion From Channel Numbers to Start and End Frequencies ........................................ 463
9.3.1.2 Co-Channel Overlap Calculation .............................................................................................. 464
9.3.1.3 Adjacent Channel Overlap Calculation ..................................................................................... 464
9.3.1.4 FDD – TDD Overlap Ratio Calculation ..................................................................................... 465
9.3.1.5 Total Overlap Ratio Calculation ................................................................................................ 465
9.3.2 Preamble Signal Quality Calculations ............................................................................................ 466
9.3.2.1 Preamble Signal Level Calculation ........................................................................................... 466
9.3.2.2 Preamble Noise Calculation ..................................................................................................... 467
9.3.2.3 Preamble Interference Calculation ........................................................................................... 469
9.3.2.4 Preamble C/N Calculation ........................................................................................................ 470
9.3.2.5 Preamble C/(I+N) Calculation ................................................................................................... 470
9.3.3 Best Server Determination ............................................................................................................. 471
9.3.4 Service Area Calculation ................................................................................................................ 471
9.3.5 Permutation Zone Selection (WiMAX 802.16e).............................................................................. 471
9.3.6 Traffic and Pilot Signal Quality Calculations................................................................................... 472
9.3.6.1 Traffic and Pilot Signal Level Calculation (DL) ......................................................................... 473
9.3.6.2 Noise Calculation (DL).............................................................................................................. 474
9.3.6.3 Traffic and Pilot Interference Calculation (DL).......................................................................... 475
9.3.6.3.1 Traffic and Pilot Interference Signal Levels Calculation (DL).............................................. 475
9.3.6.3.2 Effective Traffic and Pilot Interference Calculation (DL) ..................................................... 478
9.3.6.4 Traffic and Pilot C/N Calculation (DL)....................................................................................... 482
9.3.6.5 Traffic C/(I+N) Calculation (DL) ................................................................................................ 483
9.3.6.6 Traffic Signal Level Calculation (UL) ........................................................................................ 484
9.3.6.7 Noise Calculation (UL).............................................................................................................. 485
9.3.6.8 Traffic Interference Calculation (UL)......................................................................................... 486
9.3.6.8.1 Traffic Interference Signal Levels Calculation (UL)............................................................. 486
9.3.6.8.2 Noise Rise Calculation (UL) ................................................................................................ 487
9.3.6.9 Traffic C/N Calculation (UL)...................................................................................................... 487
9.3.6.10 Traffic C/(I+N) Calculation (UL) ................................................................................................ 490
9.3.7 Channel Throughput Calculation.................................................................................................... 493
9.3.7.1 Calculation of Cell Capacity (Total Resources) ........................................................................ 493
9.3.7.1.1 Calculation of Sampling Frequency .................................................................................... 493
9.3.7.1.2 Calculation of Symbol Duration........................................................................................... 493
9.3.7.1.3 Calculation of Cell Capacity - TDD Networks...................................................................... 494
9.3.7.1.4 Calculation of Cell Capacity - FDD Networks...................................................................... 495
9.3.7.2 Bearer Determination................................................................................................................ 496
9.3.7.3 Channel Throughput Calculation .............................................................................................. 496
9.3.8 Scheduling and Radio Resource Management .............................................................................. 498
9.3.8.1 Scheduling and Radio Resource Allocation.............................................................................. 499
9.3.8.2 User Throughput Calculation .................................................................................................... 504
9.3.9 Optimum Beamformer Model ......................................................................................................... 505
9.3.9.1 Introduction ............................................................................................................................... 505
9.3.9.2 Downlink ................................................................................................................................... 506
9.3.9.3 Uplink........................................................................................................................................ 507
9.4 Automatic Allocation Algorithms........................................................................................................... 509
9.4.1 Automatic Neighbour Allocation ..................................................................................................... 509
9.4.2 Automatic Preamble Index Allocation............................................................................................. 511
9.4.2.1 Constraint and Relationship Weights........................................................................................ 512
9.4.2.2 Calculation of Cost Between TBA and Related Cells ............................................................... 512
9.4.2.3 Automatic Allocation Algorithm ................................................................................................. 514
9.5 Glossary of WiMAX Terms................................................................................................................... 514
10 Repeaters and Remote Antennas................................................. 519

10.1 Modelling Repeaters ............................................................................................................................ 519
10.1.1 CDMA Documents.......................................................................................................................... 519
10.1.1.1 Over the Air............................................................................................................................... 519
10.1.1.1.1 Signal Level Received From Repeaters.............................................................................. 519
10.1.1.1.2 Gain Automatic Calculation................................................................................................. 520
10.1.1.1.3 Donor Side Parameter Automatic Calculation..................................................................... 521
10.1.1.2 Microwave Link ......................................................................................................................... 522
10.1.1.2.1 Signal Level Received From Repeaters.............................................................................. 522
© Forsk 2009 AT271_TRG_E6 xvii

10.1.1.2.2 Gain Automatic Calculation .................................................................................................523

10.1.1.3 Fibre Link...................................................................................................................................524
10.1.1.3.1 Signal Level Received From Repeaters ..............................................................................524
10.1.1.3.2 Gain Automatic Calculation .................................................................................................524
10.1.1.4 Appendices................................................................................................................................525
10.1.1.4.1 Automatic Controls ..............................................................................................................525
10.1.1.4.2 Carrier Power and Interference Calculation.........................................................................526
10.1.1.4.3 Consideration of Repeater Noise Figure .............................................................................528
10.1.2 GSM Documents.............................................................................................................................528
10.1.2.1 Over the Air ...............................................................................................................................528
10.1.2.1.2 EIRP Automatic Calculation.................................................................................................529
10.1.2.1.3 Donor Side Parameter Automatic Calculation .....................................................................530
10.1.2.2 Microwave Link..........................................................................................................................531
10.1.2.3 Fibre Link...................................................................................................................................532
10.1.2.4 Appendices................................................................................................................................533
10.1.2.4.1 Automatic Controls ..............................................................................................................533
10.2 Modelling Remote Antennas.................................................................................................................534
10.2.1 CDMA Documents ..........................................................................................................................534
10.2.1.1 Signal Level Received From Repeaters ....................................................................................534
10.2.1.2 Gain Automatic Calculation .......................................................................................................534
10.2.2 GSM Documents.............................................................................................................................535
10.2.2.1 Signal Level Received From Repeaters ....................................................................................535
10.2.2.2 EIRP Automatic Calculation ......................................................................................................536
xviii AT271_TRG_E6 © Forsk 2009

List of Figures
List of Figures
Figure 2.1: Digital Terrain Model.................................................................................................................................. 35

Figure 2.2: Schematic view of a DTM file .................................................................................................................... 35
Figure 2.3: Clutter Classes .......................................................................................................................................... 36
Figure 4.1: Example 1: Single Calculation Area .......................................................................................................... 75
Figure 4.2: Example 2: Multiple Calculation Areas ...................................................................................................... 75
Figure 4.3: Ground Altitude Determination - 1 ............................................................................................................. 77
Figure 4.7: Clutter Height............................................................................................................................................. 78
Figure 4.8: Radial calculation method.......................................................................................................................... 79
Figure 4.9: Site-bin centre profile................................................................................................................................. 79
Figure 4.10: Radial calculation method.......................................................................................................................... 80
Figure 4.11: Enhanced Slope at Receiver ..................................................................................................................... 87
Figure 4.12: Losses due to Clutter................................................................................................................................. 90
Figure 4.13: Tx-Rx profile .............................................................................................................................................. 91
Figure 4.14: Knife-Edge Diffraction.............................................................................................................................. 105
Figure 4.15: Deygout Construction – 1 Obstacle ......................................................................................................... 106
Figure 4.16: Deygout Construction – 3 Obstacles ....................................................................................................... 106
Figure 4.17: Epstein-Peterson Construction ................................................................................................................ 107
Figure 4.18: Millington Construction ............................................................................................................................ 108
Figure 4.19: Azimuth and Tilt Computation.................................................................................................................. 109
Figure 4.20: Vertical Pattern Transformation due to Electrical Downtilt....................................................................... 111
Figure 4.21: Vertical Antenna Pattern.......................................................................................................................... 112
Figure 4.22: Peaks and Nulls in the Antenna Pattern .................................................................................................. 112
Figure 4.23: Log-normal Probability Density Function ................................................................................................. 113
Figure 4.24: Normalised Margin .................................................................................................................................. 117
Figure 4.25: Margin - Probability (Case of 2 Signals) .................................................................................................. 123
Figure 4.26: Margin - Probability (Case of 3 Signals with sigma = 8dB, delta1 = 1dB) ............................................... 123
Figure 4.27: Margin - Probability (Case of 3 Signals with sigma = 8dB, delta1 = 2dB) ............................................... 123
Figure 4.28: Reference Point - Location of the Transmission/Reception parameters ................................................. 124
Figure 5.1: Representation of a Concentric Cell TXi.................................................................................................. 138
Figure 5.2: Representation of Micro and Macro Layers............................................................................................. 139
Figure 5.3: Concentric Cells....................................................................................................................................... 140
Figure 5.4: Concentric Cells....................................................................................................................................... 144
Figure 5.5: Reduction of Throughput per Timeslot .................................................................................................... 147
Figure 5.6: Reduction Factor for Different Packet Switched Traffic Loads (Lp, X-axis)............................................. 148
Figure 5.7: Blocking Probability for Different Packet Switched Traffic Loads (Lp, X-axis)......................................... 150
Figure 5.8: Network Dimensioning Process............................................................................................................... 150
Figure 5.9: Minimum Throughput Reduction Factor .................................................................................................. 154
Figure 5.10: Overlapping Zones .................................................................................................................................. 158
Figure 5.11: FER vs. C/I Graphs.................................................................................................................................. 175
Figure 5.12: BER vs. C/I Graphs ................................................................................................................................. 175
Figure 5.13: MOS vs. C/I Graphs................................................................................................................................. 175
Figure 6.1: Description of a Packet Session .............................................................................................................. 195
Figure 6.2: UMTS HSPA Power Control Algorithm.................................................................................................... 200
Figure 6.3: Connection status of HSDPA users......................................................................................................... 208
Figure 6.4: Bearer Allocation Process for Scheduled Users...................................................................................... 208
Figure 6.5: HSDPA Radio Bearers Table .................................................................................................................. 212
Figure 6.6: HSDPA UE Categories Table .................................................................................................................. 212
Figure 6.7: HSUPA UE Categories Table .................................................................................................................. 217
Figure 6.8: HSUPA Radio Bearers Table .................................................................................................................. 218
Figure 6.9: HSUPA Bearer SelectionTable................................................................................................................ 218
Figure 6.10: OVSF Code Tree Indices (Not OVSF Code Numbers)............................................................................ 227
Figure 6.11: Overlapping Zone for Intra-carrier Neighbours ........................................................................................ 255
© Forsk 2009 AT271_TRG_E6 xix

Figure 6.12: Overlapping Zone for Inter-carrier Neighbours - 1st Case ....................................................................... 256
Figure 6.13: Overlapping Zone for Inter-carrier Neighbours - 2nd Case ...................................................................... 256
Figure 6.14: Neighbourhood Constraints...................................................................................................................... 262
Figure 6.15: Primary Scrambling Codes Allocation ...................................................................................................... 264
Figure 6.16: Inter-Transmitter Distance Computation .................................................................................................. 270
Figure 7.1: IS-95 cdmaOne Power Control Algorithm ................................................................................................ 298
Figure 7.2: CDMA2000 1xRTT Power Control Algorithm ........................................................................................... 303
Figure 7.3: CDMA2000 1xEVDO Power Control Algorithm ........................................................................................ 310
Figure 7.4: Walsh Code Tree Indices (Not Walsh Code Numbers) ............................................................................ 317
Figure 7.5: Overlapping Zones - 1st Case.................................................................................................................. 348
Figure 7.6: Overlapping Zones - 2nd Case ................................................................................................................ 348
Figure 7.8: PN Offset Allocation ................................................................................................................................. 356
Figure 8.1: Description of a Packet Session............................................................................................................... 379
Figure 8.2: TD-SCDMA Power Control Algorithm ...................................................................................................... 383
Figure 8.3: Grid Of Beams Modelling ......................................................................................................................... 406
Figure 8.4: GOB Modelling - Determination of the Best Beam................................................................................... 406
Figure 8.5: Adaptive Beam Modelling - Determination of the Best Beam .................................................................. 407
Figure 8.6: Linear Adaptive Array System .................................................................................................................. 408
Figure 8.7: Downlink Beamforming ............................................................................................................................ 408
Figure 8.8: Uplink Adaptive Algorithm ........................................................................................................................ 410
Figure 8.9: Construction of the Geographic Distribution of Downlink Traffic Power ................................................... 412
Figure 8.10: Geographic Distribution of Downlink Traffic Power .................................................................................. 412
Figure 8.11: Geographic Distribution of downlink traffic power and uplink load ........................................................... 413
Figure 8.12: Intra-carrier Neighbours ........................................................................................................................... 415
Figure 8.13: Overlapping Coverages............................................................................................................................ 415
Figure 8.15: Scrambling Code Allocation Example ...................................................................................................... 423
Figure 8.16: Scrambling Code Allocation to All Carriers .............................................................................................. 426
Figure 8.17: Inter-Transmitter Distance Computation .................................................................................................. 429
Figure 9.1: WiMAX Simulation Algorithm ................................................................................................................... 456
Figure 9.2: Victim and Interfering Mobiles .................................................................................................................. 457
Figure 9.3: Simulation Convergence Stability Factor ................................................................................................. 457
Figure 9.4: Co-Channel and Adjacent Channel Overlaps .......................................................................................... 463
Figure 9.5: Downlink C/(I+N) calculation in Simulations............................................................................................. 477
Figure 9.6: Downlink C/(I+N) calculation in Coverage Predictions ............................................................................. 478
Figure 9.7: Segmentation ........................................................................................................................................... 479
Figure 9.8: Segmentation Interference Scenarios ...................................................................................................... 481
Figure 9.9: Uplink Subchannelization in WiMAX 802.16e Networks (Example) ......................................................... 489
Figure 9.10: Uplink Subchannelization in WiMAX 802.16e Networks (Example) ......................................................... 492
Figure 9.11: Linear Adaptive Array System .................................................................................................................. 505
Figure 9.12: Downlink Beamforming ............................................................................................................................ 506
Figure 9.13: Uplink Adaptive Algorithm ........................................................................................................................ 507
Figure 9.14: Determination of Adjacent Cells ............................................................................................................... 509
Figure 9.15: Overlapping Zones ................................................................................................................................... 510
Figure 9.16: Weighted Distance Between Cells ........................................................................................................... 513
Figure 9.17: Importance Based on Distance Relation .................................................................................................. 514
Figure 9.18: WiMAX Frame .......................................................................................................................................... 515
Figure 10.1: CDMA Documents - Over the Air Repeater ............................................................................................. 519
Figure 10.2: Over the Air Repeater - Downlink Total Gain ........................................................................................... 520
Figure 10.3: Over the Air Repeater - Uplink Total Gain................................................................................................ 521
Figure 10.4: Angle from North (Azimuth) ...................................................................................................................... 521
Figure 10.5: Positive/Negative Mechanical Downtilt..................................................................................................... 522
Figure 10.6: Tilt Angle Computation ............................................................................................................................. 522
Figure 10.7: CDMA Documents - Microwave Link Repeater ........................................................................................ 522
Figure 10.8: Microwave Link Repeater - Downlink Total Gain ..................................................................................... 523
Figure 10.9: Microwave Link Repeater - Uplink Total Gain .......................................................................................... 523
Figure 10.10: CDMA Documents - Fibre Link Repeater ................................................................................................. 524
Figure 10.11: Fibre Link Repeater - Downlink Total Gain .............................................................................................. 525
Figure 10.12: Fibre Link Repeater - Uplink Total Gain ................................................................................................... 525
Figure 10.13: GSM Documents - Over the Air Repeater ................................................................................................ 529
Figure 10.14: Over the Air Repeater - EIRP................................................................................................................... 529
xx AT271_TRG_E6 © Forsk 2009

List of Figures
Figure 10.15: Angle from North (Azimuth) ..................................................................................................................... 530

Figure 10.16: Positive/Negative Mechanical Downtilt .................................................................................................... 530
Figure 10.17: Tilt Angle Computation ............................................................................................................................ 530
Figure 10.18: GSM Documents - Microwave Link Repeater.......................................................................................... 531
Figure 10.19: Microwave Link Repeater - EIRP............................................................................................................. 532
Figure 10.20: GSM Documents - Fibre Link Repeater................................................................................................... 532
Figure 10.21: Fibre Link Repeater - EIRP...................................................................................................................... 533
Figure 10.22: CDMA Documents - Remote Antenna Signal Level ................................................................................ 534
Figure 10.23: Remote Antennas - Downlink Total Gain................................................................................................. 534
Figure 10.24: Remote Antennas - Uplink Total Gain ..................................................................................................... 535
Figure 10.25: GSM Documents - Remote Antenna Signal Level................................................................................... 535
Figure 10.26: Remote Antennas - EIRP ........................................................................................................................ 536
© Forsk 2009 AT271_TRG_E6 xxi

xxii AT271_TRG_E6 © Forsk 2009

Chapter 1
Coordinate Systems and Units
This chapter presents the different coordinate systems available in Atoll by default. It describes the projection,
display, and internal coordinate systems, and describes the format of the coordinate systems files. This chapter
also provides details of the different power and length units available in Atoll.
Atoll
Atoll
Microwave
RF PlanningLink
Microwave andPlanning
Optimisation Software
Software
24 AT271_TRG_E6 © Forsk 2009

Chapter 1: Coordinate Systems and Units
1 Coordinate Systems and Units

1.1 Coordinate Systems
A map or a geo-spatial database is a flat representation of data collected from a curved surface. A projection is a means
for producing all or part of a spheroid on a flat sheet. This projection cannot be done without distortion. Therefore, the
cartographer must choose the characteristic (distance, direction, scale, area, or shape) that he wants to be shown accu-
rately at the expense of the other characteristics, or compromise on several characteristics [1-3]. The projected zones are
referenced using cartographic coordinates (meter, yard, etc.). Two projection methods are widely used:
• The Lambert Conformal-Conic Method: A portion of the earth is mathematically projected on a cone conceptu-
ally secant at one or two standard parallels. This projection method is useful for representing countries or regions
that have a predominant east-west expanse.
• The Universal Transverse Mercator (UTM) Method: A portion of the earth is mathematically projected on a cyl-
inder tangent to a meridian (which is transverse or crosswise to the equator). This projection method is useful for
mapping large areas that are oriented north-south.
The geographic system is not a projection. It is only a representation of a location on the surface of the earth in geographic
coordinates (degree-minute-second, grade) giving the latitude and longitude in relation to the meridian origin (e.g., Paris
for NTF system and Greenwich for ED50 system). The locations in the geographic system can be converted into other
projections.
References:
[1] Snyder, John. P., Map Projections Used by the US Geological Survey, 2nd Edition, United States Government
Printing Office, Washington, D.C., 313 pages, 1982.
[2] http://www.colorado.edu/geography/gcraft/notes/gps/gps_f.html
[3] http://www.posc.org/Epicentre.2_2/DataModel/ExamplesofUsage/eu_cs34.html
[4] http://www.ign.fr/telechargement/Pi/SERVICES/transfo.pdf (Document in French)
1.1.1 Description of Coordinate Systems

A Geographic coordinate system is a latitude and longitude coordinate system. The latitude and longitude are related to
an ellipsoid, a geodetic datum, and a prime meridian. The geodetic datum provides the position and orientation of the ellip-
soid relative to the earth.
Cartographic coordinate systems are obtained by transforming each (latitude, longitude) value into an (easting, northing)
value. A projection coordinate system is obtained by transforming each (latitude, longitude) value into an (easting, north-
ing) value. Projection coordinate systems are geographic coordinate systems that provide longitude and latitude, and the
transformation method characterised by a set of parameters. Different methods may require different sets of parameters.
For example, the parameters required for Transverse Mercator coordinate systems are:
• The longitude of the natural origin (central meridian)
• The latitude of the natural origin
• The False Easting value
• The False Northing value
• A scaling factor at the natural origin (central meridian)
Basic definitions are presented below.
1.1.1.1 Geographic Coordinate System

The geographic coordinate system is a datum and a meridian. Atoll enables you to choose the most suitable geographic
coordinate system for your geographic data.
1.1.1.2 Datum
The datum consists of the ellipsoid and its position relative to the WGS84 ellipsoid. In addition to the ellipsoid, translation,
rotation, and distortion parameters define the datum.
1.1.1.3 Meridian
The standard meridian is Greenwich, but some geographic coordinate systems are based on other meridians. These
meridians are defined by the longitude with respect to Greenwich.
1.1.1.4 Ellipsoid
The ellipsoid is the pattern used to model the earth. It is defined by its geometric parameters.
© Forsk 2009 AT271_TRG_E6 25

1.1.1.5 Projection
The projection is the transformation applied to project the ellipsoid of the earth on to a plane. There are different projection
methods that use specific sets of parameters.
1.1.1.6 Projection Coordinate System

The projection coordinate system is the result of the application of a projection to a geographic coordinate system. It asso-
ciates a geographic coordinate system and a projection. Atoll enables you to choose the projection coordinate system
matching your geographic data.
1.1.2 Coordinate Systems in Atoll

Depending on the working environment, there can be either two or four coordinate systems used in Atoll. If you are working
with stand-alone documents, i.e., documents not connected to databases, there are two coordinate systems used in Atoll:
• Projection coordinate system
• Display coordinate system
If you are working in a multi-user environment, Atoll uses four coordinate systems:
• Projection coordinate system for the Atoll document
• Display coordinate system for the Atoll document
• Internal projection coordinate system for the database
• Internal display coordinate system for the database
1.1.2.1 Projection Coordinate System

The projection coordinate system is the coordinate system of the available raster geographic data files. You should set the
projection coordinate system of your Atoll document so that it corresponds to the coordinate system of the available raster
geographic data. You can set the projection coordinate system of your document in the Options dialog.
All the raster geographic data files that you want to import and use in an Atoll document must have the same coordinate
system. You cannot work with raster geographic data files with different coordinate systems in the same document.
Note:
• If you import vector geographic data (e.g., traffic, measurements, etc.) with different
coordinate systems, it is possible to convert the coordinate systems of these data into the
projection coordinate system of your Atoll document.
The projection coordinate system is used to keep the coordinates of sites (radio network data) consistent with the
geographic data.
When you import a raster geographic data file, Atoll reads the geo-referencing information from the file (or from its header
file, depending on the geographic data file format), i.e., its Northwest pixel, to determine the coordinates of each pixel. Atoll
does not use any coordinate system during the import process. However, the geo-referencing information of geographic
data files are considered to be provided in the projection coordinate system of the document.
1.1.2.2 Display Coordinate System

The display coordinate system is the coordinate system used for the display, e.g., in dialogs, in the Map window rulers, in
the status bar, etc. The coordinates of each pixel of geographic data are converted to the display coordinate system from
the projection coordinate system for display. The display coordinate system is also used for sites (radio network data). You
can set the display coordinate system of your document in the Options dialog.
If you import sites data, the coordinate system of the sites must correspond to the display coordinate system of your Atoll
document.
If you change the display coordinate system in a document which is not connected to a database, the coordinates of all
the sites are converted to the new display system.
Note:
• If the coordinate systems of all your geographic data files and sites (radio network data) are
the same, you do not have to define the projection and display coordinate systems
separately. By default, the two coordinate systems are the same.
1.1.2.3 Internal Coordinate Systems

The internal coordinate systems are the projection and the display coordinate systems stored in a database. The projection
and display coordinate systems set by the administrator in the central Atoll project are stored in the database when the
database is created, and cannot be modified by users. Only the administrator can modify the internal coordinate systems
manually by editing the entries in the CoordSys and the Units tables. All Atoll documents opened from a database will have
the internal coordinate systems of the database as their default projection and display coordinate systems.
When exporting an Atoll project to a database, the currently chosen display coordinate system becomes the internal
display coordinate system for the database, and the currently chosen projection coordinate system becomes the internal
projection coordinate system for the database.
26 AT271_TRG_E6 © Forsk 2009

Although Atoll stores both the coordinate systems in the database, i.e., the projection and the display coordinate systems,
the only relevant coordinate system for the database is the internal display coordinate system because this coordinate
system is the one used for the coordinates of sites (radio network data).
Users working on documents connected to a database can modify the coordinate systems in their documents locally, and
save these changes in their documents, but they cannot modify the coordinate systems stored in the database.
If you change the display coordinate system in a document which is not connected to a database, the coordinates of all
the sites are converted to the new display system.
If you change the display coordinate system in a document which is connected to a database, the coordinates of all the
sites are converted to the new coordinate system in the Atoll document locally but not in the database because the internal
coordinate systems cannot be changed.
Atoll uses the internal coordinates systems in order to keep the site coordinates consistent in the database which is usually
accessed by a large number of users in a multi-user environment.
1.1.3 File Formats

The Coordsystems folder located in the Atoll installation directory contains all the coordinate systems, both geographic
and cartographic, offered in the tool. Coordinate systems are grouped by regions. A catalogue per region and a "Favour-
ites" catalogue are available in Atoll. The Favourites catalogue is initially empty and can be filled by the user by adding
coordinate systems to it. Each catalogue is described by an ASCII text file with .cs extension. In a .cs file, each coordinate
system is described in one line. The line syntax for describing a coordinate system is:
Code = "Name of the system"; Unit Code; Datum Code; Projection Method Code,
Projection Parameters; "Comments"
Examples:
4230 = "ED50"; 101; 230; 1; "Europe - west"
32045 = "NAD27 / Vermont"; 2; 267; 6, -72.5, 42.5, 500000, 0, 0.9999643; "United

States - Vermont"
You should keep the following points in mind when editing or creating .cs files:
• The identification code enables Atoll to differentiate coordinates systems. In case you create a new coordinate
system, its code must be an integer value higher than 32767.
• When describing a new datum, you must enter the ellipsoid code and parameters instead of the datum code in
brackets. There can be 3 to 7 parameters defined in the following order: Dx, Dy, Dz, Rx, Ry, Rz, S. The syntax of
the line in the .cs file will be:
Code = "Name of the system"; Unit Code; {Ellipsoid Code, Dx, Dy, Dz, Rx, Ry,
Rz, S}; Projection Method Code, Projection Parameters; "Comments"
• There can be up to seven projection parameters. These parameters must be ordered according to the parameter
index (see "Projection Parameter Indices" on page 30). Parameter with index 0 is the first one. Projection param-
eters are delimited by commas.
• For UTM projections, you must provide positive UTM zone numbers for north UTM zones and negative numbers
for south UTM zones.
• You can add all other information as comments (such as usage or region).
Codes of units, data, projection methods, and ellipsoids, and projection parameter indices are listed in the tables below.
1.1.3.1 Unit Codes

Code Cartographic Units Code Geographic Units
0 Metre 100 Radian
1 Kilometre 101 Degree
2 Foot 102 Grad
3 Link 103 ArcMinute
4 Chain 104 ArcSecond
5 Yard
6 Nautical mile
7 Mile
-1 Unspecified -1 Unspecified
© Forsk 2009 AT271_TRG_E6 27

1.1.3.2 Datum Codes

Code Datum Code Datum
121 Greek Geodetic Reference System 1987 260 Manoca
125 Samboja 261 Merchich
126 Lithuania 1994 262 Massawa
130 Moznet (ITRF94) 263 Minna
131 Indian 1960 265 Monte Mario
201 Adindan 266 M'poraloko
202 Australian Geodetic Datum 1966 267 North American Datum 1927
203 Australian Geodetic Datum 1984 268 NAD Michigan
204 Ain el Abd 1970 269 North American Datum 1983
205 Afgooye 270 Nahrwan 1967
206 Agadez 271 Naparima 1972
207 Lisbon 272 New Zealand Geodetic Datum 1949
208 Aratu 273 NGO 1948
209 Arc 1950 274 Datum 73
210 Arc 1960 275 Nouvelle Triangulation Française
211 Batavia 276 NSWC 9Z-2
212 Barbados 277 OSGB 1936
213 Beduaram 278 OSGB 1970 (SN)
214 Beijing 1954 279 OS (SN) 1980
215 Reseau National Belge 1950 280 Padang 1884
216 Bermuda 1957 281 Palestine 1923
217 Bern 1898 282 Pointe Noire
218 Bogota 283 Geocentric Datum of Australia 1994
219 Bukit Rimpah 284 Pulkovo 1942
221 Campo Inchauspe 285 Qatar
222 Cape 286 Qatar 1948
223 Carthage 287 Qornoq
224 Chua 288 Loma Quintana
225 Corrego Alegre 289 Amersfoort
226 Cote d'Ivoire 290 RT38
227 Deir ez Zor 291 South American Datum 1969
228 Douala 292 Sapper Hill 1943
229 Egypt 1907 293 Schwarzeck
230 European Datum 1950 294 Segora
231 European Datum 1987 295 Serindung
232 Fahud 296 Sudan
233 Gandajika 1970 297 Tananarive 1925
234 Garoua 298 Timbalai 1948
235 Guyane Francaise 299 TM65
236 Hu Tzu Shan 300 TM75
237 Hungarian Datum 1972 301 Tokyo
238 Indonesian Datum 1974 302 Trinidad 1903
239 Indian 1954 303 Trucial Coast 1948
240 Indian 1975 304 Voirol 1875
241 Jamaica 1875 305 Voirol Unifie 1960
242 Jamaica 1969 306 Bern 1938
243 Kalianpur 307 Nord Sahara 1959
244 Kandawala 308 Stockholm 1938
245 Kertau 309 Yacare
247 La Canoa 310 Yoff
248 Provisional South American Datum 1956 311 Zanderij
28 AT271_TRG_E6 © Forsk 2009

249 Lake 312 Militar-Geographische Institut

250 Leigon 313 Reseau National Belge 1972
251 Liberia 1964 314 Deutsche Hauptdreiecksnetz
252 Lome 315 Conakry 1905
253 Luzon 1911 322 WGS 72
254 Hito XVIII 1963 326 WGS 84
255 Herat North 901 Ancienne Triangulation Française
256 Mahe 1971 902 Nord de Guerre
NAD 1927 Guatemala/Honduras/Salvador
257 Makassar 903
(Panama Zone)
258 European Reference System 1989
1.1.3.3 Projection Method Codes

Code Projection Method Code Projection Method
0 Undefined 8 Oblique Stereographic
1 No projection > Longitude / Latitude 9 New Zealand Map Grid
2 Lambert Conformal Conical 1SP 10 Hotine Oblique Mercator
3 Lambert Conformal Conical 2SP 11 Laborde Oblique Mercator
4 Mercator 12 Swiss Oblique Cylindrical
5 Cassini-Soldner 13 Oblique Mercator
6 Transverse Mercator 14 UTM Projection
7 Transverse Mercator South Oriented
1.1.3.4 Ellipsoid Codes

Code Name Major Axis Minor Axis
1 Airy 1830 6377563.396 6356256.90890985
2 Airy Modified 1849 6377340.189 6356034.44761111
3 Australian National Spheroid 6378160 6356774.71919531
4 Bessel 1841 6377397.155 6356078.96261866
5 Bessel Modified 6377492.018 6356173.50851316
6 Bessel Namibia 6377483.865 6356165.38276679
7 Clarke 1858 6378293.63924683 6356617.98173817
8 Clarke 1866 6378206.4 6356583.8
9 Clarke 1866 Michigan 6378693.7040359 6357069.45104614
10 Clarke 1880 (Benoit) 6378300.79 6356566.43
11 Clarke 1880 (IGN) 6378249.2 6356515
12 Clarke 1880 (RGS) 6378249.145 6356514.86954978
13 Clarke 1880 (Arc) 6378249.145 6356514.96656909
14 Clarke 1880 (SGA 1922) 6378249.2 6356514.99694178
15 Everest 1830 (1937 Adjustment) 6377276.345 6356075.41314024
16 Everest 1830 (1967 Definition) 6377298.556 6356097.5503009
17 Everest 1830 (1975 Definition) 6377301.243 6356100.231
18 Everest 1830 Modified 6377304.063 6356103.03899315
19 GRS 1980 6378137 6356752.31398972
20 Helmert 1906 6378200 6356818.16962789
21 Indonesian National Spheroid 6378160 6356774.50408554
22 International 1924 6378388 6356911.94612795
23 International 1967 6378160 6356774.71919530
24 Krassowsky 1940 6378245 6356863.01877305
25 NWL 9D 6378145 6356759.76948868
26 NWL 10D 6378135 6356750.52001609
27 Plessis 1817 6376523 6355862.93325557
28 Struve 1860 6378297 6356655.84708038
© Forsk 2009 AT271_TRG_E6 29

29 War Office 6378300.583 6356752.27021959

30 WGS 84 6378137 6356752.31398972
31 GEM 10C 6378137 6356752.31398972
32 OSU86F 6378136.2 6356751.51667196
33 OSU91A 6378136.3 6356751.61633668
34 Clarke 1880 6378249.13884613 6356514.96026256
35 Sphere 6371000 6371000
1.1.3.5 Projection Parameter Indices

Index Projection Parameter Index Projection Parameter
0 UTM zone number 4 Scale factor at origin
0 Longitude of origin 4 Latitude of 1st parallel
1 Latitude of origin 5 Azimuth of central line
2 False Easting 5 Latitude of 2nd parallel
3 False Northing 6 Angle from rectified to skewed grid
1.1.4 Creating a Coordinate System

Atoll provides a large catalogue of default coordinate systems. Nevertheless, it is possible to add the description of
geographic and cartographic coordinate systems. New coordinate systems can be created from scratch or initialised on
the basis of an existing one.
To create a new coordinate system from scratch:
1. Select Tools > Options. The Options dialog opens.
2. Select the Coordinates tab.
3. Click the browse button (...) on the right of the Projection field.
4. Click the New button. The Coordinate System dialog opens.
5. In the Coordinate System dialog,
a. Select the coordinate systems catalogue to which you want to add the new coordinate system.
b. In the General properties section: Enter a name for the new coordinate system, select a unit. You can also
enter any comments about its usage. Atoll assigns the code automatically.
c. In the Category section: Select the type of coordinate system. Enter the longitude and latitude for a geograph-
ic coordinate system, or the type of projection and its set of associated parameters for a cartographic coordi-
nate system (false easting and northing, and the first and second parallels).
d. In the Geo section: Specify the meridian and choose a datum for the coordinate system. The associated el-
lipsoid is automatically selected. You can also describe a geodetic datum by selecting "..." in the Datum list.
In this case, you must provide parameters (Dx, Dy, Dz, Rx, Ry, Rz, and S) needed for the transformation of
the datum into WGS84, and an ellipsoid.
6. Click OK. The new coordinate system is added to the selected coordinate system catalogue.
To create a new coordinate system based on an existing system, select a coordinate system in the Coordinate Systems
dialog before clicking New in step 4. The new coordinate system is initialised with the values of the selected coordinate
system.
1.2 Units
1.2.1 Power Units
Depending on the working environment, there can be either one or two types of units for transmission and reception
powers. If you are working with stand-alone documents, i.e., documents not connected to databases, there is only one unit
used in Atoll:
• Display power units
If you are working in a multi-user environment, Atoll uses two type of units:
• Display power units for the Atoll document
• Internal power units for the database
The display units are used for the display in dialogs and tables, e.g., reception thresholds (coverage prediction properties,
etc.), and received signal levels (measurements, point analysis, coverage predictions etc.). You can set the display units
for your document in the Options dialog.
30 AT271_TRG_E6 © Forsk 2009

The internal units are the power units stored in a database. The power units set by the administrator in the central Atoll
project are stored in the database when the database is created, and cannot be modified by users. Only the administrator
can modify the internal units manually by editing the entries in the Units tables. All Atoll documents opened from a data-
base will have the internal units of the database as their default power units.
Users working on documents connected to a database can modify the units in their documents locally, and save these
changes in their documents, but they cannot modify the units stored in the database.
1.2.2 Length Units

There are two types of units for distances, heights, and offsets:
• Display length units
• Internal length units
The display length units are used to display distances, heights, and offsets in dialogs, tables, and the status bar. You can
set the display units for your document in the Options dialog.
The internal unit for lengths is metre for all Atoll documents whether they are connected to databases or not. The internal
unit is not stored in the databases. The internal unit cannot be changed.
1.3 BSIC Format

Depending on the working environment, there can be either one or two types of BSIC formats. If you are working with
stand-alone documents, i.e., documents not connected to databases, there is only one BSIC format:
• Display BSIC format
If you are working in a multi-user environment, Atoll uses two type of formats:
• Display BSIC format for the Atoll document
• Internal BSIC format for the database
The display format is used for the display in dialogs and tables. You can set the display format for your document from the
Transmitters folder’s context menu.
The internal format is the BSIC format stored in a database. The BSIC format set by the administrator in the central Atoll
project is stored in the database when the database is created, and cannot be modified by users. Only the administrator
can modify the internal format manually by editing the corresponding entry in the Units tables. All Atoll documents opened
from a database will have the internal format of the database as their default BSIC format.
Users working on documents connected to a database can modify the format in their documents locally, and save this
change in their documents, but they cannot modify the format stored in the database.
© Forsk 2009 AT271_TRG_E6 31

32 AT271_TRG_E6 © Forsk 2009

Chapter 2
Geographic and Radio Data
This chapter defines the different types of data with which you can work in Atoll. These data can be
geographic data, such as maps, and radio network data, such as sites, antennas, other equipment and
parameters.
Atoll
Atoll
Microwave
RF PlanningLink
Software
34 AT271_TRG_E6 © Forsk 2009

Chapter 2: Geographic and Radio Data
2 Geographic and Radio Data

2.1 Geographic Data
2.1.1 Data Type
Atoll manages several geographic data types; DTM (Digital Terrain Model), clutter (Land-Use), scanned images, vector
data, traffic data, population, and any other generic data.
2.1.1.1 Digital Terrain Model (DTM)

The DTM (Digital Terrain Model or height) files describe the ground elevation above the sea level. DTM files supported by
Atoll are 16 bits/pixel relief maps in .tif, .bil, Planet© and Erdas Imagine formats and 8 bits/pixel relief maps in .tif, .bil, Erdas
Imagine and .bmp formats. DTM maps are taken into account in path loss calculations by Atoll propagation models.
DTM file provides altitude value (z stated in metre) on evenly spaced points. Abscissa and ordinate axes are respectively
oriented in right and downwards directions. Space between points is defined by pixel size (P stated in metre). Pixel size
must be the same in both directions. First point given in the file corresponds to the centre of the upper-left pixel of the map.
This point refers to the northwest point geo-referenced by Atoll. Four points (hence, four altitude values) are necessary to
describe a “bin”; these points are bin vertices.
Figure 2.1: Digital Terrain Model
Therefore, a n*n bin DTM file requires (n)2 points (altitude values).
Figure 2.2: Schematic view of a DTM file
Notes:
• Altitude values differ within a bin. Method used to calculate altitudes is described in the
Path loss calculations: Altitude determination part. Concerning DTM map display, Atoll
takes altitude of the southwest point of each bin to determine its colour.
• In most documents, Digital Elevation Model (DEM) and Digital Terrain Model (DTM) are
differentiated and do not have the same meaning. By definition, DEM refers to altitude
above sea level including, both, ground and clutter while DTM just corresponds to the
ground height above sea level. In Atoll, the DEM term may be used instead of DTM term.
© Forsk 2009 AT271_TRG_E6 35

2.1.1.2 Clutter (Land Use)

You may import two types of clutter files in ATL documents. These files indicate either the clutter class or the clutter height
on each bin of the map.
2.1.1.2.1 Clutter Classes

Atoll supports 8 bits/pixel (255 classes) raster maps in .tif, .bil, .bmp, Erdas Imagine formats or 16 bits/pixel raster maps
in Planet© format. This kind of clutter file describes the land cover (dense urban, buildings, residential, forest, open,
villages, …). A grid map represents ground and each bin of the map is characterised by a code corresponding to a main
type of cover (a clutter class). Atoll automatically lists all the clutter classes of the map. It is possible to specify an average
clutter height for each clutter class manually during the map description step. Clutter maps are taken into account in path
loss calculations by Atoll propagation models.
Clutter file provides a clutter code per bin. Bin size is defined by pixel size (P stated in metre). Pixel size must be the same
in both directions. Abscissa and ordinate axes are respectively oriented in right and downwards directions. First point given
in the file corresponds to the centre of the upper-left pixel of the image. This point refers to the northwest point geo-refer-
enced by Atoll.
Figure 2.3: Clutter Classes
Therefore, a n*n bin Clutter file requires (n)2 code values.

Note:
• The clutter code is the same inside a bin.
2.1.1.2.2 Clutter Heights

Files supported by Atoll for clutter heights are 8 or 16 bits/pixel raster maps in .tif, .bil and Erdas Imagine formats. The file
provides clutter height value on evenly spaced points. Abscissa and ordinate axes are respectively oriented in right and
downwards directions. Space between points is defined by pixel size (P in metre). Pixel size must be the same in both
directions. First point given in the file corresponds to the centre of the upper-left pixel of the map. This point refers to the
northwest point geo-referenced by Atoll.
These maps are taken into account in path loss calculations by Atoll propagation models.
Note:
• Atoll considers the clutter height of the nearest point in calculations (see Path loss
calculations: Clutter determination part). For map display, Atoll takes clutter height of the
southwest point of each bin to determine its colour.
2.1.1.3 Traffic Data

Atoll offers different kinds of traffic data:
2.1.1.3.1 Environment Traffic Maps

Atoll supports 8 bits/pixel (256 class) traffic raster maps in .tif, .bil, .bmp, Erdas Imagine formats. These maps provide
macroscopic traffic estimation. Each pixel is assigned an environment class, which is a list of user profiles with a defined
mobility type and a density.
2.1.1.3.2 User Profile Traffic Maps

Atoll supports vector traffic maps with .dxf®, Planet©, .shp, .mif and .agd formats. These maps are detailed traffic estima-
tions (lines, polygons or points carrying a specific traffic). Each polygon, line or point is assigned a specific user profile with
associated mobility type and density. They can be built from population density vector maps.
2.1.1.3.3 Live Traffic Maps

Atoll supports maps with .agd format. This kind of map is based on the network feedback. It provides actual information
on connections (and not just subscriber estimation) from the network. It is built from a coverage by transmitter prediction
36 AT271_TRG_E6 © Forsk 2009

study that defines sector boundaries for the traffic distribution in each sector. In UMTS, CDMA2000 and IS95-CDMA, either
data rates or the number of users per service are indicated for each transmitter service area. In GSM/TDMA, Atoll expects
a number of Erlangs in case of voice service and data rate values for packet-switched services for each transmitter service
area.
2.1.1.3.4 User Density Traffic Maps

This kind of map is only available in GSM/TDMA documents. Atoll supports 16 and 32 bits/pixel traffic raster maps in .tif,
.bil, .bmp, Planet© and Erdas Imagine formats. This map is also based on the network feedback as it deals with network
users information as well. Each pixel is assigned a number of users with a given service, terminal and mobility type.
In GSM documents, traffic maps are taken into account for traffic analysis and network dimensioning. In UMTS,
CDMA2000 and IS95-CDMA documents, they are used by the Monte-Carlo simulator to model user distributions and eval-
uate related network parameters (cell power, mobile terminal power, …).
2.1.1.4 Vector Data

These data represent either polygons (regions, etc.), lines (roads, coastlines, etc.) or points (towns, etc.). Atoll supports
vector data files in .dxf®, Planet©, .shp, .mif and .agd formats. These maps are only used for display and provide informa-
tion about the geographic environment.
2.1.1.5 Scanned Images

These geographic data include the road maps and the satellite images. They are only used for display and provide infor-
mation about the geographic environment. Atoll supports scanned image files in .tif (1, 4, 8, 24-bits/pixel), .bil (1, 4, 8, 24-
bits/pixel), Planet© (1, 4, 8, 24-bits/pixel), .bmp (1-24-bits/pixel), Erdas Imagine (1, 4, 8, 24-bits/pixel) and .ecw (24-bits/
pixel) formats.
2.1.1.6 Population
Atoll deals with vector population files (polygons, lines or points) in .mif, .shp and .agd formats or 8, 16, 32 bits/pixel raster
population files in .tif, .bil, .bmp and Erdas Imagine formats. Population map describes the population distribution. They
are considered in clutter statistics and in coverage prediction reports.
2.1.1.7 Other Geographic Data

It is possible to import generic geographic data types, other than those listed above, (Customer density, revenue density,
…) in Atoll. These data can be either vector files in .mif, .shp and .agd formats or 8, 16, 32 bits/pixel raster files in .tif, .bil,
.bmp and Erdas Imagine formats. These maps are taken into account in clutter statistics and in coverage prediction
reports.
The ArcView Grid format (.txt) is an ASCII format dedicated to define raster maps. It may be used to export any raster map
such as DTM, images, Clutter Classes and/or Heights, Population, Generic data maps and even coverage predictions.
The contents of an ArcView Grid file are in ASCII and consist of a header, describing the content, followed by the content
in the form of cell values.
Notes:
• The minimum resolution supported by Atoll is 1m for any raster maps, excepted for
scanned images, for which it is unlimited.
• DTM and clutter map resolution must be an integer.
• All the raster maps you want to import in an ATL document must be represented in the
same projection system.
2.1.2 Supported Geographic Data Formats

Atoll offers Import/Export filters for the most commonly used geographic data formats. The different filters are:
Import/
File format Can contain Georeferenced
Export
DTM, Clutter classes and heights, Traffic,
.bil Both Yes via .hdr files
Image, Population, Other data
DTM, Clutter classes and heights, Traffic, Yes via associated .tfw files if
.tif Both
Image, Population, Other data they exist
Planet© Both DTM, Clutter classes, Image, Vector data Yes via index files
DTM, Clutter heights, Clutter classes, Traffic,
.bmp Both Yes via .bpw (or .bmw) files
.dxf® Import Only Vector data, Vector traffic Yes
Vector data, Vector traffic, Population, Other
.shp Both Yes
data
Vector data, Vector traffic, Population, Other
.mif/.mid Both Yes
data
© Forsk 2009 AT271_TRG_E6 37

DTM, Clutter classes and heights, Traffic,

Erdas Imagine Import Only Yes
DTM, Clutter classes and heights, Traffic, Yes automatically embedded in
ArcView Grid Export Only
Image, Population, Other data the data file
Vector data, Vector traffic, Population, Other Yes automatically embedded in
.agd Both
data the data file
Vertical Mapper DTM, Clutter classes and heights, Traffic, Yes automatically embedded in
Both
(.grd, .grc) Image, Population, Other data the data file
.ecw Import Only Images Yes via ers file (not mandatory)
Note:
• The .wld files may be used as georeferencement file for any type of binary raster file.
• Tiled .tif format is not supported.
Thus, to sum up, you can import:

• DTM files in .tif (16-bits, 8-bits), .bil (16-bits, 8-bits), Planet© (16-bits), Erdas Imagine (16-bits, 8-bits), Vertical
Mapper (.grd, .grc) and .bmp (8-bits) formats.
• Clutter heights files in .tif (16-bits, 8-bits), .bil (16-bits, 8-bits), Planet© (16-bits), Erdas Imagine (16-bits, 8-bits),
Vertical Mapper (.grd, .grc) and .bmp (8-bits) formats.
• Clutter classes and traffic raster files in .tif (8-bits), .bil (8-bits), .bmp (8-bit), Erdas Imagine (8-bits) and Vertical
Mapper (.grd, .grc) and Planet© format (16-bits) are also supported.
• Vector data files in .dxf®, Planet©, .shp, .mif and .agd formats.
• Vector traffic files in .dxf®, Planet©, .shp, .mif and .agd formats.
• Scanned image files in .tif (1, 4, 8, 24-bits), .bil (1, 4, 8, 24-bits), Planet© (1, 4, 8, 24-bits), .bmp (1-24-bits), Erdas
Imagine (1, 4, 8, 24-bits), Vertical Mapper (.grd, .grc) and .ecw (Enhanced Compressed Wavelet) (24 bits) formats.
• Population files in .mif, .shp, .agd, .tif (8, 16, 32-bits), .bil (8, 16, 32-bits), .bmp (8, 32-bits), Vertical Mapper (.grd,
.grc) and Erdas Imagine (8, 16, 32-bits) formats.
• Other generic data types in .mif, .shp, .agd, .tif (8, 16, 32-bits), .bil (8, 16, 32-bits), .bmp (8, 32-bits), Vertical
Mapper (.grd, .grc) and Erdas Imagine (8, 16, 32-bits) formats.
Note:
• It is possible to import Packbit, FAX-CCITT3 and LZW compressed .tif files. However, in
case of DTM and clutter, we recommend not to use compressed files in order to avoid poor
performances. If uncompressed files are too big, it is better to split them.
2.2 Radio Data

Atoll manages several radio data types; sites, transmitters, antennas, stations and hexagonal designs. Data definition in
Atoll is detailed hereafter.
2.2.1 Site
A site is a geographical point where one or several transmitters (multi-sectored site or station) equipped with antennas are
located.
2.2.2 Antenna
An antenna is a device used for transmitting or receiving electromagnetic waves.
2.2.3 Transmitter
A transmitter is a group of radio devices located at a site. Transmitters are equipped with antenna(s) and other equipment
such as feeder, tower mounted amplifiers (TMA) and BTS.
2.2.4 Repeater
A repeater is a device that receives, amplifies and transmits the radiated or conducted RF carrier both in downlink and
uplink. It comprises a donor side and a server side. The donor side receives the signal from a donor transmitter. This signal
may be carried by different types of links such as radio link, microwave link, or optic fibre. The server side transmits the
repeated signal.
38 AT271_TRG_E6 © Forsk 2009

2.2.5 Remote Antenna

The use of remote antennas allows antenna positioning at locations that would normally require prohibitively long runs of
feeder cable. A remote antenna is connected to the base station via an optic fibre. The main difference from a repeater is
that a remote antenna generates its own cell whereas a repeater extends the coverage of an existing cell.
2.2.6 Station
A station can represent one transmitter on a site or a group of transmitters on a same site sharing the same properties.
You can define station templates and build your network from stations instead of single transmitters.
2.2.7 Hexagonal Design

A hexagonal design is a group of stations created from the same station template.
2.2.8 GSM GPRS EGPRS Documents

2.2.8.1 TRX
A base station (transmitter) consists of several transceivers or TRXs. One TRX supports as many timeslots as the multi-
plexing factor defined in properties of your frequency band (8 timeslots in GSM networks). Three types of TRXs are
modelled in Atoll:
• The BCCH TRX type: carries the BCCH,
• The TCH TRX type: which is the default traffic carrier,
• The TCH_INNER TRX type: this TRX type is an inner traffic carrier.
2.2.8.2 Subcell
A subcell corresponds to a group of TRXs having the same radio characteristics, the same quality (C/I) requirements, and
common settings. A subcell is characterised by the ‘transmitter-TRX type’ pair. Each transmitter may have one or more
subcells. The most common configurations are the {BCCH, TCH} configuration or the {BCCH, TCH, TCH_INNER} one.
2.2.8.3 Cell Type

A cell type describes the subcells (types of TRXs) that a cell can use and their parameters, which can be different. In the
current Atoll version, the cell type definition must include a TRX type as the BCCH carrier (BCCH TRX type) and another
TRX type as the default traffic carrier (TCH TRX type). Only one TRX type carrying the broadcast and only one TRX type
carrying the default TCH are supported.
2.2.9 All CDMA and WiMAX BWA Documents

2.2.9.1 Cell
Cell comprises the carrier characteristics of a transmitter. Cell is characterised by the ‘transmitter-carrier’ pair. The trans-
mitter-carrier pair must be unique.
© Forsk 2009 AT271_TRG_E6 39

40 AT271_TRG_E6 © Forsk 2009

Chapter 3
File Formats
Atoll supports a set of file formats for each type of data, may it be geographic data or calculation results.
This chapter contains details of these file formats, their usage, availability, and limitations.
Atoll
Atoll
Microwave
RF PlanningLink
Software
42 AT271_TRG_E6 © Forsk 2009

Chapter 3: File Formats
3 File Formats
3.1 BIL Format
Band Interleaved by Line is a method of organizing image data for multi-band images. It is a schema for storing the actual
pixel values of an image in a file. The pixel data is typically preceded by a file header that contains auxiliary data about the
image, such as the number of rows and columns in the image, a colour map, etc. .bil data stores pixel information band
by band for each line, or row, of the image. Although .bil is a data organization schema, it is treated as an image format.
An image description (number of rows and columns, number of bands, number of bits per pixel, byte order, etc.) has to be
provided to be able to display the .bil file. This information is included in the header .hdr file associated with the .bil file. A
.hdr file has the same name as the .bil file it refers to, and should be located in the same directory as the source file. The
.hdr structure is simple; it is an ASCII text file containing eleven lines. You can open a .hdr file using any ASCII text editor.
Atoll supports the following objects in .bil format:
• Digital Terrain Model (8 or 16 bits)
• Clutter heights (8 or 16 bits)
• Clutter classes and Environment traffic maps (8 bits)
• Traffic density maps (16 or 32 bits)
• Raster images (1, 4, 8, 24 bits)
• Population maps (8, 16, 32 bits)
• Other generic geographic data (8, 16, 32 bits)
• Path loss or received signal level value matrices (16 bits)
3.1.1 HDR Header File

3.1.1.1 Description
The header file is a text file that describes how data are organised in the .bil file. The header file is made of rows, each
row having the following format:
keyword value
where ‘keyword’ corresponds to an attribute type, and ‘value’ defines the attribute value.
Keywords required by Atoll are described below. Other keywords are ignored.
nrows Number of rows in the image.

ncols Number of columns in the image.
nbands Number of spectral bands in the image, (1 for DTM data and
8 bit pictures).
nbits Number of bits per pixel per band; 8 or 16 for DTMs or
Clutter heights (altitude in metres), 8 for clutter class-
es file (clutter code), 16 for path loss matrices (path
loss in dB, field value in dBm, dBµV and DBµV/m).
byteorder Byte order in which image pixel values are stored. Accept-
ed values are M (Motorola byte order) or I (Intel byte or-
der).
layout Must be ‘bil’.
skipbytes Byte to be skipped in the image file in order to reach the
beginning of the image data. Default value is 0.
ulxmap x coordinate of the centre of the upper-left pixel.
ulymap y coordinate of the centre of the upper-left pixel.
xdim x size in metre of a pixel.
ydim y size in metre of a pixel.
Four additional keywords may be optionally managed.
datatype Type of data read (in addition to the length)
It can be:
© Forsk 2009 AT271_TRG_E6 43

I1 Integer 1 bit
I2 Integer 2 bits
I4 Integer 4 bits
I8 Integer 8 bits
I16 Integer 16 bits
I32 Integer 32 bits
R32 Real 32 bits
R64 Real 64 bits
RGB24 Integer 3 colour components on 24 bits
By default, integer data types are chosen with respect to the pixel length (nbits).
valueoffset Real value to be added to the read value (Vread)

valuescale Scaling factor to be applied to the read value
So, we have V = V read  valuescale + valueoffset
nodatavalue Value corresponding to “NO DATA”
3.1.1.2 Samples
Here, the data is 20m.
3.1.1.2.1 Digital Terrain Model

nrows 1500
ncols 1500
nbands 1
nbits 8 or 16
byteorder M
layout bil
skipbytes 0
ulxmap 975000
ulymap 1891000
xdim 20.00
ydim 20.00
3.1.1.2.2 Clutter Classes File

nrows 1500
ncols 1500
nbands 1
nbits 8
byteorder M
layout bil
skipbytes 0
ulxmap 975000
ulymap 1891000
xdim 20.00
ydim 20.00
3.1.1.2.3 BIL File

.bil files are usually binary files without header. Data are stored starting from the Northwest corner of the area. The skip-
bytes value defined in the header file allows to skip records if the data do not start at the beginning of the file.
44 AT271_TRG_E6 © Forsk 2009

3.2 TIF Format

Tagged Image File Format graphics filter supports all image types (monochrome, greyscale, palette colour, and RGB full
colour images) and Packbit, LZW or fax group 3-4 compressions. .tif files are not systematically geo-referenced. You have
to enter spatial references of the image manually during the import procedure (x and y-axis map coordinates of the centre
of the upper-left pixel, pixel size); an associated file with .tfw extension will be simultaneously created with the same name
and in the same directory as the .tif file it refers to. Atoll will then use the .tfw file during the import procedure for an auto-
matic geo-referencing.
Note:
• Atoll also supports .tif files using the Packbit, FAX-CCITT3 and LZW compression modes.
You can modify the colour palette convention used by Atoll when exporting .tif files. This can be helpful when working on
.tif files exported by Atoll in other tools. In the default palette, the first colour indexes represent the useful information and
the remaining colour indexes represent the background. It is possible to export .tif files with a palette which defines the
background colour at the colour index 0, and then the colour indexes necessary to represent useful information. Add the
following lines in the Atoll.ini file to set up the new palette convention:
[TiffExport]
PaletteConvention=Gis
Please refer to the Administrator Manual for more details about the Atoll.ini file.
Notes:
• Using compressed geo data formats (compressed .tif, Erdas Imagine, or .ecw) can cause
performance loss due to real-time decompression. However, you can recover this loss in
performance by:
- Either, hiding the status bar, which provides geographic data information in real time, by
unchecking the Status Bar item in the View menu.
- Or, not displaying some of the information, such as altitude, clutter class and clutter
height, in the status bar. This can be done through the Atoll.ini file, by adding the following
lines:
[StatusBar]
DisplayZ=0
DisplayClutterClass=0
DisplayClutterHeight=0
• You can also save the produced map in an uncompressed format.
• Please refer to the Administrator Manual for more details about the Atoll.ini file.
Atoll supports the following objects in .tif format:

• Digital Terrain Model (8 or 16 bits)
• Clutter heights (8 or 16 bits)
• Clutter classes and Environment traffic maps (8 bits)
• Traffic density maps (16 or 32 bits)
• Population maps (8, 16, 32 bits)
• Other generic geographic data (8, 16, 32 bits)
.tfw file contains the spatial reference data of an associated .tif file. The .tfw file structure is simple; it is an ASCII text file
that contains six lines. You can open a .tfw file using any ASCII text editor.
3.2.1 TFW Header File

The .tfw files contain spatial reference data for the associated .tif file. The header file is a text file that describes how data
are organised in the .tif file. You can open a .tfw file using any ASCII text editor. The header file consists of six lines, with
each line having the following description:
Line Description
1 x dimension of a pixel in map units
2 amount of translation
3 amount of rotation
4 negative of the y dimension of a pixel in map units
5 x-axis map coordinate of the centre of the upper-left pixel
6 y-axis map coordinate of the centre of the upper-left pixel
© Forsk 2009 AT271_TRG_E6 45

Note:
• Atoll does not use the lines 2 and 3 when importing a .tif format geographic file.
3.2.2 Sample
3.2.2.1 Clutter Classes File
100.00
0.00
0.00
-100.00
60000.00
2679900.00
3.3 BMP Format

This is the MS-Windows standard format. It holds black & white, 16-, 256- and True-colour images. The palletized 16-
colour and 256-colour images may be compressed via run length encoding (though compressed .bmp files are quite rare).
The image data itself can either contain pointers to entries in a colour table or literal RGB values. .bmp files are not system-
atically geo-referenced. You have to enter spatial references of the image manually during the import procedure (x and y-
axis map coordinates of the centre of the upper-left pixel, pixel size). When exporting (saving) a .bmp file, an associated
file with .bpw extension is created with the same name and in the same directory as the .bmp file it refers to. Atoll stores
the georeferencing information in this file for future imports of the .bmp so that the .bpw file can be used during the import
procedure for automatic geo-referencing. Atoll also supports .bmw extension for the .bmp related world files.
Atoll supports the following objects in .tif format:
• Digital Terrain Model (8 bits)
• Clutter Heights (8 bits)
• Clutter classes and traffic density maps (8 bits)
• Population maps (8, 32 bits)
• Other generic geographic data (8, 32 bits)
3.3.1 BMP File Description

A .bmp file contains of the following data structures:
• BITMAPFILEHEADER bmfh Contains some information about the bitmap file (about the file, not
about the bitmap itself).
• BITMAPINFOHEADER bmih Contains information about the bitmap (such as size, colours, etc.).
• RGBQUAD aColors[] Contains a colour table.
• BYTE aBitmapBits[] Image data (whose format is specified by the bmih structure).
3.3.1.1 BMP File Structure

The following tables give exact information about the data structures. The Start-value is the position of the byte in the file
at which the explained data element of the structure starts, the Size-value contains the number of bytes used by this data
element, the Name column contains both generic name and the name assigned to this data element by the Microsoft API
documentation, and the Description column gives a short explanation of the purpose of this data element.
• BITMAPFILEHEADER (Header - 14 bytes):
Name
Start Size Description
Generic MS API
1 2 Signature bfType Must always be set to 'BM' to declare that this is a .bmp-file.
3 4 FileSize bfSize Specifies the size of the file in bytes.
7 2 Reserved1 bfReserved1 Unused. Must be set to zero.
9 2 Reserved2 bfReserved2 Unused. Must be set to zero.
Specifies the offset from the beginning of the file to the bitmap (raster)
11 4 DataOffset bfOffBits
data.
• BITMAPINFOHEADER (InfoHeader - 40 bytes):
46 AT271_TRG_E6 © Forsk 2009

Name
Generic MS API
Specifies the size of the BITMAPINFOHEADER structure, in
15 4 Size biSize
bytes (= 40 bytes).
19 4 Width biWidth Specifies the width of the image, in pixels.
23 4 Height biHeight Specifies the height of the image, in pixels.
Specifies the number of planes of the target device, must be
27 2 Planes biPlanes
set to zero or 1.
Specifies the number of bits per pixel.
1 = monochrome pallete. # of colours = 1
4 = 4-bit palletized. # of colours = 16
29 2 BitCount biBitCount
24 = 24-bit palletized. # of colours = 16M
Specifies the type of compression, usually set to zero.
0 = BI_RGB no compression
31 4 Compression biCompression
1 = BI_RLE8 8-bit RLE encoding
2 = BI_RLE4 4-bit RLE encoding
Specifies the size of the image data, in bytes. If there is no
35 4 ImageSize biSizeImage
compression, it is valid to set this element to zero.
39 4 XpixelsPerM biXPelsPerMeter Specifies the the horizontal pixels per meter.
43 4 YpixelsPerM biYPelsPerMeter Specifies the the vertical pixels per meter.
Specifies the number of colours actually used in the bitmap. If
47 4 ColoursUsed biClrUsed set to zero the number of colours is calculated using the
biBitCount element.
Specifies the number of colour that are 'important' for the
51 4 ColoursImportant biClrImportant
bitmap. If set to zero, all colours are considered important.
Note:
• biBitCount actually specifies the colour resolution of the bitmap. It also decides if there is a
colour table in the file and how it looks like.
- In 1-bit mode the colour table has to contain 2 entries (usually white and black). If a bit in
the image data is clear, it points to the first palette entry. If the bit is set, it points to the
second.
- In 4-bit mode the colour table must contain 16 colours. Every byte in the image data
represents two pixels. The byte is split into the higher 4 bits and the lower 4 bits and each
value of them points to a palette entry.
- In 8-bit mode every byte represents a pixel. The value points to an entry in the colour
table which contains 256 entries.
- In 24-bit mode three bytes represent one pixel. The first byte represents the red part, the
second the green and the third the blue part. There is no need for a palette because every
pixel contains a literal RGB-value, so the palette is omitted.
• RGBQUAD array (ColorTable):
Name
Generic MS API
1 1 Blue rgbBlue Specifies the blue part of the colour.
2 1 Green rgbGreen Specifies the green part of the colour.
3 1 Red rgbRed Specifies the red part of the colour.
4 1 Reserved rgbReserved Must always be set to zero.
Note:
• In a colour table (RGBQUAD), the specification for a colour starts with the blue byte, while
in a palette a colour always starts with the red byte.
• Pixel data:
The interpretation of the pixel data depends on the BITMAPINFOHEADER structure. It is important to know that the rows
of a .bmp are stored upside down meaning that the uppermost row which appears on the screen is actually the lowermost
row stored in the bitmap. Another important thing is that the number of bytes in one row must always be adjusted by
appending zero bytes to fit into the border of a multiple of four (16-bit or 32-bit rows).
© Forsk 2009 AT271_TRG_E6 47

3.3.1.2 BMP Raster Data Encoding

Depending on the image BitCount and on the Compression flag there are 6 different encoding schemes. In all of them,
• Pixels are stored bottom-up, left-to-right.
• Pixel lines are padded with zeros to end on a 32-bit boundary.
• For uncompressed formats every line will have the same number of bytes.
• Colour indices are zero based, meaning a pixel colour of 0 represents the first colour table entry, a pixel colour of
255 (if there are that many) represents the 256th entry. For images with more than 256 colours there is no colour
table.
BitCoun Compressio
Encoding type Remarks
t n
Every byte holds 8 pixels, its highest order bit representing the
leftmost pixel of these 8. There are 2 colour table entries. Some
1-bit readers assume that 0 is black and 1 is white. If you are storing
1 0
B&W images black and white pictures you should stick to this, with any other 2
colours this is not an issue. Remember padding with zeros up to a
32-bit boundary.
Every byte holds 2 pixels, its high order 4 bits representing the left of
those. There are 16 colour table entries. These colours do not have
4-bit
4 0 to be the 16 MS-Windows standard colours. Padding each line with
16 colour images
zeros up to a 32-bit boundary will result in up to 28 zeros = 7 'wasted
pixels'.
Every byte holds 1 pixel. There are 256 colour table entries.
8-bit
8 0 Padding each line with zeros up to a 32-bit boundary will result in up
256 colour images
to 3 bytes of zeros = 3 'wasted pixels'.
Every 2 bytes hold 1 pixel. There are no colour table entries.
16-bit
16 0 Padding each line with zeros up to a 16-bit boundary will result in up
High colour images
to 2 zero bytes.
Every 4 bytes hold 1 pixel. The first holds its red, the second its
24-bit green, and the third its blue intensity. The fourth byte is reserved
24 0
True colour images and should be zero. There are no colour table entries. No zero
padding necessary.
Pixel data is stored in 2-byte chunks. The first byte specifies the
number of consecutive pixels with the same pair of colour. The
second byte defines two colour indices. The resulting pixel pattern
4-bit will have interleaved high-order 4-bits and low order 4 bits
4 2
16 colour images (ABABA...). If the first byte is zero, the second defines an escape
code. The End-of-Bitmap is zero padded to end on a 32-bit
boundary. Due to the 16bit-ness of this structure this will always be
either two zero bytes or none.
The pixel data is stored in 2-byte chunks. The first byte specifies the
number of consecutive pixels with the same colour. The second byte
8-bit defines their colour indices. If the first byte is zero, the second
8 1
256 colour images defines an escape code. The End-of-Bitmap is zero padded to end
on a 32-bit boundary. Due to the 16bit-ness of this structure this will
always be either two zero bytes or none.
3.3.1.2.1 Raster Data Compression Descriptions

• 4-bit / 16 colour images
n (Byte 1) c (Byte 2) Description

n pixels to be drawn. The 1st, 3rd, 5th, ... pixels' colour is in c's high-order 4 bits, the
>0 any even pixels' colour is in c's low-order 4 bits. If both colour indices are the same, it
results in just n pixels of colour c.
0 0 End-of-line
0 1 End-of-Bitmap
Delta. The following 2 bytes define an unsigned offset in x and y direction (y being up).
0 2
The skipped pixels should get a colour zero.
The following c bytes will be read as single pixel colours just as in uncompressed files.
0 >=3
Up to 12 bits of zeros follow, to put the file/memory pointer on a 16-bit boundary again.
• 8-bit / 256 colour images
n (Byte 1) c (Byte 2) Description

>0 any n pixels of colour number c
48 AT271_TRG_E6 © Forsk 2009

0 0 End-of-line
0 1 End-of-Bitmap
Delta. The following 2 bytes define an unsigned offset in x and y direction (y being up).
0 2
The skipped pixels should get a colour zero.
The following c bytes will be read as single pixel colours just as in uncompressed files.
0 >=3
A zero follows, if c is odd, putting the file/memory pointer on a 16-bit boundary again.
3.3.2 BPW/BMW Header File Description

The header file is a text file that describes how data are organised in the .bmp file. The header file is made of rows, each
row having the following description:
Line Description
Atoll supports .bpw and .bmw header file extensions for Import, but exports headers with .bpw file extensions.
3.3.3 Sample
100.00
0.00
0.00
-100.00
60000.00
2679900.00
3.4 Generic Raster Header File (.wld)

.wld is a new Atoll specific header format that can be used for any raster data file for georeferencing. At the time of import
of any raster data file, Atoll can use the corresponding .wld file to read the georeferencing information related to the raster
data file. The .wld file contains the spatial reference data of any associated raster data file. The .wld file structure is simple;
it is an ASCII text file containing six lines. You can open a .wld file using any ASCII text editor.
3.4.1 WLD File Description

The .wld file is a text file that describes how data are organised in the associated raster data file. The header file is made
of rows, each row having the following description:
Line Description
3.4.2 Sample
100.00
© Forsk 2009 AT271_TRG_E6 49

0.00
0.00
-100.00
60000.00
2679900.00
3.5 DXF Format

Atoll is capable of importing and working with AutoCAD® drawings in the Drawing Interchange Format (DXF). .dxf files
can have ASCII or binary formats. But only the ASCII .dxf files can be used in Atoll.
.dxf files are composed of pairs of codes and associated values. The codes, known as group codes, indicate the type of
value that follows. .dxf files are organized into sections of records containing the group codes and their values. Each group
code and value is a separate line.
Each section starts with a group code 0 followed by the string, SECTION. This is followed by a group code 2 and a string
indicating the name of the section (for example, HEADER). Each section ends with a 0 followed by the string ENDSEC.
3.6 SHP Format

ESRI (Environmental Systems Research Institute, Inc.) ArcView® GIS Shapefiles have a simple, non-topological format
for storing geometric locations and attribute information of geographic features. A shapefile is one of the spatial data
formats that you can work with in ArcExplorer. .shp data files usually have associated .shx and .dbf files.
Among these three files:
• The .shp file stores the feature geometry
• The .shx file stores the index of the feature geometry.
• The .dbf (dBASE) file stores the attribute information of features. When a shapefile is added as a theme to a view,
this file is displayed as a feature table.
You can define mappings between the coordinate system used for the ESRI vector files, defined in the corresponding .prj
files, and Atoll. In this way, when you import a vector file, Atoll can detect the correct coordinate system automatically. For
more information about defining the mapping between coordinate systems, please refer to the Administrator Manual.
3.7 MIF Format

MapInfo Interchange Format (.mif) allows various types of data to be attached to a variety of graphical items. These ASCII
files are editable, easy to generate, and work on all platforms supported by MapInfo. Vector objects with a .mif extension
may be imported in Atoll.
Two files, a .mif and a .mid, contain MapInfo data. Graphics reside in the .mif file while the text contents are stored in the
.mid file. The text data is delimited with one row per record, and Carriage Return, Carriage Return plus Line Feed, or Line
Feed between lines. The .mif file has two sections, the file header and the data section. The .mid file is optional. When
there is no .mid file, all fields are blank.
You can find more information at http://www.mapinfo.com.
You can define mappings between the coordinate system used for the MapInfo vector files, defined in the corresponding
.mif files, and Atoll. In this way, when you import a vector file, Atoll can detect the correct coordinate system automatically.
For more information about defining the mapping between coordinate systems, please refer to the Administrator Manual.
3.8 TAB Format

TAB files (MapInfo Tables) are the native format of MapInfo. They actually consist of a number of files with extensions
such as .TAB, .DAT and .MAP. All of these files need to be present and kept together for the table to work. These are
defined as follows:
• .TAB: table structure in ASCII format
• .DAT: table data storage in binary format
• .MAP: storage of map objects in binary format
• .ID: index to the MapInfo graphical objects (.MAP) file
• .IND: index to the MapInfo tabular (DAT) file
You can find more information at http://www.mapinfo.com.
You can define mappings between the coordinate system used for the MapInfo vector files, defined in the corresponding
.mif files, and Atoll. In this way, when you import a vector file, Atoll can detect the correct coordinate system automatically.
For more information about defining the mapping between coordinate systems, please refer to the Administrator Manual.
TAB files are also supported as georeference information files for raster files (.bmp and .tif).
50 AT271_TRG_E6 © Forsk 2009

3.9 ECW Format

The Enhanced Compressed Wavelet file format is supported in Atoll. .ecw files are geo-referenced image files, which can
be imported in Atoll. This is an Open Standard wavelet compression technology, developed by Earth Resource Mapping,
which can compress images with up to a 100-to-1 compression ratio. Each compressed image file contains a header carry-
ing the following information about the image:
• The image size expressed as the number of cells across and down
• The number of bands (RGB images have three bands)
• The image compression rate
• The cell measurement units (meters, degrees or feet)
• The size of each cell in measurement units
• Coordinate space information (Projection, Datum etc.)
3.10 Erdas Imagine Format

Atoll supports Erdas Imagine data files in order to import DTM (8 or 16 bit/pixel), clutter (8 bit/pixel), traffic (8 bit/pixel), and
image (1-24 bit/pixel) files with the .img format. These files use the Erdas Imagine Hierarchical File Format (HFA) structure.
For any type of file, if there are pyramids (storage of different resolution layers), they are used to enhance performance
when decreasing the resolution of the display. Some aspects of working with Erdas Imagine format in Atoll are:
• Atoll supports uncompressed as well as compressed (or partially compressed) DTM .img files.
• You can create a .mnu file to improve the clutter class map loading.
• The colour-to-code association (raster maps) may be automatically imported from the .img file.
• These files are automatically geo-referenced, i.e., they do not require any additional file for geo-reference.
For image files, the number of supported bands is either 1 (colour palette is defined separately) or 3 (no colour palette but
direct RGB information for each pixel). In case of 3 bands, only 8 bit per pixel format is supported. Therefore, 8-bit images,
containing RGB information (three bands are provided: the first band is for Blue, the second one is for Green and the third
for Red), can be considered as 24 bit per pixel files. 32 bit per pixel files are not supported.
Notes:
• Using compressed geo data formats (compressed .tif, Erdas Imagine, or .ecw) can cause
performance loss due to real-time decompression. However, you can recover this loss in
performance by:
- Either, hiding the status bar, which provides geographic data information in real time, by
unchecking the Status Bar item in the View menu.
- Or, not displaying some of the information, such as altitude, clutter class and clutter
height, in the status bar. This can be done through the Atoll.ini file, by adding the following
lines:
[StatusBar]
DisplayZ=0
DisplayClutterClass=0
DisplayClutterHeight=0
• You can also save the produced map in an uncompressed format.
• Please refer to the Administrator Manual for more details about the Atoll.ini file.
3.11 Planet EV/Vertical Mapper Geographic Data Format

Vertical Mapper offers two types of grids:
• Numerical continuous grids, which contain numerical information (such as DTM), and are stored in files with the
.grd extension.
• Classified grids, which contain alphanumeric (characters) information, and are stored in files with the .grc exten-
sion.
Atoll is capable of supporting the Vertical Mapper Classified Grid (GRC) and Vertical Mapper Continuous Grid (GRD) file
formats in order to import and export:
• GRD: DTM, image, population, traffic density, and other data types.
• GRC: DTM, clutter classes, clutter heights, environment traffic, image, population, and other data types.
It is also possible to export coverage prediction studies in GRD and GRC formats.
This is the geographic data format used by Planet EV. So, it is possible to directly import geographic data from Planet EV
to Atoll using this format.
© Forsk 2009 AT271_TRG_E6 51

3.12 ArcView Grid Format

The ArcView Grid format (.txt) is an ASCII format dedicated to defining raster maps. It may be used to export any raster
map such as DTM, images, clutter classes and/or heights, population, other data maps, and even coverage predictions.
The contents of an ArcView Grid file are in ASCII and consist of a header, describing the content, followed by the content
in the form of cell values.
3.12.1 ArcView Grid File Description

The format of this file is as follows:
ncols XXX Number of columns of the grid (XXX columns).

nrows XXX Number of rows of the grid (XXX rows).
xllcenter XXX OR
xllcorner XXX Significant value relative to the bin centre or corner.
yllcenter OR
yllcorner XXX Significant value relative to the bin centre or corner.
cellsize XXX Grid resolution.
nodata_value XXX Optional value corresponding to no data (no information).
//Row 1 Top of the raster. Description of the first row. Syntax:
ncols number of values separated by spaces.
:
:
//Row N Bottom of the raster.
3.12.2 Sample
ncols 303
nrows 321
xllcorner 585300.000000
yllcorner 5615700.000000
cellsize 100.000000
nodata_value 0
...
3.13 Other Supported Geographic Data File Formats

Other than the .bil, .tif, Planet, .dxf, .shp, .mif, .img, and .ecw formats, Atoll supports 3 other formats.
The .ist and .dis formats are ASCII files used for Digital Terrain Model only. .ist images come from Istar, whereas .dis
images come from IGN (Institut Géographique National). The .ist format works in exactly the same way as the .bil format,
except for DTM images. For DTM images, the .ist format uses a decimetric coding for altitudes, whereas .bil images use
only a metric coding.
3.14 Planet Format

The Planet geographic data are described by a set of files grouped in a Planet directory. The directory structure depends
on the geographic data type.
Atoll supports the following objects in Planet format:
• Digital Terrain Model (8 and 16 bits)
• Clutter class maps (16 bits)
• Raster images (1, 4, 8 and 24 bits)
• Vector data
• Text data
52 AT271_TRG_E6 © Forsk 2009

3.14.1 DTM File

3.14.1.1 Description
The DTM directory consists of three files; the height file and two other files detailed below:
• The index file structure is simple; it is an ASCII text file that holds position information about the file. It contains
five columns. You can open an index file using any ASCII text editor. The format of the index file is as follows:
Field Acceptable values Description

File name Text Name of file referenced by the index file
East min Float x-axis map coordinate of the centre of the upper-left pixel in meters
East max Float x-axis map coordinate of the centre of the upper-right pixel in meters
North min Float y-axis map coordinate of the centre of the lower-left pixel in meters
North max Float y-axis map coordinate of the centre of the upper-left pixel in meters
Square size Float Dimension of a pixel in meters
• The projection file provides information about the projection system used. This file is optional. It is an ASCII text
file with four lines maximum.
Line Description
Spheroid
Zone
Projection
Latitude and longitude of projection central meridian and equivalent x and y coordinates in meters
Central meridian
(optional)
Note:
• In the associated binary file, the value -9999 corresponds to ‘No data’ which is supported
by Atoll.
3.14.1.2 Sample
Index file associated with height file (DTM data):
sydney1 303900 343900 6227900 6267900 50
Projection file associated with height file (DTM data):
Australian-1965
56
UTM
0 153 500000 10000000
3.14.2 Clutter Class Files

The Clutter directory consists of three files; the clutter file and two other files detailed below:
• The menu file, an ASCII text file, defines the feature codes for each type of clutter. It consists of as many lines
(with the following format) as there are clutter codes in the clutter data files. This file is optional.
Field Type Description

Clutter-code Integer (>1) Identification code for clutter class
Name associated with the clutter-code. (It may contain
Feature-name Text (up to 32 characters in length)
spaces)
• The index file gives clutter spatial references. The structure of clutter index file is the same as the structure of DTM
index file.
© Forsk 2009 AT271_TRG_E6 53

Note:
• In the associated binary file, the value -9999 corresponds to ‘No data’ which is supported
by Atoll.
3.14.2.2 Sample
Menu file associated with the clutter file:
1 open
2 sea
3 inlandwater
4 residential
5 meanurban
6 denseurban
7 buildings
8 village
9 industrial
10 openinurban
11 forest
12 parks
13 denseurbanhigh
14 blockbuildings
15 denseblockbuild
16 rural
17 mixedsuburban
3.14.3 Vector Files

Vector data comprises terrain features such as coastlines, roads, etc. Each of these features is stored in a separate vector
file. Four types of files are used, the vector file, where x and y coordinates of vector paths are stored, and three other files
detailed below:
• The menu file, an ASCII text file, lists the vector types stored in the database. The menu file is composed of one
or more records with the following structure:

Vector type code Integer > 0 Identification code for the vector type
Vector type name Text (up to 32 characters in length) Name of the vector type
The fields are separated by space character.

• The index file, an ASCII text file, lists the vector files and associates each vector file with one vector type, and
optionally with one attribute file. The index file consists of one or more records with the following structure:

Vector file name Text (up to 32 characters in length) Name of the vector file
Name of attribute file associated with the vector file
Attribute file name Text (up to 32 characters in length)
(optional)
vector file eastmin: minimum x-axis coordinate of all
vector path points in the vector file
vector file eastmax: maximum x-axis coordinate of all
Dimensions Real
vector file northmin: minimum y-axis coordinate of all
vector file northmax: maximum y-axis coordinate of all
vector path points
Name of the vector type with which the vector file is
Vector type name Text (up to 32 characters in length) associated. This one must match exactly a vector type
name field in the menu file.
54 AT271_TRG_E6 © Forsk 2009

The fields are separated by spaces.

• The attribute file stores the height and description properties of vector paths. This file is optional.
3.14.3.2 Sample
Index file associated with the vector files
sydney1.airport 313440 333021 6239426 6244784 airport

sydney1.riverlake 303900 342704 6227900 6267900 riverlake
sydney1.coastline 322837 343900 6227900 6267900 coastline
sydney1.railways 303900 336113 6227900 6267900 railways
sydney1.highways 303900 325155 6240936 6267900 highways
sydney1.majstreets 303900 342770 6227900 6267900 majstreets
sydney1.majorroads 303900 342615 6227900 6267900 majorroads
3.14.4 Image Files

The image directory consists of two files, the image file with .tif extension and an index file with the same structure as the
DTM index file structure.
3.14.5 Text Data Files

The text data directory consists of:
• The text data files are ASCII text files with the following format:
Airport
637111.188 3094774.00
Airport
628642.688 3081806.25
Each file contains a line of text followed by easting and northing of that text, etc.
• The index file, an ASCII text file, stores the position of each text file. It consists of one or more records with the
following structure:

File name Text (up to 32 characters in length) File name of the text data file
Minimum x-axis coordinate of all points listed in the text
East Min Real
data file
Maximum x-axis coordinate of all points listed in the text
East Max Real
data file
Minimum y-axis coordinate of all points listed in the text
North Min Real
data file
Maximum y-axis coordinate of all points listed in the text
North Max Real
data file
Text feature Text (up to 32 characters in length) This field is omitted in case no menu file is available.
The fields are separated by spaces.
railwayp.txt -260079 693937 2709348 3528665 Railway_Station

airport.txt -307727 771663 2547275 3554675 Airport
ferryport.txt 303922 493521 2667405 3241297 Ferryport
• The menu file, an ASCII text file, contains the text features. This file is optional.
1 Airport
2 Ferryport
3 Railway_Station
© Forsk 2009 AT271_TRG_E6 55

3.15 MNU Format

3.15.1 Description
A .mnu file is useful when importing clutter classes or raster traffic files in .tif, .bil and .img formats. It gives the correspond-
ence between the clutter (or traffic) code and the class name. It is a text file with the same name as the clutter (or traffic)
file with .mnu extension. It must be stored at the same location as the clutter (or traffic) file. It has the same structure as
the menu file used in the Planet format.

Class code Integer > 0 Identification code for the clutter (or traffic) class
Class name Text (up to 50 characters in length) Name of the clutter (or traffic) class. It may contain spaces.
Separator used can either be a space character or a tab.
3.15.2 Sample
A .mnu file associated to a clutter classes file:
0 none
1 open
2 sea
3 inland_water
4 residential
5 meanurban
3.16 XML Table Export/Import Format

All the data tables in an Atoll document can be exported to XML files.
Atoll creates the following files when exporting data tables to XML files:
• One index.xml file which contains the mapping between the data tables in Atoll and the corresponding XML file
created by the export.
• One XML file per data table which contains the data table format (schema) and the data.
The XML import does not modify the active document table and field definitions. Therefore, the Networks and Custom-
Fields tables, although exported, are not imported.
The following sections describe the structures of these two types of XML files created at export.
3.16.1 Index.xml File

The index.xml file stores the system (GSM, UMTS, etc.) and the technology (TDMA, CDMA, etc.) of the document, and
the version of Atoll used for exporting the data tables to XML files. It also contains the mapping between the data tables
in the Atoll document and the XML file corresponding to each data table.
The root tag <Atoll_XML_Config...> of the index.xml file contains the following attributes:
Attribute Description
Atoll_File_System Corresponds to the SYSTEM_ field of the Networks table of the exported document
Corresponds to the TECHNOLOGY field of the Networks table of the exported
Atoll_File_Technology
document
Atoll_File_Version Corresponds to the Atoll version
The index file also contains a list of mapping between the tables exported from Atoll and the XML files corresponding to
each table. This list is sorted in the order the Atoll tables are to be imported.
The list is composed of <XML_Table.../> tags with the following attributes:
Attribute Description
XML_File Corresponds to the exported XML file name (e.g., "Sites.xml")
Atoll_Table Corresponds to the exported Atoll table name (e.g., "Sites")
A sample extract of the index.xml is given below:
56 AT271_TRG_E6 © Forsk 2009

<Atoll_XML_Config Atoll_File_System="UMTS" Atoll_File_Technology="CDMA"

Atoll_File_Version="2.7.0 build 2334">
<XML_Table XML_File="CustomFields.xml" Atoll_Table="CustomFields" />
<XML_Table XML_File="CoordSys.xml" Atoll_Table="CoordSys" />
...
</Atoll_XML_Config>
Note that no closing tag </XML_Table> is required.
3.16.2 XML File

Atoll creates an XML file per exported data table. This XML file has two sections, one for storing the description of the table
structure, and the second for the data itself. The XML file uses the standard XML rowset schema (schema included in the
XML file between <s:Schema id=’RowsetSchema’> and </s:Schema> tags).
Rowset Schema
The XML root tag for XML files using the rowset schema is the following:
<xml xmlns:s='uuid:BDC6E3F0-6DA3-11d1-A2A3-00AA00C14882'
xmlns:dt='uuid:C2F41010-65B3-11d1-A29F-00AA00C14882'
xmlns:rs='urn:schemas-microsoft-com:rowset'
xmlns:z='#RowsetSchema'>
The schema definition follows the root tag and is enclosed between the following tags:
<s:Schema id=’RowsetSchema’>
<!-Schema is defined here, using <s:ElementType> and <s:AttributeType> tags ->
</s:Schema>
In the rowset schema, after the schema description, the data are enclosed between <rs:data> and </rs:data>.
Between these tags, each record is handled by a <z:row … /> tag having its attributes set to the record field values since
in the rowset schema, values are handled by attributes. Note that no closing tag </z:row> is required.
A sample extract of a Sites.xml file containing the Sites table with only one site is given below:
<xml xmlns:s='uuid:BDC6E3F0-6DA3-11d1-A2A3-00AA00C14882'
xmlns:dt='uuid:C2F41010-65B3-11d1-A29F-00AA00C14882'
xmlns:rs='urn:schemas-microsoft-com:rowset'
xmlns:z='#RowsetSchema'>
<s:Schema id='RowsetSchema'>
<s:ElementType name='row' content='eltOnly' rs:updatable='true'>
<s:AttributeType name='NAME' rs:number='1' rs:maydefer='true' rs:writeun-
known='true' rs:basetable='Sites' rs:basecolumn='NAME' rs:keycolumn='true'>
<s:datatype dt:type='string' dt:maxLength='50'/>
</s:AttributeType>
<s:AttributeType name='LONGITUDE' rs:number='2' rs:maydefer='true' rs:wri-
teunknown='true' rs:basetable='Sites' rs:basecolumn='LONGITUDE'>
<s:datatype dt:type='float' dt:maxLength='8' rs:precision='15' rs:fix-
edlength='true'/>
</s:AttributeType>
<s:AttributeType name='LATITUDE' rs:number='3' rs:maydefer='true' rs:write-
unknown='true' rs:basetable='Sites' rs:basecolumn='LATITUDE'>
<s:datatype dt:type='float' dt:maxLength='8' rs:precision='15' rs:fix-
edlength='true'/>
</s:AttributeType>
<s:AttributeType name='ALTITUDE' rs:number='4' rs:nullable='true' rs:mayde-
fer='true' rs:writeunknown='true' rs:basetable='Sites' rs:basecolumn='ALTI-
TUDE'>
© Forsk 2009 AT271_TRG_E6 57

<s:datatype dt:type='r4' dt:maxLength='4' rs:precision='7' rs:fix-

edlength='true'/>
</s:AttributeType>
<s:AttributeType name='COMMENT_' rs:number='5' rs:nullable='true' rs:mayde-
fer='true' rs:writeunknown='true' rs:basetable='Sites' rs:basecol-
umn='COMMENT_'>
<s:datatype dt:type='string' dt:maxLength='255'/>
</s:AttributeType>
<s:extends type='rs:rowbase'/>
</s:ElementType>
</s:Schema>
<rs:data>
<rs:insert>
<z:row NAME='Site0' LONGITUDE='8301' LATITUDE='-9756'/>
</rs:insert>
</rs:data>
</xml>
3.17 Externalised Propagation Results Format

Propagation results, i.e. the path loss matrices, may be stored in an external folder. This folder consists of a dBASE III
based file named ‘pathloss.dbf’ that contains calculation parameters of all the transmitters considered and one file (or two
when calculating main and extended path loss matrices) per transmitter taken into account. This is a binary file with .los
extension and contains the path loss values for a transmitter.
Note:
• Each transmitter path loss matrix is calculated on the area where calculation radius
intersects the computation zone (see: Computation zone).
3.17.1 DBF File

dBASE III file (pathloss.dbf) has a standard .dbf format described below. Its content can be checked by opening it in MS-
Access. The format is detailed hereafter.
3.17.1.1 DBF File Format

For general information, the format of .dbf files in any Xbase language is described.
Following notations are used in tables:
FS = FlagShip D3 = dBaseIII+
Fb = FoxBase D4 = dBaseIV
Fp = FoxPro D5 = dBaseV
CL = Clipper
3.17.1.1.1 DBF Structure

Byte Description
0...n .dbf header (see next part for size, byte 8)
1st record of fixed length (see next parts)

n+1 2nd record (see next part for size, byte10) … If .dbf is not empty
last record
last optional: 0x1a (eof byte)
3.17.1.1.2 DBF Header (Variable Size - Depends on Field Count)

Byte Size Contents Description Applies for (supported by)
00 1 0x03 plain .dbf FS, D3, D4, D5, Fb, Fp, CL
0x04 plain .dbf D4, D5 (FS)
0x05 plain .dbf D5, Fp (FS)
58 AT271_TRG_E6 © Forsk 2009

0x43 with .dbv memo var size FS

0xB3 with .dbv and .dbt memo FS
0x83 with .dbt memo FS, D3, D4, D5, Fb, Fp, CL
0x8B with .dbt memo in D4 format D4, D5
0x8E with SQL table D4, D5
0xF5 with .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 0,0 Reserved All
14 1 0x01 Begin transaction D4, D5
0x00 End Transaction D4, D5
0x00 ignored FS, D3, Fb, Fp, CL
15 1 0x01 Encrypted D4, D5
0x00 normal visible All
16 12 0 (1) multi-user environment use D4,D5
28 1 0x01 production index exists Fp, D4, D5
0x00 index upon demand All
29 1 n language driver ID D4, D5
0x01 codepage437 DOS USA Fp
0x02 codepage850 DOS Multi ling Fp
0x03 codepage1251 Windows ANSI Fp
0xC8 codepage1250 Windows EE Fp
30 2 0,0 reserved All
32 n*32 Field Descriptor, (see next paragraph) all
+1 1 0x0D Header Record Terminator all
• Field descriptor array in the .dbf header (32 bytes for each field)

0 11 ASCI field name, 0x00 termin all
11 1 ASCI field type (see next paragraph) all
12 4 n,n,n,n Fld address in memory D3
n,n,0,0 offset from record begin Fp
0,0,0,0 ignored FS, D4, D5, Fb, CL
16 1 byte Field length, bin (see next paragraph) all \ FS,CL: for C field type
17 1 byte decimal count, bin all / both used for fld lng
18 2 0,0 reserved all
20 1 byte Work area ID D4, D5
0x00 unused FS, D3, Fb, Fp, CL
21 2 n,n multi-user dBase D3, D4, D5
0,0 ignored FS, Fb, Fp, CL
23 1 0x01 Set Fields D3, D4, D5
0x00 ignored FS, Fb, Fp, CL
24 7 0...0 reserved all
31 1 0x01 Field is in .mdx index D4, D5
• Field type and size in the .dbf header, field descriptor (1 byte)
Size Type Description/Storage Applies for (supported by)

ASCII (OEM code page chars)
C 1...n Char all
rest= space, not \0 term.
© Forsk 2009 AT271_TRG_E6 59

n = 1...64kb (using deci count) FS

n = 1...32kb (using deci count) Fp, CL
n = 1...254 all
D8 Date 8 ASCII digits (0...9) in the YYYYMMDD format all
ASCII digits (-.0123456789)
F 1...n Numeric variable pos. of float.point FS, D4, D5, Fp
n = 1...20
N 1...n Numeric all
fix posit/no float.point
n = 1...20 FS, Fp, CL
n = 1...18 D3, D4, D5, Fb
L1 Logical ASCII chars (YyNnTtFf space) FS, D3, Fb, Fp, CL
ASCII chars (YyNnTtFf?) D4, D5 (FS)
10 digits repres. the start block posit. in .dbt file, or 10 spaces if
M 10 Memo all
no entry in memo
Variable, bin/asc data in .dbv
4bytes bin= start pos in memo
4bytes bin= block size
V 10 Variable FS
1byte = subtype
1byte = reserved (0x1a)
10 spaces if no entry in .dbv
binary data in .ftp
P 10 Picture Fp
structure like M
binary data in .dbt
B 10 Binary D5
structure like M
OLE objects
G 10 General D5, Fp
structure like M
22 short int binary int max +/- 32767 FS
44 long int binary int max +/- 2147483647 FS
88 double binary signed double IEEE FS
3.17.1.1.3 Each DBF Record (Fixed Length)

Byte Size Description Applies for (supported by)
0 1 deleted flag "*" or not deleted " " all
x-times contents of fields, fixed length, unterminated.
1…n 1… All
For n, see (2) byte 10…11
3.17.1.2 DBF File Content

The .dbf file provides information that is needed to check validity of each path loss matrix.

TX_NAME Text Name of the transmitter
FILE_NAME Text Name (and optionally, path) of .los file
MODEL_NAME Text Name of propagation model used to calculate path loss
Signature (identity number) of model used in calculations. You may check it in the
propagation model properties (General tab).
The Model_SIG is used for the purpose of validity. A unique Model_SIG is
MODEL_SIG Text assigned to each propagation model. When model parameters are modified, the
associated model ID changes. This enables Atoll to detect path loss matrix
invalidity. In the same way, two identical propagation models in different projects
do not have the same model IDa.
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
60 AT271_TRG_E6 © Forsk 2009

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
Logical number referring to antenna pattern. Antennas with the same pattern will
ANTENNA_SI Float
have the same number.
Maximum path loss stated in 1/16 dB. This information is used, when no
MAX_LOS Float
calculation radius is set, to check the matrix validity.
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
Locking status
LOCKED Boolean 0: path loss matrix is not locked
1: path loss matrix is locked.
Atoll indicates if losses due to the antenna pattern are taken into account in the
path loss matrix.
INC_ANT Boolean
0: antenna losses not taken into account
1: antenna losses included
a. In order to benefit from the calculation sharing feature, users must retrieve the propagation models from the same
central database. This can be done using the Open from database command for a new document or the Refresh
command for an existing one. Otherwise, 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.
b. These coordinates enable Atoll to determine the area of calculation for each transmitter.
c. These coordinates enable Atoll to determine the rectangle including the computation zone.
3.17.2 LOS File

The data file is a 16 bits binary row file organized in a standard row-column structure. It contains an integer path loss value,
with a 1/16 dB unit. Data are stored starting from the southwest to the northeast corner of the area.
3.18 Externalised Tuning Files

Atoll can tune path loss matrices obtained from propagation results by the use of real measurements (CW Measurements
or Test Mobile Data). For each measured transmitter, 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. A transmitter path loss
matrix can be tuned several times by the use of several measurement paths. All these tuning paths are stored in a cata-
logue. This catalogue is stored under a .tuning folder containing a .dbf file and one .pts file per corrected transmitter. Since
a tuning file can contain several measurement paths, all these measurements are added to the tuning file.
For more information on the path loss tuning algorithm, See "Path Loss Tuning" on page 108.
3.18.1 DBF File

dBASE III file (pathloss.dbf) has a standard .dbf format described below. Its content can be checked by opening it in MS-
Access. The format is detailed hereafter.
3.18.1.1 DBF File Format

For general information, the format of .dbf files in any Xbase language is described.
Following notations are used in tables:
FS = FlagShip D3 = dBaseIII+
Fb = FoxBase D4 = dBaseIV
Fp = FoxPro D5 = dBaseV
CL = Clipper
© Forsk 2009 AT271_TRG_E6 61

3.18.1.1.1 DBF Structure

Byte Description
0...n .dbf header (see next part for size, byte 8)
1st record of fixed length (see next parts)

n+1 2nd record (see next part for size, byte10) … If .dbf is not empty
last record
last optional: 0x1a (eof byte)
3.18.1.1.2 DBF Header (Variable Size - Depends on Field Count)

00 1 0x03 plain .dbf FS, D3, D4, D5, Fb, Fp, CL
0x04 plain .dbf D4, D5 (FS)
0x05 plain .dbf D5, Fp (FS)
0x43 with .dbv memo var size FS
0xB3 with .dbv and .dbt memo FS
0x83 with .dbt memo FS, D3, D4, D5, Fb, Fp, CL
0x8B with .dbt memo in D4 format D4, D5
0x8E with SQL table D4, D5
0xF5 with .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 0,0 Reserved All
14 1 0x01 Begin transaction D4, D5
0x00 End Transaction D4, D5
15 1 0x01 Encrypted D4, D5
0x00 normal visible All
16 12 0 (1) multi-user environment use D4,D5
28 1 0x01 production index exists Fp, D4, D5
0x00 index upon demand All
29 1 n language driver ID D4, D5
0x01 codepage437 DOS USA Fp
0x02 codepage850 DOS Multi ling Fp
0x03 codepage1251 Windows ANSI Fp
0xC8 codepage1250 Windows EE Fp
30 2 0,0 reserved All
32 n*32 Field Descriptor, (see next paragraph) all
+1 1 0x0D Header Record Terminator all
• Field descriptor array in the .dbf header (32 bytes for each field)

0 11 ASCI field name, 0x00 termin all
11 1 ASCI field type (see next paragraph) all
12 4 n,n,n,n Fld address in memory D3
n,n,0,0 offset from record begin Fp
0,0,0,0 ignored FS, D4, D5, Fb, CL
16 1 byte Field length, bin (see next paragraph) all \ FS,CL: for C field type
17 1 byte decimal count, bin all / both used for fld lng
18 2 0,0 reserved all
20 1 byte Work area ID D4, D5
62 AT271_TRG_E6 © Forsk 2009

0x00 unused FS, D3, Fb, Fp, CL

21 2 n,n multi-user dBase D3, D4, D5
0,0 ignored FS, Fb, Fp, CL
23 1 0x01 Set Fields D3, D4, D5
0x00 ignored FS, Fb, Fp, CL
24 7 0...0 reserved all
31 1 0x01 Field is in .mdx index D4, D5
• Field type and size in the .dbf header, field descriptor (1 byte)
Size Type Description/Storage Applies for (supported by)

ASCII (OEM code page chars)
C 1...n Char all
rest= space, not \0 term.
n = 1...64kb (using deci count) FS
n = 1...32kb (using deci count) Fp, CL
n = 1...254 all
D8 Date 8 ASCII digits (0...9) in the YYYYMMDD format all
F 1...n Numeric variable pos. of float.point FS, D4, D5, Fp
n = 1...20
N 1...n Numeric all
fix posit/no float.point
n = 1...20 FS, Fp, CL
n = 1...18 D3, D4, D5, Fb
L1 Logical ASCII chars (YyNnTtFf space) FS, D3, Fb, Fp, CL
ASCII chars (YyNnTtFf?) D4, D5 (FS)
10 digits repres. the start block posit. in .dbt file, or 10 spaces if
M 10 Memo all
no entry in memo
Variable, bin/asc data in .dbv
4bytes bin= start pos in memo
4bytes bin= block size
V 10 Variable FS
1byte = subtype
1byte = reserved (0x1a)
10 spaces if no entry in .dbv
binary data in .ftp
P 10 Picture Fp
structure like M
binary data in .dbt
B 10 Binary D5
structure like M
OLE objects
G 10 General D5, Fp
structure like M
22 short int binary int max +/- 32767 FS
44 long int binary int max +/- 2147483647 FS
88 double binary signed double IEEE FS
3.18.1.1.3 Each DBF Record (Fixed Length)

Byte Size Description Applies for (supported by)
0 1 deleted flag "*" or not deleted " " all
x-times contents of fields, fixed length, unterminated.
1…n 1… All
For n, see (2) byte 10…11
3.18.1.2 DBF File Content

The .dbf file provides information about the measured transmitters participating in the tuning.

TX_NAME Text Name of the transmitter
FILE_NAME Text Name (and optionally, path) of .pts file
AREA_XMIN Float Not used
AREA_XMAX Float Not used
© Forsk 2009 AT271_TRG_E6 63

AREA_YMIN Float Not used

AREA_YMAX Float Not used
3.18.2 PTS File

The tuning file contains a header and the list of points.
The contents of the header is:
• 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 measurements name (with prefix Num : for test mobile data and CW: for CW measurements)
• 256 bytes : comment
• 4 bytes : X_RADIUS
• 4 bytes : Y_RADIUS
• 4 bytes : Gain : measurement gain - losses
• 4 bytes : Global error
• 4 bytes : Rx height
• 4 bytes : Frequency
• 8 bytes : Tx Position
The list of points contains following 4-uplet for all points
• 4 bytes : X
• 4 bytes : Y
• 4 bytes : Measurement value
• 4 bytes : Incidence angle.
3.19 Interference Histograms File Formats

Interference histograms required by automatic frequency planning tools can be imported and exported.
Notes:
• No validity check is carried out when importing an interference histogram file.
• Atoll only imports interference histograms related to loaded transmitters.
• The lines starting with the symbol "#" are considered as comments.
• The interferer TRX type is not specified. In fact, the subcells of the interferer transmitter
differ by their power offsets. If the power offset of a subcell is X with respect to the BCCH,
then its interference C/I histogram will be shifted by X with respect to the BCCH
interference histogram. It contains no further information; therefore, the interferer TRX type
is always BCCH.
• For each interfered subcell-interferer subcell pair, Atoll saves probabilities for several C/I
values (between 6 to 24 values). Five of these values are fixed; probabilities are calculated
for C/I values equal to –9, 1, 8, 14, and 22 dB. Then, between each fixed C/I value, there
can be up to three additional values (this number depends on the probability variation
between the fixed values). The C/I values have 0.5 dB accuracy and probability values are
calculated and stored with an accuracy of 0.002 for probabilities between 1 and 0.05, and
with an accuracy of 0.0001 for probabilities lower than 0.05.
• If no power offset is defined on the Interfered TRX type, it is possible to use the "All" value.
• The values of probability should be absolute (between 0 and 1), and not in precentage
(between 0 and 100%).
3.19.1 One Histogram per Line (.im0) Format

This file contains one histogram per line for each interfered/interfering subcell pair. The histogram is a list of C/I values
with associated probabilities.
The .im0 file consists of two parts:
• The first part is a header used for format identification. It must start with and contain the following lines:
# Calculation Results Data File.

# Version 1.1, Tab separated format. Commented lines start with #.
• The second part details interference histogram of each interfered subcell-interferer subcell pair.
The lines after the header are considered as comments if they start with the symbol "#". If not, they must have the following
format:
64 AT271_TRG_E6 © Forsk 2009

<Column1><tab><Column2><tab><Column3><tab><Column4><newline>
The 4 tab-separated 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.
Interfered subcell. In order to save storage, all subcells with no power
Column3 Interfered TRX type
offset are not duplicated (e.g. BCCH, TCH).
C/I value and the probability associated to this value separated by a space
Column4 C/I Probability
character. This entry cannot be null.
3.19.1.1 Sample
# Remark: C/I results do not incorporate power offset values.
# Fields are:
#------------------------------------------------------------------------
#Transmitter Interferer TRX type {C/I Probability} values
#------------------------------------------------------------------------
#
# Warning, The parameter settings of this header can be wrong if
# the "export" is performed following an "import". They
# are correct when the "export" follows a "calculate".
#
# Service Zone Type is "Best signal level of the highest priority HCS layer".
# Margin is 5.
# Cell edge coverage probability 75%.
# Traffic spreading was Uniform
##---------------------------------------------------------------------#
#
Site0_2 Site0_1 BCCH,TCH -10 1 -9 0.996 -6 0.976 -4 0.964 -1 0.936
0 0.932 1 0.924 4 0.896 7 0.864 8 0.848
9 0.832 10 0.824 11 0.804 14 0.712 17 0.66
Site0_2 Site0_3 BCCH,TCH -10 1 -9 0.996 -6 0.976 -4 0.972 -1 0.948
0 0.94 1 0.928 4 0.896 7 0.856 8 0.84
11 0.772 13 0.688 14 0.636 15 0.608 18 0.556
Site0_3 Site0_1 BCCH,TCH -10 1 -9 0.996 -6 0.98 -3 0.948 0 0.932
1 0.924 4 0.892 7 0.852 8 0.832 9 0.816
10 0.784 11 0.764 14 0.644 15 0.616 18 0.564
Site0_3 Site0_2 BCCH,TCH -9 1 -6 0.972 -3 0.964 -2 0.96 0 0.94
1 0.932 4 0.904 7 0.876 8 0.86 9 0.844
11 0.804 13 0.744 14 0.716 15 0.692 18 0.644
3.19.2 One Value per Line with Dictionary File (.clc) Format
Atoll creates two ASCII text files in a specified directory: xxx.dct and xxx.clc (xxx is the user-specified name).
Note:
• When importing interference histograms with standard format, you must specify the .clc file
to be imported. Atoll looks for the associated .dct file in the same directory and uses it to
decode transmitter identifiers. If this file is unavailable, Atoll assumes that the transmitter
identifiers are the transmitter names. In this case, the columns 1 and 2 of the .clc file must
contain the names of the interfered and interferer transmitters instead of their identification
numbers.
© Forsk 2009 AT271_TRG_E6 65

3.19.2.1 CLC File

3.19.2.1.1 Description
The .clc file consists of two parts:

format:
<Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline>

Identification number of the interfered transmitter. If the column is empty,
Column1 Interfered transmitter
its value is identical to the one of the line above.
Identification number of the interferer transmitter. If the column is null, its
Column2 Interfering transmitter
value is identical to the one of the line above.
Interfered subcell. If the column is null, its value is identical to the one of
Column3 Interfered TRX type the line above. In order to save storage, all subcells with no power offset
are not duplicated (e.g. BCCH, TCH).
Column4 C/I threshold C/I value. This column cannot be null.
Probability to have C/I the value specified in column 4 (C/I threshold). This
Column5 Probability C/I > Threshold
field must not be empty.
Note:
• The columns 1, 2, and 3 must be defined only in the first line of each histogram.
3.19.2.1.2 Sample
# Fields are:
##------------#------------#------------#-----------#------------------#
#| Interfered | Interfering| Interfered | C/I | Probability |
#| Transmitter| Transmitter| Trx type | Threshold | C/I >= Threshold |
##------------#------------#------------#-----------#------------------#
#
#
# Margin is 5.
##---------------------------------------------------------------------#
1 2 TCH_INNER 8 1
9 0.944
10 0.904
11 0.892
14 0.844
66 AT271_TRG_E6 © Forsk 2009

15 0.832
16 0.812
17 0.752
22 0.316
25 0.292
1 2 BCCH,TCH 8 1
9 0.944
10 .904
13 0.872
14 0.84
17 0.772
Note:
• A new interference matrix histograms format has been introduced in Atoll 2.3.1 to improve
the import and export features and the overall performance. In this format, if the TCH and
BCCH histograms are the same, they are no longer duplicated. Atoll keeps a single record
of these histograms indicating that they belong to TCH and BCCH both. For example,
- Old format histograms between victim 1 and interferer 2:
1 2 TCH -9.5 1 - 9 1 - 6 1
1 2 BCCH -9.5 1 - 9 1 - 6 1
- New format histograms between victim 1 and interferer 2:
1 2 TCH,BCCH -9.5 1 - 9 1 - 6 1
3.19.2.2 DCT File

3.19.2.2.1 Description
The .dct file is divided into two parts:
# Calculation Results Dictionary File.
• The second part provides information about transmitters taken into account in AFP.
format:
<Column1><tab><Column2><newline>
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. One transmitter per line is described separated with a tab
character.
3.19.2.2.2 Sample
# Calculation Results Dictionary File.
# Fields are:
##-----------#-----------#-----------#-----------#---------#---------#
#|Transmitter|Transmitter|BCCH during|BSIC during|% of vic'|% of int'|
#|Name |Identifier |calculation|calculation|coverage |coverage |
##-----------#-----------#-----------#-----------#---------#---------#
© Forsk 2009 AT271_TRG_E6 67

#
#
# Service Zone Type is "Best signal level per HCS layer".
# Margin is 5.
# Cell edge coverage probability is 75%.
# 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
3.19.3 One Value per Line (Transmitter Name Repeated) (.im1)

Format
This file contains one C/I threshold and probability pair value per line for each interfered/interfering subcell pair. The histo-
gram is a list of C/I values with associated probabilities.
The .im1 file consists of two parts:

format:
<Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline>

Interfered subcell. In order to save storage, all subcells with no power
Column3 Interfered TRX type
offset are not duplicated (e.g. BCCH, TCH).
Column4 C/I threshold C/I value. This column cannot be null.
Probability to have C/I the value specified in column 4 (C/I threshold). This
Column5 Probability C/I > Threshold
field must not be empty.
3.19.3.1 Sample
# Fields are:
#------------------------------------------------------------------------
#Transmitter Interferer TRX type C/I Probability
68 AT271_TRG_E6 © Forsk 2009

#------------------------------------------------------------------------
#
#
# Margin is 5.
##---------------------------------------------------------------------#
Site0_2 Site0_1 BCCH,TCH -10 1
Site0_2 Site0_1 BCCH,TCH -9 0.996
Site0_2 Site0_1 BCCH,TCH 0 0.932
...
3.19.4 Only Co-Channel and Adjacent Values (.im2) Format

In this case, there is only one .im2 file containing co-channel and adjacent channel interference probabilities specified for
each interfered transmitter – interferer transmitter pair. There is only one set of values for all the subcells of the interfered
transmitter.
Each line must have the following format:
<Column1><SEP><Column2><SEP><Column3><SEP><Column4><newline>
Where the separator (<SEP>) can either be a tab or a semicolon.

The four columns are defined in the table below:

Co-channel interference Probability of having C  I  Max  C  I req 
Column3
probability BCCH ,TCH
Adjacent channel Probability of having C  I  Max  C  I req  – F

Column4
interference probability BCCH ,TCH
C  I req corresponds to the required C/I threshold. This parameter is defined for each subcell.
F is the adjacent channel protection level.
3.19.4.1 Sample
# Fields are:
#------------------------------------------------------------------------
© Forsk 2009 AT271_TRG_E6 69

#Transmitter Interferer Co-channel Adjacent channel

#------------------------------------------------------------------------
#
#
# Margin is 5.
##---------------------------------------------------------------------#
Site0_2 Site0_1 0.226667 0.024
Site0_2 Site0_3 0.27 0.024
Site0_3 Site0_1 0.276 0.02
Site0_3 Site0_2 0.226 0.028
The columns in the sample above are separated with a tab. These columns can also be separated with a semilcolon:
Site0_2;Site0_1;0.226667;0.024
Site0_2;Site0_3;0.27;0.024
Site0_3;Site0_1;0.276;0.02
Site0_3;Site0_2;0.226;0.028
70 AT271_TRG_E6 © Forsk 2009

Chapter 4
Calculations
This chapter describes in detail the calculation of path losses, the propagation models implemented in Atoll
by default, the calculation of antenna attenuation according to antenna patterns, and other calculation
algorithms in Atoll.
Atoll
Atoll
Microwave
RF PlanningLink
Software
72 AT271_TRG_E6 © Forsk 2009

Chapter 4: Calculations
4 Calculations
4.1 Overview
Three kinds of predictions are available in Atoll:
• Point analysis enables you to visualise transmitter-receiver profile and to get predictions for a user-defined
receiver in real time anywhere on a geographic map (Point analysis window: Profile tab).
• Coverage studies consider each bin of calculation areas as a potential receiver you can define. Therefore, covered
bins correspond to areas where a criterion on the predicted received signal is fulfilled.
• Point analysis based on path loss matrices enables you to get parameters derived from predicted values in cov-
erage studies (field received, path loss, C/I, UMTS parameters) for a receiver anywhere inside a calculation area
(Point analysis window: Reception, Interference, AS analysis tabs).
An overview of different analysis methods is presented in the table below:
Point analysis based on path loss

Coverage studies Point analysis
matrices
Reception, Results,
Any study Profile
Interference, AS analysis
At the centre of each
Receiver Anywhere. Even beyond
calculation bin within Anywhere inside the calculation areas
position calculation areas
computation zone
Path loss matrix No calculation: result coming from path

Calculation Real time
calculation loss matrices
Profile Radial except when Method used for coverage studies: radial
Systematic
extractiona using SPM except when using SPM
One value inside a Different values inside a

Result One value inside a calculation bin
calculation bin calculation bin
a. When using SPM, you can choose either radial or systematic calculation option.
Notes:
• In coverage studies, Atoll calculates path loss for every bin within calculation areas.
However, only results on calculation bins inside the computation zone are displayed.
• Profile point analysis is calculated in real time. Therefore, prediction is always consistent
with the network. On the other hand, if you modify any parameter (radio or geo), which may
make matrices invalid, consider updating the matrices before using point analysis based on
path loss matrices.
• Due to different calculation methods, you can get different results at a same point when
performing a point analysis in profile or reception mode.
In any case, prediction is performed in three steps:
1st step: First of all, Atoll calculates the path loss ( L path ), using the selected propagation model.
L path = L model + L ant + L ant

Tx Rx
L model is the loss on the transmitter-receiver path calculated through the propagation model. L model value depends on
the selected propagation model.
L ant is the transmitter antenna attenuation (from antenna patterns).

Tx
L ant is the receiver antenna attenuation ( L ant = 0 ) (from antenna patterns).

Rx Rx
Notes:
• In any project, Atoll considers that the receiver antenna is in the transmitter antenna axis.
Therefore, the receiver antenna attenuation is supposed to be zero.
• Transmitter antenna attenuation may not be considered in this step. It depends on
propagation model provider, who may choose to include this parameter in L path
calculation. However, all the propagation models available in Atoll calculate L path by
considering transmitter antenna attenuation.
2nd step: When the option “Shadowing taken into account” is selected, Atoll evaluates a shadowing margin,
M Shadowing – model , from the user-defined model standard deviation at the receiver and the cell edge coverage probability.
© Forsk 2009 AT271_TRG_E6 73

Note:
• For a cell edge coverage probability of 50%, the shadowing margin is always zero. In this
case, Atoll still works as above.
3rd step: Then, Atoll determines the prediction criterion and displays coverage.
For a signal level study,
The signal level at the receiver ( P Rec ) is calculated. We have (in dBm):
P Rec = EIRP – L path – M Shadowing – model – L Indoor +  G ant – L Rx 

Rx
Where EIRP = P Tx + G ant – L Tx

Tx
EIRP is the effective isotropic radiated power of the transmitter.
P Tx is the transmitter power.
G ant is the transmitter antenna gain.

Tx
L Tx are transmitter losses.
M Shadowing – model is the shadowing margin.
L Indoor are the indoor losses.These losses are defined for each clutter class.They are taken into account when the option
“Indoor coverage” is selected,
L Rx are receiver losses.
G ant is the receiver antenna gain.

Rx
Notes:
• In UMTS, CDMA2000 and IS95-CDMA documents, P Tx = P Pilot and L Tx = L total – DL .
• In UMTS, CDMA2000 and IS95-CDMA documents, Atoll considers that G ant and L Rx
Rx
equal zero when calculating the received signal level (in point analysis, Profile and
Reception tabs, and in common coverage studies such as Coverage per transmitter,
Coverage by field level, Overlapping).
• In GSM_EGPRS documents, L Tx = L total – DL .
• In GSM_EGPRS documents, receiver is equipped with an antenna with zero gain.
The prediction is performed for a user-defined cell edge coverage probability (x%). This means that the measured criterion
exceeds the predicted criterion for x% of time. The prediction is reliable during x% of time.
Note:
• In case of interference studies, only signal from interfered transmitter (C) is downgraded by
the shadowing margin. We consider that interference value (I) is not altered by the
shadowing margin.
4.2 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), and a second matrix over a larger radius computed with a low resolution
and another propagation model (extended matrix).
To be considered for calculations, a transmitter must fulfil the following conditions:
• It must be active,
• It must satisfy filter criteria defined in the Transmitters folder, and
• It must have a calculation area.
In the rest of the document, a transmitter fulfilling the conditions detailed above will be called TBC transmitter.
The path loss matrix size of a TBC transmitter depends on its calculation area. 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. You may
have one or two path loss matrices per TBC transmitter.
74 AT271_TRG_E6 © Forsk 2009

4.2.1 Calculation Area Determination

4.2.1.1 Computation Zone
Transmitter calculation area is made of a rectangle or a square depending on transmitter calculation radius and the compu-
tation zone.
Calculation radius enables Atoll to define a square around the transmitter. One side of the square equals twice the entered
calculation radius.
Since the computation zone can be made of one or several polygons, transmitter calculation area corresponds to the inter-
section area between its calculation square and the rectangle containing the computation zone area(s).
Figure 4.1: Example 1: Single Calculation Area
Figure 4.2: Example 2: Multiple Calculation Areas
Computation zone(s)
Rectangle containing the computation zone(s)
Calculation area defined (square)
Transmitter
Calculation area: real area for which Atoll calculates path losses
4.2.1.2 Use of Polygonal Zones in Coverage Prediction Reports

Prediction statistics are evaluated over the focus zone, if existing, then over the computation zone, if existing, or over the
whole covered area. The area of the focus and computation zones are calculated by decomposition in triangles.
The area of each prediction is calculated by counting its pixels inside the focus (resp. computation) zone. This number of
pixels multiplied by the area of one of its pixels gives the total area.
This area depends on the study resolution. At the border of the focus (resp. computation) zone, pixels are considered
either IN or OUT of the zone. A pixel is IN if its centre is inside the focus zone.
If a prediction covers the entire focus (resp. computation) zone, its area should be equal to the focus (resp. computation)
zone area, but as these 2 different methods differ, the results may be slightly different. If it happens that the value of the
prediction area is higher than the focus zone area, then the calculated percentage value is higher than 100%. In that case,
Atoll automatically replaces it by 100%.
© Forsk 2009 AT271_TRG_E6 75

4.2.2 Calculate / Force Calculation Comparison

4.2.2.1 Calculate
The Calculate feature (F7) enables you:
1. To calculate prediction studies
The first time you click Calculate (no path loss matrices exist), Atoll computes path loss matrices for each TBC
transmitter. Then, it calculates created and unlocked coverage prediction studies inside the computation zone.
2. To check result validity and update calculations
If calculations have been performed once and you have changed some parameters such as radio data or calcu-
lation area, Atoll automatically detects path loss matrices to be recalculated. These are either one or several path
loss matrices that become invalid due to certain modifications. Then Atoll calculates the prediction study, or just
the prediction study if matrices were all still valid.
4.2.2.2 Force Calculation

With the Force calculation feature (Ctrl+F7), Atoll deletes all the path loss matrices even if they are valid, recalculates them
and then updates the results of prediction studies.
Note:
• Geographic data (DTM, clutter) modification makes path loss matrices invalid. However,
Atoll does not detect this invalidity just by using Calculate. Therefore, to update
calculations, you must click the Force calculation command.
4.2.3 Matrix Validity

Atoll manages path loss matrix validity transmitter by transmitter, even in case of transmitters with two path loss matrices
(main and extended matrices). Therefore, even if only one path loss matrix of the transmitter is invalid, Atoll will recalculate
both of them. All the geographic data modifications and some radio data changes can make matrices invalid. This table
lists these modifications and also changes that have an impact only on prediction studies.
Force
Modification Matrix validity Impact on Calculate
calculation
Frequency Invalid Path loss matrices Sufficient Not necessary
Antenna* coordinates (site coordinate:
Invalid Path loss matrices Sufficient Not necessary
X and Y, Dx and Dy)
Antennaa height Invalid Path loss matrices Sufficient Not necessary

a Invalid Path loss matrices Sufficient Not necessary
Antenna pattern
Downtilt
Azimuth
% Power (when there is other
antennas)
Site position/altitude Invalid Path loss matrices Sufficient Not necessary
Grid resolution (main or/and
extended)
Propagation model (main or/and
extended)
Propagation model parameters Invalid Path loss matrices Sufficient Not necessary
Calculation areas
Valid Prediction study Sufficient Not necessary
1. Calculation areas gets smaller
Calculation areas
2. Calculation areas gets larger
Receiver height Invalid Path loss matrices Sufficient Not necessary
Receiver losses Valid Prediction study Sufficient Not necessary
Receiver gain Valid Prediction study Sufficient Not necessary
Receiver antenna Valid because L ant = 0 Prediction study Sufficient Not necessary
Rx
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
76 AT271_TRG_E6 © Forsk 2009

New clutter class edition Invalid Path loss matrices Insufficientb Necessary
Coverage study resolution Valid Prediction study Sufficient Not necessary
Cell edge coverage probability Valid Prediction study Sufficient Not necessary
Coverage study conditions Valid Prediction study Sufficient Not necessary
Coverage study display options Valid Prediction study Sufficient Not necessary
a.Modification of any parameter related to main or other antennas makes matrix invalid.
b.Except if this action has an impact on the site positions/altitudes.
Tip 1
Calculate or Force Calculation?
If you modify radio data or calculation areas, use the Calculate button. On the other hand, if you change geographic
data, it is necessary to use Force calculation.
Tip 2
Calculation area management
When performing prediction studies, it is recommended to follow this methodology to minimise recalculations:
1st step: Calculate without computation zone.
2nd step: Draw a computation zone and calculate.
3rd step: Decrease the calculation radius and calculate.
4.3 Path Loss Calculations

4.3.1 Ground Altitude Determination
Atoll determines reception and transmission site altitude from Digital Terrain Model map. The method used to evaluate site
altitude is based on a bilinear interpolation. It is described below.
Let us suppose a site S located inside a bin. Atoll knows the altitudes of four bin vertices, S’1, S’’1, S’2 and S’’2, from the
DTM file (Centre of each DTM pixel).
Figure 4.3: Ground Altitude Determination - 1
1st step: Atoll draws a vertical line through S. This line respectively intersects (S’1,S’’1) and (S’2, S’’2) lines at S1 and S2.
2nd step: Atoll determines the S1 and S2 altitudes using a linear interpolation method.
© Forsk 2009 AT271_TRG_E6 77

3rd step: Atoll performs a second linear interpolation to evaluate the S altitude.
4.3.2 Clutter Determination

Some propagation models need clutter class and clutter height as information at receiver or along a transmitter-receiver
profile.
4.3.2.1 Clutter Class

Atoll uses clutter classes file to determine the clutter class.
4.3.2.2 Clutter Height

To evaluate the clutter height, Atoll uses clutter heights file if available in the .atl document; clutter height of a site is the
height of the nearest point in the file.
Example: Let us suppose a site S. In the clutter heights file, Atoll reads clutter heights of four points around the site, S’1,
S’’1, S’2 and S’’2. Here, the nearest point to S is S”2; therefore Atoll takes the S”2 clutter height as clutter height of S.
Figure 4.7: Clutter Height
If you do not have any clutter height file, Atoll takes clutter height information in clutter classes file. In this case, clutter
height is an average height related to a clutter class.
4.3.3 Geographic Profile Extraction

Geographic profile extraction is needed in order to calculate diffraction losses. Profiles can be based on DTM only or on
DTM and clutter both. In fact, it depends on the selected propagation model.
4.3.3.1 Extraction Methods

4.3.3.1.1 Radial Extraction
Atoll draws radials from the site (where transmitter is located) to each calculation bin located along the transmitter calcu-
lation area border. In other words, Atoll determines a geographic profile between site and each bin centre.
78 AT271_TRG_E6 © Forsk 2009

Figure 4.8: Radial calculation method
Transmitter
Radial: Atoll will extract a geographic profile for each radial
Centre of a bin located on the calculation border
Receiver: it may be anywhere in point analysis or at the centre of each calculation bin in coverage studies
Figure 4.9: Site-bin centre profile
The receiver may be located either anywhere within a calculation bin (Point prediction) or at the centre of a calculation bin
(Coverage study). Therefore, according to the receiver position, Atoll chooses the nearest profile and uses it (receiver is
considered as located on the profile) to perform prediction study at the receiver.
4.3.3.1.2 Systematic Extraction

In this case, Atoll systematically extracts a geographic profile between the site (where transmitter resides) and the receiver.
© Forsk 2009 AT271_TRG_E6 79

Figure 4.10: Radial calculation method
Transmitter
Geographic profiles
Receiver: it may be anywhere in point analysis or at the centre of each calculation bin in coverage studies
4.3.3.2 Profile Resolution: Multi-Resolution Management

Geographic profile resolution depends on resolution of geographic data used by the propagation model (DTM and/or clut-
ter).
1. 1st case: If the chosen propagation model considers both DTM and clutter heights along the profile, the profile
resolution will be the highest of the two.
Example 1: Standard Propagation Model is used to perform predictions. A DTM map with a 40 m resolution and
a clutter heights map with a 20 m resolution are available.
Both DTM and clutter maps are considered when using the Standard propagation model. Therefore, here, the
profile resolution will be 20 m. It means that Atoll will extract geographic information, ground altitude and clutter
height, every 20 m. To get ground altitude every 20m, Atoll uses the bilinear interpolation method described in
"Ground Altitude Determination" on page 77. Clutter heights are read from the clutter heights map. Atoll takes the
clutter height of the nearest point every 20m (see Path loss calculations: Clutter determination).
Example 2: Standard Propagation Model is used to perform predictions. A DTM map with a 40 m resolution and
a clutter classes map with a 20 m resolution are available. No clutter height file has been imported in .atl document.
Both DTM and clutter maps are considered when using the Standard propagation model. Therefore, here, the
profile resolution will be 20 m. It means that Atoll will extract geographic information, ground altitude and clutter
height, every 20 m. To get ground altitude every 20 m, Atoll uses the bilinear interpolation method described in
"Ground Altitude Determination" on page 77. Atoll uses the clutter classes map to determine clutter height. Every
20 m, it determines clutter class and takes associated average height.
2. 2nd case: If the chosen propagation model takes into account only DTM map along the profile, profile resolution
will be the highest resolution among the DTM files.
Example: Cost-Hata is used to perform predictions. Both DTM maps with 40 m and 25 m resolutions and a clutter
map with a 20 m resolution are available.
80 AT271_TRG_E6 © Forsk 2009

Explorer window Work space
DTM
• DTM 1 (25m)
• DTM 2 (40m)
Clutter
• Clutter (20m)
Only DTM maps are considered along the whole profile when using Cost-Hata model. Therefore, here, the profile
resolution will be 25 m. It means that Atoll will extract geographic information, only the ground altitude, every 25 m.
DTM 1 is on the top of DTM 2. Thus, Atoll will consider ground elevation read from DTM 1 in the definition area of
DTM 1 and DTM 2 elsewhere. To get ground altitude every 25 m, Atoll uses the bilinear interpolation method
described in "Ground Altitude Determination" on page 77.
Notes:
• The selected profile resolution does not depend on the geographic layer order. In the last
example, whatever the DTM file order you choose, profile resolution will always be 25m.
On the other hand, the geographic layer order will influence the usage of data to establish
the profile.
• The calculation bin of path loss matrices defined by the grid resolution is independent of
geographic file resolution.
© Forsk 2009 AT271_TRG_E6 81

82
4.4
Propagation ITU 370-7 Erceg-Greenstein COST-Hata

ITU 1546 ITU 526-5 WLL Standard Propagation Model ITU 529-3
model (Vienna 93) (SUI) Okumura-Hata
Frequency
100-400 MHz 30-3000 MHz 30-10000 MHz 30-10000 MHz 150-3500 MHz 300-1500 MHz 1900-6000 MHz 150-2000 MHz
band
Free space loss L(d, f, HRx) L(d, f, HTx, HRx) L(d, f, HRx)
Physical Corrected Free space loss + Free space loss Free space loss
L(d, HTxeff, HRxeff, Diff loss, clutter) (per environment) (per environment) (per environment)
phenomena standard Corrections Diffraction loss Diffraction loss
Diffraction loss Diffraction loss Diffraction loss
loss
Diffraction Deygout Deygout (3 obstacles)
(3 obstacles) Deygout Epstein-Peterson (3 obstacles) Deygout Deygout Deygout
calculation - -
Deygout corrected (3 obstacles) Deygout corrected (3 obstacles) (1 obstacle) (1 obstacle) (1 obstacle)
method (3 obstacles) Millington (1 obstacle)
Propagation Models
Profile DTM DTM

- - DTM DTM DTM DTM
based on Clutter Clutter
AT271_TRG_E6
Profile
Radial
extraction - - Radial Radial Radial Radial Radial
Systematic
mode
Macro cell Macro cell Macro cell Macro cell

Cell size Macro cell Macro cell Macro cell -
Mini cell Mini cell Mini cell Mini cell
Receiver Street Street

Rooftop Rooftop Street Street Street Street
location Rooftop Rooftop
Receiver Fixed Mobile Fixed Fixed Mobile and Fixed Mobile Fixed Mobile
Propagation models available in Atoll are listed in the table below along with their main characteristics.
GSM900
GSM1800 GSM900
Broadcast
WLL UMTS GSM900 WiMAX in Urban GSM1800
Use Broadcast Land and maritime WLL
WiMAX CDMA2000 CDMA2000 and Suburban UMTS
Mobile
WiMAX CDMA2000
© Forsk 2009
Notes:
• In formulas described above, L model is stated in dB.
• Under Physical phenomena, L(...) expressions refer to formulas customisable in Atoll.

• SUI stands for Stanford University Interim models.
4.4.1 Okumura-Hata and Cost-Hata Propagation Models

4.4.1.1 Hata Path Loss Formula
Hata formula empirically describes the path loss as a function of frequency, receiver-transmitter distance and antenna
heights for an urban environment. This formula is valid for flat, urban environments and 1.5 metre mobile antenna height.
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   log d
f is the frequency (MHz).

hTx is the transmitter antenna height above ground (m) (Hb notation is also used in Atoll).
d is the distance between the transmitter and the receiver (km).

The parameters A1, A2, A3, B1 and B2 can be user-defined. Default values are proposed in the table below:
Okumura-Hata Cost-Hata
Parameters
f 1500 MHz f > 1500 MHz
A1 69.55 46.30
A2 26.16 33.90
A3 -13.82 -13.82
B1 44.90 44.90
B2 -6.55 -6.55
Default Hata parameters
4.4.1.2 Corrections to the Hata Path Loss Formula

As described above, the Hata formula is valid for urban environment and a receiver antenna height of 1.5m. For other envi-
ronments and mobile antenna heights, corrective formulas must be applied.
L model1 = Lu – a  h Rx  for large city and urban environments
f 2
L model1 = Lu – a  h Rx  – 2  log  ------  – 5.4 for suburban area
  28 
2
L model1 = Lu – a  h Rx  – 4.78  log  f   + 18.33 log  f  – 40.94 for rural area
a(hRx) is a correction for a receiver antenna height different from 1.5m.
Environment a(Hr)
Rural/Small city  1.1 log  f  – 0.7 h Rx –  1.56 log  f  – 0.8 
2
Large city 3.2  log  11.75h Rx   – 4.97
Note:
• When receiver antenna height equals 1.5m, a(hRx) is close to 0 dB regardless of
frequency.
4.4.1.3 Calculations in Atoll

Hata models take into account topo map (DTM) between transmitter and receiver and morpho map (clutter) at the receiver.
1st step: For each calculation bin, Atoll determines the clutter bin on which the receiver is located. This clutter bin corre-
sponds to a clutter class. Then, it uses the Hata formula assigned to this clutter class to evaluate L model1 .
2nd step: This step depends on whether the ‘Add diffraction loss’ option is checked.
• If the ‘Add diffraction loss’ option is unchecked, Atoll stops calculations.
L model = L model1
© Forsk 2009 AT271_TRG_E6 83

• If the ‘Add diffraction loss’ option is selected, Atoll proceeds as follows:

a. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation mode.
b. It determines the largest obstacle along the profile in accordance with the Deygout method and evaluates loss-
es due to diffraction L model2 .
L model = L model1 + L model2
Note:
• Like for any Hata-based model, L model is, by default, limited to the computed free space
loss value. It is also possible to avoid this option (option in the related scrolling menu of
Configuration tab).
4.4.2 ITU 529-3 Propagation Model

4.4.2.1 ITU 529-3 Path Loss Formula
The ITU 529.3 model is a Hata-based model. For this reason, its formula empirically describes the path loss as a function
of frequency, receiver-transmitter distance and antenna heights for a urban environment. This formula is valid for flat,
urban environments and 1.5 metre mobile antenna height.
The standard ITU 529-3 formula, for a receiver located on a urban environment, is given by:
b
E = 69.82 – 6.16 log f + 13.82 log h Tx –  44.9 – 6.55 log h Tx   log d 
where:
E is the field strength for 1 kW ERP
f is the frequency (MHz).
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, a conversion has to be made. One can find the following conversion formula:
Lu = 139.37 + 20 log f – E
which gives the following path loss formula for the ITU 529-3 model:
b
Lu = 69.55 + 26.16 log f – 13.82 log h Tx +  44.9 – 6.55 log h Tx   log d 
4.4.2.2 Corrections to the ITU 529-3 Path Loss Formula

4.4.2.2.1 Environment Correction
As described above, the Hata formula is valid for urban environment. For other environments and mobile antenna heights,
corrective formulas must be applied.
L model1 = Lu – a  h Rx  for large city and urban environments
f 2
L model1 = Lu – a  h Rx  – 2  log  ------  – 5.4 for suburban area
  28 
2
L model1 = Lu – a  h Rx  – 4.78  log f  + 18.33 log f – 40.94 for rural area
4.4.2.2.2 Area Size Correction

In the formulas above, a  h Rx  is the environment correction and is defined according to the area size
Environment a(Hr)
Rural/Small city  1.1 log f – 0.7 h Rx –  1.56 log f – 0.8 
2
Large city 3.2  log  11.75h Rx   – 4.97
84 AT271_TRG_E6 © Forsk 2009

4.4.2.2.3 Distance Correction

The distance correction refers to the term b above.
Distance b
d<20 km 1
d 0.8
b = 1 +  0.14 + 1.87  10 f + 1.07  10 h' Tx    log ------
–4 –3
 20
d>20 km
h Tx
where h' Tx = --------------------------------------------
–6 2
1 + 7  10 h Tx

Hata-based models take into account topo map (DTM) between transmitter and receiver and morpho map (clutter) at the
receiver.
1st step: For each calculation bin, Atoll determines the clutter bin on which the receiver is located. This clutter bin corre-
sponds to a clutter class. Then, it uses the ITU 529-3 formula assigned to this clutter class to evaluate L model1 .
2nd step: This step depends on whether the ‘Add diffraction loss’ option is checked.
• If the ‘Add diffraction loss’ option is unchecked, Atoll stops calculations.
L model = L model1

a. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation mode.
es due to diffraction  L model2  .
L model = L model1 + L model2
Note:
Configuration tab)
4.4.3 Standard Propagation Model (SPM)

4.4.3.1 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,
K1: constant offset (dB).
K2: multiplying factor for log(d).
d: distance between the receiver and the transmitter (m).
K3: multiplying factor for log(HTxeff).
HTxeff: effective height of the transmitter antenna (m).

K4: multiplying factor for diffraction calculation. K4 has to be a positive number.
Diffraction loss: loss due to diffraction over an obstructed path (dB).
K5: multiplying factor for log  d   log  H Txeff 
K6: multiplying factor for H Rxeff .
K7: multiplying factor for log  H Rxeff  .
H Rxeff : effective mobile antenna height (m).
Kclutter: multiplying factor for f(clutter).

f(clutter): average of weighted losses due to clutter.
© Forsk 2009 AT271_TRG_E6 85


4.4.3.2.1 Visibility and Distance Between Transmitter and Receiver
For each calculation bin, Atoll determines:
• The distance between the transmitter and the receiver.
If the distance Tx-Rx is less than the maximum user-defined distance (break distance), the receiver is considered to be
near the transmitter. Atoll will use the set of values marked “Near transmitter”.
If the distance Tx-Rx is greater than the maximum distance, receiver is considered far from transmitter. Atoll will use the
set of values “Far from transmitter”.
• Whether the receiver is in the transmitter line of sight or not.
If the receiver is in the transmitter line of sight, Atoll will take into account the set of values (K1,K2)LOS.
If the receiver is not in the transmitter line of sight, Atoll will use the set of values (K1,K2)NLOS.
4.4.3.2.2 Effective Transmitter Antenna Height

Effective transmitter antenna height (HTxeff) may be calculated with six different methods.
Height Above Ground
The transmitter antenna height is above the ground (HTx in m).
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. The profile length depends on distance min and distance max values and is limited by the trans-
mitter and receiver locations. Distance min and Distance max are minimum and maximum distances from the transmitter
respectively.
H Txeff = H Tx +  H 0Tx – H 0 
where,
H 0Tx is the ground height (ground elevation) above sea level at transmitter (m).
H 0 is the average ground height above sea level along the profile (m).
Note:
• If the profile is not located between the transmitter and the receiver, HTxeff equals HTx only.
Slope at Receiver Between 0 and Minimum Distance
The transmitter antenna height is calculated using the ground slope at receiver.
H Txeff =  H Tx + H 0Tx  – H 0Rx + K  d
where,
H 0Rx is the ground height (ground elevation) above sea level at receiver (m).
K is the ground slope calculated over a user-defined distance (Distance min). In this case, Distance min is a distance from
receiver.
Notes:
• If H Txeff  20m then, Atoll uses 20m in calculations.
• If H Txeff  200m then, Atoll takes 200m.
Spot Ht
If H 0Tx  H 0Rx then, H Txeff = H Tx +  H 0Tx – H 0Rx 
If H 0Tx  H 0Rx then, H Txeff = H Tx
86 AT271_TRG_E6 © Forsk 2009

Absolute Spot Ht
H Txeff = H Tx + H 0Tx – H 0Rx
Note:
• 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).
These values are only used in the two last methods and have different meanings according to the method.
Enhanced Slope at Receiver
Atoll offers a new method called “Enhanced slope at receiver” to evaluate the effective transmitter antenna height.
Figure 4.11: Enhanced Slope at Receiver
Let x-axis and y-axis respectively represent positions and heights. We assume that x-axis is oriented from the transmitter
(origin) towards the receiver.
This calculation is achieved in several steps:
1st step: Atoll determines line of sight between transmitter and receiver.
The LOS line equation is:
  H 0Tx + H Tx  –  H 0Rx + H Rx  
Los  i  =  H 0Tx + H Tx  – ------------------------------------------------------------------------------- Res  i 
d
where,
H Rx is the receiver antenna height above the ground (m).
i is the point index.

Res is the profile resolution (distance between two points).
2nd step: Atoll extracts the transmitter-receiver terrain profile.
3rd step: Hills and mountains are already taken into account in diffraction calculations. Therefore, in order for them not to
unfavourably influence the regression line calculation, Atoll filters the terrain profile.
Atoll calculates two filtered terrain profiles; one established from the transmitter and another from the receiver. It deter-
mines filtered height of every profile point. Profile points are evenly spaced on the basis of profile resolution. To determine
filtered terrain height at a point, Atoll evaluates ground slope between two points and compares it with a threshold set to
0.05; where three cases are possible.
Some notations defined hereafter are used in next part.
H filt is the filtered height.
H orig is the original height. Original terrain height is determined from extracted ground profile.
- Filter starting from transmitter
Let us assume that H filt – Tx  Tx  = H orig  Tx 
For each point, we have three different cases:
H orig  i  – H orig  i – 1 
1st case: If H orig  i   H orig  i – 1  and ------------------------------------------------------  0.05 ,
Res
Then, H filt – Tx  i  = H filt – Tx  i – 1  +  H orig  i  – H orig  i – 1  
© Forsk 2009 AT271_TRG_E6 87

H orig  i  – H orig  i – 1 
2nd case: If H orig  i   H orig  i – 1  and ------------------------------------------------------  0.05
Res
Then, H filt – Tx  i  = H filt – Tx  i – 1 
3rd case: If H orig  i   H orig  i – 1 
Then, H filt – Tx  i  = H filt – Tx  i – 1 
If H filt  i   H orig  i  additionally
Then, H filt – Tx  i  = H orig  i 
- Filter starting from receiver
Let us assume that H filt  Rx  = H orig  Rx 
For each point, we have three different cases:
H orig  i  – H orig  i + 1 
1st case: If H orig  i   H orig  i + 1  and -------------------------------------------------------  0.05 ,
Res
Then, H filt – Rx  i  = H filt – Rx  i + 1  +  H orig  i  – H orig  i + 1  
H orig  i  – H orig  i + 1 
2nd case: If H orig  i   H orig  i + 1  and -------------------------------------------------------  0.05
Res
Then, H filt – Rx  i  = H filt – Rx  i + 1 
3rd case: If H orig  i   H orig  i + 1 
Then, H filt – Rx  i  = H filt – Rx  i + 1 
If H filt  i   H orig  i  additionally
Then, H filt – Rx  i  = H orig  i 
Then, for every point of profile, Atoll compares the two filtered heights and chooses the higher one.
H filt  i  = max  H filt – Tx  i  H filt – Rx  i  
4th step: Atoll determines the influence area, R. 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).
The influence area must satisfy additional conditions:
• R  3000m ,
• R  0.01  d ,
• R must contain at least three bins.
Notes:
• When several influence areas are possible, Atoll chooses the highest one.
• If d < 3000m, R = d.
5th step: Atoll performs a linear regression on the filtered profile within R in order to determine a regression line.
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,
1
H m = ---
n  Hfilt  i 
i
i is the point index. Only points within R are taken into account.
R
d m = d – ----
2
d(i) is the distance between i and the transmitter (m).
Then, Atoll extends the regression line to the transmitter location. Therefore, its equation is:
88 AT271_TRG_E6 © Forsk 2009

regr  i  = a   i  Res  + b
6th step: Then, Atoll calculates effective transmitter antenna height, H Txeff (m).
H 0Tx + H Tx – b
H Txeff = -------------------------------------
-
2
1+a
If HTxeff is less than 20m, Atoll recalculates it with a new influence area, which begins at transmitter.
Notes:
• In case H Txeff  1000m , 1000m will be used in calculations.
• If H Txeff is still less than 20m, an additional correction is taken into account (7th step).
7th step: If H Txeff is still less than 20m (even negative), Atoll evaluates path loss using H Txeff = 20m and applies a
correction factor.
Therefore, if H Txeff  20m ,
L model = L model   H Txeff = 20m  d f  + K lowant
d 20   1 –  H Txeff – 20  
where, K lowant = --------- –  0.3   H Txeff – 20   – ------------------------------------------------------------------------------
5 d -   6.93 + ------------ d -
10  9.63 + ------------
 1000  1000
4.4.3.2.3 Effective Receiver Antenna Height

H Rxeff =  H Rx + H 0Rx  – H 0Tx
where,
H Rx is the receiver antenna height above the ground (m).
H 0Rx is the ground height (ground elevation) above sea level at the receiver (m).
H 0Tx is the ground height (ground elevation) above sea level at the transmitter (m).
Note:
• The calculation of effective antenna heights ( H Rxeff and H Txeff ) is based on extracted
DTM profiles. They are not properly performed if you have not imported heights (DTM file)
beforehand.
4.4.3.2.4 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.
Therefore, if the receiver is in the transmitter line of sight and the Hilly terrain correction option is active, 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, 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. Influence area, R, and regression line are supposed available.
1st step: For every profile point within influence area, Atoll calculates height deviation between the original terrain profile
and regression line. Then, it sorts points according to the deviation and draws two lines (parallel to the regression line),
one which is exceeded by 10% of the profile points and the other one by 90%.
2nd step: Atoll evaluates the terrain roughness, h; it is the distance between the two lines.
3rd step: Atoll calculates K hill LOS .
We have K hill LOS = K h + K hf
If 0  h  20m , K h = 0
2
Else K h = 7.73  log  h   – 15.29 log  h  + 6.746
If 0  h  10m , K hf = – 2  0.1924   H 0Rx + H Rx – regr  i Rx  
2 H 0Rx + H Rx – regr  i Rx 
Else K hf = – 2   – 1.616  log  h   + 14.75 log  h  – 11.21   ------------------------------------------------------------
h
© Forsk 2009 AT271_TRG_E6 89

iRx is the point index at receiver.
4.4.3.2.5 Diffraction
Four methods are available to calculate diffraction loss over the transmitter-receiver profile. They are detailed in the
Appendices.
Along the transmitter-receiver profile, you may consider:
• Either ground altitude and clutter height (Consider heights in diffraction option),
In this case, Atoll uses clutter height information from clutter heights file if available in the .atl document. Otherwise,
it considers average clutter height specified for each clutter class in the clutter classes file description.
• Or only ground altitude.
4.4.3.2.6 Losses due to Clutter

n
Atoll calculates f(clutter) over a maximum distance from receiver: f  clutter  =  Li wi

i=1
where,
L: loss due to clutter defined in the Clutter tab by the user (in dB).
w: weight determined through the weighting function.
n: number of points taken into account over the profile. Points are evenly spaced depending on the profile resolution.
Four weighting functions are available:
1
• Uniform weighting function: w i = ---
n
di
• Triangular weighting function: w i = -------------
n
-
 dj
j=1
• d i = D – d' i , where d’i is the distance between the receiver and the ith point and D is the maximum distance
defined.
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 dj
----
e
D
–1
j=1
The chart below shows the weight variation with the distance for each weighting function.
Figure 4.12: Losses due to Clutter
90 AT271_TRG_E6 © Forsk 2009

4.4.3.2.7 Recommendations
Beware that the clutter influence may be taken into account in two terms, Diffraction loss and f(clutter) at the same time.
To avoid this, we advise:
1. Not to consider clutter heights to evaluate diffraction loss over the transmitter-receiver profile if you specify losses
per clutter class.
This approach is recommended if the clutter height information is statistical (clutter roughly defined, no alti-
tude).
Or
2. Not to define any loss per clutter class if you take clutter heights into account in the diffraction loss.
In this case, f(clutter)=0. Losses due to clutter are only taken into account in the computed Diffraction loss term.
This approach is recommended if the clutter height information is either semi-deterministic (clutter roughly
defined, altitude defined with an average height per clutter class) or deterministic (clutter sharply defined, altitude
defined with an average height per clutter class or - even better - via a clutter height file).
In case of semi-deterministic clutter information, specify receiver clearance (m) per clutter class. Both ground altitude and
clutter height are considered along the whole transmitter-receiver profile except over a specific distance around the
receiver (clearance), where Atoll proceeds as if there was only the DTM map. The clearance information is used to model
streets.
Figure 4.13: Tx-Rx profile
In the above figure, the ground altitude and clutter height (in this case, average height specified for each clutter class in
the clutter classes map description) are taken into account along the profile.
Clearance definition is not necessary in case of deterministic clutter height information. Clutter height information is accu-
rate enough to be used directly without additional information such as clearance. Two cases can be considered:
1. If the receiver is in the street (clutter height lower than receiver height), Atoll calculates the path loss by considering
potentially some diffraction loss at reception.
2. If the receiver is supposed to be inside a building (clutter height higher than receiver height), Atoll does not con-
sider any difraction (and clearance) from the building but takes into account the clutter class indoor loss as an addi-
tional penetration loss.
Notes:
• To consider indoor losses in building only when using a deterministic clutter map (clutter
height map), 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).
Configuration tab)
• Even with no clearance, the clutter height (extracted either from clutter class or clutter
height folders) is never considered at the last profile point.
4.4.3.3 Automatic SPM Calibration

The goal of this tool is to calibrate parameters and methods of the SPM formula in a simple and reproducible way. Cali-
bration is based on imported CW measurement data. It is the process of limiting the difference between predicted and
measured values. For a complete description of the calibration procedure (including the very important prerequisite filtering
work on the CW measurement points), please refer to the User Manual and the SPM Calibration Guide.
The following SPM formula parameters can be estimated:
• K1, K2, K3, K4, K5, K6 and K7
© Forsk 2009 AT271_TRG_E6 91

• 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.
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. A calibrated model must restore the behaviour of CW measurements depend-
ing on their configuration on a large scale, and not just totally coincide with a few number of CW measurements. The cali-
brated model has to give correct results for every new CW measurement point in the same geographical zone, without
having been calibrated on these new CW measurements.
4.4.3.3.1 General Algorithm

Propagation model calibration is a special case of the more general Least-Square problems, i.e. given a real m x n matrix
A, and a real m-vector b, find a real n-vector x0 that minimises the Euclidean length of Ax - b.
Here,
m is the number of measurement points,
n is the number of parameters to calibrate,
A is the values of parameter associated variables (log(d), log(heff), etc.) at each measurement point, and
b is the vector of measurement values.
The vector x0 is the set of parameters found at the end of the calibration.
The theoretical mathematical solution of this problem was found by Gauss (around 1830). Further enhancements to the
original method were proposed in the 60's in order to solve the numerical instability problem.
In 1974, Lawson & Hanson [2] proposed a theoretical solution of the least-square problem with general linear inequality
constraints on the vector x0. Atoll implementation is based on this method, which is explained in detail in [1].
References:
[1] Björck A. “Numerical Methods for Least Square Problems”, SIAM, 1996.
[2] Lawson C.L., Hanson R.J. “Solving Least Squares Problems”, SIAM, 1974.
4.4.3.3.2 Sample Values for SPM Path Loss Formula Parameters

The following tables list some sample orders of magnitudes for the different parameters composing the Standard Propa-
gation Model formula.
Minimum Typical Maximum

K1 Variable Variable Variable
K2 20 44.9 70
K3 -20 5.83 20
K4 0 0.5 0.8
K5 -10 -6.55 0
K6 -1 0 0
K7 -10 0 0
K1 depends on the frequency and the technology. Here are some sample values:
Project type Frequency (MHz) K1

GSM 900 935 12.5
GSM 1800 1805 22
GSM 1900 1930 23
UMTS 2110 23.8
1xRTT 1900 23
2300 24.7
2500 25.4
WiMAX 2700 26.1
3300 27.8
3500 28.3
92 AT271_TRG_E6 © Forsk 2009

The above K1 values for WiMAX are extrapolated estimates for different frequency ranges. It is highly recommended to
calibrate the SPM using measurement data collected on the field for WiMAX networks before using the SPM for predic-
tions.
All K paramaters can be defined by the automatic calibration wizard. Since Kclutter is a constant, its value is strongly
dependant on the values given to the losses per clutter classes. From experienced users, the typical losses (in dB) per
clutter class are:
Dense urban From 4 to 5

Woodland From 2 to 3
Urban 0
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.
If you want to calibrate the losses per clutter class (Kclutter != 0) and press the Identify (Clutter row selected), a warning
message appears asking you to force the Max distance to 0 (in the Clutter tab), if it is not already set to 0. In fact, Atoll
uses the following process on these constants:
• 1st step: Atoll makes groups of measurement points according to the clutter class on which they are located.
• 2nd step: Atoll calculates the mean error for each of this group with Max distance = 0 (i.e. the clutter loss is applied
only on the reception bin).
• 3rd step: For each group, the mean error is then automatically shifted to 0 by manipulating the corresponding loss.
For example, if the mean error on a certain clutter class is 5 dB and if the initial loss for this clutter class is 2 dB,
then the calibrated loss becomes 7 dB.
Therefore, the global mean error on all the measurement points is null. In short, it is not possible to calibrate the clutter
losses if Max distance != 0. This parameter must be forced to 0.
Note:
• The Standard Propagation Model is deduced from the Hata formulae, valid in the case of
an urban environment. The above values are consistent since they are normalized with
respect to the urban clutter class (0 dB for urban clutter class). Positive values correspond
to denser clutter classes and negative values to less dense clutter classes.
4.4.3.4 Unmasked Path Loss Calculation

You can use the SPM to calculate unmasked path losses. Unmasked path losses are calculated by not taking into account
the transmitter antenna patterns, i.e., the attenuation due to the transmitter antenna pattern is not included. Such path
losses are useful when using path loss matrices calculated by Atoll with automatic optimisation tools.
The instance of the SPM available by default, under the Propagation Models folder in the Modules tab, has the following
characteristics:
• Signature: {D5701837-B081-11D4-931D-00C04FA05664}
• Type: Atoll.StdPropagModel.1
You can access these parameters in the Propagation Models table by double-clicking the Propagation Models folder in
the Modules tab.
To make the SPM calculate path losses excluding the antenna pattern attenuation, you have to change the type of the
SPM to:
• Type: Atoll.StdPropagModelUnmasked.1
However, changing the type only does not invalidate the already calculated path loss matrices, because the signature of
the propagation model is still the same. If you want Atoll to recognize that the SPM has changed, and to invalidate the path
loss matrices calculated with this model, you have to change the signature of the model as well. 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. Atoll automatically creates different signatures for different instances of the
same propagation model. Therefore, it is possible to create different instances of the SPM, with different parameter
settings, and create unmasked versions of these instances.
You can change the signature and type of the original instance of the SPM, but it is recommended to make a copy, i.e., a
new instance, of the SPM in order not to lose the original SPM parameters. So, you will be able to keep different versions
of the SPM, those that calculate path losses with antenna pattern attenuation, and others that calculate path losses without
it.
© Forsk 2009 AT271_TRG_E6 93

The usual process flow of an ACP working on an Atoll document through the API would be to:
1. Backup the storage directory of path loss matrices.
2. Set a different storage directory for calculating and storing unmasked path loss matrices.
3. Select the SPM used, backup it’s signature, and change its signature and type as shown above.
4. Perform optimisation using the path loss matrices calculated by the unmasked version of the SPM.
5. Restore the type and the signature of the SPM.
6. Reset the path loss storage directory to the original one.
Notes:
• It is not possible to calibrate the unmasked version of the SPM using measurement data.
• You can also use Atoll.ini options, AngleCalculation = 2000 and AngleCalculation = 3000,
for calculating unmasked path losses and angles of incidence, respectively. These options
are only available for the propagation models available with Atoll by default. Please refer to
the Administrator Manual for details.
• Using the SPM, you can also calculate the angles of incidence by creating a new instance
of the SPM with the following characteristics:
Type: Atoll.StdPropagModelIncidence.1
Signature: {659F0B9E-2810-4e59-9F0D-DA9E78E1E64B}
Important:
• The "masked" version of the algorithm has not been changed. It still takes into account
Atoll.ini options. However, the "unmasked" version does not take Atoll.ini options into
account.
• It’s highly recommended to use one method (Atoll.ini options) or the other one (new
identifier & signature) but not to combine both.
4.4.4 WLL Propagation Model

4.4.4.1 WLL Path Loss Formula
L model = FreeSpaceLoss + DiffractionLoss

4.4.4.2.1 Free Space Loss
Please refer to the Appendices for further details about free space loss calculation.
Atoll calculates diffraction loss along the transmitter-receiver profile built from DTM and clutter maps. Therefore, losses
due to clutter are taken into account in diffraction losses. Atoll takes clutter height information from the clutter heights file
if available in the .atl document. Otherwise, it considers average clutter height specified for each clutter class in the clutter
classes file description.
The Deygout construction (considering 3 obstacles) is used. This method is detailed in the Appendices.
Receiver Clearance
Define receiver clearance (m) per clutter class when clutter height information is either statistical or semi-deter-
ministic. Both ground altitude and clutter height are considered along the whole profile except over a specific distance
around the receiver (clearance), where Atoll proceeds as if there was only the DTM map (see SPM part). Atoll uses the
clearance information to model streets.
If the clutter is deterministic, do not define any receiver clearance (m) per clutter class. In this case, clutter height
information is accurate enough to be used directly without additional information such as clearance (Atoll can locate
streets).
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, Atoll computes Lmodel on each calculation bin using the formula defined
above. When selecting the option ‘Line of sight only’, Atoll checks for each calculation bin if the Diffraction loss (as defined
in the Diffraction loss: Deygout part) calculated along profile equals 0.
94 AT271_TRG_E6 © Forsk 2009

• In this case, receiver is considered in ‘line of sight’ and Atoll computes Lmodel on each calculation bin using the
formula defined above.
• Otherwise, Atoll considers that Lmodel tends to infinity.
4.4.5 ITU-R P.526-5 Propagation Model

L model = FreeSpaceLoss + DiffractionLoss

Atoll calculates diffraction loss along the transmitter-receiver profile is built from the DTM map. The Deygout construction
(considering 3 obstacles), with or without correction, is used. These methods are detailed in the Appendices.

If d<1 km, L model = FreeSpaceLoss
If d>1000 km, L model = 1000
If 1<d<1000 km, L model = max  FreeSpaceLoss CorrectedS tan dardLoss 
d is the distance between the transmitter and the receiver (km).

4.4.6.2.2 Corrected Standard Loss

This formula is given for a 60 dBm (1kW) transmitter power.
CorrectedS tan dardLoss = 60 – C n – A H – A cl – 108.75 + 31.54 – 20 log f

Rxeff
where,
Cn is the field strength received in dBV/m,
AH is a correction factor for effective receiver antenna height (dB),

Rxeff
Acl is the correction for terrain clearance angle (dB),

f is the frequency in MHz.
Cn Calculation
The Cn value is determined from charts Cn=f(d, HTxeff).

In the following part, let us assume that Cn=En(d,HTxeff) (where En(d,HTxeff) is the field received in dBV/m) is read from
charts for a distance, d (in km), and an effective transmitter antenna height, HTxeff (in m).
First of all, Atoll evaluates the effective transmitter antenna height, H Txeff , as follows:
If 0  d  3km , H Txeff = H 0Tx + H Tx – H 0Rx
If 3  d  15km , H Txeff = H 0Tx + H Tx – H 0  3 ;d 
If 15  d , H Txeff = H 0Tx + H Tx – H 0  3 ;15 
where,
H Tx is the transmitter antenna height above the ground (m).
© Forsk 2009 AT271_TRG_E6 95

H 0Tx is the ground height (ground elevation) above sea level at the transmitter (m).
H 0  3 ;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).
H 0  3 ;15  is the average ground height (m) above sea level for the profile between a point 3 km and another 15 km from
transmitter.
Then, depending on d and HTxeff, Atoll determines Cn using bilinear interpolation as follows.
If 37.5 HTxeff 1200, Cn= En(d,HTxeff)
Otherwise, Atoll considers d horizon = 4.1  H Txeff (d is stated in km)
Therefore,
If HTxeff < 37.5
If d  d horizon , we have C n = E n  d + 25 – d horizon 37.5 
Else Cn=En(d, 37.5) – En(dhorizon, 37.5) + En(25, 37.5)

If HTxeff > 1200
If d  d horizon , we have C n = E n  d + 142 – d horizon 1200 
Else Cn=En(d, 1200) – En(dhorizon, 1200) + En(142, 1200)
AHRxeff Calculation
c H Rx
AH = ---  20  log  ----------
Rxeff 6  10 
where,
HRx is the user-defined receiver height,
c is the height gain factor.

Note:
• c values are provided in the recommendation 370-7; for example, c=4 in a rural case.
Acl Calculation
2
If f  300 MHz, A cl = 8.1 –  6.9 + 20 log     – 0.1  + 1  +   – 0.1   
2
Otherwise, A cl = 14.9 –  6.9 + 20 log     – 0.1  + 1  +   – 0.1   
f
With  = –   4000  ----------
300
where,
 is the clearance angle (in radians) determined according to the recommendation 370-7 (figure 19),
f is the frequency stated in MHz.
4.4.7 Erceg-Greenstein (SUI) Propagation Model

Erceg-Greenstein propagation model is a statistical path loss model derived from experimental data collected at 1.9 GHz
in 95 macrocells. The model is for suburban areas, and it distinguishes between different terrain categories called the Stan-
ford University Interim Terrain Models. This propagation model is well suited for distances and base station antenna
heights that are not well-covered by other models. The path loss model applies to base antenna heights from 10 to 80 m,
base-to-terminal distances from 0.1 to 8 km, and three distinct terrain categories.
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  ------------- . This is a fixed quantity which depends upon the frequency of operation. d is the distance
  
between the base station antenna and the receiver terminal and d0 is a fixed reference distance (100 m). a(HBS) is the
correction factor for base station antenna heights, HBS:
c
a  HBS  = a – b  H BS + ----------
H BS
96 AT271_TRG_E6 © Forsk 2009

Where 10 m  HBS  80 m , and a, b, and c are correction coefficients which depend on the SUI terrain type.
The Erceg-Greenstein propagation model is further developed through the correction factors introduced by the Stanford
University Interim model. The standards proposed by the IEEE working group 802.16 include channel models developed
by Stanford University. 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 
 d 0
f
Where a(f) is the correction factor for the operating frequency, a  f  = 6  Log 10  ------------- , with f being the operating
 2000
HR
frequency in MHz. a(HR) is the correction factor for the receiver antenna height, a  H R  = X  Log 10  ------- , where d
 2
depends on the terrain type.
Note:
• a(HR) = 0 for HR = 2 m.
References:
[1] V. Erceg et. al, “An empirically based path loss model for wireless channels in suburban environments,” IEEE J.
Select Areas Commun., vol. 17, no. 7, July 1999, pp. 1205-1211.
[2] Abhayawardhana, V.S.; Wassell, I.J.; Crosby, D.; Sellars, M.P.; Brown, M.G.; "Comparison of empirical propagation
path loss models for fixed wireless access systems," Vehicular Technology Conference, 2005. IEEE 61st Volume 1, 30
May-1 June 2005 Page(s):73 - 77 Vol. 1
4.4.7.1 SUI Terrain Types

The SUI models are divided into three types of terrains1, namely A, B and C.
• Type A is associated with maximum path loss and is appropriate for hilly terrain with moderate to heavy tree
densities.
• Type B is characterised with either mostly flat terrains with moderate to heavy tree densities or hilly terrains
with light tree densities.
• Type C is associated with minimum path loss and applies to flat terrain with light tree densities.
The constants used for a, b, and c are given in the table below.
Model Parameter Terrain A Terrain B Terrain C

a 4.6 4.0 3.6
b (m-1) 0.0075 0.0065 0.005
c (m) 12.6 17.1 20

X 10.8 10.8 20
4.4.7.2 Erceg-Greenstein (SUI) Path Loss Formula

The Erceg-Greenstein (SUI) propagation model formula can be simplified from the following equation:
4d 0 d
PL = 20  Log 10  ------------- + 10  a  H BS   Log 10  ------ + a  f  – a  H R  (1)
    d 0
to the equation below:
PL = – 7.366 + 26  Log 10  f  + 10  a  H BS    1 + Log 10  d   – a  H R  (2)
Where,
• 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,
Lu = – 7.366 + 26  Log 10  f  + 10  a  H BS    1 + Log 10  d  
1. The word ‘terrain’ is used in the original definition of the model rather than ‘environment’. Hence it is used
interchangeably with ‘environment’ in this subsection.
© Forsk 2009 AT271_TRG_E6 97

The above path loss formulas are valid for d > d0, i.e. d > 100 m. For d < 100 m, 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. Simplifying the above equation, we get,
PL = 12.634 + 26  Log 10  f  + 20  Log 10  d  – a  H R  , or Lu = 12.634 + 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  , i.e. 26. Atoll uses the same coef-
ficient as the one you enter for Log 10  f  in Atoll for the case d > d0.
Note:
• You can get the same equation, i.e., Lu = 12.634 + 26  Log 10  f  + 20  Log 10  d  , by
setting a(hBS) = 2.

The Erceg-Greenstein (SUI) propagation model takes DTM into account between the transmitter and the receiver, and it
can also take clutter into account at the receiver location.
1st step: For each pixel in the calculation radius, Atoll determines the clutter bin on which the receiver is located. This clutter
bin corresponds to a clutter class. Atoll uses the Erceg-Greenstein (SUI) path loss formula assigned to this clutter class to
evaluate path loss.
2nd step: This step depends on whether the ‘Add diffraction loss’ option is selected or not.
• If the ‘Add diffraction loss’ option is not selected, 1st step gives the final path loss result.
a. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation meth-
od.
es due to diffraction L Diffraction . For more information on the Deygout method, see "3 Knife-Edge Deygout
Method" on page 105.
The final path loss is the sum of the path loss determined in 1st step and L Diffraction .
Shadow fading is computed in Atoll independent of the propagation model. For more information on the shadow fading
calculation, see "Shadowing Model" on page 113.

This propagation model is based on the P.1546-2 recommendations of the ITU-R. These recommendations extend the
P.370-7 recommendations, and are suited for operating frequencies from 30 to 3000 MHz. The path loss is calculated by
this propagation model with the help of graphs available in the recommendations. The graphs provided in the recommen-
dations represent field (or signal) strength, given in db  V  m  , as a function of distance for:
• Nominal frequencies, f n : 100, 600, and 1000 MHz
The graphs provided for 100 MHz are applicable to frequencies from 30 to 300 MHz, those for 600 MHz are appli-
cable to frequencies from 300 to 1000 MHz, and the graphs for 1000 MHz are applicable to frequencies from 1000
to 3000 MHz. The method for interpolation is described in the recommendations (Annex 5, § 6).
• Transmitter antenna heights, h 1 : 10, 20, 37.5, 75, 150, 300, 600, and 1200 m
For any values of h 1 from 10 to 3000 m, an interpolation or extrapolation from the appropriate two curves is used,
as described in the recommendations (Annex 5, § 4.1). For h 1 below 10 m, the extrapolation to be applied is given
in Annex 5, § 4.2. It is possible for the value of h 1 to be negative, in which case the method is given in Annex 5,
§ 4.3.
• Time variability, t : 1, 10, and 50 %

The propagation curves represent the field strength values exceeded for 1, 10 and 50 % of time.
• Receiver antenna height, h 2 : 10 m
For land paths, the graphs represent field strength values for a receiver antenna height above ground, equal to
the representative height of the clutter around the receiver. The minimum value of the representative height of clut-
ter is 10 m. For sea paths, the graphs represent field strength values for a receiver antenna height of 10 m.
For other values of receiver antenna height, a correction is applied according to the environment of the receiver.
The method for calculating this correction is given in Annex 5, § 9.
98 AT271_TRG_E6 © Forsk 2009

These recommendations are not valid for transmitter-receiver distances less than 1 km or greater than 1000 km. Therefore
in Atoll, the path loss between a transmitter and a receiver over less than 1 km is the same as the path loss over 1 km.
Similarly, the path loss between a transmitter and a receiver over more than 1000 km is the same as the path loss over
1000 km.
Moreover, these recommendations are not valid for transmitter antenna heights less than the average clutter height
surrounding the transmitter.
Notes:
• The cold sea graphs are used for calculations over warm and cold sea both.
• The mixture of land and sea paths is not supported by Atoll.

The input to the propagation model are the transmission frequency, transmitter and receiver heights, the distance between
the transmitter and the receiver, the precentage of time the field strength values are exceeded, the type of environment
(i.e., land or sea), and the clutter at the receiver location.
In the following calculations, f is the transmission frequency, d is the transmitter-receiver distance, and t is the percent-
age of time for which the path loss has to be calculated.
The following calculations are performed in Atoll to calculate the path loss using this propagation model.
4.4.8.1.1 Step 1: Determination of Graphs to be Used

First of all, the upper and lower nominal frequencies are determined for any given transmission frequency. The upper and
lower nominal frequencies are the nominal frequencies (100, 600, and 2000 MHz) between which the transmission
frequency is located, i.e., f n1  f  f n2 .
Once f n1 and f n1 are known, along with the information about the percentage of time t and the type of path (land or sea),
the sets of graphs which will be used for the calculation are also known.
4.4.8.1.2 Step 2: Calculation of Maximum Field Strength

A field strength must not exceed a maximum value, E Max , which is given by:
E Max = E FS = 106.9 – 20  Log  d  for land paths, and
E Max = E FS + E SE = 106.9 – 20  Log  d  + 2.38  1 – exp  – d  8.94    Log  50  t  for sea paths.
Where E FS is the free space field strength for 1 kW ERP, E SE is an enhancement for sea graphs.
4.4.8.1.3 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.
• Land paths
h 1 = h eff
• Sea paths
h 1 = Max  1 h a 
Here, all antenna heights (i.e., h 1 , h eff , and h a ) are in expressed in m. h a is the antenna height above ground and h eff
is the effective height of the transmitter antenna, which is its height over the average level of the ground between distances
of 0.2  d and d km from the transmitter in the direction of the receiver.
4.4.8.1.4 Step 4: Interpolation/Extrapolation of Field Strength

The interpolations are performed in series in the same order as described below. The first interpolation/extrapolation is
performed over the field strength values, E , from the graphs for transmitter antenna height to determine E h1 . The second
interpolation/extrapolation is performed over the interpolated/extrapolated values of E h1 to determine E d . And, the thrid
and final interpolation/extrapolation is performed over the interpolated/extrapolated values of E d to determine E f .
Step 4.1: 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, namely 10, 20,
37.5, 75, 150, 300, 600, and 1200 m, the required field strength is obtained directly from the corresponding graph. Other-
wise:
• 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:
© Forsk 2009 AT271_TRG_E6 99

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 , otherwise h Low is the nearest nominal effective height below h 1 ,
h Up = 1200 m if h 1  1200 m , otherwise h Up is the nearest nominal effective height above h 1 , E Low is the field
strength value for h Low at the required distance, and E Up is the field strength value for h Up at the required
distance.
• 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  = 4.1  h 1 , i.e., if d  4.1  h 1 ,
E h1 = E 10  d H  10   + E 10  d  – E 10  d H  h 1   , or
E h1 = E 10  12.9 km  + E 10  d  – E 10  d H  h 1   because d H  10  = 12.9 km
- For land path if the transmitter-receiver distance is greater than or equal to the smooth-Earth horizon distance
d H  h 1  = 4.1  h 1 , i.e., if d  4.1  h 1 ,
E h1 = E 10  d H  10  + d – d H  h 1   , or E h1 = E 10  12.9 km + d – d H  h 1   because d H  10  = 12.9 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.
If in the above equation, d H  10  + d – d H  h 1   1000 km even though d  1000 km , the field strength is de-
termined 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), D Up is the final tabulation distance (km), E Low is the
field strength value for D Low , and E Up is the field strength value for D Up .
- For sea path, h 1 should not be less than 1 m. This calculation requires the distance at which the path has 0.6
of the first Fresnel zone just unobstructed by the sea surface. This distance is given by:
D h1 = D 0.6  f h 1  h 2 = 10 m   (km)
Df  Dh
Where D 0.6 = Max  0.001 ------------------- (km) with D f = 0.0000389  f  h 1  h 2 (frequency-dependent term),
 D f + D h
and D h = 4.1   h 1 + h 2  (asymptotic term defined by the horizon distance).
If d  D h1 the 0.6 Fresnel clearance distance for the sea path where the transmitter antenna height is 20 m is
also calculated as:
D 20 = D 0.6  f  h 1 = 20 m   h 2 = 10 m   (km)
Once D h1 and D 20 are known, the field strength for the required distance is given by:
 E Max for d  D h1

 Log  d  D h1 
E h1 =  E D +  E D – E D   --------------------------------------
- for D h1  d  D 20
 h1 20 h1 Log  D 20  D h1 

 E'   1 – F S  + E''  F S 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 99, E D is E Max for d = D h1 ,
h1
Log  h1  10  Log  h1  10 
ED = E 10  D 20  +  E 20  D 20  – E 10  D 20    ---------------------------------- , E' = E 10  d  +  E 20  d  – E 10  d    ---------------------------------- ,
20 Log  20  10  Log  20  10 
and E'' is the field strength calculated as described for land paths. E 10  y  and E 20  y  are field strengths
interpolated for distance y and h 1 = 10 m and 20 m , respectively, and F S =  d – D 20   d .
• If h 1  0 m
A correction is applied to the field strength, E h1 , calculated in the above description in order to take into account
the diffraction and tropospheric scattering. This correction is the maximum of the diffraction correction,, and trop-
ospheric scattering correction, .
100 AT271_TRG_E6 © Forsk 2009

C h1 = Max  C h1d C h1t 
2
Where C h1d = 6.03 – J    with J    =  6.9 + 20  Log    – 0.1  + 1 +  – 0.1   and  = K    eff2 ,
–h1
 eff2 = arc tan  ------------- , and K  is 1.35 for 100 MHz, 3.31 for 600 MHz, 6.00 for 2000 MHz.
 9000
e 180  d
C h1t = 30  Log  ------------------------ with  e = ---------------------- , a = 6370 km (radius of the Earth), and k = 4  3 is the effec-
  e +  eff2 ak
tive Earth radius factor for mean refractivity conditions.
Step 4.2: Interpolation/Extrapolation of Field Strength for Transmitter-Receiver Distance
In the field strength graphs in the recommendations, the field strength is plotted against distance from 1 km to 1000 km.
The distance values for which field strengths are tabulated are 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18,
19, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200,
225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675, 700, 725, 750, 775, 800,
825, 850, 875, 900, 925, 950, 975, 1000. If the transmitter-receiver distance is a value from this list, then interpolation of
field strength is not required and the field strength can be directly read from the graphs.
If the transmitter-receiver distance does not coincide with the list of distances for which the field strengths are accurately
available from the graphs, 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 , d Up is the higher value of the nearest tabulated
distance to d , E Low is the field strength value for d Low , and E Up is the field strength value for d Up .
Step 4.3: 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, 600 MHz otherwise), f Up is the higher nominal
frequency (600 MHz if f < 600 MHz, 2000 MHz otherwise), E Low is the field strength value for f Low , and E Up is the field
strength value for f Up .
In the case of transmission frequencies below 100 MHz or above 2000 MHz, the field strength values are extrapolated
from the two nearer nominal frequency values. The above equation is used for all land paths and sea paths.
4.4.8.1.5 Step 5: Calculation of Correction Factors

Step 5.1: Correction for Receiver Antenna Height
The receiver antenna height correction depends on the type of path and clutter in which the receiver is located. The field
strength values given by the graphs for land paths are for a reference receiver antenna at a height, R (m), representative
of the height of the clutter surrounding the receiver, subject to a minimum height value of 10 m. Examples of reference
heights are 20 m for an urban area, 30 m for a dense urban area, and 10 m for a suburban area. For sea paths the notional
value of R is 10 m.
For land paths, 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' , given by R' = Max  1 --------------------------------------------------------------- .
 1000  d – 15 
Note that for h 1  6.5  d + R , R'  R .
The different correction factors are calculated as follows:

• For land path in urban and suburban zones
 6.03 – J    for h 2  R'


C Receiver =  h
  3.2 + 6.2  Log  f    Log  -----2- for h 2  R'
  R' 
R' – h 2
With J    =  6.9 + 20  Log    – 0.1  + 1 +  – 0.1   and  = 0.0108  f   R' – h 2   arc tan  ----------------- .
2
 27 
© Forsk 2009 AT271_TRG_E6 101

10
If R'  10 m , C Receiver is reduced by  3.2 + 6.2  Log  f    Log  ------ .
 R' 
• For land path other zones
h2
C Receiver =  3.2 + 6.2  Log  f    Log  ------
 10
• For sea path

d 10 and d h2 are determined as distances at which at which the path has 0.6 of the first Fresnel zone just unob-
structed by the sea surface with h 2 = 10 m and variable h 2 , respectively. These distances are given by
Df  Dh
d 10 = D 0.6  f h 1  h 2 = 10 m   and d h2 = D 0.6  f h 1 h 2  (km), respectively. Here D 0.6 = Max  0.001 -------------------
 D f + D h
as explained earlier.
h2
- If h 2  10 m , C Receiver =  3.2 + 6.2  Log  f    Log  ------
 10
h2
- If h 2  10 m and d  d 10 , C Receiver =  3.2 + 6.2  Log  f    Log  ------
 10
- If h 2  10 m and d  d 10 and d  d h2 , C Receiver = 0
h2 Log  d  d h2 
- If h 2  10 m and d  d 10 and d  d h2 , C Receiver =  3.2 + 6.2  Log  f    Log  ------   -------------------------------------
 10  Log  d 10  d h2 
Step 5.2: Correction for Short Urban/Suburban Paths
This correction is only applied when the path loss is to be calculated over land paths, over a transmitter-receiver distance
less than 15 km, in urban and suburban zones. This correction takes into account the presence of buildings in these zones.
The buildings are assumed to be of uniform height.
The correction represents a reduction in the field strength due to building clutter. It is added to the field strength and is
given by:
C Building = – 3.3  Log  f    1 – 0.85  Log  d    1 – 0.46  Log  1 + h a – R  
Where h a is the antenna height above the ground, and R is the clutter height of the clutter class where the receiver is
located. This correction is only applied when d  15 km and h 1 – R  150 m .
Step 5.3: Correction for Receiver Clearance Angle
This correction is only applied when the path loss is to be calculated over land paths, and over a transmitter-receiver
distance less than 16 km. This correction gives more precise field strength prediction over small reception areas. The
correction is added to the field strength and is given by:
C Clearance = J  '  – J   
2
Where J    =  6.9 + 20  Log    – 0.1  + 1 +  – 0.1   , ' = 0.036  f , and  = 0.065   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.
h 1S – h 2S
•  Ref : The reference angle,  Ref = arc tan  ------------------------ .
 1000  d 
Where h 1S and h 2S are the heights of the transmitter and the receiver above sea level, respectively.
4.4.8.1.6 Step 6: Calculation of Path Loss

First, the final field strength is calculated from the interpolated/extrapolated field strength, E f , by applying the corrections
calculated earlier. 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  , from which the path loss (basic transmission loss, L B ) is
calculated as follows:
L B = 139 – E + 20  Log  f 
102 AT271_TRG_E6 © Forsk 2009

4.4.9 Sakagami Extended Propagation Model

The Sakagami extended propagation model is based on the simplification of the extended Sakagami-Kuboi propagation
model. The Sakagami extended propagation model is valid for frequencies above 3 GHz. Therefore, it is only available in
WiMAX 802.16d and WiMAX 802.16e documents by default.
The Sakagami-Kuboi propagation model requires detailed information about the environment, such as widths of the streets
where the receiver is located, the angles formed by the street axes and the directions of the incident waves, heights of the
buildings close to the receiver, etc. The path loss formula for the Sakagami-Kuboi propagation model is [1]:
H 2
L Model = 100 – 7.1  Log  W  + 0.023   + 1.4  Log  h s  + 6.1  Log  H 1  – 24.37 – 3.7   --------  Log  h b  +
 h b0
13   Log  f  – 3.23 
 43.2 – 3.1  Log  h b    Log  d  + 20  Log  f  + e
Where,
• 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
0.5 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,
which also allows a simplification in terms of the input required by the model.
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:
H0 W hm
a = a  H 0  + a  W  + a  h m  = 11  Log  ------- – 7.1  Log  ------ – 5  Log  --------
 20  20  1.5
• 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 km <d< 3 km
0.8 GHz <f< 8 GHz
1.5 m < hm < 5m
Studies also show that above 3 GHz, the path loss predicted by the extended model is almost independant of the input
parameters such as street widths and angles. Therefore, the extended Sakagami-Kuboi propagation model can be simpli-
fied to the extended Sakagami propagation model:
© Forsk 2009 AT271_TRG_E6 103

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 < 100 m

0.1 km <d< 3 km
3 GHz <f< 8 GHz
1.5 m < hm < 5m
The path loss calculation formula of the Sakagami extended propagation model resembles the formula of the Standard
Propagation Model. In Atoll, this model is in fact a copy of the Standard Propagation Model with the following values
assigned to the K coefficients:
K1 65.4 (calculated for 3.5 GHz)

K2 40
K3 -30
K4 0
K5 0
K6 0
K7 -5
For more information on the Standard Propagation Model, see "Standard Propagation Model (SPM)" on page 85.
References:
[1] Manuel F. Catedra, Jesus Perez-Arriaga, "Cell Planning for Wireless Communications," Artech House Publishers,
1999.
[2] Koshiro Kitao, Shinichi Ichitsubo, "Path Loss Prediction Formula for Urban and Suburban Areas for 4G Systems,"
IEEE, 2006.
4.4.10 Appendices
4.4.10.1 Free Space Loss
The calculation of free space loss is based on ITU 525 recommendations.
FreeSpaceLoss = 32.4 + 20 log  f  + 20 log  d 

where,
f is the frequency in MHz,
d is the Tx-Rx distance in km,
Free space loss is stated in dB.
4.4.10.2 Diffraction Loss

The calculation of diffraction is based on ITU 526-5 recommendations. General method for one or more obstacles (knife-
edge diffraction) is used to evaluate diffraction losses (Diffraction loss in dB). Four construction modes are implemented
in Atoll. All of them are based on this same physical principle presented hereafter, but differ in the way they consider one
or several obstacles. Calculations take the earth curvature into account through the effective Earth radius concept (K
factor=1.333).
4.4.10.2.1 Knife-Edge Diffraction

The procedure checks whether a knife-edge obstructs the first Fresnel zone constructed between the transmitter and the
receiver. The diffraction loss, J(), depends on the obstruction parameter (), which corresponds to the ratio of the obstruc-
tion height (h) and the radius of the Fresnel zone (R).
104 AT271_TRG_E6 © Forsk 2009

Figure 4.14: Knife-Edge Diffraction
c0  n  d1  d2
R = ---------------------------------
-
f   d1 + d2 
where,
n is the Fresnel zone index,
c0 is the speed of light (2.99792 x108 ms-1),

f is the frequency in Hz
d1 is the distance from the transmitter to obstacle in m,
d2 is the distance from obstacle to receiver in m.
We have:  = h
---
r
where,
R
r = -------
2
h is the obstruction height (height from the obstacle top to the Tx-Rx axis).
Hence,
2
For 1 knife-edge method, if   – 0.7 , J    = 6.9 + 20  log    – 0.1  + 1 +   – 0.1  
Else, J    = 0
Note:
• In case of multiple-knife edge method, the minimum  required to estimate diffraction loss
is -0.78.
4.4.10.2.2 3 Knife-Edge Deygout Method

The Deygout construction, limited to a maximum of three edges, is applied to the entire profile from transmitter to receiver.
This method is used to evaluate path loss incurred by multiple knife-edges. Deygout method is based on a hierarchical
knife-edge sorting used to distinguish the main edges, which induce the largest losses, and secondary edges, which have
a lesser effect. The edge hierarchy depends on the obstruction parameter () value.
© Forsk 2009 AT271_TRG_E6 105

1 Obstacle
Figure 4.15: 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 calcu-
lated. The obstruction position, di, is also recorded. i are evaluated from these data. The point with the highest  value is
termed the principal edge, p, and the corresponding loss is J(p).
Therefore, we have
DiffractionLoss = J   P 
3 Obstacles
Then, the main edge (point p) is considered as a secondary transmitter or receiver. Therefore, the profile is divided in two
parts: one half profile, between the transmitter and the knife-edge section, another half, constituted by the knife-edge-
receiver section.
Figure 4.16: Deygout Construction – 3 Obstacles
The same procedure is repeated on each half profile to determine the edge with the higher . The two obstacles found,
(points t and r), are called ‘secondary edges’. Losses induced by the secondary edges, J(t) and J(r), are then calculated.
Once the edge hierarchy is determined, the total loss is evaluated by adding all the intermediary losses obtained.
Therefore, if  P  0
we have DiffractionLoss = J   P  + J   t  + J   r 
Otherwise, If  P  – 0.7 , DiffractionLoss = J   P 
Note:
• In case of ITU 526-5 and WLL propagation models, Diffraction loss term is determined as
follows:
- If  P  – 0.78 , we have DiffractionLoss = J   P  +  J   t  + J   r    t
J  P 
Where, t = min  -------------- 1
 6 
- Otherwise DiffractionLoss = 0
106 AT271_TRG_E6 © Forsk 2009

4.4.10.2.3 Epstein-Peterson Method

The Epstein-Peterson construction is limited to a maximum of three edges. First, Deygout construction is applied to deter-
mine the three main edges over the whole profile as described above. Then, the main edge height, hp, is recalculated
according to the Epstein-Peterson construction. hp is the height above a straight line connecting t and r points. The main
edge position dp is recorded and p and J(p) are evaluated from these data.
Figure 4.17: Epstein-Peterson Construction
Therefore, we have
DiffractionLoss = J   P  + J   t  + J   r 
4.4.10.2.4 Deygout Method with Correction

The Deygout method with correction (ITU 526-5) is based on the Deygout construction (3 obstacles) plus an empirical
correction, C.
Therefore, If  P  0 ,
we have DiffractionLoss = J   P  + J   t  + J   r  + C
Otherwise DiffractionLoss = J   P  + C
Note:
• In case of ITU 526-5 propagation model, Diffraction loss term is determined as follows:
- If  P  – 0.78 , we have DiffractionLoss = J   P  + t   J   t  + J   r  + C 
J  P 
Where, t = min  -------------- 1
 6 
C = 8.0 + 0.04d (d: distance stated in km between the transmitter and the receiver).
- Otherwise DiffractionLoss = 0
4.4.10.2.5 Millington Method

The Millington construction, limited to a single edge, is applied over the entire profile. Two horizon lines are drawn at the
transmitter and at the receiver. A straight line between the transmitter and the receiver is defined and the height of the
intersection point between the two horizon lines above the Tx-Rx axis, hh, is calculated. The position dh is recorded and
then, from these values, h and J(h) are evaluated using the same previous formulas.
Therefore, we have
DiffractionLoss = J   h 
© Forsk 2009 AT271_TRG_E6 107

Figure 4.18: Millington Construction
4.5 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). For each measured transmitter, 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. A transmitter path loss
matrix can be tuned several times by the use of several measurement paths. All these tuning paths are stored in a cata-
logue. This catalogue is stored under a .tuning folder containing a .dbf file and one .pts file per corrected transmitter. Since
a tuning file can contain several measurement paths, all these measurements are added to the tuning file.
For more information on the tuning files, See "Externalised Tuning Files" on page 61.
The same algorithm is used for CW Measurement and Test Mobile Data. It is also the same for main and extended matri-
ces.
Path Losses tuning will be done using two steps.
1. Global matrix correction
A mean error is calculated between each measured value and the corresponding bin in the pathloss matrix. Mean error is
calculated for each pathloss matrix (main and extended) of each transmitter. This mean error is then applied to all the
matrix bins. This correction is done to smooth the local corrections (step 2) of measured values and not the corrected bins.
2. Correction for each measured value
For each measured value, an ellipse is used to define the pathloss area which has to be tuned. The main axis of the ellipse
is oriented to the 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, before any correction.
The squared elliptic distance between i and M is given by:

2 2
 Xi – XM   Yi – YM 
D i = -------------------------- + --------------------------
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 correction is given by:
1
E =  ---   ei
 n
i
where:
e i is the error between measurement and prediction at point i
Then, the path loss value is corrected using E:
Pi = Pi +E
new old
Finally, a second correction ( R i ) is applied where:
Ri =  1 – Di    M – g – Pi  so R i =  1 – D i    M – g –  P i + E
new old
108 AT271_TRG_E6 © Forsk 2009

where g is (measurement gain - losses).

So, the final corrected path loss is:
Pi = Pi + R i so P i = Pi + E + Ri
tuned new tuned old
When several ellipses overlap a pathloss bin, the final corrected path loss is given by:
 
  1 – d j P j
  tuned

j
Pi = ----------------------------------------------------
tuned  
n –
 
d j

j
Where:
n is the number of overlapping ellipses
4.6 Antenna Attenuation Calculation

The modelling method used to evaluate transmitter antenna attenuation, L antTx , is described below. Atoll calculates the
accurate azimuth and tilt angles and then, performs a 3-D interpolation of horizontal and vertical patterns to determine the
attenuation of antenna.
Furthermore, you will find explanations about the remote electrical downtilt modelling.
4.6.1 Calculation of Azimuth and Tilt Angles

From the direction of the transmitter antenna and the receiver position relative to the transmitter, Atoll determines the
receiver position relative to the direction of the transmitter antenna (i.e. 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  .
aRx and eRx are respectively the azimuth and tilt of the receiver (Rx) in the coordinate system S 0  x y z  .
d is the distance between the transmitter (Tx) and the receiver (Rx).
Figure 4.19: Azimuth and Tilt Computation
In the coordinate system S 0  x y z  , the receiver coordinates are:
x Rx cos  e Rx   sin  a Rx   d
y Rx = cos  e Rx   cos  a Rx   d (1)
z Rx – 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''  . These angles describe the direction of the transmitter-receiver path in the transmitter antenna coordinate
system. Therefore, the receiver coordinates in S Tx  x'' y'' z''  are:
© Forsk 2009 AT271_TRG_E6 109

x'' Rx cos  el   sin  az   d

y'' Rx = cos  el   cos  az   d (2)
z'' Rx – sin  el   d
According to the figure above, we have the following relations:
x' cos  a Tx  – sin  a Tx  0 x

y' = sin  a Tx  cos  a Tx  0  y (3)
z' 0 0 1 z
and
x'' 1 0 0 x'
y'' = 0 cos  e Tx  – sin  e Tx   y' (4)
z'' 0 sin  e Tx  cos  e Tx  z'
Therefore, the relation between the system S 0  x y z  and the transmitter antenna system S Tx  x'' y'' z''  is:
x'' 1 0 0 cos  a Tx  – sin  a Tx  0 x

y'' = 0 cos  e Tx  – sin  e Tx   sin  a Tx  cos  a Tx  0  y (5)
z'' 0 sin  e Tx  cos  e Tx  0 0 1 z
We get,
x'' cos  a Tx  – sin  a Tx  0 x

y'' = cos  e Tx   sin  a Tx  cos  e Tx   cos  a Tx  – sin  e Tx   y (6)
z'' sin  e Tx   sin  a Tx  sin  e Tx   cos  a Tx  cos  e Tx  z
Then, substituting the receiver coordinates in the system S0 from Eq. (1) and the receiver coordinates in the system STx
from Eq. (2) in Eq. (6) leads to a system where two solutions are possible:
1st solution: If a Rx = a Tx , then az = 0 and el = e Rx – e Tx
2nd solution: If a Rx  a Tx , then
1
az = atan ------------------------------------------------------------------------------------------------
cos  e Tx  sin  e Tx   tan  e Rx 
-------------------------------------- + --------------------------------------------------
tan  a Rx – a Tx  sin  a Rx – a Tx 
and
 – sin  e Tx  cos  e Tx   tan  e Rx  

el = atan sin  az    -------------------------------------
- + ---------------------------------------------------
-
 tan  a Rx – a Tx  sin  a Rx – a Tx  
If sin  az   sin  a Rx – a Tx   0 , then az = az + 180
4.6.2 Antenna Pattern 3-D Interpolation

The direction of the transmitter-receiver path in the transmitter antenna coordinate system is given by angle values, az and
el. Atoll considers these values in order to determine transmitter antenna attenuations in the horizontal and vertical
patterns. It reads the attenuation H(az) in the horizontal pattern for the calculated azimuth angle az and the attenuation
V(el) in the vertical pattern for the calculated tilt angle el. Then, it calculates the antenna total attenuation, L antTx  az el  .
180 – az az
L antTx  az el  = H  az  – -------------------------   H  0  – V  el   + ----------   H  180  – V  180 – el  
180 180
Notes:
• We assume that the horizontal and vertical patterns are two cross-sections of the 3-D
pattern. In other words, the description of the antenna pattern must satisfy the following:
H(0)=V(0) and H()=V()
In case of an electrical tilt, , the horizontal pattern is a conical section with a  degrees
elevation off the horizontal plane. Here, horizontal and vertical patterns must satisfy the
following:
H(0)=V() and H()=V(-)
If the constraints listed above are satisfied, this implies that:
110 AT271_TRG_E6 © Forsk 2009

1. Interpolated horizontal and vertical patterns respectively fit in with the entered horizontal
and vertical patterns, even in case of electrical tilt,
2. The contribution of both the vertical pattern back and front parts are taken into account.
Otherwise, only the second point is guaranteed.
• Atoll uses this modelling method from the Atoll 2.1 version (inclusive) and above. In Atoll’s
versions prior to the 2.1, another modelling method was available to evaluate angles and
losses due to antenna pattern. The user has the option to choose between these two
methods through Atoll.ini file (see Atoll administration files). For further information about
the old modelling method, please refer to the Technical Reference Guide 2.2.
• The above interpolation is performed in dBs.
• Angle values in formulas are stated in degrees.
• The above interpolation is not used in case the transmitter antenna is described by a 3-D
antenna pattern.
4.6.3 Additional Electrical Downtilt Modelling

The additional electrical downtilt, AEDT, also referred to as remote electrical downtilt or REDT, introduces a conical trans-
formation of the 3-D antenna pattern in the vertical axis. In order to take it into account, the vertical pattern is transformed
as follows:
V  x  = V  x – AEDT  when x  [– 90,90]
V  x  = V  x + AEDT  when x  [90,270]

Where, the angle values are in degrees.
The vertical pattern transformation is represented below. 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°.
Then, Atoll proceeds as explained in the previous section. It determines the antenna attenuation in the transformed vertical
pattern for the calculated tilt angle (V(el)) and applies the 3-D interpolation formula in order to calculate the antenna total
attenuation, L antTx  az el  .
Figure 4.20: Vertical Pattern Transformation due to Electrical Downtilt
4.6.4 Antenna Pattern Smoothing

Empirical propagation models, like the Standard Propagation Model (SPM), require antenna pattern smoothing in the verti-
cal plane to simulate the effects of reflections and diffractions. Signal level predictions can be improved by smoothing the
high-attenuation points of the vertical pattern. You can smooth vertical as well as horizontal antenna patterns in Atoll.
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. Peaks (P) are the lowest attenuation angles and nulls (N) are the highest attenua-
tion angles in the pattern. Then, it determines the nulls to be smoothed (NSmoothing) and their corresponding angles accord-
ing to the defined Peak-to-Null Deviation (DPeak-to-Null). DPeak-to-Null is the minimum difference of attenuation in dBs
between two peaks and a null between them. Finally, Atoll smooths the pattern between 0 and the smoothing angle
(ASmoothing) by applying the smoothing to a certain smoothing factor (FSmoothing) defined by the user.
Let’s take an example of an antenna pattern to be smoothed, as shown in Figure 4.21: on page 112. Let DPeak-to-Null be
10 dB, ASmoothing = 90 degrees, and FSmoothing = 0.5.
© Forsk 2009 AT271_TRG_E6 111

Figure 4.21: 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.
Figure 4.22: Peaks and Nulls in the Antenna Pattern
Peaks (P) and Nulls (N):
Angle (°) Attenuation (dB)

1 0.1
15 33.5
21 13.2
30 37.6
38 16.9
49 32.2
67 15.6
Then, Atoll verifies whether the difference of attenuation at a given angle is DPeak-to-Null less than the before and after it.
This comparison determines the nulls to be smoothed (NSmoothing).
Nulls to be smoothed (NSmoothing):
Angle (°) Attenuation (dB)

15 33.5
30 37.6
49 32.2
Once the nulls are known, Atoll applies the smoothing algorithm to all the attenuation values at all the angles between the
first peak, the null, and the last peak.
112 AT271_TRG_E6 © Forsk 2009

4.6.4.1 Smoothing Algorithm

For all nulls n  N Smoothing surrounded by two peaks P1 and P2 at angles  1 and  2 ,
  A 2 – A 1 
-   i –  1  
A i Smoothed = A i – F Smoothing A i –  A  +  -----------------------
 1
 2 – 1  
Where,
i is the angle in degrees from  1 to  2 incremented by 1 degree,
AAngle is the attenuation at any given angle which can be i,  1 or  2 , and
FSmoothing is the smoothing factor defined by the user.
4.7 Shadowing Model

Propagation models predict the mean path loss as a function of transmission and reception parameters such as frequency,
antenna heights, and distance, etc. Therefore, the predicted path loss between a transmitter and a receiver is constant, in
a given environment and for a given distance. However, in reality different types of clutter may exist in the transmitter-
receiver path. Therefore, the path losses for the same distance could be different along paths that pass throught different
types of environments. 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. Some paths undergo more loss while others are less obstructed and may
have higher received signal strength. The variation of path loss with respect to the mean path loss values predicted by the
propagation models, depending on the type of environment is called shadow fading (shadowing) or slow fading. "Slow"
fading implies that the variations in the path loss due to shadow fading occur comparatively slower than the fast fading
effect (Rayleigh fading), which is due to the mobile receiving multipath copies of a signal.
Different types of clutter (buildings, hills, etc.) make large shadows that cause variations in the path loss over long
distances. As a mobile passes under a shadow, the path loss to the mobile keeps varying from point to point. Shadow
fading varies as the mobile moves, while fast fading can vary even if the mobile remains at the same location or moves
over very small distances. It is crucial to account for the shadow fading in order to predict the reliability of coverage
provided by any mobile cellular system.
The shadowing effect is modelled by a log-normal (Gaussian) distribution, as shown in Figure 4.23: on page 113, whose
standard deviation  depends on the type of clutter.
Figure 4.23: Log-normal Probability Density Function
Different clutter types have different shadowing effects. Therefore, each clutter type in Atoll can have a different standard
deviation representing its shadowing characteristics. For different standard deviations, the shape of the Gaussian distri-
bution curve remains similar, as shown in Figure 4.23: on page 113.
The accuracy of this model depends upon:
• The suitability of the range of standard deviation used for each clutter class,
• The definition (bin size) of the digital map,
• How up-to-date the digital map is,
• The number of clutter classes,
• The accuracy of assignment of clutter classes.
© Forsk 2009 AT271_TRG_E6 113

Shadowing is applied to the predicted path loss differently depending on the technology, and whether it is applied to predic-
tions or simulations. The following sections explain how shadowing margins are calculated and applied to different tech-
nology documents.
Shadowing margins are calculated for a given cell edge coverage probability. The cell edge coverage probability is the
probability of coverage at a pixel located at the cell edge, and corresponds to the reliability of coverage that you are plan-
ning to achieve at the cell edge. For example, 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. Therefore, 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, and the overall
predicted coverage area is reliable at least x % of the time.
References:
[1] Saunders S. “Antennas and propagation for Wireless Communication Systems” pp. 180-198
[2] Holma H., Toskala A. “WCDMA for UMTS”
[3] Jhong S., Leonard M. “CDMA systems engineering handbook” pp. 309-315, 1051-1053”
[4] Remy J.G., Cueugnet J., Siben C. “Systèmes de radiocommunications avec les mobiles” pp. 309-310
[5] Laiho J., Wacker A., Novosad T. “Radio network planning and optimisation for UMTS” pp. 80-81
GSM GPRS EGPRS Documents
The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 116, and
applied to signal level or C/I as explained below.
• Signal Level-Based Predictions
Signal level-based predictions include coverage predictions (Coverage by Transmitter, Coverage by Signal Level,
and Overlapping Zones) and calculations in point analysis tabs (Profile and Reception) that require calculation of
the received signal level only, and do not depend on interference.
In these calculations (signal level calculations), a shadowing margin ( M Shadowing – model ) is added to the path loss
( L path ) calculated for each pixel. 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.
• Interference-Based Predictions
Interference-based predictions include coverage predictions (Coverage by C/I Level, Interfered Zones, GPRS/
EGPRS Coding Schemes, RLC/MAC Throughout/Timeslot, Application Throughput/Timeslot, Circuit Quality Indi-
cators) and calculations in point analysis window’s Interference tab that require calculation of the received signal
level and interference received from other base stations.
In these calculations, ( C  I calculations), the shadowing margin ( M Shadowing – C  I ) is added to the ratio of the
carrier power (C) and the interfering signal levels (I) received from the interfering base stations. 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.
UMTS HSPA, IS-95 cdmaOne, and CDMA2000 1xRTT 1xEV-DO Documents
The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 116 and
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 117, and applied to signal level, Ec/I0, or Eb/Nt as
explained below.
located.
• Interference+noise-Based Predictions
Interference+noise-based predictions include coverage predictions (Pilot Reception Analysis, Downlink Total
Noise, Service Area Analyses, Handoff Status, etc.) and point analysis (AS Analysis tab) that require calculation
of the received signal level and intra-cellular interference and noise received from other base stations.
In these calculations, the shadowing margins ( M Shadowing – Ec  Io and M Shadowing –  Eb  Nt  ) , or

DL
M Shadowing –  Eb  Nt  ) are added to Ec/I0 or Eb/Nt. This shadowing margin is calculated for a given cell edge
UL
114 AT271_TRG_E6 © Forsk 2009

coverage probability and depends on the Ec/I0 or Eb/Nt standard deviations (  Ec  Io ,   Eb  Nt  , or   Eb  Nt  ,

DL UL
in dB) associated to the clutter class where the receiver is located.
• 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. 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. For detailed description of the calculation of macro-
diversity gains, please refer to "Macro-Diversity Gains Calculation" on page 118.
• Monte-Carlo Simulations
Random values for shadowing margins are calculated for each transmitter-receiver link and added to the predicted
path loss. 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 log-
normal distribution as explained above.
TD-SCDMA Documents
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 117, and applied to signal level or interference+noise
predictions as explained below.
Signal level-based predictions include coverage predictions (Best Server and RSCP P-CCPCH Coverages, P-
CCPCG Pollution, Baton Handover Coverage, DwPCH and UpPCH Coverages, Cell to Cell Interference, and
Scrambling Code Interference) and calculations in point analysis tabs (Profile and Reception) that require calcu-
lation of the received signal level only, and do not depend on interference.
located.
Interference+noise-based predictions include coverage predictions (P-CCPCH Eb/Nt and C/I Coverages, Service
Area Analsyses for downlink and uplink Eb/Nt and C/I, etc.) that require calculation of the received signal level and
interference received from other base stations.
In these calculations, the shadowing margins ( M Shadowing –  Eb  Nt  , M Shadowing –  Eb  Nt  , or

P – CCPCH DL
M Shadowing –  Eb  Nt  ) are added to Eb/Nt. This shadowing margin is calculated for a given cell edge coverage
UL
probability and depends on the Eb/Nt standard deviations (   Eb  Nt  ,   Eb  Nt  , or   Eb  Nt  , in dB)

P – CCPCH DL UL
associated to the clutter class where the receiver is located.
WiMAX 802.16d and WiMAX 802.16e Documents
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 117 , and applied to signal level or C/(I+N) as
explained below.
located.
Interference-based predictions include coverage predictions (Coverage by C/(I+N) Level, Coverage by Best
Bearer, Coverage by Channel Throughput) that require calculation of the received signal level and interference
received from other base stations.
© Forsk 2009 AT271_TRG_E6 115

In these calculations, (C/(I+N) calculations), the shadowing margin ( M Shadowing – C  I ) is added to the ratio of the
carrier power (C) and the interfering signal levels (I) and noise received from the interfering base stations. This
shadowing margin is calculated for a given cell edge coverage probability and depends on the C/I standard devi-
ation (  C  I in dB) associated to the clutter class where the receiver is located.
4.7.1 Shadowing Margin Calculation

The following sections describe the calculation method used for determining different shadowin margins.
The following shadowing margins are calculated using the method described below:
Standard
Network Type MShadowing Applied to
Deviation
 model M Shadowing – model C
GSM GPRS EGPRS
C  I M Shadowing – C  I C/I
 Ec  Io M Shadowing – Ec  Io Ec/I0
UMTS HSPA
  Eb  Nt  M Shadowing –  Eb  Nt  Eb/Nt (DL)
DL DL
  Eb  Nt  M Shadowing –  Eb  Nt  Eb/Nt (UL)

UL UL
IS-95 cdmaOne  Ec  Io M Shadowing – Ec  Io Ec/I0
CDMA2000 1xRTT 1xEV-DO   Eb  Nt  M Shadowing –  Eb  Nt  Eb/Nt (DL)

DL DL

UL UL
  Eb  Nt  M Shadowing –  Eb  Nt  Eb/Nt P-CCPCH

P – CCPCH P – CCPCH
TD-SCDMA
  Eb  Nt  M Shadowing –  Eb  Nt  Eb/Nt (DL)
DL DL

UL UL

WiMAX 802.16d
WiMAX 802.16e C  I M Shadowing – C  I C/(I+N)
4.7.1.1 Shadowing Margin Calculation in Predictions

Shadowing margins, MShadowing, are calculated from standard deviation values defined for the clutter class where the pixel
(probe mobile) is located, and required cell edge coverage probability, and added to the path loss, Lpath.
Shadowing Error PDF (1 Signal)
The measured path loss in dB can be expressed as a Gaussian random variable:
L = L path +  dB  G  0 1 
where,
• Lpath is the predicted path loss,
• dB is the user-defined standard deviation of the error,
• G(0,1) is a zero-mean unit-variance Gaussian random variable.
Therefore, the probability density function (pdf) for the random (shadowing) part of path loss is:
2
x -
– -------------
2
1 2 dB
p L  x  = ---------------------  e
 dB 2
116 AT271_TRG_E6 © Forsk 2009

The probability that the shadowing error exceeds z dB is

2
 x -
 – -------------
2
1 2 dB
PL  x  z  =  p L  x  dx = 
---------------------  e
 dB 2
dx
z z
Normalising x by dividing it bydB:
 2
x
– ------
1 z
 dx = Q  ---------
2
P L  x  z  = -----------  e
2   dB
z-
---------
 dB
where Q is the complementary cumulative function.
To ensure a given cell edge coverage probability, R L , for the predicted value, a shadowing margin, M Shadowing , is added
to the link budget.
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,
• P rec is the signal level predicted at the receiver. 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.
• L Rx are receiver losses.
• G antRx is the receiver antenna gain.
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. a set of cell edge coverage probabilities.
M Shadowing
Figure 4.24: Normalised Margin M arg in = -----------------------------
-
 dB
In interference-based predictions, where signal to noise ratio is calculated, the shadowing margin is only applied to the
signal from the interfered transmitter (C). We consider that the interference value is not altered by the shadowing margin.
Random variations also exist in the interfering signals, but taking only the average interference gives accurate results. [3]
explains how a certain level of interference is maintained by congestion control in CDMA-based networks.
4.7.1.2 Shadowing Margin Calculation in Monte-Carlo Simulations

Shadowing margins, MShadowing, are calculated from standard deviation values defined for the clutter class where the pixel
(probe mobile) is located, and required cell edge coverage probability, and added to the path loss, Lpath.
Random values are generated during Monte-Carlo simulation. Each user is assigned a service, a mobility type, an activity
status, a geographic position and a random shadowing value.
For each link, path loss (L) can be broken down to L = L path +  .
Here,  is a zero mean gaussian random variable G  0  dB  representing variation due to shadowing. It can be
expressed as the sum of two uncorrelated zero mean gaussian random variables,  L and  P .  L models the error related
to the receiver’s location (surrounding environment), 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.
Therefore, in case of two links, we have:
© Forsk 2009 AT271_TRG_E6 117

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 , the model standard deviation   model  , and the
correlation coefficient    between  1 and  2 .
Assuming all  P have the same standard deviations, we have:
2 2 2
 model =  L +  P
2
L
 = ----------------
-
2
 model
Therefore,
2 2
 P =  model   1 –  
2 2
 L =  model  
 is set to 0.5 in Atoll, which gives:
 model  model
 L = ----------------
- and  P = ----------------
-
2 2
Receiver
Therefore, to model shadowing error common to all the signals received at a receiver ( E Shadowing – model ), values are
randomly generated for each receiver. These values have a zero-mean gaussian distribution with a standard deviation of
 model
- , where   model  is the model standard deviation associated with the receiver’s clutter class.
----------------
 2 
Next, Atoll generates another random value for each transmitter-receiver pair. This values represents the shadowing error
Path
not related to the location of the receiver ( E Shadowing – model ). These values also have a zero-mean gaussian distribution
 model
with a standard deviation  ----------------- .
 2 
So, we have:
Receiver Path
E Shadowing – model = E Shadowing – model + E Shadowing – model
Random shadowing error has its mean value at zero. Hence, this shadowing modelling method has no impact on the simu-
lated network load. On the other hand, as shadowing errors on the transmitter-receiver links are uncorrelated, the method
influences the calculated macro-diversity gain in case the mobile is in soft handover.
4.7.2 Macro-Diversity Gains Calculation

The following sections explain how uplink and downlink macro-diversity gains are calculated in UMTS HSPA, IS-95
cdmaOne, and CDMA2000 1xRTT 1xEV-DO documents for predictions and AS Analysis tab of the point analysis tool.
4.7.2.1 Uplink Macro-Diversity Gain Evaluation

In UMTS, CDMA2000 and IS95-CDMA, mobiles may be in soft handoff (mobile connected to cells located on different
sites). In this case, we can consider the shadowing error pdf described below.
4.7.2.1.1 Shadowing Error PDF (n Signals)

For each link, 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. It can be expressed as
the sum of two uncorrelated zero mean gaussian random variables,  L and  P .  L models error related to the receiver
local environment; it is the same whichever the link.  P models error related to the path between transmitter and receiver.

1
 1 =  L +  P for the link 1
2
118 AT271_TRG_E6 © Forsk 2009

Knowing  i , the uplink Eb/Nt standard deviation    Eb  Nt   and the correlation coefficient  between  1 and  2 , we
UL
can calculate standard deviations of  L   L  and  P   P  (assuming all  P have the same standard deviations).
We have:
2 2 2
  Eb  Nt  = L + P
UL
2
L
 = -------------------------
-
2
  Eb  Nt 
UL
Therefore,
2 2
 P =   Eb  Nt   1 – 
UL
2 2
 L =   Eb  Nt  
UL
2 Signals Without Recombination
In technologies supporting soft handoff (UMTS, CDMA2000, IS95-CDMA), cell is interference limited. As for one link, to
ensure a required cell edge coverage probability R L for the prediction, we add to each link budget a shadowing margin,
2signals
M Shadowing –  Eb  Nt  .
UL
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
N1 1
or
Cd 1 1
--------2- = P' Tx2 – L 2 – N 2  CI pred   2  P' Tx2 – L path – N 2 – CI pred
N2 2
where
i
CI pred is the quality level (signal to noise ratio) predicted at the receiver for link i.
Ni is the noise level for link i.

We note:
2signals i
M Shadowing –  Eb  Nt  = P' Txi – L path – N i – CI pred
UL i
and
2 1 2
 1 = CI pred – CI pred
2
 1 is the minimum needed margin on each link.
Therefore, the probability of having a quality at least equal to the best predicted one is:
noMRC 2signals  Cd 1
Cd 1 
RL  M Shadowing –  Eb  Nt   = 1 – P L1 L2  --------1-  CI pred --------2-  CI pred
UL
 1N N 2 
noMRC 2signals 2signals 2signals 2

RL  M Shadowing –  Eb  Nt   = 1 – P    1  M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt  – 1 
UL 1  2 UL UL
1 2
We can express it using  L ,  P and  P
2signals 2signals 2
P   1 M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt  –  1  L =  L
1 2 UL UL
1 2signals 2 2signals 2
= P  L   P 1 2   P  M Shadowing –  Eb  Nt  –  L  P  M Shadowing –  Eb  Nt  – 1 – L 
L  P  P UL UL
2signals 2signals 2
P   1 M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt  –  1  L =  L
1 2 UL UL
= P    L   P    P  M Shadowing –  Eb  Nt  –  L  P    P  M Shadowing –  Eb  Nt  – 1 –  L 
L P UL P UL
© Forsk 2009 AT271_TRG_E6 119

noMRC 2signals
RL  M Shadowing –  Eb  Nt  
UL

 
=  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  – L 
P UL
2
  –x
---------- 2signals

 2 P
2
 M Shadowing –  Eb  Nt  UL –  L 
1
=  ------------------  e dx = Q  ---------------------------------------------------------------------
  2  P  
 P
  M 2signals –  L 
 Shadowing –  Eb  Nt  UL
Then, we have:
noMRC 2signals
UL
 2signals 2signals 2
  M Shadowing –  Eb  Nt UL –  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 2signals
UL
2
 – xL
 ---------  M 2signals
Shadowing –  Eb  Nt  UL – x L   Eb  Nt    M 2signals
Shadowing –  Eb  Nt  UL – x L   Eb  Nt  UL  –  1
2
 
1
=  1 – ----------- e   Q  ------------------------------------------------------------------------------------------------------- - dx L
-  Q  -------------------------------------------------------------------------------------------------------------------
2 UL
 2    Eb  Nt  1–     Eb  Nt  1–  
 –  UL   UL  
n Signals Without Recombination
We can generalize the previous expression to n signals (n is the number of available signals - Atoll may consider up to 3
signals):
noMRC nsignals
UL
2
 – xL
 ---------  M nsignals
Shadowing –  Eb  Nt  UL – x L   Eb  Nt    M nsignals
Shadowing –  Eb  Nt  UL – x L   Eb  Nt  UL  –  1
2
 
1
=  1 – ----------- e   Q  ------------------------------------------------------------------------------------------------------- - dx L
-  Q  -------------------------------------------------------------------------------------------------------------------
2 UL
 2    Eb  Nt  1–     Eb  Nt  1–  
 –  UL   UL  
The case where softer handoff occurs (two signals from co-site cells) is equivalent to the one signal case. The Softer/soft
case is equivalent to the two signals case. For the path associated with the softer recombination, we will use combined
SNR to calculate the availability of the link.
Correlation Coefficient Determination
There is currently no agreed model for predicting correlation coefficient    between  1 and  2 . Two key variables influ-
ence correlation:
• The angle between the two signals. If this angle is small, correlation is high.
• The relative values of the two signal lengths. If angle is 0 and lengths are the same, correlation is zero. Correlation
is different from zero when path lengths differ.
A simple model has been found [1]:
 T  D1
 =  ------ -------- when  T    
   D2
 T is a function of the mean size of obstacles near the receiver and  is also linked to the receiver environment.
In a normal handover status, assuming a hexagonal design for sites,  is close to  (+/- /3) and D1/D2 is close to 1.

In [1,5],  = 0.5 when  = 0.3 and  T = ------ .
10
In Atoll,  is set to 0.5.
120 AT271_TRG_E6 © Forsk 2009

4.7.2.1.2 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.
Therefore, we have:
UL nsignals
G macro – diversity = M Shadowing –  Eb  Nt  – M Shadowing –  Eb  Nt 
UL UL
Where n is the number of cell-mobile signals.
4.7.2.2 Downlink Macro-Diversity Gain Evaluation

In UMTS, CDMA2000 and IS95-CDMA, in case of soft handoff, mobiles are able to switch from one cell to another if the
best pilot drastically fades. To model this function, we have to consider the probability of fading over the shadowing margin,
both for the best signal and for all the other available signals, in the shadowing margin calculation.
Let us consider the shadowing error pdf described below.
4.7.2.2.1 Shadowing Error PDF (n Signals)

For each link, 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. It can be expressed as
the sum of two uncorrelated zero mean gaussian random variables,  L and  P .  L models the error related to the receiver
local environment, which is the same for all links.  P models the error related to the path between the transmitter and the
receiver.
1
2
Knowing  i , the Ec/Io standard deviation   Ec  I o  and the correlation coefficient  between  1 and  2 , 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,
2 2
 P =  Ec  I o   1 –  
2 2
 L =  Ec  I o  
2 Available Signals
In technologies supporting soft handoff (UMTS, CDMA2000 and IS95-CDMA), cells are interference limited. As for one
link, to ensure a required cell edge coverage probability R L for the prediction, we add a shadowing margin,
2signals
M Shadowing – Ec  Io , to each link budget.
Ec Ec
Prediction reliability to have -------   ------- for the best server can be expressed as:
Io  Io  pred
Ec Ec 1 Ec 1
---------1- = P pilot – L 1 – Io   -------   1  P pilot – L m – Io –  -------
Io 1  Io  pred 1 1  Io  pred
Or
Ec Ec 1 Ec 1
---------2- = P pilot – L 2 – Io   -------   2  P pilot – L m – Io –  -------
Io 2  Io  pred 2 2  Io  pred
We note:
Ec 1
M Shadowing – Ec  Io = P pilot – L m – Io –  -------
2signals
i i  Io  pred
© Forsk 2009 AT271_TRG_E6 121

Ec 1 Ec 2
 1 =  ------- –  -------
2
 Io  pred  Io  pred
2
 1 is the minimum needed margin on each link.
Therefore, probability of having a quality at least equal to the best predicted one is:
Ec 1 Ec 1 Ec 2 Ec 1 
 M Shadowing – Ec  Io  = 1 – P L1 L2  ----------   -------  ----------   -------
noMRC 2signals
RL
 Io  Io  pred Io  Io  pred
noMRC 2signals 2signals 2signals 2

RL  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 ,  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 1 2   P  M Shadowing – Ec  Io –  L  P  M Shadowing – Ec  Io –  1–  L 
L  P  P
2signals 2signals 2
P 1 2   1  M Shadowing – Ec  Io  2  M Shadowing – Ec  Io –  1  L =  L
= P    L  P    P  M Shadowing – Ec  Io – L  P    P  M Shadowing – Ec  Io – 1 –  L 
L P P
noMRC 2signals
RL  M Shadowing – Ec  Io 

 P  L   P  P  MShadowing – Ec  Io – L   P  P  MShadowing – Ec  Io – 1 – L  dL

= 1–
L P P
–
2
 –x
---------- 2signals
2
1 2 P  M Shadowing – Ec  Io –  L

i 2signals
P  P  M Shadowing – Ec  Io –  L  = ------------------ e dx = Q  -----------------------------------------------------------
P
 P 2  P 
 SHO –  L
Then, we have:
 2signals 2signals 2
 M Shadowing – Ec  Io –  L  M Shadowing – Ec  Io –  1 –  L

noMRC 2signals
RL  M Shadowing – Ec  Io  = 1 – 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 2signals
2
 –xL 2signals 2signals 2
1
---------  M Shadowing – Ec  Io – x L  Ec  I o   M Shadowing – Ec  Io –  1 – x L  Ec  I o 

2
= 1 – ----------- e  Q  ------------------------------------------------------------------------------------  Q  ------------------------------------------------------------------------------------------------ dx L
2   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 - Atoll may consider up to 3
signals):
noMRC nsignals
2
 –xL nsignals n nsignals i
1
---------  M Shadowing – Ec  Io – x L  Ec  I o   M Shadowing – Ec  Io –  1 – x L  Ec  I o 
 
2
= 1 – ----------- e  Q  ------------------------------------------------------------------------------------ x Q  ------------------------------------------------------------------------------------------------ dx L
2   Ec  I o 1 –     Ec  I o 1 –  
– i=2
122 AT271_TRG_E6 © Forsk 2009

2
 1 =1 dB
2
 1 =5 dB
2
 1 =10 dB
Figure 4.25: Margin - Probability (Case of 2 Signals)
2 signals
3
 1 =5 dB
3
 1 =10 dB
Figure 4.26: Margin - Probability (Case of 3 Signals with sigma = 8dB, delta1 = 1dB)
2 signals
3
 1 =5 dB
3
 1 =10 dB
Figure 4.27: Margin - Probability (Case of 3 Signals with sigma = 8dB, delta1 = 2dB)
Correlation Coefficient Determination
For further information about determination of the correlation coefficient, please see "Correlation Coefficient Determina-
tion" on page 123.
4.7.2.2.2 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.
Therefore, we have:
DL nsignals
G macro – diversity = M Shadowing – Ec  Io – M Shadowing – Ec  Io
Where n is the number of available signals.

Note:
• Atoll uses the DL macro-diversity gain to calculate Ec/Io. You can force Atoll not to take it
into account through the Atoll.ini file (see Atoll administration files). You must create this file
and place it in the Atoll installation directory.
© Forsk 2009 AT271_TRG_E6 123

4.8 Appendices
4.8.1 Transmitter Radio Equipment
Radio equipment such as TMA, feeder and BTS, 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,
CDMA2000 1xRTT 1xEV-DO, IS-95 cdmaOne, TD-SCDMA, WiMAX 802.16d, and WiMAX 802.16e documents,
• Transmitter total losses  L Total  in GSM GPRS EGPRS documents.
In Atoll, the transmitter-equipment pair is modelled a single entity. The entry to the BTS is considered the reference point
which is the location of the transmission/reception parameters.
Figure 4.28: Reference Point - Location of the Transmission/Reception parameters
4.8.1.1 UMTS HSPA, CDMA2000 1xRTT 1xEV-DO, IS-95 cdmaOne, and

TD-SCDMA Documents
As the reference point is the BTS entry, the transmitter noise figure corresponds to the BTS noise figure. Therefore, we
have:
NF TX = NF BTS
where NF BTS is the BTS noise figure.
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 (in dB)
where,
UL
L Misc are the miscellaneous reception losses (Transmitter property),
UL UL UL UL UL
L Feeder are the feeder reception losses ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and
UL
L Connector are respectively the feeder loss per metre (Feeder property), the reception feeder length in metre (Transmitter
property) and the connector reception losses,
UL
L BTS – Conf are the losses due to BTS configuration (BTS property),
UL
G Ant – div is the antenna diversity gain (Transmitter property),
NR Repeaters is the noise rise at transmitter due to repeaters. This parameter is taken into account only if the transmitter
has active repeater(s),
G TMA is the gain due to TMA.
The noise rise at transmitter due to repeaters is calculated as follows:
For each active repeater ( k ), Atoll calculates a noise injection margin ( NIM Rp ). This is the difference between the donor
k
transmitter noise figure ( NF TX ) and the repeater noise figure received at the donor.
Rp k TX – Rp k
NIM Rp = NF TX –  NF Rp + G amp – L  (in dB)
r  k 
where,
NF Rp is the repeater noise figure,

k
Rp k
G amp is the repeater amplification gain (repeater property),
124 AT271_TRG_E6 © Forsk 2009

TX – R p k
L are the losses between the donor transmitter and the repeater (repeater property).
For each active repeater ( k ), Atoll converts the noise injection margin ( NIM Rp ) to Watt. Then, it uses the values to calcu-
k
late the noise rise at the donor transmitter due to active repeaters ( NR Repeaters ).
 1 
NR Repeaters = 10  Log  1 +
  ------------------
NIM Rp 
r
- (in dB)
r
The gain due to TMA is calculated as follows:

WithoutTMA WithTMA
G TMA = NF Composite – NF Composite (in dB)
where,
WithTMA WithoutTMA
NF Composite and NF Composite are the composite noise figures with and without TMA respectively.
Friis' equation is used to calculate the composite noise figure when there is a TMA.
WithTMA  NF Feeder – 1   NF BTS – 1 

NF Composite = NF TMA + -------------------------------------- - (not in dB)
+ ---------------------------------------
UL UL UL
G TMA G TMA  G Feeder
And,
WithoutTMA
NF Composite = NF BTS + NF Feeder (in dB)
where,
NF Feeder is the feeder noise figure,
NF TMA is the TMA noise figure,
NF BTS is the BTS noise figure,
UL
G TMA is the TMA reception gain,
UL UL UL
G Feeder is the feeder UL gain; G Feeder = – L Feeder .
UL UL UL UL UL UL
L Feeder is the feeder reception loss ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and L Connector
are respectively the feeder loss per metre, the reception feeder length in metre and the connector reception loss),
Notes:
• According to the book “Radio network planning and optimisation for UMTS” by Laiho J.,
Wacker A., Novosad T., the noise figure corresponds to the loss in case of passive
components. Therefore, feeder noise figure is equal to the cable uplink losses.
UL
NF Feeder = L Feeder (in dB)
• Loss and gain inputs specified in .atl documents must be positive values.
Atoll calculates total DL losses as follows.

DL DL DL DL
L Total – DL = L TMA + L Feeder + L Misc + L BTS – Conf (in dB)
where,
DL
L TMA is the TMA transmission loss,
DL DL DL DL DL
L Feeder is the feeder transmission loss ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and
DL
L Connector are respectively the feeder loss per metre, the transmission feeder length in metre and the connector trans-
mission losses),
DL
L Misc are the miscellaneous transmission losses,
DL
L BTS – Conf are the losses due to BTS configuration (BTS property).
4.8.1.2 GSM GPRS EGPRS Documents

Atoll calculates DL total losses as follows:
DL DL DL DL
L Total – DL = L TMA + L Feeder + L Misc + L BTS – Conf (in dB)
where,
© Forsk 2009 AT271_TRG_E6 125

DL
DL DL DL DL DL
DL
mission loss),
DL
DL
4.8.1.3 WiMAX 802.16d and WiMAX 802.16e Documents

As the reference point is the BTS entry, the transmitter noise figure corresponds to the BTS noise figure. Therefore, we
have:
NF TX = NF BTS
where NF BTS is the BTS noise figure.
Atoll calculates total UL losses as follows:

UL UL UL UL
L Total – UL = L Misc + L Feeder + L BTS – Conf – G Ant – div – G TMA (in dB)
where,
UL
L Misc are the miscellaneous reception losses (Transmitter property),
UL UL UL UL UL
L Feeder are the feeder reception losses ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and
UL
L Connector are respectively the feeder loss per metre (Feeder property), the reception feeder length in metre (Transmitter
property) and the connector reception losses,
UL
L BTS – Conf are the losses due to BTS configuration (BTS property),
UL
G Ant – div is the antenna diversity gain (Transmitter property),
G TMA is the gain due to TMA, which is calculated as follows:
WithoutTMA WithTMA
G TMA = NF Composite – NF Composite (in dB)
where,
WithTMA WithoutTMA
NF Composite and NF Composite are the composite noise figures with and without TMA respectively.
Friis' equation is used to calculate the composite noise figure when there is a TMA.
WithTMA  NF Feeder – 1   NF BTS – 1 

NF Composite = NFTMA + -------------------------------------- - (not in dB)
+ ---------------------------------------
UL UL UL
G TMA G TMA  G Feeder
WithoutTMA
And NF Composite = NF BTS + NF Feeder (in dB)
where,
NF Feeder is the feeder noise figure,
NF TMA is the TMA noise figure,
NF BTS is the BTS noise figure,
UL
G TMA is the TMA reception gain,
UL UL UL
G Feeder is the feeder UL gain; G Feeder = – L Feeder .
UL UL UL UL UL UL
L Feeder is the feeder reception loss ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and L Connector
are respectively the feeder loss per metre, the reception feeder length in metre and the connector reception loss),
Notes:
• According to the book “Radio network planning and optimisation for UMTS” by Laiho J.,
Wacker A., Novosad T., the noise figure corresponds to the loss in case of passive
components. Therefore, feeder noise figure is equal to the cable uplink losses.
UL
NF Feeder = L Feeder (in dB)
126 AT271_TRG_E6 © Forsk 2009

• Loss and gain inputs specified in .atl documents must be positive values.
Atoll calculates total DL losses as follows.

DL DL DL DL
L total – DL = L TMA + L Feeder + L Misc + L BTS – Conf (in dB)
where,
DL
DL DL DL DL DL
DL
mission losses),
DL
DL
4.8.2 Secondary Antennas

When secondary antennas are installed on a transmitter, the signal level received from it is calculated as follows:
 P  1 –  G ant – mTx G ant – i 

 Tx 
   X i  -----------------------
 L Tx
-
P Tx  X i  ---------------------
L Tx 
Tx

i
 ----------------------------------------------------------------------- + ----------------------------------------------
 L ant – m  az m el m  L ant – i  az i el i  
Tx Tx
 i 
 
P rec = ------------------------------------------------------------------------------------------------------------------------------------------ (not in dB2)
L model
Where,
PTx is the transmitter power (Ppilot in UMTS, CDMA2000 and IS95-CDMA documents),
i is the secondary antenna index,

xi is the percentage of power dedicated to the secondary antenna, i,
G ant – m is the gain of the main antenna installed on the transmitter,

Tx
LTx are transmitter losses (LTx=Ltotal-DL),
G ant – i is the gain of the secondary antenna, i, installed on the transmitter,

Tx
Lmodel is the path loss calculated by the propagation model,
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, i.

Tx
The definition of angles, az and el, depends on the used calculation method.
• Method 1 (must be indicated in an Atoll.ini file):
- azm is the difference between the receiver antenna azimuth and azimuth of the transmitter main antenna,
- elm is the difference between the receiver antenna tilt and tilt of the transmitter main antenna,
- azi is the difference between the receiver antenna azimuth and azimuth of the transmitter secondary antenna,
i,
- eli is the difference between the receiver antenna tilt and tilt of the transmitter secondary antenna, i,
• Method 2 (default):
- azm is the receiver azimuth in the coordinate system of the transmitter main antenna,
- elm is the receiver tilt in the coordinate system of the transmitter main antenna,
- azi is the receiver azimuth in the coordinate system of the transmitter secondary antenna, i,
- eli is the receiver tilt in the coordinate system of the transmitter secondary antenna, i,
2. Formula cannot be directly calculated from components stated in dB and must be converted in linear values.
© Forsk 2009 AT271_TRG_E6 127

128 AT271_TRG_E6 © Forsk 2009

Chapter 5
GSM GPRS EGPRS Networks
This chapter provides descriptions of all the algorithms for calculations, analyses, automatic allocations and
prediction studies available in GSM GPRS EGPRS projects.
Atoll
130 AT271_TRG_E6 © Forsk 2009

Chapter 5: GSM GPRS EGPRS Networks
5 GSM GPRS EGPRS Networks

5.1 General Prediction Studies
5.1.1 Calculation Criteria
Three criteria can be studied in point analysis (Profile tab) and in general coverage studies. Study criteria are detailed in
the table below.
Study criteria Formulas

Signal level received from a transmitter on a TRX type
Txi
Signal level ( P rec ) Txi Txi
P rec  tt  = EIRP  tt  – P  tt  – L path – M Shadowing – model – L Indoor +  G ant – L Rx
Rx
Txi Txi
Path loss ( L path ) L path = L model + L ant
Tx
Txi Txi Txi

Total losses ( L total ) L total =  L path + M Shadowing – model + L Indoor + L Tx + L Rx  –  G ant + G ant 
Tx Rx
where,
EIRP is the effective isotropic radiated power of the transmitter,
L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model,
L ant is the transmitter antenna attenuation (from antenna patterns),

Tx
M Shadowing – model is the shadowing margin. This parameter is taken into account when the option “Shadowing taken into
account” is selected,
L Indoor are the indoor losses. These losses are defined for each clutter class. They are taken into account when the option
L Rx are the receiver losses,
G ant is the receiver antenna gain,

Rx
P is the power offset defined for the selected TRX type in the transmitter property dialog,
tt is the TRX type (in the GSM GPRS EGPRS.mdb document template, there are three possible TRX types, BCCH, TCH
and inner TCH).
5.1.2 Point Analysis

5.1.2.1 Profile Tab
Txi
Atoll displays the signal level received from the selected transmitter on a TRX type ( P rec  tt  ).
Notes:
• If power offsets of subcells are identical, field level received from a selected transmitter will
be the same for all the studied TRX types.
Txi
• For a selected transmitter, it is also possible to study the path loss, L path , or the total
Txi
losses, L total . Path loss and total losses are the same on any TRX type.
5.1.2.2 Reception Tab

Analysis provided in the Reception tab is based on path loss matrices. So, you can study reception from TBC transmitters
for which path loss matrices have been computed on their calculation areas.
Txi
For each transmitter, Atoll displays the signal level received on a TRX type, ( P rec  tt  ).
Reception bars are displayed in a decreasing signal level order. The maximum number of reception bars depends on the
signal level received from the best server. Only reception bars of transmitters whose signal level is within a 30 dB margin
from the best server can be displayed.
© Forsk 2009 AT271_TRG_E6 131

Notes:
• If power offsets of subcells are identical, field level received from a given transmitter will be
the same whichever the studied TRX type.
Txi Txi
• It is also possible to study the path loss, L path , or the total losses, L total of each
transmitter. Path loss and total losses are the same on any TRX type.
• You can use a value other than 30 dB for the margin from the best server signal level, for
example a smaller value for improving the calculation speed. For more information on
defining a different value for this margin, see the Administrator Manual.
5.1.3 Coverage Studies

For each TBC transmitter, Txi, Atoll determines the selected criterion on each bin inside the Txi calculation area. In fact,
each bin within the Txi calculation area is considered as a potential (fixed or mobile) receiver.
Coverage study parameters to be set are:
• The study conditions in order to determine the service area of each TBC transmitter,
• The display settings to select how to colour service areas.
5.1.3.1 Service Area Determination

Atoll uses parameters entered in the Condition tab of the coverage study property dialog to predetermine areas where it
will display coverage.
We can distinguish seven cases as below. Let us assume that:
• Each transmitter, Txi, belongs to a Hierarchical Cell Structure (HCS) layer, k, with a defined priority and a defined
reception threshold.
• The maximum range option (available in the System tab of the Predictions property dialog) is inactive.
5.1.3.1.1 All Servers

For each HCS layer, k, the service area of Txi corresponds to the bins where:
Txi Txi Txi
MinimumThreshold  P rec  tt   or L tot orTotal – Losses   MaximumThreshold
Note:
• The minimum threshold is either globally defined or specifically for each subcell (subcell
reception threshold)
5.1.3.1.2 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
P rec  ic   Best  P rec  ic   – M

Txi Txj
ji
M is the specified margin (dB).

Best function: considers the highest value.
Notes:
• If the margin equals 0 dB, Atoll will consider bins where the signal level received from Txi is
the highest.
• If the margin is set to 2 dB, Atoll will consider bins where the signal level received from Txi
is either the highest or 2dB lower than the highest.
• 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, which are 2nd best servers.
5.1.3.1.3 Second Best Signal Level and a Margin

Txi Txi Txi
And
Txi nd
P rec  ic   2 Best  P Txj  ic   – M
rec
ji
132 AT271_TRG_E6 © Forsk 2009

2nd Best function: considers the second highest value.

Notes:
the second highest.
is either the second highest or 2dB lower than the second highest.
is 2dB higher than the signal levels from transmitters, which are 3rd best servers.
5.1.3.1.4 Best Signal Level per HCS Layer and a Margin

Txi Txi Txi
And
P rec  BCCH   Best  P rec  BCCH   – M

Txi Txj
ji

Notes:
the highest.
is 2dB higher than the signal levels from transmitters that are the 2nd best servers.
5.1.3.1.5 HCS Servers and a Margin

Txi Txi Txi
And

Txi Txj
ji
Txi
The received P rec  tt  exceeds the reception threshold defined per HCS layer

Notes:
the highest.
5.1.3.1.6 Highest Priority HCS Server and a Margin

In this case, the service area of Txi corresponds to the bins where:
Txi Txi Txi
And

Txi Txj
ji
And
© Forsk 2009 AT271_TRG_E6 133

Txi belongs to the HCS layer with the highest priority. The highest priority is defined by the priority field (0: lowest) assum-
Txi
ing the received P rec  tt  exceeds the reception threshold defined per HCS layer.

Notes:
the highest.
• In the case two layers have the same priority, the traffic is served by the transmitter for
which the difference between the received signal strength and the HCS threshold is the
highest. The way the competition is managed between layers with the same priority can be
modified. For more information, see the Administrator Manual.
5.1.3.1.7 Second Best Signal Level per HCS Layer and a Margin
Txi Txi Txi
And
Txi nd
P rec  BCCH   2 Best  P Txj  BCCH   – M
rec
ji

Notes:
the second highest.
is 2dB higher than the signal levels from transmitters that are the 3rd best servers.
5.1.3.1.8 Best Idle Mode Reselection Criterion (C2)

Such type of coverage would is useful :
• To compare Idle and Dedicated mode best servers for Voice traffic
• Display the GPRS/EGPRS best server map (based on GSM idle mode)
The path loss criterion parameter C1 used for cell selection and reselection is defined by :
Txi
C1 = P rec  ic  – MinimumThreshold
The path loss criterion (GSM03.22) is satisfied if C1  0 .

The reselection criterion C2 is used for cell reselection only and is defined by :
C2 = C1 + CELL_RESELECT_OFFSET
where CELL_RESELECT_OFFSET is the reselection value (in dB) defined for at the transmitter level.
Txi Txi Txi
And
 BCCH  = Best  C2  BCCH  

Txi Txj
C2
j

On each bin, the best C2 value is kept. It corresponds to the best server in Idle Mode. Since the C2 value is an integer
value, so must be rounded.
134 AT271_TRG_E6 © Forsk 2009

5.1.3.2 Coverage Display

5.1.3.2.1 Plot Resolution
Prediction plot resolution is independent of the matrix resolutions and can be defined on a per study basis. Prediction plots
are generated from multi-resolution path loss matrices using bilinear interpolation method (similar to the one used to eval-
uate site altitude).
5.1.3.2.2 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, dBµV, dBµV/m)
Atoll calculates signal level received from the transmitter on each bin of each transmitter service area. A bin of a service
area is coloured if the signal level exceeds (  ) the defined minimum thresholds (bin colour depends on signal level).
Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as transmitter service areas. Each layer shows the different signal levels available in the transmitter service area.
Best Signal Level (in dBm, dBµV, dBµV/m)
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. When other service-
When other service areas overlap the studied one, Atoll chooses the highest value. A bin of a service area is coloured if
the signal level exceeds (  ) the defined thresholds (the bin colour depends on the signal level). Coverage consists of
several independent layers whose visibility in the workspace can be managed. There are as many layers as defined
thresholds. Each layer corresponds to an area where the signal level from the best server exceeds a defined minimum
threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each bin of each transmitter service area. A bin of a service area is
coloured if path loss exceeds (  ) the defined minimum thresholds (bin colour depends on path loss). Coverage consists
of several independent layers whose visibility in the workspace can be managed. There are as many layers as service
areas. Each layer shows the different path loss levels in the transmitter service area.
Total Losses (dB)
Atoll calculates total losses from the transmitter on each bin of each transmitter service area. A bin of a service area is
coloured if total losses exceed (  ) the defined minimum thresholds (bin 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
service areas. Each layer shows the different total losses levels in the transmitter service area.
Best Server Path Loss (dB)
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. When other service
areas overlap the studied one, Atoll determines the best transmitter and evaluates path loss from the best transmitter. A
bin of a service area is coloured if the path loss exceeds (  ) the defined thresholds (bin colour depends on path loss).
layers as defined thresholds. Each layer corresponds to an area where the path loss from the best server exceeds a
defined minimum threshold.
Best Server Total Losses (dB)
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where service areas
overlap the studied one, Atoll determines the best transmitter and evaluates total losses from the best transmitter. A bin
of a service area is coloured if the total losses exceed (  ) the defined thresholds (bin colour depends on total losses).
layers as defined thresholds. Each layer corresponds to an area where the total losses from the best server exceed a
Number of Servers
Atoll evaluates how many service areas cover a bin in order to determine the number of servers. The bin colour depends
on the number of servers. Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the number of serv-
ers exceeds (  ) a defined minimum threshold.
Cell Edge Coverage Probability (%)
On each bin of each transmitter service area, the coverage corresponds to the pixels where the signal level from this trans-
mitter fulfils signal conditions defined in Conditions tab with different cell edge coverage probabilities. There is one cover-
age area per transmitter in the explorer.
© Forsk 2009 AT271_TRG_E6 135

Best Cell Edge Coverage Probability (%)
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. There is one coverage area per cell edge coverage probability in the
explorer.
Best C2 (dBm)
Atoll calculates C2 values received from transmitters on each bin of each transmitter service area. When other service
areas overlap the studied one, Atoll chooses the highest value. A bin of a service area is coloured if the C2 value exceeds
(  ) the defined thresholds (the bin colour depends on the C2 value). Coverage consists of several independent layers
whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer corre-
sponds to an area where the best C2 value exceeds a defined minimum threshold.
5.2 Traffic Analysis

When starting a traffic analysis, Atoll distributes the traffic from maps to transmitters of each layer according to the compat-
ibility criteria defined in the transmitter, services, mobility type, terminal type properties. Transmitters considered in traffic
analysis are the active and filtered transmitters that belong to the focus zone.
Notes:
• If no focus zone exists in the .atl document, Atoll takes into account the computation zone.
• For details of the average timeslot capacity calculation, see the Network Dimensioning
section (calculation of minimum reduction factor).
5.2.1 Traffic Distribution

5.2.1.1 Normal Cells (Nonconcentric, No HCS Layer)
5.2.1.1.1 Circuit Switched Services
A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the BCCH and
TCH subcells of a transmitter if:
• The terminal, t, works on the frequency band used by the BCCH subcell,
• The terminal, t, works on the frequency band used by the TCH subcell.
5.2.1.1.2 Packet Switched Services

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the BCCH and
TCH subcells of a transmitter if:
• The transmitter is an GPRS/EGPRS station (option specified in the transmitter property dialog),
• The terminal, t, is technologically compatible with the transmitter,
• The terminal, t, works on the frequency band used by the TCH subcell.
5.2.1.2 Concentric Cells

In case of concentric cells, TCH_INNER TRX type has the highest priority to carry traffic.

A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the TCH_INNER,
BCCH and TCH subcells of a transmitter if:
• The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells.

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the TCH_INNER,
BCCH and TCH subcells of a transmitter if:
• The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells.
5.2.1.3 HCS Layers

For each HCS layer, k, you may specify the maximum mobile speed supported by the transmitters of the layer.
136 AT271_TRG_E6 © Forsk 2009


A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the BCCH and
TCH subcells (and TCH_INNER in case of concentric cells) of a transmitter if:
• The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells,
• The user’s mobility, m, is less than the maximum speed supported by the layer, k.

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the BCCH and
TCH subcells (and TCH_INNER in case of concentric cells) of a transmitter if:
• The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells,
• The user mobility, m, is less than the maximum speed supported by the layer, k.
5.2.2 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, serv-
ices, mobility, terminal properties as explained above.
• Atoll distributes traffic on subcell service areas, which are determined using the option “Best signal level per HCS
layer” with a 0dB margin and the subcell reception threshold as lower threshold.
• Same traffic is distributed to the BCCH and TCH subcells.
5.2.2.1 Traffic Maps Based on Environments and User Profiles

5.2.2.1.1 Normal Cells (Nonconcentric, No HCS Layer)
Number of subscribers ( X up m ) for each TCH subcell (Txi, TCH), per user profile up with a given mobility m, is inferred as:
X up m  Txi TCH  = S up m  Txi TCH   D
Where Sup,m is the TCH service area containing the user profile up with the mobility m and D is the user profile density.
For each behaviour described in the user profile up, Atoll calculates the probability for the user to be connected with a
given service using a terminal t.
Circuit Switched Services
For a circuit switched service c, we have:
N call  d
p up  c t  = ---------------------
-
3600
Where Ncall is the number of calls per hour and d is the average call duration (in seconds).
Then, Atoll evaluates the traffic demand, D up  c t  m , in Erlangs for the subcell (Txi, TCH) service area.
D up  c t  m  Txi TCH  = X up m  Txi TCH   p up  c t 
Packet Switched Services
For a packet switched service p, we have:
N call  V  8
p up  p t  = -------------------------------
-
3600
Where Ncall is the number of calls per hour and V is the transmitted data volume per call (in Kbytes).
Then, Atoll evaluates the traffic demand, D up  p t  m , in kbits/s for the subcell (Txi, TCH) service area.
D up  p t  m  Txi TCH  = X up m  Txi TCH   p up  p t 
5.2.2.1.2 Concentric Cells

In case of concentric cells, 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. The traffic spread over the
TCH_INNER subcell may overflow to the TCH subcell. In this case, the traffic demand is the same on the TCH_INNER
subcell but increases on the TCH subcell.
© Forsk 2009 AT271_TRG_E6 137

Note:
• Traffic overflowing from the TCH_INNER to the TCH is not uniformly spread over the TCH
service area. It is still located on the TCH_INNER service area.
Number of subscribers ( X up m ) for each TCH_INNER (Txi, TCH_INNER) and TCH (Txi, TCH) subcell, per user profile up
with a given mobility m, is inferred as:
X up m  Txi,TCH_INNER  = S up m  Txi,TCH_INNER   D
X up m  Txi,TCH  =  S up m  Txi,TCH  – S up m  Txi,TCH_INNER    D
S up m  Txi,TCH_INNER  and S up m  Txi,TCH  respectively refer to the TCH_INNER and TCH subcell service areas
containing the user profile up with the mobility m. D is the user profile density.
Figure 5.1: Representation of a Concentric Cell TXi
Circuit Switched Services
For each user of the user profile up using a circuit switched service c with a terminal t, Atoll calculates the probability
( p up  c t  ) of the user being connected. Calculations are detailed in "Circuit Switched Services" on page 136.
Then, Atoll evaluates the traffic demand, D up  c t  m , in Erlangs in the (Txi, TCH_INNER) and (Txi, TCH) subcell service
areas.
D up  c t  m  Txi,TCH_INNER  = X up m  Txi,TCH_INNER   p up  c t 
D up  c t  m  Txi,TCH  = X up m  Txi,TCH   p up  c t  + D up  c t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
Where O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell.
Packet Switched Services
For each user of the user profile up using a packet switched service p with a terminal t, probability of the user being
connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 136.
Atoll evaluates the traffic demand, D up  p t  m , in kbits/s in the (Txi, TCH_INNER) and (Txi, TCH) subcell service areas.
D up  p t  m  Txi,TCH_INNER  = X up m  Txi,TCH_INNER   p up  p t 
D up  p t  m  Txi,TCH  = X up m  Txi,TCH   p up  p t  + D up  p t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
Where O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell.
5.2.2.1.3 HCS Layers

We assume two HCS layers: the micro layer has a higher priority than the macro layer. Txi belongs to the micro layer and
Txj to the macro.
Normal Cells
Atoll distributes traffic on the TCH service areas. 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. 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). Traffic on the overlapping
area is distributed to the TCH subcell of the micro layer because it has a higher priority. On this area, traffic of the micro
layer may overflow to the macro layer. In this case, the traffic demand is the same on the TCH subcell of the micro layer
but increases on the TCH subcell of the macro layer.
Note:
• Traffic overflowing to the macro layer is not uniformly spread over the TCH service area of
Txj. It is only located on the overlapping area.
138 AT271_TRG_E6 © Forsk 2009

Figure 5.2: Representation of Micro and Macro Layers
Atoll evaluates the traffic demand on the micro layer (higher priority) as explained above. For further details, please refer
to formulas for normal cells. Then, it proceeds with the macro layer (lower priority).
macro
Number of subscribers ( X up m ) for each TCH subcell (Txj, TCH) of the macro layer, per user profile up with the mobility
m, is inferred as:
macro macro macro
X up m  Txj TCH  =  S up m  Txj TCH  – S up m – overlapping  Txj TCH    D
macro
Where 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.
For each user described in the user profile up with the circuit switched service c and the terminal t, the probability for the
user being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 136.
macro
Then, Atoll evaluates the traffic demand, D up  c t  m , in Erlangs in the subcell (Txj, TCH) service area.
macro
macro macro micro S upm – overlapping  Txj TCH 
D up  c t m  Txj TCH  = X up m Txj TCH   p up  c t  + D up  c t m Txi TCH   --------------------------------------------------------------------------
micro
-  Omax Txi TCH 
S up m  Txi TCH 
For each user described in the user profile up with the packet switched service p and the terminal t, probability for the user
to be connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 136.
macro
Then, Atoll evaluates the traffic demand, D up  p t  m , in kbits/s in the subcell (Txj, TCH) service area.
macro
macro macro micro S upm – overlapping  Txj TCH 
D up  p t m  Txj TCH  = X up m Txj TCH   p up  p t  + D up  p t m Txi TCH   --------------------------------------------------------------------------
micro
-  Omax Txi TCH 
S up m  Txi TCH 
Where 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.
Concentric Cells
Atoll evaluates the traffic demand on the micro layer (higher priority HCS layer) as explained above. For further details,
please refer to formulas given in case of concentric cells. Then, it proceeds with the macro layer (lower priority HCS layer).
The traffic capture is calculated with the option “Best signal level per HCS layer”. It means that there are overlapping areas
between HCS layers where traffic is spread according to the layer priority. On these areas, traffic of the higher priority layer
may overflow.
© Forsk 2009 AT271_TRG_E6 139

Figure 5.3: Concentric Cells
The TCH_INNER service area of the macro layer is overlapped by the micro layer. This area consists of two parts: an area
macro
overlapped by the TCH service area of the micro layer S overlapping –  Txi TCH   Txj,TCH_INNER  and another overlapped
macro
by the TCH_INNER service area of the micro layer S overlapping –  Txi,TCH_INNER   Txj,TCH_INNER  .
Let us consider three areas, S1, S2 and S3.
macro macro
S 1 = S up m  Txj,TCH_INNER  – S up m – overlapping –  Txi TCH   Txj,TCH_INNER 
macro
S 2 = S up m – overlapping –  Txi,TCH_INNER   Txj,TCH_INNER 
macro
S 3 = S up m – overlapping –  Txi TCH   Txj,TCH_INNER  – S 2
macro
Where S up m  Txj,TCH_INNER  is the TCH_INNER subcell service area of Txj containing the user profile up with the
mobility m. We only consider the overlapping areas containing the user profile up with the mobility m.
macro
On S1, 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  = S 1  D
Where D is the user profile density.

The traffic spread over the TCH_INNER service area of the micro layer may overflow on the TCH subcell. The traffic over-
flowing to the TCH subcell is located on the TCH_INNER service area. On S2, the TCH subcell traffic coming from the
TCH_INNER subcell traffic overflow may overflow proportional to R2.
S2
R 2 = -----------------------------------------------------------------
micro
S up m  Txi,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.
S3
R 3 = -------------------------------------------------------------------------------------------------------------------
micro micro
S up m  Txi,TCH  – S up m  Txi,TCH_INNER 
micro micro
Where S up m  Txi,TCH  and S up m  Txi,TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively
containing the user profile up with the mobility m.
For each user described in the user profile up with a circuit switched service c and a terminal t, the probability for the user
being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 136. Then, Atoll evaluates
macro
the traffic demand, D up  c t  m , in Erlangs in the subcell (Txj, TCH_INNER) service area.
macro
X up m  Txj,TCH_INNER   p up  c t  +
macro
D up  c t  m  Txj,TCH_INNER  = R  D micro
2 up  c t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi,TCH  +
micro
R 3  X up m  Txi TCH   p up  c t   O max  Txi TCH 
140 AT271_TRG_E6 © Forsk 2009

For each user described in the user profile up with a packet switched service p and a terminal t, probability for the user to
be connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 136.
macro
Then, Atoll evaluates the traffic demand, D up  p t  m , stated in kbits/s in the subcell (Txj, TCH_INNER) service area.
macro
X up m  Txj,TCH_INNER   p up  p t  +
macro
D up  p t  m  Txj,TCH_INNER  = R  D micro
2 up  p t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi,TCH  +
micro
R 3  X up m  Txi TCH   p up  p t   O max  Txi TCH 
Where O max  Txi TCH  and O max  Txi,TCH_INNER  are the maximum rates of traffic overflow (stated in %) specified
for the TCH and TCH_INNER subcells of Txi respectively.
The area of the TCH ring of the macro layer is overlapped by the micro layer. There are two parts: an area overlapped by
macro
the TCH service area of the micro layer S overlapping –  Txi TCH   Txj,TCH -- TCH_INNER  and another one by the
macro
TCH_INNER service area of the micro layer S overlapping –  Txi,TCH_INNER   Txj,TCH -- TCH_INNER  .
Let us consider three areas, S’1, S’2 and S’3.
macro macro macro

S' 1 = S up m  Txj,TCH  – S up m  Txj,TCH_INNER  – S up m – overlapping –  Txi TCH   Txj,TCH -- TCH_INNER 
macro
S' 2 = S up m – overlapping –  Txi,TCH_INNER   Txj,TCH -- TCH_INNER 
macro
S' 3 = S up m – overlapping –  Txi TCH   Txj,TCH -- TCH_INNER  – S' 2
macro macro
Where S up m  Txj,TCH  and S up m  Txj,TCH_INNER  are the TCH and TCH_INNER subcell service areas of Txj
respectively. We only consider the overlapping areas containing the user profile up with the mobility m.
macro
On 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  = S' 1  D
Where D is the user profile density.

The traffic spread over the TCH_INNER service area of the micro layer may overflow on the TCH subcell. The traffic over-
flowing on the TCH subcell is located on the TCH_INNER service area. On S’2, the TCH subcell traffic coming from the
TCH_INNER subcell traffic overflow may overflow proportionally to R’2.
S' 2
R' 2 = -----------------------------------------------------------------
micro
S up m  Txi,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.
S' 3
R' 3 = -------------------------------------------------------------------------------------------------------------------
micro micro
S up m  Txi,TCH  – S up m  Txi,TCH_INNER 
micro micro
Where S up m  Txi,TCH  and S up m  Txi,TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively
containing the user profile up with the mobility m.
For each user described in the user profile up with a circuit switched service c and a terminal t, the probability for the user
being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 136.
macro
Then, Atoll evaluates the traffic demand, D up  c t  m , in Erlangs in the subcell (Txj, TCH) service area.
macro
X up m  Txj TCH   p up  c t  +
macro
macro D up  c t  m  Txj,TCH_INNER   O max  Txj,TCH_INNER  +
D up  c t  m  Txj TCH  =
micro
R' 2  D up  c t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi,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, the probability for the user
being connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 136.
macro
Then, Atoll evaluates the traffic demand, D up  p t  m , in kbits/s in the subcell (Txj, TCH) service area.
© Forsk 2009 AT271_TRG_E6 141

macro
X up m  Txj TCH   p up  p t  +
macro
macro D up  p t  m  Txj,TCH_INNER   O max  Txj,TCH_INNER  +
D up  p t  m  Txj TCH  =
micro
R' 2  D up  p t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi,TCH  +
micro
R' 3  X up m  Txi TCH   p up  p t  m  O max  Txi TCH 
Where O max  Txi,TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi (micro
layer), O max  Txi,TCH_INNER  the maximum rate of traffic overflow indicated for the TCH_INNER subcell of Txi (macro
layer), O max  Txj,TCH_INNER  the maximum rate of traffic overflow indicated for the TCH_INNER subcell of Txj (macro
micro
layer) and 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 136).
5.2.2.2 Traffic Maps Based on Transmitters and Services

We assume that the traffic map is built from a coverage by transmitter prediction study calculated for the TCH subcells
with options:
• “HCS Servers” and no margin if the network only consists of normal cells and concentric cells,
• “Highest Priority HCS Server” and no margin in case of HCS layers.
When creating the traffic map, you have to specify the traffic demand per transmitter and per service (throughput for a
packet switched service and Erlangs for a circuit switched service) and the global distribution of terminals and mobility
types.
Let E c  Txi TCH  denote the Erlangs for the circuit switched service, c, on the TCH subcell of Txi.
Let T p  Txi TCH  denote the throughput of the packet switched service, p, on the TCH subcell of Txi.
We assume that 100% of users have the terminal, t, and the mobility type, m.
5.2.2.2.1 Normal Cells (Nonconcentric, No HCS Layer)

For each circuit switched service, c, Atoll evaluates the traffic demand, Dc,t,m, in Erlangs in the subcell (Txi, TCH) service
area.
D c t m  Txi TCH  = E c  Txi TCH 
For each packet switched service, p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell (Txi, TCH) service
area.
D p t m  Txi TCH  = T p  Txi TCH 
5.2.2.2.2 Concentric Cells

In case of concentric cells, 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 ring served by the TCH subcell only. The traffic spread over the
TCH_INNER subcell may overflow to the TCH subcell. In this case, the traffic demand is the same on the TCH_INNER
subcell and rises on the TCH subcell.
Note:
• Traffic overflowing from the TCH_INNER to the TCH is not uniformly spread over the TCH
service area. It is only located on the TCH_INNER service area.
For each circuit switched service, c, Atoll evaluates the traffic demand, Dc,t,m, in Erlangs in the subcell, (Txi, TCH_INNER)
and (Txi, TCH), service areas.
S  Txi,TCH_INNER 
D c t m  Txi,TCH_INNER  = -----------------------------------------------------  E c  Txi TCH 
S  Txi TCH 
and
 S  Txi,TCH  – S  Txi,TCH_INNER  
-----------------------------------------------------------------------------------------------  E c  Txi TCH  +
D c t m  Txi,TCH  = S  Txi TCH 
D c t m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
For each packet switched service, p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell, (Txi, TCH_INNER)
and (Txi, TCH), service areas.
D p t m  Txi,TCH_INNER  = -----------------------------------------------------  T p  Txi TCH 
and
142 AT271_TRG_E6 © Forsk 2009

 S  Txi,TCH  – S  Txi,TCH_INNER  
-----------------------------------------------------------------------------------------------  T p  Txi TCH  +
D p t m  Txi,TCH  = S  Txi TCH 
D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
Where O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell,
S  Txi,TCH  and S  Txi,TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively.
5.2.2.2.3 HCS Layers

We assume we have two HCS layers: the micro layer has a higher priority and the macro layer has a lower one. Txi belongs
to the micro layer and Txj to the macro one.
Normal Cells
Atoll distributes traffic on the TCH service areas. The traffic capture is calculated with the option “HCS Servers”. It means
macro
that there is an overlapping area between HCS layers. Let S overlapping  Txj TCH  denote the TCH service area of the
macro layer overlapped by the TCH service area of the micro layer. Traffic on the overlapping area is distributed to the
TCH subcell of the micro layer (higher priority layer). On this area, traffic of the micro layer may overflow to the macro layer.
In this case, the traffic demand is the same on the TCH subcell of the micro layer but rises on the TCH subcell of the macro
layer.
Note:
• Traffic overflowing on the macro layer is not uniformly spread over the TCH service area of
Txj. It is only located on the overlapping area.
Atoll starts evaluating the traffic demand on the micro layer (highest priority HCS layer).
micro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txi, TCH) service
area.
micro
D c t m  Txi TCH  = E c  Txi TCH 
micro
For each packet switched service, p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txi, TCH) service
area.
micro
D p t m  Txi TCH  = T p  Txi TCH 
Then, Atoll proceeds with the macro layer (lower priority HCS layer). For each circuit switched service, c, Atoll calculates
macro
the traffic demand, D c t m , in Erlangs in the subcell (Txj, TCH) service area.
macro
macro micro S overlapping  Txj TCH 
D c t m  Txj TCH  = E c  Txj TCH  + D c t m  Txi TCH   ------------------------------------------------------------
micro
-  O max  Txi TCH 
macro
For each packet switched service, p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj, TCH) service
area.
macro
macro micro S overlapping  Txj TCH 
D p t m  Txj TCH  = T p  Txj TCH  + D p t m  Txi TCH   ------------------------------------------------------------
micro
-  O max  Txi TCH 
Where O max  Txi TCH  is the maximum rate of traffic overflow (in %) specified for the TCH subcell of Txi (micro cell) and
micro
S  Txi TCH  the TCH service area of Txi.
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).
The traffic capture is calculated with the option “HCS Servers”. It means that there is overlapping areas between HCS
layers where traffic is spread over according to the layer priority. On these areas, traffic of the higher priority layer may
overflow.
© Forsk 2009 AT271_TRG_E6 143

Figure 5.4: Concentric Cells
The TCH_INNER service area of the macro layer is overlapped by the micro layer. This area consists of two parts: an area
macro
overlapped by the TCH service area of the micro layer S overlapping –  Txi TCH   Txj,TCH_INNER  and another overlapped
macro
by the TCH_INNER service area of the micro layer S overlapping –  Txi,TCH_INNER   Txj,TCH_INNER  .
Let us consider three areas, S1, S2 and S3.
macro macro
S1 = S  Txj,TCH_INNER  – S overlapping –  Txi TCH   Txj,TCH_INNER 
macro
S 2 = S overlapping –  Txi,TCH_INNER   Txj,TCH_INNER 
macro
S 3 = S overlapping –  Txi TCH   Txj,TCH_INNER  – S 2
macro
Where S  Txj,TCH_INNER  is the TCH_INNER subcell service area of Txj.
The traffic specified for Txj in the map description ( E c  Txj TCH  ) is spread over S1 proportionally to R1.
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”.
The traffic spread over the TCH_INNER service area of the micro layer may overflow to the TCH subcell. The traffic over-
flowing to the TCH subcell is located on the TCH_INNER service area. On S2, the TCH subcell traffic coming from the
TCH_INNER subcell traffic overflow may overflow proportional to R2.
S2
R 2 = -----------------------------------------------------------------
micro
The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S3 proportional to R3.
S3
R 3 = -------------------------------------------------------------------------------------------------------------------
micro micro
S  Txi,TCH  – S  Txi,TCH_INNER 
macro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txj,
TCH_INNER) service area.
R 1  E c  Txj TCH  +
micro
macro R 2  D c t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
D c t m  Txj,TCH_INNER  =
micro micro
S  Txi TCH  – S  Txi,TCH_INNER  
R 3  ---------------------------------------------------------------------------------------------------------------------------
micro
 E c  Txi TCH   O max  Txi TCH 
macro
For each packet switched service, p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj,
TCH_INNER) service area.
144 AT271_TRG_E6 © Forsk 2009

R 1  T p  Txj TCH  +
micro
macro R 2  D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
D p t m  Txj,TCH_INNER  =
micro micro
S  Txi TCH  – S  Txi,TCH_INNER  
R 3  ---------------------------------------------------------------------------------------------------------------------------
micro
 T p  Txi TCH   O max  Txi TCH 
Where O max  Txi TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi,
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi
micro
and S  Txi TCH  is the TCH subcell service area of Txi.
The area of the TCH ring of the macro layer is overlapped by the micro layer. There are two parts: an area overlapped by
macro
the TCH service area of the micro layer S overlapping –  Txi TCH   Txj,TCH -- TCH_INNER  and another overlapped by the
macro
TCH_INNER service area of the micro layer S overlapping –  Txi,TCH_INNER   Txj,TCH -- TCH_INNER  .
Let us consider three areas, S’1, S’2 and S’3.
macro macro macro

S' 1 = S  Txj TCH  – S  Txj,TCH_INNER  – S overlapping –  Txi TCH   Txj,TCH -- TCH_INNER 
macro
S' 2 = S overlapping –  Txi,TCH_INNER   Txj,TCH -- TCH_INNER 
macro
S' 3 = S overlapping –  Txi TCH   Txj,TCH -- TCH_INNER  – S' 2
macro macro
Where S  Txj TCH  and S  Txj,TCH_INNER  are the TCH and TCH_INNER subcell service areas of Txj
respectively.
The traffic specified for Txj in the map description ( E c  Txj TCH  ) is spread over S’1 proportional to R’1.
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”.
The traffic spread over the TCH_INNER service area of the micro layer may overflow to the TCH subcell. The traffic over-
flowing to the TCH subcell is located on the TCH_INNER service area. On S’2, the TCH subcell traffic coming from the
TCH_INNER subcell traffic overflow may overflow proportional to R’2.
S' 2
R' 2 = -----------------------------------------------------------------
micro
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.
S' 3
R' 3 = -------------------------------------------------------------------------------------------------------------------
micro micro
S  Txi,TCH  – S  Txi,TCH_INNER 
macro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txj, TCH) service
area.
R' 1  E c  Txj TCH  +

macro
D c t m  Txj,TCH_INNER   O max  Txj,TCH_INNER  +
macro
D c t m  Txj TCH  = micro
R' 2  D c t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
micro micro
S  Txi,TCH  – S  Txi,TCH_INNER  
R' 3  ------------------------------------------------------------------------------------------------------------------------
micro
 E c  Txi TCH   O max  Txi TCH 
S  Txi,TCH 
macro
For each packet switched service, p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj, TCH) service
area.
R' 1  T p  Txj TCH  +

macro
D c t m  Txj,TCH_INNER   O max  Txj,TCH_INNER  +
macro
D p t m  Txj TCH  = micro
R' 2  D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
micro micro
S  Txi,TCH  – S  Txi,TCH_INNER  
R' 3  ------------------------------------------------------------------------------------------------------------------------
micro
 T p  Txi TCH   O max  Txi TCH 
S  Txi,TCH 
© Forsk 2009 AT271_TRG_E6 145

Where O max  Txj,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell
of Txj, O max  Txi TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi,
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi,
micro micro
S  Txi,TCH  is the TCH subcell service area of Txi and S  Txi,TCH_INNER  is the TCH_INNER subcell service
area of Txi.
5.3 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.
5.3.1 Dimensioning Models and Quality Graphs

In Atoll, a dimensioning model is an entity utilized by the dimensioning engine along with other inputs (traffic, limitations,
criteria, etc.) in the process of dimensioning. A dimensioning model defines the QoS KPIs to be taken into account when
dimensioning a network for both circuit and packet switched traffic. 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. The dimensioning engine will only utilize the quality curves of the KPI selected. The KPIs not
selected are supposed to be either already satisfactory or not relatively important.
5.3.1.1 Circuit Switched Traffic

The network dimensioning for circuit switched traffic is performed using the universally accepted and adopted Erlang B
and Erlang C formulas. The dimensioning criterion in these formulas is the Grade of Service or the allowed blocking prob-
ability of the circuit switched traffic.
In the Erlang B approach, this Grade of Service is defined as the percentage of incoming circuit switched calls that are
blocked due to lack of resources or timeslots. This formula implies a loss system. The blocked calls are supposed to be
lost and the caller has to reinitiate it.
In the Erlang C approach, the Grade of Service is the percentage of incoming calls that are placed in a waiting queue when
there are no resources available, until some resources or timeslots are liberated. This queuing system has no lost calls.
As the load on the system increases, the average waiting time in the queue also increases.
These formulas and their details are available in many books. For example, Wireless Communications Principles and
Practice by Theodore S. Rappaport, Prentice Hall.
Following the common practice, network dimensioning in Atoll is based on the principle that a voice or GSM call has priority
over data transmission. Therefore, as explained later in the network dimensioning steps, 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.
5.3.1.2 Packet Switched Traffic

Since packet switched traffic does not occupy an entire timeslot the whole time, it is much more complicated to study than
circuit switched traffic. Packet traffic is intermittent and bursty. Whenever there is packet data to be transferred, a Tempo-
rary Block Flow (TBF) is initiated for transferring these packets. Multiple TBFs can be multiplexed on the same timeslot.
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.
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. 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.
5.3.1.2.1 Throughput
Throughput is defined as the amount of data delivered to the Logical Link Control Layer in a given unit of time. Each tempo-
rary block flow (TBF), and hence each user, has an associated measured throughput sample in a given network. Each
network will have a different throughput probability distribution depending on the load and network configuration. Instead
of using the precise probability distributions, it is more practical to compute the average and percentile throughput values.
In GPRS, the resources are shared between the users being served, and consequently, the throughput is reduced as the
number of active users increases. This reduction in user perceived throughput is modelled through a reduction factor. 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
146 AT271_TRG_E6 © Forsk 2009

Timeslot Capacity
The timeslot capacity is the average throughput per fully utilized timeslot. It represents the average throughput from the
network point of view. It mainly depends on the network’s propagation conditions and criteria in the coverage area of a
transmitter (carrier power, carrier-to-interference distribution, etc.). It is a measure of how much data the network is able
to transfer with 1 data Erlang, or in other words, how efficiently the hardware resources are being utilized by the network.
It may also depend on the RLC protocol efficiency.
Atoll computes the average timeslot capacity during the traffic analysis and is used to determine the minimum throughput
reduction factor. But since this information is displayed in the network dimensioning results (only due to relevance), this
information has been considered as a part of the network dimensioning process in this document.
Timeslot Utilisation
Timeslot utilization takes into account the average number of timeslots that are available for packet switched traffic. It is a
measure of how much the network is loaded with data services. Networks with timeslot utilisation close to 100% are close
to saturation and the end-user performance is likely to be very poor.
In Atoll this parameter is termed as the Load (Traffic load for circuit switched traffic and packet switched traffic load for
packet switched traffic). It is described in more detail in the Network dimensioning steps section.
Reduction Factor
Reduction factor takes into account the user throughput reduction due to timeslot sharing among many users. The figure
below shows how the peak throughput available per timeslot is reduced by interference and sharing.Reduction factor is a
function of the number of timeslots assigned to a user (Nu), number of timeslots available in the system (Ns) and the aver-
age system packet switched traffic load (Lp) (utilization of resources in the system). Data Erlangs or data traffic is given by:
Data Erlangs = L P  N S
Figure 5.5: Reduction of Throughput per Timeslot
More precisely, the reduction factor is a function of the ratio Ns/Nu (Np). Np models the equivalent timeslots that are avail-
able for the packet switched traffic in the system. For example, 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.
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).
Let X be a random variable that measures the reduction factor in a certain system state:
0 if n = 0
1 if 0 < n  N P
X
N
------P- if n > N P
n
Where n is the instantaneous number of connections in the system. The throughput reduction factor is defined as:

PX= n
RF   X  ----------------------
PX= 0
-
n=0
Or,

PX= n
RF =  X  ------------------------------

-
 PX= i
n=0
i=0
Here, P(X=n) is the probability function of having n connections in the system. Under the same assumptions as those of
the Erlang formulas, the probability function can be written as:
© Forsk 2009 AT271_TRG_E6 147

n
 LP  NP 
-------------------------
-
n!
P  X = n  = ----------------------------------------------------------------------------------------------
- if 0  n  N P
N P 
i i
 LP  NP   LP  NP 
 i!
-+
-----------------------  --------------------------------
N !  NP
 i – NP 
-
i=0 i = NP +1 P
n
 LP  NP 
--------------------------------
-
 i – NP 
N P!  N P
P  X = n  = ----------------------------------------------------------------------------------------------
N
- if n > N P
P 
i i
 LP  NP   LP  NP 
 i!
-+
-----------------------  --------------------------------
N !  NP
 i – NP 
-
i=0 i = NP +1 P
Hence the reduction factor can finally be written as:

NP 
i i
 LP  NP   LP  NP  NP
 -----------------------
-+  ---------------------------------   -------
i!  i – NP   i 
i=1 i = NP + 1 P
N !  NP
RF = ----------------------------------------------------------------------------------------------------------------
N
-
P 
i i
 LP  NP   LP  NP 
 i!
-+
-----------------------  --------------------------------
N !  N
 i – NP 
-
i=1 i = NP +1 P P
This formula is not directly applicable in any software application due to the summations up to infinity. Atoll uses the follow-
ing version of this formula that is exactly the same formula without the summation overflow problem.
NP  NP + 1  NP
 LP  NP 
n
NP  L 
n
 - – ---------------------   ln  1 – L P  +
-------------------------  -----P-
n! N P !  n
 
RF = n----------------------------------------------------------------------------------------------------------------------------------
=1
N
n=1 -
P
n NP
 LP  NP   LP  NP  LP
 n!
- + -----------------------------  ---------------
-------------------------
N P! 1 – LP
n=1
The default quality curves for the Reduction Factor have been derived using the above formula. 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). The figure below contains all the reduction factor quality curves in Atoll. The Maximum reduction
factor can be 1, implying a maximum throughput, and the minimum can be 0, implying a saturated system with no data
throughput.
Figure 5.6: Reduction Factor for Different Packet Switched Traffic Loads (Lp, X-axis)
Each curve in the above figure represents an equivalent number of packet switched timeslots, NP.
148 AT271_TRG_E6 © Forsk 2009

5.3.1.2.2 Delay
Delay is the time required for an LLC PDU to be completely transferred from the SGSN to the MS, or vice versa. Its model-
ling in an RF planning tool is a difficult task. Currently, study on this subject is underway at Forsk. Models for different traffic
types (HTTP, FTP, SMTP, SMS, etc.) are being studied to search for a possible analytical solution for this problem.
As the delay is a function of the delays and the losses incurred at the packet level, the network parameters, such as the
packet queue length, and different protocol properties, such as the size of the LLC PDU, become important. It is also quite
dependent upon the radio access round trip time (RA RTT) and has a considerable impact on the application level perform-
ance viewed by the user.
The delay parameter is a user level parameter rather than being a network level quantity, like throughput per cell, timeslot
capacity, TBF blocking and reduction factor, hence it is difficult to model and is currently under study. Hence, no default
curve is presently available for delay in Atoll.
5.3.1.2.3 Blocking Probability

In GPRS, there is no blocking as in circuit switched connections. If a new temporary block flow (TBF) establishment is
requested and there are already M users per timeslot, M being the maximum limit of multiplexing per timeslot (Multiplexing
factor), the request is queued in the system to be established later when resources become available.
Supposing that M number of users can be multiplexed over a single timeslot (PDCH), we can have a maximum of M * Np
users in the system. This implies that if a new TBF is requested when there are already M * Np users active, 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, meaning,
PX= n for n =  M  N P  + 1
as in this case n is always greater than Np, we have,
n
 LP  NP 
---------------------------------
 i – NP 
N P!  N P
P  X = n  = ----------------------------------------------------------------------------------------------
N
-
P 
i i
 LP  NP   LP  NP 
 ------------------------ +
i!  --------------------------------
N !  N
 i – NP 
-
i=0 i = NP +1 P P
So, the Blocking Probability can be given as:


n
 LP  NP 
  ----------------------------------
N !  NP
 1 – NP 
-
n = M  NP + 1 P
BP =  P  X = n  = ------------------------------------------------------------------------------------------------
N P
i

i
n = MN+1  LP  NP   LP  NP 
 i!
-+
-----------------------  ----------------------------------
N !  N
 1 – NP 
-
i=0 i = NP +1 P P
Eliminating the summations to infinity, the blocking probability can be stated in a simpler form:
M  NP
 LP  NP  LP
-  ---------------
--------------------------------------------
 M  N P – NP  1 – L
N P!  N P P
BP = ------------------------------------------------------------------------------------------
N
-
P
i NP
 LP  NP   LP  NP  LP
 ------------------------ + -----------------------------  ---------------
i! N P! 1 – LP
i=0
The above formula has been used to generate the default quality curves for blocking probability in Atoll.
These graphs are generated for a user multiplexing factor of 8 users per timeslot. Each curve represents an equivalent
number of packet switched timeslots, NP.
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. The figure below contains all the blocking probability curves for
packet switched traffic dimensioning in Atoll. The blocking probability increases with the packet switched traffic load, which
implies that as the packet switched traffic increases for a given number of timeslots, the system starts to get more and
more loaded, hence there is higher probability of having a temporary block flow placed in a waiting queue.
© Forsk 2009 AT271_TRG_E6 149

Figure 5.7: Blocking Probability for Different Packet Switched Traffic Loads (Lp, X-axis)
Reference:
T. Halonen, J. Romero, J. Melero; GSM, GPRS and EDGE performance – Evolution towards 3G/UMTS, John Wiley
and Sons Ltd.
5.3.2 Network Dimensioning Process

The network dimensioning process is described below in detail. As the whole dimensioning process is in fact a chain of
small processes that have there respective inputs and outputs, with outputs of a preceding one being the inputs to the
next, the best method is to detail each process individually in form of steps of the global dimensioning process.
5.3.2.1 Network Dimensioning Engine

During the dimensioning process, Atoll first computes the number of timeslots required to accommodate the circuit
switched traffic. Then it calculates the number of timeslots to add in order to satisfy the demand of packet switched traffic.
This is performed using the quality curves entered in the dimensioning model used. If the dimensioning model has been
indicated to take all three KPIs in to account (throughput reduction factor, delay and blocking probability), the number of
timeslots to be added is calculated such that:
1. The throughput reduction factor is greater than the minimum throughput reduction factor,
2. Delay is less than the maximum permissible delay defined in the service properties, and
3. The blocking probability is less than the maximum allowable blocking probability defined in the service properties.
The figure below depicts a simplified flowchart of the dimensioning engine in Atoll.
Figure 5.8: Network Dimensioning Process
On the whole, following are the inputs and outputs of the network dimensioning process:
5.3.2.1.1 Inputs
• Circuit switched traffic demand
• Packet switched traffic demand
150 AT271_TRG_E6 © Forsk 2009

• 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, maximum permissible delay, maximum allow-
able blocking probability etc.
• Dimensioning model parameters: Maximum number of TRXs per transmitter, dimensioning model for circuit
switched traffic, number of minimum dedicated packet switched timeslots per transmitter, maximum number of
TRXs added for packet switched services, KPIs to consider, and their quality curves.
5.3.2.1.2 Outputs
• Number of required TRXs per transmitter
• Number of required shared, 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, delay and blocking probability
5.3.2.2 Network Dimensioning Steps

This section describes the entire process step by step as it is actually performed in Atoll. Details of the calculations of the
parameters that are calculated during each step are described as well.
5.3.2.2.1 Step 1: Timeslots Required for CS Traffic

Atoll computes the number of timeslots required to accommodate the circuit switched traffic assigned to each subcell. Atoll
takes the circuit switched traffic demand (Erlangs), calculated in the traffic analysis and assigned to the current subcell,
and the maximum blocking probability defined for the circuit switched service, and computes the required number of times-
lots to satisfy this demand using the Erlang B or Erlang C formula (as defined by the user).
If the user-defined target rate of traffic overflow per subcell, OTarget, is greater than the maximum blocking rate defined in
the services properties, it is going to be taken as the Grade of Service required for that subcell instead of the maximum
blocking rate of the service.
For the blocking probability GoS and circuit switched traffic demand TDC, Atoll determines the required number of times-
lots TSreq. C for each subcell using formulas described below. In fact, Atoll searches for TSreq. C value until the defined
grade of service is reached.
For Erlang B, we have:
TS reqC
 TD C 
--------------------------------
 TS reqC !
GoS = ----------------------------------
TS
-
reqC
k
 TD C 
 ------------------
k!
-
k=0
For Erlang C, we have:

TS reqC
 TD C 
GoS = -----------------------------------------------------------------------------------------------------------------------------------------------------
TS –1
-
reqC
k
TS reqC TD C   TD C 
 TD C  +  TS reqC !   1 – ------------------
-   ------------------
-
 TS  k!
reqC
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 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.
If the number of timeslots required to accommodate the full-rate circuit switched traffic is TSreq. FR, and the percentage of
half-rate traffic within the subcell is defined by HR, then the effective number of equivalent full-rate circuit switched times-
lots TSeff. that can carry this traffic mix is calculated by:
HR
TS eff = TS reqFR   1 – ---------
 2 
Atoll employs this simplified approach to integrating half-rate circuit switched traffic, which provides approximately the
same results as obtained by using the half-rate traffic charts.
5.3.2.2.2 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. 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.
If a timeslot configuration defines a certain number of dedicated packet switched timeslots pre-allocated in certain TRXs,
those timeslots will not be considered capable of carrying circuit switched traffic and hence will not be allocated. For exam-
ple, if 4 timeslots have been marked as packet switched timeslots in the first TRX and Atoll computes 8 timeslots for carry-
© Forsk 2009 AT271_TRG_E6 151

ing 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.
The total numbers of timeslots that carry circuit switched and packet switched traffic respectively are the sums of respec-
tive dedicated and shared timeslots:
TS P = TS S + TS P dedicated and TS C = TS S + TS C dedicated
5.3.2.2.3 Step 3: Effective CS Blocking, Effective CS Traffic Overflow and Served CS Traffic
In this step, the previously calculated number of required TRXs is used to compute the effective blocking rate for the circuit
switched traffic. 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). It then calculates the
effective traffic overflow rate, Oeff..
In case of Erlang B formula, the effective rate of traffic overflow for the circuit switched traffic is the same as the circuit
switched blocking rate. While in case of the Erlang C model, the circuit switched traffic is supposed to be placed in an
infinite-length waiting queue. This implies that there is no overflow in this case.
From this data, it also computes the served circuit switched traffic. 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. Hence, for an
effective traffic overflow rate of Oeff. and the circuit switched traffic demand of TDC, the served circuit switched traffic STC
is computed as:
ST C = TD C   1 – O eff 
5.3.2.2.4 Step 4: TRXs to Add for PS Traffic

This step is the core of the dimensioning process for packet switched services. First of all, Atoll computes the number of
TRXs to be added to carry the packet switched traffic demand. This is the number of TRXs that contain dedicated packet
switched and shared timeslots.
To determine this number of TRXs, Atoll calculates the equivalent average packet switched traffic demand in timeslots by
studying each pixel covered by the transmitter. This calculation is in fact performed in the traffic analysis process. Knowing
the traffic demand per pixel of the covered area in terms of kbps and the maximum attainable throughput per pixel (accord-
ing to the C and/or C/I conditions and the coding scheme curves in the GPRS/EDGE equipment), Atoll calculates the aver-
age traffic demand in packet switched timeslots by:
Traffic demand per pixel (kbps)

TD P
Timeslots
=  ----------------------------------------------------------------------------------
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.
With the number of timeslots required to serve the circuit switched traffic, the timeslots required for packet switched traffic
and their respective distributions according to the timeslot configurations being known, Atoll calculates the number of
timeslots available for carrying the packet switched traffic demand. These timeslots can be dedicated packet switched
timeslots and the shared ones. So, following the principle that shared timeslots are potential carriers of both traffic types,
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, Delay and Blocking Probability is the equivalent
number of available timeslots for packet switched traffic, i.e. NP. This is computed by dividing the total number of timeslots
available for carrying packet switched traffic by the number of downlink timeslots defined in the mobile terminal properties.
So, NP is calculated at this stage as:
TS P
N P = ---------------------------
-
TS Terminal
Where, TSTerminal is the number of timeslots that a terminal will use in packet switched calls. This is determined by taking
the lower of the maximum number of timeslots for packet switched service defined in the service properties and the maxi-
mum number of timeslots that a mobile terminal can use for packet switched services defined in the terminal type proper-
ties.
TS Terminal = min  TS Max Service TS Max TerminalType 
Here, the min(X,Y) function yields the lower value among X and Y as result.
Now, knowing the packet switched traffic load, LP, and the equivalent number of available timeslots, NP, Atoll finds out the
KPIs that have been selected before launching the dimensioning process using the quality curves stored in the dimension-
ing model.
This particular part of this step can be iterative if the KPIs to consider in dimensioning are not satisfied in the first try. If the
KPIs calculated above are within acceptable limits as defined by the user, it means that the dimensioning process has
152 AT271_TRG_E6 © Forsk 2009

acceptable results. If these KPIs are not satisfied, 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. It then recalculates the packet
switched traffic load, LP, and the equivalent number of available timeslots, NP. Then it recomputes the KPIs with these
new values of LP and NP. If the KPIs are within satisfactory limits the results are considered to be acceptable. Otherwise,
Atoll performs another iteration to find the best possible results.
The calculated values of all the KPIs are compared with the ones defined in the service properties. The values for maxi-
mum 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. This calculation is in fact
performed during the traffic analysis process, but since it is relevant to the dimensioning procedure, 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.
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, the average throughput per timeslot deduced from the throughput curves stored in the GPRS/
EDGE equipment properties for each coding scheme, the number of downlink timeslots defined in the properties of the
mobile terminal and the required availability defined in the service properties.
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. According to the measured carrier power or carrier-to-interference ratio, Atoll
deduces the maximum throughput available on that pixel through the throughput vs. C or throughput vs. C/I curves of the
GPRS/EDGE equipment.
The throughput per timeslot per pixel TPTS, Pixel can be either a function of carrier power C, or carrier power C and the
carrier-to-interference ratio C/I, depending on the user-defined traffic analysis RF conditions criteria. Therefore,
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. This parameter renders user-manageable flexibility to the throughput require-
ment constraint.
To calculate the minimum throughput reduction factor for the transmitter, Atoll computes the minimum throughput reduc-
tion factor for each pixel using the formula:
TP user min
RF min Pixel = -------------------------------------------------------------
-
TP TS Pixel  TS Terminal
Once the minimum reduction factor for each pixel is known, Atoll calculates the global minimum reduction factor that is
satisfied by the percentage of covered pixels defined in the required availability. The following example may help in under-
standing the concept and calculation method.
Example: Let the total number of pixels, covered by a subcell S, be 1050. The reliability level set to 90%. This implies that
the required minimum throughput for the given service will be available at 90% of the pixels covered. This, in turn, implies
that there will be a certain limit on the reduction factor, i.e. if the actual reduction factor in that subcell becomes less than
a minimum required, the service will not be satisfactory.
Atoll computes the minimum reduction factor at each pixel using the formula mentioned above, and outputs the following
results:
RFmin Number of pixels

0.3 189
0.36 57
0.5 20
0.6 200
0.72 473
0.9 23
0.98 87
So for a reliability level of 90%, the corresponding RFmin will be the one provided at least 90% of the pixels covered, i.e.
945 pixels. The corresponding value of the resulting RFmin in this example hence turns out to be 0.9, since this value
covers 962 pixels in total. Only 87 of the covered pixels imply an RFmin of 0.98. These will be the pixels that do not provide
satisfactory service.
This calculation is performed for each service type available in the subcell coverage area. The final minimum throughput
reduction factor is the highest one amongst all calculated for each service separately.
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.
© Forsk 2009 AT271_TRG_E6 153

Figure 5.9: Minimum Throughput Reduction Factor
5.3.2.2.5 Step 5: Served PS Traffic

Atoll calculates the served packet switched traffic using the number of timeslots available to carry the packet switched traf-
fic demand. As the result of the above iterative step, 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. Hence, there is no packet
traffic overflow unless the packet switched traffic demand requires more TRXs than the maximum allowed
5.3.2.2.6 Step 6: Total Traffic Load

This step calculates the final result of the dimensioning process, i.e. the total traffic load. The total traffic load L is calculated
as:
ST C + ST P
L = -----------------------------------------------------------------------------------------------
-
TS C dedicated + TS P dedicated + TS S
Where,
• STC is the served circuit switched traffic
• STP is the served packet switched traffic
• TSC, dedicated is the number of dedicated circuit switched timeslots
• TSP, dedicated is the number of dedicated packet switched timeslots
• TSS is the number of shared timeslots
5.4 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. 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.
The concept of this computation is the inverse of that of the dimensioning process. In this case, Atoll has the results of the
dimensioning process already committed and known. Atoll then computes the current values for all the KPIs knowing the
number of required TRXs, the respective numbers of shared and dedicated timeslots and the circuit switched and packet
switched traffic demands.
The computation algorithm utilizes the parameters set in the dimensioning model properties and the quality curves for the
throughput reduction factor, delay and the blocking probability.
The following conventional relations apply:
If,
• TSC, dedicated is the number of timeslots dedicated to the circuit switched traffic,
• TSP, dedicated is the number of timeslots dedicated to the packet switched traffic,
• TSS is the number of shared timeslots for a transmitter,
Then, the number of timeslots available for the circuit switched traffic, TSC, is defined as:
TS C = TS S + TS C dedicated
And the number of timeslots available for the packet switched traffic, TSP, is given by:
TS P = TS S + TS P dedicated
154 AT271_TRG_E6 © Forsk 2009

5.4.1 Circuit Switched Traffic

For each subcell, Atoll has already calculated the effective traffic overflow rate and the blocking rate during the dimension-
ing process. Also knowing the circuit switched traffic demand, TDC, and the number of timeslots available for circuit
switched traffic, TSC, the blocking probability can be easily computed using the Erlang formulas or tables.
5.4.1.1 Erlang B
Under the current conditions of circuit switched traffic demand, TDC, and the number of timeslots available for the circuit
switched traffic, 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, the circuit switched traffic overflow rate, OC, is the same as the
percentage of traffic blocked by the subcell calculated above.
5.4.1.2 Erlang C
Similarly, under the current conditions of circuit switched traffic demand, TDC, and the number of timeslots available for
the circuit switched traffic, TSC, the percentage of delayed circuit switched traffic can be computed through:
TS C
 TD C 
% of traffic delayed = --------------------------------------------------------------------------------------------------------------------------
TS – 1 C
k
TS C TD C  TD C 
 TD C  +  TS C !   1 – -----------   -------------------
 TS  k!
C
k=0
If the circuit switched traffic demand, TDC, is higher than the number of timeslots available to accommodate circuit
switched traffic, the column for this result will be empty signifying that there is a percentage of circuit switched traffic actu-
ally being rejected rather than just being delayed under the principle of Erlang C model.
The circuit switched traffic overflow rate, OC, will be 0 if the circuit switched traffic demand, TDC, is less than the number
of timeslots available for the circuit switched traffic, TSC.
If, on the other hand, the circuit switched traffic demand, TDC, is higher than the number of timeslots available to carry the
circuit switched traffic, TSC, then there will be a certain percentage of circuit switched traffic that will overflow from the
subcell. This circuit switched traffic overflow rate, OC, is calculated as:
TD C – TS C
O C = ----------------------------
-
TD C
5.4.1.3 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, OC. The
served circuit switched traffic, STC, is calculated as:
ST C = TD C   1 – O C 
5.4.2 Packet Switched Traffic

Identifying the total traffic demand, TDT, (circuit switched traffic demand + packet switched traffic demand) as:
TD T = TD C + TD P
The following two cases can be considered.
5.4.2.1 Case 1: Total Traffic Demand > Dedicated + Shared Timeslots

In the case where the total number of timeslots available is less than the total traffic demand, there will be packet switched
data traffic that will be rejected by the subcell as it will not be able to accommodate it. The following results are expected
in this case:
5.4.2.1.1 Traffic Load

The traffic load will be 100%, as the subcell will have more traffic to carry than it can. This implies that the system will be
loaded to the maximum and even saturated. Hence the user level quality of service is bound to be very unsatisfactory.
© Forsk 2009 AT271_TRG_E6 155

5.4.2.1.2 Packet Switched Traffic Overflow

In a 100% loaded, or even saturated subcell, the packet switched data calls will start being rejected because of shortage
of available resources. Hence there will be a perceptible packet switched traffic overflow in this subcell, OP. This overflow
rate is calculated as show below:
  TS C dedicated + TS P dedicated + TS S  – ST C 
-  100
O P = 1 – ----------------------------------------------------------------------------------------------------------------------------
TD P
5.4.2.1.3 Throughput Reduction Factor

The resulting throughput reduction factor for a 100% loaded or saturated subcell will be 0. Hence, the throughput perceived
by the packet switched service user will be 0, implying a very bad quality of service.
5.4.2.1.4 Delay
Again for a 100% loaded or saturated subcell, the delay at the packet switched service user end will be infinite as there is
no data transfer (throughput = 0).

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. Hence, the blocking probability will be 100%.
5.4.2.1.6 Served Packet Switched Traffic

With the packet switched data traffic overflowing from the subcell, there will be a part of that traffic that is not served. The
served packet switched data traffic, STP, is calculated on the same principle as the served circuit switched traffic:
ST P = TD P   1 – O P 
5.4.2.2 Case 2: Total Traffic Demand < Dedicated + Shared Timeslots

In the case where the total traffic demand is less than the number of timeslots available to carry the traffic, the subcell will
not be saturated and there will be some deducible values for all the data KPIs. In a normally loaded subcell, the packet
switched data traffic will have no overflow percentage. This is due to the fact that the packet switched data traffic is rather
placed in a waiting queue than be rejected.
Therefore, there will be a within limits packet switched traffic load, LP, 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 equiv-
alent timeslots available for the packet switched data traffic, NP, which is calculated in the same manner as in the dimen-
sioning process as well:
TS P
N P = ---------------------------
-
TS Terminal
These parameters calculated, now Atoll can compute the required KPIs through their respective quality curves.
5.4.2.2.1 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
5.4.2.2.2 Packet Switched Traffic Overflow

In a normally loaded subcell, no packet switched data calls will be rejected. The packet switched traffic overflow will, there-
fore, be 0.
5.4.2.2.3 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, LP, and number of equivalent timeslots, NP.
5.4.2.2.4 Delay
The resulting delay the subcell is calculated through the delay quality curve for given packet switched traffic load, LP, and
number of equivalent timeslots, NP.

The resulting blocking probability for a normally loaded subcell is calculated through the blocking probability quality curve
for given packet switched traffic load, LP, and number of equivalent timeslots, NP.
156 AT271_TRG_E6 © Forsk 2009

5.4.2.2.6 Served Packet Switched Traffic

As there is no overflow of the packet switched traffic demand from the subcell under consideration, the served packet
switched traffic will be the same as the packet switched traffic demand:
ST P = TD P
5.5 Neighbour Allocation

The intra-technology neighbour allocation algorithm takes into account all the TBC transmitters. It means that all the TBC
transmitters of the .atl document are potential neighbours.
The transmitters to be allocated will be called TBA transmitters. 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 folder can be either the Transmitters folder
or a group of transmitters or a single transmitter.
Only TBA transmitters may be assigned neighbours.
Note:
5.5.1 Global Allocation for All Transmitters

We assume a reference transmitter A and a candidate neighbour, transmitter B.
When automatic allocation starts, Atoll checks following conditions:
1. The distance between both transmitters must be less than the user-definable maximum inter-site distance. If the
distance between the reference transmitter and the candidate neighbour is greater than this value, then the
candidate neighbour is discarded.
2. 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. This constraints can be weighted among the others and ranks the neighbours through
the importance field (see after).
Force adjacent transmitters as neighbours: This option enables you to force transmitters geographically adjacent to the
reference transmitter in the candidate neighbour list. This constraints can be weighted among the others and ranks the
neighbours through the importance field (see after).
Notes:
• Adjacence criterion: Geographically adjacent transmitters are determined on the basis of
their Best Server coverages in 2G (GSM GPRS EGPRS) projects. More precisely, a
transmitter TXi is considered adjacent to another transmitter TXj if there exists at least one
pixel of TXi Best Server coverage area where TXj is the 2nd Best Server. The ranking of
the adjacent neighbour transmitter increases with the number of these pixels. The figure
below shows the above concept.
© Forsk 2009 AT271_TRG_E6 157

• When this option is checked, adjacent cells are sorted and listed from the most adjacent to
the least, depending on the above criterion. Adjacence is relative to the number of pixels
satisfying the criterion.
• This criteria is only applicable to transmitters belonging to the same HCS layer. The
geographic adjacency criteria is not the same in 3G (UMTS HSPA, CDMA2000) projects.
Force neighbour symmetry: This option enables user to force the reciprocity of a neighbourhood link. Therefore, if the refer-
ence transmitter is a candidate neighbour of another transmitter, the later will be considered as candidate neighbour of the
reference transmitter.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you may
force/forbid a transmitter to be candidate neighbour of the reference transmitter.
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, the existing neighbours are kept.
3. There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability where:
• SA is the area where the received signal level from the transmitter A is greater than a minimum signal level. SA is
the coverage area of reference transmitter A restricted between two boundaries; the first boundary represents the
start of the handover area (best server area of A plus the 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”)
• SB is the coverage area where the candidate transmitter B is the best server.
SA  SB
Atoll calculates either the percentage of covered area ( ----------------------  100 ) if the option “Take into account Covered Area” is
SA
selected, or the percentage of traffic covered on the overlapping area S A  S B for the option “Take into account Covered
Traffic”. Then, 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”). If this percentage is not exceeded, the candidate neighbour B is discarded.
The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see
number 4 below).
Figure 5.10: Overlapping Zones
4. The importance values are used by the allocation algorithm to rank the neighbours according to the allocation
reason, and to quantify the neighbour importance.
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. 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.
158 AT271_TRG_E6 © Forsk 2009

Among these 15 candidate neighbours, only 8 (having the highest importances) will be allocated to the reference trans-
mitter.
As indicated in the table below, the neighbour importance depends on the neighbourhood cause; this value goes from 0
to 100%.
Importance
Neighbourhood cause When
value
Only if the Delete existing neighbours option is not selected Existing
Existing neighbour
and in case of a new allocation importance
Exceptional pair Only if the Force exceptional pairs option is selected 100 %
Only if the Force co-site transmitters as neighbours option is
Co-site transmitter (IF) function
selected
Only if the Force adjacent transmitters as neighbours option is
Adjacent transmitter (IF) function
selected
Neighbourhood relationship that fulfils
Only if the % minimum covered area is exceeded (IF) function
coverage conditions
Symmetric neighbourhood
Only if the Force neighbour symmetry option is selected (IF) function
relationship
Except forced neighbour case (importance = 100%), priority assigned to each neighbourhood cause is now linked to the
(IF) Importance Function evaluation. The importance is evaluated through a function (IF), taking into account the following
3 factors:
• Co-site factor (C) which is a Boolean factor,
• Adjacency factor (A) which deals with the percentage of adjacency,
• Overlapping factor (O) meaning the percentage of overlapping
The (IF) function is user-definable using the Min importance and Max importance fields.
Min Max
Factor Default value Default value
importance importance
Overlapping factor (O) Min(O) 1% Max(O) 30%
Adjacency factor (A) Min(A) 30% Max(A) 60%
Co-site factor (C) Min(C) 60% Max(C) 100%
The (IF) function is evaluated as follows:
Neighbourhood cause (IF) function with default Min

(IF) function
Co-site Adjacent and Max default values
no no Min(O) + Delta(O)(O) 1% + 29%(O)
Min(A)+Delta(A){Max(O)(O)+(100%-
no yes 30% + 30%{30%(O) + 70%(A)}
Max(O))(A)}
Min(C)+Delta(C){Max(O)(O)+(100%-
yes yes 60% + 40%{30%(O )+ 70%(A)}
Max(O))(A)}
Where Delta(x) = Max(x) - Min(x)

Notes:
• If there is no overlapping between the range of each factor, the neighbours will be ranked
by neighbourhood cause. Using the default values for minimum and maximum importance
fields, neighbours will be ranked in this order: first co-site neighbours, then adjacent
neighbours, and finally neighbours found on overlapping criterion.
• If ranges of (IF) factors overlap each other, the neighbours may not be ranked by
neighbourhood cause.
• The ranking between neighbours from the same category will depend on (A) and (O)
factors.
• 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 overlapping.
In the Results part, Atoll provides the list of neighbours, the number of neighbours and the maximum number of neigh-
bours 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, adjacent, coverage or symmetric. For neighbours
accepted for co-site, adjacency and coverage reasons, Atoll displays the percentage of area meeting the coverage condi-
tions (or the percentage of covered traffic on this area) and the corresponding surface area (km2) (or the traffic covered
on the area in Erlangs), the percentage of area meeting the adjacency conditions and the corresponding surface area
(km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing.
© Forsk 2009 AT271_TRG_E6 159

Notes:
• 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.
• 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 (or the percentage of covered traffic) is calculated with the
resolution specified in the property dialog of the predictions folder (Default resolution
parameter).
• When the option “Force adjacent transmitters as neighbours” is used, the margin
“handover start” is not taken into account. Atoll considers a fixed value of 0 dB.
• 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.
• The force neighbour symmetry option enables the users to consider the reciprocity of a
neighbourhood link. This reciprocity is allowed only if the neighbour list is not already full.
Thus, if transmitter B is a neighbour of the transmitter A while transmitter A is not a
neighbour of the transmitter B, two cases are possible:
1st case: There is space in the transmitter B neighbour list: the transmitter A will be added
to the list. It will be the last one.
2nd case: The transmitter B neighbour list is full: Atoll will not include transmitter A in the list
and will cancel the link by deleting transmitter B from the transmitter A neighbour list.
• When the options “Force exceptional pairs” and “Force symmetry” are selected, 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 one, symmetry cannot be respected. In this case, Atoll
displays a warning in the Event viewer.
• In the Results, Atoll displays only the transmitters for which it finds new neighbours.
Therefore, if a transmitter has already reached its maximum number of neighbours before
starting the new allocation, it will not appear in the Results table.
5.5.2 Allocation for a Group of Transmitters or One Transmitter

In this case, Atoll allocates neighbours to:
• TBA transmitters,
• Neighbours of TBA transmitters marked as exceptional pair, adjacent and symmetric,
• Neighbours of TBA transmitters that satisfy coverage conditions.
Automatic neighbour allocation parameters are described in "Global Allocation for All Transmitters" on page 157.
5.6 Interference Prediction Studies

Two interference studies with predefined settings are available:
• The coverage by C/I level study: This study provides you a global analysis of the network quality.
• The interfered areas study: This study shows the areas where a transmitter is interfered by other ones.
In both cases, Atoll calculates C/I ratio on each calculation bin where conditions on signal level reception are satisfied.
Then, it either considers the bins where the calculated C/I exceeds a lower threshold in the coverage and colours these
bins depending on C/I value (coverage by C/I level study), or it considers the bins where the calculated C/I is lower than
a upper threshold in the coverage and colours them depending on colour of the interfered transmitter (interfered areas
study).
The user-defined thermal noise (N) value is used in the calculations if the corresponding calculation conditions are
selected in the conditions tab of an interference study. The thermal noise is considered to be a white guassian background
noise fixed at the user-defined value for the entire network or part of the network under consideration. This value is defined
in the document database at -121 dBm by default.
All the TBC transmitters are taken into account in these studies. Let us assume that each bin within each TBC transmitter
calculation area corresponds to a probe mobile receiver.
• The study conditions in order to determine the coverage area of each TBC transmitter
• The display settings to select how to colour coverage areas.
160 AT271_TRG_E6 © Forsk 2009

Note:
• For information on the common prediction studies (like coverage by transmitter, profile
study, …), please, refer to Common prediction studies part.

The areas, where Atoll will calculate C/I, depend on signal level reception conditions. Atoll uses the parameters entered
in the Conditions tab in order to determine service area of each TBC transmitter. We can distinguish four cases:
Here we presume that:
• Each transmitter, Txi, belongs to a hierarchical cell structure (HCS) layer, k, with a defined priority.

Txi
Minimum threshold  P rec  tt   Maximum threshold
5.6.1.1.2 Best Signal Level per HCS Layer and a Margin

Txi
and
Txi Txj
ji
where,
5.6.1.1.3 Best Signal Level of the Highest Priority HCS Layer and a Margin
In this case, the service area of Txi corresponds to the bins where:
Txi
and
Txi Txj
ji
and
Txi belongs to the HCS layer with the highest priority
where,
5.6.1.1.4 Second Best Signal Level per HCS Layer and a Margin
Txi
and
Txi nd Txj
P rec  BCCH   2 Best  P rec  BCCH   – M
ji
where,

Note:
• When the maximum range option is selected, Atoll searches for interference on the bins:
- Where the respective criteria described above are checked, and
- Located within a specified distance from the transmitter (maximum range).
5.6.1.2 Carrier to Interference Ratio Calculation

Atoll works out carrier to interference ratio on each bin of transmitter service areas.
© Forsk 2009 AT271_TRG_E6 161

In order to understand the difference between each frequency hopping mode from the mobile point of view, it is interesting
to consider the Mobile Station Allocation (MSA). MSA is characterised by the pair (Channel list, MAIO). When a non
hopping (NH) mode is used, channel list is a channel while it corresponds to the mobile allocation list (MAL) in case of
base band hopping (BBH) or synthesised frequency hopping (SFH). For BBH, channels of MAL belong to a unique TRX
type.
Examples:
For each example given below, we assume that.
In case of NH, we have:
TRX index Channel list MAIO MSA

1 53 * (53,*)
2 54 * (54,*)
In case of BBH, assuming TRXs belong to the same TRX type, we have:

1 53 * ([53,54,55],0)
2 54 * ([53,54,55],1)
3 55 * ([53,54,55],2)
In case of SFH, we have:

1 53 54 55 56 2 ([53,54,55,56],2)
2 53 54 55 56 3 ([53,54,55,56],3)
Therefore, for a mobile station, BBH and SFH work in the same way.
Consider the following notations:
v is a victim transmitter (TBC transmitter with a service area),
MSAS(v) is the set of MSAs associated to v. The number of MSAS(v) depends on TRX type(s) to be analysed (option
available in study properties): you may study a given TRX type tt (There are as many MSA(v) as TRXs allocated to the
subcell (v,tt)) or all the TRX types (The number of MSA(v) corresponds to the number of TRXs allocated to v),
i is a potential interfering transmitter (TBC transmitters which calculation area intersects service area of v),
MSAS(i) is the set of MSAs related to potential interferers i,
INT(v) is the set of transmitters that interfere v.
Several MSAs, m, are related to a transmitter. Therefore, for each victim transmitter v with MSA m (m  MSAS(v)), Atoll
 C v  m 
calculates carrier to interference ratio  ----------------
- , received at the mobile; mobile is connected to a victim transmitter, v with
 Iv  m  
v v
a given m. C  m  is the carrier power level received from v on m and I  m  corresponds to the interference received from
interfering transmitters i on m.
Atoll studies the most interfered MSA. So, it considers:
C  C v  m 
---- = Min  ----------------
- except if analysis is detailed (Detailed result option).
 I v k  Iv  m  
If the interference conditions for the prediction study are defined using the option C/(I+N), Atoll takes the total noise N tot
into account as well. The total noise is computed by adding the thermal noise N thermal (defined in the document database
at -121 dBm by default) to the noise figure NF (either defined at the terminal type properties level, if a terminal type is
defined for the study, or defined directly in the prediction study conditions). So,
N tot = N thermal + NF
C  Cv  m  
Thus, for computations based on C/(I+N),  ------------------ = Min  ------------------------------
-
 I + N tot k  I v  m  + N tot
v
Note:
• The M Shadowing used in the computations of C/I is a function of C/I standard deviation and
not the Model standard deviation.
5.6.1.2.1 Carrier Power Level

v v
C  m  = P rec  m 
162 AT271_TRG_E6 © Forsk 2009

5.6.1.2.2 Interference Calculation

Potential interferers can be transmitters i (iv), using co-channels and/or adjacent channels. Therefore, we can write:
v v v i
I  m  = I co  m  + I adj  m  + I IMPx3 – G PC
v v
Where I co  m  is the interference received at v on m due to co-channels, I adj  m  is the interference received at v on m
i
due to adjacent channels, I IMPx3 is the third order intermodulation interference, and G PC is the average power control
gain defined for the interfering transmitter i.
v
I co  m  is the interference received at v on m due to co-channels, given by:
 v i 
 
v i
I co  m  =  p m n  P rec  n   T i  n 
 
i  INT  v  n  MSAS  i  co
v
And, I adj  m  is the interference received at v on m due to adjacent channels, given by:
i
 v i  P rec  n 
 
v
I adj  m  =  p m n  -------------------  T i  n 
  F
i  INT  v  n  MSAS  i  adj
v i
p m n is the probability of having a co- or adjacent channel collision between MSAs n and m (when n and m contain co-
and adjacent channels). It depends on the used frequency hopping mode.
i
P rec  n  is the carrier power level received from i on n,
Ti(n) is occupancy of the MSA n.
i i
T i  n  = L traffic  n   f act  n 
i
If “Average” is selected in the study properties, L traffic  n  is the traffic load defined for the MSA n of i. If “Maximum” option
i
is selected, L traffic  n  = 1 .
i
f act  n  is the activity factor defined for the MSA n of i. If the subcell (i,tt) supports DTX mode, it is a global value specified
in the study properties. Otherwise, the activity factor is 1.
Note:
• Since BCCH carrier is always On Air, DTX and traffic load gains do not reduce BCCH
i i
interfering energy. In other words, f act  n  = 1 and L traffic  n  = 1 on the BCCH for the
interference estimation.
I IMPx3 has three components, i.e., intermodulation interference from frequencies used by the interfering transmitter, inter-
ference due to spurious emissions from the interfering transmitter, and the intermodulation interference received at the
mobile terminal:
TX SE Term
I IMPx3 = I IMPx3 + I IMPx3 + I IMPx3
The above components are calculated as follows:

i
TX P rec
I IMPx3 = ----------------
-
v
L IMPx3
SE SE i
I IMPx3 = P rec
i
Term 3  P rec
I IMPx3 = --------------------------
-
Prot
2  F IMPx3
i
Where P rec is the carrier power level received from the interferer i, L IMPx3 is the third order intermodulation loss at the
SE i Prot
victim transmitter v, P rec is the spurious emission power level received from the interferer i, and F IMPx3 is the third order
intermodulation protection factor for the terminal.
For a pair of frequencies, f 1 and f 2 , two third order intermodulation products are generated at frequencies f 3 = 2  f 1 – f 2
and f 4 = 2  f 2 – f 1 . If a transmitter uses f 3 or f 4 , it is interfered by transmitters using f 1 and f 2 . All interferer frequencies
are used to calculate intermodulation products. When several frequency pairs generate intermodulation products, the
IMPs are independenly calculated and added to the interference. If power received over different frequencies is not the
same for two frequencies (not the same power offset for example), the corresponding intermodulation frequencies are
© Forsk 2009 AT271_TRG_E6 163

ignored. Frequency hopping is not considered to have any impact on the intermodulation products. IMPs for hopping and
non-hopping cases are considered to be the same.
Intermodulation products generated by the adjacent frequencies of the frequencies actually being used by an interferer
are not taken into account. Similarly, intermodulation interference received on the adjacent frequencies of the frequencies
used by the victim are also ignored.
5.6.1.2.3 Collision Probability for Non Hopping Mode

v i
We have: p m n = 1
5.6.1.2.4 Collision Probability for BBH and SFH Modes

MSA m of v can be defined as the pair ([f1,f2,….fn], MAIO) and MSA n of i as the pair ([f’1,f’2,….f’n], MAIO’) (where f and f’
are channels).
v i
Now, let us consider the occurrence, OCCUR  f m f' n  , such that a channel f of m can meet a channel f’ of n during hopping
sequence.
There is a collision if f and f’ are co- or adjacent channels. Then, we can define a collision as follows:
v i v i
Collision = OCCUR  f m f' n  such that f m – f' n = 
v i
(  equals 0 if f m and f' n are co-channels or 1 if adjacent channels)
Therefore, we have:
v i n collision
p m n = --------------------------
-
n occurence
ncollision and noccurence respectively correspond to the number of collisions and the number of occurrences. They are
closely linked to the correlation between m and n. We can have two cases:
1st Case: MSAs m and n are Correlated
m and n must have identical HSN and synchronisation.

The number of occurrences depends on the MAL size, MAIO And MAIO’.
Example:
Schematic view of hopping sequences

MSA m of v
34 37 39
([34 37 39], MAIO=0)
MSA n of i
38 36 34
([38 36 34], MAIO’=2)
Here, the number of occurrences is 3; the number of co-channel collisions is 1 and the number of adjacent channel colli-
sions is 1.
So, we have:
v i 1 v i 1
 p m n  co = --- and  p m n  adj = ---
3 3
2nd Case: MSAs m and n are Not Correlated
Condition specified above is not fulfilled.

Probability to have each pair is the same. All the occurrences are possible.
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], MAIO’=2) 38 36 34
Here, the number of occurrences is 9; the number of co-channel collisions is 1 and the number of adjacent channel colli-
sions is 3.
So, we have:
v i 1 v i 1
 p m n  co = --- and  p m n  adj = ---
9 3
164 AT271_TRG_E6 © Forsk 2009

Note:
• Only the carrier power level is downgraded by the shadowing margin. The interference
level is not altered.
5.6.1.3 Coverage Area Determination

C C
For each victim transmitter v, coverage area corresponds to bins where  ---- or  ------------ is between lower and upper
 I v  I + N v
thresholds specified in study properties. There are two possibilities:
5.6.1.3.1 Interference Condition Satisfied by At Least One TRX

This criterion implies that the interference condition defined in the interference study properties dialog must be satisfied
by at least on TRX of the transmitter in order for the pixel under study to be included in the coverage area.
In this case, the coverage area of a transmitter Txi corresponds to the bins where:
C C
Minimum threshold   ----  Maximum threshold or Minimum threshold   ------------  Maximum threshold
 I v  I + N v
TRX j TRX j
where, TRXj is any TRX belonging to Txi.
This coverage area will include all the bins satisfying the above criteria even if they are only covered by the TRX with the
best C/I or C/(I+N) conditions.
5.6.1.3.2 Interference Condition Satisfied by The Worst TRX

This criterion implies that the interference condition defined in the interference study properties dialog must be satisfied
by the worst TRX of the transmitter in order for the pixel under study to be included in the coverage area.
In this case, the coverage area of a transmitter Txi corresponds to the bins where:
C C
Minimum threshold   ----  Maximum threshold or Minimum threshold   ------------  Maximum threshold
 I v  I + N v
TRX j TRX j
where, TRXj is the TRX (belonging to Txi) with the worst C/I or C/(I+N) conditions at the bin.
This coverage area will include only the bins satisfying the above criteria, i.e. covered by the TRX with the worst C/I or C/
(I+N) conditions.
5.6.1.4 Coverage Area Display

You can display the transmitter coverage area depending on the C/I (or C/(I+N)) level, prefer a display depending on trans-
mitter colour or on any other transmitter attribute.
5.6.1.4.1 C/I Level

Each bin of the transmitter coverage area is coloured if the calculated C/I (or C/(I+N)) level exceeds (  ) the specified mini-
mum thresholds (bin colour depends on C/I (or C/(I+N)) level). Coverage consists of several independent layers whose
visibility in the workspace can be managed. There are as many layers as transmitter coverage areas. Each layer shows
the different C/I levels available in the transmitter coverage area.
5.6.1.4.2 Max C/I Level

Atoll compares calculated C/I (or C/(I+N)) levels received from transmitters on each bin of each transmitter coverage area
where coverage areas overlap the studied one and chooses the highest value. A bin of a coverage area is coloured if the
C/I (or C/(I+N)) level exceeds (  ) the specified thresholds (the bin colour depends on the C/I (or C/(I+N)) level). Coverage
defined thresholds. Each layer corresponds to an area where the highest received C/I level exceeds a defined minimum
threshold.
5.6.1.4.3 Min C/I Level

Atoll compares C/I (or C/(I+N)) levels received from transmitters on each bin of each transmitter coverage area where the
coverage areas overlap the studied one and chooses the lowest value. A bin of a coverage area is coloured if the C/I (or
C/(I+N)) level exceeds (  ) the specified thresholds (the bin colour depends on the C/I (or C/(I+N)) level). Coverage
defined thresholds. Each layer corresponds to an area where the lowest received C/I level exceeds a defined minimum
threshold.
5.6.1.4.4 Transmitter
Atoll colours each bin of each transmitter coverage area. The bin colour corresponds to the transmitter colour. Coverage
interfered transmitters.
© Forsk 2009 AT271_TRG_E6 165


Analysis provided in the Interference tab is based on path loss matrices. You can study interference on:
• TBC transmitters for which path loss matrices have been computed,
• calculation areas.
Atoll indicates the following at the receiver:
• The carrier power level received from the victim transmitter v on the most interfered MAS m,
• Either the overall interference received from interfering transmitters i on MAS m (both co-channel and adjacent
channel interferers are considered), or the co-channel interference received from co-channel interfering transmit-
ters i on MAS m, or the adjacent channel interference received from adjacent channel interfering transmitters i on
MAS m (for further information about noise calculation, please refer to Signal to noise calculation: noise calculation
part)
• The interference level received from each interfering transmitter i on m. Interferers are sorted in a descending
order w.r.t. carrier power level.
Notes:
• Neither DTX nor traffic load of TRXs are taken into account to evaluate interference levels.
i i
Therefore, we have T i  n  = L traffic  n   f act  n  = 1 .
• Only carrier power level is downgraded by the shadowing margin. The interference level is
not altered.
5.7 GPRS EGPRS Coverage Studies

Atoll calculates a coverage area for all the TBC transmitters, assuming that each bin within a TBC EGPRS transmitter
calculation area corresponds to a probe mobile receiver.
• The study conditions in order to determine the coverage area of each TBC transmitter,
5.7.1 Coverage Area Determination

Let us assume that:
• Each transmitter, Txi, belongs to a HCS layer, k, with a defined priority and a reception threshold.
• Each transmitter, Txi, is a GPRS/EGPRS station (Txi-EGPRS as notation).
• GPRS/EGPRS equipment installed on each transmitter, Txi, does not support 8PSK modulation.
We can have the following four cases.
5.7.1.1 All Servers

For each HCS layer, k, the coverage area of Txi corresponds to Txi calculation area.
5.7.1.2 Best Signal Level per HCS Layer and a Margin

For each HCS layer, k, the coverage area of Txi corresponds to the bins where the signal level received from Txi,
Txi – EGPRS
P rec  BCCH  , is the highest one (Txi is the best server) or within a defined margin of the highest signal level (within
a margin of the best server).
Note:
• If the margin equals 0, the coverage area of Txi corresponds to the bins where
Txi – EGPRS
P rec  BCCH  is the highest.
5.7.1.3 Second Best Signal Level per HCS Layer and a Margin
Txi – EGPRS
P rec  BCCH  , is the second highest one (Txi is the second best server) or within a defined margin of the second
highest signal level (within a margin of the second best server).
Note:
Txi – EGPRS
P rec  BCCH  is the second highest.
166 AT271_TRG_E6 © Forsk 2009

5.7.1.4 HCS Servers and a Margin

Txi – EGPRS
Txi – EGPRS
a margin of the best server). In addition P rec  BCCH  must exceed the reception threshold defined per HCS layer.
Notes:
Txi – EGPRS
5.7.1.5 Highest Priority HCS Server and a Margin

In this case, the coverage area of Txi corresponds to the bins where the signal level received from Txi,
Txi – EGPRS
a margin of the best server). And, Txi belongs to the HCS layer with the highest priority. The highest priority is defined by
Txi – EGPRS
the priority field (0: lowest) assuming the received P rec  BCCH  exceeds the reception threshold defined per HCS
layer.
Note:
5.7.2 Calculation Options

GPRS/EGPRS studies can be based either on signal level (C), or on the carrier-to-interference ratio (C/I) with or without
considering thenoise (N).
If a reference terminal type is defined when launching a CQI study, Atoll will consider the noise figure defined for that termi-
nal type when computing the total noise ( N tot ). If no terminal type is defined, the value defined in the coverage prediction
properties (8 dB, by default) is used.
Different GPRS/EGPRS equipment can be defined for a transmitter and for the reference terminal type. In this case, Atoll
performs an intersection for the coding schemes defined in the transmitter and the reference terminal type GPRS/EGPRS
equipment to determine the coding schemes that are common in the two. Hence, Atoll creates a temporary GPRS/EGPRS
equipment comprising only these common coding schemes and uses this equipment to eventually compute the coding
scheme, throughput and other values.
If no reference terminal type is defined or if the defined reference terminal type has no GPRS/EGPRS equipment assigned
to it, Atoll will perform the computations based on the GPRS/EGPRS equipment of the transmitter. Similarly, if a transmitter
has no GPRS/EGPRS equipment defined, Atoll will compute the coverage study considering only the GPRS/EGPRS
equipment defined for the reference terminal type. If there is no GPRS/EGPRS equipment defined for the transmitter and
for the reference terminal type, there will be no coverage for the transmitter.
5.7.2.1 Calculations Based on C

In this case, only the received signal level is taken into account. Atoll evaluates the signal level received from GPRS/
Txi – EGPRS
EGPRS transmitters on TRXs (TRX) belonging to a selected TRX type (tt) or on all the TRXs, P rec  TRX  .
5.7.2.2 Calculations Based on C/I Without Considering Thermal Noise

For GPRS/EGPRS studies based on the received signal level and C/I ratio, Atoll evaluates:
• The signal level received from GPRS/EGPRS transmitters on TRXs (TRX) belonging to the selected TRX type (tt)
Txi – EGPRS
or on all the TRXs, P rec  TRX  .
And
• The carrier-to-interference ratio received on TRXs (TRX) belonging to the selected TRX type (tt) or on all the
Txi – EGPRS
P rec  TRX 
TRXs, -------------------------------------------------- .
I
© Forsk 2009 AT271_TRG_E6 167

Notes:
• When GPRS/EGPRS prediction studies calculations are based on C/I, Atoll calculates the
carrier-to-interference ratio for all the EGPRS TBC transmitters but takes into account all
the TBC transmitters (GSM and GPRS/EGPRS) to evaluate the interference.
• For further information on interference (I) calculation, please refer to Interference prediction
studies: Interference calculation part.
5.7.2.3 Calculations Based on C/I Considering Thermal Noise

For GPRS/EGPRS studies based on the received signal level and C/I ratio considering the effect of thermal noise, Atoll
evaluates:
• The received signal-level-to-thermal-noise ratio from GPRS/EGPRS transmitters on TRXs (TRX) belonging to the
Txi – EGPRS
P rec  TRX 
selected TRX type (tt) or on all the TRXs, -------------------------------------------------- .
N
And
• The carrier-to-interference-and-noise ratio received on TRXs (TRX) belonging to the selected TRX type (tt) or on
Txi – EGPRS
P rec  TRX 
all the TRXs, -------------------------------------------------- .
I+N
Where, N is the thermal noise whose value is defined in the document database at -121 dBm by default.
Notes:
• When GPRS/EGPRS prediction studies calculations are based on C/I, Atoll calculates the
carrier-to-interference ratio for all the EGPRS TBC transmitters but takes into account all
the TBC transmitters (GSM and GPRS/EGPRS) to evaluate the interference.
• For further information on interference (I) calculation, please refer to Interference prediction
studies: Interference calculation part.
5.7.2.4 Ideal Link Adaptation (ILA)

Ideal link adaptation implies that the coding scheme selected will be the one that provides the maximum throughput.
5.7.3 Coverage Study Scenarios

With the above options, there are many different possible scenarios of calculations. These scenarios are described below.
5.7.3.1 GPRS/EGPRS Studies Based on C Without ILA

5.7.3.1.1 Coding Scheme Based on C Without ILA
Atoll calculates the signal level received from Txi on each bin of the Txi coverage area. Then, selects a coding scheme,
cs, from among the coding schemes available in the GPRS/EGPRS equipment, such that:
For each TRX type, tt, cs = Lowest  CS 

 Txi – EGPRS
P rec  TRX    Reception Threshold  CS
Where, cs is the resulting coding scheme, CS is the set of all available coding schemes, and  Reception Threshold CS
are the values of reception thresholds for the coding schemes available in the GPRS/EGPRS equipment, defined in the
Reception Thresholds column of the properties dialog.
There can be more than one coding schemes whose reception thresholds are less than the received signal level. cs is the
coding scheme with the lowest coding scheme number.
A Power Backoff, in dBs, can be defined for each subcell using an 8PSK modulation (EDGE) based codec equipment.
This power backoff is taken in to account when selecting the codec mode available in the coverage area of the subcell,
such that:
Txi – EGPRS Txi – EGPRS
For each TRX type, tt, P rec  TRX  – P Backoff  TRX    Reception Threshold  CS
5.7.3.1.2 Throughput Based on C Without ILA

Txi – EGPRS
Atoll reads the throughput value for the received signal level, P rec  TRX  , from the Throughput=f(C) graph asso-
ciated to the coding scheme, cs, determined above.
168 AT271_TRG_E6 © Forsk 2009

5.7.3.2 GPRS/EGPRS Studies Based on C With ILA

5.7.3.2.1 Coding Scheme Based on C With ILA
With Ideal Link Adaptation active, Atoll selects the coding scheme that provides the highest throughput at the received
signal level value for a bin.
Atoll calculates the signal level received from Txi on each bin of the Txi coverage area. Then, selects a coding scheme,
cs, from among the coding schemes available in the GPRS/EGPRS equipment, such that:
For each TRX type, tt, cs = Lowest  CS 

 TP = Highest  TP=f  C=P rec
Txi – EGPRS
 TRX   
Where, cs is the resulting coding scheme, CS is the set of all available coding schemes, and
Txi – EGPRS
Highest  TP=f  C=P rec  TRX    is the highest throughput provided by any coding scheme at the received signal
level, from the Throughput = f(C) graphs defined in the properties dialog.
If there are more than one coding schemes providing the highest throughput at the bin, cs will be the one with the lowest
coding scheme number.
such that:
For each TRX type, tt, TP = Highest  TP=f  C=P rec  TRX -  P Backoff  TRX    
5.7.3.2.2 Throughput Based on C With ILA

The throughput at the given bin and for each TRX type, tt, is simply the throughput computed earlier for the coding scheme,
cs, determined above. This throughput is computed from the Throughput = f(C) graphs for the value of the received signal
Txi – EGPRS
level at the bin, P rec  TRX  .
5.7.3.3 GPRS/EGPRS Studies Based on C/I Without ILA and Thermal Noise
5.7.3.3.1 Coding Scheme Based on C/I Without ILA and Thermal Noise
Atoll calculates signal level and C/I level received from Txi on each bin of the Txi coverage area. Then, selects two coding
schemes from among the coding schemes available in the GPRS/EGPRS equipment, such that:
For each TRX type, tt, cs C = Lowest  CS 

 Txi – EGPRS
P rec  TRX    Reception Threshold  CS
 
And, cs C  I = Lowest  CS 
 Txi – EGPRS
P rec  TRX  
 C- 
-----------------------------------------------------   --- 
I  I Threshold CS
Where, csC is the coding scheme determined from the signal level, csC/I is the coding scheme determined from the C/I
level, and CS is the set of all available coding schemes.
 Reception Threshold  CS are the values of reception thresholds for the coding schemes available in the GPRS/EGPRS
C
equipment, defined in the Reception Thresholds column of the properties dialog. And,  ---- Threshold are the values of
I CS
C/I thresholds for the coding schemes available in the GPRS/EGPRS equipment, defined in the C/I Thresholds column of
the properties dialog.
There can be more than one coding schemes whose reception thresholds are less than the received signal level, and
whose C/I thresholds are less than the received C/I level. cs is the coding scheme with the lowest coding scheme number.
The resulting coding scheme, cs, is the coding scheme with the lowest coding scheme number from among csC and csC/I.
cs = Lowest  cs C cs C  I 
such that:
For each TRX type, tt, P rec  TRX  – P Backoff  TRX    Reception Threshold  CS
And
P rec  TRX  – P Backoff  TRX   C
-  ---- Threshold
----------------------------------------------------------------------------------------------------------
I I  CS
© Forsk 2009 AT271_TRG_E6 169

5.7.3.3.2 Throughput Based on Worst Case Between C and C/I Without ILA
Atoll determines two throughput values, TPC from the Throughput = f(C) graph corresponding to the coding scheme csC
determined above, and TPC/I from Throughput = f(C/I) graph corresponding to the coding scheme csC/I determined above.
The resulting throughput TP is the lower of the two values, TPC and TPC/I.
TP = Lowest  TP C TP C  I 
5.7.3.4 GPRS/EGPRS Studies Based on C/I With ILA and Without Thermal Noise
5.7.3.4.1 Coding Scheme Based on C/I With ILA and Without Thermal Noise
With Ideal Link Adaptation active, Atoll selects the coding scheme that provides the highest throughput at the received
signal level and C/I value for a bin.
Atoll calculates signal level and C/I level received from Txi on each bin of the Txi coverage area. Then, selects two coding
For each TRX type, tt, cs C = Lowest  CS 

 TP = Highest  TP=f  C=P rec
Txi – EGPRS
 TRX   
 
 
And, cs C  I = Lowest  CS Txi – EGPRS

   C P rec  
 TP = Highest  TP=f  ---- = -----------------------------------  TRX   
  I I 
Where, csC is the coding scheme determined from the signal level, csC/I is the coding scheme determined from the C/I
level, and CS is the set of all available coding schemes.
Txi – EGPRS
Highest  TP=f  C=P rec  TRX    is the highest throughput provided by any coding scheme at the received signal
Txi – EGPRS
  C P rec 
level, from the Throughput = f(C) graphs defined in the properties dialog. And, Highest  TP=f  ---- = --------------------------------  TRX  
 I I 
is the highest throughput provided by any coding scheme at the received C/I level, from the Throughput = f(C/I) graphs
defined in the properties dialog.
If there are more than one coding schemes providing the highest throughput at the bin, csC and csC/I will be the ones with
the lowest coding scheme numbers.
The resulting coding scheme, cs, is the coding scheme with the lowest coding scheme number from among csC and csC/I.
cs = Lowest  cs C cs C  I 
such that:
For each TRX type, tt, TP = Highest  TP=f  C=P rec  TRX -  P Backoff  TRX    
And
  C P rec  TRX -  P Backoff  TRX   
TP = Highest  TP=f  ---- = ------------------------------------------------------------------------------------------------------------- 
 I I 
5.7.3.4.2 Throughput Based on Worst Case Between C and C/I With ILA
Atoll determines two throughput values, TPC from the Throughput = f(C) graph corresponding to the coding scheme csC
determined above, and TPC/I from Throughput = f(C/I) graph corresponding to the coding scheme csC/I determined above.
The resulting throughput TP is the lower of the two values, TPC and TPC/I.
TP = Lowest  TP C TP C  I 
5.7.3.5 GPRS/EGPRS Studies Based on C/I Without ILA and With Thermal Noise
5.7.3.5.1 Coding Scheme Based on C/I Without ILA and With Thermal Noise
The reception thresholds given for signal level C are internally converted to C/N thresholds (where 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. C/I thresholds are
also indexed by the C/(I+N) value.
Atoll calculates the C/N and C/(I+N) level received from Txi on each bin of the Txi coverage area. Then, selects two coding
 
For each TRX type, tt, cs C  N = Lowest  CS 
 Txi – EGPRS
 C- 
-----------------------------------------------------   ----------- Threshold
N I + N  CS
170 AT271_TRG_E6 © Forsk 2009

 
And, cs C   I + N  = Lowest  CS 
 Txi – EGPRS
 C- 
-----------------------------------------------------   ----------- Threshold
I+N I + N  CS
Where, csC/N is the coding scheme determined from the C/N level, csC/(I+N) is the coding scheme determined from the C/
(I+N) level, and CS is the set of all available coding schemes.
C-
 ----------- Threshold
I + N  CS are the values of C/(I+N) thresholds for the coding schemes available in the GPRS/EGPRS equip-
ment, determined from the C/I threshold values defined in the C/I Thresholds column of the properties dialog.
There can be more than one coding schemes whose C/(I+N) thresholds are less than the received C/N level, whose C/
(I+N) thresholds are less than the received C/(I+N) level. cs is the coding scheme with the lowest coding scheme number.
The resulting coding scheme, cs, is the coding scheme with the highest coding scheme number from among csC/N and
csC/(I+N).
cs = Highest  cs C  N cs C   I + N  
such that:
P rec  TRX  – P Backoff  TRX  C
For each TRX type, tt, -----------------------------------------------------------------------------------------------------------   ------------ Threshold
N I + N  CS
And
P rec  TRX  – P Backoff  TRX   C
-  ------------ Threshold
----------------------------------------------------------------------------------------------------------
I+N I + N  CS
5.7.3.5.2 Throughput Based on Interpolation Between C/N and C/(I+N) Without ILA
Atoll determines two throughput values, TPC/N from the Throughput = f(C) graph corresponding to the coding scheme csC/
N determined above, and TPC/(I+N) from Throughput = f(C/I) graph corresponding to the coding scheme csC/(I+N) deter-
mined above.
The Throughput = f(C) graph is internally converted to Throughput = f(C/N) graph, in order to be indexed with the C/(I+N)
value. The Throughput = f(C/I) graph is also indexed with the C/(I+N) value. The final throughput is computed by interpo-
lating 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 resulting throughput TP is given by:
TP =   TP C  N +  1 –    TP C   I + N 
Where,
pN
 = ---------------------
pI + N
pN is the thermal noise power (value in Watts)
p(I+N) is the interferences + thermal noise power (value in Watts)
TPC/N is the throughput obtained from the C/N graph
TPC/(I+N) is the throughput obtained from the C/I+N graph
5.7.3.6 GPRS/EGPRS Studies Based on C/I With ILA and Thermal Noise
5.7.3.6.1 Coding Scheme Based on C/I With ILA and Thermal Noise
The reception thresholds given for signal level C are internally converted to C/N thresholds (where 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. C/I thresholds are
also indexed by the C/(I+N) value.
With Ideal Link Adaptation active, Atoll selects the coding scheme that provides the highest throughput at the received C/
N and C/(I+N) values for a bin.
Atoll calculates the C/N and C/(I+N) level received from Txi on each bin of the Txi coverage area. Then, selects two coding
 
 
For each TRX type, tt, cs C  N = Highest  CS Txi – EGPRS

   C P rec 
 TP = Highest  TP=f  ------------ = -----------------------------------  TRX   
 I + N N 
© Forsk 2009 AT271_TRG_E6 171

 
 
And, cs C   I + N  = Highest  CS Txi – EGPRS

   C P rec 
 TP = Highest  TP=f  ------------ = -----------------------------------  TRX   
  I + N I + N 
Where, csC/N is the coding scheme determined from the C/N level, csC/(I+N) is the coding scheme determined from the C/
(I+N) level, and CS is the set of all available coding schemes.
Txi – EGPRS
  C P rec 
Highest  TP=f  ------------ = --------------------------------  TRX   is the highest throughput provided by any coding scheme at the received C/
  I + N N 
N level, from the Throughput = f(C/(I+N)) graphs converted from the f(C/I) graphs defined in the properties dialog. And,
Txi – EGPRS
  C P rec 
Highest  TP=f  ------------ = --------------------------------  TRX   is the highest throughput provided by any coding scheme at the received C/
 I + N I+N 
(I+N) level, from the Throughput = f(C/(I+N)) graphs converted from the f(C/I) graphs defined in the properties dialog.
If there are more than one coding schemes providing the highest throughput at the bin, csC/N and csC/(I+N) will be the ones
with the highest coding scheme numbers.
The resulting coding scheme, cs, is the coding scheme with the highest coding scheme number from among csC/N and
csC/(I+N).
cs = Highest  cs C  N cs C   I + N  
such that:
For each TRX type, tt, -----------------------------------------------------------------------------------------------------------   ------------ Threshold
N I + N  CS

And -----------------------------------------------------------------------------------------------------------   ------------ Threshold
I+N I + N  CS
5.7.3.6.2 Throughput Based on Interpolation Between C/N and C/(I+N) With ILA
Atoll determines two throughput values, TPC/N from the Throughput = f(C) graph corresponding to the coding scheme csC/
N determined above, and TPC/(I+N) from Throughput = f(C/I) graph corresponding to the coding scheme csC/(I+N) deter-
mined above.
The Throughput = f(C) graph is internally converted to Throughput = f(C/N) graph, in order to be indexed with the C/(I+N)
value. The Throughput = f(C/I) graph is also indexed with the C/(I+N) value. The final throughput is computed by interpo-
lating 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 resulting throughput TP is given by:
TP =   TP C  N +  1 –    TP C   I + N 
Where,
pN
 = ---------------------
pI + N
pN is the thermal noise power (value in Watts)
p(I+N) is the interferences + thermal noise power (value in Watts)
TPC/N is the throughput obtained from the C/N graph
TPC/(I+N) is the throughput obtained from the C/I+N graph
5.7.4 Coverage Display

Coverage area can be displayed with colours depending on:
5.7.4.1 GPRS/EGPRS Coding Schemes Study Display Types

5.7.4.1.1 Coding Schemes
Only the bins with a coding scheme assigned are coloured. The bin colour depends on the assigned coding scheme.
Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many
layers as transmitter coverage areas. Each layer shows the coding schemes available in the transmitter coverage area.
172 AT271_TRG_E6 © Forsk 2009

5.7.4.1.2 Best Coding Schemes

On each bin, Atoll chooses the highest coding scheme available from the TRXs of different transmitters covering that bin.
Only the bins with a coding scheme assigned are coloured. The bin colour depends on the assigned coding scheme.
Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many
layers as possible coding schemes. Each layer shows the areas where a given coding scheme can be used.
5.7.4.2 RLC/MAC and Application Throughput/Timeslot Studies Display Types

5.7.4.2.1 Relation Between RLC/MAC and Application Throughputs
Application throughput per timeslot is deduced from the RLC/MAC (or gross) throughput per timeslot by the equation:
SF
TP Application = TP RLC  MAC  ---------- – TP Offset
100
Where, TP Application = Application throughput,
TP RLC  MAC = RLC/MAC level throughput,
TP Offset = Throughput offset (kbps) accounting for headers, guard-bits etc.,
SF = Throughput scaling factor (%) accounting for coding, redundance etc.

Note:
• The default value of the throughput offset is 0 kbps. If you are getting the same coverage
plots for the RLC/MAC and application level throughputs, try assigning a higher value to
this offset to visualise the difference.
5.7.4.2.2 Throughput/Timeslot
A bin of a coverage area is coloured if the calculated throughput exceeds the defined minimum threshold. The bin colour
depends on throughput. Coverage consists of several independent layers whose visibility in the map window can be
managed. There are as many layers as transmitter coverage areas. Each layer shows the throughputs that a transmitter
can provide on one timeslot.
5.7.4.2.3 Best Throughput/Timeslot

On each bin, Atoll chooses the highest throughput available from the TRXs of different transmitters covering that bin. A
bin of a coverage area is coloured if the best throughput exceeds the defined minimum threshold. The bin colour depends
on throughput. Coverage consists of several independent layers whose visibility in the map window can be managed.
There are as many layers as defined thresholds. Each layer shows the areas where a throughput can be provided on one
timeslot.
5.7.4.2.4 Average Throughput/Timeslot

On each bin, Atoll calculates the average throughput available from the TRXs of different transmitters covering that bin. A
bin of a coverage area is coloured if the average throughput exceeds the defined minimum threshold. The bin colour
depends on throughput. Coverage consists of several independent layers whose visibility in the map window can be
managed. There are as many layers as defined thresholds. Each layer shows the areas where a throughput can be
provided on one timeslot.
5.7.4.2.5 Block Error Rate Computation

TP
Block error rate is computed according to the formula BLER = ------------------
TP MAX
Where, TP = Throughput per timeslot calculated for a bin, and TPMAX = Maximum throughput per timeslot deduced from
the GPRS/EGPRS equipment assigned to the terminal type (or transmitter, if no terminal type has been selected for the
study) for the coding scheme calculated for a bin.
Notes:
• BLER can be displayed for RLC/MAC and application throughput according to the option
selected.
• If TP > TPMAX, then BLER = 0.
5.7.4.2.6 BLER Percentage

Atoll calculates BLER percentage by considering throughput/timeslot per bin (computed as described earlier) and the
maximum throughput/timeslot possible (deduced from the GPRS/EGPRS equipment graphs). A bin of a coverage area is
coloured if the calculated BLER percentage exceeds the defined minimum threshold. The bin colour depends on the
BLER. Coverage consists of several independent layers whose visibility in the map window can be managed. There are
as many layers as transmitter coverage areas. Each layer shows the BLERs that a transmitter experiences on one times-
lot.
© Forsk 2009 AT271_TRG_E6 173

5.7.4.2.7 Maximum BLER Percentage

On each bin, Atoll chooses the maximum BLER percentage from the BLER values corresponding to TRXs of different
transmitters covering that bin. A bin of a coverage area is coloured if the maximum BLER exceeds the defined minimum
threshold. The bin colour depends on the BLER. Coverage consists of several independent layers whose visibility in the
map window can be managed. There are as many layers as defined thresholds. Each layer shows the areas where a BLER
is observed on one timeslot.
5.8 Circuit Quality Indicators Studies

Atoll calculates a coverage area for all the TBC transmitters. Let us assume that each bin within a TBC transmitter calcu-
lation area corresponds to a probe mobile receiver. Coverage study parameters to be set are:
• The study conditions in order to determine the coverage area of each TBC transmitter,
5.8.1 Circuit Quality Indicators

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. These frames are usually discarded, in
which case this can be called the Frame Erasure Rate.
• BER or Bit Error Rate
- BER is a measurement of the raw bit error rate in reception before the decoding process begins. Any factor
that impacts the decoding performance, such as frequency hopping, will impact the correlation between BER
and FER, or the perceived end-user voice quality.
• MOS or Mean Opinion Score
- Voice quality can be quantified using mean opinion score (MOS). MOS values can only be measured in a test
laboratory environment. MOS values range from 1 (bad) to 5 (excellent). Different voice codecs have slightly
different FER to MOS correlation since the smaller the voice codec bit rate is, the more sensitive it becomes
to frame erasures.
There are various codec modes defined for AMR depending on the FR and HR channel modes that it supports. Codec
modes defined in Atoll include the basic EFR, FR and HR codec modes along with the AMR codec modes:
• AMR FR
- AMR TCH/AFS12.2
- AMR TCH/AFS10.2
- AMR TCH/AFS7.95
- AMR TCH/AFS7.4
- AMR TCH/AFS6.7
- AMR TCH/AFS5.9
- AMR TCH/AFS5.15
- AMR TCH/AFS4.75
• AMR HR
- AMR TCH/AHS7.95
- AMR TCH/AHS7.4
- AMR TCH/AHS6.7
- AMR TCH/AHS5.9
- AMR TCH/AHS5.15
- AMR TCH/AHS4.75
• EFR
• FR
• HR
A codec equipment should contain codec mode adaptation thresholds and quality graphs for the relevant circuit quality
indicators in order to be considered in circuit quality indicators coverage studies. The default codec equipment in Atoll
include default FER, BER and MOS quality graphs with respect to the carrier to interference ratio, and codec mode adap-
tation thresholds (computed from the FER vs. C/I graphs for all codec modes at 5% FER.).
Note:
• Depending on the reference CQI, codec mode that provides the lowest BER or FER, or the
highest MOS are selected during computations.
References:
The above graphs are based on:
[1] T. Halonen, J. Romero, J. Melero; GSM, GPRS and EDGE performance – Evolution towards 3G/UMTS, John Wiley
and Sons Ltd.
[2] J. Wigard, P. Mogensen; A simple mapping from C/I to FER and BER for a GSM type of air interface.
[3] 3GPP Specifications TR 26.975 V6.0.0; Performance characterization of the Adaptive Multi-Rate (AMR) speech
codec (Release 6)
174 AT271_TRG_E6 © Forsk 2009

Figure 5.11: FER vs. C/I Graphs
Figure 5.12: BER vs. C/I Graphs
Figure 5.13: MOS vs. C/I Graphs
5.8.2 Coverage Area Determination

We can have four different cases for coverage area determination. Let us assume that:
• Each transmitter, Txi, belongs to a HCS layer, k, with a defined priority.
• Each transmitter, Txi, is has a codec equipment assigned (Txi as notation).
© Forsk 2009 AT271_TRG_E6 175

5.8.2.1 All Servers

For each HCS layer, k, the coverage area of Txi corresponds to Txi calculation area.
5.8.2.2 Best Signal Level per HCS Layer and a Margin

Txi
P rec  BCCH  , is the highest one (Txi is the best server) or within a defined margin of the highest signal level (within a
margin of the best server).
Note:
Txi
5.8.2.3 Second Best Signal Level per HCS Layer and a Margin
Txi
P rec  BCCH  , is the second highest one (Txi is the second best server) or within a defined margin of the second highest
signal level (within a margin of the second best server).
Note:
Txi
P rec  BCCH  is the second highest.
5.8.2.4 HCS Servers and a Margin

Txi
P rec  BCCH  , is the highest one (Txi is the best server) or within a defined margin of the highest signal level (within a
Txi
margin of the best server). In addition P rec  BCCH  must exceed the reception threshold defined per HCS layer.
Notes:
Txi
5.8.2.5 Highest Priority HCS Server and a Margin

Txi
In this case, the coverage area of Txi corresponds to the bins where P rec  BCCH  is the highest one (Txi is the best
server) or within a defined margin of the highest signal level (within a margin of the best server). And, Txi belongs to the
HCS layer with the highest priority. The highest priority is defined by the priority field (0: lowest) assuming the received
Txi
P rec  BCCH  exceeds the reception threshold defined per HCS layer.
Note:
5.8.3 Calculation Options

CQI studies can be based either on the signal-to-noise level (C/N) or on the signal-to-noise-plus-interference ratio (C/
(I+N)).
If a reference terminal type is defined when launching a CQI study, Atoll will consider the noise figure defined for that termi-
nal type when computing the total noise (N). If no terminal type is defined, the value defined in the coverage prediction
properties (8 dB, by default) is used.
Different codec equipment can be defined for a transmitter and for the reference terminal type. In this case, Atoll performs
an intersection for the codec modes defined in the transmitter and the reference terminal type codec equipment to deter-
mine the codec modes that are common in the two. Hence, Atoll creates a temporary codec equipment comprising only
these common codec modes and uses this codec equipment to eventually compute the CQI values.
If no reference terminal type is defined or if the defined reference terminal type has no codec equipment assigned to it,
Atoll will perform the computations based on the codec equipment of the transmitter. Similarly, if a transmitter has no codec
equipment defined, Atoll will compute the coverage study considering only the codec equipment defined for the reference
176 AT271_TRG_E6 © Forsk 2009

terminal type. If there is no codec equipment defined for the transmitter and for the reference terminal type, there will be
no coverage for the transmitter.
For the case where more than one codec modes, compatible with the transmitter - reference terminal type pair, satisfy the
C or C/I conditions at a bin under study, Atoll chooses the codec mode for CQI determination according to their selection
priorities. The table below depicts the selection priorities for all codec modes modeled in Atoll. These selection priorities
are based on the chronological order of their development and on their C/I - MOS performance.
Codec Mode Selection priority

FR 1
HR 2
EFR 3
AMR TCH/AFS4.75 4
AMR TCH/AFS5.15 5
AMR TCH/AFS5.9 6
AMR TCH/AFS6.7 7
AMR TCH/AFS7.4 8
AMR TCH/AFS7.95 9
AMR TCH/AFS10.2 10
AMR TCH/AFS12.2 11
AMR TCH/AHS4.75 12
AMR TCH/AHS5.15 13
AMR TCH/AHS5.9 14
AMR TCH/AHS6.7 15
AMR TCH/AHS7.4 16
AMR TCH/AHS7.95 17
5.8.3.1 Calculations Based on C/N

Txi
In this case, only signal level received and the total noise are taken into account. Atoll evaluates P rec  TRX  , the signal
level received from transmitter Txi on TRXs (TRX) belonging to the selected TRX type (tt), or on all the TRXs, on each bin
of Txi coverage area and converts it into C/N values using the value of total noise computed as follows:
N tot = N + NF
Where, N is the thermal noise (defined in the document database at -121 dBm by default), and NF is the receiver noise
figure (either defined at the terminal type properties level, if a reference terminal type is defined for the study, or defined
in the coverage study properties).
The computed total noise N tot is then compared to the codec equipment reference noise N Ref . If the values are the same,
the defined graphs are used as is, unless the entry is downshifted by the difference N tot – N Ref .
5.8.3.2 Calculations Based on C/(I+N)

For circuit quality indicator studies based on the signal-to-noise-and-interference ratio (C/(I+N)), Atoll evaluates:
Txi
P rec  TRX 
• ----------------------------- : The signal-level-to-noise ratio received from transmitter Txi on TRXs (TRX) belonging to the selected
N tot
TRX type (tt) or on all the TRXs.
And
Txi
P rec  TRX 
• ----------------------------- : The carrier-to-interference-and-noise ratio received from transmitter Txi on TRXs (TRX) belonging
I + N tot
to the selected TRX type (tt) or on all the TRXs.
The value of total noise computed as follows:
N tot = N + NF
Where, N is the thermal noise (defined in the document database at -121 dBm by default), and NF is the receiver noise
figure (either defined at the terminal type properties level, if a terminal type is defined for the study, or defined in the cover-
age study properties).
The computed total noise N tot is then compared to the codec equipment reference noise N Ref . If the values are the same,
the defined graphs are used as is, unless the entry is downshifted by the difference N tot – N Ref .
Note:
© Forsk 2009 AT271_TRG_E6 177

• Atoll calculates the carrier-to-interference ratio for all the TBC transmitters but takes into
account all the transmitters (even the ones with no codec equipment assigned) to evaluate
the interference.
5.8.3.3 Ideal Link Adaptation (ILA)

Ideal link adaptation for circuit quality indicator studies is defined at the codec equipment level. If the ideal link adaptation
option is checked, Atoll will select the codec mode, for the transmitter under study, according to the codec quality graphs
(CQI = f(C/N) and CQI = f(C/I)) related to the defined reference CQI, which may be different from the CQI being calculated.
Otherwise, Atoll will use the adaptation thresholds defined in the Adaptation Thresholds tab to determine the codec mode
to be used in the studies.
5.8.4 Calculation Scenarios

With the above options, there are many different possible scenarios of calculations. These scenarios are described below.
5.8.4.1 CQI Study Based on C/N Without ILA

Atoll calculates signal level received from Txi on each bin of Txi coverage area and converts it into C/N values as described
earlier. Then, Atoll filters all the codec modes that satisfy the C/N values and are common between the transmitter and the
terminal type codec equipment. It then determines the codec mode for the bin, such that:
 
 
For each TRX type, tt, cm = Highest Priority  CM Txi 
 P rec  TRX 
------------------------------   Adaptation Threshold  CM

 N tot 
Where, cm is the codec mode with the highest priority among the set of codec modes CM having their adaptation thresh-
Txi
P rec  TRX 
olds less than the received C/N level, ----------------------------- .
N tot
From the CQI=f(C/N) graph associated to the selected codec mode cm, Atoll evaluates the CQI for which the study was
Txi
P rec  TRX 
performed corresponding to ----------------------------- for the selected codec mode.
N tot
5.8.4.2 CQI Study Based on C/N With ILA

Ideal link adaptation is used by a codec equipment according to a defined reference CQI (MOS by default).
Atoll calculates signal level received from Txi on each bin of Txi coverage area and converts it into C/N values as described
earlier. Then, Atoll filters all the codec modes that satisfy the C/N criterion (defined by the CQI = f(C/N) graphs for the refer-
ence CQI) and are common between the transmitter and the terminal type codec equipment.
The selected codec mode among these filtered codec modes will be,
 
 
For each TRX type, tt, cm = Highest Priority  CM Txi  , for MOS
   C P rec  TRX   
 CQI Ref = Highest  CQI=f  ---- = ------------------------------  
 N N tot 
 
 
Or, cm = Highest Priority  CM Txi  , for BER and FER
   C rec P  TRX  
 CQI Ref = Lowest  CQI=f  ---- = ------------------------------  
 N N tot 
Where, cm is the codec mode with the highest priority among the set of codec modes CM for which the reference CQI
Txi
P rec  TRX 
gives the highest or the lowest value at the received C/N level, ----------------------------- .
N tot
If more than one codec mode graphs give the same value for reference CQI, then Atoll selects the codec mode with the
highest priority.
From the CQI = f(C/N) graph associated to the selected codec mode cm, Atoll evaluates the CQI for which the study was
Txi
P rec  TRX 
performed corresponding to ----------------------------- for the selected codec mode.
N tot
5.8.4.3 CQI Study Based on C/(I+N) Without ILA

Atoll calculates the C/I level received from the transmitter on each bin of Txi coverage area, for each TRX and converts it
into C/(I+N). Then, Atoll filters all the codec modes that satisfy the C/(I+N) values and are common between the transmitter
and the terminal type codec equipment.
It then determines the codec mode for the bin, such that:
178 AT271_TRG_E6 © Forsk 2009

 
 
For each TRX type, tt, cm = Highest Priority  CM Txi 
 P rec  TRX 
------------------------------   Adaptation Threshold  CM

 I + N tot 
Where, cm is the codec mode with the highest priority among the set of codec modes CM having their adaptation thresh-
Txi
P rec  TRX 
olds less than the received C/(I+N) level, ----------------------------- .
I + N tot
From the CQI = f(C/I) graph associated to the selected codec mode cm (indexed with the C/(I+N) values), Atoll evaluates
Txi
P rec  TRX 
the CQI for which the study was performed corresponding to ----------------------------- for the selected codec mode.
I + N tot
5.8.4.4 CQI Study Based on C/(I+N) With ILA

Ideal link adaptation is used by a codec equipment according to a defined reference CQI (MOS by default).
Atoll calculates the C/I level received from the transmitter on each bin of Txi coverage area, for each TRX and converts it
into C/(I+N). Then, 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 equipment.
The selected codec mode among these filtered codec modes will be,
 
 
For each TRX type, tt, cm = Highest Priority  CM Txi  , for MOS
   C rec P  TRX  
 CQI Ref = Highest  CQI=f  ---- = ------------------------------  
 I I + N tot  
 
 
Or, cm = Highest Priority  CM Txi  , for BER and FER
   C P rec  TRX   
 CQI Ref = Lowest  CQI=f  ---- = ------------------------------  
  I I + N tot  
Where, cm is the codec mode with the highest priority among the set of codec modes CM for which the reference CQI
Txi
P rec  TRX 
gives the highest or the lowest value at the received C/(I+N) level, ----------------------------- .
I + N tot
If more than one codec mode graphs give the same value for reference CQI, then Atoll selects the codec mode with the
highest priority.
From the CQI = f(C/I) graph associated to the selected codec mode cm (indexed with the C/(I+N) values), Atoll evaluates
Txi
P rec  TRX 
the CQI for which the study was performed corresponding to ----------------------------- for the selected codec mode.
I + N tot
5.8.5 Coverage Display

Coverage area can be displayed with colours depending on:
5.8.5.1 Circuit Quality Indicators Study Display Types

5.8.5.1.1 FER/BER/MOS
Only the bins with a CQI assigned are coloured. The bin colour depends on the assigned CQI value. Coverage consists
of several independent layers whose visibility in the map window can be managed. There are as many layers as transmitter
coverage areas. Each layer shows the CQI values available in the transmitter coverage area.
5.8.5.1.2 Max FER/Max BER/Max MOS

On each bin, Atoll chooses the maximum CQI value available from the TRXs of different transmitters covering that bin.
Only the bins where the CQI values exceeds a defined threshold are coloured. The bin colour depends on the assigned
CQI value. Coverage consists of several independent layers whose visibility in the map window can be managed. There
are as many layers as the number of thresholds defined. Each layer shows the areas where a given CQI value is available.
© Forsk 2009 AT271_TRG_E6 179

180 AT271_TRG_E6 © Forsk 2009

Chapter 6
UMTS HSPA Networks
This chapter provides descriptions of all the algorithms for calculations, analyses, automatic allocations,
simulations and prediction studies available in UMTS HSPA projects.
Atoll
182 AT271_TRG_E6 © Forsk 2009

Chapter 6: UMTS HSPA Networks
6 UMTS HSPA Networks

Three criteria can be studied in point analysis (Profile tab) and in common coverage studies. Study criteria are detailed in
the table below:

Signal level received from a transmitter on a carrier (cell)
Signal level ( P rec ) in dBm
P rec  ic  = EIRP  ic  – L path – M Shadowing – model – L Indoor + G term – L term
Path loss ( L path ) in dBm L path = L model + L ant

Tx
Total losses ( L total ) in dBm L total =  L path + L Tx + L term + L indoor + M Shadowing – model  –  G Tx + G term 
where,
ic is a carrier number,

Tx
L Indoor are the indoor losses. These losses are defined for each clutter class.They are taken into account when the option
L term are the receiver losses,
G term is the receiver antenna gain,
G Tx is the transmitter antenna gain,
L Tx is the transmitter loss ( L Tx = L total – DL ). For information on calculating transmitter loss, see "UMTS HSPA,
CDMA2000 1xRTT 1xEV-DO, IS-95 cdmaOne, and TD-SCDMA Documents" on page 124.
Notes:
• EIRP  ic  = P pilot  ic  + G Tx – L Tx ( P pilot  ic  is the cell pilot power).
• It is possible to analyse all the carriers. In this case, Atoll takes the highest pilot power of
cells to calculate the signal level received from a transmitter.
• Atoll considers that G term and L term equal zero.

6.1.2.1 Profile Tab
Atoll displays either the signal level received from the selected transmitter on a carrier ( P rec  ic  ), or the highest signal
level received from the selected transmitter on all the carriers.
Note:
losses, L total . Path loss and total losses are the same on any carrier.

© Forsk 2009 AT271_TRG_E6 183

For each transmitter, Atoll displays either the signal level received on a carrier, ( P rec  ic  ), or the highest signal level
received on all the carriers.
Note:

For each TBC transmitter, Txi, Atoll determines the selected criterion on each pixel inside the Txi calculation area. In fact,
each pixel within the Txi calculation area is considered as a potential (fixed or mobile) receiver.

We can distinguish three cases:

Txi Txi Txi

Txi Txi Txi
And

Txi Txj
ji

Notes:
the highest.

Txi Txi Txi
And
Txi nd
P rec  ic   2 Best  P Txj  ic   – M
rec
ji
184 AT271_TRG_E6 © Forsk 2009

Notes:
the second highest.


such as:
Atoll calculates signal level received from the transmitter on each pixel of each transmitter service area. A pixel of a service
area is coloured if the signal level exceeds (  ) the defined minimum thresholds (pixel colour depends on signal level).
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where other service
areas overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the signal level
exceeds (  ) the defined thresholds (the pixel colour depends on the signal level). Coverage consists of several independ-
ent layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer
corresponds to an area where the signal level from the best server exceeds a defined minimum threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if path loss exceeds (  ) the defined minimum thresholds (pixel colour depends on path loss). Coverage consists
Total Losses (dB)
Atoll calculates total losses from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if total losses exceed (  ) the defined minimum thresholds (pixel colour depends on total losses). Coverage
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where other service
pixel of a service area is coloured if the path loss exceeds (  ) the defined thresholds (pixel colour depends on path loss).
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where service
areas overlap the studied one, Atoll determines the best transmitter and evaluates total losses from the best transmitter.
A pixel of a service area is coloured if the total losses exceed (  ) the defined 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 defined thresholds. Each layer corresponds to an area where the total losses from the best server
exceed a defined minimum threshold.
© Forsk 2009 AT271_TRG_E6 185

Number of Servers
Atoll evaluates how many service areas cover a pixel in order to determine the number of servers. The pixel colour
depends on the number of servers. Coverage consists of several independent layers whose visibility in the workspace can
be managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the number of
servers exceeds (  ) a defined minimum threshold.
On each pixel of each transmitter service area, 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. There is one
coverage area per transmitter in the explorer.
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the best signal level received
explorer.
6.2 Definitions and Formulas

Input parameters and formulas used in simulations and predictions (coverage predictions and point analysis) are detailed
in the tables below.
6.2.1 Inputs
This table lists simulation and prediction inputs (calculation options, quality targets, active set management conditions,
etc.).
Name Value Unit Description
F ortho Clutter parameter or global parameter (default value) None Orthogonality factor
Tx
F MUD Site equipment parameter None MUD factor
Term
F MUD Terminal parameter - HSDPA properties None MUD factor
ic Frequency band parameter None Carrier number

Threshold for macro diversity
AS_Th  Txi ic  Cell parameter None specified for a transmitter on a given
carrier ic
req
Q pilot E
-----c- Mobility parameter None
Ec/Io target on downlink for the best
 I 0  threshold server
CM – activation Pilot RSCP threshold for compressed

RSCP pilot Global parameter None
mode activation
CM – activation Ec/Io threshold for compressed mode
Q pilot Global parameter None
activation
E
DL
DL
Q req -----b- (Bearer, Mobility) parameter None Eb/Nt target on downlink
 N t  req
DL Downlink Eb/Nt target increase due to

Q req Global parameter None
compressed mode activation
E
UL
UL
Q req -----b- (Bearer, Mobility) parameter None Eb/Nt target on uplink
 N t  req
UL Uplink Eb/Nt target increase due to

Q req Global parameter None
compressed mode activation
CE – UL Number of channel elements available
N max  NI  Site parameter None
for a site on uplink
CE – DL Number of channel elements available
for a site on downlink
CE – UL Number of channel elements of a site
N  NI  Simulation result None
consumed by users on uplink
CE – DL Number of channel elements of a site
consumed by users on downlink
186 AT271_TRG_E6 © Forsk 2009

Number of channel elements used by

Overhead – C E – UL Site equipment parameter None the cell for common channels on
N
uplink
Overhead – C E – DL Site equipment parameter None the cell for common channels on
N
downlink
TCH – C E – UL Number of channel elements used for

N (Bearer, site equipment) parameter None
traffic channels on uplink
TCH – C E – DL Number of channel elements used for

N (Bearer, site equipment) parameter None
traffic channels on downlink
Codes Maximum number of 512 bit-length
N max  Txi ic  Simulation constraint None
OVSF codes available per cell (512)
Codes Number of 512 bit-length OVSF codes

N  Txi ic  Simulation result None
used by the cell
Overhead – C odes Overhead – C E – DL Number of 256 bit-length OVSF codes

N Site equipment parameter - Corresponds to N None
used by the cell for common channels
Maximum number of 16 bit-length
Codes – HS PDSCH
N max  Txi ic  Cell parameter (for HSDPA only) None OVSF codes available per cell for HS-
PDSCH
Minimum number of 16 bit-length
Codes – HS PDSCH
N min  Txi ic  Cell parameter (for HSDPA only) None OVSF codes available per cell for HS-
PDSCH
NF term Terminal parameter None Terminal Noise Figure
Transmitter parameter (user-defined or calculated from transmitter

NF Tx None Transmitter Noise Figure
equipment characteristics)
K 1.38 10-23 J/K Boltzman constant
T 293 K Ambient temperature
W 3.84 MHz Hz Spreading Bandwidth
Network parameter Interference reduction factor between

RF  ic ic adj  None
If not defined, it is assumed that there is no inter-carrier interference two adjacent carriers ic and ic adj
UL
X max Simulation constraint (global parameter or cell parameter) % Maximum uplink load factor
DL
%Power max Simulation constraint (global parameter or cell parameter) % Maximum percentage of used power
Tx
N0 NF Tx  K  T  W W Thermal noise at transmitter
Term NF Term  K  T  W
N0 W Thermal noise at terminal
Rc –3 bps Chip rate

W  10 W
UL
f rake efficiency Equipment parameter None Uplink rake receiver efficiency factor
DL Downlink rake receiver efficiency

f rake efficiency Terminal parameter None
factor
DL
R nominal Bearer parameter kbps Bearer downlink nominal bit rate
DL Downlink spreading factor for active

F spreading  Active user  Bearer parameter None
users
DL Downlink spreading factor for inactive
F spreading  Inactive user  Bearer parameter None
users
ratio between DPCCH and DPCH
transmission duration on downlink
DL DPCCH and DPCH respectively refer
rc Bearer parameter None
to the Dedicated Physical Control
Channel and Dedicated Physical
Channel
DL Maximum connection rate per user on
R max Cell parameter kbps
downlink
UL
R nominal Bearer parameter kbps Bearer uplink nominal bit rate
UL Uplink activity factor on E-DPCCH

f act – ADPCH Service parameter kbps
channels
© Forsk 2009 AT271_TRG_E6 187

DL Downlink Activity factor on A-DPCH

f act – ADPCH Service parameter kbps
channel
ratio between the DPCCH and DPCH
powers transmitted on uplink
UL DPCCH and DPCH respectively refer
rc Bearer parameter None
to the Dedicated Physical Control
Channel and Dedicated Physical
Channel
UL Maximum connection rate per user on
R max Cell parameter kbps
uplink
DL
W -
---------------------
Gp DL None Service downlink processing gain
R nominal
W
Gp
UL ---------------------- None Service uplink processing gain
UL
R nominal
DL User application throughput on

T application HSDPA study result kbps
downlink
DL RLC peak rate supported by the
R RLC – peak HSDPA study result - HSDPA Bearer parameter kbps
HSDPA bearer
DL RLC peak throughput supported by
T RLC – peak HSDPA study result kbps
the HSDPA bearer
DL Average RLC throughput supported
T RLC – Av HSDPA study result kbps
by the HSDPA bearer
DL MAC rate supported by the HSDPA
R MAC HSDPA study result kbps
bearer
DL MAC throughput supported by the
T MAC HSDPA study result kbps
HSDPA bearer
UL
T application HSUPA study result kbps User application throughput on uplink
UL RLC peak rate supported by the

R RLC – peak HSUPA study result - HSUPA Bearer parameter kbps
HSUPA bearer
UL Minimum RLC throughput supported
T RLC – Min HSUPA study result kbps
by the HSUPA bearer
UL MAC rate supported by the HSUPA
R MAC HSUPA study result kbps
bearer
R Service parameter (for HSDPA only) kbps Throughput offset
SF Rate Service parameter (for HSDPA only) % Scaling factor
Maximum shared power

P max  Txi  Transmitter parameter W Available only if the inter-carrier power
sharing option is activated
P SCH  Txi ic  Cell parameter W Cell synchronisation channel power
Cell other common channels (except

P OtherCCH  Txi ic  Cell parameter W
CPICH and SCH) powera
P pilot  Txi ic  Cell parameter W Cell pilot power
Cell parameter (user-defined or simulation result) (for HSDPA only) Available cell HSDPA power
P HSDPA  Txi ic  W HSDPA: High Speed Downlink Packet
P HS – PDSCH  Txi ic  + n HS – SCCH  P HS – SCCH  Txi ic 
Access
Cell HS-PDSCH power
P HS – PDSCH  Txi ic  Simulation result (for HSDPA only) W HS-PDSCH: High Speed Physical
Downlink Shared Channel
Cell HS-SCCH power
P HS – SCCH  Txi ic  Cell parameter (for HSDPA only) W HS-SCCH: High Speed Shared
Control Channel
number of HS-SCCH channels
n HS – SCCH Cell parameter (user-defined or simulation result) (for HSDPA only)
managed by the cell
P Headroom  Txi ic  Cell parameter (for HSDPA only) W Cell headroom power
P max  Txi ic  Cell parameter W Maximum Cell power
R99 traffic channel power transmitted

P tch  Txi ic  Simulation result W
on carrier ic
188 AT271_TRG_E6 © Forsk 2009

min Minimum power allowed on traffic data

P tch Bearer parameter W
channel
max Maximum power allowed on traffic
P tch Bearer parameter W
data channel
Cell HSUPA power
P HSUPA  Txi ic  Cell parameter W HSUPA: High Speed Uplink Packet
Access
Cell parameter or simulation result
P pilot  Txi ic  + P SCH  Txi ic  + P OtherCCH  Txi ic  +
Transmitter R99 transmitted power on
P tx – R99  Txi ic  W
 
DL
P tch  Txi ic  + P tch  Txi ic   f act –ADPCH carrier ic
tch(ic) used for tch(ic) used for
R99 users HSUPA users
Simulation result Transmitter total transmitted power on

P tx  Txi ic  W
P tx – R99  Txi ic  + P HSDPA  Txi ic  + P HSUPA  Txi ic  carrier ic
Terminal power transmitted to obtain

P term – R99 Simulation result W
the R99 radio bearer
Terminal power transmitted to obtain
P term – HSUPA Simulation result W
the HSUPA radio bearer
Simulation result
UL
P term – R99  f act –ADPCH + P term – HSUPA for HSPA users Total power transmitted by the
P term W
terminal
P term – R99 for R99 users
min
P term Terminal parameter W Minimum terminal power allowed
max
P term Terminal parameter W Maximum terminal power allowed
Percentage of BTS signal correctly

 BTS BTS parameter %
transmitted
Percentage of terminal signal correctly
 term Terminal parameter %
transmitted
Percentage of pilot finger - percentage
 Clutter parameter % of signal received by the terminal pilot
finger
G Tx Antenna parameter None Transmitter antenna gain
G Term Terminal parameter None Terminal gain
DL
G Div Bearer parameter - Depends on the transmitter Tx diversity None Gain due to transmit diversity
UL
G Div Bearer parameter - Depends on the transmitter Rx diversity None Gain due to receive diversity

L Tx None Transmitter lossb
L body Service parameter None Body loss
L Term Terminal parameter None Terminal loss
L indoor Clutter parameter Indoor loss
L path Propagation model result None Path loss
Result calculated from cell edge coverage probability and model Model Shadowing margin
M Shadowing – model None
standard deviation Only used in prediction studies
Result calculated from cell edge coverage probability and Ec/Io Ec/Io Shadowing margin
M Shadowing – Ec  Io None
DL npaths DL gain due to availability of several
DL G macro – diversity = M Shadowing – Ec  Io – M Shadowing –Ec  Io
G macro – diversity None
n=2 or 3 pilot signals at the mobile c.
M Shadowing –  Eb  Nt  Result calculated from cell edge coverage probability and DL Eb/Nt DL Eb/Nt Shadowing margin
None
DL standard deviation Only used in prediction studies
M Shadowing –  Eb  Nt  Result calculated from cell edge coverage probability and UL Eb/Nt UL Eb/Nt Shadowing margin
None
UL standard deviation Only used in prediction studies
© Forsk 2009 AT271_TRG_E6 189

UL npaths
UL UL UL UL quality gain due to signal diversity
n=2 or 3 in soft handoffd.
Global parameter (default value)
Random shadowing error drawn
E Shadowing Simulation result None during Monte-Carlo simulation
Only used in simulations
In prediction studiese
For Ec/Io 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
LT ------------------------------------------------------------------------------------------------------------------------------------------------------------
- None Transmitter-terminal total loss
G Tx  G term
For UL Eb/Nt calculation
UL
------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
In simulations
L path  L Tx  L term  L body  L indoor  E Shadowing
--------------------------------------------------------------------------------------------------------------------------------
G Tx  G term
P pilot  Txi ic 
P c  Txi ic  ----------------------------------- W Chip power received at terminal
LT
DL P tch  Txi ic  Bit power received at terminal on

P b  Txi ic  -------------------------------- W
LT carrier ic
DL P tx  Txi ic  Total power received at terminal from

P tot  Txi ic  ------------------------------ W
LT a transmitter on carrier ic
P tch  Txi ic  Total power received at terminal from

DL
P traf  Txi ic   --------------------------------
LT W traffic channels of a transmitter on
tch  ic  carrier ic
UL P term Bit power received at transmitter on

P b  ic  -------------- W
LT carrier ic used by terminal
UL
P term – R99 Bit power received at transmitter on
P b – R99  ic  --------------------------- W
LT carrier ic used by terminal
UL UL UL Bit power received at transmitter on

P b – DPDCH  ic  P b – R99  ic    1 – r c  W
DPDCH from a terminal on carrier ic
a. For the calculation of interference, P OtherCCH  Txi ic  also includes the MBMS SCCPCH channel power
when the optional MBMS feature is activated. You must modify the data structure for activating the optional MBMS feature.
For more information, see the Administrator Manual.
b. L Tx = L total – UL on uplink and L Tx = L total – DL on downlink. For information on calculating transmitter
losses on uplink and downlink, see "UMTS HSPA, CDMA2000 1xRTT 1xEV-DO, IS-95 cdmaOne, and TD-SCDMA
Documents" on page 124.
npaths
c. M Shadowing – Ec  Io corresponds to the shadowing margin evaluated from the shadowing error probability
density function (n paths) in case of downlink Ec/Io modelling.
npaths
d. M Shadowing –  Eb  Nt  corresponds to the shadowing margin evaluated from the shadowing error probability
UL
density function (n paths) in case of uplink soft handoff modelling.

e. In uplink prediction studies, only carrier power level is downgraded by the shadowing margin
( M Shadowing –  Eb  Nt  ). In downlink prediction studies, carrier power level and intra-cell interference are downgraded by
UL
the shadowing model ( M Shadowing –  Eb  Nt  or M Shadowing – Ec  Io ) while extra-cell interference level is not. Therefore,
DL
M Shadowing –  Eb  Nt  or M Shadowing – Ec  Io is set to 1 in downlink extra-cell interference calculation.

DL
6.2.2 Ec/Io Calculation

This table details the pilot quality ( Q pilot or Ec  Io ) calculations.
190 AT271_TRG_E6 © Forsk 2009

DL  DL P SCH  txi ic  Downlink intra-cell interference at

P tot  txi ic  –  BTS     P tot  txi ic  – ---------------------------------
I intra  txi ic  DL - W
L  terminal on carrier ic
T
Downlink extra-cell interference at


DL
DL
I extra  ic  P tot  txj ic  W
terminal on carrier ic
txj j  i
 Ptot  txj icadj 

DL
DL Downlink inter-carrier interference at
I inter – carrier  ic   j
W
txj
------------------------------------------------
- terminal on carrier ic
RF  ic ic adj 
DL DL DL Term
I intra  txi ic  + I extra  ic  + I inter – carrier  ic  + N 0
Without Pilot:
DL –  1 –     BTS  P c  txi ic  Total received noise at terminal on
I 0  ic  W
carrier ic
DL DL Term
Total noise: P DL  txi ic  + I extra  ic  + I inter – carrier  ic  + N 0
tot
Ec  BTS    P c  txi ic  Quality level at terminal on pilot for

Q pilot  txi ic    ------ ------------------------------------------------------- None
 I0  DL
I 0  ic  carrier ic
6.2.3 DL Eb/Nt Calculation

Eb
This table details calculations of downlink traffic channel quality ( Q tch or  ------- ). When the optional MBMS feature is
DL
 Nt  DL
activated, the MBMS Eb/Nt is also calculated in the same manner. You must modify the data structure for activating the
optional MBMS feature. For more information, see the Administrator Manual.
DL  DL P SCH  txi ic  Downlink intra-cell interference at

P tot  txi ic  –  BTS  F ortho   P tot  txi ic  – ---------------------------------
I intra  txi ic  DL - W
L  terminal on carrier ic
T


DL
DL
txj j  i

DL
W
txj
------------------------------------------------
DL DL DL DL Term Total received noise at terminal on

N tot  ic  I intra  ic  + I extra  ic  + I inter – carrier  ic  + N 0 W
carrier ic
Without useful signal:
DL
 BTS  P b  txi ic 
Eb -  G DL
--------------------------------------------------------------------------------------------------------------- DL
Div  G p
Quality level at terminal on a traffic
DL DL
Q tch  txi ic    ------
DL N tot  ic  –  1 – F ortho    BTS  P b  txi ic  channel from one transmitter on
 N t  DL None
DL carrier ic a
 BTS  P b  txi ic 
-  G DL
Total Noise: ------------------------------------------------
DL
Div  G p
DL
N tot  ic 
Quality level at terminal using carrier

DL DL
DL f rake efficiency  Q tch  tx k ic  ic due to combination of all
Q  ic  None
txk  ActiveSet transmitters of the active set (Macro-
diversity conditions).
DL
DL Q  ic 
---------------------------------------------------------
-
G SHO DL None Soft handover gain on downlink
Q tch  BestServer ic 
DL Required transmitter traffic channel

req Q req
P tch  txi ic  -  P tch  txi ic 
-------------------- W power to achieve Eb/Nt target at
DL
Q  ic  terminal on carrier ic
a. Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account
only in simulations. In point analysis and coverage studies, Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt.
© Forsk 2009 AT271_TRG_E6 191

6.2.4 UL Eb/Nt Calculation

Eb
This table details calculations of uplink traffic channel quality ( Q tch or  ------- ).
UL
 Nt  UL
Total power received at transmitter

 Pb
UL
UL intra  ic 
I tot  txi ic  W from intra-cell terminals using carrier
term
ic
txi


UL
UL extra P b  ic 
I tot  txi ic  W from extra-cell terminals using carrier
term
ic
txj j  i
 Pb
UL
 ic adj 
UL Uplink inter-carrier interference at
I inter – carrier  txi ic  term W
txj j
--------------------------------------
-
UL UL extra UL intra
Total received interference at
I tot  txi ic  I tot
Tx
 txi ic  +  1 – F MUD   term  I tot
UL
 txi ic + I inter – carrier  txi ic  W
transmitter on carrier ic
UL UL tx Total noise at transmitter on carrier ic

N tot  txi ic  I tot  txi ic  + N 0 W
(Uplink interference)
UL
 term  P b – DPDCH  ic 
-  G UL
--------------------------------------------------------------------------------------------------------------- UL
Div  G p
Eb UL Tx
N tot  txi ic  –  1 – F MUD    term  P b  ic 
UL Quality level at transmitter on a traffic
Q tch  txi ic    ------
UL
 N t  UL None
channel for carrier ic a
UL
 term  P b – DPDCH  ic 
-  G UL
Total noise: ---------------------------------------------------------
UL
Div  G p
UL
N tot  txi ic 
UL
No HO: Q tch  txi ic 

UL UL
Softer HO: f rake efficiency  Q tch  tx k ic 
tx k  ActiveSet
 samesite 
Quality level at site using carrier ic due
Soft, softer/soft HO (No MRC): to combination of all transmitters of
UL UL
the active set located at the same site
Max  Q tch  tx k ic    G macro – diversity and taking into account increasing of
UL tx  ActiveSet None
Q  ic  k the quality due to macro-diversity
Softer/soft HO (MRC): (macro-diversity gain).
 
Max  UL  UL

UL UL
 rake efficiency
f  Q tch  tx k ic   Q tch  tx l  ic  In simulations G macro – diversity = 1 .
tx ,tx  ActiveSet
k l  
txk  samesite  tx k 
tx l  othersite
UL
 G macro – diversity
UL
UL Q  ic 
G SHO ---------------------------------------------------------- None Soft handover gain on uplink
UL
Q tch  BestServer ic 
UL Required terminal power to achieve

req Q req
P term  ic  -  P term
-------------------- W Eb/Nt target at transmitter on carrier
UL
Q  ic  ic
6.3 Active Set Management

The mobile’s active set (AS) is the list of the transmitters to which the mobile is connected. The active set may consist of
one or more transmitters; depending on whether the service supports soft handover and on the terminal active set size.
The terminal frequency bands are taken into account and transmitters in the mobile’s active set must use a frequency band
supported by the terminal. Finally, the quality of the pilot (Ec⁄Io) is what determines whether or not a transmitter can belong
192 AT271_TRG_E6 © Forsk 2009

to the active set. The active set management is detailed hereafter. Cells entering a mobile’s active set must satisfy the
following conditions:
• The best server (first cell entering active set)
The pilot quality from the best serving cell must exceed the Ec/Io threshold. Best server cell is the one with the
highest pilot quality.
• Other cells in the active set

- Must use the same carrier as the best server,
- The pilot quality difference between other candidate cells and the best server must be less than the AS
threshold specified for the best server,
- Other candidate cells must belong to the neighbour list of the best server if it is located on a site where the
equipment imposes this restriction (the “restricted to neighbours” option selected in the equipment properties).
6.4 Simulations
The simulation process consists of two steps:
1. Obtaining a realistic user distribution
Atoll generates a user distribution using a Monte-Carlo algorithm, which requires traffic maps and data as input. The result-
ing user distribution complies with the traffic database and maps provided to the algorithm.
Each user is assigned a service, a mobility type, and an activity status by random trial, 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. A user may be either active or inactive. Both active and inactive users
consume radio resources and create interference.
Then, Atoll randomly assigns a shadowing error to each user using the probability distribution that describes the shadow-
ing effect.
Finally, 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. Power control simulation
6.4.1 Generating a Realistic User Distribution

During the simulation, a first random trial is performed to determine the number of users and their activity status. Four activ-
ity 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
The determination of the number of users and the activity status allocation depend on the type of traffic cartography used.
Note:
• Atoll follows a Poisson distribution to determine the total number of users attempting a
connection in each simulation. In order for Atoll to use a constant total number of users
attempting a connection, the following lines must be added to the Atoll.ini file:
[CDMA]
RandomTotalUsers=0
6.4.1.1 Simulations Based on Raster Traffic and Vector Traffic Maps

Raster traffic maps are based on environments. 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).
Vector traffic maps are directly based on user profiles. Each polygon and line of the map is assigned a density of subscrib-
ers with given user profile and mobility type. If the map is composed of points, each point is assigned a number of subscrib-
ers with given user profile and mobility type.
The user profile models the behaviour of the different subscriber categories. Each user profile contains a list of services
and their associated parameters describing how these services are accessed by the subscriber.
From environment (or polygon) surface (S) and user profile density (D), a number of subscribers (X) per user profile is
inferred.
X = SD
Notes:
© Forsk 2009 AT271_TRG_E6 193

• In case of vector traffic map composed of lines, 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 the vector traffic map is composed of
points.
For each behaviour described in a user profile, according to the service, frequency use and exchange volume, Atoll calcu-
lates the probability for the user being active in uplink and in downlink at an instant t.
6.4.1.1.1 Circuit Switched Service (i)

User profile parameters for circuit switched services are:
• The used terminal (equipment used for the service (from the Terminals table)),
• The average number of calls per hour N call ,
• The average duration of a call (seconds) d .
The number of users and their distribution per activity status is determined as follows:
1. Calculation of the service usage duration per hour ( p 0 : probability of a connection):
N call  d
p o = ---------------------
-
3600
2. 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.
3. Calculation of activity probabilities:
UL DL
Probability of being inactive on UL and DL: p inactive =  1 – f act    1 – f act 
UL DL
Probability of being active on UL only: p UL = f act   1 – f act 
DL UL
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, f act and f act are respectively the UL and DL activity factors defined for the circuit switched service i.
4. Calculation of number of users per activity status:

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, a user when he is connected can have four different activity status: either active on both links, or inactive on
both links, or active on UL only, or active on DL only.
6.4.1.1.2 Packet Switched Service (j)

User profile parameters for packet switched services are:
• The used terminal (equipment used for the service (from the Terminals table)),
• 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.
A packet session consists of several packet calls separated by a reading time. 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.
In Atoll, a packet session is described by following parameters:
UL
N packet –c all : Average number of packet calls on the uplink during a session,
DL
N packet –c all : Average number of packet calls on the downlink during a session,
UL
T packet – call : Average time (millisecond) between two packets calls on the uplink ,
DL
T packet – call : Average time (millisecond) between two packets calls on the downlink ,
194 AT271_TRG_E6 © Forsk 2009

UL
T packet : Average time (millisecond) between two packets on the uplink ,
DL
T packet : Average time (millisecond) between two packets on the downlink ,
UL
S packet : Packet size (Bytes) on uplink,
DL
S packet : Packet size (Bytes) on downlink.
Figure 6.1: Description of a Packet Session
The number of users and their distribution per activity status is determined as follows:
1. Calculation of the average packet call size (kBytes):
UL V UL DL V DL
- and S packet – c all = -------------------------------------------
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.
Note:
UL DL
• In case of HSDPA and HSUPA services, f eff and f eff are set to 1.
2. Calculation of the average number of packets per packet call:

UL DL
UL  S packet –c all   S packet –c all 
- + 1 and N DL
N packet = int  ----------------------------------- packet = int  -----------------------------------
- + 1
UL
 S packet  1024  S DLpacket  1024

Note:
• 1kBytes = 1024Bytes.
3. Calculation of the average duration of inactivity within a packet call (s):

UL UL DL DL
UL  N packet – 1   T packet DL  N packet – 1   T packet
 D Inactivity  packet – call = --------------------------------------------------------------
- and  D Inactivity  packet – call = --------------------------------------------------------------
-
1000 1000
4. 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
5. Calculation of the average duration of activity in a session (s):

UL UL DL DL
UL UL N packet  S packet  8 N packet  S packet  8
- and  D DL
 D Activity  session = N packet –c all  -----------------------------------------------------
DL
Activity  session = N packet –c all  -----------------------------------------------------
-
UL DL
R average  1000 R average  1000
UL DL
Where R average and R average are the uplink and downlink average requested rates defined for the service j.
Therefore, 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
6. Calculation of the service usage duration per hour (probability of a connection):
UL N sess UL DL N sess DL
p Connection = --------------  D Connection and p Connection = --------------  D Connection
3600 3600
© Forsk 2009 AT271_TRG_E6 195

7. Calculation of the probability of being connected:

UL DL
p Connected = 1 –  1 – p Connection    1 – p Connection 
Therefore, 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, we have to consider three possible cases when a user is connected:
• 1st case: At a given time, packets are downloaded and uploaded.

In this case, the probability of being connected is:
UL DL
UL + DL p Connection  p Connection
p Connected = ----------------------------------------------------------------
-
p Connected
• 2nd case: At a given time, packet are uploaded (no packet is downloaded).
Here, the probability of being connected is:
UL DL
UL p Connection   1 – p Connection 
p Connected = ------------------------------------------------------------------------------
-
p Connected
• 3rd case: At a given time, packet are downloaded (no packet is uploaded).
In this case, the probability of being connected is:
DL UL
DL p Connection   1 – p Connection 
p Connected = ------------------------------------------------------------------------------
-
p Connected
Now, we have to take into account activity periods during the connection in order to determine the activity status of each
user.
8. Calculation of the probability of being active:
UL DL
UL  D Activity  session DL  D Activity  session
f - and f
= ----------------------------------------------------------------------------------------------------- = -----------------------------------------------------------------------------------------------------
-
UL UL DL DL
  D Inactivity  session +  D Activity  session    D Inactivity  session +  D Activity  session 
Therefore, we have:
• 1st case: At a given time, packets are downloaded and uploaded.

The user can be active on UL and inactive on DL; this probability is:
1 UL DL UL + DL
p UL = f  1 – f   p Connected
The user can be active on DL and inactive on UL; this probability is:
1 DL UL UL + DL
p DL = f  1 – f   p Connected
The user can be active on both links; this probability is:

1 UL DL UL + DL
p UL + DL = f f  p Connected
The user can be inactive on both links; this probability is:

1 UL DL UL + DL
p inactive =  1 – f   1 – f   p Connected
• 2nd case: At a given time, packet are uploaded (no packet is downloaded).
The user can be active on UL and inactive on DL; this probability is:
2 UL UL
p UL = f  p Connected

2 UL UL
p inactive =  1 – f   p Connected
• 3rd case: At a given time, packet are downloaded (no packet is uploaded).
The user can be active on DL and inactive on UL; this probability is:
3 DL DL
p DL = f  p Connected

3 DL DL
p inactive =  1 – f   p Connected
9. Calculation of number of users per activity status

inactive 1 2 3
Number of inactive users on UL and DL: n j = n j   p inactive + p inactive + p inactive 
196 AT271_TRG_E6 © Forsk 2009

1 2
Number of users active on UL and inactive on DL: n j  UL  = n j   p UL + p UL 
1 3
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
Note:
• The user distribution per service and the activity status distribution between the users are
average distributions. And the service and the activity status of each user are randomly
drawn in each simulation. Therefore, if you compute several simulations at once, the
average number of users per service and average numbers of inactive, active on UL, active
on DL and active on UL and DL users, respectively, will correspond to calculated
distributions. But if you check each simulation, the user distribution between services as
well as the activity status distribution between users is different in each of them.
6.4.1.2 Simulations Based on Traffic Map per Service and per Transmitter
Traffic map per service and per transmitter is based on live traffic data from OMC (Operation and Maintenance Centre).
Live traffic is spread over the best pilot coverage area of each transmitter and each coverage area is assigned either the
total throughput demand or the number of active users or Erlangs.
For each transmitter Txi and each service:
• If you are creating a Map based on Transmitters and Services (Throughputs), enter the throughput demands
in the uplink and downlink for each sector and for each listed service. Atoll calculates the number of active users
on UL and DL in the Txi cell using the service (NUL and NDL) as follows:
UL DL
Rt Rt
- N UL = ---------------------- and N DL = ----------------------
UL DL
R average R average
UL
Rt is the kbits per second transmitted on UL in the Txi cell to supply the service.
DL
Rt is the kbits per second transmitted on DL in the Txi cell to supply the service.
DL
R average is the downlink average requested rate defined for the service,
UL
. R average is the uplink average requested rate defined for the service.
• If you are creating a Map based on Transmitters and Services (# Active Users), enter the active users in the
uplink and downlink for each sector and for each listed service (NUL and NDL).
For each transmitter Txi and each circuit-switched service:

• If you are creating a Map based on Transmitters and Services (Erlangs), enter Erlangs, i.e., the total number
of users, active and inactive, for each sector and for each circuit-switched service (n).

There can be two cases for circuit-switched service:
1. Number of active users on uplink and downlink (NUL and NDL),
2. Erlangs, i.e, a total number of users (n).
Active Users on Uplink and Downlink
NUL and NDL values include:
• Users active on uplink and inactive on downlink (ni(UL)),

• Users active on downlink and inactive on uplink (ni(DL)),
• And users active on both links (ni(UL+DL)).
They are calculated as follows:

Calculation of activity probabilities
UL DL
DL UL
UL DL
© Forsk 2009 AT271_TRG_E6 197

UL DL
Probability of being inactive both on UL and DL: p inactive =  1 – f act    1 – f act 
UL DL
active
Calculation of the total number of active users, n i , attempting to access the circuit switched service i
We have:
active
 p UL + p UL + DL   n i = N UL
active
 p DL + p UL + DL   n i = N DL
Therefore, we have:
N UL  p UL + DL N DL  p UL + DL
n i  UL + DL  = min  ------------------------------------- -------------------------------------
 P UL + p UL + DL P DL + p UL + DL
n i  UL  = N UL – n i  UL + DL 
n i  DL  = N DL – n i  UL + DL 
And
active
ni = n j  UL  + n j  DL  + n j  UL + DL 
inactive
Calculation of the number of inactive users, n i , attempting to access the circuit switched service i
The number of inactive users on uplink and downlink is calculated from the total number of active users as follows:
active
inactive ni
ni = ------------------------------  p inactive
1 – p inactive
Erlangs
n i is the total number of users trying to access the service i. This figure includes both active and inactive users. They are
determined as follows:
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
UL DL
UL DL
Calculation of number of users per activity status:

inactive
Number of inactive users on UL and DL: n i = n i  p inactive
Number of users active on UL only: n i  UL  = n i  p UL
Number of users active on DL only: 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

Here, Atoll considers all the users as active. Activity probabilities are not calculated.
NUL and NDL values include:
• Users active on uplink and inactive on downlink (ni(UL)),

• Users active on downlink and inactive on uplink (ni(DL)),
• And users active on both links (ni(UL+DL)).
If N UL  N DL
198 AT271_TRG_E6 © Forsk 2009

n j  UL + DL  = N UL
n j  UL  = 0
n j  DL  = N DL – N UL
If N UL  N DL
n j  UL + DL  = N DL
n j  DL  = 0
n j  UL  = N UL – N DL
Therefore, we have:
n j = n j  UL  + n j  DL  + n j  UL + DL 
where n j is the total number of users in the Txi cell attempting to access the packet switched service j.
Note:
• The activity status distribution between users is an average distribution. In fact, in each
simulation, the activity status of each user is randomly drawn. Therefore, if you compute
several simulations at once, average numbers of inactive, active on UL, active on DL and
active on UL and DL users correspond to the calculated distribution. But if you check each
simulation, the activity status distribution between users is different in each of them.
6.4.2 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.
HSDPA users are linked to the A-DPCH radio bearer (an R99 radio bearer). Therefore, 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 HSUPA
users, the network first uses a E-DPCCH/A-DPCH power control on UL and DL, 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. Atoll simulates
these network regulation mechanisms with an iterative algorithm and calculates, for each user distribution, network param-
eters such as cell power, mobile terminal power, active set and handoff status for each terminal. During each iteration of
the algorithm, all the users (R99, HSDPA, and HSUPA service users) selected during the user distribution generation (1st
step) attempt to connect one by one to network transmitters. The process is repeated until the network is balanced, i.e.,
until the convergence criteria (on UL and DL) are satisfied.
As shown in Figure 6.2: on page 200, the simulation algorithm is divided in three parts. All users (R99, HSDPA, and
HSUPA service users) are evaluated by the R99 part of the algorithm. HSDPA and HSUPA users active on DL, unless
they have been rejected during the R99 part of the algorithm, are then evaluated by the HSDPA part of the algorithm.
Finally, HSUPA users active on uplink, unless they have been rejected during the R99 or HSDPA parts of the algorithm,
are then evaluated by the HSUPA part of the algorithm.
© Forsk 2009 AT271_TRG_E6 199

Initialisation
R99 part
Mi Best Server Determination
Mi Active Set Determination
For HSDPA users, this part of UL Power Control

the algorithm is performed for For each R99, HSDPA
the A-DPCH bearer (R99 bearer) and HSUPA mobile, Mi
For HSUPA users, this part is
performed for the E-DPCCH/A-
DPCH bearer (R99 bearer) DL Power Control
UL and DL Interference Update
Congestion and Radio Resource Control
HSDPA part
For each HSDPA and

HSUPA mobile, Mi Fast Link Adaptation
Mobile Scheduling
Radio Resource Control
HSUPA part
Admission Control
For each HSUPA

mobile, Mi
Noise Rise Scheduling
Radio Resource Control
Convergence Study
Figure 6.2: UMTS HSPA Power Control Algorithm
The steps of this algorithm are detailed below.
6.4.2.1 Algorithm Initialization

The total power transmitted by the base station txi on the carrier ic m , P Tx  txi 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  . Uplink received powers by
UL intra UL extra UL
the base station txi on carrier ic m , I tot  txi ic m  , I tot  txi ic m  and I inter – carrier  txi ic m  are initialised to 0 W (i.e.
no connected mobile).
UL
UL I tot  txi ic m 
 X k  txi ic m  = ---------------------------------
- = 0
UL
N tot  txi ic m 
6.4.2.2 R99 Part of the Algorithm

UL
The algorithm is detailed for any iteration k. Xk is the value of the X (variable) at the iteration k. In the algorithm, all Q req
DL
and Q req thresholds depend on user mobility and are defined in Service and Mobility parameters tables. All variables are
described in Definitions and formulas part.
Here, the rate downgrading is not taken into account.
For each mobile Mb
200 AT271_TRG_E6 © Forsk 2009

Determination of Mb’s Best Server (SBS(Mb))
The algorithm used in this step depends on the type of network you plan, a single frequency band network (with f1 as
frequency band) or a dual-band network (with f1 and f2 as frequency bands). For dual-band networks, you can activate
the dual-band dedicated carrier selection mode in the atoll.ini file and model different configurations of dual-band termi-
nals.
To activate the dual-band dedicated carrier selection mode, add the following lines in the Atoll.ini file:
[CDMA]
MultiBandSimu=1
Dual-band terminals can have the following configurations:

• Configuration 1: The terminal can work on f1 and f2 without any priority (select "All" as main frequency band in the
terminal property dialogue).
• Configuration 2: The terminal can work on f1 and f2 but f1 has a higher priority (select "f1" as main frequency band
and "f2" as secondary frequency band in the terminal property dialogue).
1. Default mode (the dual-band dedicated carrier selection mode is not active)
This part describes the way the algorithm works when the dual-band dedicated carrier selection mode is not active. It
applies to single frequency band networks as well as dual-band networks.
For each station txi containing Mb in its calculation area and using the main frequency band supported by the Mb’s terminal
(i.e. either f1 for a single frequency band network, or f1 or f2 for a dual-band terminal with the configuration 1, or f1 for a
dual-band terminal with the configuration 2).
Determination of BestCarrier k  txi M b  .
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.

If carrier selection mode is “UL min noise”
For each carrier ic used by txi, we calculate current loading factor:
UL
UL I tot  txi ic  UL
- + X
X k  txi ic  = -----------------------------
UL
EndFor
UL
BestCarrier k  txi M b  is the carrier with the lowest X k  txi ic 
Else if carrier selection mode is “DL min 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
Q pilot  txi BestCarrier 

k
   BTS  P c  txi M b BestCarrier 

= --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
DL DL DL Term
P tot  txi BestCarrier k  txi M b   + I extra  BestCarrier k  txi M b   + I inter – carrier  BestCarrier k  txi M b   + N 0
If user selects “without Pilot”
Q pilot  txi M b BestCarrier 

k

= ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
DL DL
I intra  BestCarrier k  txi M b   + I extra  BestCarrier k  txi M b   +
DL Term
I inter – carrier  BestCarrier k  txi M b   + N 0 –  1 –     BTS  P c  txi M b BestCarrier 
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
© Forsk 2009 AT271_TRG_E6 201

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).
UL UL
If X k  txi BestCarrier  txi M b    X max , 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 and Mb’s terminal can work on one frequency band only, Mb has failed to be connected to
the network and is rejected.
If no TxBS has been selected and Mb’s terminal can work on another frequency band.
Determination of BestCarrier k  txi M b  for each station txi containing Mb in its calculation area and using another
frequency band supported by the Mb’s terminal (i.e. f1 or f2 for a dual-band terminal with the configuration 1, or f2 for a
dual-band terminal with the configuration 2)
If carrier selection mode for txi is “UL min noise”
UL
- + X
X k  txi ic  = -----------------------------
UL
EndFor
UL

UL UL
Calculation of
Q pilot  txi BestCarrier 

k

= --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
DL DL DL Term
P tot  txi BestCarrier k  txi M b   + I extra  BestCarrier k  txi M b   + I inter – carrier  BestCarrier k  txi M b   + N 0
Q pilot  txi M b BestCarrier 

k

= ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
DL DL
I intra  BestCarrier k  txi M b   + I extra  BestCarrier k  txi M b   +
DL Term
I inter – carrier  BestCarrier k  txi M b   + N 0 –  1 –     BTS  P c  txi M b BestCarrier 

pilot
If Q pilot  txi M b BestCarrier   Q req  Mobility  M b   then txi is rejected by Mb
k
max
k k
Admission control (If simulation respects a loading factor constraint and Mb was not connected in previous iteration).
UL UL
If X k  txi BestCarrier  txi M b    X max , then txi is rejected by Mb
Else
max
k k
202 AT271_TRG_E6 © Forsk 2009

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.
Then we will consider BestCell k  M b  as the best serving cell and ic = BestCarrier k  Tx BS  M b  M b  .
2. Dual-band dedicated carrier selection mode

This part describes the way the algorithm works when the dual-band dedicated carrier selection mode is active. It applies
to dual-band networks when dual-band terminals have the first congiguration (i.e. they can work on two frequency bands
without any priority).
For each transmitter txi containing Mb in its calculation area and using one of the frequency bands supported by the Mb’s
terminal (i.e. either f1 or f2)
   BTS  P c  txi M b ic 
Calculation of Q pilot  txi ic Mb  = ---------------------------------------------------------------------------------------------------------------------------------
-
k DL DL DL Term
P tot  txi ic  + I extra  ic  + I inter – carrier  ic  + N 0
   BTS  P c  txi M b ic 
Q pilot  txi ic Mb  = ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
k DL DL DL Term
I intra  ic  + I extra  ic  + I inter – carrier  ic  + N 0 –  1 –     BTS  P c  txi M b ic 
Rejection of bad candidate cells when the pilot is not received and when 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  txi M b ic   Q req  Mobility  M b   then (txi,ic) is rejected by Mb
k
UL UL
If X k  txi ic   X max , then (txi,ic) is rejected by Mb
Else
Keep (txi,ic) as good candidate cell
If no good candidate cell has been selected, Mb has failed to be connected to the network and is rejected.
Compare the pilot quality received from good candidate cells
The site containing the cell (txi, ic) with the highest Q pilot  txi M b ic  is the best site, N BS  M b  .
k
For each cell (txi, ic) on the best site N BS  M b  and using one of the frequency bands supported by the Mb’s terminal (i.e.
either f1 or f2)
Rejection of bad candidate cells when the pilot is not received and when 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  txi M b ic   Q req  Mobility  M b   then (txi,ic) is rejected by Mb
k
UL UL
If X k  txi ic   X max , then (txi,ic) is rejected by Mb
Determination of the best cell (txi, ic) on the best site N BS  M b 
If a given carrier is specified for the service requested by Mb
BestCell k  Ni M b  is the transmitter associated to the carrier specified for the service
Else the carrier selection mode defined for Ni is considered.

For each pair (txi,ic) on Ni, we calculate the current loading factor:
UL
- + X
X k  txi ic  = -----------------------------
UL
EndFor
UL
BestCell k  Ni M b  is the cell with the lowest X k  txi ic 
BestCell k  Ni M b  is the cell with the lowest P tx  txi ic  k
BestCell k  Ni M b  is randomly selected
© Forsk 2009 AT271_TRG_E6 203

Note:
• The Sequential carrier selection mode is not implemented with the dual-band dedicated
carrier selection algorithm. It only works with the default algorithm.
Endif
max
BestCell k  M b  is the best serving cell and its pilot quality is Q pilot  M b 
k
In the following lines, we will consider ic as the carrier used by the best serving cell
Active Set Determination
For each station txi containing Mb in its calculation area, using ic , and, if neighbours are used, neighbour of
BestCell k  M b 
   BTS  P c  txi M b ic 
Calculation of Q pilot  txi M b ic  = ---------------------------------------------------------------------------------------------------------------------------------
-
k DL DL DL Term
P tot  txi ic  + I extra  ic  + I inter – carrier  ic  + N 0
   BTS  P c  txi M b ic 
Q pilot  txi M b ic  = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
k DL DL DL Term
I intra  txi ic  + I extra  ic  + I inter – carrier  ic  + N 0 –  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  M b ic  k
For each cell (txi,ic) of the Mb active set
Calculation of quality level on Mb traffic channel at (txi,ic), with the minimum power allowed on traffic channel for the Mb
service
req
UL P term – R99  M b ic  k – 1
P b – R99  txi M b ic  = --------------------------------------------------------
-
L T  txi M b 
UL UL UL
P b – DPDCH  txi M b ic  = P b – R99  txi M b ic    1 – r c 
UL UL UL
P b – DPCCH  txi M b ic  = P b – R99  txi M b ic   r c
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,
UL UL
P b – R99  txi M b ic  = P b – DPCCH  txi M b ic  if the user is inactive,
UL
UL  term  P b – DPDCH  txi M b ic  k
-  G UL
Q tch  txi M b ic  k = -------------------------------------------------------------------------------------------------------------------------------------------------------
UL
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
UL  term  P b – DPDCH  txi M b ic  k UL UL
Q tch  txi M b ic  k = ---------------------------------------------------------------------------------
UL
 G p  Service  M b    G div
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)
204 AT271_TRG_E6 © Forsk 2009


UL UL UL
Q k  M b  = f rake efficiency  Q tch  txi M b ic  k
txi  ActiveSet
Else if (Mb is in soft, or softer/soft without MRC)
UL UL UL
Q k  M b  = Max  Q tch  txi M b ic  k    G macro – diversity  2 links
txi  ActiveSet
Else if (Mb is in soft/soft)
UL UL UL
Q k  M b  = 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 

UL UL UL UL
Q k  M b  = Max  f rake efficiency  Q tch  ic  Q tch  ic    G macro – diversity  2 links
 other site 
 txi  ActiveSet 
 samesite 
End If
UL
req Q req  Service  M b  Mobility  M b   req
P term – R99  M b ic  k = --------------------------------------------------------------------------------------------
UL
 P term – R99  M b ic  k – 1
Qk  Mb 
If compressed mode is operated,

Note:
• Compressed mode is operated if:
- Mi and Sj support compressed mode,
And
Resulting CM – activation
- Either Q pilot  txi M b ic   Q pilot if the Ec/Io Active option is selected,
k
CM – activation
- Or P c  txi M b ic   RSCP pilot if the RSCP Active option is selected.
UL UL
req Q req  Service  M b  Mobility  M b    Q req   Service  M b  Mobility  M b   
-  P req
P term – R99  M b ic  k = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- term – R99  M b ic  k – 1
UL
Qk  Mb 
req min 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 
req max
If P term – R99  M b ic  k  P term  M b  then Mb cannot select any cell and its active set is cleared
UL UL
If R nominal  M b   R max  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,ic) in Mb active set
Calculation of quality level on (txi,ic) traffic channel at Mb with the minimum power allowed on traffic channel for the Mb
service
min
DL P tch  Service  M b  
P b  txi M b ic  = ----------------------------------------------------
-
DL
DL  BTS  P b  txi M b ic  k
-  G DL
Q tch  txi M b ic  k = -----------------------------------------------------------------------------------------------------------------------------------
DL
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
DL  BTS  P b  txi M b ic  k DL DL
Q tch  txi M b ic  k = --------------------------------------------------------------
DL
 G p  Service  M b    G div
N tot  ic 
End For
© Forsk 2009 AT271_TRG_E6 205


DL DL DL
txi  ActiveSet
Do
Calculation of the required power for DL traffic channel between (txi,ic) and Mb:
DL
req Q req  Service  M b  Mobility  M b   min
P tch  txi M b ic  k = --------------------------------------------------------------------------------------------
DL
 P tch  Service  M b  
Qk  Mb 
If compressed mode is operated.

DL DL
req Q req  Service  M b  Mobility  M b    Q req   Service  M b  Mobility  M b   
-  P min
P tch  txi M b ic  k = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- tch  Service  M b  
DL
Qk  Mb 
Note:
• Compressed mode is operated if:
- Mi and Sj support compressed mode,
And
- Either Q pilot  txi M b ic   Q pilot if the Ec/Io Active option is selected,
k
CM – activation
- Or P c  txi M b ic   RSCP pilot if the RSCP Active option is selected.
req max 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
DL P tch  Service  M b  
P b  txi M b ic  = ---------------------------------------------------
-
DL
DL  BTS  P b  txi M b ic 
-  G DL
Q tch  txi M b ic  k = --------------------------------------------------------------------------------------------------------------------------
DL
DL DL
N tot  ic  –  1 – F ortho    BTS  P b  txi M b ic 
DL DL DL
If the user is inactive, then his contribution to interference in the calculation of N tot  ic  is P b  txi M b ic   r c .
EndFor

DL DL DL
txi  ActiveSet
DL DL
While Q k  M b   Q req  Service  M b  Mobility  M b   and Mb active set is not empty
DL DL
If R nominal  M b   R max  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).
For each cell (txi,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, Cell Power, Channel Elements)

P tx  txi ic  k DL
While ------------------------------  %Power max
P max
206 AT271_TRG_E6 © Forsk 2009

Rejection of the mobile with the lowest service priority starting from the last admitted
EndFor
Codes Codes
While N  txi ic  k  N max  txi ic 
EndFor
For each site (Node B) Ni
CE – DL CE – DL
While N  N i  k  N max  Ni 
CE – UL CE – UL
While N  N i  k  N max  Ni 
EndFor
Uplink Load Factor Control
UL UL
For each cell (txi,ic) with X  txi ic   X max
EndFor
UL UL
While at least one cell with X  txi ic   X max exists
6.4.2.3 HSDPA Part of the Algorithm

HSDPA and HSUPA users active on DL, unless they have been rejected during the R99 part of the algorithm, are then
evaluated by the HSDPA part of the algorithm.
6.4.2.3.1 HSDPA Power Allocation

The total transmitted power of the cell ( P tx  ic  ) is the sum of the R99 transmitted power, the HSUPA power and the
HSDPA power.
P tx  ic  = P tx – R99  ic  + P HSDPA  ic  + P HSUPA  ic 
• In case of a static HSDPA power allocation strategy, 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.
Therefore, 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, Atoll checks in the simulation that:
DL
P tx – R99  ic  + P HSUPA  ic   P max  ic   %Power max
And it calculates the HSDPA power as follows:
P HSDPA  ic  = P max  ic  – P Headroom  ic  – P tx – R99  ic  – P HSUPA  ic 
6.4.2.3.2 Number of HS-SCCH Channels and Maximum Number of HSDPA Users

The number of HS-SCCH channels ( n HS – SCCH ) is the maximum number of HS-SCCH channels that the cell can manage.
Each HSDPA user consumes one HS-SCCH channel. Therefore, at a time (over a transmission time interval), the number
of connected HSDPA users cannot exceed the number of HS-SCCH channels per cell.
The maximum number of HSDPA users ( n max ) corresponds to the maximum number of HSDPA users that the cell can
support.
Let us assume there are 20 HSDPA users in the cell. All of them are active on DL and they are connected to the A-DCH
R99 bearer. Finally, the number of HS-SCCH channels and the maximum number of HSDPA users respectively equal 4
and 15.
In the HSDPA part, HSDPA users are ranked by the scheduler and treated as described in the figure below.
© Forsk 2009 AT271_TRG_E6 207

Figure 6.3: Connection status of HSDPA users
• The first four users may be simultaneously served if there are enough HSDPA power and OVSF codes available
in order for them to obtain a HSDPA bearer. In this case, they will be connected. Else, they will be delayed.
• The next eleven ones will be delayed since there are no longer HS-SCCH channels available. Their connection
status will be "HSDPA Delayed".
• Finally, the last five users will be rejected beacuse the maximum number of HSDPA user has been fixed to 15.
Their connection status will be "HSDPA Scheduler Saturation".
6.4.2.3.3 HSDPA Bearer Allocation Process

Let us focus on the first four users mentionned in the example of the previous paragraph, "Number of HS-SCCH Channels
and Maximum Number of HSDPA Users" on page 207. Once scheduled, a new fast link adaptation is carried out on these
users in order to determine if they can obtain a HSDPA bearer. They are processed in the order defined by the scheduler
and the cell HSDPA power is shared between them as explained below.
In the bearer allocation process shown below, the 4 HSDPA users are represented by Mj, with j = 1 to 4. And, the initial
values of their respective HSDPA powers is 0, i.e. PHSDPA(B(MX)) = 0, where X = 0 to 4. These power values are assigned
one by one by the scheduler, so that with their allocated values, looped back to the starting point, are used in successive
steps.
For the user, Mj, with j varying from 1 to 4:
Figure 6.4: Bearer Allocation Process for Scheduled Users
208 AT271_TRG_E6 © Forsk 2009

6.4.2.3.4 Fast Link Adaptation Modelling

Fast link adaptation (or Adaptive Modulation and Coding) is used in HSDPA. The power on the HS-DSCH channel is trans-
mitted at a constant power while the modulation, the coding and the number of codes are changed to adapt to the radio
conditions variations. Based on the reported channel quality indicator (CQI), the node-B may change every 2ms the modu-
lation (QPSK and optionally 16QAM), 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); 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). Then, it determines the HS-PDSCH CQI, 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. Once the bearer selected, Atoll finds the highest downlink rate that can be provided to the user and may calcu-
late the application throughput.
CQI Based on CPICH Quality
When the option “CQI based on CPICH quality” is selected, Atoll proceeds as follows.
1. CPICH Quality Calculation
Ec
Let us assume the following notation:  -------  ic  corresponds to the CPICH quality.
 Nt  pilot
Two options, available in Global parameters, may be used to calculate Nt: option Without useful signal or option Total
noise.
Therefore, we have:
 BTS    P c  ic 
 Ec
-------  ic  i
= ---------------------------------------------- for the total noise option,
 Nt  pilot DL
N tot  ic 
And
 BTS    P c  ic 
 Ec
-------  ic  i
= --------------------------------------------------------------------------------------- for the without useful signal option.
 Nt  pilot DL
N tot  ic  –  1 –     BTS  P c  ic 
i
With
DL DL DL DL term
N tot  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + N 0
DL  DL P SCH  ic   DL P SCH  ic 

I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 –     P tot  ic  – -----------------------
- –  BTS   P tot  ic  – -----------------------
DL term
-
 LT   LT 
txi  txi   txi 

DL DL
I extra  ic  = P tot  ic 
txj j  i
 Ptot  icadj 
DL
DL txj j
I inter – carrier  ic  = --------------------------------------
-
icadj is a carrier adjacent to ic.
RF  ic ic adj  is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
P pilot  ic 
P c  ic  = -----------------------
i LT
i

L T = --------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
term
 BTS ,  and N 0 are defined in "Inputs" on page 186.
Note:
• Atoll performs intra-cell interference computations based on the total power. You can
instruct Atoll to use maximum power by adding the following lines in the Atoll.ini file:
[CDMA]
PmaxInIntraItf = 1
In this case, Atoll considers the following formula:
P max  ic  P max  ic  – P SCH  ic  P max  ic  – P SCH  ic 

- +  BTS  1 – F MUD  1 –    ------------------------------------------------------ –  BTS   ------------------------------------------------------
DL term
I intra  ic  = ----------------------
LT  LT   LT 
© Forsk 2009 AT271_TRG_E6 209

2. CPICH CQI Determination
Let us assume the following notation:  CQI  pilot corresponds to the CPICH CQI.  CQI  pilot is read in the table
Ec
 CQI  pilot = f   -------  ic   . This table is defined for the terminal reception equipment and the selected mobility.
  Nt  pilot
3. HS-PDSCH Quality Calculation

Atoll proceeds as follows:
1st step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ).
P HSDPA  ic  is the power available for HSDPA on the carrier ic. This parameter is either a simulation output, or a user-
defined cell input.
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. It is either fixed
by the user (when the option “HS-SCCH Power Dynamic Allocation”in the cell property dialog is unchecked) or dynamically
calculated (when the option “HS-SCCH Power Dynamic Allocation” is selected).
Ec req
In this case, the HS-SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (noted  -------  ic  ). It
Nt HS – SCCH
is specified in mobility properties.

We have:
 BTS  P c  ic 
 Ec
-------  ic  i
= ------------------------------------ for the total noise option,
 Nt  HS – SCCH DL
N tot  ic 
And
 BTS  P c  ic 
 Ec
-------  ic  i
- for the without useful signal option.
= --------------------------------------------------------------------------------------------------------------------------------------
 Nt  HS – SCCH DL term
N tot  ic  –  1 – F ortho    1 – F MUD    BTS  P c  ic 
i
With
DL DL DL DL term

I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – -----------------------
- –  BTS   P tot  ic  – -----------------------
DL term
-
 LT   LT 

DL DL
txj j  i
DL
DL  j
I inter – carrier  ic  = txj
--------------------------------------
-
P HS – SCCH  ic 
P c  ic  = --------------------------------------
-
i LT
i
and

L T = --------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
term term
 BTS , F ortho , F MUD and N 0 are defined in "Inputs" on page 186.
Therefore,
req
  Ec
-------  ic   N tot  ic 
DL
  Nt  HS – SCCH 
P HS – SCCH  ic  =  ---------------------------------------------------------------------------  L T for the total noise option,
  BTS  i
 
And
210 AT271_TRG_E6 © Forsk 2009

req
  Ec
-------  ic  DL
 N tot  ic  
P HS – SCCH  ic  =  ---------------------------------------------------------------------------------------------------------------------------------------------------------
req
-  L T for the without useful signal option.
  term  Ec-  ic   i
 BTS   1 +  1 – F ortho    1 – F MUD    ------ Nt  HS – SCCH 
2nd step: Then, Atoll evaluates the HS-PDSCH quality
Ec
Let us assume the following notation:  -------  ic  corresponds to the HS-PDSCH quality.
 Nt  HS – PDSCH
Therefore, we have:
 BTS  P c  ic 
 Ec
-------  ic  i
 Nt  HS – PDSCH DL
N tot  ic 
And
 BTS  P c  ic 
 Ec
-------  ic  i
= --------------------------------------------------------------------------------------------------------------------------------------
 Nt  HS – PDSCH P c  ic 
DL term i
N tot  ic  –  1 – F ortho    1 – F MUD    BTS  ----------------
-
n
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5).
With
DL DL DL DL term

- –  BTS   P tot  ic  – -----------------------
DL term
-
 LT   LT 

DL DL
I extra  ic   = P tot  ic 
txj j  i
DL
DL txj j
-
P HS – PDSCH  ic 
P c  ic  = -----------------------------------------
-
i LT
i
And

L T = --------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
term term
Note:
[CDMA]
PmaxInIntraItf = 1

-+  BTS  1 – F MUD   1 – Fortho   ------------------------------------------------------ –  BTS  ------------------------------------------------------
DL term
I intra  ic  = ----------------------
LT  LT   LT 
4. HS-PDSCH CQI Determination

The best bearer that can be used depends on the HS-PDSCH CQI. Let us assume the following notation:
 CQI  HS – PDSCH corresponds to the HS-PDSCH CQI. Atoll calculates  CQI  HS – PDSCH as follows:
 CQI  HS – PDSCH =  CQI  pilot – P pilot + P HS – PDSCH
5. HSDPA Bearer Selection

Knowing the HS-PDSCH CQI, Atoll finds the best bearer that can be used in the table Best Bearer=f(HS-PDSCH CQI).
This table is defined for the terminal reception equipment and the specified mobility.
© Forsk 2009 AT271_TRG_E6 211

Then, Atoll checks if best bearer characteristics are compliant with cell and user equipment category capabilities. If they
are compliant, Atoll selects the best bearer. Otherwise, it searches for so as to find the suitable bearer.
Bearer characteristics are provided in the HSDPA Bearer table. We assume the best bearer index is 23. Characteristics
of this bearer are:
• Transport block size: 9719 bits
• Number of HS-PDSCH channels used: 7
• 16QAM modulation used: Yes
• RLC Peak Rate: 4.48 Mb/s
Figure 6.5: HSDPA Radio Bearers Table
We assume that the terminal user equipment category is 3. Its capabilities are:
• Maximum transport block size: 7298 bits
• Maximum number of HS-PDSCH channels: 5
• 16QAM modulation supported: Yes
• Minimum number of TTI between two TTI used: 2
Figure 6.6: HSDPA UE Categories Table
HSDPA cell capabilities are:

• Maximum number of HS-PDSCH channels: 15.
The bearer 23 cannot be selected because:
212 AT271_TRG_E6 © Forsk 2009

• The number of HS-PDSCH channels (7) exceeds the maximum number of HS-PDSCH channels the terminal can
use (5),
• And the transport block size (9719 bits) exceeds the maximum transport block size (7298 bits) the terminal can
carried.
In the Bearer table, Atoll searches a suitable bearer and selects the bearer index 22.
• The number of HS-PDSCH channels (5) does not exceed the maximum number of HS-PDSCH channels the ter-
minal can use (5) and the maximum number of HS-PDSCH channels available at the cell level (15),
• The transport block size (7168 bits) does not exceed the maximum transport block size (7298 bits) the terminal
can carried.
• 16QAM modulation is supported by the terminal.
6. HS-PDSCH Quality Update

Once the bearer selected, Atoll exactly knows the number of HS-PDSCH channels. Therefore, when the method “Without
useful signal” is used, 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).
CQI Based on HS-PDSCH Quality
When the option “CQI based on HS-PDSCH quality” is selected, Atoll proceeds as follows.
1. HS-PDSCH Quality Calculation
Atoll proceeds as follows:
1st step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ).
P HSDPA  ic  is the power available for HSDPA on the carrier ic. This parameter is either a simulation output, or a user-
defined cell input.
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. It is either fixed
by the user (when the option “HS-SCCH Power Dynamic Allocation”in the cell property dialog is unchecked) or dynamically
calculated (when the option “HS-SCCH Power Dynamic Allocation” is selected).
Ec req
In this case, the HS-SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (noted  -------  ic  ). It
Nt HS – SCCH
is specified in mobility properties.

We have:
 BTS  P c  ic 
 Ec
-------  ic  i
N tot  ic 
And
 BTS  P c  ic 
 Ec
-------  ic  i
= --------------------------------------------------------------------------------------------------------------------------------------
i
With
DL DL DL DL term

- –  BTS   P tot  ic  – -----------------------
DL term
-
 LT   LT 

DL DL
txj j  i
DL
DL txj j
-
© Forsk 2009 AT271_TRG_E6 213

P c  ic  = --------------------------------------
-
i LT
i
And

L T = --------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
term term
Therefore,
req
  Ec
-------  ic   N tot  ic 
DL
P HS – SCCH  ic  =  ---------------------------------------------------------------------------  L T for the total noise option,
  BTS  i
 
And
req
  Ec
-------  ic  DL
 N tot  ic  
P HS – SCCH  ic  =  ---------------------------------------------------------------------------------------------------------------------------------------------------------
req
-  L T for the without useful signal option.
  term  Ec    i
 BTS   1 +  1 – F ortho    1 – F MUD    ------ Nt
-  ic 
 HS – SCCH 
2nd step: Then, Atoll evaluates the HS-PDSCH quality
Ec
Let us assume the following notation:  -------  ic  corresponds to the HS-PDSCH quality.
Two options, available in Global parameters, may be used to calculate Nt: option Without useful signal or option Total
noise.
We have:
 BTS  P c  ic 
 Ec
-------  ic  i
 Nt  HS – PDSCH DL
N tot  ic 
And
 BTS  P c  ic 
 Ec
-------  ic  i
= --------------------------------------------------------------------------------------------------------------------------------------
 Nt  HS – PDSCH P c  ic 
DL term i
N tot  ic  –  1 – F ortho    1 – F MUD    BTS  -----------------
n
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5). Then, it calculates the HS-PDSCH
CQI and the bearer to be used. Once the bearer selected, Atoll exactly knows the number of HS-PDSCH channels and
recalculates the HS-PDSCH quality with the real number of HS-PDSCH channels.
With
DL DL DL DL term

- –  BTS   P tot  ic  – -----------------------
DL term
-
 LT   LT 

DL DL
I extra  ic   = P tot  ic 
txj j  i
DL
DL  j
I inter – carrier  ic  = txj
--------------------------------------
-
P HS – PDSCH  ic 
P c  ic  = -----------------------------------------
-
i LT
i
And

L T = --------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
term term
214 AT271_TRG_E6 © Forsk 2009

Note:
[CDMA]
PmaxInIntraItf = 1

-+ BTS 1 – F MUD  1 – F ortho  ------------------------------------------------------ –  BTS  ------------------------------------------------------
DL term
I intra  ic  = ----------------------
LT  L T
  LT 
2. HS-PDSCH CQI Determination
Let us assume the following notation:  CQI  HS – PDSCH corresponds to the HS-PDSCH CQI.  CQI  HS – PDSCH is read in
Ec
the table  CQI  HS – PDSCH = f   -------  ic   . This table is defined for the terminal reception equipment and the
  Nt  HS – PDSCH
specified mobility.
3. HSDPA Bearer Selection
The bearer is selected as described in "HSDPA Bearer Selection" on page 211.
6.4.2.3.5 Scheduling Algorithms

Three scheduling algorithms are available, Max C/I, Round Robin and Proportional Fair. Impact they have on the simula-
tion result is described in the tables below.
Let 16 HSDPA users in the cell. All of them are active on DL and connected to the A-DCH R99 bearer. The number of HS-
SCCH channels and the maximum number of HSDPA users have been respectively set to 4 and 15.
• Max C/I: 15 HSDPA users (where 15 corresponds to the maximum number of HSDPA users defined) enters the
scheduler in the same order as in the simulation. Then, they are sorted in descending order by the channel quality
indicator (CQI), i.e. in a best bearer descending order.
Best Bearer DL Obtained Rate Connection

Mobiles Simulation Rank
(kbps) (kbps) Status
M1 2 2400 2400 Connected
M4 9 2080 3.4 Delayed
M5 10 2080 3.4 Delayed
M6 12 2080 3.4 Delayed
M7 13 2080 3.4 Delayed
M8 14 2080 3.4 Delayed
M9 7 1920 3.4 Delayed
M10 1 1600 3.4 Delayed
M11 3 1600 3.4 Delayed
M12 4 1600 3.4 Delayed
M13 5 1600 3.4 Delayed
M14 6 1600 3.4 Delayed
M15 11 1440 3.4 Delayed
M16 16 2080 0 Scheduler Saturation
• Round Robin: HSDPA users are taken into account in the same order than the one in the simulation (random
order).
Best Bearer DL Obtained Rate Connection

Mobiles Simulation Rank
(kbps) (kbps) Status
M3 3 1600 3.4 Delayed
M4 4 1600 3.4 Delayed
M5 5 1600 3.4 Delayed
M6 6 1600 3.4 Delayed
© Forsk 2009 AT271_TRG_E6 215

M7 7 1920 3.4 Delayed

M8 8 2080 3.4 Delayed
M9 9 2080 3.4 Delayed
M10 10 2080 3.4 Delayed
M11 11 1440 3.4 Delayed
M12 12 2080 3.4 Delayed
M13 13 2080 3.4 Delayed
M14 14 2080 3.4 Delayed
M15 15 2400 3.4 Delayed
M16 16 2080 0 Scheduler Saturation
• Proportional Fair: 15 HSDPA users (where 15 corresponds to the maximum number of HSDPA users defined)
enters the scheduler in the same order as in the simulation. Then, 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).
For a user i, the random parameter RP i is calculated as follows:
Simu CQI
RP i = 50  R i + 50  R i
Where,
Simu
Ri is the user rank in the simulation.
CQI
Ri is the user rank according to the CQI.
Note:
• You can change the default weights by editing the atoll.ini file. For more information, see
the Administrator Manual.
Simulation Best Bearer DL Obtained Rate

Mobiles CQI Rank RP Connection Status
Rank (kbps) (kbps)
M1 2 1 150 2400 2400 Connected
M2 1 10 550 1600 960 Connected
M3 8 3 550 2080 160 Connected
M4 9 4 650 2080 3.4 Delayed
M5 3 11 700 1600 3.4 Delayed
M6 10 5 750 2080 3.4 Delayed
M7 4 12 800 1600 3.4 Delayed
M8 7 9 800 1920 3.4 Delayed
M9 15 2 850 2400 3.4 Delayed
M10 5 13 900 1600 3.4 Delayed
M11 12 6 900 2080 3.4 Delayed
M12 6 14 1000 1600 3.4 Delayed
M13 13 7 1000 2080 3.4 Delayed
M14 14 8 1100 2080 3.4 Delayed
M15 11 15 1300 1440 3.4 Delayed
M16 16 - - 2080 0 Scheduler Saturation
6.4.2.4 HSUPA Part of the Algorithm

Finally, HSUPA users active on uplink, unless they have been rejected during the R99 or HSDPA parts of the algorithm,
are then evaluated by the HSUPA part of the algorithm.
Let us assume there are eight HSUPA users within an HSUPA cell and all of them are active on UL. The first four have
been connected to an HSDPA bearer, the last one has been rejected and the remaining three have been delayed in the
HSDPA part. Finally, the maximum number of HSUPA users within each cell has been set to six.
216 AT271_TRG_E6 © Forsk 2009

HSDPA
Mobiles Simulation Rank Connection
Status
M1 1 Connected
M2 2 Connected
M3 3 Connected
M4 4 Connected
M5 5 Delayed
M6 6 Delayed
M7 7 Delayed
M8 8 Rejected
6.4.2.4.1 HSUPA Users

Atoll processes all the HSUPA users who are connected to an HSDPA bearer or were delayed in the HSDPA part. It
considers each user in the order established during the generation of the user distribution without exceeding the maximum
number of HSUPA users within each cell.
In this case, Atoll will consider the first six HSUPA users only and will reject the seventh user in order not to exceed the
maximum number of HSUPA users allowed in the cell (his connection status is "HSUPA scheduler saturation").
6.4.2.4.2 Admission Control

During admission control, Atoll selects a list of HSUPA bearers that are compatible with the user equipment capabilities
for each HSUPA user.
Let us focus on one HSUPA user with category 3 user equipment and a 50km/h mobility.
HSUPA user equipment categories are provided in the HSUPA User Equipment Categories table. 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
Figure 6.7: HSUPA UE Categories Table
HSUPA bearer characteristics are provided in the HSUPA Bearer table. An HSUPA bearer is described with following char-
acteristics:
• Radio Bearer Index: The bearer index number.
• TTI Duration (ms): The TTI duration in ms. The TTI can be 2 or 10 ms.
• Transport Block Size (Bits): The transport block size in bits.
• Number of E-DPDCH Codes: The number of E-DPDCH channels used.
• Minimum Spreading Factor: The smallest spreading factor used.
• RLC Peak Rate (bps): The RLC peak rate represents the peak rate without coding (redundancy, overhead,
addressing, etc.).
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 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).
• The minimum spreading factor used by the bearer is not less than the smallest spreading factor supported by the
terminal (4).
The HSUPA bearers compatible with category 3 user equiment are framed in red:
© Forsk 2009 AT271_TRG_E6 217

Figure 6.8: HSUPA Radio Bearers Table
Then, during admission control, 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. Among the compatible HSUPA bearers, Atoll chooses the one with the
lowest required Ec/Nt threshold.
Here, this is the index 1 HSUPA bearer; the required Ec/Nt threshold to obtain this bearer is -21.7dB.
Ec req
Then, from the required Ec/Nt threshold,  -------
req
, Atoll calculates the required terminal power, P term – HSUPA .
 Nt  E – DPDCH
Ec req
P term – HSUPA =  -------
req UL
 L T  N tot
With
UL intra UL extra tx
UL tx UL
N tot  ic  =  1 – F MUD   term   I tot  ic  + I tot  ic  + I inter – carrier  ic  + N 0

UL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
tx UL intra UL extra UL tx
 term , F MUD , I tot , I tot , I inter – carrier and N 0 are defined in "Inputs" on page 186.
Figure 6.9: HSUPA Bearer SelectionTable
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").
At the end of this step, the number of non-rejected HSUPA users is n HSUPA . All of them will be connected to an HSUPA
bearer at the end.
6.4.2.4.3 HSUPA Bearer Allocation Process

Noise Rise Scheduling and Determination of the Obtained HSUPA Bearer
The obtained HSUPA radio bearer is the bearer that the HSUPA user obtains after noise rise scheduling and radio
resource control. The noise rise scheduling algorithm attempts to evenly share the remaining cell load between the users
admitted in admission control; in terms of HSUPA, each user is allocated a right to produce interference.
UL
The remaining cell load factor on uplink ( X  txi ic  ) depends on the maximum load factor allowed on uplink and how
much uplink load is produced by the served R99 traffic. It can be expressed as follows:
218 AT271_TRG_E6 © Forsk 2009

UL UL UL
X  txi ic  = X max  txi ic  – X R99  txi ic 
Then, Atoll evenly shares the remaining cell load factor between the HSUPA users admitted during the previous step.
UL
UL X  txi ic 
X user  txi ic  = ----------------------------------
n HSUPA
Ec max
From this value, Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------- ). For further information on the
calculation, see "Uplink Load Factor Due to One User" on page 229.
max
 Ec
------- 1
= ----------------------------------------------- for the Without useful signal option
 Nt  E – DPDCH UL
F  txi ic 
-------------------------------------- – 1
UL
X user  txi ic 
UL
 Ec
max X user
------- - for the Total noise option
= -----------------
 Nt  E – DPDCH UL
F
Then, it selects an HSUPA bearer. The allocation depends on the maximum E-DPDCH Ec⁄Nt allowed and on UE capabil-
ities. Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. This is the HSUPA bearer
UL
R RLC – peak  Index HSUPABearer 
( Index HSUPABearer ) with the highest potential throughput ( ------------------------------------------------------------------------------------ ) where:
N Rtx  Index HSUPABearer 
req Ec max
•  Ec
-------   -------
 Nt  E – DPDCH  Nt  E – DPDCH
req max
• And P term – HSUPA  P term
Ec req
When several HSUPA bearers are available, Atoll selects the one with the lowest  ------- .
If we keep the same example as previously, we have six HSUPA users in the cell. The remaining cell load factor is shared
between the users. If the remaining cell load factor is 0.6, the UL load factor alloted to each user is 0.1. Let’s take the cell
UL reuse factor equal to 1.5. Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (the Without useful signal option is
selected).
Ec max
We have:  ------- = -11.5 dB
Here, the obtained HSUPA bearer is the index 5 HSUPA bearer. It provides a potential throughput of 128 kbps and
requires E-DPDCH Ec⁄Nt of -13 dB (lower than -11.5 dB) and a terminal power lower than the maximum terminal power
allowed.
. Potential
HSUPA Bearers Required Ec/Nt Nb of RLC Peak Rate
Throughput
Index Threshold (dB) Retransmissions (kbps)
(kbps)
1 -21.7 2 32 16
2 -19 2 64 32
3 -16.1 2 128 64
4 -13.9 2 192 96
5 -13 2 256 128
6 -10.1 2 512 256
7 -8 2 768 384
UL
After selecting the bearer, Atoll reads the corresponding RLC peak rate, R RLC – peak  Index HSUPABearer  and checks that
each user has obtained the average requested rate (defined in the properties of the service). A user is considered as
"happy" if the RLC peak rate provided by the HSUPA bearer exceeds the average requested rate and "unhappy" if not.
Atoll collects the unused load of "happy" users and redistributes it among the "unhappy" users. This process is repeated
until there is no more available load.
UL
The unused load (  X user  unused ) of an "happy" user corresponds the difference of load between the user load allowing
Ec max
the maximum E-DPDCH Ec⁄Nt (  ------- ) and the user load required to reach the required E-DPDCH Ec⁄Nt
Ec req
(  ------- ). It is determined as follows:
UL UL UL
 X user  unused = X user –  X user  req
© Forsk 2009 AT271_TRG_E6 219

Where
UL
X user is the UL load factor alloted to the user after equally sharing the remaining cell UL load between the HSUPA users.
Ec req
 X user  req is the UL load factor required in order to reach the required E-DPDCH Ec⁄Nt (  -------
UL
).
Determination of the Requested HSUPA Bearer
The requested HSUPA radio bearer is selected from the HSUPA bearers compatible with the user equipment. Atoll deter-
mines the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. The HSUPA user is
treated as if he is the only user in the cell. Therefore, if we go on with the previous example, the maximum E-DPDCH
Ec⁄Nt allowed is equal to -1.8 dB and the requested HSUPA bearer is the index 7 HSUPA bearer. It requires E-DPDCH
Ec⁄Nt of -8 dB (lower than -1.8 dB) and a terminal power lower than the maximum terminal power allowed.
6.4.2.5 Convergence Criteria

The convergence criteria are evaluated for each iteration, and can be written as follow:
  max P tx  ic  k – P tx  ic  k – 1   max N DL DL
user  ic  k – N user  ic  k – 1

 DL = max  int  ---------------------------------------------------------------------------------------
Stations
-  100  int  ----------------------------------------------------------------------------------------------------
Stations
 100 
  P tx  ic  k   DL 
  N user  ic  k 
  max I UL UL
tot  ic  k – I tot  ic  k – 1
  max N UL UL
user  ic  k – N user  ic  k – 1

 UL = max  int  ---------------------------------------------------------------------------------------
Stations
-  100  int  ----------------------------------------------------------------------------------------------------
Stations
 100 
  UL   UL 
  I tot  ic  k   N user  ic  k 
Atoll stops the algorithm if:
1st case: Between two successive iterations,  UL and  DL are lower (  ) than their respective thresholds (defined when
creating a simulation).
The simulation has reached convergence.
Example: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5.
If  UL  5 and  DL  5 between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th iteration. Convergence
has been reached.
2nd case: After 30 iterations,  UL and/or  DL are still higher than their respective thresholds and from the 30th iteration,
 UL and/or  DL do not decrease during the next 15 successive iterations.
The simulation has not reached convergence (specific divergence symbol).

Examples: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5.
1. After the 30th 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. Convergence has not been reached.
2. After the 30th iteration,  UL and/or  DL equal 80, they start decreasing slowly until the 40th iteration (without going
under the thresholds) and then, do not change during 15 successive iterations: Atoll stops the algorithm at the 56th
iteration without reaching convergence.
3rd case: After the last iteration.
If  UL and/or  DL are still strictly higher than their respective thresholds, the simulation has not reached convergence
(specific divergence symbol).
If  UL and  DL are lower than their respective thresholds, the simulation has reached convergence.
6.4.3 Results
6.4.3.1 R99 Related Results
This table contains some R99 specific simulation results provided in the Cells and Mobiles tabs of the simulation property
dialog.
220 AT271_TRG_E6 © Forsk 2009

 DL 
 P  txi ic  – P SCH  txi ic 
P tot  txi ic  – F ortho   BTS   tot
DL ---------------------------------
- Downlink intra-cell interference at
DL
I intra  txi ic  LT  None
 txi  terminal on carrier ic
DL
–  1 – F ortho    BTS  P b  txi ic 


DL
DL
txj j  i

DL
W
txj
------------------------------------------------
DL DL DL DL Total effective interference at terminal

I tot  ic  I intra  ic  + I extra  ic  + I inter – carrier  ic  W
on carrier ic (after unscrambling)
DL DL Term Total received noise at terminal on

N tot  ic  I tot  ic  + N 0 W
carrier ic

 Pb
UL
UL intra  ic 
term
ic
txi


UL
term
ic
txj j  i
 Pb
UL
 ic adj 
txj j
--------------------------------------
-
UL UL extra UL intra
Total received interference at
Tx
 txi ic +  1 – F MUD   term  I tot
UL
 txi ic  +I inter – carrier  txi ic  W

UL
UL
I tot  txi ic 
X  txi ic  -----------------------------
- None Cell uplink load factor on carrier ic
UL
UL
F
UL
 txi ic  ------------------------------------------------------------------------------------------- None Cell uplink reuse factor on carrier ic
UL intra Tx
I tot  txi ic    1 – F MUD   term 
1 Cell uplink reuse efficiency factor on

E
UL
 txi ic  ------------------------------ None
UL
F  txi ic  carrier ic
Simulation result available per cell

DL DL
 I extra  ic  + I inter – carrier  ic    L T
---------------------------------------------------------------------------------------- + 1 – F ortho   BTS
DL
P Tx  txi ic 
 ------------------------------------------------------------------------------------------------------------------------------------------
1
------------- +  1 – F ortho   BTS 
-
tch DL
CI req
DL None
X  txi ic  DL Downlink load factor on carrier ic
DL Q req
with CI req = -----------
-
DL
Gp
Simulation result available per mobile
DL
I tot  ic 
--------------------
DL
N tot  ic 
DL
DL
I tot  ic 
F  txi ic  -------------------------------
- None Downlink reuse factor on a carrier ic
DL
I intra  txi ic 
DL DL dB Noise rise on downlink

NR  txi ic  – 10 log  1 – X  txi ic  
UL UL dB Noise rise on uplink
© Forsk 2009 AT271_TRG_E6 221

6.4.3.2 HSPA Related Results

6.4.3.2.1 Statistics Tab
In the Statistics tab, Atoll displays as results:
• The number of non-connected users: This figure includes rejected and delayed users.
• The number of R99 bearer users connected to a cell (result of the R99 part). This figure includes R99 users as
well as HSDPA and HSUPA users since all of them request a R99 bearer.
- The number of R99 bearer users per frequency band.
- The number of R99 bearer users per activity status.
DL UL
- The downlink and uplink rates ( R R99 and R R99 ) generated by their connection to R99 bearers. Only active
users are considered.
 
DL DL UL UL
R R99 = R nominal  R99 Bearer  and R R99 = R nominal  R99 Bearer 
Active Active
users users
DL UL
R nominal  R99 Bearer  is the downlink nominal rate of the user R99 radio bearer and R nominal  R99 Bearer  is the uplink
nominal rate of the user R99 radio bearer.
• The number of connected users with an HSDPA bearer (result of the HSDPA part) and the downlink rate they gen-
erate. Both HSDPA and HSUPA users are considered since they both request an HSDPA bearer. On the other
DL
hand, only active users are taken into consideration in the downlink rate calculation ( R HSDPA ).

DL DL
R HSDPA = R RLC – peak  Index HSDPABearer 
Active
users
DL
R RLC – peak  Index HSDPABearer  is the RLC peak rate provided by the HSUPA bearer.
• The number of connected HSUPA users (result of the HSUPA part). Only HSUPA users are considered.
At the end of the R99 part, HSDPA and HSUPA users can be:
- Either connected and in this case, they obtain the requested R99 bearer (ADPCH-UL64 for HSDPA users and
ADPCH-EDPCCH for HSUPA users),
- Or rejected exactly for the same reasons as R99 users.
Then, only connected HSDPA and HSUPA users are considered in the HSDPA part. At the end of the HSDPA
part, they can be:
- Either connected if they obtain an HSDPA bearer,

- Or rejected if the maximum number of HSDPA users per cell is exceeded,
- Or delayed in case of lack of resources (HSDPA power, HS-SCCH power, HS-SCCH channels, OVSF codes).
In the HSUPA part, Atoll processes all the HSUPA users who are connected to an HSDPA bearer or were delayed
in the previous step. At the end, they can be:
- Either connected if they obtain an HSUPA bearer,

- Or rejected if the maximum number of HSUPA users per cell is exceeded or if the terminal power required to
obtain the lowest compatible HSUPA bearer exceeds the maximum terminal power.
UL
In addition, Atoll indicates the uplink rate the active HSUPA users connected with an HSUPA bearer ( R HSUPA )
generate:

UL UL
R HSUPA = R RLC – peak  Index HSUPABearer 
Active
users
UL
R RLC – peak  Index HSUPABearer  is the RLC peak rate provided by the HSUPA bearer.
6.4.3.2.2 Mobiles Tab

In the Mobiles tab, Atoll indicates for each user:
UL DL
• The uplink and downlink total requested rates in kbps (respectively, R requested  M b  and R requested  M b  )
For an R99 user, the DL and UL total requested rates correspond to the DL and UL nominal rates of the R99 bearer asso-
ciated to the service.
For an HSDPA user, the uplink requested rate corresponds to the nominal rate of ADPCH R99 radio bearer and the down-
link requested rate is the sum of the ADPCH radio bearer nominal rate and the RLC peak rate that the selected HSDPA
radio bearer can provide. Here, the HSDPA user is treated as if he is the only user in the cell and then, Atoll determines
the HSDPA bearer he would obtain by considering the entire HSDPA power available of the cell.
For an HSUPA user, the uplink requested rate is equal to the sum of the ADPCH-EDPCCH radio bearer nominal rate and
the RLC peak rate of the requested HSUPA radio bearer. The requested HSUPA radio bearer is selected from the HSUPA
222 AT271_TRG_E6 © Forsk 2009

bearers compatible with the user equipment. Here, the HSUPA user is treated as if he is the only user in the cell and then,
Atoll determines the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. The down-
link requested rate is the sum of the ADPCH-EDPCCH radio bearer nominal rate and the RLC peak rate that the requested
HSDPA radio bearer can provide. The requested HSDPA radio bearer is determined as explained in the previous para-
graph.
UL DL
• The uplink and downlink total obtained rates in kbps (respectively, R obtained  M b  and R obtained  M b  )
For an R99 user, the obtained rate is the same as the requested rate if he is connected without being downgraded. Other-
wise, the obtained rate is lower (it corresponds to the nominal rate of the selected R99 bearer). If the user is rejected, the
obtained rate is zero.
For an HSDPA user connected to an HSDPA bearer, the uplink obtained rate equals the requested one and the downlink
obtained rate corresponds to the instantaneous rate; this is the sum of the A-DPCH radio bearer nominal rate and the RLC
peak rate provided by the selected HSDPA radio bearer after scheduling and radio resource control. If the HSDPA user is
delayed (he is only connected to an R99 radio bearer), uplink and downlink obtained rates correspond to the uplink and
downlink nominal rates of ADPCH radio bearer. Finally, if the HSDPA user is rejected either in the R99 part or in the
HSDPA part (i.e., because the HSDPA scheduler is saturated), the uplink and downlink obtained rates are zero.
For a connected HSUPA user, on uplink, if the user is connected to an HSUPA bearer, the obtained uplink rate is the sum
of the ADPCH-EDPCCH radio bearer nominal rate and the RLC peak rate provided by the selected HSUPA radio bearer
after noise rise scheduling. On downlink, if the user is connected to an HSDPA bearer, the obtained downlink rate corre-
sponds to the instantaneous rate. The instantaneous rate is the sum of the ADPCH-EDPCCH radio bearer nominal rate
and the RLC peak rate provided by the selected HSDPA radio bearer after scheduling and radio resource control. If the
user is delayed, the obtained downlink rate corresponds to the downlink nominal rate of ADPCH-EDPCCH radio bearer.
If the HSUPA user is rejected, the obtained uplink and downlink rates are "0."
• The mobile total power ( P term )
UL
P term = P term – R99  f act – EDPCCH + P term – HSUPA for HSUPA users
And
P term = P term – R99 in case of a R99 user
DL
• The HSDPA application throughput in kbps ( T application  M b  )
This is the net HSDPA throughput without coding (redundancy, overhead, addressing, etc.).
DL
DL R obtained  M b    1 – BLER HSDPA   SF Rate – R
T application  M b  = ----------------------------------------------------------------------------------------------------------------------------------
-
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). This graph describes the variation of BLER as a function of
the measured quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll calculates the corresponding BLER.
SF Rate and R respectively represent the scaling factor between the application throughput and the RLC (Radio Link
Control) throughput and the throughput offset. These two parameters model the header information and other supplemen-
tary data that does not appear at the application level. They are defined in the service properties.
TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal user
equipment category properties.
• The number of OVSF codes
This is the number of 512-bit length OVSF codes consumed by the HSDPA or the HSUPA user for ADPCH and HS-
PDSCH channels.
• The required HSDPA power in dBm
It corresponds to the HSDPA power required to provide the HSDPA or HSUPA user with the downlink requested rate.
If the HSDPA bearer allocated to the user is the best one, the required HSDPA power corresponds to the available HSDPA
power of the cell. On the other hand, if the HSDPA has been downgraded in order to be compliant with cell and UE capa-
bilities, the required HSDPA power will be lower than the available HSDPA power of the cell.
• The served HSDPA power in dBm
This is the HSDPA power required to provide the HSDPA or HSUPA user with the downlink obtained rate.
• The No. of HSUPA Retransmissions (Required)
The number of retransmissions performed by the requested HSUPA radio bearer.
• The No. of HSUPA Retransmissions (Obtained)
The number of retransmissions performed by the obtained HSUPA radio bearer.
UL
• The HSUPA application throughput in kbps ( T application  M b  )
This is the net HSUPA throughput without coding (redundancy, overhead, addressing, etc.).
© Forsk 2009 AT271_TRG_E6 223

UL
UL R obtained  M b    1 – BLER HSUPA   SF Rate – R
T application  M b  = ----------------------------------------------------------------------------------------------------------------------------------
-
N Rtx
Where:
BLER HSUPA is the residual BLER after N Rtx retransmissions. It is read in the quality graph defined for the quartet “recep-
tion equipment-selected bearer-number of retransmissions-mobility” (HSUPA Quality Graphs tab in the Reception equip-
ment properties). This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt).
Knowing the E-DPDCH Ec/Nt, Atoll calculates the corresponding BLER.
N Rtx is the number of retransmissions for the obtained HSUPA bearer. This figure is read in the HSUPA Bearer Selection
table.
The following columns appear if, when creating the simulation, you select "Detailed information about mobiles":
• The uplink and downlink requested RLC peak rates (kbps)
For an R99 user, the uplink and downlink requested RLC peak rates are 0.
For an HSDPA user, the uplink RLC peak rate is 0 and the downlink requested RLC pear rate is the rate that the selected
HSDPA radio bearer can provide. Here, the HSDPA user is treated as if he is the only user in the cell and then, Atoll deter-
mines the HSDPA bearer he would obtain by considering the entire HSDPA power available of the cell.
For an HSUPA user, the requested uplink RLC peak rate is the rate of the requested HSUPA radio bearer. The requested
HSUPA radio bearer is selected from the HSUPA bearers compatible with the user equipment. Here, the HSUPA user is
treated as if he is the only user in the cell and then, Atoll determines the HSUPA bearer the user would obtain by consid-
ering the entire remaining load of the cell. If the user is connected to an HSDPA bearer in the downlink, the downlink
requested RLC peak rate is the rate that the requested HSDPA radio bearer can provide. The requested HSDPA radio
bearer is determined as explained in the previous paragraph.
• The uplink and downlink obtained RLC peak rate (kbps)
For an R99 user, the uplink and downlink obtained RLC peak rates are 0.
For an HSDPA user connected to an HSDPA bearer, the uplink obtained RLC peak rate is 0, and the downlink obtained
RLC peak rate is the rate provided by the selected HSDPA radio bearer after scheduling and radio resource control.
For a connected HSUPA user, on uplink, if the user is connected to an HSUPA bearer, the obtained uplink RLC peak rate
is the rate provided by the selected HSUPA radio bearer after noise rise scheduling. On downlink, if the user is connected
to an HSDPA bearer, the downlink obtained RLC peak rate is the rate provided by the selected HSDPA radio bearer after
scheduling and radio resource control.
6.4.3.2.3 Cells Tab

In the Cells tab, Atoll gives:
• The available HSDPA power in dBm ( P HSDPA  ic  )
This is:
- Either a fixed value in case of a static HSDPA power allocation strategy,
- Or a simulation result when the option "HSDPA Power Dynamic Allocation" is selected. We have:
P HSDPA  ic  = P max  ic  – P Headroom  ic  – P tx – R99  ic  – P HSUPA  ic 
 
DL
with P tx – R99  ic  = P pilot  ic  + P SCH  ic  + P OtherCCH  ic  + P tch  ic  + P tch  ic   f act –ADPCH
tch used for tch used for
R99 users HSPA users
• The number of HSDPA users in the cell

They are the connected and delayed HSDPA bearer users. This figure includes HSDPA and HSUPA users.
• 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, i.e. within one transmis-
sion time interval. At the end of the HSDPA part of the simulation, all these users are connected to the cell, with a connec-
tion with the R99 radio bearer (ADPCH-UL64 for HSDPA users and ADPCH-EDPCCH bearer for HSUPA users) and a
HSDPA bearer.
DL
• The instantaneous HSDPA rate in kbps ( R Inst  cell  )
This is the number of kilobits per second that the cell supports on downlink to provide simultaneous connected HSDPA
bearer users with a HSDPA bearer.

DL DL
R Inst  cell  = R obtained  M b 
M b  cell
224 AT271_TRG_E6 © Forsk 2009

DL
• The instantaneous HSDPA MAC Throughput in kbps ( T MAC  cell  )
S block  M b 

DL
T MAC  cell  = -------------------------------------------
-
T TTI   TTI  M b 
M b  cell
Where,
S block  M b  is the transport block size (in kbits) of the HSDPA bearer selected by the user; it is defined for each HSDPA
bearer in the HSDPA Radio Bearers table.
TTI  M b  is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal
user equipment category properties.
–3
T TTI is the TTI duration, i.e. 2 10 s (2000 TTI in one second). This value is specified by the 3GPP.
DL
• The average instantaneous HSDPA rate in kbps ( R Av – Inst  cell  )

DL
R obtained  M b 
DL M  cell
R Av – Inst  cell  = --------------------------------------------------------
b -
nM
b
DL
• The HSDPA application throughput in kbps ( T application  cell  )

DL DL
Either T application  cell  = T application  M b  if the scheduling algorithm is Round Robin or Proportional Fair,
M b  cell
DL DL
Or T application  cell  = T application  M b  maxC  I   if the scheduling algorithm is Max C/I.
M b  maxC  I  is the user with the highest C  I in the cell.
DL
• The minimum HSDPA RLC peak rate in kbps ( min  R obtained  M b   )
M b  cell
It corresponds to the lowest of RLC peak rates obtained by HSDPA bearer users connected to the cell.
DL
• The maximum HSDPA RLC peak rate in kbps ( max  R obtained  M b   )
M b  cell
It corresponds to the highest of RLC peak rates obtained by HSDPA bearer users connected to the cell.
• The number of HSUPA users in the cell ( n M ):

c
They are the connected HSUPA users.

UL
• The HSUPA application throughput in kbps ( T application  cell  )

UL UL
T application  cell  = T application  M b 
M b  cell
UL
• The uplink cell load factor due to HSUPA traffic ( X HSUPA  cell  ):
UL
UL  I tot  cell   HSUPA
X HSUPA  cell  = --------------------------------------------
-
UL
N tot  cell 
Where
UL
 I tot  cell   HSUPA is the total interference at transmitter received from HSUPA bearer users.
6.4.3.2.4 Sites Tab

In the Sites tab, Atoll displays:
DL
• The instantaneous HSDPA rate carried by the site in kbps ( R Inst  site  )

DL DL
R Inst  site  = R Inst  cell 
cell  site
DL
• The instantaneous HSDPA MAC Throughput carried by the site in kbps ( T MAC  site  in kbps)

DL DL
T MAC  site  = T MAC  cell 
cell  site
© Forsk 2009 AT271_TRG_E6 225

UL
• The HSUPA rate carried by the site in kbps ( R  site  )

UL UL
R  site  = R obtained  M c 
M c  site
6.4.4 Appendices
6.4.4.1 Admission Control in the R99 Part
During admission control of the R99 part of the simulation, Atoll calculates the uplink load factor of a considered cell
assuming the mobile concerned is connected with it. Here, activity status assigned to users is not taken into account. So
even if the mobile is not active on UL, it can be rejected due to cell load saturation. To calculate the cell UL load factor,
either Atoll takes into account the mobile power determined during power control if mobile was connected in previous iter-
UL
ation, or it estimates a load rise due to the mobile and adds it to the current load. The load rise ( X ) is calculated as
follows:
UL 1
X = --------------------------------------------------
W
1 + ----------------------------------------
UL UL
Q req  R nominal
6.4.4.2 Resources Management

6.4.4.2.1 OVSF Codes Management
OVSF codes are managed on the downlink during the simulation since this resource is downlink limited only. Atoll checks
the availability of this resource during the simulation, first in the R99 part and then in the HSDPA part. It determines the
number of codes that will be consumed by each cell.
OVSF codes form a binary tree. Codes of longer lengths are generated from codes of a shorter length. Length-k OVSF
codes are generated from length-k/2 OVSF codes. Therefore, if one channel needs 1 length-k/2 OVSF code, it is equiva-
lent to use 2 length-k OVSF codes, or 4 length-2k OVSF codes and so on.
512 512-bit-length codes per cell are available in UMTS HSPA projects.
In the R99 part, during the resource control, Atoll determines the number of 512 bit-length codes that will be consumed for
each cell.
If the cell supports HSUPA, Atoll allocates codes for the DL channels used for HSUPA:
• A 128 bit-length code for the E-HICH and E-RGCH channels (i.e. four 512 bit-length OVSF codes), for each cell.
Therefore, Atoll will take four 512-bit-length codes,
• A 256 bit-length code for the E-AGCH channel (i.e. two 512 bit-length OVSF codes), for each cell. Therefore, Atoll
will take two 512-bit-length codes,
If the cell supports HSDPA, Atoll reserves for potential HSDPA users:
Codes – HS – PDSCH
• The minimum number of HS-PDSCH codes defined for the cell, N min . They are 16-bit length OVSF
codes (i.e. thirty-two 512 bit-length OVSF codes). Therefore, Atoll will take 32  N min 512-bit-
length codes,
• A 128 bit-length code per HS-SCCH channel (i.e. four 512 bit-length OVSF codes), for each cell. Therefore, Atoll
will take 4  n HS – SCCH 512-bit-length codes,
Then, it allocates to the cell OVSF codes to support R99 bearers required by R99 and HSDPA users:
• A 256 bit-length code per common channel (i.e. two 512 bit-length OVSF codes), for each cell. Therefore, Atoll will
Overhead – Codes
take 2  N 512-bit-length codes,
• A code per cell-receiver link, for TCH (traffic channels). The length of code to be allocated, Code_Length, depends
on the user activity. We have:
DL
Either Code_Length = F spreading  Active user  when the user is active,
DL
Or Code_Length = F spreading  Inactive user  if the user is inactive.
Codes-TCH
The number of 512 bit-length OVSF codes needed N is calculated from the length of the code to be allocated
as follows:
Codes-TCH 512
N = ------------------------------------
Code_Length
226 AT271_TRG_E6 © Forsk 2009

Figure 6.10: OVSF Code Tree Indices (Not OVSF Code Numbers)
The OVSF code allocation follows the “Buddy” algorithm, which guarantees that:
• If a k-length OVSF code is used, all of its children with lengths 2k, 4k, …, cannot be used as they will not be orthog-
onal.
• If a k-length OVSF code is used, all of its ancestors with lengths k/2, k/4, …, cannot be used as they will not be
orthogonal.
Example: We consider a user with a service requiring the UDD64 R99 radio bearer. This user is active on DL while
connected to a cell (which does not support HSDPA). The spreading factor for active users has been set to 64 and site
equipment requires four overhead downlink channel elements per cell. Atoll will consume four 256 bit-length OVSF codes
for common channels (i.e. eight 512 bit-length OVSF codes) and a 64 bit-length OVSF code for traffic channels (i.e. eight
additional 512 bit-length OVSF codes).
Notes:
• In the R99 part, the OVSF code allocation follows the mobile connection order (mobile
order in the Mobiles tab).
• The OVSF code and channel element management is differently dealt with in case of
“softer” handover. Atoll allocates OVSF codes for each cell-mobile link while it globally
assigns channel elements to a site.
In the HSDPA part, each HSDPA user is assigned a HSDPA bearer (Fast link adaptation). Therefore, Atoll allocates to the
cell:
• A number of 16-bit length OVSF codes per cell-HSDPA receiver, for HS-PDSCH N . This
number depends on the HSDPA bearer assigned to the user; this is the number of HS-PDSCH channels required
by the HSDPA bearer. Therefore, Atoll will take 32  N 512-bit-length codes for each HSDPA
user connected to the cell
Notes:
• In the HSDPA part, Atoll sorts HSDPA mobiles according the selected scheduling
technique. Then, the OVSF code allocation follows this order.
• When HSDPA users (at least one) are connected to the cell, Atoll gives the cell back the
minimum number of OVSF codes reserved for HS-PDSCH ( N min ). On the
other hand, if no HSDPA user is connected, Atoll still keeps these codes and the codes for
HS-SCCH too. This is the same with HSUPA users. Even if no HSUPA user is connected
to the cell, Atoll still keeps the codes for E-HICH, E-RGCH and E-AGCH channels.
6.4.4.2.2 Channel Elements Management

Channel elements are controlled in the simulation. Atoll checks the availability of this resource on uplink and downlink.
In the R99 part, during the resource control, Atoll determines the number of channel elements required by each site on
uplink and downlink.
CE – UL
On uplink, Atoll consumes N  j  channel elements for each cell j on a site NI. This figure includes:
Overhead – CE – UL
• N channel elements for control channels,
TCH – CE – UL
• N per cell-receiver link, for TCH (traffic channels).
CE – UL
Therefore, the number of channel elements required on uplink at the site level, N  N I  , is:
N
CE – UL CE – UL
N  NI  = j
j  NI
© Forsk 2009 AT271_TRG_E6 227

CE – DL
On downlink, Atoll consumes N  j  channel elements for each cell j on a site NI. This figure includes:
Overhead – CE – DL
• N channel elements for control channels (Pilot channel, Synchronisation channel, common
channels),
TCH – CE – DL
• N per cell-receiver link, for TCH (traffic channels).
CE – DL
Therefore, the number of channel elements required on downlink at the site level, N  N I  , is:
N
CE – DL CE – DL
N  NI  = j
j  NI
Note:
• In case of “softer” handover (the mobile has several links with co-site cells), Atoll allocates
channel elements for the best serving cell-mobile link only.
6.4.4.3 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).
6.4.4.3.1 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”.
DL
Q req
Let CI req = -----------
- be the required quality.
DL
Gp
DL DL
Gp and Q req are the processing gain on downlink and the Eb/Nt target on downlink respectively.
In case of soft-handoff, required quality is limited to the effective contribution of the transmitter.
 Ptch  ic 
DL
P tx  ic  = P pilot  ic  + P SCH  ic  + P otherCCH  ic  +
tch
 Ptch  ic 
DL ortho nonOrtho
P tx  ic  = P CCH  ic  + P CCH  ic  +
tch
where
ortho
P CCH  ic  = P pilot  ic  + P otherCCH  ic 
nonOrtho
P CCH  ic  = P SCH  ic 
At mobile level, we have a required power, Ptch:
term
P tch  ic  = CI req   I extra  ic  + I inter – carrier  ic  + I intra  ic  + 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. In case of an HSDPA bearer
DL
user, r = f act – ADPCH .
 DL nonOrtho 
  P tx  ic  – P CCH  ic  – P tch  ic   
 I extra  ic  + I inter – carrier  ic  +  1 – F ortho   BTS   --------------------------------------------------------------------------------------------- 
 LT  L r
P tch  ic  = CI req 
 nonOrtho  T
 P CCH  ic  term 
 - + N0
+ ---------------------------------- 
 LT 
DL
 I extra  ic  + I inter – carrier  ic    L T  r +  1 – F ortho   BTS   P tx  ic   r
+
nonOrtho term
F ortho   BTS  P CCH  ic   r + N 0  LT  r
P tch  ic  = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1
---------------------- +  1 – F ortho   BTS 
CI req  r
I intra  ic  is the total power received at the receiver from the cell with which it is connected.
I extra  ic  is the total power received at the receiver from other cells.
I inter – carrier  ic  is the inter-carrier interference received at the terminal.
We have:
228 AT271_TRG_E6 © Forsk 2009

DL
 I extra  ic  + I inter – carrier  ic    L T  r +  1 – F ortho   BTS   P tx  ic   r
+
nonOrtho term
F ortho   BTS  P CCH  ic   r + N 0  L T  r

DL ortho nonOrtho
P tx  ic  = P CCH  ic  + P CCH  ic  + ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1
tch ---------------------- +  1 – F ortho   BTS 
CI req  r
 I extra  ic  + I inter – carrier  ic    L T  r

-  P DL
------------------------------------------------------------------------------------------------ tx  ic 
DL
DL
P tx  ic  P  ic 
= P ortho  ic  + P nonOrtho  ic  + ---------------------------------------------------------------------------------------------------------------------------
CCH CCH 
tx
- +
1
tch ---------------------
- +  1 – F ortho   BTS 
CI req  r
DL nonOrtho term
 1 – F ortho   BTS   P tx  ic   r F ortho   BTS  P CCH  ic   r + N 0  LT  r
 ------------------------------------------------------------------------------------ +
1  ----------------------------------------------------------------------------------------------------------------------------
1
-
tch ---------------------- +  1 – F ortho   BTS  tch ---------------------- +  1 – F ortho   BTS 
CI req  r CI req  r

------------------------------------------------------------------------------------------------
- + 1 – F ortho   BTS  r 
 P
DL
 ic  
-----------------------------------------------------------------------------------------------------------------------------------------------------------
-   P DL

DL tx
P tx  ic  – tx  ic 
 1 
tch  ---------------------- +  1 – F ortho   BTS  
 CI req  r 
nonOrtho term
 ----------------------------------------------------------------------------------------------------------------------------
ortho nonOrtho
= P CCH  ic  + P CCH  ic  + -
1
tch ---------------------- +  1 – F ortho  BTS 
CI req  r
nonOrtho term

ortho nonOrtho
P CCH  ic  + P CCH  ic  + ----------------------------------------------------------------------------------------------------------------------------
-
1
tch ---------------------- +  1 – F ortho   BTS 
DL CI req  r
P tx  ic  = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
------------------------------------------------------------------------------------------------- + 1 – F ortho   BTS  r 
 DL
P tx  ic  
------------------------------------------------------------------------------------------------------------------------------------------------------------
1–
  1
---------------------- +  1 – F ortho   BTS 

tch  CI req  r

 
Therefore, the downlink load factor can be expressed as:

- + 1 – F ortho   BTS  r
------------------------------------------------------------------------------------------------
DL
P tx  ic 

DL
X = ------------------------------------------------------------------------------------------------------------------------------------------------------------
1
tch ---------------------- +  1 – F 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).
6.4.4.3.2 Downlink Load Factor per Mobile

Atoll evaluates the downlink load factor for any connected mobile as follows:
DL
DL I tot  ic 
X = --------------------
DL
N tot  ic 
6.4.4.4 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 ).
UL
In this calculation, we assume that the cell UL reuse factor ( F  txi ic  ) is constant.
The result depends on the option used to calculate Nt (Without useful signal or Total noise that you may select in Global
parameters).
Without Useful Signal Option
UL
UL W  P b  k   req
UL
-  ------------------------------------------------------------------------------------------------------------------------
Q req  k  = ----------------------------- UL tx
-
R nominal  k  I intra –  P b  k   req + I extra + I inter – carrier + N 0
UL
Q req  k  = -----------------------------
UL
-  -------------------------------------------------------------------------------
UL UL tx
-
R nominal  k  I intra  F –  P b  k   req + N 0
© Forsk 2009 AT271_TRG_E6 229

UL UL
UL  UL R nominal  k  UL R nominal  k  UL tx
 P b  k   req   1 + Q req  k   ------------------------------ = Q req  k   ------------------------------   I intra  F + N 0 
 W  W
UL UL
UL R nominal  k  UL UL R nominal  k  tx
Q req  k   ------------------------------  I intra  F Q req  k   ------------------------------  N 0
UL W W
 P b  k   req = -------------------------------------------------------------------------------------------- + ------------------------------------------------------------------------
UL UL
UL R nominal  k  UL R nominal  k 
1 + Q req  k   ------------------------------ 1 + Q req  k   ------------------------------
W W
UL
Ec req R nominal  k 
We note  -------  k 
UL
= Q req  k   ------------------------------
 Nt  E – DPDCH W
UL tx
UL I intra  F N0
 P b  k   req = -------------------------------------------------------------
- + -------------------------------------------------------------
-
   
 1   1 
 ---------------------------------------------
req
- + 1  ---------------------------------------------- + 1
  Ec     Ec  req 
  ------ - k    ------ - k 
Nt  E – DPDCH Nt  E – DPDCH
  Pb
UL
As I intra =  k   req , we have:
K
1 1
 -------------------------------------------------------------
- + N 0   --------------------------------------------------------------
UL tx
I intra = I intra  F 
   
K  1  K  1 
 ---------------------------------------------
req
- + 1  ---------------------------------------------
req
- + 1
  Ec -  k     Ec -  k  
  ------
Nt  E – DPDCH    ------
Nt  E – DPDCH 
1
 -------------------------------------------------------------
tx
N0  -
 
K  1 
 ---------------------------------------------
req
- + 1
  Ec -  k  
  ------
I intra = ------------------------------------------------------------------------------------------------
1

UL
1–F  --------------------------------------------------------------
 
K  1 
 ---------------------------------------------
req
- + 1
  Ec - k  
  ------
tx UL
N0  F
I intra = -----------------------------------------------------------------------------------------------
-
1
-------------------------------------------------------------------------------------- – 1
1

UL
F  --------------------------------------------------------------
 
K  1 
 ---------------------------------------------
req
- + 1
  Ec -  k  
  ------
UL
UL I intra + I extra + I inter – carrier I intra  F 1
X = -----------------------------------------------------------------------------------
- = ------------------------------------------
- = --------------------------------------
tx UL 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:
1
 -------------------------------------------------------------
UL UL
X = F  -
 
K  1 
 ---------------------------------------------
req
- + 1
  Ec -  k  
  ------
So, we can conclude that the contribution of one user to the UL load is defined as:
UL UL 1
X k = F  --------------------------------------------------------------
 
 1 
 ---------------------------------------------
req
- + 1
  Ec -  k  
  ------
230 AT271_TRG_E6 © Forsk 2009

Total Noise Option
UL
Q req  k  = -----------------------------
UL
-  -----------------------------------------------------------------------------------
-
tx
R nominal  k  I intra + I extra + I inter – carrier + N 0
UL
UL
-  ------------------------------------------
Q req  k  = ----------------------------- UL tx
-
R nominal  k  I intra  F + N 0
UL
UL UL R nominal  k  UL tx
 P b  k   req = Q req  k   ------------------------------   I intra  F + N 0 
W
UL
Ec req R nominal  k 
We note  -------  k 
UL
= Q req  k   ------------------------------
 Nt  E – DPDCH W
Ec req
 P b  k   req =  -------  k 
UL UL tx
  I intra  F + N 0 
  Pb
UL
As I intra =  k   req , we have:
K
Ec req
  ------
-  k 
UL tx
I intra =  I intra  F + N0  
Nt  E – DPDCH
K
req
 Ec
-------  k 

tx
N0 
K
I intra = --------------------------------------------------------------------
-
UL
1–F
UL
UL I intra + I extra + I inter – carrier I intra  F 1
X = -----------------------------------------------------------------------------------
- = ------------------------------------------
- = --------------------------------------
tx UL 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:
Ec req
  ------
-  k 
UL UL
X = F 
Nt  E – DPDCH
K
So, we can conclude that the contribution of one user to the UL load is defined as:
Ec req
  -------  k 
UL UL
X k = F
6.4.4.5 Inter-carrier Power Sharing Modelling

Inter-carrier power sharing enables the network to dynamically allocate available power from R99-only and HSDPA carri-
ers among HSDPA carriers.
In this part, we will consider the most common scenario, a network consisting of a R99-only carrier (c1) and an HSDPA
carrier with dynamic power allocation (c2) (c2 does not support HSUPA).
As explained in The User Manual, 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. In this case, the HSDPA cell can use 100% of the available power,
i.e, all of the R99-only cell’s unused power can be allocated to the HSDPA cell.
Let’s take the following example to measure the impact of the inter-carrier power sharing.
• 1st case: Inter-carrier power sharing is not activated
On c1, we have: P max  Tx c 1  = 43dBm and P tx – R99  Tx c 1  = 39.1dBm .
On c2, we have: P max  Tx c 2  = 43dBm , P tx – R99  Tx c 2  = 36.1dBm and P Headroom  Tx c 2  = 0dB .
Therefore, 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, we have: P max  Tx c 1  = 43dBm and P tx – R99  Tx c 1  = 39.1dBm .
On c2, we have: P max  Tx c 2  = 46dBm , P tx – R99  Tx c 2  = 36.1dBm and P Headroom  Tx c 2  = 0dB .
Therefore, 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  = 44.4dBm
© Forsk 2009 AT271_TRG_E6 231

6.5 UMTS HSPA Prediction Studies

6.5.1.1 AS Analysis Tab
Let us suppose a receiver with a terminal, a service and a mobility type. This receiver does not create any interference.
You can make the prediction for a specific carrier or for all carriers of the main frequency band for the selected terminal.
If you have selected a dual-band terminal, you can make the coverage prediction on a specific carrier or on all carriers of
any frequency band for the selected terminal, or for all the carriers of all the frequency bands. The analysis is based on
the following parameters:
• The uplink load factor and the downlink total power of cells,
• The cell HSDPA power in case of a HSDPA user,
• The cell UL reuse factor, the cell UL load factor due to HSUPA and the maximum cell UL load factor for HSUPA
users.
These parameters can be results of a given simulation, average values calculated from a group of simulations, or user-
defined cell inputs. In the last case, when no value is defined in the Cells table, Atoll uses the following default values:
• Total transmitted power = 50% of the maximum power (i.e, 40 dBm if the maximum power is set to 43 dBm)
• Uplink load factor = 50%.
• Uplink reuse factor = 1
• Uplink load factor due to HSUPA = 0%
• Maximum uplink load factor = 75%
On the other hand, no default value is used for the HSDPA power; this parameter must be defined by the user.
6.5.1.1.1 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. In this case, for each transmitter i containing the receiver in its calcu-
lation area and using the selected carrier, Atoll calculates the pilot quality at the receiver on this carrier. Then, it determines
the best serving transmitter using the selected carrier ic.
2nd case: Analysis based on all carriers

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. The best carrier selection depends on the option selected for the site
equipment (UL minimum noise, DL minimum power, random, sequential). Then, 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).
3rd case: Analysis based on all carriers of any frequency band (for dual-band terminals with priority defined on frequency
bands only)
The frequency band that can be used is fixed. Atoll determines the best carrier for each transmitter i containing the receiver
in its calculation area and using the selected frequency band. The best carrier selection depends on the option selected
for the site equipment (UL minimum noise, DL minimum power, random, sequential). Then, 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).
Ec/Io (or Q pilot  ic  ) Evaluation
Let us assume that ic is either the best carrier or the selected carrier of a transmitter i containing the receiver in its radius
calculation and icadj is another carrier adjacent to ic. An interference reduction factor, RF  ic ic adj  , is defined between ic
and icadj and set to a value different from 0.
Two ways may be used to calculate Io.
Option Total noise: Atoll considers the noise generated by all the transmitters and the thermal noise.
Option Without pilot: Atoll considers the total noise deducting the pilot signal.
Calculation option may be selected in Global parameters.
Therefore, we have:
 BTS    P c  i ic 
Q pilot  i ic  = -------------------------------------------------
-
DL
I 0  ic 
With,
DL DL DL DL term
I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + N 0 for the total noise option,
And
DL DL DL DL term
I 0  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + N 0 –  1 –     BTS  P c  i ic  for the without pilot option.
232 AT271_TRG_E6 © Forsk 2009

1st step: P c  i ic  calculation for each cell (i,ic)
P c  i ic  is the pilot power of a transmitter i on carrier ic at the receiver.
P pilot  i ic 
P c  i ic  = ---------------------------
-
LT
I
L T is the total loss between transmitter i and receiver.

I
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  , P tot  i ic  and P tot  j ic adj  calculations
We have:

DL DL
I extra  ic  = P tot  j ic 
txj j  i
P SCH  ic 
I intra  ic  = P tot  i ic  –  BTS     P tot  i ic  – -----------------------
DL DL DL
-
 L 
T
And
 Ptot  j icadj 
DL
DL txj j
I inter – carrier  ic  = -------------------------------------------
DL
For each transmitter of the network, P tot  ic  is the total power received at the receiver from the transmitter on the best
carrier ic of the transmitter i.
DL P Tx  ic 
P tot  ic  = ------------------
-
LT
P Tx  ic  is the total power transmitted by the transmitter on the best carrier. 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.
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. This carrier is adjacent to ic.
DL P Tx  ic adj 
P tot  ic adj  = -------------------------
-
LT
P Tx  ic adj  is the total power transmitted by 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.
term
3rd step: N 0 calculation
term
N0 = NF term  K  T  W
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, G macro – diversity , models the decrease in shadowing margin due to the fact there are several
available pilot signals at the mobile.
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).
Note:
• This parameter is determined from cell edge coverage probability and Ec/Io standard
deviation. When the Ec/Io standard deviation is set to 0, the macro-diversity gain equals 0.
6th step: 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
Resulting
coverage probability, Q pilot  ic  .
© Forsk 2009 AT271_TRG_E6 233

Resulting DL
Q pilot  ic  = G macro – diversity  max  Q pilot  i ic  
Resulting req Resulting

If Q pilot  Q pilot , it means pilot quality at the receiver exceeds Q pilot  ic  x% of time (x is the fixed cell edge cover-
age probability). The cell whose Q pilot  i ic  is the highest one enters the active set as best server ( Q pilot  BS ic  ) and
the best carrier (icBS) of the best server, BS, will be the carrier used by other transmitters of the active set (when active
set size is greater than 1). Pilot is available.
Resulting req
If Q pilot  Q pilot , no cell (i,ic) can enter the active set. Pilot is unavailable.
Then, pilot qualities at the receiver from transmitters i (except the best server) on the best carrier of the best server, icBS,
are recalculated to determine the entire receiver active set (when active set size is greater than 1). Same formulas and
DL
calculation method are used to update I 0  ic BS  value and determine Q pilot  i ic BS  .
We have:
 BTS    P c  i ic 
Q pilot  i ic  = -------------------------------------------------
-
DL
I 0  ic 
With,
DL DL DL DL term
I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + N 0 for the total noise option,
And
DL DL DL DL term
I 0  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + N 0 –  1 –     BTS  P c  i ic  for the without pilot option.
Other cells (i,icBS) in the active set must satisfy the following criteria:
Q pilot  i ic BS  – Q pilot  BS ic BS   AS_threshold  i BS ic BS 
 i ic BS   neighbour list  i BS ic BS  (optionally)
Number of Cells in Active Set
This is a user-specified input in the Terminal properties. It corresponds to the active set size.
Thermal Noise
This parameter is calculated as described above (3rd step).
Io (Best Server)
Io (Best server) is the total noise received at the receiver on icBS. The notation “Best server” refers to the best server of
active set. This is relevant when using the calculation option “Without pilot”. In this case, it informs that the pilot signal of
the best server (BS,icBS) is deducted from the total noise.
Downlink Macro-Diversity Gain
This parameter is calculated as described above (5th step).
6.5.1.1.2 Downlink Sub-Menu

The Downlink sub-menu may contain R99-related results and HSDPA-related results when a HSPA user is modelled.
• R99-related Results
Atoll calculates the traffic channel quality from each cell (k,icBS) of the receiver’s active set at the receiver. No power
control is performed as in simulations. Here, Atoll determines the downlink traffic channel quality at the receiver for the
maximum allowed traffic channel power per transmitter. Then, after combination, the total downlink traffic channel quality
is evaluated and compared with the specified target quality.
Eb/Nt Target
DL
Eb/Nt target ( Q req ) is a user-defined parameter for a given service, mobility and reception equipment. This parameter is
available in the R99 Radio Bearers table.
234 AT271_TRG_E6 © Forsk 2009

Notes:
• 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
And
- Either the received Ec/Io is lower than the Ec/Io activation threshold (Global parameters):
Q pilot  Q pilot ,
- Or the pilot RSCP is lower than the pilot RSCP activation threshold (Global parameters):
CM – activation
P c  RSCP pilot .
• When compressed mode is activated, the downlink Eb/Nt target is increased by the value
DL
user-defined for the DL Eb/Nt target increase field (Global parameters), Q req .
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.
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, for each cell (k,icBS), we have:
DL
DL  BTS  P b – max  k ic BS 
-  G DL
Q max  k ic BS  = -------------------------------------------------------------
DL
p  G Div
DL
N tot  ic BS 
max
DL P tch DL DL DL DL term
With P b – max  k ic BS  = ------------
- 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 tch is the maximum power allowed on traffic channel. This parameter is user-defined in the R99 Radio Bearers table.
DL
N tot  ic BS  is the total noise at the receiver on the best carrier of the best server.
DL
I intra  ic BS  is the intra-cell interference at the receiver on the best carrier of the best server.
P SCH  k ic BS 
I intra  ic BS  = P DL  k ic  –  BTS  F ortho   P DL  k ic  – ------------------------------------
DL
-
tot BS  tot BS L 
T
DL
I extra  ic BS  is the extra-cell interference at the receiver on the best carrier of the best server.
 Ptot  j icBS 
DL DL
I extra  ic BS  =
j j  k
DL
I inter – carrier  ic BS  is the inter-carrier interference at the receiver on the best carrier of the best server.
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, defined between ic and icadj and set to a value different from 0.
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.
On downlink, if there is no handoff, we have:

DL DL
Q MAX  ic BS  = Q max  k ic BS 
For any other handoff status, we have:
 Qmax  k icBS 
DL DL DL
Q MAX  ic BS  = f rake efficiency 
k
© Forsk 2009 AT271_TRG_E6 235

Where
DL
f rake efficiency is the downlink rake efficiency factor defined in Terminal properties.
req
3rd step: P tch calculation
DL
req Q req
-  P max
P tch = ------------------------------- tch
DL
Q MAX  ic BS 
Notes:
compressed-mode-capable equipment
And
- Either the received Ec/Io is lower than the Ec/Io activation threshold (Global parameters):
Q pilot  Q pilot .
- Or the pilot RSCP is lower than the pilot RSCP activation threshold (Global parameters):
CM – activation
P c  RSCP pilot
• When compressed mode is activated, the downlink Eb/Nt target is increased by the value
DL
user-defined for the DL Eb/Nt target increase field (Global parameters), Q req . In this
DL DL
req Q req  Q req
-  P max
case, we have: P tch = --------------------------------- tch
DL
Q MAX  ic BS 
Eb/Nt Max for Each Cell of Active Set
For each cell (k,icBS), we have:
DL
-  G DL
Q max  k ic BS  = -------------------------------------------------------------
DL
p  G Div
DL
N tot  ic BS 
max
DL P tch
With P b – max  k ic BS  = ------------
-
LT
k
DL DL DL DL term
N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + N 0
max req
P SCH  k ic BS  P tch – P tch
I intra  ic BS  = P DL  k ic  –  BTS  F ortho   P DL  k ic  – ------------------------------------
DL
- –  1 –  BTS   max (------------------------------ ,0)
tot BS  tot BS L  L
T Tk
DL DL
j j  k
DL
DL txj j
Where
req
P tch is the required transmitter power on traffic channels.
Eb/Nt Max
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.

DL DL
DL DL DL
k
Where
DL
236 AT271_TRG_E6 © Forsk 2009

DL DL DL DL DL
Therefore, 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).
Effective Eb/Nt
DL
Q eff is the effective traffic channel quality at the receiver on icBS.
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).
Downlink Soft Handover Gain
DL
G SHO corresponds to the DL soft handover gain.
DL
DL Q MAX  ic BS 
G SHO = -----------------------------------------------------
-
DL
max  Q max  k ic BS  
DL DL
max  Q max  k ic BS   corresponds to the highest Q max  k icBS  value.
• HSDPA-related Results
Atoll determines the best HSDPA bearer that the user can obtain. The HSDPA user is processed as if he is the only user
in the cell i.e. he uses the entire HSDPA power available in the cell.
For further information on the fast link adaptation modelling, see "Fast Link Adaptation Modelling" on page 209.
HS-PDSCH Ec/Nt
Atoll calculates the best HS-PDSCH quality (HS-PDSCH Ec/Nt). 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 further details on the HS-PDSCH quality calculation, see either "HS-PDSCH Quality Calculation" on page 210 if the
selected option is "CQI based on CPICH quality" or "HS-PDSCH Quality Calculation" on page 213 if the selected option
is "CQI based on HS-PDSCH quality".
HS-SCCH Ec/Nt
When the HS-SCCH power allocation strategy is dynamic, this parameter corresponds to the HS-SCCH Ec/Nt threshold
defined for the selected mobility type.
When the HS-SCCH power allocation strategy is static, the HS-SCCH Ec/Nt is calculated from the fixed HS-SCCH power.
We have:
 BTS  P c  ic 
 Ec
-------  ic  i
N tot  ic 
And
 BTS  P c  ic 
 Ec
-------  ic  i
= --------------------------------------------------------------------------------------------------------------------------------------
i
With
DL DL DL DL term

- –  BTS   P tot  ic  – -----------------------
DL term
-
 LT   LT 

DL DL
txj j  i
DL
DL txj j
-
P c  ic  = --------------------------------------
-
i LT
i
© Forsk 2009 AT271_TRG_E6 237

And

L T = --------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
term term
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.
For further details on the HS-PDSCH quality calculation, see either "HS-PDSCH CQI Determination" on page 211 if the
selected option is "CQI based on CPICH quality" or "HS-PDSCH CQI Determination" on page 215 if the selected option is
"CQI based on HS-PDSCH quality".
RLC Peak Rate
Knowing the HS-PDSCH CQI, Atoll calculates the best HSDPA bearer that can be used and selects the suitable bearer
so as to comply with cell and terminal user equipment HSDPA capabilities. Once the bearer selected, Atoll reads the RLC
peak rate that can be provided to the user.
For further details of the HSDPA bearer selection, see "HSDPA Bearer Selection" on page 211.
6.5.1.1.3 Uplink Sub-Menu

The Uplink sub-menu may contain R99-related results and HSUPA-related results when a HSPA user is modelled.
• R99-related Results
For each cell (k,icBS) in the receiver’s active set, Atoll calculates uplink traffic channel quality from receiver. No power
control is performed as in simulations. Here, Atoll determines the uplink traffic channel quality at the cell for the maximum
terminal power allowed. Then, the total uplink traffic channel quality is evaluated with respect to the receiver handover
status. From this value, Atoll calculates the terminal power required to obtain the R99 bearer and compares it to the maxi-
mum terminal power allowed.
Max Terminal Power
max
Max terminal power ( P term ) is an input user-defined for each terminal. It corresponds to the terminal’s maximum power.
Required Terminal Power
req
The calculation of the terminal power required to obtain a R99 bearer ( P term – R99 ) may be divided into three steps.
UL
1st step: Q max  k ic BS  evaluation for each cell
For each cell (k,icBS) in the receiver’s active set, we have:
UL
UL  term  P b – max  k ic BS 
-  G UL
Q max  k ic BS  = --------------------------------------------------------------
UL
p  G Div
UL
N tot  k ic BS 
max UL
UL P term   1 – r c 
With P b – max  k ic BS  = ------------------------------------------
-
LT
k
UL
N tot  k ic BS  is the total noise at the transmitter on the best carrier of the best server. This value is calculated from the
UL
cell uplink load factor X  k ic BS  .
tx
UL N0
N tot  k ic BS  = -----------------------------------------
-
UL
1 – X  k ic BS 
tx
N 0 is the transmitter thermal noise.

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.
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  
238 AT271_TRG_E6 © Forsk 2009

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. 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
For soft-soft handoffs (3/3):

UL UL UL
UL
For softer and softer-softer handoffs (1/2 and 1/3):
  Qmax  k icBS  
UL UL UL
k
For softer-soft handoffs (2/3), there are two possibilities. If the MRC option is selected (option available in Global param-
eters), we have:
 UL 

UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency   Q max  k ic BS   Q max  k ic BS 
 k on the same site

k on the same site
Else,
UL UL UL
req
3rd step: P term – R99 calculation
req
P term – R99 is the required terminal power.
UL
req Q req
-  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 service and mobility. This parameter is available in
the R99 Radio Bearers table.
Notes:
compressed-mode-capable equipment, and
- The received Ec/Io is lower than the Ec/Io activation threshold (Global parameters):
Q pilot  Q pilot .
- The pilot RSCP is lower than the pilot RSCP activation threshold (Global parameters):
CM – activation
P c  RSCP pilot
• When compressed mode is activated, the uplink Eb/Nt target is increased by the value
UL
user-defined for the UL Eb/Nt target increase field (Global parameters), Q req . In this
UL UL
-  P max
case, we have: P term – R99 = --------------------------------- term
UL
Q MAX  ic BS 
req max
Therefore, the service on the uplink traffic channel is available if P term – R99  P term .
Eb/Nt Max
For each cell (k,icBS) in the receiver’s active set, we have:
UL
UL  term  P b – max  k ic BS 
-  G UL
Q max  k ic BS  = --------------------------------------------------------------
UL
p  G Div
UL
N tot  k ic BS 
max UL
UL P term   1 – r c 
With P b – max  k ic BS  = ------------------------------------------
-
LT
k
© Forsk 2009 AT271_TRG_E6 239

UL
N tot  k ic BS  is the total noise at the transmitter on the best carrier of the best server. This value is calculated from the
UL
cell uplink load factor X  k ic BS  .
tx max req
UL N0 P term – P term – R99
- +  1 –  term   max (----------------------------------------------
N tot  k ic BS  = ----------------------------------------- - , 0)
UL
1 – X  k ic BS  LT
k
tx
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.
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
UL
UL UL
max  Q max  k ic BS   corresponds to the highest Q max  k ic BS  value.
For soft-soft handoffs (3/3):

UL UL UL
UL
For softer and softer-softer handoffs (1/2 and 1/3):
  Qmax  k icBS  
UL UL UL
k
For softer-soft handoffs (2/3), there are two possibilities. If the MRC option is selected (option available in Global param-
eters), we have:
 UL 

UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency   Q max  k ic BS   Q max  k ic BS 
 k on the same site

k on the same site
Else,
UL UL UL
Effective Eb/Nt
UL
Q eff is the effective traffic channel quality at the transmitter on icBS.
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).
Uplink Soft Handover Gain
UL
G SHO corresponds to the uplink soft handover gain.
UL
UL Q MAX  ic BS 
G SHO = -----------------------------------------------------
-
UL
UL UL
• HSUPA-related Results
Atoll determines the best HSUPA bearer that the user can obtain. The HSUPA user is processed as if he is the only user
in the cell i.e. he uses the entire remaining load of the cell.
For further information on the HSUPA bearer selection, see "HSUPA Bearer Allocation Process" on page 218.
240 AT271_TRG_E6 © Forsk 2009

Required E-DPDCH Ec/Nt
Ec req
It corresponds to the E-DPDCH Ec/Nt required to obtain the HSUPA bearer (  ------- ). This value is defined for a
HSUPA bearer ( Index HSUPABearer ) and a certain number of retransmissions ( N Rtx ) in the HSUPA Bearer Selection
table.
Required Terminal Power
Ec req
From  -------
req
, Atoll calculates the terminal power required to obtain the HSUPA bearer, P term – HSUPA .
Ec req
P term – HSUPA =  -------
req UL
 L T  N tot
With
UL intra UL extra tx
UL tx UL
N tot  ic  =  1 – F MUD   term   I tot  ic  + I tot  ic  + I inter – carrier  ic  + N 0

UL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
tx UL intra UL extra UL tx
 term , F MUD , I tot , I tot , I inter – carrier and N 0 are defined in "Inputs" on page 186.
RLC Peak Rate
Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. This is the HSUPA bearer with the highest po-
UL
R RLC – peak  Index HSUPABearer 
tential throughput ( ------------------------------------------------------------------------------------ ) where:
N Rtx  Index HSUPABearer 
req Ec max
•  Ec
-------   -------
 Nt  E – DPDCH  Nt  E – DPDCH
req max
• And P term – HSUPA  P term
With
max
 Ec
------- : the maximum E-DPDCH Ec⁄Nt allowed.
max
P term : the maximum terminal power allowed.
UL
After selecting the HSUPA bearer, Atoll reads the corresponding RLC peak rate, R RLC – peak  Index HSUPABearer  .
Application Throughput
UL
Atoll displays the provided application throughput ( T application ). The application throughput represents the net throughput
after deduction of coding (redundancy, overhead, addressing, etc.). This one is calculated as follows:
UL
UL R RLC –p eak   1 – BLER HSUPA   SF Rate – R
T application  M b  = -------------------------------------------------------------------------------------------------------------------------
-
N Rtx
Where:
Knowing the E-DPDCH Ec/Nt, Atoll finds the corresponding BLER.

Let us assume each pixel on the map corresponds to a probe receiver with a terminal, a mobility type and a service. This
receiver does not create any interference. You can make the coverage prediction for a specific carrier or for all carriers of
the main frequency band for the selected terminal. If you have selected a dual-band terminal, you can make the coverage
prediction on a specific carrier or on all carriers of any frequency band for the selected terminal, or for all the carriers of all
the frequency bands. Coverage predictions are based on parameters that can be either simulation results, or user-defined
cell inputs.
© Forsk 2009 AT271_TRG_E6 241

6.5.2.1 Pilot Reception Analysis

For further details of calculation formulas and methods, please refer to Definitions and formulas part, and Point analysis
– AS analysis tab – Pilot sub-menu part.
We consider the following cases:
1st case: Analysis Based on a Specific Carrier

lation area and using the selected carrier, Atoll calculates the pilot quality at the receiver on this carrier icgiven. Then, it
determines the best serving transmitter BS using the carrier icgiven ( Q pilot  ic given  ) and calculates the best pilot quality
BS
Resulting
received with a fixed cell edge coverage probability, Q pilot  ic given  .
Atoll displays the best pilot quality received with a fixed cell edge coverage probability.
2nd case: Analysis Based on All Carriers

Atoll proceeds as in point predictions. It determines the best carrier of each transmitter i containing the receiver in its calcu-
lation area and using a frequency band supported by the receiver’s terminal. The best carrier selection depends on the
option selected for the site equipment (UL minimum noise, DL minimum power, random, sequential) and is based on the
UL load percentage and the downlink total power of cells (simulation results or cell properties). Atoll calculates the pilot
quality at the receiver from these transmitters on their best carriers and determines the best serving transmitter BS on its
best carrier icBS ( Q pilot  ic BS  ). Then, it calculates the best pilot quality received with a fixed cell edge coverage proba-
BS
Resulting
bility, Q pilot  ic BS  .
3rd case: Analysis based on all carriers of any frequency band (for dual-band terminals with priority defined on frequency
bands only)
The frequency band that can be used is fixed. Atoll determines the best carrier of each transmitter i containing the receiver
in its calculation area and using the selected frequency band. The best carrier selection depends on the option selected
for the site equipment (UL minimum noise, DL minimum power, random, sequential) and is based on the UL load percent-
age 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 carriers and determines the best serving transmitter BS on its best carrier
icBS ( Q pilot  ic BS  ). Then, it calculates the best pilot quality received with a fixed cell edge coverage probability,
BS
Resulting
Q pilot  ic BS  .
6.5.2.1.1 Prediction Study Inputs

The Pilot Reception Analysis depends on the downlink total transmitted power of cells. This parameter can be either a
simulation output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the total
transmitted power, Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power is set
to 43 dBm).
6.5.2.1.2 Study Display Options

Single colour
Resulting req
Atoll displays a coverage if Q pilot  ic   Q pilot . Coverage consists of a single layer with a unique colour
req
( ic = ic BS or ic given ). Q pilot is a target value defined in the Mobility table by the user.
Colour per transmitter
Resulting req
Atoll displays a coverage if Q pilot  ic   Q pilot ( ic = ic BS or ic given ). Coverage consists of several layers with asso-
ciated colours. There is a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to
the best serving transmitter BS.
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned.

Coverage consists of several layers with a layer per user-defined mobility defined in Mobility sub-folder. For each layer,
Resulting req
area is covered if Q pilot  ic   Q pilot ( ic = ic BS or ic given ). Each layer is assigned a colour and displayed with inter-
sections between layers.
Colour per probability
This display option is available only if analysis is based on all simulations in a group (i.e. if you select a group of simulations
and the “All” option in the Simulation tab of prediction properties). Coverage consists of several layers with a layer per
242 AT271_TRG_E6 © Forsk 2009

user-defined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting req
Q pilot  ic   Q pilot ( ic = ic BS or ic given ) in the required number of simulations. Each layer is assigned a colour and
displayed with intersections between layers.
Colour per cell edge coverage probability
Coverage consists of several layers with a layer per user-defined cell edge coverage probability, p, defined in the Display
Resulting req
tab (Prediction properties). For each layer, area is covered if Q pilot  ic p   Q pilot ( ic = ic BS or ic given ). Each layer
is assigned a colour and displayed with intersections between layers.
Colour per quality level (Ec/Io)
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
Resulting
properties). For each layer, area is covered if Q pilot  ic    Q pilot  threshold ( ic = ic BS or ic given ). Each layer is
assigned a colour and displayed with intersections between layers.
Colour per quality margin (Ec/Io margin)
Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction prop-
Resulting req
erties). For each layer, area is covered if Q pilot  ic  – Q pilot   Q pilot  m arg in ( ic = ic BS or ic given ). Each layer is
6.5.2.2 Downlink Service Area Analysis

As in point predictions, Atoll calculates traffic channel quality at the receiver for each cell (k,ic) (with ic=icBS or icgiven) in
the receiver’s active set. No power control is performed as in simulations. Here, Atoll determines downlink traffic channel
quality at the receiver for a maximum allowed traffic channel power for transmitters. Then, the total downlink traffic channel
DL
quality ( Q MAX  ic  ) is evaluated after recombination.
Note:
• Best server and active set determination is performed as in point prediction (AS analysis).
Atoll displays traffic channel quality at the receiver for transmitters in active set on the carrier ic ( ic BS or ic given ).
For further details of calculation formulas and methods, see "Downlink Sub-Menu" on page 234.

The Downlink Service Area Analysis depends on the downlink total transmitted power of cells. This parameter can be
either a simulation output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the
total transmitted power, Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power
is set to 43 dBm).

Single colour
DL DL DL DL DL
Atoll displays a coverage with a unique colour if Q MAX  ic   Q req (or Q MAX  ic   Q req  Q req if compressed mode is
activated).
DL
Q req is a user-defined parameter for a service and mobility. It is available in the R99 Radio Bearers table.
DL
Q req is the DL Eb/Nt target increase; this parameter is user-defined in the Global parameters.
DL DL DL DL DL
Atoll displays a coverage if Q MAX  ic   Q req (or Q MAX  ic   Q req  Q req if compressed mode is activated). Coverage
consists of several layers with associated colours. There is a layer per transmitter with no intersection between layers.
Layer colour is the colour assigned to best serving transmitter.
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned. Coverage consists of several layers with a
DL DL
layer per user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if Q MAX  ic   Q req (or
DL DL DL
Q MAX  ic   Q req  Q req if compressed mode is activated). Each layer is assigned a colour and displayed with intersec-
tions between layers.
© Forsk 2009 AT271_TRG_E6 243

Colour per service3
In this case, receiver is not completely defined and no service is assigned. Coverage consists of several layers with a layer
DL DL
per user-defined service defined in Services sub-folder. For each layer, area is covered if Q MAX  ic   Q req (or
DL DL DL
DL DL
Q MAX  ic   Q req in the required number of simulations. Each layer is assigned a colour and displayed with intersections
between layers.
Colour per cell edge coverage probability
DL DL DL DL DL
tab (Prediction properties). For each layer, area is covered if Q MAX  ic p   Q req (or Q MAX  ic   Q req  Q req if
compressed mode is activated). Each layer is assigned a colour and displayed with intersections between layers.
Colour per maximum quality level (max Eb/Nt)
DL
properties). For each layer, area is covered if Q MAX  ic   Threshold . Each layer is assigned a colour and displayed with
intersections between layers.
Colour per effective quality level (Effective Eb/Nt)
DL
properties). For each layer, area is covered if Q eff  ic   Threshold . Each layer is assigned a colour and displayed with
DL DL DL DL DL DL DL
intersections between layers. Q eff  ic  = min  Q MAX  ic  Q req  (or Q eff  ic  = min  Q MAX  ic  Q req  Q req  when
compressed mode is activated).
Colour per quality margin (Eb/Nt margin)
DL DL DL DL DL
erties). For each layer, area is covered if Q MAX  ic  – Q req  M arg in (or Q MAX  ic  – Q req  Q req  M arg in when
3. Service oriented studies (effective service area, DL or UL service area analysis) with per service display are based
on a calculation and display optimisation method. Atoll considers that a calculation pixel covered for the upper service of
the list is automatically covered for the lower services. Sometimes, this optimisation is not possible. In such a case, the
composite coverage is reliable but single service coverage layers might be incorrect. Atoll detects when this optimisation
may involve errors. In this case, it advises the user to perform a study for each service to get a reliable service coverage.
Different cases where optimisation method does not work correctly are detailed below:
- Any uplink or downlink quality study must be performed for each service if handover is not available for all the services
and the service order is different from the one found when considering handover for all the services. A pixel could be
covered by a service requiring a high quality target, due to handover (Eb/Nt combination), while it would not be covered
by a service requiring a lower quality target but not allowing handover.
UL
For uplink, services are sorted according to a decreasing quality indicator ( I Q ):
UL
UL Q req
I Q = -----------
-
UL
Gp
DL
For downlink, services are sorted according to a decreasing quality indicator ( I Q ):
DL
DL Q req
I Q = -----------------------------
-
DL max
G p  P tch
- Effective service area study must be performed for each service if uplink and downlink orders of services (explained
above) are not the same.
244 AT271_TRG_E6 © Forsk 2009

Colour per required power
req
Atoll calculates the downlink required power, P tch  ic  , as follows:
DL
req Q req
-  P max
P tch  ic  = ------------------------ tch
DL
Q MAX  ic 
Where
DL
Q req is the Eb/Nt target on downlink. This parameter, available in the R99 Radio Bearers table, is user-defined for given
service and mobility.
max
P tch is a user-defined input for each bearer related to a service. It corresponds to the maximum allowable traffic channel
power for a transmitter.
DL DL
-  P max
When compressed mode is activated, we have: P tch  ic  = --------------------------------- tch .
DL
Q MAX  ic 
Coverage consists of several layers with a layer per user-defined required power threshold defined in the Display tab
req
(Prediction properties). For each layer, area is covered if P tch  ic   Threshold . Each layer is assigned a colour and
Colour per required power margin
Coverage consists of several layers with a layer per user-defined power margin defined in the Display tab (Prediction prop-
req max
erties). For each layer, area is covered if P tch  ic  – P tch  M arg in . Each layer is assigned a colour and displayed with
6.5.2.3 Uplink Service Area Analysis

As in point prediction, Atoll calculates uplink traffic channel quality from receiver for each cell (k,ic) (with ic=icBS or icgiven)
in receiver active set. No power control simulation is performed. Atoll determines uplink traffic channel quality at the trans-
UL
mitter for the maximum terminal power allowed. Then, the total uplink traffic channel quality ( Q MAX  ic  ) is evaluated with
respect to receiver handover status.
Note:
Atoll displays traffic channel quality at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the
receiver.
For further details of calculations formulas and methods, see "Uplink Sub-Menu" on page 238.

The Uplink Service Area Analysis depends on the UL load factor of cells. This parameter can be either a simulation output,
or a user-defined cell input. In the last case, when no value is defined in the Cells table for the uplink load factor, Atoll uses
50% as default value.

Single colour
UL UL UL UL UL
colour is unique.
UL
Q req is a user-defined parameter for a service and mobility. It is available in the R99 Radio Bearers table.
UL
Q req is the UL Eb/Nt target increase; this parameter is user-defined in the Global parameters.
UL UL UL UL UL
consists of several layers with associated colours. There is a layer per transmitter with no intersection between layers.
Layer colour is the colour assigned to best server transmitter.
© Forsk 2009 AT271_TRG_E6 245

Colour per mobility
UL UL
layer per user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if Q MAX  ic   Q req (or
UL UL UL
Colour per service
UL UL
per user-defined service defined in Services sub-folder. For each layer, area is covered if Q MAX  ic   Q req (or
UL UL UL
UL UL UL UL UL
Q MAX  ic   Q req (or Q MAX  ic   Q req  Q req if compressed mode is activated) in the required number of simulations.
Each layer is assigned a colour and displayed with intersections between layers.
Colour per maximum quality level (Max Eb/Nt)
UL
Colour per effective quality level (Effective Eb/Nt)
UL
properties). For each layer, area is covered if Q effective  ic   Threshold . Each layer is assigned a colour and displayed
with intersections between layers.
UL UL UL UL UL UL UL
Q eff  ic  = min  Q MAX  ic  Q req  (or Q eff  ic  = min  Q MAX  ic  Q req  Q req  when compressed mode is activated).
Colour per quality margin (Eb/Nt margin)
UL UL UL UL UL
erties). For each layer, area is covered if Q MAX  ic  – Q req  M arg in (or Q MAX  ic  – Q req  Q req  M arg in if
Colour per required power
Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction
req
properties). For each layer, area is covered if P term – R99  ic   Threshold . Each layer is assigned a colour and displayed
Colour per required power margin
req max
erties). For each layer, area is covered if P term – R99  ic  – P term  M arg in . Each layer is assigned a colour and displayed
Colour per soft handover gain
Coverage consists of several layers with a layer per soft handover gain value defined in the Display tab (Prediction prop-
UL
erties). For each layer, area is covered if G SHO  Threshold . Each layer is assigned a colour and displayed with inter-
6.5.2.4 Downlink Total Noise Analysis

Atoll determines the downlink total noise generated by cells.
246 AT271_TRG_E6 © Forsk 2009

DL
 Ptot  ic  + txj--------------------------------------
DL DL  j term
N tot  ic  = - + N0
txj j
DL
Downlink noise rise, NR DL  ic  , is calculated from the downlink total noise, N tot , as follows:
term
 N0 
NR DL  ic  = – 10 log  -------------
-
 N DL tot

6.5.2.4.1 Study Inputs

The Downlink Total Noise Analysis depends on the downlink total transmitted power of cells. This parameter can be either
a simulation output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the total
transmitted power, Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power is set
to 43 dBm).
6.5.2.4.2 Analysis on All Carriers

If all the carriers are selected, Atoll determines DL total noise for all the carriers. Then, allows the user to choose different
colours.
Colour per minimum noise level
Coverage consists of several layers with a layer per user-defined noise level defined in the Display tab (Prediction prop-
DL
erties). For each layer, area is covered if minN tot ic   Threshold . Each layer is assigned a colour and displayed with
ic
Colour per maximum noise level
DL
erties). For each layer, area is covered if maxN tot ic   Threshold . Each layer is assigned a colour and displayed with
ic
Colour per average noise level
DL
erties). For each layer, area is covered if averageN tot ic   Threshold . Each layer is assigned a colour and displayed
ic
Colour per minimum noise rise
Atoll displays bins where minNR DL ic   Threshold . Coverage consists of several areas with an area per user-defined
ic
noise rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
Colour per maximum noise rise
Atoll displays bins where maxNR DL ic   Threshold . Coverage consists of several areas with an area per user-defined
ic
Colour per average noise rise
Atoll displays bins where averageNR DL ic   Threshold . Coverage consists of several areas with an area per user-
ic
defined noise rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
6.5.2.4.3 Analysis on a Specific Carrier

When only one carrier is analysed, Atoll determines DL total noise or DL noise rise on this carrier. In this case, the
displayed coverage is the same for any selected display per noise level (average, minimum, maximum) or any display per
noise rise (average, minimum, maximum).
Colour per noise level
DL
erties). For each layer, area is covered if N tot  ic   Threshold . Each layer is assigned a colour and displayed with inter-
© Forsk 2009 AT271_TRG_E6 247

Colour per noise rise
Atoll displays bins where NR DL  ic   Threshold . Coverage consists of several areas with an area per user-defined noise
rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
6.5.2.5 HSDPA Prediction Study

When calculating the HSDPA coverage prediction, either you can take all the possible HSDPA radio bearers into consid-
eration, or you can study a certain HSDPA radio bearer. Then, available display options depend on what you have
selected.
When considering all the HSDPA radio bearers, you can set display parameters:
• To analyse the uplink and downlink A-DPCH qualities on the map,
• To analyse the HS-SCCH quality/power,
• To model fast link adaptation for a single HSDPA user or for a defined number of HSDPA users.
When studying a certain HSDPA radio bearer, you can display areas where a certain RLC peak rate is available with differ-
ent cell edge coverage probabilities (i.e. the probability of having a certain RLC peak rate).
Let us assume each pixel on the map corresponds to one or several users with HSDPA capable terminal, mobility and
HSDPA service. Each user may be using a specific carrier or all of them. Moreover, he does not create any interference.
Note that the HSDPA service area is limited by the pilot quality and the A-DPCH quality.

Parameters used as input for the HSDPA prediction study are:
• The cell HSDPA power,
• The downlink total transmitted power of the cell,
• The number of HSDPA users within the cell if the study is calculated for several users.
These parameters can be either simulation outputs, or user-defined cell inputs. In the last case, when no value is defined
in the Cells table for the total transmitted power and the number of HSDPA users, Atoll uses the following default values:
• Total transmitted power = 50% of the maximum power (i.e, 40 dBm if the maximum power is set to 43 dBm)
• Number of HSDPA users = 1
On the other hand, no default value is used for the HSDPA power; this parameter must be defined by the user.

When considering all the HSDPA radio bearers, several display options are available in the study properties dialogue.
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,
• To analyse the HS-SCCH quality/power,
• To model fast link adaptation for a single HSDPA user
• To model fast link adaptation for a defined number of HSDPA users.
When studying a certain HSDPA radio bearer, only one display option is available. It allows you to display where a certain
RLC peak rate is available with different cell edge coverage probabilities.
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  ic  ) for the best server on the carrier ic ( ic BS or ic given ). No
power control is performed as in simulations. Here, Atoll determines downlink traffic channel quality at the receiver for a
maximum traffic channel power allowed for the best server.
For further details of calculation formulas and methods, please refer to Prediction studies: Point analysis – AS analysis tab
– Downlink sub-menu part.
DL
• Colour per Max A-DPCH Eb/Nt UL
UL
Atoll displays the A-DPCH quality at the best server ( Q MAX  ic  ) on the carrier ic ( ic BS or ic given ). No power control is
performed as in simulations. Here, Atoll determines uplink traffic channel quality at the receiver for a maximum terminal
power allowed.
For further details of calculations formulas and methods, please refer to Point analysis – AS analysis tab – Uplink sub-
menu part.
UL
248 AT271_TRG_E6 © Forsk 2009

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. In this case, Atoll displays on each pixel
the HS-SCCH power per HS-SCCH channel. Coverage consists of several layers with a layer per threshold. For each
layer, area is covered if P HS – SCCH  ic   Threshold . Each layer is assigned a colour and displayed with intersections
between layers.
• Colour per HS-SCCH Ec/Nt
This display option is relevant in case of static HS-SCCH power allocation only. In this case, Atoll displays on each pixel
the HS-SCCH quality per HS-SCCH channel. Coverage consists of several layers with a layer per threshold. For each
Ec
layer, area is covered if  -------  ic   Threshold . Each layer is assigned a colour and displayed with intersections
 Nt  HS – SCCH
between layers.
Fast Link Adaptation Modelling For A Single User
When you calculate the study with the following display options, Atoll considers one HSDPA user on each pixel and deter-
mines the best HSDPA bearer that the user can obtain. On each pixel, the HSDPA user is processed as if he is the only
user in the cell i.e. he uses the entire HSDPA power available in the cell.
For further information on the fast link adaptation modelling, see "Fast Link Adaptation Modelling" on page 209.
• Colour per HS-PDSCH Ec/Nt
Atoll displays on each pixel the HS-PDSCH quality. Coverage consists of several layers with a layer per threshold. For
Ec
each layer, area is covered if  -------  ic   Threshold . Each layer is assigned a colour and displayed with inter-
• Colour per CQI
Atoll displays either the CPICH CQI (see the calculation detail in "CPICH CQI Determination" on page 210) when the
selected option in Global parameters (HSDPA part) is CQI based on CPICH quality, or the HS-PDSCH CQI (see the calcu-
lation detail in the section 10.7.1.2.2) when considering the CQI based on HS-PDSCH quality option.
Coverage consists of several layers with a layer per CQI threshold (  CQI  threshold ). For each layer, area is covered if
CQI   CQI  threshold . Each layer is assigned a colour and displayed with intersections between layers.
• Colour per MAC Rate

DL
Atoll displays the MAC rate ( R MAC ) provided on each pixel. The MAC rate is calculated as follows:
DL S block
R MAC = ---------------
-
T TTI
Where,
S block is the transport block size (in kbits) of the selected HSDPA bearer; it is defined for each HSDPA bearer in the
related table.
–3
DL
Coverage consists of several layers with a layer per possible MAC rate ( R MAC ). For each layer, area is covered if the
MAC rate exceeds the user-defined thresholds. Each layer is assigned a colour and displayed with intersections between
layers.
• Colour per MAC Throughput
DL
Atoll displays the MAC throughput ( T MAC ) provided on each pixel. The MAC throughput is calculated as follows:
DL S block
T MAC = ------------------------------
-
T TTI   TTI
Where,
S block is the transport block size (in kbits) of the selected HSDPA bearer; it is defined for each HSDPA bearer in the
HSDPA Radio Bearers table.
–3
© Forsk 2009 AT271_TRG_E6 249

DL
Coverage consists of several layers with a layer per possible MAC throughput ( T MAC ). For each layer, area is covered if
the MAC throughput exceeds the user-defined thresholds. Each layer is assigned a colour and displayed with intersections
between layers.
• Colour per RLC Peak Rate
After selecting the bearer, Atoll reads the corresponding RLC peak rate. This is the highest rate that the bearer can provide
on each pixel.
DL
Coverage consists of several layers with a layer per possible RLC peak rate ( R RLC –p eak ). For each layer, area is covered
if the RLC peak rate can be provided. Each layer is assigned a colour and displayed with intersections between layers.
• Colour per RLC Peak Throughput
DL
Atoll displays the RLC peak throughput ( T RLC –p eak ) provided on each pixel. The RLC peak throughput is calculated as
follows:
DL
DL R RLC –p eak
T RLC –p eak = ---------------------------
TTI
Where TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the termi-
nal user equipment category properties.
DL
Coverage consists of several layers with a layer per possible RLC peak throughput ( T RLC –p eak ). For each layer, area is
covered if the RLC peak throughput exceeds the user-defined thresholds. Each layer is assigned a colour and displayed
• Colour per Average RLC Throughput
DL
Atoll displays the average RLC throughput ( T RLC – Av ) provided on each pixel. The average RLC throughput is calculated
as follows:
DL
DL R RLC –p eak   1 – BLER HSDPA 
T RLC – Av = ---------------------------------------------------------------------------------
-
TTI
Where,
the measured quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll finds the corresponding BLER.
DL
Coverage consists of several layers with a layer per possible average RLC throughput ( T RLC – Av ). For each layer, area
is covered if the average RLC throughput exceeds the user-defined thresholds. Each layer is assigned a colour and
• Colour per Application Throughput
DL
Atoll displays the application throughput ( T application ) provided on each pixel. The application throughput represents the
net throughput after deduction of coding (redundancy, overhead, addressing, etc.). This one is calculated as follows:
DL
DL R RLC –p eak   1 – BLER HSDPA   SF Rate – R
T application = -------------------------------------------------------------------------------------------------------------------------
-
TTI
Where:
the measured quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll finds the corresponding BLER.
DL
Coverage consists of several layers with a layer per possible application throughput ( T application ). For each layer, area is
covered if the application throughput exceeds the user-defined thresholds. Each layer is assigned a colour and displayed
250 AT271_TRG_E6 © Forsk 2009

Fast Link Adaptation Modelling For Several Users
When you calculate the study with the following display options, Atoll considers several HSDPA users per pixel and deter-
mines the best HSDPA bearer that each user can obtain. In this case, the cell HSDPA power is shared between HSDPA
users. When the coverage prediction is not based on a simulation, the number of HSDPA users is taken from the cell prop-
erties. The displayed results of the coverage prediction will be an average result for one user.
For further information on the HSDPA bearer allocation process when there are several users, see "HSDPA Bearer Allo-
cation Process" on page 208 For further information on the fast link adaptation modelling, see "Fast Link Adaptation
Modelling" on page 209.
• Colour per MAC Throughput Per Mobile
DL
Atoll displays the average MAC throughput per mobile (  T MAC average ) provided on each pixel. The average MAC
throughput per mobile is calculated as follows:
n HSDPA

DL
T MAC  x 
DL x=1
 T MAC  average = ---------------------------------------
-
n HSDPA
Where,
n HSDPA is the number of HSDPA users within the cell.
DL
T MAC  x  is the MAC throughput of each HSDPA user. For further information on the calculation of the MAC throughput,
see "Colour per MAC Throughput" on page 249.
DL
Coverage consists of several layers with a layer per possible average MAC throughput per mobile (  T MAC  average ). For
each layer, area is covered if the average MAC throughput per mobile exceeds the user-defined thresholds. Each layer is
• Colour per RLC Throughput Per Mobile
DL
Atoll displays the average RLC throughput per mobile (  T RLC  average ) provided on each pixel. The average RLC through-
put per mobile is calculated as follows:
n HSDPA

DL
T RLC –p eak  x 
DL x=1
 T RLC  average = ----------------------------------------------------
n HSDPA
Where,
DL
T RLC –p eak  x  is the RLC peak throughput of each HSDPA user. For further information on the calculation of the RLC peak
throughput, see "Colour per RLC Peak Throughput" on page 250.
DL
Coverage consists of several layers with a layer per possible average RLC throughput per mobile (  T RLC  average ). For
each layer, area is covered if the average RLC throughput per mobile exceeds the user-defined thresholds. Each layer is
• Colour per ApplicationThroughput Per Mobile
DL
Atoll displays the average application throughput per mobile (  T application  average ) provided on each pixel. The average
application throughput per mobile is calculated as follows:
n HSDPA

DL
T application  x 
DL x=1
 T application  average = -----------------------------------------------------
-
n HSDPA
Where,
DL
T application  x  is the application throughput of each HSDPA user. For further information on the calculation of the appli-
cation throughput, see "Colour per Application Throughput" on page 250.
Coverage consists of several layers with a layer per possible average application throughput per mobile
DL
(  T application average ). For each layer, area is covered if the average application throughput per mobile exceeds the user-
defined thresholds. Each layer is assigned a colour and displayed with intersections between layers.
© Forsk 2009 AT271_TRG_E6 251

Probability of Having a Certain RLC Peak Rate
This result can be obtained only if you have selected an HSDPA radio bearer in the Condition tab.
• Colour per Cell Edge Coverage Probability
Atoll shows areas where the selected HSDPA radio bearer is available with different cell edge coverage probabilities.
Coverage consists of several layers with a layer per cell edge coverage probability defined in the Display tab. For each
layer, area is covered if the selected HSDPA radio bearer is available. Each layer is assigned a colour and displayed with
6.5.2.6 HSUPA Prediction Study

A dedicated HSUPA study is available with different calculation and display options. Atoll determines on each pixel the
best HSUPA bearer that can be obtained; it can consider either a single HSUPA user or several ones on each pixel. For
further information on the HSUPA bearer selection, see "HSUPA Bearer Allocation Process" on page 218. By caclulating
this study with suitable display options, it is possible:
• To analyse the power required by the selected terminal,
• To analyse the required E-DPDCH quality,
• To analyse rates and throughputs.
Let us assume each pixel on the map corresponds to one or several users with HSUPA capable terminal, mobility and
HSUPA service. Each user may be using a specific carrier or all of them. Moreover, he does not create any interference.

Parameters used as input for the HSUPA prediction study are:
• The cell UL load factor,
• The cell UL reuse factor,
• The cell UL load factor due to HSUPA,
• The maximum cell UL load factor,
• The number of HSUPA users within the cell if the study is calculated for several users.
These parameters can be either simulation outputs, or user-defined cell inputs. In the last case, When no value is defined
in the Cells table, Atoll uses the following default values:
• Uplink load factor = 50%
• Uplink reuse factor = 1
• Uplink load factor due to HSUPA = 0%
• Maximum uplink load factor = 75%
• Number of HSUPA users = 1
6.5.2.6.2 Calculation Options

Atoll can calculate the HSUPA coverage prediction in one of two ways:
• HSUPA resources can be dedictated to a single user: On each pixel, the HSUPA user is processed as if he is
the only user in the cell i.e he will use the entire remaining load after allocating capacity to all R99 users.
• HSUPA resources can be shared by HSUPA users defined or calculated per cell: Atoll considers several
HSUPA users per pixel. After allocating capacity to all R99 users, the remaining load of the cell will be shared
equally between all the HSUPA users. When the coverage prediction is not based on a simulation, the number of
HSUPA users is taken from the cell properties. The displayed results of the coverage prediction will be an average
result for one user.
6.5.2.6.3 Display Options

The following display options are available in the prediction property dialogue.
Colour per Required E-DPDCH Ec/Nt
Atoll displays on each pixel the E-DPDCH Ec/Nt required to obtain the selected HSUPA bearer. Coverage consists of
Ec req
several layers with a layer per threshold. For each layer, area is covered if  -------  Threshold . Each layer is
Colour per Required Terminal Power
Atoll displays on each pixel the terminal power required to obtain the selected HSUPA bearer. The required terminal power
is calculated from the required E-DPDCH Ec/Nt. Coverage consists of several layers with a layer per threshold. For each
req
layer, area is covered if P term  Threshold . Each layer is assigned a colour and displayed with intersections between
layers.
Colour per MAC Rate
UL
Atoll displays the MAC rate ( R MAC ) provided on each pixel. The MAC rate is calculated as follows:
252 AT271_TRG_E6 © Forsk 2009

UL
UL S block
R MAC = ---------------
-
T TTI
Where,
UL
S block is the transport block size (in kbits) for the selected HSUPA bearer; 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; it is defined for each HSUPA bearer in the HSUPA Radio
Bearers table.
UL
Coverage consists of several layers with a layer per possible MAC rate ( R MAC ). For each layer, area is covered if the
MAC rate exceeds the user-defined thresholds. Each layer is assigned a colour and displayed with intersections between
layers.
Colour per RLC Peak Rate
After selecting the HSUPA bearer, Atoll reads the corresponding RLC peak rate. This is the highest rate that the selected
HSUPA bearer can provide on each pixel.
UL
Coverage consists of several layers with a layer per possible RLC peak rate ( R RLC –p eak ). For each layer, area is covered
if the RLC peak rate can be provided. Each layer is assigned a colour and displayed with intersections between layers.
Colour per Minimum RLC Throughput
UL
Atoll displays the minimum RLC throughput ( T RLC – Min ) provided on each pixel. The minimum RLC throughput is calcu-
lated as follows:
UL
UL R RLC –p eak   1 – BLER HSUPA 
T RLC – Min = ---------------------------------------------------------------------------------
-
N Rtx
Where,
N Rtx is the number of retransmissions for the selected HSUPA bearer. This figure is read in the HSUPA Bearer Selection
table.
DL
Coverage consists of several layers with a layer per possible average RLC throughput ( T RLC – Av ). For each layer, area
is covered if the minimum RLC throughput exceeds the user-defined thresholds. Each layer is assigned a colour and
Colour per Application Throughput
UL
Atoll displays the application throughput ( T application ) provided on each pixel. The application throughput represents the
net throughput after deduction of coding (redundancy, overhead, addressing, etc.). This one is calculated as follows:
UL
UL R RLC –p eak   1 – BLER HSUPA   SF Rate – R
T application  M b  = -------------------------------------------------------------------------------------------------------------------------
-
N Rtx
Where:
N Rtx is the number of retransmissions for the selected HSUPA bearer. This figure is read in the HSUPA Bearer Selection
table.
UL
Coverage consists of several layers with a layer per possible application throughput ( T application ). For each layer, area is
covered if the application throughput exceeds the user-defined thresholds. Each layer is assigned a colour and displayed
© Forsk 2009 AT271_TRG_E6 253

6.6 Automatic Neighbour Allocation

Atoll permits the automatic allocation of intra-technology neighbours in the current network. Two allocation algorithms are
available, one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours.
The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all
the cells of TBC transmitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfil following conditions:
Only TBA cells may be assigned neighbours.
Note:

We assume a reference cell A and a candidate neighbour, cell B. When automatic allocation starts, Atoll checks the follow-
ing 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.
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: This option enables you to force cells located on the reference cell site in the candidate
neighbour list. This constraints can be weighted among the others and ranks the neighbours through the importance field
(see after).
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.This constraints can be weighted among the others and ranks
the neighbours through the importance field (see after).
Notes:
• Adjacence criterion: Let CellA be a candidate neighbour cell of CellB. 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).
• When this option is checked, adjacent cells are sorted and listed from the most adjacent to
the least, depending on the above criterion. Adjacence is relative to the number of pixels
ence cell is a candidate neighbour of another cell, this one will be considered as candidate neighbour of the reference cell.
force/forbid a cell to be candidate neighbour of the reference cell.
254 AT271_TRG_E6 © Forsk 2009

3. There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability:
• Intra-carrier neighbours: intra-carrier handover is a soft handover.
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).
SA is the area where the cell A is the best serving cell. It means that the cell A is the first one in the active set.
- The pilot signal received from the cell A is greater than the minimum pilot signal level.
- The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0).
- The pilot quality from A is the best.
SB is the area where the cell B can enter the active set.
- The pilot signal received from the cell B is greater than the minimum pilot signal level.
- The pilot quality from B is greater than the pilot quality from A minus the Ec/I0 margin. The Ec/I0 margin has
the same meaning as the AS-threshold defined in the Cell properties. So, it should logically have the same
value.
Figure 6.11: Overlapping Zone for Intra-carrier Neighbours
• Inter-carrier neighbours: inter-frequency handover is a hard handover. It is needed in a multi-carrier W-CDMA net-
work:
- To balance loading between carriers and layers (1st case),
- To make a coverage reason handover from micro cell frequency to macro cells (2nd case).
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.
SA is the area where the cell A is not the best serving cell of its layer but can enter the active set.
- The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0).
- The pilot quality from A is not the highest one. It is strictly lower than the best pilot quality received and greater
than the best pilot quality minus the Ec/I0 margin.
SB is the area where the cell B is the best serving cell of its layer.
- The pilot quality from B exceeds a user-definable minimum value (minimum Ec/I0).
- The pilot quality from B is the highest one.
© Forsk 2009 AT271_TRG_E6 255

Figure 6.12: Overlapping Zone for Inter-carrier Neighbours - 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.
SA is the area where the pilot quality from the cell A starts significantly decreasing but the cell A is still the best serving cell
of its layer (since it is on the border).
- The pilot quality from A is the highest one
- The pilot quality from A is lower than a user-definable minimum value (minimum Ec/I0) plus the Ec/I0 margin.
SB is the area where the cell B is the best serving cell of its layer.
- The pilot quality from B exceeds a user-definable minimum value (minimum Ec/I0).
- The pilot quality from B is the highest one.
Figure 6.13: Overlapping Zone for Inter-carrier Neighbours - 2nd Case
Note:
• Two ways enable you to determine the I0 value:
1 - A reduction factor (% of maximum powers contributing to I0) may be applied to cell
maximum powers (defined in Cell properties) to customize their contribution to I0. Thus, I0
represents the sum of effective powers received from the other cells. The entered
percentage is a kind of downlink load factor estimation.
If the % of maximum powers contributing to I0 is too low, i.e. if %  P max  P pilot , Atoll
takes into account the pilot powers to evaluate the I0 value.
2 - Atoll takes into account load parameters defined per cell (such as the total downlink
power used). I0 represents the sum of total transmitted powers.
256 AT271_TRG_E6 © Forsk 2009

SA  SB
Atoll calculates the percentage of covered area ( ----------------------  100 ) and compares this value to the % minimum covered
SA
area. If this percentage is not exceeded, the candidate neighbour B is discarded.
after).
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 importances) will be allocated to the reference cell. Note that
specific maximum numbers of neighbours (maximum number of intra-carrier neighbours, maximum number of inter-carrier
neighbours) can be defined at the cell level (property dialog or cell table). If defined there, this value is taken into account
instead of the default one available in the dialog.
to 100%.
Importance
value
Existing neighbour
Co-site transmitter Only if the Force co-site cells as neighbours option is selected (IF) function
Only if the Force adjacent cells as neighbours option is
selected
coverage conditions
relationship
(IF) Importance Function evaluation. The importance is evaluated through a function (IF), taking into account 3 factors:
• The co-site factor (C) which is a Boolean factor,
• The adjacency factor (A) which deals with the percentage of adjacency,
• The overlapping factor (O) meaning the percentage of overlapping
Min Max

(IF) function
no yes 30% + 30%{30%(O) + 70%(A)}
Max(O))(A)}
yes yes 60% + 40%{30%(O) + 70%(A)}
Max(O))(A)}
Where Delta(x)= Max(x) - Min(x)

Notes:
factors.
© Forsk 2009 AT271_TRG_E6 257

tions and the corresponding surface area (km2), the percentage of area meeting the adjacency conditions and the corre-
sponding surface area (km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing.
Notes:
• No simulation or 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.
• Even if no specific terminal, mobility or service is selected in the automatic allocation, 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.
Mobility: no impact on the allocation, no specific selection.
Terminal: selection of the one with the greatest (Gain - Loss) value, and, if equal, the one
with the lowest noise figure.
• The neighbour lists may be optionally used in the power control simulations to determine
the mobile's active set.
• The percentage of area is calculated with the resolution specified in the properties dialog of
the predictions folder (default resolution parameter).
Thus, if the cell B is a neighbour of the cell A while the cell A is not a neighbour of the cell
B, two cases are possible:
1st case: There is space in the cell B neighbour list: the cell A will be added to the list. It will
be the last one.
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.
• 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.

Atoll allocates neighbours to:
• TBA cells,
• Neighbours of TBA cells marked as exceptional pair, adjacent and symmetric,
• Neighbours of TBA cells that satisfy coverage conditions.
6.7 Primary Scrambling Code Allocation

Downlink primary scrambling codes enable you to distinguish cells from one another (cell identification).
By default, there are 512 primary scrambling codes numbered (0...511).
The cells to which Atoll allocates scrambling codes are referred to as the TBA cells (cells to be allocated). TBA cells fulfil
following conditions:
- They are active,
- They satisfy the filter criteria applied to the Transmitters folder,
258 AT271_TRG_E6 © Forsk 2009

- 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.
Note:
6.7.1 Automatic Allocation Description

6.7.1.1 Options and Constraints
The scrambling code allocation algorithm can take into account following constraints and options:
1. Neighbourhood between cells,
You may consider:
• First order neighbours: The neighbours of TBA cells listed in the Intra-technology neighbours table,
• Second order neighbours: The neighbours of neighbours,
• Third order neighbours: The neighbour’s neighbour’s neighbours.
Notes:
• In the context of the primary scrambling code allocation, the term "neighbours" refers to
intra-carrier neighbours.
• Atoll can take into account inter-technology neighbour relations as constraints to allocate
different scrambling codes to the UMTS neighbours of a GSM transmitter. In order to
consider inter-technology neighbour relations in the scrambling code allocation, you must
make the Transmitters folder of the GSM .atl document accessible in the UMTS .atl
document. For information on making links between GSM and UMTS .atl documents, see
the User Manual.
• Atoll considers symmetry relationship between a cell, its first order neighbours, its second
order neighbours and its third order neighbours.
2. Cells fulfilling a criterion on Ec/Io (option “Additional Overlapping Conditions”),

For a reference cell “A”, 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/Io)A exceeds the minimum Ec/Io and is the highest one and (Ec/Io)B is within a Ec/Io
margin of (Ec/Io)A).
Note:
• Atoll takes the total downlink power used by the cell into account in order to evaluate Io. Io
equals the sum of total transmitted powers. In case this parameter is not specified in the
cell properties, Atoll uses 50% of the maximum power.
3. Reuse distance,
Notes:
• Reuse distance is a constraint on the allocation of scrambling codes. 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.
• Scrambling code reuse distance can be defined at cell level. If this value is not defined,
then Atoll will use the default reuse distance defined in the Scrambling Code Automatic
Allocation dialog.
4. Exceptional pairs,
5. Domains of scrambling codes,
Note:
• When no domain is assigned to cells, Atoll considers the 512 primary scrambling codes
available.
6. The number of primary scrambling codes per cluster. In Atoll, we call "cluster", a group of scrambling codes as
defined in 3GPP specifications. 3GPP specifications define 64 clusters consisting of 8 scrambling codes (in this
case, clusters are numbererd from 0 to 63). However, it is possible to define another value (e.g. if you set the
number of codes per cluster to 4, scrambling codes will be distributed in 128 clusters).
7. The carrier on which the allocation is run: It can be a given carrier or all of them. In this case, either Atoll
independently plans scrambling codes for the different carriers, or it allocates the same primary scrambling code
to each carrier of a transmitter if the option "Allocate carriers identically" is selected.
8. 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.
© Forsk 2009 AT271_TRG_E6 259

9. The "Delete All Codes" option: When selecting this option, Atoll deletes all the current scrambling codes and
carries out a new scrambling code allocation. If not selected, the existing scrambling codes are kept.
In addition, it depends on the selected allocation strategy. Allocation strategies can be:
• Clustered allocation: The purpose of this strategy is to choose for a group of mutually constrained cells, scrambling
codes among a minimum number of clusters. In this case, Atoll will preferentially allocate all the codes within the
same cluster.
• Distributed per cell allocation: This strategy consists in using as many clusters as possible. Atoll will preferentially
allocate codes from different clusters.
• One cluster per site allocation: This strategy allocates one cluster to each site, then, one code from the cluster to
each cell of each site. When all the clusters have been allocated and there are still sites remaining to be allocated,
Atoll reuses the clusters as far as possible at another site.
• Distributed per site allocation: This strategy allocates a group of adjacent clusters to each site, then, one cluster
to each transmitter on the site according to its azimuth and finally, one code from the cluster to each cell of each
transmitter. The number of adjacent clusters per group depends on the number of transmitters per site you have
in your network; this information is required to start allocation based on this strategy. When all the groups of adja-
cent clusters have been allocated and there are still sites remaining to be allocated, Atoll reuses the groups of
adjacent clusters as far as possible at another site.
In the Results table, Atoll only displays scrambling codes allocated to TBA cells.
6.7.1.2 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/Io condition (option “Additional Overlapping Conditions”),
• The cells with distance from the TBA cell less than the reuse distance,
• The cells that make exceptional pairs with the TBA cell.
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
.atl document is accessible in the UMTS .atl document),
• The neighbour cells cannot share the same cluster (for the "Distributed per site" allocation strategy only).
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. During the allocation, Atoll tries to assign different scrambling codes to the TBA
cell and its near cells. If it respects all the constraints, the cost of the scrambling code plan is 0. When a cell has too many
constraints and there are not anymore scrambling codes available, Atoll breaks the constraint with the lowest cost so as
to generate the scrambling code plan with the lowest cost. For information on the cost generated by each constraint, see
"Cell Priority" on page 261.
6.7.1.2.1 Single Carrier Network

The allocation process depends on the selected strategy. Algorithm works as follows:
Strategies: Clustered and Distributed per Cell
Atoll processes TBA cells according to their priority. It allocates scrambling codes 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 calcu-
lating cell priority, see "Cell Priority" on page 261.
Strategy: One Cluster per Site
All sites which have constraints with the studied site are referred to as near sites.
Atoll assigns a 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 Reuse Distance option is selected, the algo-
rithm reuses the clusters as soon as the reuse distance is exceeded. Otherwise, when the option is not selected, the algo-
rithm tries to assign reused clusters as spaced out as possible.
Then, Atoll allocates a primary scrambling code from the cluster to each cell located on the sites (codes belong to the
assigned 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.
For information on calculating site priority, see "Site Priority" on page 264. For information on calculating cell priority, see
Strategy: Distributed per Site
Atoll assigns a group of adjacent clusters 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 groups of adjacent clusters have been
260 AT271_TRG_E6 © Forsk 2009

allocated and there are still sites remaining to be allocated, Atoll reuses the groups of adjacent clusters at another site.
When the Reuse Distance option is selected, the algorithm reuses the groups of adjacent clusters as soon as the reuse
distance is exceeded. Otherwise, when the option is not selected, the algorithm tries to assign reused groups of adjacent
clusters as spaced out as possible. Then, 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 "Second-
ary Neighbours in Other Clusters"). Then, Atoll allocates a primary scrambling code to each cell located on the transmitters
(codes belong to the assigned 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.
Determination of Groups of Adjacent Clusters
In order to determine the groups of adjacent clusters to be used, Atoll proceeds as follows: It defines theoretical groups of
adjacent clusters, independently of the defined domain, considering the 512 primary scrambling codes available and the
specified number of codes per cluster (if this one is set to 8, 64 clusters are supposed to be available). 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, the number of possible groups.
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. In this case, we have the following theoretical groups:
Group 1 Group 2 Group 3 Group 4 ... Group 21

Cluster 0 Cluster 3 Cluster 6 Cluster 9 Cluster 61
Cluster 1 Cluster 4 Cluster 7 Cluster 10 ... Cluster 62
If no domain is assigned to cells, Atoll can use all these groups for the allocation. On the other hand, if a domain is used,
the tool compares adjacent clusters really available in the assigned domain to the theoretical groups and only keeps adja-
cent clusters mapping the theoretical groups.
Let us assume that we have a domain consisted of 12 clusters: clusters 1 to 8 and clusters 12 to 15.
Therefore, Atoll will be able to use the following groups of adjacent clusters:
• Group 2 with cluster 3, 4 and 5,
• Group 3 with cluster 6, 7 and 8,
• Group 6 with cluster 12, 13 and 14.
• The clusters 1, 2 and 15 will not be used.
If a domain does not contain any adjacent clusters, the user is warned through the 'Event Viewer'.
6.7.1.2.2 Multi-Carrier Network

In case you have a multi-carrier network and you run the scrambling code allocation on all the carriers, the allocation proc-
ess depends on the allocation strategy as detailed above and in addition, 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 (Clustered and Distributed per cell), 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 same scrambling code is assigned to each cell of the transmitter.
In case of the "One cluster per site" strategy, Atoll assigns a cluster to each site and then, allocates a scrambling code 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 scrambling code is assigned to each cell of the trans-
mitter.
In case of the "Distributed per site" strategy, Atoll assigns a group of adjacent clusters to each site, then a cluster to each
transmitter and finally, allocates a scrambling code 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
scrambling code is assigned to each cell of the transmitter.
For information on calculating transmitter priority, see "Transmitter Priority" on page 264.
Note:
• When cells, transmitters or sites have the same priority, processing is based on an
alphanumeric order.
6.7.1.3 Priority Determination

6.7.1.3.1 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. A cell without
© Forsk 2009 AT271_TRG_E6 261

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 scrambling
code allocation process.
There are six criteria employed to determine the cell priority:
• Scrambling Code Domain Criterion
The cost due to the domain constraint, C i  Dom  , depends on the number of scrambling codes available for the allocation.
The domain constraint is mandatory and cannot be broken.
When no domain is assigned to cells, 512 scrambling codes are available and we have:
C i  Dom  = 0
When domains of scrambling codes are assigned to cells, each unavailable scrambling code 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 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. 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. 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 excep-
tional 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 6.14: 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. This value can be defined in the Constraint Cost dialogue.
262 AT271_TRG_E6 © Forsk 2009

Because two first order neighbours must not have the same scrambling code, 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:
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 scrambling code, 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 scrambling code, Atoll considers the cost created by two third
C l  N3  i   + C l  N3  i  
C l – l  N3  i   = ----------------------------------------------------------
-
2
Note:
• 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
• GSM Neighbour Criterion

This criterion is considered when the co-planning mode is activated (i.e. the Transmitters folder of the GSM .atl document
is made accessible in the UMTS .atl document) and inter-technology neighbours have been allocated. If the cell i is neigh-
bour of a GSM transmitter, the cell constraint level depends on how many cells j are neighbours of the same GSM trans-
mitter. The total cost due to GSM neighbour constraint is given as:
C i  N 2G  =  cN 2G
 j – Tx 2G 
j
Where
cN is the cost of the GSM neighbour constraint. This value can be defined in the Constraint Cost dialogue.
2G
• Cluster Criterion
When the "Distributed per Site" allocation strategy is used, you can consider additional constraints on allocated clusters
(one cell, its first order neighbours and its second order neighbours must be assigned scrambling codes from different clus-
ters). In this case, the constraint level of any cell i depends on the number of first and second order neighbours, j and k.
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. 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
© Forsk 2009 AT271_TRG_E6 263

C i  U  = C i  Dist  + C i  EP  + C i  N  + C i  N 2G  + C i  Cluster 
6.7.1.3.2 Transmitter Priority

In case you have a multi-carrier network and you run scrambling code allocation on "all" the carriers with the option "allo-
cate 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 scrambling code 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 
With C Tx  U  = Max  C  U   and C  Dom  = 512 – Number of scrambling codes in the domain
i  Tx i Tx
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.
6.7.1.3.3 Site Priority

In case of "Per Site" allocation strategies (One cluster per site and Distributed per site), algorithm in Atoll allots priorities
to sites. Priorities assigned to sites depend on how much constrained each site is and the cost defined for each constraint.
The higher the cost on a site, the higher the priority it has for the scrambling code allocation process.
Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. 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
Tx  S Tx S
Here, the domain considered for the site is the intersection of domains available for transmitters of the site. The domain
constraint is mandatory and cannot be broken.
6.7.2 Allocation Examples

6.7.2.1 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, let us consider the following sample scenario:
Figure 6.15: Primary Scrambling Codes Allocation
Let Site0, Site1, 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. This implies that the domain of scrambling codes for the four
sites is from 0 to 23 (cluster 0 to cluster 2). The reuse distance is supposed to be less than the inter-site distance. Only
co-site neighbours exist.
The following section lists the results of each combination of options with explanation where necessary.
264 AT271_TRG_E6 © Forsk 2009

6.7.2.1.1 Strategy: Clustered

Since the restrictions of neighbourhood only apply to co-sites with the same importance and sites distances are greater
than reuse distances, every cell has the same priority. Then, scrambling code allocation to cells is performed in an alpha-
numeric order.
Without ‘Use a Maximum of Codes’ 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, Atoll starts
each site. allocation at the start of a different cluster at each site.
When a cluster is reused, and there are non allocated
codes left in the cluster, Atoll first allocates those codes
before reusing the already used ones.
6.7.2.1.2 Strategy: Distributed

than reuse distances, every cell has the same priority. Then, scrambling code allocation to cells is performed in an alpha-
numeric order.
Atoll allocates codes from different clusters to each cell of Atoll allocates codes from different clusters to each site’s
the same site. Under given constraints of neighbourhood cells. As it is possible to use a maximum of codes, Atoll allo-
and reuse distance, same codes can be allocated to each cates the codes so that there is least repetition of codes.
site’s cells.
© Forsk 2009 AT271_TRG_E6 265

6.7.2.1.3 Strategy: ‘One Cluster per Site

than reuse distances, every site has the same priority. Then, cluster allocation to sites is performed in an alphanumeric
order.
In this strategy, a cluster of codes is limited to be used at When it is possible to use a maximum of codes, Atoll can
just one site at a time unless all codes and clusters have allocate different codes from a reused cluster at another
been allocated and there are still sites remaining to be allo- site.
cated. In this case Atoll reuses the clusters as far as possi-
ble at another site.
6.7.2.1.4 Strategy: ‘Distributed per Site

than reuse distances, every site has the same priority. Then, the group of adjacent clusters allocation to sites is performed
in an alphanumeric order.
In this strategy, a group of adjacent clusters is limited to be When it is possible to use a maximum of codes, Atoll can
used at just one site at a time unless all codes and groups allocate different codes from a reused group of adjacent
of adjacent clusters have been allocated and there are still cluster at another site.
sites remaining to be allocated. In this case (here only one
group of adjacent clusters (clusters 0, 1 and 2) is available),
Atoll reuses the group at another site.
6.7.2.2 Allocate Carriers Identically

In order to understand the behaviour of algorithm when using the option "Allocate Carriers Identically" or not, let us
consider the following sample scenario:
Let Site0, Site1, Site2 and Site3 be four sites with 3 cells using carrier 0 and 3 cells using carrier 1. Scrambling codes have
to be allocated out of 3 clusters consisted of 8 primary scrambling codes. This implies that the domain of scrambling codes
for the five sites is from 0 to 23 (cluster 0 to cluster 2). The reuse distance is supposed to be less than the inter-site
266 AT271_TRG_E6 © Forsk 2009

distance. Only co-site neighbours exist. Allocation algorithm will be based on the "One Cluster per Site" strategy and the
option "Use a Maximum of Codes" is selected.
Without ‘Allocate Carriers Identically’ With ‘Allocate Carriers Identically’
Atoll allocates one cluster at each site as detailed in the In this case, Atoll allocates one cluster at each site and
previous section. Then, it allocates a code from the cluster then, one code to each transmitter so as to use a maximum
to each cell of the site so as to use a maximum of codes. of codes. Then, the same code is given to each cell of the
transmitter.
In both cases (with and without ’Allocate Carriers Identically’), every site has the same priority. Then, cluster allocation to
sites is performed in an alphanumeric order.
6.8 Automatic GSM-UMTS Neighbour Allocation

6.8.1 Overview
It is possible to automatically calculate and allocate neighbours between GSM and UMTS networks. In Atoll, it is called
inter-technology neighbour allocation.
Inter-technology handover is used in two cases:
• When the UMTS coverage is not continuous. In this case, the UMTS coverage is extended by UMTS-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 network, GSM.atl, and another one describing the UMTS network,
UMTS.atl,
• An existing link on the Transmitters folder of GSM.atl into UMTS.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
UMTS.atl, satisfy following conditions:
• They satisfy the filter criteria applied to Transmitters folder,
• 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 UMTS TBA cells may be assigned neighbours.

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 UMTS reference cell, A, and a GSM candidate neighbour, transmitter B.
© Forsk 2009 AT271_TRG_E6 267

6.8.2.1 Algorithm Based on Distance

1. The distance between the UMTS reference cell and the GSM neighbour must be less than the user-definable
maximum inter-site distance. If the distance between the UMTS reference cell and the GSM neighbour is greater
than this value, then the candidate neighbour is discarded.
Candidate neighbours are sorted in descending order with respect to distance.
Note:
• Transmitter azimuths are taken into account to evaluate the inter-transmitter distance (for
further information on inter-transmitter distance calculation, please refer to "Calculation of
Inter-Transmitter Distance" on page 270)

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. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference UMTS cell.
and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
3. Atoll lists all candidate neighbours and sorts them by priority 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. The candidate neighbour
priority depends on the neighbourhood cause. Priority assigned to each neighbourhood cause is listed in the table
below (1 is a higher priority than 2 and so on).
Neighbourhood cause When Priority

Only if the Delete existing neighbours option is not selected
Existing neighbour 1
and in case of a new allocation
Exceptional pair Only if the Force exceptional pairs option is selected 2
Co-site cell Only if the Force co-site cells as neighbours option is selected 3
Only if the Max inter-site distance is not exceeded 4
distance conditions
If there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8.
Among 15 candidate neighbours, only 8 (those with the highest priority) will be allocated to the reference cell.
bours allowed for each cell. In addition, it indicates the allocation cause for each neighbour. Therefore, a neighbour may
be marked as exceptional pair or co-site. If the neighbour is not forced but fulfils distance conditions, Atoll displays the
distance from the reference cell. Finally, if cells have previous allocations in the list, neighbours are marked as existing.
6.8.2.2 Algorithm Based on Coverage Overlapping

1. The distance between the UMTS reference cell and the GSM neighbour must be less than the user-definable
maximum inter-site distance. If the distance between the UMTS reference cell and the GSM neighbour is greater
Note:
• Here, real inter-transmitter distance is considered.

the reference UMTS cell in the candidate neighbour list. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference UMTS cell.
3. There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability.
Four different cases may be considered for SA:
268 AT271_TRG_E6 © Forsk 2009

- 1st case: SA is the area where the cell A is the best serving cell of the UMTS 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/Io 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 UMTS network.
The Ec/I0 margin must be equal to 0dB, the max Ec/Io option selected and a maximum Ec/Io user-defined.
- The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/Io.
- 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/Io option disabled.
- The pilot quality from A is within a margin from the best Ec/Io, where the best Ec/Io exceeds the minimum
Ec/Io.
- 4th case: SA represents the area where:
- The pilot quality from A is within a margin from the best Ec/Io (where the best Ec/Io exceeds the minimum
Ec/Io) and lower than the maximum Ec/Io.
In this case, the margin must be different from 0dB, the max Ec/Io option selected and a maximum Ec/Io user-
defined.
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:
is within a margin from the best BCCH signal level.
SA  SB
SA
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of covered
area.
Guidelines for the automatic allocation

When the automatic allocation is based on coverage overlapping, we recommend you to perform two successive auto-
matic allocations:
- A first allocation in order to find handovers due to non-continuous UMTS 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.
below (1 is a higher priority than 2 and so on…).

Only if the % minimum covered area is exceeded 4
coverage conditions
Therefore, among 15 candidate neighbours, only 8 (those with the highest priority) will be allocated to the reference cell.
© Forsk 2009 AT271_TRG_E6 269

be marked as exceptional pair or co-site. If the neighbour is not forced but fulfils coverage conditions, Atoll displays the
percentage of covered area and the overlap area (km2) in brackets. Finally, if cells have previous allocations in the list,
neighbours are marked as existing.
Notes:
not found.
• The percentage of covered area is calculated with the resolution specified in the properties
dialog of the predictions folder (default resolution parameter).
6.8.2.3 Appendices
6.8.2.3.1 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 UMTS.atl. Therefore, if you start a new allocation without selecting the Delete existing neigh-
bours 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 trans-
mitter) 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.
6.8.2.3.2 Calculation of Inter-Transmitter Distance

When allocation algorithm is based on distance, Atoll takes into account the real distance and azimuths of antennas in
order to calculate the effective inter-transmitter distance.
Dist  CellA CellB  = D   1 + x  cos  – x  cos  

where x = 0.5% so that the maximum D variation not to exceed 1% and D is stated in m.
Figure 6.16: 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. It is this effective distance that will be taken into account rather than the real distance.
Note:
• This formula is not used when allocation algorithm is based on coverage overlapping. In
this case, real inter-transmitter distance is considered.
270 AT271_TRG_E6 © Forsk 2009

Chapter 7
IS-95 cdmaOne and
CDMA2000 Networks
simulations and prediction studies available in IS-95 cdmaOne and CDMA2000 projects.
Atoll
272 AT271_TRG_E6 © Forsk 2009

Chapter 7: IS-95 cdmaOne and CDMA2000 Networks
7 IS-95 cdmaOne and CDMA2000 Networks

Three criteria can be studied in point analysis (Profile tab) and in common coverage studies. Study criteria are detailed in
the table below:

Signal level ( P rec ) in dBm
P rec  ic  = EIRP  ic  – L path – M Shadowing – model – L Indoor + G term – L term
Path loss ( L path ) in dBm L path = L model + L ant

Tx
Total losses ( L total ) in dBm L total =  L path + L Tx + L term + L indoor + M Shadowing – model  –  G Tx + G term 
where,
ic is a carrier number,

Tx
L term are the receiver losses,
G term is the receiver antenna gain,
G Tx is the transmitter antenna gain,
L Tx is the transmitter loss ( L Tx = L total – DL ). For information on calculating transmitter loss, "UMTS HSPA, CDMA2000
1xRTT 1xEV-DO, IS-95 cdmaOne, and TD-SCDMA Documents" on page 124.
Notes:
• For IS-95 cdmaOne and CDMA2000 1xRTT systems, EIRP  ic  = P pilot  ic  + G Tx – L Tx
(where, P pilot  ic  is the cell pilot power).
• For CDMA2000 1xEV-DO systems, EIRP  ic  = P max  ic  + G Tx – L Tx (where P max  ic 

is the maximum cell power).
• It is also possible to analyse all the carriers at once. In this case, Atoll displays the best
signal level received from a transmitter. Therefore, if the network consists of 1xRTT and
1xEV-DO carriers, Atoll takes the highest power of both cells for each transmitter (i.e. 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.
• Atoll considers that G term and L term equal zero.

7.1.2.1 Profile Tab
Atoll displays either the signal level received from the selected transmitter on a carrier ( P rec  ic  ), or the highest signal
level received from the selected transmitter on all the carriers.
© Forsk 2009 AT271_TRG_E6 273

Note:

For each transmitter, Atoll displays either the signal level received on a carrier, ( P rec  ic  ), or the highest signal level
received on all the carriers.
Note:

For each TBC transmitter, Txi, Atoll determines the selected criterion on each bin inside the Txi calculation area. In fact,
each bin within the Txi calculation area is considered as a potential (fixed or mobile) receiver.

We can distinguish three cases:

Txi Txi Txi
Minimum threshold  P rec  ic   or L total or L path   Maximum threshold

Txi Txi Txi
And
Txi Txj
ji

Notes:
the highest.

274 AT271_TRG_E6 © Forsk 2009

Txi Txi Txi

And
Txi nd Txj
P rec  ic   2 Best  P rec  ic   – M
ji

Notes:
the second highest.


such as:
Atoll calculates signal level received from the transmitter on each bin of each transmitter service area. A bin of a service
area is coloured if the signal level exceeds (  ) the defined minimum thresholds (bin colour depends on signal level).
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where other service
areas overlap the studied one, Atoll chooses the highest value. A bin of a service area is coloured if the signal level
exceeds (  ) the defined thresholds (the bin colour depends on the signal level). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer
corresponds to an area where the signal level from the best server exceeds a defined minimum threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each bin of each transmitter service area. A bin of a service area is
coloured if path loss exceeds (  ) the defined minimum thresholds (bin colour depends on path loss). Coverage consists
Total Losses (dB)
Atoll calculates total losses from the transmitter on each bin of each transmitter service area. A bin of a service area is
coloured if total losses exceed (  ) the defined minimum thresholds (bin colour depends on total losses). Coverage
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where other service
bin of a service area is coloured if the path loss exceeds (  ) the defined thresholds (bin colour depends on path loss).
© Forsk 2009 AT271_TRG_E6 275

Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where service areas
overlap the studied one, Atoll determines the best transmitter and evaluates total losses from the best transmitter. A bin
of a service area is coloured if the total losses exceed (  ) the defined thresholds (bin colour depends on total losses).
layers as defined thresholds. Each layer corresponds to an area where the total losses from the best server exceed a
Number of Servers
Atoll evaluates how many service areas cover a bin in order to determine the number of servers. The bin colour depends
on the number of servers. Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the number of serv-
ers exceeds (  ) a defined minimum threshold.
On each bin of each transmitter service area, the coverage corresponds to the pixels where the signal level from this trans-
mitter fulfils signal conditions defined in Conditions tab with different Cell edge coverage probabilities. There is one cover-
age area per transmitter in the explorer.
On each bin of each transmitter service area, the coverage corresponds to the pixels where the best signal level received
explorer.

7.2.1 Parameters Used for IS-95 cdmaOne and CDMA2000 1xRTT
Modelling
7.2.1.1 Inputs
etc.)
Tx
ic Global parameter None Carrier number

Active set upper threshold
req req req
Q pilot Q pilot  txi ic  + Q pilot None (used to determine the best server in
the active set)
Active set lower threshold
min min min
Q pilot Q pilot  txi ic  + Q pilot None (used to determine other members of
the active set)
req Minimum Ec/I0 required from the cell
Q pilot  txi ic  Min. Ec/I0 - Cell parameter None
to be the best server in the active set
min Minimum Ec/I0 required from the cell
Q pilot  txi ic  T_Drop - Cell parameter None
not to be rejected from the active set
Variation of the minimum Ec/I0
req
Q pilot Delta Min. Ec/I0 - Mobility parameter None required from the cell to be the best
server in the active set
min
Q pilot Delta T_Drop - Mobility parameter None required from the cell not to be
rejected from the active set
E
FCH – DL
Eb/Nt target for FCH channel on
DL
 Q req  FCH -----b- (Service, Mobility) parameter None
 N t  req downlink
276 AT271_TRG_E6 © Forsk 2009

E
SCH – DL
DL -----b- Eb/Nt target for SCH channel on
 Q req  SCH  N t  req None
downlink
(Service, Mobility, SCH rate multiple) parameter
E
FCH – UL
Eb/Nt target for FCH channel on
UL
 Q req  FCH -----b- (Service, Terminal) parameter None
 N t  req uplink
E
SCH – UL
UL -----b- Eb/Nt target for SCH channel on
 Q req  SCH  N t  req None
uplink
(Service, Terminal, SCH rate multiple) parameter
CE – U L Number of channel elements available
N max  N I  Site parameter None
for a site on uplink
CE – D L Number of channel elements available
for a site on downlink
CE – U L Number of channel elements of a site
consumed by users on uplink
CE – D L Number of channel elements of a site
consumed by users on downlink
Overhead – C E – UL Site equipment parameter None the cell for common channels on
N
uplink
Overhead – C E – DL Site equipment parameter None the cell for common channels on
N
downlink
FCH – C E – UL Number of channel elements used for

N (Terminal, site equipment) parameter None
FCH on uplink
FCH – C E – DL Number of channel elements used for

FCH on downlink
Codes Maximum number of Walsh codes
N max  txi ic  Simulation constraint None
available per cell (128)
Codes Number of Walsh codes used by the

N  txi ic  Simulation result None
cell


UL
DL
%Power max Simulation constraint (global parameter or cell parameter) % Maximum percentage of used power
Tx
Term
N0 NF Term  K  T  W W Thermal noise at terminal
Rc W bps Chip rate
UL
DL Downlink rake receiver efficiency

f rake efficiency Terminal parameter None
factor
DL SCH rate factor (drawn following the
Frate SCH Simulation result None
SCH probabilities of the service)
DL
R FCH Terminal parameter bps Downlink FCH nominal rate
DL DL DL
R SCH R FCH  Frate SCH bps Downlink SCH bit rate
UL SCH rate factor (drawn following the

Frate SCH Simulation result None
SCH probabilities of the service)
© Forsk 2009 AT271_TRG_E6 277

UL
R FCH Terminal parameter bps Uplink FCH nominal rate
UL UL UL
R SCH R FCH  Frate SCH bps Uplink SCH bit rate
W Downlink service processing gain on

Gp
FCH – DL -------------- None
DL
R FCH FCH
W Downlink service processing gain on

Gp
SCH – DL -------------- None
DL
R SCH SCH
FCH – UL
W -
------------- Uplink service processing gain on
Gp UL None
R FCH FCH
W Uplink service processing gain on

Gp
SCH – UL -------------- None
UL
R SCH SCH
DL
AF FCH Service parameter None Downlink activity factor on FCH
UL
AF FCH Service parameter None Uplink activity factor on FCH
P Sync  txi ic  Cell parameter W Cell synchronisation channel power
Cell other common channels (except

P paging  txi ic  Cell parameter W
CPICH and SCH) power
P pilot  txi ic  Cell parameter W Cell pilot power
P max  txi ic  Cell parameter W Maximum cell power
Maximum amount of power reserved

M pooling  txi ic  Cell parameter dB
for pooling
min
P FCH Service parameter W Minimum power allowed for FCH
max
P FCH Service parameter W Maximum power allowed for FCH
min
P SCH Service parameter W Minimum power allowed for SCH
max
P SCH Service parameter W Maximum power allowed for SCH
DL Cell FCH power for a traffic channel

P FCH  txi ic tch  Simulation result including the term AFFCH  Serv  W
on carrier ic
P FCH  txi ic   P FCH  txi ic tch  W Total FCH power on carrier ic
tch  FCH  ic  
Transmitter SCH power for a traffic

P SCH  txi ic tch  Simulation result W
channel on carrier ic
P SCH  txi ic   P SCH  ic tch  W Total SCH power on carrier ic

tch  SCH  ic  
P pilot  txi ic  + P Sync  txi ic  + P paging  txi ic  + P SCH  txi ic  Transmitter total transmitted power on
P tx  txi ic  W
+ P FCH  txi ic  carrier ic
min
max
FCH UL Terminal FCH power transmitted in

P term  ic  Simulation result including the term AFFCH  Serv  W
carrier ic
SCH Terminal SCH power transmitted on
P term  ic  Simulation result W
carrier ic
transmitted
transmitted
finger
278 AT271_TRG_E6 © Forsk 2009


L Tx None Transmitter lossa
f Terminal parameter None Number of fingers
p Terminal parameter % Pilot power percentage

Result calculated from cell edge coverage probability and Ec/I0 Ec/I0 Shadowing margin
n=2 or 3 pilot signals at the mobile b.
M Shadowing –  Eb  Nt  Result calculated from cell edge coverage probability and DL Eb/Nt DL Eb/Nt Shadowing margin
None
DL standard deviation Only used in prediction studies
None
UL npaths
G macro – diversity = M Shadowing –  Eb  Nt  – M Shadowing – Eb  Nt 
n=2 or 3 in soft handoffc.
In prediction studiesd
For Ec/I0 calculation

---------------------------------------------------------------------------------------------------------------------------------------------------
G Tx  G term
For DL Eb/Nt calculation
DL
LT ------------------------------------------------------------------------------------------------------------------------------------------------------------
- None Transmitter-terminal total loss
G Tx  G term
UL
------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
In simulations
--------------------------------------------------------------------------------------------------------------------------------
G Tx  G term
P pilot  txi ic 
P c  txi ic  --------------------------------- W Chip power received at terminal
LT
FCH – DL P FCH  txi ic tch  Bit received power at terminal for FCH
Pb  txi ic tch  ---------------------------------------------- W
LT on carrier ic
SCH – DL P SCH  txi ic tch  Bit received power at terminal for SCH
Pb  txi ic tch  ---------------------------------------------- W
LT on carrier ic
DL FCH – DL SCH – DL Bit received power at terminal for

P b  txi ic tch  Pb  txi ic tch  + P b  txi ic tch  W
FCH+SCH on carrier ic
DL P tx  txi ic  Total received power at terminal from

P tot  txi ic  ---------------------------- W
P FCH  txi ic  + P SCH  txi ic  Total power received at terminal from

DL
P traf  txi ic   ---------------------------------------------------------------------------
LT W traffic channels of a transmitter on
tch  ic  carrier ic
© Forsk 2009 AT271_TRG_E6 279

FCH
FCH – UL P term Bit received power at transmitter for
Pb  ic  -------------- W
LT FCH on carrier ic
SCH
SCH – UL P term Bit received power at transmitter for
Pb  ic  -------------- W
LT SCH on carrier ic
UL FCH – UL SCH – UL Bit received power at transmitter for

P b  ic  Pb  ic  + P b  ic  W
SCH+FCH on carrier ic
UL
UL UL UL P b  ic  Total power transmitted by the
P tot  ic  P b  ic  + P c  ic  = -------------------
- W
1 – p terminal on carrier ic
UL UL
P c  ic  p  P tot  ic  W Chip received power at transmitter
a. L Tx = L total – UL on uplink and L Tx = L total – DL on downlink. For information on calculating transmitter

losses on uplink and downlink, see "UMTS HSPA, CDMA2000 1xRTT 1xEV-DO, IS-95 cdmaOne, and TD-SCDMA
Documents" on page 124.
npaths
b. 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.
npaths
c. M Shadowing –  Eb  Nt  corresponds to the shadowing margin evaluated from the shadowing error probability
UL

d. In uplink prediction studies, only carrier power level is downgraded by the shadowing margin
UL
the shadowing model ( M Shadowing –  Eb  Nt  or M Shadowing – Ec  Io ) while extra-cell interference level is not. Therefore,
DL
M Shadowing –  Eb  Nt  or M Shadowing – Ec  Io is set to 1 in downlink extra-cell interference calculation.

DL
7.2.1.2 Ec/I0 Calculation

This table details the pilot quality ( Q pilot or Ec  Io ) calculations.
DL DL Downlink intra-cell interference at

I intra  txi ic  P tot  txi ic  W

DL Downlink extra-cell interference at
DL
txj j  i

DL
W
txj
------------------------------------------------

I 0  ic  I intra  txi ic  + I extra  ic  + I inter – carrier  ic  + N 0 W
carrier ic
Ec  BTS    P c  txi ic  Quality level at terminal on pilot for

Q pilot  txi ic    ------ ------------------------------------------------------- None
 I0  DL
I 0  ic  carrier ic
7.2.1.3 DL Eb/Nt Calculation

Eb
This table details calculations of downlink traffic channel quality ( Q tch (tch could be FCH or SCH) or  ------- ).
DL
 Nt  DL
DL Downlink intra-cell interference at

I intra  txi ic   1 –  BTS  F ortho   P DL  txi ic  W
tot terminal on carrier ic

DL
txj j  i

DL
W
txj
------------------------------------------------
280 AT271_TRG_E6 © Forsk 2009


N tot  ic  I intra  ic  + I extra  ic  + I inter – carrier  ic  + N 0 W
carrier ic
FCH – DL
 BTS  Pb  txi ic tch 
-  G FCH
--------------------------------------------------------------------------------------------------------------- p
– DL
E b DL DL
N tot  ic  –  1 – F ortho    BTS  P b  txi ic 
DL
ic    ------
DL
Q FCH  txi None channel from one transmitter for a
 N t  FCH
 BTS  P b
FCH – DL
 txi ic tch  FCH channel on carrier ic a
-  G FCH
Total noise: --------------------------------------------------------------------------- p
– DL
DL
N tot  ic 
Quality level at terminal for FCH using


DL DL carrier ic due to combination of all
DL f rake efficiency  Q FCH  tx k ic 
Q FCH  ic  None
txk  ActiveSet  FCH 
transmitters of the active set (Macro-
SCH – DL
 BTS  Pb  txi ic tch 
-  G SCH
--------------------------------------------------------------------------------------------------------------- p
– DL
E b DL DL
N tot  ic  –  1 – F ortho    BTS  P b  txi ic 
DL
ic    ------
DL
Q SCH  txi None channel from one transmitter for a
 N t  SCH
 BTS  P b
SCH – DL
 txi ic tch  SCH channel on carrier icb
-  G SCH
Total noise: --------------------------------------------------------------------------- p
– DL
DL
N tot  ic 
Quality level at terminal for SCH using


DL DL carrier ic due to combination of all
DL f rake efficiency  Q SCH  tx k ic 
Q SCH  ic  None
txk  ActiveSet  SCH 
transmitters of the active set (Macro-
DL
DL Q FCH  ic  Downlink soft handover gain for FCH
 G SHO  FCH ------------------------------------------------------------- None
DL
Q FCH  BestServer ic  channel on carrier ic
DL
DL Q SCH  ic  Downlink soft handover gain for SCH
 G SHO  SCH -------------------------------------------------------------
- None
DL
Q SCH  BestServer ic  channel on carrier ic
DL
 Q req  FCH Required transmitter FCH traffic
req
P FCH  txi ic  ---------------------------  P FCH  txi ic  W channel power to achieve Eb/Nt target
DL
Q FCH  ic  at terminal on carrier ic
DL
 Q req  SCH Required transmitter SCH traffic
req
P SCH  txi ic  ---------------------------  P SCH  txi ic  W channel power to achieve Eb/Nt target
DL
Q SCH  ic  at terminal on carrier ic
req req req Required transmitter traffic channel

P tch  txi ic  P FCH  txi ic  + P SCH  txi ic  W
power on carrier ic
b. Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account
7.2.1.4 UL Eb/Nt Calculation

Eb
This table details calculations of uplink traffic channel quality ( Q tch (tch could be FCH or SCH) or  ------- ).
UL
 Nt  UL

  Pb
UL UL
UL intra  ic  + P c  ic  
term
ic
txi


UL UL
UL extra  P b  ic  + P c  ic  
term
ic
txj j  i
  Pb
UL UL
 ic adj  + P c  ic adj  
txj  j
-----------------------------------------------------------------------------
-
UL UL extra UL intra Total received interference at

Tx
 txi ic  +  1 – F MUD   term I tot
UL
© Forsk 2009 AT271_TRG_E6 281


FCH – UL
 term  P b  ic 
-  G FCH
--------------------------------------------------------------------------------------------------------------- p
– UL
Quality level at transmitter on a traffic
Eb UL Tx
UL
Q FCH  txi ic    ------
UL
None channel for the FCH channel on
 N t  UL
 term  P b
FCH – UL
 ic  carrier ica
-  G FCH
Total noise: ----------------------------------------------------- p
– UL
UL

SCH – UL
-  G SCH
--------------------------------------------------------------------------------------------------------------- p
– UL
Quality level at transmitter on a traffic
Eb UL Tx
UL
Q SCH  txi ic    ------
UL
None channel for the SCH channel on
 N t  UL
 term  P b
SCH – UL
 ic  carrier icb
-  G SCH
Total noise: ------------------------------------------------------ p
– UL
UL
UL
No HO: Q tch  txi ic 

UL UL
tx k  ActiveSet
 samesite 
Soft, Softer/Soft HO (No MRC): to combination of all transmitters of
the active set located at the same site
UL UL
Max  Q tch  tx k ic    G macro – diversity and taking into account increase of the
None quality due to macro-diversity (macro-
UL tx  ActiveSet
k
Q tch  ic 
diversity gain).
Softer/Soft HO (MRC): tch could be FCH or SCH
 
Max  UL  UL

UL UL
 f rake efficiency  Q tch  tx k ic  Q tch  tx l ic  In simulations, G macro – diversity = 1 .
k l  
txk  samesite  tx k 
tx  othersite
l
UL
UL
UL Q FCH  ic  Uplink soft handover gain for FCH
 G SHO  FCH ------------------------------------------------------------- None
UL
Q FCH  BestServer ic  channel on carrier ic
UL
UL Q SCH  ic  Uplink soft handover gain for SCH
 G SHO  SCH -------------------------------------------------------------
- None
UL
Q SCH  BestServer ic  channel on carrier ic
UL
 Q req  FCH Required terminal power to achieve
FCH – req
P term  ic  ---------------------------  P FCH
term  ic  W Eb/Nt target at transmitter for FCH on
UL
Q FCH  ic  carrier ic
UL
 Q req  SCH Required terminal power to achieve
SCH – req
P term  ic  ---------------------------  P SCH
term  ic  W Eb/Nt target at transmitter for SCH on
UL
Q SCH  ic  carrier ic
req FCH – req SCH – req
P term  ic  P term  ic  + P term  ic  W Required terminal power on carrier ic
b. Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account
7.2.1.5 Simulation Results

This table contains some simulation results provided in the Cells and Mobiles tabs of the simulation property dialog.
P tot  txi ic  – F ortho   BTS  P tot  txi ic 

DL DL
DL Downlink intra-cell interference at
I intra  txi ic  None
DL
–  1 – F ortho   BTS   P b  txi ic 
282 AT271_TRG_E6 © Forsk 2009


DL
txj j  i

DL
W
txj
------------------------------------------------
DL DL DL DL Total effective interference at terminal

I tot  ic  I intra  ic  + I extra  ic  + I inter – carrier  ic  W
on carrier ic (after unscrambling)
DL DL Term Total received noise at terminal on
N tot  ic  I tot  ic  + N 0 W
carrier ic

  Pb
UL UL
UL intra  ic  + P c  ic  
term
ic
txi


UL UL
UL extra  P b  ic  + P c  ic  
term
ic
txj j  i
  Pb
UL UL
 ic adj  + P c  ic adj  
txj  j
-----------------------------------------------------------------------------
-

Tx
UL

UL
UL
X  txi ic  -----------------------------
UL
UL
F
UL
UL intra Tx

E
UL
 txi ic  ------------------------------ None
UL carrier ic
F  txi ic 
DL
P tx  txi ic  
 -------------------------------- Percentage of max transmitter power
%Power  txi ic  -  100 None
 P max  txi ic  used.
Simulation result available per cell

DL DL
---------------------------------------------------------------------------------------- + 1 – F ortho   BTS
P tx  txi ic 
 ------------------------------------------------------------------------------------------------------------------------------------------
1
------------- +  1 – F ortho   BTS 
-
tch DL
CI req
DL None Downlink load factor on carrier ic
X  txi ic  SCH – DL FCH – DL
DL Q req Q req
with CI req = ------------------------
- + ------------------------
-
SCH – DL FCH – DL
Gp Gp
DL
I tot  ic 
Simulation result available per mobile: --------------------
DL
N tot  ic 
DL
DL
I tot  ic 
F  txi ic  -------------------------------
- None Downlink reuse factor on a carrier ic
DL
I intra  txi ic 

7.2.2 Parameters Used for CDMA2000 1xEV-DO Modelling

7.2.2.1 Inputs
etc.)
© Forsk 2009 AT271_TRG_E6 283

Tx
ic Global parameter None Carrier number

Active set upper threshold
req req req
Q pilot Q pilot  txi ic  + Q pilot None (used to determine the best server in
the active set)
Active set lower threshold
min min min
Q pilot Q pilot  txi ic  + Q pilot None (used to determine other members of
the active set)
req Minimum Ec/I0 required from the cell
Q pilot  txi ic  Min. Ec/I0 - Cell parameter None
to be the best server in the active set
min Minimum Ec/I0 required from the cell
Q pilot  txi ic  T_Drop - Cell parameter None
not to be rejected from the active set
req
Q pilot Delta Min. Ec/I0 - Mobility parameter None required from the cell to be the best
server in the active set
min
Q pilot Delta T_Drop - Mobility parameter None required from the cell not to be
rejected from the active set
Mobility parameter for 1xEV-DO Rev. 0 users
E
UL
-----c- Parameter read in the 1xEV-DO Rev. A Radio Bearer Selection None Minimum pilot quality level on uplink
 N t  min
(Uplink) table for 1xEV-DO Rev. A users
Number of subframes associated to
n SF 1xEV-DO Rev. A Radio Bearer Selection (Uplink) table None
uplink 1xEV-DO Rev. A bearer
UL Uplink RLC peak rate provided by the
R RLC – peak Uplink 1xEV-DO Rev. A Radio Bearer table None
1xEV-DO Rev. A bearer
Mobility parameter for 1xEV-DO Rev. 0 users
E
DL
-----c- Minimum pilot quality level required to
 N t  min Parameter read in the 1xEV-DO Rev. A Radio Bearer Selection None
obtain a data rate on downlink
(Downlink) table for 1xEV-DO Rev. A users
Number of timeslots associated to
n TS 1xEV-DO Rev. A Radio Bearer Selection (Downlink) table None
downlink 1xEV-DO Rev. A bearer
DL Downlink RLC peak rate provided by
R RLC – peak Downlink 1xEV-DO Rev. A Radio Bearer Table None
the 1xEV-DO Rev. A bearer
Number of EVDO channel elements
EVDO – CE
N max  NI  Site parameter None available for a site on uplink and
downlink
Total number of EVDO channel
EVDO – CE
N  NI  Simulation result None elements of a site consumed by users
on uplink and downlink
TCH – C E – UL Number of channel elements used for

TCH on uplink
MacIndexes Maximum number of MAC indexes
N max  txi ic  Simulation constraint None
available per cell (59)
MacIndexes Number of MAC indexes used by the

N  txi ic  Simulation result None
cell
EVDO Maximum number of EVDO users that
n max  txi ic  Simulation constraint (cell parameter) None
can be connected to the cell
EVDO Number of EVDO users connected to

n  txi ic  Simulation result None
the cell


284 AT271_TRG_E6 © Forsk 2009

UL
Tx
Term
N0 NF Term  K  T  W W Thermal noise at terminal
Rc W bps Chip rate
UL
UL Simulation result bps Uplink data rate

R
UL Uplink data rate due to TCP

R TCP – ACK Simulation result bps
aknowledgements
Downlink data rate for Broadcast/

R BCMCS Cell parameter bps
Multicast services
DL Downlink maximum data rate supplied

R max Simulation result bps
to the terminal
DL
R avg Simulation result bps Downlink average cell data rate
DL DL
R application SF rate  R max – R bps Downlink user application throughput
SF Rate Service parameter % Scaling factor
R Service parameter kbps Offset
W Uplink service processing gain on

Gp
UL ---------- None
UL FCH
R
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
Pilot burst transmitted by the

P tx  txi ic b pilot  P max  txi ic  W
transmitter on carrier ic.
P max  txi ic  if users to support Traffic burst transmitted by the
P tx  txi ic b traffic  W
P max  txi ic   G idle – power if no user to support transmitter on carrier ic.
ER DRC Cell parameter % Error rate on the DRC channel
Pourcentage of EVDO timeslots

TS BCMCS Cell parameter % dedicated to Broadcast/Multicast
services
Pourcentage of EVDO timeslots

TS EVDO – CCH Cell parameter %
dedicated to control channels
Terminal power transmitted on carrier

P term  ic  Simulation result W
ic
min
max

transmitted
transmitted
finger
© Forsk 2009 AT271_TRG_E6 285


L Tx None Transmitter lossa
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
Result calculated from cell edge coverage probability and Ec/I0 Ec/I0 Shadowing margin
n=2 or 3 pilot signals at the mobile b.
None
UL npaths
n=2 or 3 in soft handoffc.
In prediction studiesd
For Ec/I0 and Ec/Nt calculations
---------------------------------------------------------------------------------------------------------------------------------------------------
G Tx  G term
LT None Transmitter-terminal total loss
UL
------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
In simulations
--------------------------------------------------------------------------------------------------------------------------------
G Tx  G term
DL P tx  txi ic b pilot  Pilot bust received at terminal from a

P tot  txi ic b pilot  -------------------------------------------- W
LT transmitter on carrier ic
DL P tx  txi ic b traffic  Traffic burst received at terminal from

P tot  txi ic b traffic  ----------------------------------------------- W
UL P term Bit received power at transmitter on

P b  ic  -------------- W
LT carrier ic
UL
NR threshold  txi ic  Cell parameter dB Cell uplink noise rise threshold
UL Cell uplink noise rise upgrading/

NR threshold  txi ic  Cell parameter dB
downgrading delta
a. L Tx = L total – UL on uplink and L Tx = L total – DL on downlink.

npaths
b. 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.
npaths
c. M Shadowing –  Eb  Nt  corresponds to the shadowing margin evaluated from the shadowing error probability
UL
286 AT271_TRG_E6 © Forsk 2009

d. In uplink prediction studies, only carrier power level is downgraded by the shadowing margin
UL
the shadowing model ( M Shadowing – Ec  Io ) while extra-cell interference level is not. Therefore, M Shadowing – Ec  Io is set to
1 in downlink extra-cell interference calculation.
7.2.2.2 Ec/I0 and Ec/Nt Calculations

Ec Ec Ec
This table details ------  txi ic b pilot  , ------  txi ic b pilot  and ------  txi ic b traffic  calculations.
I0 Nt Nt
Downlink intra-cell interference at

DL  txi ic 
I intra   0 W terminal on carrier ic (only one mobile
b
 pilot or b traffic  is served at a time)

DL Downlink extra-cell interference based
DL
I extra  ic b pilot  P tot  txj ic b pilot  W
on pilot at terminal on carrier ic
txj j  i

DL
I extra  ic b traffic  P tot  txj ic b traffic  W
on traffic at terminal on carrier ic
txj j  i
 Ptot  txj icadj bpilot 

DL
DL Downlink inter-carrier interference
I inter – carrier  ic b pilot   j
W
txj
----------------------------------------------------------------- based on pilot at terminal on carrier ic
Downlink inter-carrier interference

 Ptot  txj icadj btraffic 
DL
DL based on traffic at terminal on carrier
I inter – carrier  ic b traffic  txj  j
W
--------------------------------------------------------------------
- ic
DL DL DL DL term Total noise based on pilot received at

I 0  ic b pilot  P tot  txi ic b pilot  + I extra  ic b pilot  + I inter – carrier  ic b pilot  + N 0 W
DL DL DL DL term Total noise based on traffic received

I 0  ic b traffic  P tot  txi ic b traffic  + I extra  ic b traffic  + I inter – carrier  ic b traffic  + N 0 W
at terminal on carrier ic
DL DL term Total noise based on pilot received at

N tot  ic b pilot  I extra  ic b pilot  + N 0 W
DL DL term Total noise based on traffic received
N tot  ic b traffic  I extra  ic b traffic  + N 0 W
Q pilot  txi ic  DL
 BTS    P tot  txi ic b pilot  Pilot quality level at terminal on carrier
Ec --------------------------------------------------------------------------- None
 ------  txi ic b pilot 
DL
I 0  ic b pilot  ic
I0
DL
Ec  BTS    P tot  txi ic b pilot  Pilot quality level at terminal on carrier
------  txi ic b pilot  ------------------------------------------------------------------------------------------------------------------------
- None
Nt DL
N tot  ic b pilot  +  1 –  BTS   P tot  txi ic b pilot 
DL ic
DL
Ec  BTS    P tot  txi ic b traffic  Traffic quality level at terminal on
------  txi ic b traffic  -------------------------------------------------------------------------------------------------------------------------------
- None
Nt DL
N tot  ic b traffic  +  1 –  BTS   P tot  txi ic b traffic 
DL carrier ic
7.2.2.3 UL Eb/Nt Calculation

Eb
or  ------- ).
UL
This table details calculations of uplink quality ( Q
 Nt  UL

 Pb
UL
UL intra  ic 
term
ic
txi


UL
term
ic
txj j  i
© Forsk 2009 AT271_TRG_E6 287

 Pb
UL
 ic adj 
txj j
--------------------------------------
-

Tx
UL

N tot  txi ic  I tot  ic  + N 0 W
UL
-  G UL
--------------------------------------------------------------------------------------------------------------- p
Eb UL Tx
UL Quality level at transmitter on carrier
 txi ic    ------
UL
Q None
 N t  UL ica
UL
-  G UL
Total noise: --------------------------------------- p
UL
UL
No HO: Q  txi ic 

UL UL
tx k  ActiveSet
 samesite 
Soft, Softer/Soft HO (No MRC):
to combination of all transmitters of
UL
Max  Q tch  tx k ic    G macro – diversity
UL the active set located at the same site
tx k  ActiveSet and taking into account increase of the
UL
Q total  ic  None quality due to macro-diversity (macro-
Softer/Soft HO (MRC): diversity gain).
 
Max  UL 

UL UL UL
 f rake efficiency  Q tch  tx k ic  Q tch  tx l ic  In simulations, G macro – diversity = 1 .
k l  
tx  samesite  tx k 
k
tx  othersite
l
UL
UL
UL Q total  ic 
G SHO --------------------------------------------------------- None Uplink soft handover gain on carrier ic
UL
Q  BestServer ic 
For 1xEV-DO Rev 0 terminal
E
 ------
c
UL
UL
 G p   1 + G ACK + G DRC + G TCH 
 N t  min
For 1xEV-DO Rev A terminalb

UL
When the acknoledgement signal is considered
Q req None Eb/Nt target on uplink
E
UL
-----c- 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
E
 ------
c
UL
UL
 G p   1 + G RRI + G DRC + G TCH + G Auxiliary – Pilot 
 N t  min
UL
req Q req Required terminal power to achieve
P term  ic  ------------------------  P term W
UL
Q total  ic  Eb/Nt target at transmitter on carrier ic
b. In simulations, the uplink Eb/Nt target is calculated whithout considering the aknoledgement signal.
7.2.2.4 Simulation Results

This table contains some simulation results provided in the Cells and Mobiles tabs of the simulation property dialog.

Downlink intra-cell interference at
DL DL
I intra  txi ic b traffic   1 – F ortho   BTS   P tot  txi ic b traffic  = 0 W terminal on carrier ic (only one mobile
is served at a time)
288 AT271_TRG_E6 © Forsk 2009


DL
I extra  ic b traffic  P tot  txj ic b traffic  W
on traffic at terminal on carrier ic
txj j  i
Downlink inter-carrier interference

DL
DL based on traffic at terminal on carrier
I inter – carrier  ic b traffic  txj  j
W
--------------------------------------------------------------------
- ic
Total effective interference based on

DL DL DL DL
I tot  ic b traffic  I intra  ic b traffic  + I extra  ic b traffic  + I inter – carrier  ic b traffic  W traffic at terminal on carrier ic (after
unscrambling)
DL DL term Total noise based on traffic received
N tot  ic b traffic  I tot  ic b traffic  + N 0 W

 Pb
UL
UL intra  ic 
term
ic
txi


UL
term
ic
txj j  i
 Pb
UL
 ic adj 
txj j
--------------------------------------
-

Tx
UL

Number of mobiles connected to
N mobiles  txi ic  Simulation result None
transmitter txi on carrier ic
DL
DL
I tot  ic b traffic 
X  txi ic  ---------------------------------------
- None Cell downlink load factor on carrier ic
DL
N tot  ic b traffic 
UL
UL
X  txi ic  -----------------------------
UL
UL
F
UL
UL intra Tx

E
UL
 txi ic  ------------------------------ None
UL carrier ic
F  txi ic 
7.3 Active Set Management

Active set (AS) management is detailed hereafter. The active set may consist of one or more cells; depending on whether
the service supports soft handoff and on the terminal active set size. The quality of the pilot (Ec⁄I0) is what determines
whether or not a cell can belong to the active set. Cells entering the mobile’s active set must fulfill the following conditions:
1. The best server (first cell entering active set)
The pilot quality from the best server cell must exceed an upper threshold equal to the sum of the Min. Ec/I0
defined in the properties of the best serer cell and the Delta Min. Ec/I0 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. Best server cell is the
cell with the highest pilot quality.
2. Other cells of active set

- They must use the same carrier as the best server cell,
- The pilot quality from other candidate cells must exceed a lower threshold. The lower threshold depends both
on the type of carrier and the mobility type. 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.
© Forsk 2009 AT271_TRG_E6 289

- If you have selected to restrict the active set to neighbours, the cell must be a neighbour of the best server
(the “restricted to neighbours” option is selected in the equipment properties).
7.4 Simulations
The simulation process is divided into two steps:
1. Obtaining a realistic user distribution
Atoll generates a user distribution using a Monte-Carlo algorithm, which requires traffic maps and data as input. The result-
ing user distribution complies with the traffic database and maps provided to the algorithm.
Each user is assigned a service, a mobility type, and an activity status by random trial, 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. A user may be either active or inactive. Both active and inactive users
consume radio resources and create interference.
Additionally, each 1xEV-DO Rev. 0 user is assigned a transition flag ("True" or "False") for each possible data rate transi-
tion (from 9.6 to 19.2 kbps, 19.2 to 38.4 kbps, 38.4 to 76.8 kbps, and 76.8 to 153.6 kbps for data rate upgrading and from
153.6 to 76.8 kbps, 76.8 to 38.4 kbps, 38.4 to 19.2 kbps, and 19.2 to 9.6 kbps for data rate downgrading). These transition
flags are based on the data rate downgrading and upgrading probabilities. If a transition flag is "True," the user data rate
can be downgraded or upgraded if necessary.
Then, Atoll randomly assigns a shadowing error to each user using the probability distribution that describes the shadow-
ing effect.
Finally, 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. Modelling the network regulation mechanism
This algorithm depends on the network. Atoll uses a power control algorithm in case of cdmaOne/CDMA2000 1xRTT
networks and a different algorithm, which mixes data rate control on downlink and power control on uplink, for CDMA2000
1xEV-DO networks.

7.4.1.1 Number of Users, User Activity Status and User Data Rate
During the simulation, a first random trial is performed to determine the number of users and their activity status. The deter-
mination of the number of users and the activity status allocation depend on the type of traffic cartography used.
Note:
• Atoll follows a Poisson distribution to determine the number of total connected users before
each simulation. In order to make Atoll use a constant number of total connected users, the
following lines must be added to the Atoll.ini file:
[CDMA]
RandomTotalUsers=0
7.4.1.1.1 Simulations Based on Raster Traffic and Vector Traffic Maps

a list of user profiles with an associated mobility type and a given density (number of subscribers with the same profile per
km²).
Vector traffic maps are directly based on user profiles. Each polygon and line of the map is assigned a density of subscrib-
ers with given user profile and mobility type. If the map is composed of points, each point is assigned a number of subscrib-
ers with given user profile and mobility type.
The user profile models the behaviour of the different subscriber categories. Each user profile contains a list of services
and their associated parameters describing how these services are accessed by the subscriber.
From environment (or polygon) surface (S) and user profile density (D), a number of subscribers (X) per user profile is
inferred.
X = SD
Notes:
• In case of vector traffic map composed of lines, 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 the vector traffic map is composed of
points.
290 AT271_TRG_E6 © Forsk 2009

For each behaviour described in a user profile, according to the service, frequency use and exchange volume, Atoll calcu-
lates the probability for the user being connected in uplink and in downlink at an instant t.
Calculation of the service usage duration per hour ( p 0 : probability of being connected):
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).
Then, Atoll deduces the total number of users trying to access a certain service.
Calculation of the total number of users trying to access the service j ( n j ):
nj = X  p0
The next step determines the activity status of each user. This one depends on the type of service (Voice, 1xRTT data,
1xEV-DO data…).
IS-95 cdmaOne and CDMA2000 1xRTT Services
Activity status of voice and data service users is determined as follows.

Users are always active on FCH in both directions, uplink and downlink. Therefore, 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, for voice and data services, 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 Service Users

Voice service users are active on uplink and downlink. However, the FCH can have inactivity periods on both links. This
UL DL
is modelled by the FCH activity factor, AFFCH and AF FCH . Therefore, all voice service users try to access the service
UL UL DL DL
with the following FCH rates, R FCH  AF FCH on uplink and R FCH  AF FCH on downlink.
UL DL
R FCH and R FCH are respectively the uplink and downlink FCH nominal rates.
• Data Service Users

Data service users are active on uplink and downlink. FCH is always allocated but can have inactivity periods on both links;
UL DL
this is modelled by the FCH activity factor, AFFCH and AF FCH . SCH may be allocated with four possible rates (2x, 4x,
8x and 16xFCH nominal rate).
Therefore, data service users can access the service with different rates. Possible rates are detailed in the table below:
Allocated rates
SCH rate factor r k
On UL On DL
UL UL DL DL
Only FCH is used - R FCH  AF FCH R FCH  AF FCH
UL UL DL DL
2x R FCH   AF FCH + 2  R FCH   AF FCH + 2 
UL UL DL DL
Both FCH and SCH are used
UL UL DL DL
UL UL DL DL
UL DL
R FCH and R FCH are respectively the uplink and downlink FCH nominal rates.
© Forsk 2009 AT271_TRG_E6 291

Then, Atoll determines the distribution of users between the different possible rates.
UL DL
In case of a data service, j, several data rate probabilities, P k and P k , can be assigned to different rate factors, r k , for
SCH channel.
Note:
• For non-data services, these probabilities are 0.
For data service users, a random trial compliant with data rate probabilities is performed for each link in order to determine
the rate for each user.
On uplink, we have:
rk UL UL
For each SCH rate factor, r k , the number of users n j with the data rate R FCH   AF FCH + r k  is calculated as follows,
rk UL
nj = Pr  nj
k
FCH UL UL
Therefore, the number of users n j with the data rate, R FCH  AF FCH , is:
rk
 nj
FCH
nj = nj –
rk
On downlink, we have:
rk DL DL
For each SCH rate factor, r k , the number of users, n j with the data rate, R FCH   AF FCH + r k  , is calculated as follows,
rk DL
nj = Pr  nj
k
FCH DL DL
Therefore, the number of users n j with the data rate, R FCH  AF FCH , is:
rk
 nj
FCH
nj = nj –
rk
CDMA2000 1xEV-DO Services
As power control is performed on uplink only, 1xEV-DO data service users will be considered either active on uplink or
inactive. 1xEV-DO data Rev. 0 service users can access the service with uplink rates of 9.6, 19.2, 38.4, 76.8 and 153.6
kbps. 1xEV-DO data Rev. A service users can access the service with uplink rates of 4.8, 9.6, 19.2, 38.4, 76.8, 115.2,
153.6, 230.4, 307.2, 460.8, 614.4, 921.6, 1,228.8 and 1,848.2 kbps.
UL UL
For each service, j, several data rate probabilities, P k , can be assigned to different rates R k . The number of users
active on uplink ( n j  UL  ) and the number of inactive users ( n j  inactive  ) are calculated as follows:
 Pk
UL UL
Probability of being active on UL: p UL =  Rk 
UL
Rk
 Pk
UL UL
Probability of being inactive: p inactive = 1 –  Rk 
UL
Rk
Probability of being active on UL and DL both: p UL + DL = 0
Therefore, 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
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, Atoll determines the distribution of users between the different possible rates, R k . The number of users with the
UL UL
data rate R k , n j  R k  , is calculated as follows:
UL UL
nj  Rk  = Pk  nj
Inactive users have a requested data rate equal to 0.
292 AT271_TRG_E6 © Forsk 2009

Notes:
• The user distribution per service is an average distribution and the service of each user is
randomly drawn in each simulation. Therefore, if you compute several simulations at once,
the average number of users per service will correspond to the calculated distribution. But if
you check each simulation, the user distribution between services is different in each of
them.
It is the same for the SCH rate distribution between 1xRTT data service users and the
traffic data rate distribution between 1xEV-DO data service users.
• In calculations detailed above, we assume that the sum of data rate probabilities is less
than or equal to 1. If the sum of data rate probabilities exceeds 1, Atoll considers
 
normalised data rate probabilities values, P r  
k
  Pr  , instead of specified data rate
k
rk
probabilities P r .
k
7.4.1.1.2 Simulations Based on Traffic Maps per Service and per Transmitter
IS-95 cdmaOne and CDMA2000 1xRTT Services
• Voice Service Users

This calculation can be divided into several steps. First of all, Atoll determines the number of users active on UL and DL
using the service j in the Txi cell.
For each transmitter, Txi, and each service j:
- Either Atoll deduces the number of users active on UL and DL using the service j in the Txi cell, if you have
selected the throughput map option.
UL DL
Rt Rt
- and N DL = ---------
N UL = --------- -
UL DL
Rj Rj
Where,
UL
Rt is the number of kbits per second transmitted on UL in the Txi cell to provide the service j to the users
(user-defined value in the traffic map properties)
DL
Rt is the number of kbits per second transmitted on DL in the Txi cell to provide the service j to the users
(user-defined value in the traffic map properties).
UL DL
Rj and R j correspond to uplink and downlink rates of a user. FCH is always allocated to active users but
UL UL UL UL
can have inactivity periods on both links. Therefore, we have R j = R FCH  AF FCH (where R FCH is the serv-
UL
ice FCH nominal rate on UL and AF FCH corresponds to the FCH activity factor on UL) and
DL DL DL DL DL
Rj = R FCH  AFFCH (where R FCH is the service FCH nominal rate on DL and AF FCH corresponds to the
FCH activity factor on DL).
- Or Atoll directly uses the defined NUL and NDL values (number of users active on UL and DL using the service
j in the Txi cell), if you have selected the user map option.
Users active on UL and DL both are included in the NUL and NDL values. Therefore, it is necessary to accurately determine
the number of users active on UL ( n j  UL  ), on DL ( n j  DL  ) and on UL and DL ( n j  UL + DL  ) both.
Users are always active on FCH for both links. Therefore, we have following activity probabilities.
Let us consider n j  active  , the total number of active users using the service j in the Txi cell
We have,
 p UL + p UL + DL   n j  active  = N UL
 p DL + p UL + DL   n j  active  = N DL
Thus,
© Forsk 2009 AT271_TRG_E6 293

n j  UL + DL  = min  ------------------------------------------------------------------ ------------------------------------------------------------------
 p UL   1 – p DL  + p UL + DL p DL   1 – p UL  + p UL + DL
n j  UL  = N UL – n j  UL + DL 
n j  DL  = N DL – n j  UL + DL 
and, n j  active  = n j  UL  + n j  DL  + n j  UL + DL 
The number of inactive users trying to access the voice service j, n j  inactive  , is,
n j  active 
n j  inactive  = ------------------------------  p inactive = 0
1 – p inactive
Therefore, the total number of users trying to access the voice service j, n j , is,
n j = n j  active 
UL UL DL DL
Voice service users try to access the service with the FCH rates, R FCH  AF FCH on uplink and R FCH  AF FCH on down-
link.
All user characteristics determined, a second random trial is performed to obtain their geographical positions.
• Data Service Users
FCH is always allocated to active users but can have inactivity periods on both links. This is modelled by the FCH activity
UL DL
factors, AFFCH and AF FCH . SCH may be allocated with four possible rates (2x, 4x, 8x, 16xFCH nominal rate). Several
UL DL
data rate probabilities, P k and P k , can be assigned to different rates factor, r k , for SCH channel.
Note:
• For non-data services, these probabilities are 0.
Atoll determines the number of users active on UL and DL using the service j in the Txi cell. For each transmitter, Txi, and
each service j:
- Either Atoll deduces the number of users active on UL and DL using the service j in the Txi cell, if you have
selected the throughput map option.
UL DL
Rt Rt
- and N DL = ---------
N UL = --------- -
UL DL
Rj Rj
Where,
UL
Rt is the number of kbits per second transmitted on UL in the Txi cell to provide the service j (user-defined
value in the traffic map properties).
DL
Rt is the number of kbits per second transmitted on DL in the Txi cell to provide the service j (user-defined
value in the traffic map properties).
UL DL
Rj and R j correspond to uplink and downlink rates of a user.
 UL
  rk + AFFCH   RFCH  Pr  Pr
UL UL UL UL UL UL
Rj = + 1 –   R FCH  AF FCH
k
 k

rk rk
 DL
  rk + AFFCH   RFCH  Pr  Pr
DL DL DL DL DL DL
Rj = + 1 –   R FCH  AF FCH
k
 k

rk rk
UL DL
R FCH and R FCH are the uplink and downlink FCH nominal rates respectively.
Note:
 
k
  Pr  , k
instead of specified data rate
rk
probabilities P r .
k
- Or Atoll directly uses the defined values for NUL and NDL (number of users active on UL and DL using the
service j in the Txi cell), if you have selected the user map option.
294 AT271_TRG_E6 © Forsk 2009

Users active on UL and DL both are included in the NUL and NDL values. Therefore, it is necessary to accurately determine
the number of users active on UL ( n j  UL  ), on DL ( n j  DL  ) and on UL and DL ( n j  UL + DL  ) both.
Users are always active on FCH for both links. Therefore, we have following activity probabilities.
Let us consider n j  active  , the total number of active users using the service j in the Txi cell
We have,
 p UL + p UL + DL   n j  active  = N UL
 p DL + p UL + DL   n j  active  = N DL
Thus,
n j  UL + DL  = min  ------------------------------------------------------------------ ------------------------------------------------------------------
 p UL   1 – p DL  + p UL + DL p DL   1 – p UL  + p UL + DL
n j  UL  = N UL – n j  UL + DL 
n j  DL  = N DL – n j  UL + DL 
and, n j  active  = n j  UL  + n j  DL  + n j  UL + DL 
The number of inactive users trying to access the service j, n j  inactive  , is,
n j  active 
n j  inactive  = ------------------------------  p inactive = 0
1 – p inactive
Therefore, the total number of users trying to access the service j, n j , is,
n j = n j  active 
As explained above, data service users can access the service with different rates. Possible rates are detailed in the table
below:
Allocated rates
SCH rate factor r k
On UL On DL
UL UL DL DL
Only FCH is used - R FCH  AF FCH R FCH  AF FCH
UL UL DL DL
UL UL DL DL
Both FCH and SCH are used
UL UL DL DL
UL UL DL DL
Atoll determines the distribution of users with the different possible rates. A random trial compliant with data rate probabil-
ities is performed for each link in order to determine the data rate of each user.
On uplink, we have,
rk UL UL
For each SCH rate factor, r k , the number of users n j with the data rate R FCH   AF FCH + r k  is calculated as follows,
rk UL
nj = Pr  nj
k
FCH UL UL
Therefore, the number of users n j with the data rate, R FCH  AF FCH , is,
rk
 nj
FCH
nj = nj –
rk
On downlink, we have,
rk DL DL
For each SCH rate factor, r k , the number of users, n j with the data rate, R FCH   AF FCH + r k  , is calculated as follows,
© Forsk 2009 AT271_TRG_E6 295

k DL
nj = Pk  nj
FCH DL DL
Therefore, the number of users n j with the data rate, R FCH  AF FCH , is,
rk
 nj
FCH
nj = nj –
rk
CDMA2000 1xEV-DO Services
As power control is performed on uplink only, 1xEV-DO data service users will be considered either active on uplink or
inactive. 1xEV-DO data Rev. 0 service users can access the service with uplink rates of 9.6, 19.2, 38.4, 76.8 and 153.6
kbps. 1xEV-DO data Rev. A service users can access the service with uplink rates of 4.8, 9.6, 19.2, 38.4, 76.8, 115.2,
153.6, 230.4, 307.2, 460.8, 614.4, 921.6, 1,228.8 and 1,848.2 kbps.
UL UL
For each service, j, several data rate probabilities, P k , can be assigned to different uplink rates R k . The number of
users active on uplink ( n j  UL  ) and the number of inactive users ( n j  inactive  ) are calculated into several steps. First
of all, Atoll determines the number of users active on UL using the service j in the Txi cell.
For each transmitter, Txi, and each service j:
- Either Atoll deduces the number of users active on UL using the service j in the Txi cell, if you have selected
the throughput map option.
UL
Rt
N UL = ---------
-
UL
Rj
Where:
UL
Rt is the number of kbits per second transmitted on UL in the Txi cell to provide the service j (user-defined value in the
traffic map properties).
UL
Rj corresponds to the uplink data rate for a user.
 Pk
UL UL UL
Rj =  Rk
k
Note:
 
k
  Pr  , instead of specified data rate
k
rk
probabilities P r .
k
- Or Atoll directly uses the defined values of NUL (number of users active on UL using the service j in the Txi
cell), if you have selected the user map option.
Activity probabilities are:
 Pk
UL UL
Probability of being active on UL: p UL =  Rk 
UL
Rk
 Pk
UL UL
Probability of being inactive: p inactive = 1 –  Rk 
UL
Rk
Therefore, we have:
Number of users active on UL: n j  UL  = N UL  p UL
Number of inactive users: n j  inactive  = N UL  p inactive
Number of users active on DL: n j  DL  = 0
Number of users active on UL and DL both: n j  UL + DL  = 0
n j = n j  UL  + n j  inactive 
296 AT271_TRG_E6 © Forsk 2009

Then, Atoll determines the distribution of users with the different possible rates. The number of users with the data rate
UL UL
R k , n j  R k  , is calculated as follows:
UL UL
nj  Rk  = Pk  nj
Inactive users have a requested data rate equal to 0.

Note:
• The user distribution per service is an average distribution and the service of each user is
randomly drawn In each simulation. Therefore, if you compute several simulations at once,
the average number of users per service will correspond to the calculated distribution. But if
you check each simulation, the user distribution between services is different in each of
them.
It is the same for the SCH rate distribution between 1xRTT data service users and the
traffic data rate distribution between 1xEV-DO data service users.
7.4.1.2 Transition Flags for 1xEV-DO Rev.0 User Data Rates

For 1xEV-DO Rev. 0 services supporting data rate downgrading, you can define the probability of the service being
UL UL UL UL UL
upgraded ( P Upg – k  R k  ) or downgraded ( P Downg – k  R k  ) on the uplink (reverse link) for each data rate ( R k ). The
probabilities are taken into account in order to determine if a user with a certain data rate can be upgraded or downgraded.
User data rate downgrading and upgrading occur during congestion control when the cell is over- or underloaded.
The following table shows the data rate changes that are possible when a data rate is upgraded or downgraded. The prob-
abilities are defined with a number from 1 to 255 for each data rate.
Possible Data Rate Changes Possible Data Rate Changes

During Upgrading During Downgrading
From To From To
9.6 kbps 19.2 kbps 153.6 kbps 76.8 kbps
During the generation of the user distribution, each 1xEV-DO Rev. 0 user is assigned a random number between 1 and
255 for each possible data rate transition. When this number is lower or equal to the value of the probability, the transition
flag for this data rate transition is set to "True" meaning that this data rate transition can be performed if necessary.
UL
The number of 1xEV-DO Rev. 0 users with a certain data rate that can be downgraded ( n j  R k  Downg ) and upgraded
UL
( n j  R k  Upg ) are calculated as follows:
UL UL UL
UL P Upg – k  R k   n j  R k 
n j  R k  Upg = -------------------------------------------------------------
-
255
And
UL UL UL
UL P Downg – k  R k   n j  R k 
n j  R k  Downg = --------------------------------------------------------------------
-
255
Note:
• The number of users with a certain data rate that can be downgraded or upgraded is an
average. Therefore, if you compute several simulations at once, the average number of
users with a certain data rate that can be downgraded or upgraded will correspond to the
calculated value. But if you check each simulation, this number is different in each of them.
7.4.1.3 User Geographical Position

Once all the user characteristics determined, 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.
7.4.2 Network Regulation Mechanism

7.4.2.1 IS-95 cdmaOne Power Control Simulation Algorithm
The IS-95 cdmaOne network automatically regulates itself by using uplink and downlink power control in order to minimize
interference and maximize capacity. Atoll simulates the power control mechanism for each user distribution. Simulations
use an iterative algorithm, where in each iteration, all the mobiles selected during the user distribution generation (1st step)
try to connect to the network’s active transmitters with a calculation area. The process is repeated for each iteration and
© Forsk 2009 AT271_TRG_E6 297

ends once the network is balanced, i.e. when the convergence criteria (on UL and DL) are satisfied. Algorithm steps are
detailed below.
Figure 7.1: IS-95 cdmaOne Power Control Algorithm
7.4.2.1.1 Algorithm Initialization

The total power transmitted by the base station txi on carrier ic P Tx  txi ic  is initialised to
P pilot  txi ic  + P sync  txi ic  + P paging  txi ic  .
Uplink received powers by the base station txi on carrier ic, I tot  txi ic  , I tot  txi ic  and I inter – carrier  txi ic  , are
initialised to 0 W (no connected mobile).
UL
UL I tot  txi ic 
 X k  txi ic  = -----------------------------
- = 0
UL
7.4.2.1.2 Presentation of the Algorithm

UL
The algorithm is detailed for any iteration k. Xk is the value of the variable X at the iteration k. In the algorithm, all Q req
DL
For each mobile Mb
Determination of Mb’s Best Server (SBS(Mb))
For each station txi containing Mb in its calculation area,
Determination of BestCarrier k  txi M b  .
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.

UL
- + X
X k  txi ic  = -----------------------------
UL
EndFor
UL
298 AT271_TRG_E6 © Forsk 2009


UL UL

Calculation of Q pilot  txi M b BestCarrier  = ------------------------------------------------------------------------------------------------
k DL
I 0  BestCarrier k  txi M b  

req
If Q pilot  txi M b BestCarrier   Q pilot then txi is rejected by Mb
k
max
k k
Admission control (If simulation respects a load factor constraint and Mb was not connected in previous iteration).
UL UL
If X K  txi BestCarrier k  txi M b    X max , then txi is rejected by Mb
Else
max
k k
tx BS  M b  = txi
Endif
EndFor
If no txBS has been selected, Mb cannot get a connection to the network
In the following lines, we will consider ic = BestCarrier k  tx BS  M b  M b 
Determination of the Active Set
For each station txi containing Mb in its calculation area, using ic, and if neighbours are used, neighbour of txBS(Mi)
   BTS  P c  txi M b ic 
Calculation of Q pilot  txi M b ic  = -----------------------------------------------------------------
-
k DL
I 0  ic 

pilot
If Q pilot  txi M b ic   Q min then txi is rejected by Mb
k
Else txi is included in the Mb active set

Rejection of txi if the Mb active set is full

k
EndFor
req
Calculation of the required power for Mb, P term  M b ic  k
For each cell (txi,ic) present in the Mb active set
Calculation of quality level on Mb traffic channel at (txi,ic), with the minimum power allowed on traffic channel for the Mb
service
FCH –r eq
FCH – UL P term  M b ic  k – 1
Pb  txi M b ic  = -----------------------------------------------------
-
FCH –U L
UL  term  P b  txi M b ic  FCH – UL
Q FCH  txi M b ic  k = -------------------------------------------------------------------------------------------------------------------------------------------------
UL FCH – U L
 Gp  Service 
N tot  txi ic  –  1 – F MUD    term  P b  txi M b ic 
If the user selects the option “Total noise”

FCH – U L
UL  term  P b  txi M b ic 
-  G FCH
Q FCH  txi M b ic  k = ------------------------------------------------------------------------ p
– UL
 Service 
UL
© Forsk 2009 AT271_TRG_E6 299

End For
If (Mb is not in handoff)
UL UL
Q FCH  M b  = Q FCH  txi M b ic  k
k
Else if (Mb is in softer handoff)

UL UL UL
Q FCH  M b  = f rake efficiency  Q FCH  txi M b ic  k
k
txi  ActiveSet
Else if (Mb is in soft or softer/soft without MRC)
UL UL UL
Q FCH  M b  = Max  Q FCH  txi M b ic  k    G macro – diversity  2 links
k
txi  ActiveSet
Else if (Mb is in soft/soft)
UL UL UL
Q FCH  M b  = Max  Q FCH  txi M b ic  k    G macro – diversity  3 links
k
txi  ActiveSet
Else if (Mb is in softer/soft with MRC)
 
 UL 

UL UL UL UL
Q FCH  M b  = Max  f rake efficiency  Q FCH  txi M b ic  Q FCH  txi M b ic    G macro – diversity  2 links
k
 other site 
 txi  ActiveSet 
(same site)
EndIf
UL
FCH – r eq  Q req  Service  M b  Mobility  M b    FCH FCH –r eq
P term  M b ic  k = -----------------------------------------------------------------------------------------------------------
UL
 P term  M b ic  k – 1
Q FCH  M b 
k
req FCH –r eq min

P term  M b ic  k = Max  P term  M b ic  k P term  txi M b  
FCH – r eq max
If P term  M b ic  k  P term  M b  then Mb cannot select any station and its active set is cleared

Calculation of quality level on (txi,ic) FCH at Mb, with the minimum power allowed on FCH for the Mb service
min
FCH – D L P FCH  Service  M b  
Pb  txi M b ic  = ------------------------------------------------------
-
FCH – D L
DL  BTS  P b  txi M b ic  FCH – DL
Q FCH  txi M b ic  k = ---------------------------------------------------------------------------------------------------------------------------------------
DL FCH – D L
 Gp  Service  M b  
N tot  ic  –  1 – F ortho    BTS  P b  txi M b ic 

FCH –D L
DL  BTS  P b  txi M b ic 
-  G FCH
Q FCH  txi M b ic  k = ------------------------------------------------------------------------ p
– DL
 Service  M b  
DL
N tot  ic 
End For

DL DL DL
k
txi  ActiveSet  FCH 
Do
Calculation of the required power for DL traffic channel between (txi,ic) and Mb:
DL
req Q req  Service  M b  Mobility  M b   FCH
-  P min
P FCH  txi M b ic  k = ------------------------------------------------------------------------------------------------------ FCH  Service  M b  
DL
Q FCH  M b 
k
req max
If P FCH  txi M b ic  k  P FCH  Service  M b   then  txi ic  is excluded from Mb active set
DL
Recalculation of a decreased Q req
300 AT271_TRG_E6 © Forsk 2009

EndFor

DL DL DL
k
txi  ActiveSet  FCH 
DL DL
While Q FCH  M b   Q req  Service  M b  Mobility  M b   FCH and Mb active set is not empty
k
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)
UL
Update of N tot  S j ic 
EndFor
For each mobile Mi
DL
EndFor
Control of Radio Resource Limits (Walsh Codes, Cell Power and Site Channel Elements)
For each cell (txi,ic) on a site Nl
P tx  txi ic  k DL
While ------------------------------  %Power max
P max
req
Rejection of mobile with highest P FCH  txi 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 (txi,ic) containing rejected users
Control of the available power for the other cells (txj,ic) of the site where power pooling between transmitters is not acti-
vated

DL
If  %Power max  P max – P tx  txj ic  k   0
 txj ic 
txj  N l
Then, the power unused by the cells (txj,ic) of the site can be allocated to cells (txi,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  txi ic  k + P FCH  txi M b ic  k  %Power max  P max + M Pooling  txi ic 
Mb is reconnected
EndIf
EndFor
EndIf
EndFor
Codes Codes
While N  txi ic  k  N max  txi ic 
Rejection of last admitted mobile

EndFor
For each site (Node B) Nl
CE – DL CE – DL
While N  N I  k  N max  NI k
req
Rejection of mobile with highest P tch  txi M b  k for the lowest service priority
© Forsk 2009 AT271_TRG_E6 301

CE – UL CE – UL
While N  N I  k  N max  NI k
req
Rejection of mobile with highest P term  M b ic  k for the lowest service priority
EndFor
UL UL
For each cell (txi,ic) with X  txi ic   X max
Rejection of a mobile with the lowest service priority

EndFor
UL UL
While at least one cell with X  txi ic   X max exists
7.4.2.1.3 Convergence Criterion

The convergence criteria are evaluated at each 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 
-  100  int  ma x ------------------------------------------------------------------
 DL = max  int  ma x -----------------------------------------------------  100 
  Stations P tx  ic  k   Stations N
DL
 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 
 UL = max  int  ma x -----------------------------------------------------
-  100  int  ma x ------------------------------------------------------------------  100 
  Stations UL
I  ic    Stations N
UL
 ic  
tot k user k
1st case: Between two successive iterations,  UL and  DL are lower (  ) than their respective thresholds (defined
when creating a simulation).
has been achieved.
2nd case: After 30 iterations,  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.

1. After the 30th 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. Convergence has not been achieved.
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.
If  UL and/or  DL are still strictly higher than their respective thresholds, the simulation has not converged (specific diver-
gence symbol).
If  UL and  DL are lower than their respective thresholds, the simulation has converged.
7.4.2.2 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. Atoll simulates this network regulation mechanism with an iterative algorithm and calculates, for each user distri-
bution, network parameters such as base station power, mobile terminal power, active set and handoff status for each
terminal.
The power control simulation is based on an iterative algorithm, where in each iteration, all the mobiles selected during
the user distribution generation (1st step) try to connect to network active transmitters with a calculation area. The process
is repeated from iteration to iteration until convergence is achieved. The algorithm steps are detailed below.
302 AT271_TRG_E6 © Forsk 2009

Figure 7.2: CDMA2000 1xRTT Power Control Algorithm

Total power on carrier ic, P Tx  ic  , of base station Sj is initialised to P pilot  ic  + P sync  ic  + P paging  ic  .
Uplink received powers on carrier ic, I tot  ic  , I tot  ic  and I inter – carrier  ic  , at base station Sj are initialised to 0 W
(no connected mobile).
UL
UL I tot  S j ic 
 X k  S j ic  = ----------------------------
- = 0
UL
N tot  S j ic 

UL
The algorithm is detailed for any iteration k. Xk is the value of the variable X at the iteration k. In the algorithm, all Q req
DL
For each mobile Mi
Determination of Mi’s Best Server (SBS(Mi))
For each station Sj containing Mi in its calculation area,
Determination of BestCarrier k  S j M i  .
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.
Else the carrier selection mode defined for Sj is considered.

For each carrier ic used by Sj, we calculate current loading factor:
UL
UL I tot  S j ic  UL
X k  S j ic  = ----------------------------
- + X
UL
EndFor
UL
BestCarrier k  S j M i  is the carrier with the lowest X k  S j ic 
BestCarrier k  S j M i  is the carrier with the lowest P tx  S j ic  k
© Forsk 2009 AT271_TRG_E6 303

BestCarrier k  S j M i  is randomly selected

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  = --------------------------------------------------------------------------------------------
-
k DL
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).
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, Mi cannot get a connection to the network
In the following lines, we will consider ic = BestCarrier k  S BS  M i  M i 
For each station Sj containing Mi in its calculation area, using ic, and if neighbours are used, neighbour of SBS(Mi)
   BTS  P c  M i S j 
Calculation of Q pilot  M i S j ic  = ------------------------------------------------------
-
k DL
I 0  ic 
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

k
EndFor
req
Calculation of the required power for Mi, P term  M i ic  k
For each cell (Sj,ic) present in the Mi active set
Calculation of quality level on Mi traffic channel at (Sj,ic), with the minimum power allowed on traffic channel for the Mi
service
FCH – r eq SCH – r eq
FCH – U L P term  M i ic  k – 1 SCH – U L P term  M i ic  k – 1
Pb  M i S j ic  = ---------------------------------------------------- and P b  M i S j ic  = ----------------------------------------------------
-
L T  M i S j  L T  M i S j 
FCH –U L
UL  term  P b  M i S j ic  FCH – UL
Q FCH  M i S j ic  k = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
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  
SCH – U L
UL  term  P b  M i S j ic  SCH – UL
Q SCH  M i S j ic  k = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
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  
304 AT271_TRG_E6 © Forsk 2009

FCH – U L
UL  term  P b  M i S j ic 
-  G FCH
Q FCH  M i S j ic  k = --------------------------------------------------------------------- p
– UL
 Service 
UL
N tot  ic 
SCH –U L
UL  term  P b  M i S j ic 
-  G SCH
Q SCH  M i S j ic  k = --------------------------------------------------------------------- p
– UL
 Service 
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

UL UL UL
Q FCH  M i  = f rake efficiency  Q FCH  M i S j ic  k
k
S j  ActiveSet

UL UL UL
Q SCH  M i  = f rake efficiency  Q SCH  M i S j ic  k
k
S j  ActiveSet
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
S j  ActiveSet
UL 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
S j  ActiveSet
UL 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)
 
 
= Max  f rake efficiency    ic    G macro – diversity  2 links
UL UL UL UL UL
Q FCH  M i  Q FCH  ic  Q FCH
k  
 i AS  ActiveSet other site 
 
(same site)
 
 
= Max  f rake efficiency    ic    G macro – diversity  2 links
UL UL UL UL UL
Q SCH  M i  Q SCH  ic  Q SCH
k  
 i AS  ActiveSet other site 
 
(same site)
EndIf
UL
FCH –r eq  Q req  Service  M i  Mobility  M i    FCH FCH – r eq
P term  M i ic  k = --------------------------------------------------------------------------------------------------------
UL
 P term  M i ic  k – 1
Q FCH  M i 
k
UL
SCH – r eq  Q req  Service  M i  Mobility  M i  SCH_rate_multiple   SCH
P term -  P SCH
 M i ic  k = ------------------------------------------------------------------------------------------------------------------------------------------------------------- term
–r eq
 M i ic  k – 1
UL
Q SCH  M i 
k
req FCH – r eq SCH – r eq

P term  M i ic  k = P term  M i ic  k + P term  M i ic  k
req min
If P term  M i ic  k  P term  M i  then
min
FCH – r eq P term  M i S j 
P term -  P FCH
 M i ic  k = --------------------------------- term
– r eq
 M i ic  k
req
P term  M i  k
min
SCH – r eq P term  M i S j 
P term -  P SCH
 M i ic  k = --------------------------------- term
– r eq
 M i ic  k
req
P term  M i  k
EndIf
FCH – r eq max
If P term  M i ic  k  P term  M i  then Mi cannot select any station and its active set is cleared
© Forsk 2009 AT271_TRG_E6 305

req max
If P term  M i ic  k  P term  M i  and Mi uses SCH then:
Downgrading the service SCH rate:

req max UL UL
While P term  M i ic  k  P term  M i  and R SCH  Service  M i    R FCH  Service  M i    2
UL
UL R SCH  Service  M i  
R SCH  Service  M i    ------------------------------------------------------
2
SCH – r eq UL UL
SCH – r eq P term  M i ic  k  Q req  Service  M i  Mobility  M i  R SCH  Service  M i     SCH
P term  M i ic  k = ----------------------------------------------  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 UL
 Q  Service  M  Mobility  M  R
UL
 Service  M   2   
req i i SCH i SCH

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 rate:
CE –U L CE – U L UL UL
While N  M i   N max  S j  and R SCH  Service  M i    R FCH  Service  M i    2
UL
UL R SCH  Service  M i  
R SCH  Service  M i    ------------------------------------------------------
2
CE – U L
CE – U L N SCH  M i  k
N SCH  M i  k = ---------------------------------
-
2
SCH – r eq SCH – UL UL
SCH – r eq P term  M i ic  k Q req  Service  M i  Mobility  M i  R SCH  Service  M i   
P term  M i ic  k = ----------------------------------------------  -------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 Q
SCH – UL
 Service  M  Mobility  M  R
UL
 Service  M   2  
req i i SCH i

CE – U L CE – U L CE –U L
N  M i  k = N SCH  M i  k + N FCH  Mi k
EndWhile
Endif
If Mi uses an SCH on the downlink
For each cell (Sj,ic) in Mi FCH active set
Calculation of quality level on (Sj,ic) FCH at Mi, with the minimum power allowed on FCH for the Mi service
min
FCH – D L P FCH  Service  M i  
Pb  M i S j ic  = -----------------------------------------------------
L T  M i S j 
FCH – D L
DL  BTS  P b  M i S j 
-  G FCH
Q FCH  M i S j ic  k = ----------------------------------------------------------------------------------------------------------------------- p
– DL
 Service  M i  
DL DL
N tot  ic  –  1 – F ortho    BTS  P b  M i S j ic 

FCH –D L
Q FCH  M i S j ic  k = -------------------------------------------------------------
-
DL
N tot  ic 
If cell (Sj,ic) in Mi SCH active set

Calculation of quality level on (Sj,ic) SCH at Mi, with the minimum power allowed on SCH for the Mi service
min
SCH – D L P SCH  Service  M i  
Pb  M i S j ic  = -----------------------------------------------------
-
L T  M i S j 
306 AT271_TRG_E6 © Forsk 2009

SCH – D L
-  G SCH
Q SCH  M i S j ic  k = ----------------------------------------------------------------------------------------------------------------------- p
– DL
 Service  M i  
DL DL
N tot  ic  –  1 – F ortho    BTS  P b  M i S j ic 

SCH –D L
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.

DL DL DL
Q FCH  M i  = f rake efficiency  Q FCH  M i S j ic  k
k
S f  ActiveSet  FCH 

DL DL DL
Q SCH  M i  = f rake efficiency  Q SCH  M i S j ic  k
k
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,ic) and Mi:
DL DL
req  Q req  Service  M i  Mobility  M i  R FCH  Service  M i     FCH
-  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,ic) in Mi SCH active set
Calculation of the required power for DL traffic channel between (Sj,ic) and Mi:
DL DL
req  Q req  Service  M i  Mobility  M i  R SCH  Service  M i     SCH
-  P min
P SCH  M i S j ic  k = ----------------------------------------------------------------------------------------------------------------------------------------------------------------- SCH  Service  M i  
DL
Q SCH  M i 
k
Downgrading the service SCH rate (only for (Sj,ic) best server cell of Mi):
req max DL
While P SCH  M i S j ic  k  P SCH  Service  M i  R SCH  Service  M i   
req DL DL
Or P tx  S j ic  k + P tch  M i S j ic  k  P max  S j ic  and R SCH  Service  M i    R FCH  Service  M i    2
DL
DL R SCH  Service  M i  
R SCH  Service  M i   = -----------------------------------------------------
-
2
req DL DL
req P SCH  M i S j ic  k  Q req  Service  M i  Mobility  M i  R SCH  Service  M i     SCH
P SCH  M i S j ic  k = --------------------------------------------  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 DL
DL
 Service  M   2   
req i i SCH i SCH
req req 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
CE – D L CE –D L DL DL
While N  M i   N max  S j  and R SCH  Service  M i    R FCH  Service  M i    2
DL
R SCH  Service  M i   = -----------------------------------------------------
-
2
CE –D L
CE –D L N SCH  M i  k
N SCH  M i  k = ---------------------------------
-
2
req DL DL
P SCH  M i S j ic  k = --------------------------------------------  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 DL
DL
 Service  M   2   
req i i SCH i SCH
© Forsk 2009 AT271_TRG_E6 307

req req req

CE – D L CE – D L CE –D L
N  M i  k = N SCH  M i  k + N FCH  Mi k
EndWhile
CE –D L CE – D L
If N  M i   N max  S j  then Mi will not use SCH
Endif
Codes Codes DL DL
While N  M i   N max  S j ic  and R SCH  Service  M i    R FCH  Service  M i    2
DL
R SCH  Service  M i   = -----------------------------------------------------
-
2
Codes
Codes N SCH  M i  k
N SCH  M i  k = --------------------------------
2
req DL DL
P SCH  M i S j ic  k = --------------------------------------------  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 DL
DL
 Service  M   2   
req i i SCH i SCH
req req req

Codes Codes Codes

N  M i  k = N FCH  M i  k + N SCH  M i  k
EndWhile
Codes Codes
If N  M i   N max  S j  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.

DL DL DL
Q FCH  M i  = f rake efficiency  Q FCH  M i S f ic  k
k
S f  ActiveSet  FCH 

DL DL DL
Q SCH  M i  = f rake efficiency  Q SCH  M i S f ic  k
k
S f  ActiveSet  SCH 
DL DL
While Q k  M i   Q req  Service  M i  Mobility  Mi   and Mi FCH active set is not empty
DL DL
And Q k  M i   Q req  Service  M i  Mobility  Mi   (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)
UL
EndFor
For each mobile Mi
DL
EndFor
Control of Radio Resource Limits (Walsh Codes, Cell Power and Site Channel Elements)
For each cell (Sj,ic) on a site Nl
P tx  S j ic  k DL
While ----------------------------  %Power max
P max
req
Rejection of mobile with highest P tch  S j M b ic  k for the lowest service priority
308 AT271_TRG_E6 © Forsk 2009

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) containing rejected users
Control of the available power for the other cells (Si,ic) of the site where power pooling between transmitters is not acti-
vated

DL
If  %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) of the site can be allocated to cells (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
Codes Codes
While N  S j ic  k  N max  S j ic 
Rejection of last admitted mobile

EndFor
CE – DL CE – DL
While N  N I  k  N max  NI 
req
Rejection of mobile with highest P tch  M i S j  k for the lowest service priority
CE – UL CE – UL
req
Rejection of mobile with highest P term  M i ic  k for the lowest service priority
EndFor
UL UL
For each cell (Sj,ic) with X  S j ic   X max

EndFor
UL UL
While at least one cell with X  S j ic   X max exists

The convergence criteria are evaluated at each 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 
-  100  int  ma x ------------------------------------------------------------------
 DL = max  int  ma x -----------------------------------------------------  100 
  Stations P tx  ic  k   Stations N
DL
 ic  
user k
UL UL UL UL
 UL = max  int  ma x -----------------------------------------------------
-  100  int  ma x ------------------------------------------------------------------  100 
  Stations UL
I  ic    Stations
N
UL
 ic  
tot k user k
© Forsk 2009 AT271_TRG_E6 309

1st case: Between two successive iterations,  UL and  DL are lower (  ) than their respective thresholds (defined
when creating a simulation).
has been achieved.
2nd case: After 30 iterations,  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.

1. After the 30th 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. Convergence has not been achieved.
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.
If  UL and/or  DL are still strictly higher than their respective thresholds, the simulation has not converged (specific diver-
gence symbol).
If  UL and  DL are lower than their respective thresholds, the simulation has converged.
7.4.2.3 CDMA2000 1xEV-DO Power/Data Rate Control Simulation Algorithm

In a CDMA2000 1xEV-DO system, the UL supports power control but the DL does not. When a DL connection is estab-
lished, the transmitter transmits at full power (Pmax). Instead of power control, there is a data rate control based on the C/
I ratio computed at the mobile. For each distribution of users, Atoll simulates the power control mechanism for the UL and
the data rate control for the DL.
The simulation uses an iterative algorithm, where in each iteration, 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 process is
repeated from iteration to iteration until convergence is achieved. The algorithm steps are detailed below.
Figure 7.3: CDMA2000 1xEVDO Power Control Algorithm

Uplink received powers on carrier ic, I tot  ic  , I tot  ic  and I inter – carrier  ic  , at base station Sj are initialised to 0 W
(no connected mobile).
310 AT271_TRG_E6 © Forsk 2009

UL
UL I tot  S j ic 
 X k  S j ic  = ----------------------------
- = 0
UL

E c UL
The algorithm is detailed for any iteration k. Xk is the value of the variable X at the iteration k. In the algorithm,  ------ is
 N t  min
the minimum pilot quality level on uplink.
In case of 1xEV-DO Rev. 0 users, this threshold depends on the user mobility and is defined in the Mobility parameters
table.
Ec UL
For 1xEV-DO Rev. A users, the value of  ------ depends on the user requested data rate. This data rate can be obtained
 N t  min
by using a certain uplink 1xEV-DO Rev. A radio bearer ( Index UL – RevA – Bearer ) in a certain number of subframes ( n SF ).
E
UL
-----c- is the value defined in the 1xEV-DO Rev. A Radio Bearer Selection (Uplink) table for the combination (radio
 N t  min
bearer Index, mobility and number of subframe) providing the user requested data rate. Two values are available for this
parameter, one when the service uplink mode is "Low Latency" and another one for high capacity services.
All variables are described in Definitions and formulas part (see "Definitions and Formulas" on page 276).
For each mobile Mi
Determination of Mi’s Best Server (SBS(Mi))
For each station Sj containing Mi in its calculation area,
Determination of BestCarrier k  S j M i  .
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.
Else the carrier selection mode defined for Sj is considered.

For each carrier ic used by Sj, we calculate current loading factor:
UL
UL I tot  S j ic  UL
X k  S j ic  = ----------------------------
- + X
UL
EndFor
UL
BestCarrier k  S j M i  is the carrier with the lowest X k  S j ic 
BestCarrier k  S j M i  is the carrier with the lowest P tx  S j ic  k
BestCarrier k  S j M i  is randomly selected

UL UL
BestCarrier k  S j M i  is the first carrier so that X k  S j ic   X max
DL
 BTS    P tot  M i S j BestCarrier b pilot 
Calculation of Q pilot  M i S j BestCarrier  = ----------------------------------------------------------------------------------------------------------------
-
k DL
I 0  BestCarrier k  S j M i  b pilot 
req
If Q pilot  M i S j BestCarrier   Q pilot 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).
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
© Forsk 2009 AT271_TRG_E6 311

S BS  M i  = S j
Endif
EndFor
If no SBS has been selected, Mi cannot get a connection to the network
In the following lines, we will consider ic = BestCarrier k  S BS  M i  M i 
For each station Sj containing Mi in its calculation area, using ic, and if neighbours are used, neighbour of SBS(Mi)
DL
 BTS    P tot  M i S j ic b pilot 
Calculation of Q pilot  M i S j ic  = ----------------------------------------------------------------------------------
-
k DL
I 0  ic b pilot 
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

k
EndFor
req
Calculation of the required power for Mi, P term  M i ic  k
For each cell (Sj,ic) present in the Mi active set
Calculation of quality level on Mi traffic channel at (Sj,ic), with the minimum power allowed on traffic channel for the Mi
service
req
UL P term  M i ic  k – 1
P b  M i S j ic  = ------------------------------------------
-
L T  M i S j 
UL
UL  term  P b  M i S j ic  UL
Q  M i S j ic  k = -----------------------------------------------------------------------------------------------------------------------
UL Tx UL
 G p  Service 
N tot  ic  –  1 – F MUD   term  P b  M i S j ic 

UL
UL  term  P b  M i S j ic  UL
Q  M i S j ic  k = ---------------------------------------------------------
UL
 G p  Service 
N tot  ic 
End For
If (Mi is not in handoff)
UL UL
Q total  M i  = Q  M i S j ic 
k

UL UL UL
Q total  M i  = f rake efficiency  Q  M i S j ic  k
k
S j  ActiveSet
Else if (Mi is in soft or softer/soft without MRC)
UL UL UL
Q total  M i  = Max  Q  M i S j ic  k    G macro – diversity  2 links
k
I AS  ActiveSet
Else if (Mi is in soft/soft)
UL UL UL
Q total  M i  = Max  Q  M i S j ic  k    G macro – diversity  3 links
k
I AS  ActiveSet
Else if (Mi is in softer/soft with MRC)
 
 
= Max  f rake efficiency      G UL
UL UL UL UL
Q total  M i  Q  M i S j ic  k Q  M i S j ic  k macro – diversity  2 links
k  
 i AS  ActiveSet othersite 
 
(same site)
312 AT271_TRG_E6 © Forsk 2009

EndIf
UL
req Q req  Service  M i  Mobility  M i   req
P term  M i ic  k = -----------------------------------------------------------------------------------------
UL
 P term  M i ic  k – 1
Q total  M i 
k
If the service of Mi uses Transmission Control Protocol (TCP)

For the best server cell (Sk,ic) of Mi
Calculation of the Mi downlink application throughput
DL
Calculation of N tot  ic b traffic 

DL
 Ptot  Sj i c btraffic  + txj--------------------------------------------------------------------

DL DL  j term
N tot  ic b traffic  = - + N0
j j  k
DL
Calculation of the maximum data rate supplied to Mi, R max  M i S k 
Calculation of pilot quality level at Mi
DL
Ec P tot  M i S k ic b pilot 
------  M i S k ic b pilot  = ------------------------------------------------------
-
Nt DL
N tot  ic b pilot 
If Mi is an 1xEV-DO Rev. 0 service user, determination of the maximum data rate from the graph (Max rate=f(C/I)) specified
for the mobility type of Mi
Ec
R max  M i S k  = f  ------  M i S k ic b pilot 
DL
 Nt 
If Mi is an 1xEV-DO Rev. A service user, selection of the downlink 1xEV-DO Rev. A radio bearer ( Index DL –R evA – Bearer ):
E E DL
Index DL –R evA – Bearer where -----c-  M i S k ic b pilot    -----c-  Index DL –R evA – Bearer 
Nt  Nt  min
DL
DL R RLC – Peak  Index DL –R evA – Bearer 
Determination of the maximum data rate: R max  M i S k  = --------------------------------------------------------------------------------------------
n TS  Index DL –R evA – Bearer 
DL DL
R application  M i S k  = R max  M i S k   SF rate  Service  Mi   – R  Service  Mi  
UL
Determination of the uplink data rate due to TCP acknowledgements, R 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
R TCP – ACK  M i S k  = f  R application  M i S k  
UL UL UL
Determination of the nearest lower and higher supported rates ( R low and R high ) for R TCP – ACK  M i S k 
UL UL UL UL UL UL
For R low and R high , calculation of CI req  R low  and CI req  R high 
Ec UL
CI req =  ------
UL
  1 + G DRC + G TCH  for DO Rev.0 terminals
 N t  min
And
E c UL
CI req =  ------
UL
  1 + G DRC + G TCH + G RRI + G Auxiliary – pilot  for DO Rev.A terminals
 N t  min
EndFor
UL UL UL UL UL UL
Linear interpolation of CI req  R TCP – ACK  between CI req  R low  and CI req  R high 
UL UL UL UL UL
CI req = CI req  R  + CI req  R TCP – ACK 
UL UL W
Q req = CI req  -------------------------------------------------
UL UL
-
 R + R TCP – ACK 
EndIf
req req min
P term  M i ic  k = Max  P term  M i ic  k P term  M i S j  
req max
If P term  M i ic  k  P term  M i  then:
Downgrading the traffic data channel data rate:
© Forsk 2009 AT271_TRG_E6 313

req max
While P term  M i ic  k  P term  M i 
UL UL
And R  Service  M i    9.6kbps for DO-Rev.0 users or R  Service  M i    4.8kbps for DO-Rev.A users
req
P term  M i ic  k UL
req
-  R UL  Service  M i   ( R low  Service  M i   is the nearest lower supported data
P term  M i ic  k = --------------------------------------------------
UL
R low  Service  M i  
rate)
UL UL
R  Service  M i   = R low  Service  M i  
EndWhile
req max
If P term  M i ic  k  P term  M i  then Mi is rejected
Endif
Endif
Uplink Interference Updates
Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones)
UL
EndFor
Control of Radio Resource Limits (Number of EVDO users, MAC Indices and Site Channel Elements)
EVDO EVDO
While n  S j ic   n max  S j ic 
Rejection of the last admitted mobile

EndFor
MacIndexes MacIndexes
While N  S j ic   N max  S j ic 

EndFor
EVDO – CE EVDO – CE

EndFor
UL UL UL
For each cell (Sj,ic) with NR  S j ic   NR threshold  S j ic  + NR threshold  S j ic 
UL UL UL
While NR  S j ic   NR threshold  S j ic  + NR threshold  S j ic  and there is at least one mobile that can be downgraded
Downgrading the traffic data channel rate for all 1xEV-DO Rev. 0 mobiles for which the data rate transition flag is set to
"True".
UL
Endwhile
UL UL UL
For each cell (Sj,ic) with NR  S j ic   NR threshold  S j ic  – NR threshold  S j ic 
UL UL UL
While NR  S j ic   NR threshold  S j ic  – NR threshold  S j ic  and there is at least one mobile that can be upgraded
Upgrading the traffic data channel rate for all 1xEV-DO Rev. 0 mobiles for which the data rate transition flag is set to "True".
UL
Endwhile
UL UL UL
For each cell (Sj,ic) with NR  S j ic   NR threshold  S j ic  + NR threshold  S j ic 
314 AT271_TRG_E6 © Forsk 2009

UL UL UL
While NR  S j ic   NR threshold  S j ic  + NR threshold  S j ic  and there is at least one mobile to be downgraded
Downgrading the traffic data channel rate for the 1xEV-DO Rev. A mobile with the highest data rate
UL
Endwhile
UL UL UL
For each cell (Sj,ic) with NR  S j ic   NR threshold  S j ic  – NR threshold  S j ic 
UL UL UL
While NR  S j ic   NR threshold  S j ic  – NR threshold  S j ic  and there is at least one mobile to be upgraded
Upgrading the traffic data channel rate for the 1xEV-DO Rev. A mobile with the lowest data rate (only 1xEV-DO Rev. A
mobiles which have not been downgraded can be upgraded. In addition, the upgraded data rate cannot exceed the initial
user data rate drawn by the Monte-Carlo algorithm. This means that only mobiles downgraded during the uplink power
control step can be upgraded)
UL
Endwhile
UL UL
For each cell (Sj,ic) with X  S j ic   X max

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,ic)
DL
Calculation of N tot  ic b traffic 
For each cell (Sj,ic) ( k  j )
Determination of the number of mobiles connected to the cell (Sj,ic), N mobiles  S j ic 
If N mobiles  S j ic  = 0 then, 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
 Ptot  Sj ic btraffic  + N0

DL DL term
N tot  ic b traffic  =
j j  k
EndFor
DL
Calculation of the maximum data rate supplied to Mi, R max
For the best server cell (Sk,ic) of Mi
Calculation of pilot quality level at Mi
DL
Ec P tot  M i S k ic b pilot 
------  M i S k ic b pilot  = ------------------------------------------------------
-
Nt DL
N tot  ic b pilot 
If Mi is an 1xEV-DO Rev. 0 service user, determination of the maximum data rate from the graph (Max rate=f(C/I)) specified
for the mobility type of Mi
Ec
R max  M i S k  = f  ------  M i S k ic b pilot 
DL
 Nt 
If Mi is an 1xEV-DO Rev. A service user, selection of the downlink 1xEV-DO Rev. A radio bearer ( Index DL –R evA – Bearer ):
Ec Ec DL
Index DL –R evA – Bearer for which ------  M i S k ic b pilot    ------  Index DL –R evA – Bearer 
Nt  Nt  min
DL
DL R RLC – Peak  Index DL –R evA – Bearer 
Determination of the maximum data rate: R max  M i S k  = --------------------------------------------------------------------------------------------
n TS
EndFor
DL
Calculation of the average cell data rate, R av
© Forsk 2009 AT271_TRG_E6 315

DL DL
If N mobiles  S j ic  = 1 , then R av  S j ic  = R max  M i S j ic 
Else if N mobiles  S j ic   1 , determination of the multi-user gain G MU from the graph (MUG table=f(nb users)) specified
for (Sj,ic)

DL
R max  M i S j ic 
DL N mobiles
R av  S j ic  = M N  S  ic 
G MU  N mobiles  S j ic    ---------------------------------------------------------------------
i mobiles j -   1 –  ER DRC  S j ic   
N mobiles  S j ic 
  1 – TS BCMCS  S j ic  – TS EVDO – CCH  S j ic   + R BCMCS  S j ic   TS BCMCS  S j ic 
EndIf
EndFor

The algorithm convergence is studied on uplink only. The uplink convergence criterion is evaluated at each iteration, and
can be written as follow:
UL UL UL UL
 UL = max  int  ma x -----------------------------------------------------
-  100  int  ma x ------------------------------------------------------------------  100 
  Stations UL
I  ic    Stations
N
UL
 ic  
tot k user k
1st case: Between two successive iterations,  UL is lower (  ) than the threshold (defined when creating a simulation).
Example: Let us assume that the maximum number of iterations is 100, UL convergence threshold is set to 5. If  UL  5
between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th iteration. Convergence has been achieved.
2nd case: After 30 iterations,  UL is still higher than the threshold and from the 30th iteration,  UL does not decrease
during the next 15 successive iterations.
Examples: Let us assume that the maximum number of iterations is 100, UL convergence threshold is set to 5.
1. After the 30th iteration,  UL equals 100 and do not decrease during the next 15 successive iterations: Atoll stops the
algorithm at the 46th iteration. Convergence has not been achieved.
2. After the 30th iteration,  UL equals 80, it starts decreasing slowly until the 40th iteration (without going under the thresh-
old) and then does not change during the next 15 successive iterations: Atoll stops the algorithm at the 56th iteration with-
out achieving convergence.
If  UL is still strictly higher than the threshold, the simulation has not converged (specific divergence symbol).
If  UL is lower than the threshold, the simulation has converged.
7.4.3 Appendices
7.4.3.1 Admission Control
During admission control, Atoll calculates the uplink load factor of a considered cell assuming the mobile concerned is
connected with it. Here, activity status assigned to users is not taken into account. So even if the mobile is not active on
UL, it can be rejected due to cell load saturation. To calculate the cell UL load factor, either Atoll takes into account the
mobile power determined during power control if mobile was connected in previous iteration, or it estimates a load rise due
UL
to the mobile and adds it to the current load. The load rise ( X ) is calculated as follows:
UL 1
X = ---------------------------------------
W
1 + -----------------------------
UL UL
Q req  R
In case of IS-95 cdmaOne and CDMA2000 1xRTT networks, we have:

UL UL UL UL UL UL
Q req =  Q req  FCH +  Q req  SCH and R = R FCH + R SCH
316 AT271_TRG_E6 © Forsk 2009

7.4.3.2 Resources Management

7.4.3.2.1 Walsh Code Management
Walsh codes are managed on the downlink during a simulation in case of IS-95 cdmaOne and CDMA2000 1xRTT
networks. Atoll performs Walsh code allocation during the radio resource control step.
Walsh codes form a binary tree with codes of a longer length generated from codes of a shorter length. Length-k Walsh
codes are generated from length-k/2 Walsh codes. Therefore, if a channel needs 1 length-k/2 Walsh code, it is equivalent
to using 2 length-k Walsh codes, or 4 length-2k Walsh codes and so on.
Figure 7.4: Walsh Code Tree Indices (Not Walsh Code Numbers)
128 128-bit-length Walsh codes per cell are available in CDMA2000 and IS-95 cdmaOne documents.
During the resource control, Atoll determines the number of 128-bit-length Walsh codes that will be consumed by each
cell. Therefore, it allocates :
• A code with the longest length (i.e. a 128 bit-length code) per common channel for each cell,
• Two 128 bit-length codes per cell-receiver link for FCH in RC1, RC2, RC3 or RC5 and only one for FCH in RC4.
• The number of 128 bit-length codes to be allocated per cell-receiver link for SCH (in case SCH is supported by the
Walsh codes
user radio configuration), N 128 bits , is determined as follows:
Walsh codes DL
N 128 bits = Frate SCH  2 for RC1, RC2, RC3 and RC5,
And
Walsh codes DL
N 128 bits = Frate SCH for RC4.
Where
DL
Frate SCH is the SCH rate factor.
The Walsh code allocation follows the “Buddy” algorithm, which guarantees that:
• If a k-length Walsh code is used, all of its children with lengths 2k, 4k, …, cannot be used as they are not orthog-
onal.
• If a k-length Walsh code is used, all of its ancestors with lengths k/2, k/4, …, cannot be used as they are not orthog-
onal.
Notes:
• The Walsh code allocation follows the mobile connection order (mobile order in the Mobiles
tab).
• The Walsh code and channel element management is dealt with differently in case of
“softer” handoff. Atoll allocates Walsh codes for each transmitter-receiver link while it
assigns channel elements globally to a site.
7.4.3.2.2 Channel Element Management

Channel elements are controlled in the simulation.
IS95 cdmaOne and CDMA2000 1xRTT networks
Atoll checks the availability of this resource on uplink and downlink.

CE – UL
Overhead – CE – UL
• N channel elements for control channels (Pilot channel),
FCH – CE – UL UL
• N   1 + Frate SCH  per cell-receiver link, for TCH (TCH correspond to Traffic channels i.e. FCH and
SCH).
© Forsk 2009 AT271_TRG_E6 317

CE – UL
N
CE – UL CE – UL
N  NI  = j
j  NI
CE – DL
On downlink, Atoll consumes N  j  channel elements for each cell j on a site NI. This figure includes:
Overhead – CE – DL
• N channel elements for control channels (Pilot channel, Synchronisation channel, Paging
channel),
FCH – CE – DL DL
• N   1 + Frate SCH  per cell-receiver link, for TCH (TCH correspond to Traffic channels i.e. FCH and
SCH).
CE – DL
Therefore, the number of channel elements required on downlink at the site level, N  N I  , is:
N
CE – DL CE – DL
N  NI  = j
j  NI
Note:
• In case of “softer” handover (the mobile has several links with co-site cells), Atoll allocates
channel elements for the best serving cell-mobile link only.
CDMA2000 1xEV-DO networks
CE – UL
• 2 channel elements for control channels (Pilot channel, Data Rate Control channel, etc ). This value is fixed and
hard-coded.
TCH – CE – UL
• N per cell-receiver link, for TCH (Traffic channels).
CE – UL
N
CE – UL CE – UL
N  NI  = j
j  NI
On downlink, only one user can be served by a cell at a time, so this resource is not limited.
7.4.3.3 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).
7.4.3.3.1 Downlink Load Factor per Cell

The downlink load factor is calculated for each CDMA2000 1xRTT and IS-95 cdmaOne 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”.
DL – FCH DL – SCH
Q req Q req
Let CI req = ------------------------ - be the required quality.
- + ------------------------
DL – FCH DL – SCH
Gp Gp
FCH SCH
So, we have CI req = CI req + CI req
In case of soft handoff, required quality is limited to the effective contribution of the transmitter.
 Ptch  ic 
DL ortho
P tx  ic  = P pilot  ic  + P sync  ic  + P paging  ic  + P SCH  ic  + P FCH  ic  = P CCH  ic  +
tch
where
ortho
P CCH  ic  = P pilot  ic  + P sync  ic  + P paging  ic 
 Ptch  ic  = PSCH  ic  + PFCH  ic 

tch
At mobile level, we have a required power, Ptch:
term
P tch  ic  = CI req   I extra  ic  + I intra  ic  + I inter – carrier  ic  + N 0   LT
318 AT271_TRG_E6 © Forsk 2009

DL
  P tx  ic  – P tch  ic  term
P tch  ic  = CI req   I extra  ic  + I inter – carrier  ic  +  1 – F ortho   BTS    ----------------------------------------------- + N 0   L T
  LT  
DL term
 I extra  ic  + I inter – carrier  ic    L T +  1 – F ortho   BTS   P tx  ic  + N 0  LT
P tch  ic  = -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
1
------------- +  1 – F ortho   BTS 
CI req
where
DL
I intra  ic  is the total power received at receiver from the cell to which it is connected.
DL
I extra  ic  is the total power received at receiver from other cells.
I inter – carrier  ic  is the inter-carrier interference received at receiver.
DL term
 I extra  ic  + I inter – carrier  ic    L T +  1 – F ortho   BTS   P tx  ic  + N 0  LT

DL ortho
P tx  ic  = P CCH  ic  + -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
1
tch ------------- +  1 – F ortho   BTS 
CI req
We have:
 I extra  ic  + I inter – carrier  ic    L T DL term

---------------------------------------------------------------------------------------- +  1 – F ortho   BTS   P tx  ic  + N 0  LT
DL
P tx  ic 

DL ortho
P tx  ic  = P CCH  ic  + ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1
tch ------------- +  1 – F ortho   BTS 
CI req
  I extra  ic  + I inter – carrier  ic    L T 

  ---------------------------------------------------------------------------------------
DL
P  ic 
- + 1 – F ortho   BTS

DL tch tx DL
P tx  ic  – ---------------------------------------------------------------------------------------------------------------------------------------------------------  P tx  ic 
1
------------- +  1 – F ortho   BTS 
CI req
 term 
 ortho N0  LT 
=  P CCH  ic  +
  -
--------------------------------------------------------------------
1 
 tch ------------- +  1 – F ortho   BTS 
CI req
term
N0  LT

ortho
P CCH  ic  + --------------------------------------------------------------------
-
1
tch ------------- +  1 – F ortho   BTS 
DL CI req
P tx  ic  = -------------------------------------------------------------------------------------------------------------
-
I extra  ic   L T
------------------------------------ + 1 – F ortho   BTS
DL
P tx  ic 
1– 
--------------------------------------------------------------------------------------
1 + 1 – F
tch ------------- ortho   BTS 
CI req
Therefore, the downlink load factor can be expressed as:

---------------------------------------------------------------------------------------- + 1 – F ortho   BTS
DL
P tx  ic 

DL
X = ------------------------------------------------------------------------------------------------------------------------------------------
-
1
tch ------------- +  1 – F ortho   BTS 
CI req
The downlink load factor represents the signal degradation in relative to the reference interference (thermal noise).
7.4.3.3.2 Downlink Load Factor per Mobile

Atoll evaluates the downlink load factor for any connected mobile (CDMA2000 1xRTT 1xEV-DO or IS-95 cdmaOne user)
as follows,
DL
DL I tot  ic 
X = --------------------
DL
N tot  ic 
7.5 IS-95 cdmaOne and CDMA2000 Prediction Studies

7.5.1 Point Analysis: The AS Analysis Tab
Let us imagine a probe receiver with associated terminal, mobility and service with certain UL and DL rates. This receiver
does not create any interference. The analysis is based on the uplink load percentage and the downlink total power of
© Forsk 2009 AT271_TRG_E6 319

cells. These parameters can be either outputs of a given simulation, average values calculated from a group of simulations,
or user-defined cell inputs.
7.5.1.1 Bar Graph and Pilot Sub-Menu

Atoll provides the same outputs in the bar graph and pilot sub-menu whichever the studied network, IS-95 cdmaOne,
CDMA2000 1xRTT or 1xEV-DO.
Atoll proceeds as in power control simulation. It determines the best carrier of each transmitter i (its best cell) that contains
the receiver in its calculation area. The best carrier selection depends on the option chosen in Equipment (UL minimum
noise, DL minimum power, random, sequential). Then, Atoll calculates pilot quality at the receiver from these transmitters
on their best carriers and defines the best server (on its best carrier).
1. 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.
For, IS-95 cdmaOne and CDMA2000 1xRTT users we have,
 BTS    P c  i ic  DL DL DL DL term
Q pilot  i ic  = -------------------------------------------------
- with I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + N 0
DL
I 0  ic 
For CDMA2000 1xEV-DO users, we have,

DL
 BTS    P tot  i ic b pilot 
Q pilot  i ic  = ---------------------------------------------------------------------
-
DL
I 0  ic b pilot 
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  + N 0
The calculation of Q pilot  i ic  can be divided into 6 steps explained in the table below.
IS-95 cdmaOne and CDMA2000 1xRTT users CDMA2000 1xEV-DO users

DL
P tot  i ic b pilot  calculation for each cell (i,ic)
P c  i ic  calculation for each cell (i,ic) DL
P tot  i ic b pilot  is the pilot burst from the transmitter i on the carrier
P c  i ic  is the pilot power from a transmitter i on the carrier ic at the ic at the receiver.
receiver.
1st step
DL P tx  i ic b pilot 
P pilot  i ic  P tot  i ic b pilot  = -------------------------------------
-
P c  i ic  = ---------------------------
- LT
I
LT
I and
P tx  i ic b pilot  = P max  i ic 

L T is the total loss between the transmitter i and the receiver: ---------------------------------------------------------------------------------------------------------------------------------------------------
I G Tx  G term
DL DL
I extra  ic  and I inter – carrier  ic  calculation
We have,
 Ptot  j ic 
DL DL
I extra  ic  = DL DL
I extra  ic b pilot  and I inter – carrier  ic b pilot  calculation
j j  i
DL
For each transmitter of the network, P tot  j ic  is the total power We have,
received at the receiver from the transmitter j on the best carrier ic of
2nd step
the transmitter i.
 Ptot  j ic bpilot 
DL DL
I extra  ic b pilot  =
DL P tx  j ic 
P tot  j ic  = ---------------------
- j j  i
LT and
P tx  j ic  is the total power transmitted by the transmitter j on the best
 Ptot  j icadj bpilot 
DL
carrier of the transmitter i. DL
I inter – carrier  ic j j
b pilot  = -------------------------------------------------------
-
Finally, we have,

DL
P tot  j ic adj 
DL j j
I inter – carrier  ic  = ---------------------------------------
term
3rd step
N0 calculation
term
N0 = F term  K  T  W
320 AT271_TRG_E6 © Forsk 2009

4th step
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, G macro – diversity , models the decrease in shadowing margin due to the fact there are several pilot signals at the
mobile.
5th step
DL npaths
G macro – diversity = 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).
Note: This parameter is determined from the fixed cell edge coverage probability and the model standard deviation. When the model standard
deviation is set to 0, the macro-diversity gain equals 0.
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  ic  .
Resulting DL
Q pilot  ic  = G macro – diversity  max  Q pilot  i ic  
Resulting req Resulting
Q pilot  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). 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 best
server
6th step
BS will be the one used by other transmitters of active set (when active set size is greater than 1). Pilot is available.
Resulting req
If Q pilot  ic   Q pilot , no cell (i,ic) can enter the active set. Pilot is unavailable.
Then, pilot qualities at the receiver from transmitters i (other than the best server) on the best carrier of the best server, icBS, are recalculated
to determine the entire receiver active set (when active set is greater than 1). Same formulas and calculation method are used to update
DL
I 0  ic BS  and determine Q pilot  i ic BS  .
Other cells (i,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)
2. Number of cells in active set

This is a user-defined input in the terminal properties. It corresponds to the active set size.
3. Number of fingers
The number of fingers, f, of the rake receiver. This parameter is defined in the terminal properties. It is relevant in
CDMA2000 1xRTT only4. This is the maximum number of active set links that the terminal (rake) can combine.
4. Thermal noise
This parameter is calculated as described above (3rd step).

5. Io (Best server)
Io (Best server) is the total noise received at the receiver on icBS.
6. Downlink macro-diversity gain
This parameter is calculated as described above (5th step).
7.5.1.2 Downlink Sub-Menu

Outputs calculated by Atoll depend on the studied network (IS-95 cdmaOne, CDMA2000 1xRTT or CDMA2000 1xEV-DO).
7.5.1.2.1 IS-95 cdmaOne

Let m denote the number of cells in the receiver active set and f be the number of fingers defined for the terminal. We
assume that f is less than or equal to m5.
Among the m cells of the receiver active set, only the first f cells will be considered in order to determine the traffic channel
availability on downlink. Each of these cells is noted (k,icBS).
Atoll calculates the traffic channel quality from each cell (k,icBS). No power control is performed as in simulations. Here,
Atoll determines the downlink traffic channel quality at the receiver for the maximum traffic channel power per transmitter
4. In IS-95 cdmaOne, the number of fingers is the same than the active set size and CDMA2000 1xEV-DO systems
do not support soft handover on downlink.
5. In IS-95 cdmaOne, the number of fingers is normally the same than the active set size.
© Forsk 2009 AT271_TRG_E6 321

allowed on the fundamental channel (FCH). Then, after combination, the total downlink traffic channel quality is evaluated
and compared with the specified target quality.
1. Eb/Nt target
The Eb/Nt target parameter is available in the Services table. This is the user-defined downlink traffic data quality target
DL
on FCH (  Q req  FCH ) for a given service and a terminal.
2. Required transmitter power on FCH

req
The calculation of the required transmitter power on FCH ( P FCH ) may be divided into three steps.
1st step: Eb/Nt max for the first f (number of fingers) cells of active set
DL
Let us assume the following notation: Eb/Nt max corresponds to Q max .
DL
-  G DL
Q max  k ic BS  = ------------------------------------------------------------- p
DL
N tot  ic BS 
max
DL P FCH DL DL DL DL term
With P b – max  k ic BS  = -------------
LT
k
Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table.
L T is the total loss between the transmitter i and the receiver.

k

DL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
DL
With
DL
I intra  ic BS  =  1 –  BTS  F ortho   P DL  k ic 
tot BS
DL DL
j j  k
DL
For each transmitter in the network, P tot  ic BS  is the total power received at the receiver from this transmitter on icBS.
DL
DL
DL txj j

DL
Q MAX is the traffic channel quality at the receiver on icBS after combining the signal from each cell (k,icBS).

DL DL
DL DL DL
k
Where
DL
f rake efficiency is the downlink rake efficiency factor defined in the Terminal properties.
req
3rd step: P FCH calculation
322 AT271_TRG_E6 © Forsk 2009

DL
req  Q req  FCH
-  P max
P FCH = ------------------------------- FCH
DL
Q MAX  ic BS 
3. Eb/Nt max for the first f (number of fingers) cells of active set
DL
Let us assume the following notation: Eb/Nt max corresponds to Q max .
DL
-  G DL
Q max  k ic BS  = ------------------------------------------------------------- p
DL
N tot  ic BS 
max
DL P FCH DL DL DL DL term
With P b – max  k ic BS  = -------------
LT
k
Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table.

k

DL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
DL
With
max req
DL
I intra  ic BS  =  1 –  BTS  F ortho   P DL  k ic  –  1 –  P FCH – P FCH
tot BS BTS   max ( ----------------------------------,0)
LT
k
And
DL DL
j j  k
DL
For each transmitter in the network, P tot  ic BS  is the total power received at the receiver from the transmitter on icBS.
DL
DL
DL txj j
4. Eb/Nt max
DL
Q MAX is the traffic channel quality at the receiver on icBS after combining the signal from each cell (k,icBS).

DL DL
DL DL DL
k
Where
DL
f rake efficiency is the downlink rake efficiency factor defined in the Terminal properties.
DL DL
Therefore, the service on the downlink traffic channel is available if Q MAX  ic BS    Q req  FCH .
5. Effective Eb/Nt
DL
Q eff is the downlink effective traffic channel quality at the receiver on icBS.
DL DL DL
Q eff = min  Q MAX  Q req  FCH 
© Forsk 2009 AT271_TRG_E6 323

6. Downlink soft handover gain

DL
G SHO corresponds to the DL soft handover gain.
DL
DL Q MAX  ic BS 
G SHO = --------------------------------------------------------
DL
k
DL DL
k
7.5.1.2.2 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. We assume that f is less
than or equal to mFCH and mSCH.
Among the mFCH cells of the receiver active set, only the first f cells will be considered in order to determine the FCH avail-
ability on downlink. In the same way, 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. Each of these cells is noted (k,icBS).
Atoll calculates the traffic channel quality on FCH from each cell (k,icBS). No power control is performed as in simulations.
Here, Atoll determines the downlink traffic channel quality on FCH at the receiver for the maximum traffic channel power
per transmitter allowed on FCH. Then, after combination, the total downlink traffic channel quality on FCH is evaluated and
compared with the specified target quality.
Atoll calculates the traffic channel quality on SCH from each cell (k,icBS). No power control is performed as in simulations.
Here, Atoll determines the downlink traffic channel quality on SCH at the receiver for the maximum traffic channel power
per transmitter allowed on SCH. This value depends on the downlink data rate specified in the analysis. Then, after combi-
nation, the total downlink traffic channel quality on SCH is evaluated and compared with the specified target quality.
1. 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). This value
is user-defined for a given service and terminal.
DL
Eb/Nt target on SCH (  Q req  SCH ) is the downlink traffic data quality target on the supplemental channel (SCH). This value
is specified for a given service, terminal and SCH rate.
2. 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.
1st step: Eb/Nt max for the first f (number of fingers) cells of active set
DL
Let us assume the following notations: Eb/Nt max on FCH and SCH respectively correspond to  Q max  FCH and
DL
 Q max  SCH .
DL – FCH
DL  BTS  P b –max  k ic BS 
-  G DL
 Q max  k ic BS   FCH = ----------------------------------------------------------------- p
– FCH
DL
N tot  ic BS 
And
DL – SCH
-  G DL
 Q max  k ic BS   SCH = ------------------------------------------------------------------ p
– SCH
DL
N tot  ic BS 
max max
DL – FCH P FCH DL – SCH P SCH
With P b  k ic BS  = -------------
- , P b –max  k ic BS  = -------------
-
LT LT
k k
DL DL DL DL term
And N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + N 0
Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table for a certain terminal.
max
P SCH is the maximum power allowed on SCH for the specified downlink data rate. This parameter is user-defined in the
Services table for a certain terminal and SCH rate.

k

DL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
324 AT271_TRG_E6 © Forsk 2009

DL
With
DL
I intra  ic BS  =  1 –  BTS  F ortho   P DL  k ic 
tot BS
And
DL DL
j j  k
DL
For each transmitter in the network, P tot  ic BS  is the total power received at the receiver from this transmitter on icBS.
DL
DL
DL txj j
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,icBS).

DL DL
 Q MAX  ic BS   FCH =  Q max  k ic BS   FCH
  Qmax  k icBS  FCH

DL DL DL
 Q MAX  ic BS   FCH = f rake efficiency 
k
Where
DL
DL
 Q MAX  SCH is the traffic channel quality on SCH at the receiver on icBS after combining the signal from each cell (k,icBS).

DL DL
 Q MAX  ic BS   SCH =  Q max  k ic BS   SCH
  Qmax  k icBS  SCH

DL DL DL
 Q MAX  ic BS   SCH = f rake efficiency 
k
req req
3rd step: P FCH and P SCH calculation
DL
-  P max
P FCH = ---------------------------------------------- FCH
DL
 Q MAX  ic BS   FCH
DL
req  Q req  SCH
-  P max
P SCH = ---------------------------------------------- SCH
DL
 Q MAX  ic BS   SCH
3. 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 .
DL – FCH
-  G DL
 Q max  k ic BS   FCH = ----------------------------------------------------------------- p
– FCH
DL
N tot  ic BS 
max
DL – FCH P FCH DL DL DL DL term
With P b –max  k ic BS  = -------------
LT
k
Where
© Forsk 2009 AT271_TRG_E6 325

max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table for a certain terminal.

k

DL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
DL
With
DL max req
I intra  ic BS  =  1 –  BTS  F ortho   P DL  k ic  –  1 –  P FCH – P FCH
tot BS BTS   max (----------------------------------,0)
LT
k
And
DL DL
j j  k
DL
DL
DL
DL txj j
4. 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 .
DL – SCH
-  G DL
 Q max  k ic BS   SCH = ------------------------------------------------------------------ p
– SCH
DL
N tot  ic BS 
max
DL – SCH P SCH DL DL DL DL term
With P b –max  k ic BS  = -------------
LT
k
Where
max
P SCH is the maximum power allowed on SCH for the specified downlink data rate. This parameter is user-defined in the
Services table for a certain terminal and SCH rate.

k

DL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
DL
With
max req
DL
I intra  ic BS  =  1 –  BTS  F ortho   P DL P SCH – P SCH
tot  k ic BS  –  1 –  BTS   max (----------------------------------,0)
LT
k
And
DL DL
j j  k
DL
DL
326 AT271_TRG_E6 © Forsk 2009

DL
DL txj j
5. 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,icBS).

DL DL
 Q MAX  ic BS   FCH =  Q max  k ic BS   FCH
  Qmax  k icBS  FCH

DL DL DL
k
Where
DL
DL
 Q MAX  SCH is the traffic channel quality on SCH at the receiver on icBS after combining the signal from each cell (k,icBS).

DL DL
 Q MAX  ic BS   SCH =  Q max  k ic BS   SCH
  Qmax  k icBS  SCH

DL DL DL
 Q MAX  ic BS   SCH = f rake efficiency 
k
DL DL
Therefore, the service on the downlink traffic channel is available if  Q MAX  ic BS   FCH   Q req  FCH and
DL DL
 Q MAX  ic BS   SCH   Q req  SCH .
6. 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.
DL DL DL
 Q eff  FCH = min   Q MAX  FCH  Q req  FCH 
And
DL DL DL
 Q eff  SCH = min   Q MAX  SCH  Q req  SCH 
7. 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.
DL
DL  Q MAX  ic BS   FCH
 G SHO  FCH = -----------------------------------------------------------------------
DL
max   Q max  k ic BS   FCH 
k
And
DL
DL  Q MAX  ic BS   SCH
 G SHO  SCH = -----------------------------------------------------------------------
-
DL
max   Q max  k ic BS   SCH 
k
DL DL
k
7.5.1.2.3 CDMA2000 1xEV-DO

Atoll calculates the effective pilot quality level at the receiver and compares this value with the required quality level.
1. Required rate
DL
The required rate, R req , is the downlink data rate selected for the analysis.
2. Required C/I
© Forsk 2009 AT271_TRG_E6 327

 C
For 1xEV-DO Rev. 0 users, the required C/I (  ---I- ) is determined from the graph “Max Rate=f(C/I)” defined for the mobil-
req
ity type selected in the analysis. It corresponds to the value read in the graph “Max Rate=f(C/I) (Rev0)” for the specified
DL
required rate, R req .
DL
For 1xEV-DO Rev. A users, the required data rate ( R req ) is obtained by using a certain downlink transmission format (i.e.
a 1xEV-DO Rev. A radio bearer ( Index DL –R evA – Bearer ) with a certain number of timeslots ( n TS )). It is calculated as
follows:
DL
DL R RLC – peak  Index DL – RevA – Bearer 
R req = ---------------------------------------------------------------------------------------------
-
n TS
C
---- is the value defined in the 1xEV-DO Rev. A Radio Bearer Selection (Downlink) table for this downlink transmission
 I  req
format (radio bearer Index, mobility and number of timeslots). It corresponds to the C/I required to obtain the defined
DL
required rate, R req .
3. Effective C/I
Ec
Let ------  ic BS b pilot  be the effective C/I at the receiver on icBS.
Nt
For the best cell (BS,icBS) of the receiver active set, we have:
pilot
Ec   Q resulting  ic BS 
------  ic BS b pilot  = --------------------------------------------------
-
Nt pilot
 – Q resulting  ic BS 
Where
pilot DL
Q resulting  ic BS  = G macro – diversity  Q pilot  ic BS 
BS
4. Effective data rate

DL
For 1xEV-DO Rev. 0 users, the effective data rate, R , is determined from the graph “Max Rate=f(C/I) (Rev0)” defined
DL
for the mobility type selected in the analysis. R is the value read in the graph “Max Rate=f(C/I) (Rev0)” for the calculated
effective C/I, E c .
------  ic BS b pilot 
Nt
DL
For 1xEV-DO Rev. A users, the effective data rate ( R ) provided on downlink depends on the downlink transmission
format, i.e the radio bearer index ( Index DL –R evA – Bearer ) with the number of timeslots ( n TS ). For the defined mobility
Ec C
type, Atoll selects the downlink transmission format where ------  ic BS b pilot    ---- .
Nt  I  req Then, it determines the downlink
effective data rate as follows:
DL
DL R RLC – peak  Index DL –R evA – Bearer 
R -.
= -------------------------------------------------------------------------------------------
n TS
DL DL
The traffic data channel on downlink is available if R  R req .
7.5.1.3 Uplink Sub-Menu

Outputs calculated by Atoll depend on the studied network (IS-95 cdmaOne, CDMA2000 1xRTT or CDMA2000 1xEV-DO).
7.5.1.3.1 IS-95 cdmaOne

For each cell (i,icBS) in the receiver active set, Atoll calculates the uplink traffic channel quality on FCH from the receiver.
No power control is performed as in simulations. Here, Atoll determines the uplink traffic channel quality at the cell for the
maximum terminal power allowed. Then, the total uplink traffic channel quality is evaluated with respect to the receiver
handover status. From this value, Atoll calculates the required terminal power and compares it to the maximum terminal
power allowed.
1. Max terminal power
max
The Max terminal power parameter ( P term ) is user-defined for each terminal. It corresponds to the maximum terminal
power allowed.
2. Required terminal power
req
The required terminal power ( P term ) calculation may be divided into three steps:
328 AT271_TRG_E6 © Forsk 2009

UL
1st step: Evaluation of the uplink traffic channel quality, Q max  i ic BS  , for each cell of active set
For each cell (i,icBS), we have:
UL
UL  term  P b –max  i ic BS 
-  G UL
Q max  i ic BS  = ---------------------------------------------------------- p
UL
N tot  i ic BS 
max
UL P term
With P b –max  i ic BS  = -------------
-
LT
i

i

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. This value is deduced from the cell
UL
uplink load factor X  i ic BS  .
tx
UL N0
N tot  i ic BS  = ---------------------------------------
-
UL
1 – X  i ic BS 
tx
UL
2nd step: Calculation of the total traffic channel quality at the transmitter on icBS ( Q MAX ) based on the receiver handover
status.
If there is no handover, we have:
UL UL
Q MAX  ic BS  = Q max  i ic BS 
For soft handover, we have:

UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  i ic BS  
i
UL
UL UL
max  Q max  i ic BS   corresponds to the highest Q max  i ic BS  value.
i
For soft-soft handover, we have:

UL UL UL
i
UL
For softer and softer-softer handovers, we have:
 Qmax  i icBS 
UL UL UL
i
For softer-soft handover, there are two possibilities. If the MRC option is selected (option available in Global parameters),
we have:
 UL 

UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency  Q max  i ic BS  Q max  i ic BS 
 i on the other site 
i on the same site
otherwise,
UL UL UL
i
req
3rd step: Calculation of P term
UL
-  P max
P term = ------------------------------- term
UL
Q MAX  ic BS 
Where
© Forsk 2009 AT271_TRG_E6 329

UL
 Q req  FCH is the user-defined uplink data traffic quality target on FCH for a given service and a terminal. This parameter
is available in the Services table.
req max
Therefore, the service on the uplink data traffic channel is available if P term  P term .
3. Eb/Nt max on FCH

UL
-  G UL
Q max  i ic BS  = ---------------------------------------------------------- p
UL
max
UL P term
With P b –max  i ic BS  = -------------
-
LT
i

i

UL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx max req
UL N0 P FCH – P FCH
- +  1 –  term   max (---------------------------------
N tot  i ic BS  = --------------------------------------- -,0)
UL
1 – X  i ic BS  LT
i
tx
UL
Q MAX  ic BS  is the traffic channel quality on FCH at the transmitter on icBS after signal combination of all the transmitters
of the active set.
If there is no handover, we have:
UL UL

UL UL UL
i
UL
UL UL
max  Q max  i ic BS   corresponds to the highest Q max  i icBS  value.
i

UL UL UL
i
UL
UL UL UL
i
we have:
 UL 

UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency  Q max  i ic BS  Q max  i ic BS 
i on the same site
otherwise,
UL UL UL
i
4. Effective Eb/Nt
330 AT271_TRG_E6 © Forsk 2009

UL
Q eff is the uplink effective traffic channel quality at the receiver on icBS.
UL UL UL
Q eff = min  Q MAX  Q req  FCH 
5. Uplink soft handover gain

UL
G SHO corresponds to the UL soft handover gain.
UL
G SHO = -----------------------------------------------------
UL
max  Q max  i ic BS  
i
UL UL
i
7.5.1.3.2 CDMA2000 1xRTT

For each cell (i,icBS) in the receiver active set, Atoll calculates the uplink traffic channel quality on FCH and SCH from the
receiver. No power control is performed as in simulations. Here, Atoll determines the uplink traffic channel quality on FCH
at the cell for the maximum terminal power allowed on FCH. In the same way, it evaluates the uplink traffic channel quality
on SCH at the cell for the maximum terminal power allowed on SCH. Then, total uplink traffic channel qualities on FCH
and SCH are evaluated with respect to the receiver handover status. From these values, Atoll deduces required terminal
powers on FCH and SCH, calculates the total terminal power required and compares this value with the maximum terminal
power allowed.
1. Max terminal power on FCH and SCH
max
power allowed. On uplink, the terminal power is shared between pilot, FCH and SCH channels. So, 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. This parameter is user-defined in the terminal proper-
ties.
And
max UL UL UL
 P term  FCH  Q req  FCH R FCH  AF FCH
-  ---------------------------------------
- = --------------------------
----------------------------
max UL UL
 P term  SCH  Q req  SCH R SCH
Therefore,
max
max  1 – p   P term
 P term  FCH = ---------------------------------------------------------------------------------
-
UL UL
 Q req  SCH  R SCH
1 + ------------------------------------------------------------------------
UL UL UL
 Q req  FCH  R FCH  AF FCH
And
max
max  1 – p   P term
 P term  SCH = ---------------------------------------------------------------------------------
-
UL UL UL
 Q req  FCH  R FCH  AF FCH
1 + ------------------------------------------------------------------------
UL UL
 Q req  SCH  R SCH
2. Required terminal power on FCH and SCH

req req
The required terminal powers on FCH and SCH, respectively  P term FCH and  P term  SCH , are calculated as follows:
UL UL
1st step: Evaluation of uplink traffic channel qualities on FCH and SCH,  Q max  ic BS   and  Q max  ic BS   , for
i FCH i SCH
each cell of active set.
UL – FCH
-  G UL
 Q max  i ic BS   FCH = ---------------------------------------------------------------- p
– FCH
UL
And
UL – SCH
UL  term  P b –max  i ic BS  UL – SCH
 Q max  i ic BS   SCH = -----------------------------------------------------------------
UL
 Gp
© Forsk 2009 AT271_TRG_E6 331

max max
UL – FCH  P term  FCH UL – SCH  P term  SCH
With P b –max  i ic BS  = ----------------------------
- and P b –max  i ic BS  = ----------------------------
-
LT LT
i i

i

UL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx
UL N0
N tot  i ic BS  = ---------------------------------------
-
UL
1 – X  i ic BS 
tx
UL
2nd step: Calculation of FCH and SCH total traffic channel qualities at the transmitter on icBS,  Q MAX  FCH and
UL
 Q max  SCH , based on the receiver handover status.
If there is no handoff, 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

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
UL UL
i

UL UL UL
i
And
UL UL UL
i
UL
 G macro – diversity  3 links is the uplink macro-diversity gain.This parameter is determined from the fixed cell edge coverage
  Qmax  i icBS  FCH

UL UL UL
i
  Qmax  i icBS  SCH

UL UL UL
And  Q MAX  ic BS   SCH = f rake efficiency 
i
we have:
 UL 

UL UL UL UL
 Q MAX  ic BS   FCH =  G macro – diversity  2 links  max  f rake efficiency   Q max  i ic BS   FCH  Q max  i ic BS   
FCH
i on the same site
And
 UL 

UL UL UL UL
 Q MAX  ic BS   SCH =  G macro – diversity  2 links  max  f rake efficiency   Q max  i ic BS   SCH  Q max  i ic BS   
SCH
i on the same site
otherwise,
332 AT271_TRG_E6 © Forsk 2009

UL UL UL
i
And
UL UL UL
i
req req
3rd step: Calculation of  P term  FCH and  P term  SCH
UL UL
req  Q req  FCH  Q req  SCH
-   P max
 P term  FCH = ----------------------------------------------
req
-   P max
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. This parameter
is available in the Services table.
UL
 Q req  SCH is the user-defined uplink data traffic quality target on SCH for a given service, terminal and SCH rate. This
parameter is available in the Services table.
req
Then, from the required terminal power on FCH and SCH, Atoll determines the total terminal power required ( P term ).
req req req req

P term =  P term  FCH +  P term  SCH +  P term  pilot
req req
As  P term pilot = p  P term , we have:
req req
req  P term  FCH +  P term  SCH
P term = ----------------------------------------------------------------
-
1–p
req max
Therefore, the service on the uplink data traffic channel is available if P term  P term .
3. Eb/Nt max on FCH for each cell in active set

UL – FCH
-  G UL
 Q max  i ic BS   FCH = ---------------------------------------------------------------- p
– FCH
UL
max
UL – FCH  P term FCH
With P b –max  i ic BS  = ----------------------------
-
LT
i

i

UL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx max req
UL N0 P FCH – P FCH
N tot  i ic BS  = ---------------------------------------
- +  1 –  term   max (---------------------------------
-,0)
UL
1 – X  i ic BS  LT
i
tx
4. Eb/Nt max on SCH for each cell in active set

UL – SCH
UL  term  P b –max  i ic BS  UL – SCH
 Q max  i ic BS   SCH = -----------------------------------------------------------------
UL
 Gp
max
UL – SCH  P term  SCH
With P b –max  i ic BS  = ----------------------------
-
LT
i

i

UL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
© Forsk 2009 AT271_TRG_E6 333

UL
UL
tx max req
UL N0 P SCH – P SCH
- +  1 –  term   max (----------------------------------
N tot  i ic BS  = --------------------------------------- ,0)
UL
1 – X  i ic BS  LT
i
tx
5. 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.

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

UL UL UL
i
And
UL UL UL
i
UL
UL UL
i

UL UL UL
i
And
UL UL UL
i
UL
  Qmax  i icBS  FCH

UL UL UL
i
  Qmax  i icBS  SCH

UL UL UL
And  Q MAX  ic BS   SCH = f rake efficiency 
i
we have:
 UL 

UL UL UL UL
 Q MAX  ic BS   FCH =  G macro – diversity  2 links  max  f rake efficiency   Q max  i ic BS   FCH  Q max  i ic BS   
FCH
i on the same site
And
 UL 

UL UL UL UL
 Q MAX  ic BS   SCH =  G macro – diversity  2 links  max  f rake efficiency   Q max  i ic BS   SCH  Q max  i ic BS   
SCH
i on the same site
otherwise,
UL UL UL
i
And
UL UL UL
i
6. Effective Eb/Nt on FCH and SCH
334 AT271_TRG_E6 © Forsk 2009

UL
 Q eff  FCH is the uplink effective traffic channel quality on FCH at the receiver on icBS.
UL
 Q eff  SCH is the uplink effective traffic channel quality on SCH at the receiver on icBS.
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 
7. Uplink soft handover gain FCH and SCH

UL
 G SHO  FCH corresponds to the UL soft handover gain on FCH.
UL
 G SHO  SCH corresponds to the UL soft handover gain on SCH.
UL UL
UL  Q MAX  ic BS   FCH UL  Q MAX  ic BS   SCH
 G SHO  FCH = --------------------------------------------------------------------
- and  G SHO  SCH = --------------------------------------------------------------------
-
UL UL
max   Q max  i ic BS   FCH  max   Q max  i ic BS   SCH 
I I
UL UL
I
7.5.1.3.3 CDMA2000 1xEV-DO

For each cell (l,icBS) in the receiver active set, Atoll calculates the uplink quality level from the receiver. No power control
is performed as in simulations. Here, Atoll determines the uplink quality level at the cell for the maximum terminal power
allowed. Then, the total uplink quality level is evaluated with respect to the receiver handover status. From this value, Atoll
calculates the required terminal power and compares it with the maximum terminal power allowed.
1. Max terminal power
max
power allowed.
2. 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, Q max  i ic BS  , for each cell of active set
UL
-  G UL
Q max  i ic BS  = ---------------------------------------------------------- p
UL
max
UL P term
With P b –max  i ic BS  = -------------
-
LT
i

i

UL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx
UL N0
N tot  i ic BS  = ---------------------------------------
-
UL
1 – X  i ic BS 
tx
UL
2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status.

UL UL

UL UL UL
i
UL
 G macro – diversity  2 links is the uplink macro-diversity gain.This parameter is determined from the fixed cell edge coverage
© Forsk 2009 AT271_TRG_E6 335

UL UL
i

UL UL UL
i
UL
UL UL UL
i
we have:
 UL 
 Qmax  i icBS  Qmax
UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency   i ic BS 
 i on the other site

i on the same site
otherwise,
UL UL UL
i
UL
3rd step: Evaluation of the required quality level on uplink, Q req
In case of a 1xEV-DO Rev. 0 capable terminal, we have:
E c UL
Q req =  ------
UL UL
 G p   1 + G ACK + G DRC + G TCH 
 N t  min
Where
E
UL
-----c- is the minimum pilot quality level on uplink. This parameter is available in the Mobility types table.
 N t  min
G ACK , G DRC and G TCH are respectively acknowledgement, data rate control and traffic data gains relative to the pilot.
They are defined in the terminal properties (1xEV-DO Rev. 0 tab).
In case of a 1xEV-DO Rev. A capable terminal, we have:
E c UL
Q req =  ------
UL UL
 G p   1 + G ACK + G DRC + G TCH + G RRI + G Auxiliary – pilot 
 N t  min
Where
E
UL
-----c- UL
is the minimum pilot quality level required on uplink to obtain the defined data rate, R req . The required data rate,
 N t  min
UL
R req (i.e. the uplink data rate selected for the analysis) is obtained by using a certain uplink transmission format (i.e. 1xEV-
DO Rev. A radio bearer ( Index UL – RevA – Bearer ) with a certain number of subframes ( n SF )) and calculated as follows:
UL
UL R RLC – peak  Index UL – RevA – Bearer 
R req = ---------------------------------------------------------------------------------------------
-
n SF
E
UL
-----c- is the value defined in the 1xEV-DO Rev. A Radio Bearer Selection (Uplink) table for this uplink transmission
 N t  min
format (radio bearer Index, mobility and number of subframe). Two values are available for this parameter, one when the
service uplink mode is "Low Latency" and another one for high capacity services.
G ACK , G DRC , G TCH , G RRI and G Auxiliary – pilot are respectively acknowledgement, data rate control, traffic data chan-
nel, reverse rate indicator and auxiliary pilot channel gains relative to the pilot. They are defined in the terminal properties
(1xEV-DO Rev. A tab). Two values of G TCH are available, one when the service uplink mode is "Low Latency" and another
one for high capacity services.
req
4th step: Calculation of P term
UL
req Q req
-  P max
P term = ------------------------------- term
UL
Q MAX  ic BS 
req max
Therefore, the service on the uplink traffic data channel is available if P term  P term .
336 AT271_TRG_E6 © Forsk 2009

3. Required terminal power without ACK

Atoll also calculates the required terminal power without taking into account the ACK channel contribution. Calculations
are quite similar to those detailed in the previous paragraph, only the evaluation of the required quality on uplink is different.
E c UL
 Q req  withoutACK =  ------
UL UL
 G p   1 + G DRC + G TCH  for 1xEV-DO Rev. 0 capable terminals
 N t  min
And
Ec UL
 Q req  withoutACK =  ------
UL UL
 G p   1 + G DRC + G TCH + G RRI + G Auxiliary – pilot  for 1xEV-DO Rev. A capable terminals
 N t  min
And then,
UL
req  Q req  withoutACK max
 P term  withoutACK = -------------------------------------------
UL
 P term
Q MAX  ic BS 
4. UL SHO gain
UL
1st step: Evaluation of the uplink quality, Q max  i ic BS  , for each cell of active set.
UL
-  G UL
Q max  i ic BS  = ---------------------------------------------------------- p
UL
max
UL P term
With P b –max  i ic BS  = -------------
-
LT
i

i

UL
L T = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx max req
UL N0 P term – P term
- +  1 –  term   max (---------------------------------
N tot  i ic BS  = --------------------------------------- -,0)
UL
1 – X  i ic BS  LT
i
tx
UL
2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status.
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.
UL UL

UL UL UL
i
UL
UL UL
i

UL UL UL
i
UL
© Forsk 2009 AT271_TRG_E6 337

UL UL UL
i
we have:
 UL 
 Qmax  i icBS  Qmax
UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency   i ic BS 
 i on the other site

i on the same site
otherwise,
UL UL UL
i
3rd step: Calculation of the UL SHO gain

UL
G SHO corresponds to the uplink soft handover gain.
UL
G SHO = -----------------------------------------------------
UL
max  Q max  i ic BS  
i

Let us assume each pixel of the map corresponds to a probe receiver with associated terminal, mobility and service. This
receiver may be using a specific carrier or all of them. Moreover, it does not create any interference. Coverage studies are
based on the uplink load percentage and the downlink total power of cells. These parameters can either be outputs of a
simulation, average values calculated from a group of simulations or user-defined cell inputs.
7.5.2.1 Pilot Reception Analysis

For further details on calculation formulas, see "Definitions and Formulas" on page 276 For further details on calculations,
see "Bar Graph and Pilot Sub-Menu" on page 320
7.5.2.1.1 1st Case: Analysis Based on all Carriers

Atoll proceeds as in point analysis. It determines the best carrier of each transmitter i containing the receiver in its calcu-
lation area. The best carrier selection depends on the option chosen in Equipment (UL minimum noise, DL minimum
power, random, sequential) and is based on the UL load percentage and the downlink total power of cells (simulation
results or cell properties). 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 ( Q pilot  ic BS  ). Then, it deduces the best pilot quality
BS
Resulting
received with a fixed cell edge coverage probability, Q pilot  ic BS  .
7.5.2.1.2 2nd Case: Analysis Based on a Specific Carrier

lation area that may use the specified carrier (carrier specified in Cell Properties), Atoll calculates pilot quality at the
receiver on this carrier icgiven. Then, 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 cell edge coverage probability,
BS
Resulting
Q pilot  ic given  .
1. Single colour
Resulting req
Atoll displays a coverage if Q pilot  ic   Q pilot . Coverage consists of a single layer with a unique colour.
ic = ic BS or ic given
2. Colour per transmitter

Resulting req
Atoll displays a coverage if Q pilot  ic   Q pilot ( ic = ic BS or ic given ). Coverage consists of several layers with asso-
ciated colours. There is a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to
the best serving transmitter BS.
3. Colour per mobility
In this case, the receiver is not completely defined and no mobility assigned. Coverage consists of several layers with a
layer per user-defined mobility type defined in the Mobility Types sub-folder. For each layer, area is covered if
338 AT271_TRG_E6 © Forsk 2009

Resulting req
Q pilot  ic   Q pilot ( ic = ic BS or ic given ). Each layer is assigned a colour and displayed with intersections between
layers.
4. Colour per probability
and the “All” option in the Condition tab of prediction properties).
Coverage consists of several layers with a layer per user-defined probability level defined in the Display tab (Prediction
Resulting req
properties). For each layer, area is covered if Q pilot  ic   Q pilot ( ic = ic BS or ic given ) in the required number of
simulations. Each layer is assigned a colour and displayed with intersections between layers.
5. Colour per cell edge coverage probability
Resulting req
tab (Prediction properties). For each layer, area is covered if Q pilot  ic p   Q pilot ( ic = ic BS or ic given ). Each layer
6. Colour per quality level (Ec/I0)
Resulting
properties). For each layer, area is covered if Q pilot  ic    Q pilot  threshold ( ic = ic BS or ic given ). Each layer is
7. Colour per quality margin (Ec/I0 margin)
Resulting req
erties). For each layer, area is covered if Q pilot  ic  – Q pilot   Q pilot  m arg in ( ic = ic BS or ic given ). Each layer is
7.5.2.2 Downlink Service Area Analysis

The downlink service area analysis depends on the studied network (IS-95 cdmaOne, CDMA2000 1xRTT or CDMA2000
1xEV-DO). Several display options are available when calculating this study, some of which are dedicated to IS-95
cdmaOne and CDMA2000 1xRTT networks while others are relevant when analysing CDMA2000 1xEV-DO systems only.
7.5.2.2.1 IS-95 cdmaOne and CDMA2000 1xRTT

As in point analysis, Atoll calculates downlink quality on FCH at the receiver for each cell (k,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). No power
control is performed as in simulations. Here, Atoll determines the downlink quality on FCH at the receiver for a maximum
traffic channel power per transmitter allowed on the fundamental channel (FCH). Then, the total downlink quality on FCH
DL
(  Q MAX  ic  FCH ) is evaluated after recombination.
Note:
• Best server and active set determination is performed as in point prediction.
Atoll displays total traffic channel quality at the receiver on the carrier ic ( ic BS or ic given ).
For further details on formulas, see "Definitions and Formulas" on page 276. For further details on calculation, see "Down-
link Sub-Menu" on page 321.
You may choose following display options:
1. Single colour
DL DL DL
Atoll displays a coverage with a unique colour if  Q MAX  ic   FCH   Q req  FCH .  Q req  FCH is the downlink traffic data qual-
ity target on the fundamental channel (FCH). This parameter is user-defined for a given service and a terminal in the Serv-
ices sub-folder.
DL DL
Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH . Coverage consists of several layers with associated colours.
There is a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to best serving
transmitter.
layer per user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if
DL DL
 Q MAX  ic   FCH   Q req  FCH . Each layer is assigned a colour and displayed with intersections between layers.
4. Colour per service

DL DL
per user-defined service defined in Services sub-folder. For each layer, area is covered if  Q MAX  ic   FCH   Q req  FCH .
© Forsk 2009 AT271_TRG_E6 339


and the “All” option in the Condition tab of prediction properties).
Coverage consists of several layers with a layer per user-defined probability level defined in the Display tab (Prediction
DL DL
properties). For each layer, area is covered if  Q MAX  ic   FCH   Q req  FCH in the required number of simulations. Each
layer is assigned a colour and displayed with intersections between layers.
DL DL
tab (Prediction properties). For each layer, area is covered if  Q MAX  ic p   FCH   Q req  FCH . Each layer is assigned a
colour and displayed with intersections between layers.
7. Colour per maximum quality level (max Eb/Nt)
DL
properties). For each layer, area is covered if  Q MAX  ic   FCH  Threshold . Each layer is assigned a colour and displayed
8. Colour per effective quality level (Effective Eb/Nt)
DL
properties). For each layer, area is covered if  Q eff  ic   FCH  Threshold . Each layer is assigned a colour and displayed
9. Colour per quality margin (Eb/Nt margin)
DL DL
erties). For each layer, area is covered if  Q MAX  ic   FCH –  Q req  FCH  M arg in . Each layer is assigned a colour and
10. Colour per required power
req
Atoll calculates the downlink power required on FCH, P FCH  ic  , as follows:
DL
-  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. It corresponds to the maximum traffic data power
allowed on FCH for a transmitter.
Coverage consists of several layers with a layer per user-defined required power threshold defined in the Display tab
req
(Prediction properties). For each layer, area is covered if P FCH  ic   Threshold . Each layer is assigned a colour and
11. Colour per required power margin
max req
erties). For each layer, area is covered if P FCH – P FCH  ic   M arg in . Each layer is assigned a colour and displayed with
12. Colour per data rate
DL
This display option is relevant for CDMA2000 1xRTT data services only. For each possible data rate, R
DL DL DL DL DL DL DL DL DL DL
( R FCH  AF FCH , R FCH   AF FCH + 2  , R FCH   AF FCH + 4  , R FCH   AF FCH + 8  , R FCH   AF FCH + 16  ), Atoll calcu-
lates traffic channel quality at the receiver for each cell (k,ic) (with ic=icBS or icgiven). Downlink traffic channel quality at the
receiver is evaluated from a maximum traffic channel power per transmitter allowed for the corresponding data rate. Then,
DL DL
the total downlink traffic channel quality ( Q MAX  ic R  ) is calculated after recombination.
DL
Coverage consists of several layers with a layer per possible data rate, R . For each layer, area is covered if
DL DL DL DL
Q MAX  ic R  Q req  R  . Each layer is assigned a colour and displayed with intersections between layers.
DL DL DL
Q req  R  is the downlink traffic data quality target for the data rate, R . This parameter is user-defined for a given
service, terminal and data rate in the Services sub-folder.
7.5.2.2.2 CDMA2000 1xEV-DO

Ec
As in point analysis, Atoll calculates the effective pilot quality level at the receiver from the best server cell, ------  ic b pilot  .
Nt
Best server and active set determination is performed as in point prediction (AS analysis).Then, from this value, it deter-
DL
mines the effective downlink data rate received, R .
340 AT271_TRG_E6 © Forsk 2009

For further details on formulas, see "Definitions and Formulas" on page 276. For further details on calculations, see
"Downlink Sub-Menu" on page 321.
1xEV-DO Rev. 0 Users
For 1xEV-DO Rev. 0 users (users with EV-DO Rev. 0-capable terminals and EV-DO Rev. 0 services), the effective data
DL
rate ( R ) provided on downlink is determined from the graph “Max Rate=f(C/I) (Rev0)” defined for the mobility type
DL
selected in the Condition tab (Prediction properties). R is the value read in the graph “Max Rate=f(C/I) (Rev0)” for the
calculated effective pilot quality level, E c .
------  ic BS b pilot 
Nt
1xEV-DO Rev. A Users
For 1xEV-DO Rev. A users (users with EV-DO Rev. A-capable terminals and EV-DO Rev. A services), the effective data
DL
rate ( R ) provided on downlink depends on the downlink transmission format, i.e the radio bearer index
( Index DL –R evA – Bearer ) with the number of timeslots ( n TS ). Atoll selects the downlink transmission format where
Ec C
------  ic BS b pilot    ---- .
Nt  I  req Then, it determines the downlink effective data rate as follows:
DL
R -.
= ---------------------------------------------------------------------------------------------
n TS
The effective data rate corresponds to the guaranted data rate after a certain number of retransmissions (i.e. the number
of timeslots, n TS ).
When HARQ (Hybrid Automatic Repeat Request) is used, the required average number of retransmissions is smaller and
DL
the data rate is an average data rate ( R av ) calculated as follows:
DL
R av = -----------------------------------------------------------------------------------------------
-
DL
 n Rtx (Index DL – RevA – Bearer,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. The Early Termination Probability graph shows the probability of early termination
DL
( p ) as a function of the number of retransmissions ( n Rtx ). Atoll calculates the average number of retransmissions
DL
(  n Rtx  av ) as follows:
 n Rtx  max

DL DL DL
 p  n Rtx  – p  n Rtx – 1    n Rtx
DL
DL n =1
 n Rtx  av = -------------------------------------------------------------------------------------------------
Rtx
DL
p   n Rtx  max 
Display Options
You may choose following display options:

1. Colour per C/I
Coverage consists of several layers with a layer per quality threshold defined in the Display tab (Prediction properties).
Ec
For each layer, area is covered if ------  ic b pilot   Threshold . Each layer is assigned a colour and displayed with intersec-
Nt
2. Colour per rate
DL
Coverage consists of several layers with a layer per possible data rate ( R ). For each layer, area is covered if the data
DL
rate, R , can be obtained. Each layer is assigned a colour and displayed with intersections between layers.
3. Colour per average rate
This display option is available for 1xEV-DO Rev. A users only. It enables you to view the obtained downlink data rate
DL
when HARQ is used. Coverage consists of several layers with a layer per possible average data rate ( R av ). For each
DL
layer, area is covered if the average data rate, R av , can be obtained. Each layer is assigned a colour and displayed with
© Forsk 2009 AT271_TRG_E6 341

7.5.2.3 Uplink Service Area Analysis

The results displayed when calculating the uplink service area analysis depend on the studied network (IS-95 cdmaOne,
CDMA2000 1xRTT or CDMA2000 1xEV-DO).
7.5.2.3.1 IS-95 cdmaOne and CDMA2000 1xRTT

As in point analysis, Atoll calculates uplink quality on FCH from receiver for each cell (l,ic) (with ic=icBS or icgiven) in receiver
active set. No power control simulation is performed. Atoll determines uplink quality on FCH at the transmitter for the maxi-
UL
mum terminal power. Then, the total uplink traffic channel quality (  Q MAX  ic   FCH ) is evaluated with respect to the
receiver handover status.
Note:
Atoll displays uplink quality on FCH at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the
receiver.
For further details on formulas, see "Definitions and Formulas" on page 276. For further details on calculations, see "Uplink
Sub-Menu" on page 328.
1. Single colour
UL UL UL
Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH . Coverage colour is unique.  Q req  FCH is the uplink data traffic
quality target on the fundamental channel (FCH). This parameter is user-defined for a given service and a terminal in the
Services sub-folder.
UL UL
Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH . Coverage consists of several layers with associated colours.
There is a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to best server
transmitter.
layer per user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if
UL UL
 Q MAX  ic   FCH   Q req  FCH . Each layer is assigned a colour and displayed with intersections between layers.

UL UL
per user-defined service defined in Services sub-folder. For each layer, area is covered if  Q MAX  ic   FCH   Q req  FCH .
and the “All” option in the Condition tab of prediction properties). Coverage consists of several layers with a layer per user-
defined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if
UL UL
 Q MAX  ic   FCH   Q req  FCH in the required number of simulations. Each layer is assigned a colour and displayed with
UL UL
tab (Prediction properties). For each layer, area is covered if  Q MAX  ic p   FCH   Q req  FCH . Each layer is assigned a
7. Colour per maximum quality level (Max Eb/Nt)
UL
properties). For each layer, area is covered if  Q MAX  ic   FCH  Threshold . Each layer is assigned a colour and displayed
UL
properties). For each layer, area is covered if  Q effective  ic   FCH  Threshold . Each layer is assigned a colour and
UL UL
erties). For each layer, area is covered if  Q MAX  ic   FCH –  Q req  FCH  M arg in . Each layer is assigned a colour and
342 AT271_TRG_E6 © Forsk 2009


FCH – re q
The required terminal power, P term , is calculated as described in the Point analysis – AS analysis tab – Uplink sub-
menu part. Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab
FCH – re q
(Prediction properties). For each layer, area is covered if P term  ic   Threshold . Each layer is assigned a colour and
erties). For each layer, area is covered if P max – P FCH –re q  ic   M arg in . Each layer is assigned a colour and displayed
term term
12. Colour per rate
UL
This display option is relevant for CDMA2000 1xRTT data services only. For each possible data rate, R
UL UL UL UL UL UL UL UL UL UL
( R FCH  AF FCH , R FCH   AF FCH + 2  , R FCH   AF FCH + 4  , R FCH   AF FCH + 8  , R FCH   AF FCH + 16  ), Atoll calcu-
UL UL
lates the total uplink traffic channel quality ( Q MAX  ic R  ). Coverage consists of several layers with a layer per possible
UL UL UL UL UL
data rate, R . For each layer, area is covered if Q MAX  ic R   Q req  R  . Each layer is assigned a colour and
UL UL UL
displayed with intersections between layers. Q req  R  is the uplink traffic data quality target for the data rate, R . This
parameter is user-defined for the service, a given terminal and data rate in the service properties.
7.5.2.3.2 CDMA2000 1xEV-DO

As in point analysis, Atoll calculates uplink quality from receiver for each cell (l,ic) (with ic=icBS or icgiven) in receiver active
set. No power control simulation is performed. For 1xEV-DO Rev. 0 users, Atoll determines uplink quality at the transmitter
for the maximum terminal power allowed and an uplink data channel rate of 9.6 kbps. For 1xEV-DO Rev. A users, Atoll
determines uplink quality at the transmitter for the maximum terminal power allowed and an uplink data channel rate of
UL
4.8 kbps. Then, the total uplink quality ( Q MAX  ic  ) is evaluated with respect to the receiver handover status.
Note:
Atoll displays uplink quality at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the receiver.
For further details on formulas, see "Definitions and Formulas" on page 276. For further details on calculations, see "Uplink
Sub-Menu" on page 328.
1. Single colour
UL UL UL
Atoll displays a coverage if Q MAX  ic   Q req . Coverage colour is unique. For 1xEV-DO Rev. 0 users, Q req is the quality
UL
required on uplink for a 9.6 kbps data channel rate. For 1xEV-DO Rev. A users, Q req is the quality required on uplink for
a 4.8 kbps data channel rate. This parameter is calculated from the minimum uplink pilot quality and gains on the different
uplink channels.
We have:
E c UL
Q req =  ------
UL UL
 G p   1 + G ACK + G DRC + G TCH  for 1xEV-DO Rev 0 terminals,
 N t  min
And
Ec UL
Q req =  ------
UL UL
 G p   1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary – Pilot  for 1xEV-DO Rev A terminals.
 N t  min

UL UL
Atoll displays a coverage if Q MAX  ic   Q req . Coverage consists of several layers with associated colours. There is a layer
per transmitter with no intersection between layers. Layer colour is the colour assigned to best server transmitter.
UL UL
layer per user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if Q MAX  ic   Q req . Each
layer is assigned a colour and displayed with intersections between layers.
UL UL
per user-defined service defined in Services sub-folder. For each layer, area is covered if Q MAX  ic   Q req . Each layer is
© Forsk 2009 AT271_TRG_E6 343

and the “All” option in the Condition tab of prediction properties). Coverage consists of several layers with a layer per user-
defined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if
UL UL
Q MAX  ic   Q req in the required number of simulations. Each layer is assigned a colour and displayed with intersections
between layers.
UL UL
tab (Prediction properties). For each layer, area is covered if Q MAX  ic p   Q req . Each layer is assigned a colour and
7. Colour per maximum quality level (Max Eb/Nt)
UL
Here, Atoll calculates the total uplink traffic channel quality (  Q MAX  ic   TCH ).
UL
UL  Q req  TCH
-  P max
 Q MAX  ic   TCH = -------------------------- term
req
P term
With
Ec UL
 Q req  TCH =  ------
UL UL
 G p  G TCH
 N t  min
UL
properties). For each layer, area is covered if  Q MAX  ic   TCH  Threshold . Each layer is assigned a colour and displayed
UL
properties). For each layer, area is covered if  Q effective  ic   TCH  Threshold . Each layer is assigned a colour and
UL UL UL
 Q effective  ic   TCH = min   Q MAX  ic   TCH  Q req  TCH 

UL UL
erties). For each layer, area is covered if Q MAX  ic  – Q req  M arg in . Each layer is assigned a colour and displayed with
TCH – re q
The required terminal power on traffic data channel, P term , is calculated as described in the Point analysis – AS anal-
ysis tab – Uplink sub-menu part.
req
TCH – re q P term
P term = -------------------------------------------------------------------  G TCH for 1xEV-DO Rev 0 terminals,
1 + G ACK + G DRC + G TCH
And
req
TCH – re q P term
P term = -------------------------------------------------------------------------------------------------------------------------------------  G TCH for 1xEV-DO Rev A terminals.
1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary – Pilot
Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction
properties). For each layer, area is covered if P TCH –re q  ic   Threshold . Each layer is assigned a colour and displayed
term
erties). For each layer, area is covered if P max – P req  ic   M arg in . Each layer is assigned a colour and displayed with
term term
12. Colour per rate
UL UL UL
For each possible data rate, R , Atoll calculates the total uplink quality ( Q MAX  ic R  ). Coverage consists of several
UL UL UL UL
layers with a layer per possible data rate. For each layer, area is covered if Q MAX  ic R   Q req  R  . Each layer is
UL UL UL
Q req  R  is the uplink quality required to obtain the data rate, R .
• 1xEV-DO Rev. 0 users
344 AT271_TRG_E6 © Forsk 2009

UL
The possible data rates on uplink, R , are: 9.6, 19.2, 38.4, 76.8 and 153.6 kbps
Ec UL
Q req =  ------
UL UL
 G p   1 + G ACK + G DRC + G TCH 
 N t  min
Where
E
UL
-----c- is the minimum pilot quality level on uplink. This parameter is available in the Mobility types table.
 N t  min
G ACK , G DRC and G TCH are respectively acknowledgement, data rate control and traffic data gains relative to the pilot.
They are defined in the terminal properties (1xEV-DO Rev. 0 tab).
• 1xEV-DO Rev. A users
UL
The data rate, R is obtained when a certain uplink transmission format (i.e. 1xEV-DO Rev. A radio bearer
( Index UL – RevA – Bearer ) with a certain number of subframes ( n SF )) is used. It is calculated as follows:
UL
UL R RLC – peak  Index UL – RevA – Bearer 
R req = ---------------------------------------------------------------------------------------------
-
n SF
E c UL
Q req =  ------
UL UL
 G p   1 + G ACK + G DRC + G TCH + G RRI + G Auxiliary – pilot 
 N t  min
Where
E
UL
-----c- UL
is the minimum pilot quality level required on uplink to obtain the data rate, R . The value is defined in the 1xEV-
 N t  min
DO Rev. A Radio Bearer Selection (Uplink) table for the uplink transmission format (radio bearer Index, mobility and
number of subframe). Two values are available for this parameter, one when the service uplink mode is "Low Latency"
and another one for high capacity services.
G ACK , G DRC , G TCH , G RRI and G Auxiliary – pilot are respectively acknowledgement, data rate control, traffic data chan-
nel, reverse rate indicator and auxiliary pilot channel gains relative to the pilot. They are defined in the terminal properties
(1xEV-DO Rev. A tab). Two values of G TCH are available, one when the service uplink mode is "Low Latency" and another
one for high capacity services.
13. Colour per average rate
This display option is available for 1xEV-DO Rev. A users only. When HARQ (Hybrid Automatic Repeat Request) is used,
UL
the required average number of retransmissions is smaller and the data rate is an average data rate ( R av ) calculated as
follows:
UL
UL R RLC – peak  Index UL –R evA – Bearer 
R av = -----------------------------------------------------------------------------------------------
-
UL
 n Rtx  Index UL –R evA – 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.e. the radio bearer index ( Index UL –R evA – Bearer ) with the number of subframes
( n SF )). The Early Termination Probability graph shows the probability of early termination ( p ) as a function of the number
UL UL
of retransmissions ( n Rtx ). Atoll calculates the average number of retransmissions (  n Rtx  av ) as follows:
 n Rtx  max

UL UL UL
 p  n Rtx  – p  n Rtx – 1    n Rtx
UL
UL n =1
 n Rtx  av = -------------------------------------------------------------------------------------------------
Rtx
UL
p   n Rtx  max 
UL UL UL
For each possible average data rate, R av , Atoll calculates the total uplink quality ( Q MAX  ic R av  ). Coverage consists of
UL UL UL UL
several layers with a layer per possible average data rate. For each layer, area is covered if Q MAX  ic R av   Q req  R av  .
UL UL UL
Q req  R av  is the uplink quality required to obtain the average data rate, R av .
7.5.2.4 Downlink Total Noise Analysis

Atoll determines downlink total noise generated by cells.
For IS-95 cdmaOne and CDMA2000 1xRTT systems, we have:
© Forsk 2009 AT271_TRG_E6 345

DL
 Ptot  ic  + txj--------------------------------------
DL DL  j term
N tot  ic  = - + N0
txj j
For CDMA2000 1xEV-DO systems, we have:
 Ptot  icadj bpilot 

DL
 Ptot  ic bpilot  + txj-------------------------------------------------------

DL DL  j term
N tot  ic  = + N0
txj j
term
DL  N0 
Downlink noise rise, NR DL  ic  , is calculated from the downlink total noise, N tot , as: NR DL  ic  = – 10 log  -------------
-
 N DL tot

7.5.2.4.1 Analysis on all Carriers

If all the carriers are selected, Atoll determines DL total noise for all the carriers. Then, allows the user to choose different
displays.
1. Colour per minimum noise level
DL
erties). For each layer, area is covered if min NR tot  ic   Threshold . Each layer is assigned a colour and displayed
ic

2. Colour per maximum noise level
DL
erties). For each layer, area is covered if max NR tot  ic   Threshold . Each layer is assigned a colour and displayed with
ic
3. Colour per average noise level
DL
erties). For each layer, area is covered if average NR tot  ic   Threshold . Each layer is assigned a colour and displayed
ic
4. Colour per minimum noise rise
Atoll displays bins where min NR DL  ic   Threshold . Coverage consists of several areas with an area per user-defined
ic
5. Colour per maximum noise rise
Atoll displays bins where max NR DL  ic   Threshold . Coverage consists of several areas with an area per user-defined
ic
6. Colour per average noise rise
Atoll displays bins where average NR DL  ic   Threshold . Coverage consists of several areas with an area per user-
ic
defined noise rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
7.5.2.4.2 Analysis on a Specific Carrier

When only one carrier is analysed, Atoll determines DL total noise or DL noise rise on this carrier. In this case, the
displayed coverage is the same for any selected display per noise level (average, minimum or maximum) or any display
per noise rise (average, minimum or maximum).
1. Colour per noise level
DL
erties). For each layer, area is covered if N tot  ic   Threshold . Each layer is assigned a colour and displayed with inter-
2. Colour per noise rise
Atoll displays bins where NR DL  ic   Threshold . Coverage consists of several areas with an area per user-defined noise
rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
7.6 Automatic Neighbour Allocation

Atoll permits the automatic allocation of intra-technology neighbours in the current network. Two allocation algorithms are
available, one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours.
346 AT271_TRG_E6 © Forsk 2009

The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all
the cells of TBC transmitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfill the following conditions:
Note:
7.6.1 Global Allocation for all Transmitters

We assume a reference cell A and a candidate neighbour, cell B. When automatic allocation starts, Atoll checks following
conditions:
is discarded.
Force co-site cells as neighbours: This option enables you to force cells located on the reference cell site in the candidate
neighbour list. This constraints can be weighted among the others and ranks the neighbours through the importance field
(see after).
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.This constraints can be weighted among the others and ranks
the neighbours through the importance field (see after).
ence cell is a candidate neighbour of another cell, this one will be considered as candidate neighbour of the reference cell.
force/forbid a cell to be candidate neighbour of the reference cell.
Note:
• Adjacence criterion: Let CellB be a candidate neighbour cell of CellA. 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).
When this option is checked, adjacent cells are sorted and listed from the most adjacent to
the least, depending on the above criterion. Adjacence is relative to the number of pixels sat-
isfying the criterion.
• Intra-carrier neighbours: intra-carrier handover is a soft handover.

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).
SA is the area where the cell A is the best serving cell. It means that the cell A is the first one in the active set.
- The pilot quality from A exceeds Min. Ec/I0.
- The pilot quality from A is the best.
SB is the area where the cell B can enter the active set.
- The pilot quality from B is greater than T_Drop.
• Inter-carrier neighbours: inter-frequency handover is a hard handover. It is needed in a multi-carrier (1xRTT and
1xEV-DO carriers) CDMA network:
- To balance loading between carriers and layers (1st case),
- To make a coverage reason handover from micro cell frequency to macro cells (2nd case).
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.
© Forsk 2009 AT271_TRG_E6 347

SA is the area where:
- The pilot signal from A is not the highest one. It is strictly lower than the best pilot signal received and higher
than the best pilot signal minus the margin.
SB is the area where:
- The pilot signal from B is the highest one.
Figure 7.5: Overlapping Zones - 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.
SA is the area where:
- 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.
SB is the area where:
- The pilot signal from B is the highest one.
Figure 7.6: Overlapping Zones - 2nd Case
Note:
• Two ways enable you to determine the I0 value:
1 - A reduction factor (% of maximum powers contributing to I0) may be applied to cell
maximum powers (defined in Cell properties) to customize their contribution to I0. Thus, I0
represents the sum of effective powers received from the other cells. The entered
percentage is a kind of downlink load factor estimation.
If the % of maximum powers contributing to I0 is too low, i.e. if %  P max  P pilot , Atoll
takes into account the pilot powers to evaluate the I0 value.
348 AT271_TRG_E6 © Forsk 2009

2 - Atoll takes into account load parameters defined per cell (such as the total downlink
power used). I0 represents the sum of total transmitted powers.
SA  SB
SA
after).
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 importances) will be allocated to the reference cell. Note that
specific maximum numbers of neighbours (maximum number of intra-carrier neighbours, maximum number of inter-carrier
neighbours) can be defined at the cell level (property dialog or cell table). If defined there, this value is taken into account
instead of the default one available in the dialog.
to 100%.
Importance
value
Existing neighbour
Co-site transmitter Only if the Force co-site cells as neighbours option is selected (IF) function
Only if the Force adjacent cells as neighbours option is
selected
coverage conditions
relationship
(IF) Importance Function evaluation. The importance is evaluated through a function (IF), taking into account 3 factors:
• The co-site factor (C) which is a Boolean factor,
• The adjacency factor (A) which deals with the percentage of adjacency,
Min Max

(IF) function
no yes 30% + 30%{30%(O) + 70%(A)}
Max(O))(A)}
yes yes 60% + 40%{30%(O) + 70%(A)}
Max(O))(A)}
Where Delta(x)= Max(x) - Min(x)

Notes:
© Forsk 2009 AT271_TRG_E6 349

factors.
tions and the corresponding surface area (km2), the percentage of area meeting the adjacency conditions and the corre-
sponding surface area (km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing.
Notes:
• No simulation or 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.
• Even if no specific terminal, mobility or service is selected in the automatic allocation, 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.
Mobility: no impact on the allocation, no specific selection.
Terminal: selection of the one with the greatest (Gain - Loss) value, and, if equal, the one
with the lowest noise figure.
• The neighbour lists may be optionally used in the power control simulations to determine
the mobile's active set.
• The percentage of area is calculated with the resolution specified in the properties dialog of
the predictions folder (default resolution parameter).
Thus, if the cell B is a neighbour of the cell A while the cell A is not a neighbour of the cell
B, two cases are possible:
1st case: There is space in the cell B neighbour list: the cell A will be added to the list. It will
be the last one.
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.

• TBA cells,
• Neighbours of TBA cells marked as exceptional pair, adjacent and symmetric,
• Neighbours of TBA cells that satisfy coverage conditions.
Automatic neighbour allocation parameters are described in "Global Allocation for all Transmitters" on page 347.
7.7 PN Offset Allocation

PN offset is used to identify a cell. It is a time offset used by a cell to shift a Pseudo Noise sequence. Mobile processes
the strongest received PN sequence and reads its phase that identifies the cell.
By default, there are 512 PN Offsets. PN Offsets are numbered (0...511).
350 AT271_TRG_E6 © Forsk 2009

The cells to which Atoll allocates PN Offsets are referred to as the TBA cells (cells to be allocated). TBA cells fulfil 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.
Note:

7.7.1.1 Options and Constraints
The PN Offset allocation algorithm can take into account following constraints and options:
1. PILOT_INC parameter,
Atoll uses this parameter to determine the pool of possible PN offsets (512 divided by PILOT_INC value). The first
PN offset is PILOT_INC and other ones are multiples of this value.
For example: When PILOT_INC is set to 4, the pool of possible PN offsets consists of PN offsets from 4 to 508
with a separation interval of 4 (i.e. [4,8,12,16,...508]).
2. Neighbourhood between cells,

You may consider:
• First order neighbours: The neighbours of TBA cells listed in the Intra-technology neighbours table,
• Second order neighbours: The neighbours of neighbours,
• Third order neighbours: The neighbour’s neighbour’s neighbours.
Notes:
• In the context of the PN Offset allocation, the term "neighbours" refers to intra-carrier
neighbours.
different PN Offsets to the CDMA neighbours of a GSM transmitter. In order to consider
inter-technology neighbour relations in the PN Offset allocation, you must make the
Transmitters folder of the GSM .atl document accessible in the CDMA .atl document. For
information on making links between GSM and CDMA .atl documents, see the User
Manual.
3. Cells fulfilling a criterion on Ec/I0 (option “Additional Ec/I0 conditions”),

For a reference cell “A”, 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. Ec/I0 and is the highest one and (Ec/I0)B exceeds T_Drop).
Note:
• Atoll takes the total downlink power used by the cell into account in order to evaluate Io. Io
equals the sum of total transmitted powers. In case this parameter is not specified in the
cell properties, Atoll uses 50% of the maximum power.
4. Co-PN Reuse distance,

Reuse distance is a constraint on the allocation 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.
Note:
• PN offset reuse distance can be defined at cell level. If this value is not defined, then Atoll
will use the default reuse distance defined in the PN offset Automatic Allocation dialog.
5. PN-cluster size. Within the context of PN offset allocation, the term "PN-cluster" refers to a sub-group of PN
offsets.
6. Exceptional pairs,
7. Domains of PN Offsets,
Note:
• When no domain is assigned to cells, Atoll considers the PILOT_INC parameter only to
determine available PN offsets (e.g., If PILOT_INC is set to 4, all PN offsets from 4 to 508
with a separation interval of 4 can be allocated).
© Forsk 2009 AT271_TRG_E6 351

8. The carrier on which the allocation is run: It can be a given carrier or all of them. In this case, either Atoll
independently plans PN Offsets for the different carriers, or it allocates the same PN Offset to each carrier of a
transmitter if the option "Allocate carriers identically" is selected.
9. 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.
10. The "Delete All Codes" option: When selecting this option, Atoll deletes all the current PN Offsets and carries out
a new PN Offset allocation. If not selected, the existing PN Offsets are kept.
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 allo-
cated 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.

and "First Order"),
• The cells that fulfil Ec/I0 condition (option “Additional Ec/I0 conditions”),
One additional constraint is considered when:
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 353.

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 353.
Strategy: Adjacent PN-Clusters per site
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. Other-
wise, 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 trans-
mitter 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.
Strategy: Distributed PN-Clusters per site
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,
352 AT271_TRG_E6 © Forsk 2009

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 assigns a PN offset from the cluster to each cell of each transmitter located on the sites according to the trans-
mitter 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.

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 353. For information on calculating transmitter prior-
ity, see "Transmitter Priority" on page 355.
Note:
alphanumeric order.
7.7.1.2.3 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}).

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
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
C i  Dist  =  Cj  Dist  i  
j
© Forsk 2009 AT271_TRG_E6 353

Where

C i  EP  =  cEP  i – j 
j
Where

     
Ci  N  = 
C j  N1  i   = I j  c N1
Where
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
Where
Because two second order neighbours must not have the same PN Offset, Atoll considers the cost created by two second
C k  N2  i   + C k  N2  i  
C k – k  N2  i   = ------------------------------------------------------------
-
2
C l  N3  i   = Max  j   c N3
Where
354 AT271_TRG_E6 © Forsk 2009

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
Note:
And

C i  N 2G  =  cN 2G
 j – Tx 2G 
j
Where
2G
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 

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.
as:
With C Tx  U  = Max  C  U   and C  Dom  = 512 – Number of PN offsets in the domain

i  Tx i Tx

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.
© Forsk 2009 AT271_TRG_E6 355

7.7.2 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 7.8: 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 sepa-
ration 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.
7.7.2.1 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 alphanu-
meric order.
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 Atoll allocates the first three PN offsets in the domain (4, 8
and 12) to the Site0’s cells. Under given constraints of and 12) to the Site0’s cells. As it is allowed to use a maxi-
neighbourhood and reuse distance, same PN offsets can mum of PN offsets, Atoll allocates different PN offsets to
be allocated to each site’s cells. each site’s cells so that there is least repetition.
356 AT271_TRG_E6 © Forsk 2009

7.7.2.2 Strategy: Adjacent PN-Clusters Per Site

meric order.
Atoll allocates a PN cluster of adjacent PN offsets to Site0 As it is possible to use a maximum of PN offsets, Atoll allo-
and then, one PN offset of the PN cluster to each cell. cates different PN clusters of adjacent PN offsets to sites so
Under given constraints of neighbourhood and reuse that there is least repetition of PN offsets.
distance, the same PN cluster can be allocated to each site
and same PN offsets to each site’s cells.
7.7.2.3 Strategy: Distributed PN-Clusters Per Site

meric order.
Atoll allocates a PN cluster of distributed PN offsets to Site0 As it is possible to use a maximum of PN offsets, Atoll allo-
and then, one PN offset of the PN cluster to each cell. cates different PN clusters of distributed PN offsets to sites
Under given constraints of neighbourhood and reuse so that there is least repetition of PN offsets.
distance, the same PN cluster can be allocated to each site
and same PN offsets to each site’s cells.
7.8 Automatic GSM-CDMA Neighbour Allocation

7.8.1 Overview
It is possible to automatically calculate and allocate neighbours between GSM/TDMA and CDMA (i.e. IS-95 cdmaOne or
CDMA2000) networks. In Atoll, it is called inter-technology neighbour allocation.
• 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.
© Forsk 2009 AT271_TRG_E6 357

Note that the automatic inter-technology neighbour allocation algorithm takes into account both cases.
• An .atl document containing the GSM/TDMA network, GSM.atl, and another one containing the CDMA (i.e. IS-95
cdmaOne or CDMA2000) network, CDMA.atl,
• An existing link on the Transmitters folder of GSM.atl into CDMA.atl.
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:
Only CDMA TBA cells may be assigned neighbours.

• The inter-transmitter distance,
• The maximum number of neighbours fixed,
• Allocation options,
• The selected allocation strategy,
We assume we have a CDMA reference cell, A, and a GSM candidate neighbour, transmitter B.

1. 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
Note:
• Transmitter azimuths are taken into account to evaluate the inter-transmitter distance. For
further information on inter-transmitter distance calculation, see "Calculation of Inter-
Transmitter Distance" on page 270.

the reference CDMA cell in the candidate neighbour list. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference CDMA cell.

distance conditions
358 AT271_TRG_E6 © Forsk 2009

be marked as exceptional pair or co-site. If the neighbour is not forced but satisfies distance conditions, Atoll displays the

1. 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
Note:
• Here, real inter-transmitter distance is considered.

the reference CDMA cell in the candidate neighbour list. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference CDMA cell.
3. There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability where:
- 1st case: SA is the area where the cell A is the best serving cell of the CDMA network.
- 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 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 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 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
user-defined.
- 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.
is the highest one.
- 2nd case: The margin is different from 0dB and SB is the area where:
SA  SB
SA
area.
© Forsk 2009 AT271_TRG_E6 359

Guidelines for the automatic allocation

When the automatic allocation is based on coverage overlapping, we recommend you to perform two successive auto-
matic allocations:
- 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.

coverage conditions
be marked as exceptional pair or co-site. If the neighbour is not forced but fulfils coverage conditions, Atoll displays the
Notes:
not found.
dialog of the predictions folder (default resolution parameter).
7.8.2.3 Delete Existing Neighbours Option

A new TBA cell i is created in CDMA.atl. Therefore, if you start a new allocation without selecting the Delete existing neigh-
bours 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 trans-
mitter) 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.
7.9 Automatic cdmaOne-CDMA2000 Neighbour

Allocation
7.9.1 Overview
It is possible to automatically calculate and allocate neighbours between cdmaOne and CDMA2000 networks. In Atoll, it
is called inter-technology neighbour allocation.
360 AT271_TRG_E6 © Forsk 2009

In order to be able to use the external neighbour allocation algorithm, you must have:
• An .atl document containing the CDMA2000 network, CDMA2000.atl, and another one containing the cdmaOne
network, CDMA.atl,
• An existing link of the Transmitters folder of CDMA.atl into CDMA2000.atl or vice-versa.
The external neighbour allocation algorithm takes into account all the CDMA TBC cells. It means that all the TBC cells of
CDMA.atl are potential neighbours. The cells to be allocated will be called TBA cells which, being cells of CDMA2000.atl,
fulfill following conditions:
Only CDMA2000 TBA cells may be assigned neighbours.

We assume we have a CDMA2000 reference cell, A, and a CDMA candidate neighbour, cell B.
1. The distance between the CDMA2000 reference cell and the CDMA neighbour must be less than the user-
definable maximum inter-site distance. If the distance between the CDMA2000 reference cell and the CDMA
neighbour is greater than this value, then the candidate neighbour is discarded.
Note:
information on inter-transmitter distance calculation, please refer to paragraph "Calculation
of Inter-Transmitter Distance" on page 270.

Force co-site cells as neighbours: It enables you to automatically include CDMA cells located on the same site as the refer-
ence CDMA2000 cell in the candidate neighbour list. This option is automatically selected.
force/forbid a CDMA cell to be candidate neighbour of the reference CDMA2000 cell.

distance conditions
be marked as exceptional pair or co-site. If the neighbour is not forced but fulfils distance conditions, Atoll displays the
Notes:
© Forsk 2009 AT271_TRG_E6 361

7.9.2.1 Delete Existing Neighbours Option

A new TBA cell i is created in CDMA2000.atl. Therefore, if you start a new allocation without selecting the Delete existing
neighbours option, Atoll determines the neighbour list of the transmitter i,
If you change some allocation criteria (e.g. increase the maximum number of neighbours or create a new CDMA TBC cell)
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 CDMA TBC cell can enter the TBA cell
neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list.
362 AT271_TRG_E6 © Forsk 2009

Chapter 8
TD-SCDMA Networks
This chapter provides descriptions of all the algorithms for calculations, analyses, automatic allocations
and prediction studies available in TD-SCDMA projects.
Atoll
364 AT271_TRG_E6 © Forsk 2009

Chapter 8: TD-SCDMA Networks
8 TD-SCDMA Networks
This chapter describes in detail the algorithms, calculation parameters, and processes of the coverage predictions and the
simulations available in TD-SCDMA documents. The first part of this chapter lists all the input and output parameters in
the TD-SCDMA documents, their significance, location in the Atoll GUI, and their usage. Detailed explanation of the basic
coverage predictions, which do not require simulation results, is provided in the second part. The third part describes the
traffic scenario generation and Montel Carlo simulation algorithms including smart antenna modelling and dynamic chan-
nel allocation. The next sections are dedicated to TD-SCDMA coverage predictions which can be based on results
obtained from simulations. The last three sections describe in detail the allocation of frequencies, i.e., master and slave
carriers, the allocation of neigbours, and the allocation of scrambling codes.

The tables in the following subsections list the input and output parameters and formulas used in simulations and other
computations.
8.1.1 Inputs
This table lists the inputs to computations, coverage predictions, and simulations.
R Ch Global parameter Mcps Chip rate (or Spreading rate) (1.28)
Spread
F Min Global parameter None Minimum spreading factor (1)
Spread
F Max Global parameter None Maximum spreading factor (16)
Proc
G P – CCPCH Global parameter None P-CCPCH processing gain (13.8 dB)
SF
N TS Global parameter None Number of timeslots per subframe (7)
SF Global parameter ms Subframe duration (5)

D
Frame Global parameter ms Frame duration (10)
D
GP Number of guard period chips per

N Ch  TS Global parameter None
timeslot (16)
Data Number of data chips per timeslot
(704)
Midamble Number of midamble chips per
timeslot (144)
GP Number of guard period chips per pilot
N Ch  PTS Global parameter None
timeslot (96)
N Ch  DwPTS Global parameter None
DwPTS timeslot (32)
SYNC_DL Number of SYNC_DL chips per
N Ch  DwPTS Global parameter None
DwPTS timeslot (64)
Global parameter Total number of chips per DwPTS
Total
N Ch  DwPTS Total GP SYNC_DL None
N Ch  DwPTS = N Ch  DwPTS + N Ch  DwPTS timeslot (96)

N Ch  UpPTS Global parameter None
UpPTS timeslot (32)
SYNC_UL Number of SYNC_UL chips per
N Ch  UpPTS Global parameter None
UpPTS timeslot (128)
Global parameter Total number of chips per UpPTS
Total
N Ch  UpPTS Total GP SYNC_UL None
N Ch  UpPTS = N Ch  UpPTS + N Ch  UpPTS timeslot (160)
Calculated global parameter

Data
W N Ch  TS bps Chip rate (140800 bps)
W = ------------------
-
SF
D
Average frequency range of the
F Avg Frequency band parameter MHz
frequency band (2010)
© Forsk 2009 AT271_TRG_E6 365

Channel bandwidth of the carriers of a

BW Frequency band parameter MHz
frequency band (1.6)
F IRF Cell parameter None Interference reduction factor
TX
F JD Site equipment parameter None Joint Detection (JD) factor
TX Transmitter parameter
NF None BTS Noise Figure
(user-defined or calculated from transmitter equipment characteristics)
Transmitter loss
TX Transmitter parameter L Tx = L Total – UL on uplink
L None
(user-defined or calculated from transmitter equipment characteristics)
L Tx = L Total – DL on downlink
TX Percentage of BTS signal correctly

 BTS parameter None
transmitted
Max
P TCH 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
P-CCPCH RSCP comparative

TComp P – CCPCH Cell parameter None
threshold for baton handover
Maximum difference between two
P Max Cell parameter None
transmitted powers
Req
RU UL Cell parameter None Required resource units in uplink
Req
RU DL Cell parameter None Required resource units in downlink
Available HS-PDSCH power available per

P HS – PDSCH Cell parameter W
downlink timeslot
P HR Cell parameter None Power headroom
HS-SCCH power per downlink

P HS – SCCH Cell parameter W
timeslot
N HS – SCCH Cell parameter None Number of HS-SCCH channels
N HS – SICH Cell parameter None Number of HS-SICH channels
Max
N HSDPA Cell parameter None Maximum number of HSDPA users
Min Minimum number of HS-PDSCH

N HS-PDSCH Codes Cell parameter None
codes
Max Maximum number of HS-PDSCH
N HS-PDSCH Codes Cell parameter None
codes
Max Maximum number of intra-technology
N Intra – Neigh Cell parameter None
neighbours
Max Maximum number of inter-technology
N Intra – Neigh Cell parameter None
neighbours
P OCCH Timeslot parameter W Other common channel powera
DL
P TCH Timeslot parameter W Downlink traffic power
Max Timeslot parameter Maximum percentage of downlink

%P DL None
(Simulation constraint) used power
Timeslot parameter
X UL None Uplink load factor
(Simulation result)
Max Timeslot parameter
X UL None Maximum uplink load factor
(Simulation constraint)
Available
P HS – PDSCH Timeslot parameter W HS-PDSCH power available
Min Minimum number of HS-PDSCH

N HS-PDSCH Codes Timeslot parameter None
codes
Max Maximum number of HS-PDSCH
N HS-PDSCH Codes Timeslot parameter None
codes
366 AT271_TRG_E6 © Forsk 2009

RU Overhead Timeslot parameter Overhead resource units
Body Service parameter None Body loss

L
Downlink activity factor for circuit-
Act
f DL Service parameter None switched services and the A-DPCH
activity factor for HSDPA services
Uplink activity factor for circuit-
Act
f UL Service parameter None switched services and the A-DPCH
activity factor for HSDPA services
Eff Downlink efficiency factor for circuit-
f DL Service parameter None
switched services
Eff Uplink efficiency factor for circuit-
f UL Service parameter None
switched services
F Scaling Service parameter None Application througput scaling factor
O TP Service parameter kbps Application throughput offset
UL Average number of packet calls on the

N PacketCall Service parameter (packet session modelling) None
uplink during a session
DL Average number of packet calls on the
N PacketCall Service parameter (packet session modelling) None
downlink during a session
UL Average time between two packet
T PacketCall Service parameter (packet session modelling) ms
calls on the uplink
DL Average time between two packet
T PacketCall Service parameter (packet session modelling) ms
calls on the downlink
UL
S Min – PacketCall Service parameter (packet session modelling) KBytes Minimum packet call size on the uplink
DL Minimum packet call size on the

S Min – PacketCall Service parameter (packet session modelling) KBytes
downlink
UL Maximum packet call size on the
S Max – PacketCall Service parameter (packet session modelling) KBytes
uplink
DL Maximum packet call size on the
S Max – PacketCall Service parameter (packet session modelling) KBytes
downlink
UL Average time between two packets on
T Packet Service parameter (packet session modelling) ms
the uplink
DL Average time between two packets on
T Packet Service parameter (packet session modelling) ms
the downlink
UL
S Packet Service parameter (packet session modelling) Bytes Packet size on uplink
DL
S Packet Service parameter (packet session modelling) Bytes Packet size on downlink
Nom
R DL R99 bearer parameter kbps Downlink nominal bit rate
Nom
R UL R99 bearer parameter kbps Uplink nominal bit rate
W
G DL
Proc R99 bearer parameter (Can be calculated as -------------
Nom
-) None Downlink processing gain
R DL
W
G UL
Proc R99 bearer parameter (Can be calculated as -------------
Nom
-) None Uplink processing gain
R UL
Min Allowed minimum downlink traffic

P TCH – DL R99 bearer parameter W
channel power
Max Allowed maximum downlink traffic
P TCH – DL R99 bearer parameter W
channel power
TS
N DL R99 bearer parameter None Number of downlink timelots
TS
N UL R99 bearer parameter None Number of uplink timelots
E b Req C Req
Req
Q TCH – UL R99 bearer parameter per mobility (  ------ or  ---- ) None Eb/Nt or C/I target on uplink
 N t  TCH – UL  I  TCH – UL
E b Req C Req
Req
Q TCH – DL R99 bearer parameter per mobility (  ------ or  ---- ) None Eb/Nt or C/I target on downlink
 N t  TCH – DL  I  TCH – DL
© Forsk 2009 AT271_TRG_E6 367

Req
RSCP TCH – UL R99 bearer parameter per mobility W Target RSCP on uplink TCH
Req
RSCP TCH – DL R99 bearer parameter per mobility W Target RSCP on downlink TCH
Div
G DL R99 bearer parameter per mobility None Downlink diversity gain
Div
G UL R99 bearer parameter per mobility None Uplink diversity gain
Term
P Max Terminal parameter W Maximum terminal power
Term
P Min Terminal parameter W Minimum terminal power
P UpPCH Terminal parameter W UpPCH power
Term Terminal parameter None Terminal Noise Figure

NF
Term
F JD Terminal parameter None Joint Detection (JD) factor
Term Percentage of terminal signal correctly

 Terminal parameter None
transmitted
Term Terminal parameter None Terminal gain
G
Term Terminal parameter None Terminal loss
L
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
RSCP DwPCH Mobility parameter W Required RSCP threshold for DwPCH
Req
RSCP UpPCH Mobility parameter W Required RSCP threshold for UpPCH
E b Req C Req Required quality threshold for P-

Req
Q P – CCPCH Mobility parameter (  ------ or  ---- ) None
 N t  P – CCPCH  I  P – CCPCH CCPCH
E c Req Required quality threshold for HS-

Req
Q HS – SCCH Mobility parameter (  ------ ) None
 N t  HS – SCCH SCCH
E c P – CCPCH Required quality threshold for P-

Req
Q HS – SICH Mobility parameter (  ------ ) None
 N t  HS – SICH CCPCH
C Req
Req
Q DwPCH Mobility parameter (  ---- ) None Required quality threshold for DwPCH
 I  DwPCH
Model Clutter class parameter None Model standard deviation


Eb/Nt
 P – CCPCH or P-CCPCH Eb/Nt or C/I standard
Clutter class parameter None
CI deviation
 P – CCPCH
Eb/Nt CI Downlink Eb/Nt or C/I standard

 DL or  DL Clutter class parameter None
deviation
Eb/Nt CI
 UL or  UL Clutter class parameter None Uplink Eb/Nt or C/I standard deviation
L Indoor Clutter class parameter None Indoor loss
Ortho
F DL Clutter class parameter None Downlink orthogonality factor
Ortho
F UL Clutter class parameter None Uplink orthogonality factor
 Spread Clutter class parameter ° Spreading angle
K 1.38 x 10-23 J/K Boltzman constant

TX
N0 NF TX  K  T  BW W Thermal noise at transmitter
Term
N0 NF Term  K  T  BW W Thermal noise at terminal
TX Antenna parameter None Transmitter antenna gain

G
368 AT271_TRG_E6 © Forsk 2009

L Path Propagation model result None Path loss
Model Result calculated from cell edge coverage probability and model Model shadowing margin used in
M Shadowing None
standard deviation coverage predictions
P – CCPCH Result calculated from cell edge coverage probability and P-CCPCH P-CCPCH Eb/Nt shadowing margin
M Shadowing None
Eb/Nt standard deviation used in coverage predictions
 Eb  Nt  DL Result calculated from cell edge coverage probability and DL Eb/Nt DL Eb/Nt shadowing margin used in
M Shadowing None
 Eb  Nt  UL Result calculated from cell edge coverage probability and UL Eb/Nt UL Eb/Nt shadowing margin used in
M Shadowing None
For RSCP calculation Transmitter-terminal total loss in
TX Term Body Model coverage predictions
Model L Path  L  L L  L Indoor  M Shadowing
LT = ------------------------------------------------------------------------------------------------------------------------------------
-
TX Term
G G In UL, only carrier power is attenuated
 Eb  Nt  UL
For P-CCPCH Eb/Nt calculation by M Shadowing .
TX Term Body P – CCPCH
 Eb  Nt  P L Path  L  L L  L Indoor  M Shadowing
LT = ------------------------------------------------------------------------------------------------------------------------------------
-
G G
TX Term In DL, carrier power and intra-cell
LT None interference are attenuated by
For DL Eb/Nt calculation  Eb  Nt  DL P – CCPCH
 Eb  Nt  DL M Shadowing or M Shadowing while
TX Term Body
 Eb  Nt  DL L Path  L  L L  L Indoor  M Shadowing
LT = ------------------------------------------------------------------------------------------------------------------------------------
- extra-cell interference is not.
TX Term
G G
 Eb  Nt  DL
For UL Eb/Nt calculation Therefore, M Shadowing or
TX Term Body  Eb  Nt  UL
 Eb  Nt  UL L Path  L  L L  L Indoor  M Shadowing P – CCPCH
M Shadowing are set to 1 in DL extra-
LT = ------------------------------------------------------------------------------------------------------------------------------------
-
TX Term
G G cell interference calculation.
a. For the calculation of interference, P OCCH also includes the MBMS SCCPCH channel power when the optional MBMS
feature is activated. You must modify the data structure for activating the optional MBMS feature. For more information, see
the Administrator Manual.
8.1.2 P-CCPCH Eb/Nt and C/I Calculation

TX i TX i  ic 
TX i  ic    RSCP P – CCPCH
E-----b- ------------------------------------------------------------ Proc
 G P – CCPCH None P-CCPCH Eb/Nt for the cell TX i  ic 
 N t  P – CCPCH TX i  ic 
N Tot – DL
TX i  ic    RSCP P – CCPCH
C
---- ------------------------------------------------------------ None P-CCPCH C/I for the cell TX i  ic 
 I  P – CCPCH TX i  ic 
N Tot – DL
TX i  ic  TX i  ic  TX i  ic  Downlink total noise for the cell

Term W
N Tot – DL I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0 TX i  ic 
TX i  ic  TX i TX i  ic  TX i
RSCP P – CCPCH     + RSCP OCCH – TS0  
TX i TX i
TX i  ic  With  =    1 – F DL
Ortho
   1 – F JD
Term
 and Downlink intra-cell interference for the
I Intra – DL W
cell TX i  ic 

 =  0 Without Useful Signal
 1 Total Noise
TX j  ic  TX j  ic  Downlink extra-cell interference for the

TX i  ic 
I Extra – DL   RSCPP – CCPCH + RSCPOCCH – TS0  W
ji
TX j  jc  TX j  jc 
I IC – DL  ic jc 
  RSCPP – CCPCH + RSCPOCCH – TS0  W Inter-carrier interference
TX
-----------------------------------------------------------------------------------------------------------
j -
F IRF  ic jc 
8.1.3 DwPCH C/I Calculation

© Forsk 2009 AT271_TRG_E6 369

TX i  ic    RSCP DwPCH
C
---- ---------------------------------------------------- None DwPCH C/I for the cell TX i  ic 
 I  DwPCH TX i  ic 
N Tot – DL

Term W
TX i  ic  TX i
RSCP DwPCH    
TX i TX i
TX i  ic  With  =    1 – F DL
Ortho
   1 – F JD
Term
 and Intra-cell interference for the cell
I Intra – DL W
TX i  ic 

 =  0 Without Useful Signal
 1 Total Noise
TX j  ic  Extra-cell interference for the cell

TX i  ic 
I Extra – DL   RSCPDwPCH  W
TX i  ic 
ji
TX j  jc 
  RSCPDwPCH  W Inter-carrier interference
TX
-----------------------------------------------
j -
8.1.4 DL TCH Eb/Nt and C/I Calculation

When the optional MBMS feature is activated, the MBMS Eb/Nt is also calculated in the same manner. You must modify
the data structure for activating the optional MBMS feature. For more information, see the Administrator Manual.

TX i  ic    RSCP TCH – DL Downlink TCH Eb/Nt for the cell
E-----b- ------------------------------------------------------- Proc
 G DL  G DL
Div
None
 N t  TCH – DL TX i  ic  TX i  ic 
N Tot – DL
TX i  ic    RSCP TCH – DL
C
---- -------------------------------------------------------  G Div None Downlink TCH C/I for the cell TX i  ic 
 I  TCH – DL TX i  ic  DL
N Tot – DL

Term W
 TX i Ortho Term TX i 
    1 – F DL    1 – F JD  +  1 –    Downlink intra-cell interference for the
TX i  ic    W
I Intra – DL cell TX i  ic 
TX i  ic  TX i  ic 
  RSCP TCH – DL + RSCP OCCH 
TX j  ic  TX j  ic  Downlink extra-cell interference for the

TX i  ic 
I Extra – DL   RSCPTCH – DL + RSCPOCCH  W
ji
  RSCPTCH – DL + RSCPOCCH  W Inter-carrier interference
TX
------------------------------------------------------------------------------------------
j -
8.1.5 UL TCH Eb/Nt and C/I Calculation

Term TX i  ic 
TX i  ic    RSCP TCH – UL
E-----b- -----------------------------------------------------------
- Proc
 G UL  G UL
Div
None Uplink TCH Eb/Nt for the cell TX i  ic 
 N t  TCH – UL TX i  ic 
N Tot – UL
Term TX i  ic 
C
---- -  G Div
----------------------------------------------------------- None Uplink TCH C/I for the cell TX i  ic 
 I  TCH – UL TX i  ic  UL
N Tot – UL
Req Req
Term Q TCH – UL Q TCH – UL
Term - or P Term
P Max  -------------------------------- Max  ------------------------------
P Req E TX i  ic  TX i  ic  W Uplink required power for the terminal
 ------
b C ----
370 AT271_TRG_E6 © Forsk 2009

8.1.6 Interference Calculation

  RSCPTCH – DL + RSCPOCCH 
I C2C  TX i TX j  TX j  ic  TX j  ic  W Cell to cell interference
  RSCPTCH – DL + RSCPOCCH  +------------------------------------------------------------------------------------------
TX j
-
TX j
TX i  ic 
TX i  ic  TX i X TS1 – UL
I TS1 – UL N0  -------------------------------------
TX i  ic 
W UpPCH interference
 1 – X TS1 – UL 
8.1.7 HSDPA Dynamic Power Calculations

TX i  ic 
E -----c-
TX i  ic 
  N Tot – DL –     RSCP HS – SCCH 
TX i  ic   N t  HS – SCCH
P HS – SCCH W HS-SCCH power
-------------------------------------------------------------------------------------------------------------------------------------------------  L Model
T
TX i

TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic 
P HS – PDSCH P Max – DL – Eff – P R99 – DL – P HR – P HS – SCCH W HS-PDSCH power
TX i  ic 
E -----c-
TX i  ic 
  N Tot – UL –     RSCP HS – SICH 
Mi Mi
Mi  N t  HS – SICH W HS-SICH power
P HS – SICH -------------------------------------------------------------------------------------------------------------------------------------------  L Model
Mi T

8.1.8 Smart Antenna Modelling

E SA Smart antenna model parameter None Number of smart antenna elements
 Calculation parameter Degrees Angle of arrival for the useful signal

Angle at which the smart antenna
 Calculation parameter Degrees
effect is calculated

--- , where  is the wavelength of the signal
Distance between two adjacent
d m
2 antenna elements
gn    Smart antenna model parameter None Gain of a single element
2 2 2 T
j  -------  d  sin  j  -------  2d  sin  j  -------   E SA – 1 d  sin 
   None Steering vector for the direction of 
S 1 e e  ... e
2
– j  -------  nd  sin 

e
wn None Complex smart antenna weight
– j    n  sin  
e with d = ---
2
H Array correlation matrix for a given

R R = S  S None
user direction 
H H H
SA
gn     S  R  S = gn     S  S  S  S
G  None Smart antenna gain in any direction 
SA
10  Log  G     in dB
J
R Avg  j  pj  Rj None Average array correlation matrix

j=1
–1 Vector of ESA complex weights for the

ŵ   RN  S None
beam former
E SA
 ----------------------------------
- None Smart antenna gain in the uplink
H –1
S  RN  S
© Forsk 2009 AT271_TRG_E6 371

J
H
 pj  Sj  Sj
2
RN Rn + RI = n I+ None Total noise correlation matrix
j=1
2
Rn n  I None Thermal noise correlation matrix
J
H
RI
 pj  Sj  Sj None Interference correlation matrix
j=1
Total noise power, including thermal

Pˆ N
2 H –1
  S  RN  S W noise and interference from all uplink
interferers
2 Total power received from the served
Pˆ  2 H –1
p     S  RN  S  W
user
2 H –1 2
Pˆ p     S  RN  S  H –1
Q UL
SA
------- = --------------------------------------------------------------- = p  S  RN  S None Signal quality in the uplink
Pˆ 2 H –1
N   S  RN  S
H
SA
S   I  S  = E SA Uplink smart antenna beam forming
G None
10  Log  E SA  in dB gain in the direction of the served user
K
–1 1 Average of the inverse noise
 RN
–1
RN ----  W
Avg K k correlation matrices
k=1
8.2 Signal Level Based Calculations

Two types of signal level based calculations are available in Atoll:
1. Point Analysis: Real-time calculations for profile and reception analysis using the mouse to move a probe mobile
on the map.
2. RSCP Based Coverage Predictions: Calculation of RSCP related parameters on each pixel and colouring
according to the selected display.

For the selected transmitted TXi and carrier (ic), you can study three parameters in point analysis Profile tab:

Signal level ( RSCP ) in dBm TX i  ic  TX i  ic  Model
RSCP = EIRP – L Path – M Shadowing – L Indoor
TX i
Path loss ( L Path ) in dB L Path = L Model + L Ant
TX i Model TX i
Total losses ( L T ) in dB L T = L Path + L + L Indoor + M Shadowing – G
Where,
RSCP is the received signal code power for the P-CCPCH.
TX i  ic  TX i  ic  TX i TX i
EIRP is the effective isotropic radiated power of the transmitter. EIRP = P P – CCPCH + G –L .
ic is a carrier number
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. This parameter is taken into account when the option “Shadowing taken into
account” is selected
L Indoor are the indoor losses. These losses are defined for each clutter class. They are taken into account when the option
"Indoor coverage" is selected
TX i
G is the transmitter antenna gain
372 AT271_TRG_E6 © Forsk 2009

TX i TX i
L is the transmitter loss ( L = L Total – DL )
Notes:
• It is possible to analyse all the carriers. In this case, Atoll takes the highest P-CCPCH
power of cells to calculate the signal level received from a transmitter.
8.2.1.1 Profile Tab

TX i  ic 
Atoll displays either the signal level received from the selected transmitter on a carrier ( RSCP P – CCPCH ), or the highest
signal level received from the selected transmitter on all the carriers.
Notes:
• For a selected transmitter, it is also possible to study the path loss, L Path , or the total
losses, L T . Path loss and total losses are the same on any carrier.

Analysis provided in the Reception tab is based on path loss matrices. You can study reception from TBC transmitters for
which path loss matrices have been calculated on their calculation areas.
TX i  ic 
For each transmitter, Atoll displays either the signal level received on a carrier, ( RSCP P – CCPCH ), or the highest signal
level received on all the carriers.
Received signal level bar graphs are displayed in a decreasing signal level order. The number of bars in the graph depends
on the signal level received from the best server. Only bars for transmitters whose signal level is within a 30 dB margin
from the best server signal are displayed.
Note:
8.2.2 RSCP Based Coverage Predictions

For each TBC transmitter, TXi, Atoll determines the value of the selected parameter on each studied pixel inside the TXi
calculation area. Each pixel within the TXi calculation area is considered a probe receiver.
• 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. Coverage prediction display resolution is independent of the path loss matrix and
geographic data resolutions, and can be different for each coverage prediction. Coverage predictions are calculated using
bilinear interpolation of multi-resolution path loss matrices (similar to the evaluation of site altitudes).
8.2.2.1 Calculation Criteria

The RSCP from a transmitter TXi and a selected carrier (ic) is given by:
TX i  ic  TX i  ic  Model Term Term

RSCP = EIRP – L Path – M Shadowing – L Body – L Indoor + G –L
Where,
RSCP is the received signal code power. RSCP can be calculated for P-CCPCH, DwPCH, or the downlink TCH.
EIRP is the effective isotropic radiated power of the transmitter. EIRP P – CCPCH = P P – CCPCH + G –L ,
TX i  ic  TX i  ic  TX i TX i TX i  ic  TX i  ic  TX i TX i
EIRP DwPCH = P DwPCH +G –L , or EIRP DL – TCH = P DL – TCH +G –L .
ic is a carrier number
TX i
L Path = L Model + L Ant
TX i
© Forsk 2009 AT271_TRG_E6 373

Model
Term
L is the terminal loss
L Body is the body loss defined in the service
Term
G is the receiver total gain
TX i
TX i TX i
8.2.2.2 P-CCPCH RSCP Coverage Prediction

8.2.2.2.1 Coverage Condition
This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the P-CCPCH. The cover-
age prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. 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. Afterwards, the coverage
prediction is calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any
pixels covered by this transmitter. If you perform this coverage prediction for all the carriers, Atoll calculates the RSCP
considering:
• the preferred carrier of the selected service, or
• the carrier with the highest P-CCPCH power, if no preferred carrier is defined for the service, or
• the master carrier in case of N-frequency mode compatible transmitters.
You can select the display colours according to the RSCP, or on any best server parameter.
TX i  ic 
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  are covered and coloured
according to the selected display parameter.
8.2.2.2.2 Coverage Display

It is possible to colour the transmitter service areas using a unique colour per transmitter, or colour the pixels in the cover-
age 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. Where other
service areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is coloured if
TX i  ic 
RSCP P – CCPCH  TAdd P – CCPCH  Mobility  . The pixel colour depends on the RSCP level. Coverage consists
of several independent layers whose visibility in the workspace can be managed. 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.
• RSCP Margin (dB)

Coverage consists of several layers with a layer per user-defined RSCP margin defined in the Display tab (Predic-
TX i  ic  RSCP
tion properties). For each layer, area is covered if RSCP P – CCPCH – TAddP – CCPCH  Mobility   M P – CCPCH .
• Cell Edge Coverage Probability (%)

TX i  ic 
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the RSCP P – CCPCH
from the transmitter exceeds TAddP – CCPCH defined in the mobility selected in the Conditions tab, with different
cell edge coverage probabilities. There is one coverage area per transmitter in the explorer.
8.2.2.3 Best Server P-CCPCH Coverage Prediction

This coverage prediction calculates and displays the best server RSCP for the P-CCPCH. The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The best server for the cover-
age 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. Afterwards, the coverage prediction is calculated
for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this
transmitter. If you perform this coverage prediction for all the carriers, Atoll calculates the RSCP considering:
374 AT271_TRG_E6 © Forsk 2009

TX i  ic 
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and where
TX i  ic  Best  RSCP TXj  jc 

RSCP P – CCPCH = P – CCPCH  will be covered and coloured according to the transmitter colour.
j = All
8.2.2.4 P-CCPCH Pollution Coverage Prediction

This coverage prediction calculates and displays the number of P-CCPCH polluters. Atoll calculates the Received Signal
Code Power (RSCP) for the P-CCPCH for each pixel in the TX i  ic  coverage area where
TX i  ic 
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 a given set of a terminal type, a mobility type, a service, a carrier, and for TS0.
The best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the high-
est P-CCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the
coverage prediction is calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will
not be any pixels covered by this transmitter. If you perform this coverage prediction for all the carriers, Atoll calculates the
RSCP considering:
Atoll determines the number of transmitters covering each pixel and colours the pixel according to the number of polluting
transmitters. Coverage consists of several independent layers whose visibility in the workspace can be managed. There
are as many layers as defined thresholds. Each layer corresponds to an area where the number of servers exceeds (  )
a defined minimum threshold.
8.2.2.5 DwPCH RSCP Coverage Prediction

This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the DwPCH. The coverage
prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for DwPTS. 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. Afterwards, the coverage
prediction is calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any
pixels covered by this transmitter. If you perform this coverage prediction for all the carriers, Atoll calculates the RSCP
considering:
TX i  ic  Req
The pixels in the TX i  ic  coverage area where RSCP DwPCH  RSCP DwPCH  Mobility  are covered and coloured
according to the selected display parameter.

• DwPCH RSCP (dBm)
TX i  ic 
Atoll calculates the best RSCP DwPCH received from each transmitter TX i  ic  on each pixel.. Where other serv-
ice areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is coloured if
TX i  ic  Req
RSCP DwPCH  RSCP DwPCH  Mobility  . The pixel colour depends on the RSCP level. Coverage consists of
several independent layers whose visibility in the workspace can be managed. 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.

TX i  ic  Req RSCP
tion properties). For each layer, area is covered if RSCP DwPCH – RSCP DwPCH  Mobility   M DwPCH . Each layer
© Forsk 2009 AT271_TRG_E6 375

• Cell edge coverage probability (%)

TX i  ic 
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the RSCP DwPCH
TX i  ic 
from the transmitter TX i  ic  exceeds RSCP DwPCH defined in the mobility selected in the Conditions tab, with
different cell edge coverage probabilities. There is one coverage area per transmitter in the explorer.
8.2.2.6 UpPCH RSCP Coverage Prediction

This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the UpPCH in the uplink.
The coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for UpPTS.
RSCP considering:
Atoll uses the UpPCH power of the selected terminal to calculate the RSCP from each pixel of each transmitter’s best
server coverage area.
Term Req
The pixels where RSCP UpPCH  RSCP UpPCH  Mobility  are covered and coloured according to the selected display
parameter.

• UpPCH RSCP (dBm)
Term
Atoll calculates the best RSCP UpPCH received from each pixel of each transmitter service area at the transmitter.
Where other service areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is
Term Req
coloured if RSCP UpPCH  RSCP UpPCH  Mobility  . The pixel colour depends on the RSCP level. Coverage
consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as defined thresholds. Each layer corresponds to an area where the RSCP at the best server exceeds a

Term Req RSCP
tion properties). For each layer, area is covered if RSCP UpPCH – RSCP UpPCH  Mobility   M UpPCH . Each layer

On each pixel of each transmitter service area, the coverage corresponds to the pixels from which the
Term Term
RSCP UpPCH at the transmitter exceeds RSCP UpPCH defined in the mobility selected in the Conditions tab, with
different cell edge coverage probabilities. There is one coverage area per transmitter in the explorer.
8.2.2.7 Baton Handover Coverage Prediction

This coverage prediction determines the pixels which receive RSCP from cells other than the best server high enough to
perform baton handovers. Received Signal Code Power (RSCP) is calculated for the P-CCPCH. The coverage prediction
is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. 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. Afterwards, the coverage prediction is
calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered
by this transmitter. If you perform this coverage prediction for all the carriers, Atoll calculates the RSCP considering:
The pixels are covered and coloured according to the selected display parameters, where
TX i  ic  TX j  jc  TX j  jc 
RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and RSCP P – CCPCH  TDrop P – CCPCH  Mobility  – TComp P – CCPCH .
376 AT271_TRG_E6 © Forsk 2009


It is possible to display the potential handover areas or the number of transmitters covering each pixel.
• Handover Areas
Atoll displays the pixels where there are transmitters other than the best server that satisfy the above criteria.
Coverage consists of a single layer with a defined colour whose visibility in the workspace can be managed.
• Number of Potential Servers

Atoll determines the number of transmitters covering each pixel and colours the pixel according to the number of
transmitters. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the number of servers
exceeds (  ) a defined minimum threshold.
8.2.2.8 Scrambling Code Interference Analysis

This coverage prediction calculates and displays the pixels covered by two cells using the same scrambling code. Atoll
calculates the Received Signal Code Power (RSCP) for the P-CCPCH for each pixel in the TX i  ic  coverage area where
TX i  ic 
RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and determines the interfering transmitters according to:
RSCP P – CCPCH  Best  RSCP P – CCPCH  – M
ji
Where M is the specified pollution margin.

The coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0.
RSCP considering:
Atoll determines whether the cells of two transmitters covering a pixel have the same scrambling code. If the pixel is inter-
fered, Atoll colours it according to the colour assigned to the scrambling code in the display parameters. Coverage consists
of several independent layers whose visibility in the workspace can be managed. There are as many layers as scrambling
codes. Each layer corresponds to the area where the corresponding scrambling code has interference. A layer corre-
sponding to areas where more than one scrambling code interferes is also available.
8.3 Monte Carlo Simulations

The simulation process is divided into two steps.
• Generating a realistic user distribution as explained in "Generating a Realistic User Distribution" on page 377.
Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data. The resulting user
distribution complies with the traffic database and maps selected when creating simulations.
• Dynamic channel allocation and power control as explained under "Power Control Simulation" on page 383.

During each simulation, Atoll performs two random trials. 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.
• "Simulations Based on Raster and Vector Traffic Maps" on page 378.
• "Simulations Based on Traffic Map Based on Transmitters and Services" on page 381.
Notes:
• Atoll follows a Poisson distribution to determine the total number of users attempting a
connection in each simulation. In order for Atoll to use a constant total number of users
attempting a connection, the following lines must be added to the Atoll.ini file:
[CDMA]
RandomTotalUsers=0
Each user is randomly assigned a service, a terminal, and a mobility type. The activity status is determined based on the
calculations of activity probabilities using the traffic inputs.
The user activity status influences the next step of the simulation, i.e., the interference in the network. Both active and
inactive users use radio resources and generate interference.
© Forsk 2009 AT271_TRG_E6 377

Once all the user characteristics have been determined, a second random trial is performed to obtain their geographical
locations weighted according to the clutter classes, and whether they are indoor or outdoor according to the percentage
of indoor users per clutter class defined for the traffic maps.
Atoll also calculates the shadowing margin for each user based on the standard deviations defined for the clutter class of
each user.
In TD-SCDMA networks users accessing packet-switched services can transmit either on uplink or on downlink, but never
on both simultaneously. Users accessing circuit-switched services transmit on both uplink and downlink simultaneously.
Circuit-switched service users, mobiles connected in uplink and downlink both, are modelled in Atoll by two mobiles gener-
ated at the same location with one connected on the uplink and the other on the downlink. If one of these two mobiles is
rejected for some reason, the other is also rejected due to the same reason.
8.3.1.1 Simulations Based on Raster and Vector Traffic Maps

a list of user profiles with an associated mobility type and a given density, i.e., number of users of a user profile per km².
Vector traffic maps are based on user profiles. 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, each point is assigned a number of users with given user
profile and mobility type.
The user profile models the behaviour of the different user categories. Each user profile contains a list of services and their
associated parameters describing how these services are accessed by the user.
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).
N Users = S Env  D UP
Notes:
• In case of vector traffic map composed of lines, 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 the vector traffic map is composed of points.
At any given instant, 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, i.e., the number of voice calls or data sessions, the average
duration of each voice call, or the volumes of the data exchanged in the uplink and the downlink in each data session.

User profile parameters for circuit switched services are:
• The user terminal equipment used for the service (from the Terminals table),
• The average number of calls per hour N Call ,
• The average duration of a call (seconds) D Call .
N Call  D Call
Calculation of the service usage duration per hour ( p 0 : probability of a connection): p 0 = --------------------------------
-
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, i.e., the uplink and downlink activity
UL DL
factors defined for the circuit switched service i, f Act and f Act .

UL DL
Probability of being inactive: p Inactive =  1 – f Act    1 – f Act 
UL UL DL
Probability of being active on UL: p Active = f Act   1 – f Act 
DL DL UL
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
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
378 AT271_TRG_E6 © Forsk 2009

Therefore, a connected user can be either active on both links, inactive on both links, active on UL only, or active on DL
only.

User profile parameters for packet switched services are:
• The user terminal equipment used for the service (from the Terminals table),
• The average number of packet sessions per hour N Sess ,
DL UL
• The volume (in kBytes) which is transferred on the downlink V and the uplink V during a session.
A packet session consists of several packet calls separated by a reading time. 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.
Figure 8.1: Description of a Packet Session
Calculation of the average packet call size (kBytes):

UL DL
UL V DL V
S PacketCall = ------------------------------------------- and S PacketCall = -------------------------------------------
UL UL DL DL
N PacketCall  f Eff N PacketCall  f Eff
UL DL
In case of HSDPA services, f Eff and f Eff are the uplink and downlink A-DPCH activity factors, respectively.
Calculation of the average number of packets per packet call:

UL DL
UL  S PacketCall   S PacketCall 
- + 1 and N DL
N Packet = Int  ------------------------------------ Packet = Int  ------------------------------------
- + 1
UL
 S Packet  1024  S DLPacket  1024

Note:
• 1 kBytes = 1024 Bytes.
Calculation of the average duration of inactivity within a packet call (c):

UL UL DL DL
UL  N Packet – 1   T Packet DL  N Packet – 1   T Packet
 D Inactivity  PacketCall = ---------------------------------------------------------------
- and  D Inactivity  PacketCall = ---------------------------------------------------------------
-
1000 1000
Calculation of the average duration of inactivity in a session (s):

UL UL UL
 D Inactivity  Session = N PacketCall   D Inactivity  PacketCall and
DL DL DL
 D Inactivity  Session = N PacketCall   D Inactivity  PacketCall
Calculation of the average duration of activity in a session (s):

UL UL
UL UL N Packet  S Packet  8
 D Activity  Session = N PacketCall  ------------------------------------------------------
UL
- and
R Nom  1000
DL DL
DL DL N Packet  S Packet  8
 D Activity  Session = N PacketCall  ------------------------------------------------------
DL
-
R Nom  1000
Therefore, 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):
© Forsk 2009 AT271_TRG_E6 379

UL N Sess UL DL N Sess 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, the number of users trying to access the service j is:
n j = N Users  p Connected
As Figure 8.1: on page 379 shows, there can be three possible cases when a user is connected:
a. 1st case: At a given time, packets are downloaded and uploaded.

UL DL
UL + DL p Connection  p Connection
The probability of being connected is: p Connected = ----------------------------------------------------------------
-
p Connected
b. 2nd case: At a given time, packet are uploaded only.

UL DL
UL p Connection   1 – p Connection 
The probability of being connected is: p Connected = ------------------------------------------------------------------------------
-
p Connected
c. 3rd case: At a given time, packet are downloaded only.

DL UL
DL p Connection   1 – p Connection 
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, the activity periods during the connection are taken into account.
UL DL
UL  D Activity  Session DL  D Activity  Session
f - and f
= -------------------------------------------------------------------------------------------------- = --------------------------------------------------------------------------------------------------
-
UL UL DL DL
 D Inactivity  Session +  D Activity  Session  D Inactivity  Session +  D Activity  Session
Therefore, we have:
a. 1st case: At a given time, packets are downloaded and uploaded.

UL UL DL UL + DL
The probability of the user being active on UL and inactive on DL: p1 Active = f  1 – f   p Connected
DL DL UL UL + DL
The probability of the user being active on DL and inactive on UL: p1 Active = f  1 – f   p Connected
UL + DL UL DL UL + DL
The probability of the user being active on both UL and DL: p1 Active = f f  p Connected
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. 2nd case: At a given time, packet are uploaded only.

UL UL UL
The probability of the user being active on UL and inactive on DL: p2 Active = f  p Connected
UL UL
The probability of the user being inactive on both UL and DL: p2 Inactive =  1 – f   p Connected
c. 3rd case: At a given time, packet are downloaded only.

DL DL DL
The probability of the user being active on DL and inactive on UL: p1 Active = f  p Connected
DL DL
The probability of the user being inactive on both UL and DL: p3 Inactive =  1 – f   p Connected
Number of inactive users on UL and DL: n j – Inactive = n j   p1 Inactive + p2 Inactive + p3 Inactive 
UL UL UL
Number of users active on UL and inactive on DL: n j – Active = n j   p1 Active + p2 Active 
DL DL DL
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, a connected user can be active on both links, inactive on both links, active on UL only, or active on DL
only.
380 AT271_TRG_E6 © Forsk 2009

Notes:
• The user distribution per service, and the activity status distribution between the users are
average distributions. The service and the activity status of each user are random in each
simulation. Therefore, if you compute several simulations at once, the average number of
users per service and average numbers of inactive, active on UL, active on DL, and active
on UL and DL users, will correspond to calculated distributions. But, if you compare each
simulation, you will observe that the user distribution between services as well as the
activity status distribution between users is different in each simulation.
8.3.1.2 Simulations Based on Traffic Map Based on Transmitters and Services

Traffic maps based on transmitters and services are also referred to as live traffic maps. Live traffic data from the OMC is
spread over the best server coverage areas of the transmitters included in the traffic map. Throughput demands per serv-
ice, the numbers of active users per service, or Erlangs per service are assigned to the coverage areas of each transmitter.
For each transmitter TXi and each service S,
• Traffic Map Based on Transmitters and Services (Throughputs)

Atoll calculates the number of active users of each service S on UL and DL in the coverage area of TXi as follows:
UL DL
UL RS DL RS
- N = -------------- and N = -------------- for R99 circuit and packet switched services
UL DL
R Nom R Nom
DL
DL RS
- N - for HSDPA service
= ------------
DL
R Avg
UL DL
RS and R S are the uplink and downlink rates for service S in the TXi cell from the traffic map.
• Traffic Map Based on Transmitters and Services (# Active Users)

UL DL
Atoll directly uses the defined N and N values, i.e., the number of active users on UL and DL in the TXi
coverage area using the service S.
• Traffic Map Based on Transmitters and Services (Erlangs)

UL DL
Atoll directly uses the defined N and N values, i.e., the number of active and inactive users on UL and DL
in the TXi coverage area using the service S.

There can be two cases for circuit-switched service:
1. Number of active users on uplink and downlink (NUL and NDL),
2. Erlangs, i.e, a total number of users (n).
Active Users on Uplink and Downlink
UL DL UL DL
N and N values include users active on uplink ( n i – Active ), on downlink ( n i – Active ) and on both links
UL + DL
( n i – Active ). They are calculated as follows:
UL DL
UL UL DL
DL DL UL
UL + DL UL DL
Calculation of the total number of active users, n i – Active , attempting to access the circuit switched service
i:
UL UL UL + DL DL DL UL + DL
We have: N =  p Active + p Active   n i – Active and N =  p Active + p Active   n i – Active
Therefore,
UL UL + DL DL UL + DL
UL + DL  N  p Active N  p Active 
- -------------------------------------------
n i – Active = Min  ------------------------------------------- -
 p Active + p Active p DL
UL UL + DL
Active + p Active
UL + DL

© Forsk 2009 AT271_TRG_E6 381

UL UL UL + DL
n i – Active = N – n i – Active
DL DL UL + DL
n i – Active = N – n i – Active
And
UL DL UL + DL
n i – Active = n i – Active + n i – Active + n i – Active
Calculation of the number of inactive users, n i – Inactive , for the circuit switched service i
The number of inactive users is calculated from the total number of active users as follows:
n i – Active
n i – Inactive = ------------------------------  p Inactive
1 – p Inactive
Erlangs
n i is the total number of users trying to access the service i. This figure includes both active and inactive users.
They are determined as follows:
UL DL
UL UL DL
DL DL UL
UL + DL UL DL
Number of inactive users: n i – Inactive = n i  p Inactive
UL UL
Number of users active on UL: n i – Active = n i  p Active
DL DL
Number of users active on DL: n i – Active = n i  p Active
UL + DL UL + DL
Number of users active on UL and DL both: n i – Active = n i  p Active
Therefore, a connected user can have four different activity status: either active on both links, active on UL only,
active on DL only, or inactive.

Atoll considers all the users as active. Activity probabilities are not calculated.
nj is the total number of users in the TXi cell using the packet switched service j.
UL DL UL DL UL + DL
N and N values include users active on uplink ( n j – Active ), on downlink ( n j – Active ) and on both links ( n j – Active ).
UL DL
If N N
UL + DL UL
n j – Active = N
UL
n j – Active = 0
DL DL UL
n j – Active = N –N
UL DL
If N N
UL + DL DL
n j – Active = N
DL
n j – Active = 0
UL UL DL
n j – Active = N –N
Therefore, we have:
382 AT271_TRG_E6 © Forsk 2009

UL DL UL + DL
n j – Active = n j – Active + n j – Active + n j – Active
Note:
• The activity status of users is based on an average distribution. The activity status of each
user is random in each simulation. Therefore, if you compute several simulations at once,
the average numbers of inactive, active on UL, active on DL, and active on UL and DL
users, will correspond to calculated distributions. But, if you compare each simulation, you
will observe that the activity status distribution between users is different in each
simulation.
8.3.2 Power Control Simulation

Based on CDMA air interface, a TD-SCDMA network automatically regulates itself by using uplink and downlink power
control in order to minimise interference and maximise capacity. For each user distribution, Atoll simulates these network
regulation mechanisms using an iterative algorithm and calculates network parameters such as traffic power per cell and
per timeslot, mobile terminal power, and handoff status for each terminal.
In each iteration, all the mobiles (R99 and HSDPA service users) selected during generation of the user distribution
attempt to connect to the network one by one. The process is repeated from iteration to iteration and ends when the
network is balanced, i.e., when the convergence criteria on uplink and downlink are satisfied.
The simulation algorithm also models the impact of smart antennas in the power control loop. The influence of smart anten-
nas is taken into account in signal quality calculations. Smart antennas improve the signal quality of each served mobile,
decrease the required powers and the loads of all the surrounding cells. Interference on the downlink and the uplink is
calculated on a per user. Power control is simulated over a sub-frame, i.e., 7 timeslots.
For HSDPA users, uplink and downlink power control is performed on the associated A-DCH bearer before fast link adap-
tation on downlink. The steps of this algorithm are detailed below.
Figure 8.2: TD-SCDMA Power Control Algorithm
8.3.2.1 Algorithm Initialisation

At the start of each simulation, 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
© Forsk 2009 AT271_TRG_E6 383

• Uplink interference powers received on all the carriers and timeslots I Intra – UL and I Extra – UL are initialised to 0
Watts (i.e., no connected mobiles)
Term
• Uplink required power for mobiles is set to P Min
8.3.2.2 R99 Part of the Algorithm

The algorithm is described for an iteration k. Here, Xk is the value of the variable X at the iteration k. In the algorithm, all
Req Req
Q UL and Q DL thresholds depend on the user mobility, and are defined in the Service and Mobility parameter tables.
All the variables used in the description below are listed in "Definitions and Formulas" on page 365.
The following calculations are made for all R99 and HSDPA mobiles (Mi) using R99 bearers.
8.3.2.2.1 Determination of Mi’s Best Server (SBS(Mi))

This step is performed for TS0 for each station TXi containing Mi in its calculation area.
The best server for Mi 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. Afterwards, the P-CCPCH RSCP is calcu-
lated for:
• the preferred carrier of the service used by Mi, or
The RSCP from a transmitter TXi and a selected carrier ic is given by:
TX i  ic  TX i  ic  TX i TX i Model Mi Mi Mi
RSCP P – CCPCH = P P – CCPCH + G –L – L Path – M Shadowing – L Body – L Indoor + G –L in dBm
Where,
TX i
TX i
Model
Mi
L is the los of the terminal used by Mi
Mi
L Body is the body loss defined in the service used by Mi
Mi
G is the receiver gain of the terminal user by Mi
TX i
TX i TX i
A cell TX i  ic  is considered the best server of a mobile Mi if it satisfies the following conditions:
TX i  ic  TX i  ic  Best  RSCP TXj  jc 

RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and RSCP P – CCPCH = P – CCPCH  .
j = All
The best server is determined once for the whole simulation during the first iteration, i.e., k = 0, because the best server
does not change during the simulation and smart antennas do not influence this step.
Mi is considered unable to connect to the network if no best server has been selected. In this case, Mi is rejected for the
reason P-CCPCH RSCP < Min P-CCPCH RSCP. If Mi has no best server, it is not taken into account in the next steps.
8.3.2.2.2 Dynamic Channel Allocation

The dynamic channel allocation is performed once for the whole simulation during the first iteration, i.e., k = 0. The DCA
controls the mobile admission. Once a mobile has been admitted for a simulation, it remains admitted for the all the itera-
tions unless there are other reasons to reject it (following steps).
The aim of Dynamic Channel Allocation (DCA) is to reduce interference in order to maximise the usage of the radio
resources. In other words, the DCA tries to find the "best carrier" and the "best timeslots", which when allocated to the
mobiles will optimise the load balance between carriers.
384 AT271_TRG_E6 © Forsk 2009

If a preferred carrier is defined for the service requested by Mi and if it is available at TX i . BestCarrier  TX i M i  = the
carrier preferred for the service. In the case of N-frequency compatible transmitters, Mi can be allocated timeslots over
more than one slave carrier.
Mi is considered unable to connect to the network if no carrier or not enough timeslots have been selected. In this case,
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, then the mobile will be rejected for the reason "DL Load
Saturation" or "Admission Rejection" respectively.
There are four strategies for the DCA available in Atoll. These strategies are described below one by one.
1. 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. The best carrier for a mobile is the one that is least loaded:
BestCarrier  TX i M i  = Carrier DCA

Min  X 
DCA DCA TX i  ic TS  M i  

Where, X = X DL = N Tot – DL if the mobile is connected in the downlink.
TX i  ic TS  M i  
DCA DCA N Tot – UL
And, X = X UL -  X DCA if the mobile is connected in the uplink.
= -----------------------------------------------------
TX i  ic TS  M i   TX i
N Tot – UL + N0
DCA
X is the load increment given by:
Mi Ortho TX i
DCA    1 – f UL    1 – f JD 
X = ------------------------------------------------------------------------
1
1 + -------------
Req
Q UL
E
Req
-----b-
C Req  N t  UL
Where Q UL =  ----
Req Proc
= --------------------- is the uplink required signal quality. The uplink processing gain G UL calcu-
 I  UL Proc
G UL
lated from the service parameters, if no smart antenna is used by the transmitter in the uplink.
If a smart antenna is used by the transmitter in the uplink, the smart antenna gain is taken into account in calcu-
Req
lating Q UL .
Notes:
TX i  ic TS  M i  
• N Tot – UL is described in "Uplink Power Control" on page 386.
TX i  ic TS  M i  
• N Tot – DL is described in "Downlink Power Control" on page 387.
• The carrier is the same in the uplink and in the downlink for mobiles accessing circuit-
switched services.
Timeslot selection by Load: From the selected carrier, Atoll selects the timeslots which are the least loaded and
have enough resource units for the service being accessed by Mi.
2. 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 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 selects the timeslots which have the high-
est numbers of available resource units.
3. 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. Atoll searches for
interfering mobiles within the angle defined by the Angular Step. For example, if you enter an angular step of 15
degrees, Atoll searches for interfering mobiles within 15 degrees to the right and to the left of the served user, and
allocates a different carrier than the ones used by any interfering mobiles found. The best carrier for a mobile is
the one which is not interfered by another mobile in the direction of the mobile Mi.
BestCarrier  TX i M i  = Carrier DoA  Mi   DoA  Mj 
In other words, the direction of arrival for the served user Mi should not be the direction of arrival of an interfering
mobile.
© Forsk 2009 AT271_TRG_E6 385

Timeslot selection by Direction of Arrival: From the selected 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.
4. Sequential
Sequential carrier selection: The DCA allocates carriers to served users Mi in a sequential order.
Sequential timeslot selection: From the selected carrier, Atoll allocates timeslots to served users Mi in a sequen-
tial order.
At the end of the DCA, each admitted mobile has an associated carrier and timeslots. In case of N-frequency mode
compatible transmitters, an admitted mobile can have associated timeslots over more than one slave carrier.
8.3.2.2.3 Uplink Power Control

For each mobile Mi, the uplink power control step calculates the uplink power required to satisfy the required quality level
on the traffic channel, which is defined for the service being accessed by Mi.
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:
Mi TX i  ic TS  M i   Mi TX i  ic TS  Mi  
TX i  ic TS  M i     RSCP TCH – UL TX i  ic TS  Mi     RSCP TCH – UL
E-----b- -  G Proc Div  C- -  G Div
 N t  TCH – UL
= ---------------------------------------------------------------
TX i  ic TS  M i   UL  G UL or  --- 
I TCH – UL
= ---------------------------------------------------------------
TX i  ic TS  Mi   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  .
TX i  ic TS  Mi   TX i  ic TS  Mi   TX i
N Tot – UL = I Tot – UL + N0
Where
TX i  ic TS  Mi   i M Mi
I Tot – UL = RSCP TCH – UL  TX i  ic TS  M i        +
Mj Mi
 RSCP TCH – UL  TX i  ic TS  M i      +
M j  TX i  ic TS  Mi  
Mj  Mi
Mj Mj
  1 –    RSCP TCH – UL  TX i  ic TS  M i    +
M j  TX i  ic TS  Mi  
Mj  Mi
Mj
 RSCP TCH – UL  TX i  ic TS  M i   
M j  TX i  ic TS  Mi  
Mi Mi TX i 
   1 – F JD  and  =  0 Without Useful Signal
Ortho
 =    1 – F UL
 1 Total Noise
The above formula gives the value of I Tot – UL for the uplink connection between Mi and TX i  ic  , taking into account the
interference received from other mobiles, Mj, which are located in the Mi best server coverage area, as well as located in
the coverage areas of other cells. The mobile Mi is the focus, i.e., the mobile that is listened to by the transmitter TX i  ic  .
The four terms comprising I Tot – UL are:
• The useful signal for which the received mobile is the focus (Mi).
• The intra-cell interference for which the best-server is the same for the received mobile Mj and the focus Mi,
TX i  ic  .
• The intra-cell interference due to distortion in the terminal transmission.

• The extra-cell interference for which the best-server for the received mobile Mj is not TX i  ic  .
Mi
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
TX i Mi Mi Model
L Path  L  L  L Body  L Indoor  M Shadowing
Model Mi
LT - and P Req  TX i  ic TS  M i    is the uplink required mobile
= ------------------------------------------------------------------------------------------------------------------------------
TX i Mi
G G
power calculated for the timeslot allocated to Mi . If Mi is an HSDPA user,
Mi Mi
P Req  TX i  ic TS  M i    = 0.1  P Req  TX i  ic TS  M i   
386 AT271_TRG_E6 © Forsk 2009

Model TX i SA TX i SA
In L T , G = G UL and L = L UL are calculated according to the smart antenna modelling method used, for
Mi TX i TX i
P Req  TX i  ic TS  M i    , if a smart antenna is available in the uplink. Otherwise, G and L are read from the main
antenna model.
Notes:
• Interference is updated only for active mobiles on the uplink for circuit- and packet-
switched services. However, if these mobiles are rejected, they are considered in the
number of rejected mobiles.
i M
Calculation of Uplink Required Power ( P Req ):
Then Atoll determines the required uplink power by:
E
Req
-----b-
Mi Mi  N t  TCH – UL
P Req  TX i  ic TS  M i    = P Req  TX i  ic TS  M i     -------------------------------------------
TX i  ic TS  Mi  
-
E
k k–1
-----b-
Req
C----
Mi Mi  I  TCH – UL
or P Req  TX i  ic TS  M i    = P Req  TX i  ic TS  M i     -----------------------------------------
TX i  ic TS  M i  
-
k k–1
C ----
 I  TCH – UL
Mi Mi Mi Mi
And if P Req  TX i  ic TS  M i     P Min then P Req  TX i  ic TS  M i    = P Min
Mi Mi
If P Req  TX i  ic TS  M i     P Max then the mobile Mi is rejected for the reason "Pmob > PmobMax", and
Mi
P Req  TX i  ic TS  M i    is set to 0.
Mi Mi
P Min and P Max are set in the properties of the terminal used by the mobile Mi.
Mi
TX i  ic TS  M i   P Req  TX i  ic TS  M i   
k–1
Where RSCP TCH – UL = -----------------------------------------------------------------------
-
Model
LT
TX i Mi Mi Model
Model L Path  L  L  L Body  L Indoor  M Shadowing Mi
LT - and P Req  TX i  ic TS  M i   
= ------------------------------------------------------------------------------------------------------------------------------ is the uplink required mobile
TX i Mi k–1
G G
power for iteration k - 1 transmitted on the timeslot allocated to Mi.
Mi TX i TX i
P Req  TX i  ic TS  M i    , if a smart antenna is available in the uplink. Otherwise, G and L are read from the main
antenna model.
Note:
• The uplink required powers for mobiles inactive in the uplink accessing circuit- or packet-
switched services are calculated for information only. However, if these mobiles are
rejected, they are considered in the number of rejected mobiles.
8.3.2.2.4 Downlink Power Control

For each mobile 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, which is defined for the service being accessed by Mi.
If the mobile Mi is connected (active or inactive) in the downlink and has a best server TX i  ic  assigned to it, Atoll calcu-
lates the signal quality on the uplink timeslots allocated to Mi by the DCA:
TX i TX i  ic TS  M i   TX i TX i  ic TS  M i  
TX i  ic TS  M i     RSCP TCH – DL TX i  ic TS  M i     RSCP TCH – DL
E-----b- -  G Proc Div  C- -  G Div
 N t  TCH – DL
= ------------------------------------------------------------------
TX i  ic TS  M i   DL  G DL or  ---
I  TCH – DL
= ------------------------------------------------------------------
TX i  ic TS  M i   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  .
TX i  ic TS  M i   TX i  ic TS  M i   Mi
N Tot – DL = I Tot – DL + I IC – DL  ic jc  + I MM  M i M j  + N 0
© Forsk 2009 AT271_TRG_E6 387

Where
TX i  ic TS  Mi   TX i  ic TS  Mi   TX i
I Tot – DL = RSCP Tot – DL  Mi      +
TX i  ic TS  Mi   TX i
 RSCP Tot – DL  Mj   +
M j  TX i  ic TS  Mi  
Mj  Mi
TX i TX i  ic TS  Mi  
 1 –    RSCP Tot – DL  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).
• The intra-cell interference for which the best-server is the same for the received mobile Mj and the focus Mi,
TX i  ic  .
• The intra-cell interference due to distortion in the transmitter.

• The extra-cell interference for which the best-server for the received mobile Mj is not TX i  ic  .
TX j  jc TS  Mi  
 RSCP Tot – DL  Mi 
All TX j
I IC – DL  ic jc  = ---------------------------------------------------------------------------
TX i TX i Mi 
Ortho
 =    1 – F DL
 1 Total Noise
I IC – DL  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink, which is reduced by
the interference reduction factor F IRF  ic jc  defined for the pair (ic, jc).
TX i  ic TS  Mi   TX i  ic TS  M i   TX i  ic TS  Mi  

RSCP Tot – DL = RSCP TCH – DL + RSCP OCCH
TX i  ic TS  M i   TX i  ic TS  M i  
TX i  ic TS  M i   P TCH – DL TX i  ic TS  Mi   P OCCH
k–1
With RSCP TCH – DL - and RSCP OCCH
= ---------------------------------------------- = -----------------------------------
-
Model Model
LT LT
TX i Mi Mi Model
Model L Path  L  L  L Body  L Indoor  M Shadowing TX i  ic TS  M i  
LT - and P TCH – DL
= ------------------------------------------------------------------------------------------------------------------------------ is the downlink traffic power transmit-
TX i Mi k–1
G G
ted on the timeslot allocated to Mi during the iteration k - 1. If Mi is an HSDPA user, P TCH – DL = 0.1  P TCH – DL
In L T , G = G DL and L = L DL are calculated according to the smart antenna modelling method used, for
TX i  ic TS  M i   TX i  ic TS  M i   TX i TX i
P TCH – DL only and not for P OCCH , if a smart antenna is available in the downlink. Otherwise, G and L
are read from the main antenna model.
Mj
 RSCPTCH – UL  Mi 
M
I MM  M i M j  = ---------------------------------------------------------
j - is the interference from each mobile Mj transmitting in the uplink on the same times-
lots as those on which the mobile Mi is receiving in the downlink. 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, and
• The downlink timeslot of Mi (TSMi) is the same as the uplink timeslot of Mj, (TSMj).
The interference received from the mobile Mj at the mobile Mi is calculated using either the free-space propagation model
or the Xia model.
Mj
Mj P TCH – UL
RSCP TCH – UL  M i  = ------------------------
L MM
388 AT271_TRG_E6 © Forsk 2009

 Mi – Mj
 32.4 + 20  Log  F Avg  + 20  Log  d  If d 3m
L MM =  with F Avg being the average frequency in MHz of
 Mi – Mj
 49 + 30  Log  F Avg  + 40  Log  d  If d 3m
the frequency band used by the best server of the mobile Mi, and d is the distance between the mobiles Mi and Mj in km.
TX  ic TS  M i  
Calculation of Downlink Required Power ( P Reqi ):
Then Atoll determines the required downlink power by:
E
Req
-----b-
TX i  ic TS  Mi   TX i  ic TS  M i    N t  TCH – DL
P Req = P Req  -------------------------------------------
TX i  ic TS  M i  
-
E
k k–1
-----b-
Req
C ----
TX i  ic TS  Mi   TX i  ic TS  M i    I  TCH – DL
or P Req = P Req  -----------------------------------------
TX i  ic TS  M i  
-
k k–1
C ----
 I  TCH – DL
TX i  ic TS  M i   Min TX i  ic TS  M i   Min

And if P Req  P TCH – DL  Service  then P Req = P TCH – DL  Service 
TX i  ic TS  M i   Max TX i  ic TS  M i  

If P Req  P TCH – DL  Service  then the mobile Mi is rejected for the reason "Ptch > PtchMax", and P Req
is set to 0.
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.
TX i  ic TS  M i   TX i  ic TS  M i   TX i  ic TS  M i  

Otherwise, the downlink traffic power is incremented P TCH – DL = P TCH – DL + P Req
For each mobile, Atoll also calculates the downlink traffic power for the different values of the Angular Step  Step .
SA
TX i  ic TS  Mi   TX i  ic TS  Mi   G DL
RSCP TCH – DL = RSCP TCH – DL  ----------------------------
SA
Step L DL   Step 
TX i  ic TS  M i  
TX i  ic TS  M i   P Req
k–1
Where RSCP TCH – DL = ----------------------------------------------
-
Model
LT
TX i Mi Mi Model
Model L Path  L  L  L Body  L Indoor  M Shadowing TX i  ic TS  M i  
LT - and P Req
= ------------------------------------------------------------------------------------------------------------------------------ is the downlink traffic power for itera-
TX i Mi k–1
G G
tion k - 1 transmitted on the timeslot allocated to Mi.
TX i  ic TS  M i   TX i TX i
P Req , if a smart antenna is available in the downlink. Otherwise, G and L are read from the main antenna
model.
Note:
• The downlink power for mobiles inactive in the downlink accessing circuit- or packet-
switched services are calculated for information only.
8.3.2.2.5 Uplink Signals Update

This step uses the uplink terminal powers calculated for each timeslot allocated to the mobiles. The Dynamic Channel Allo-
cation allocates timeslots and carriers to all the connected and active mobiles. 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.
This step updates the received signals for all the mobiles Mi interfered in the uplink by the uplink connection between inter-
fering mobiles Mj and their best servers TX j  ic  , only if TX j  ic  contain Mi in their coverage areas. TX i  ic  is the inter-
fered receiver and Mi is the focus, i.e., the mobile that is listened to by the transmitter TX i  ic  .
For each mobile Mi interfered by Mj in the uplink by the connection between Mj and TX j  ic  , Atoll updates
RSCP TCH – UL .
© Forsk 2009 AT271_TRG_E6 389

8.3.2.2.6 Downlink Signals Update

For the first iteration, i.e., k = 0, the downlink traffic powers for all the downlink timeslots are set to 0 Watts. Therefore, for
the first iteration, this step is performed for any downlink timeslot for each mobile Mi that is connected and active.
However, for the following iterations, the downlink signals update step uses the actual downlink traffic powers calculated
for each timeslot and the actual timeslots allocated to the mobiles. The Dynamic Channel Allocation allocates timeslots
and carriers to all the connected and active mobiles. 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.
Therefore, this step is performed for any downlink timeslot for each mobile Mi that is connected and active for the first iter-
ation, and this step is performed for all the downlink timeslots allocated to the mobile Mi on which it is connected and active,
for the following iterations since the DCA has been performed.
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.
TX i  ic TS  M i  
For each mobile interfered by Mi, Atoll updates RSCP TCH – DL
Where TX i  ic  is the transmitter considered and Mi is the focus, i.e., the mobile that is the target for TX i  ic  .
8.3.2.2.7 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.
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. The effective maximum cell traffic power per timeslot is calculated as:
P Max – DL – Eff = P Max – DL  %P Max – DL
TX i  ic TS  M i  
Where P Max – DL is the maximum cell power per timeslot defined per cell, 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, and downlink timeslot TS M ,

i
TX i  ic TS  M i   TX i  ic TS  M i   TX i  ic TS  M i  

P R99 – DL = P TCH – DL + P OCCH
If P R99 – DL  P Max – DL – Eff the mobile with the lowest service priority is rejected for the reason "DL Load Saturation".
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.
TX i  ic TS  M i  
The maximum allowed uplink cell load, X Max – UL , is 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, and uplink timeslot TS M ,

i
If X UL  X Max – UL the mobile with the lowest service priority is rejected for the reason "UL Load Saturation".
The uplink load is given by:

TX i  ic TS  M i   N Tot – UL
X UL - if no smart antenna is used by the transmitter in the uplink.
= -----------------------------------------------------
TX i  ic TS  M i   TX i
N Tot – UL + N0
If a smart antenna is used by the transmitter in the uplink, the smart antenna gain is taken into account in the calculation
of uplink load.
8.3.2.3 HSDPA Part of the Algorithm

The following calculations are made for all HSDPA mobiles (Mi).
8.3.2.3.1 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.
P Tot – DL = P R99 – DL + P HR + P HS – SCCH + P HS – PDSCH
390 AT271_TRG_E6 © Forsk 2009

The HSDPA powers, i.e., the HS-SCCH and HS-PDSCH powers are calculated as follows:
• HS-SCCH Power:
HS-SCCH channels are transmitted on DL traffic timeslots. The maximum number of supported HS-SCCH chan-
nels is defined per cell. 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.
- Dynamic Allocation
TX i  ic 
HS-SCCH power is calculated for E-----c- Req
= Q HS – SCCH  Mobility  so that
 N t  HS – SCCH
P HS – SCCH  P Available – HS – SCCH .
TX i  ic  TX i  ic  TX i  ic  TX i  ic 
Where P Available – HS – SCCH = P Max – DL – Eff – P R99 – DL – P HR is the power available for HS-SCCH in the
TX i  ic  TX i  ic  TX i  ic 
cell TX i  ic  , and P R99 – DL = P TCH – DL + P OCCH .
The effective maximum cell traffic power per timeslot is calculated as:
P Max – DL – Eff = P Max – DL  %P Max – DL . P Max – DL is the maximum power defined per cell, and %P Max – DL
is the maximum allowed downlink load either taken from the properties of each cell or from the simulation prop-
erties if a global value is defined.
TX i  ic 
E -----c-
TX i  ic 
  N Tot – DL –     RSCP HS – SCCH 
TX i  ic   N t  HS – SCCH
P HS – SCCH -  L Model
= ------------------------------------------------------------------------------------------------------------------------------------------------ T
TX i

TX i  ic 
Where N Tot – DL is the downlink total noise calculated in "Downlink Power Control" on page 387,
TX i TX i Mi 
Ortho
 =    1 – F DL
 1 Total Noise
TX i Mi Mi Model
Model L Path  L  L  L Body  L Indoor  M Shadowing TX i  ic 
LT - and P HS – SCCH is the HS-SCCH power calculat-
= ------------------------------------------------------------------------------------------------------------------------------
TX i Mi
G G
ed for the timeslots allocated to Mi.
In L T , G = G DL and L = L DL are calculated according to the smart antenna modelling method
TX i  ic  TX i TX i
used, for P HS – SCCH , if a smart antenna is available in the downlink. Otherwise, G and L are read from
the main antenna model.
• HS-PDSCH Power:
HS-PDSCH channels are transmitted on DL traffic timeslots. 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.
HS-PDSCH power is calculated as follows:
P HS – PDSCH = P Max – DL – Eff – P R99 – DL – P HR – P HS – SCCH
Where P R99 – DL = P TCH – DL + P OCCH . The effective maximum cell traffic power per timeslot is calculated as:
P Max – DL – Eff = P Max – DL  %P Max – DL . P Max – DL is the maximum power defined per cell, and %P Max – DL
is the maximum allowed downlink load either taken from the properties of each cell or from the simulation prop-
erties if a global value is defined.
The HS-SICH power is calculated as follows:

• HS-SICH Power:
HS-SICH channels can be transmitted on any UL traffic timeslot. The maximum number of supported HS-SICH
channels is defined per cell. 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.
© Forsk 2009 AT271_TRG_E6 391

Ec TX i  ic 
Mi TX i  ic 
HS-SICH power is calculated for  ------
Req
= Q HS – SICH  Mobility  so that P HS – SICH  P Max – HS – SICH
 N t  HS – SICH
Mi Mi
and P HS – SICH  P Max – HS – SICH .
TX i  ic 
E -----c-
TX i  ic 
  N Tot – UL –     RSCP HS – SICH 
Mi Mi
Mi  N t  HS – SICH
P HS – SICH -  L Model
= ------------------------------------------------------------------------------------------------------------------------------------------ T
Mi

TX i  ic 
Where N Tot – UL is the uplink total noise calculated in "Uplink Power Control" on page 386,
Mi Mi TX i 
Ortho
 =    1 – F UL
 1 Total Noise
TX i Mi Mi Model
Model L Path  L  L  L Body  L Indoor  M Shadowing Mi
LT - and P HS – SICH is the HS-SICH power calculated
= ------------------------------------------------------------------------------------------------------------------------------
TX i Mi
G G
for the timeslots allocated to Mi.
In L T , G = G UL and L = L UL are calculated according to the smart antenna modelling method
Mi TX i TX i
used, for P HS – SICH , if a smart antenna is available in the uplink. Otherwise, G and L are read from
the main antenna model.
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.
P Tot – DL = P R99 – DL + P HR + P HS – SCCH + P HS – PDSCH
8.3.2.3.2 Connection Status and Number of HSDPA Users

HSDPA users cannot receive HS-SCCH and HS-PDSCH powers simultaneously. HS-PDSCH arrives 3 timeslots after the
HS-SCCH. HS-SICH is 9 timeslots afte the HS-PDSCH. Atoll assumes that an active HSDPA user has the same proba-
bility of receiving HS-SCCH and HS-PDSCH, and transmitting HS-SICH because their occurence is equally likely. There-
fore, each HSDPA user is assigned a sub-connection status randomly. 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.
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. Each HSDPA user consumes one HS-SCCH and HS-SICH channels. Therefore, at a given
instance, the number of connected HSDPA users cannot exceed the number of HS-SCCH and HS-SICH channels per
cell. The maximum number of HSDPA users ( n Max ) corresponds to the maximum number of HSDPA users that the cell
can support.
8.3.2.3.3 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. Each cell is able to manage a maximum number of HS-SCCH channels,
n HS – SCCH . During the R99 part, the DCA provides a DL timeslot with one SF16 resource unit that has the downlink Ec/
Nt higher than the required quality. If no cell with such a resource unit is available, the user is rejected.
HS-SICH
HS-SICH admission control is performed for active HSDPA users connected to A-DCH bearers on the uplink and having
an HS-SICH sub-connection status. Each cell is able to manage a maximum number of HS-SICH channels, n HS – SICH .
During the R99 part, the DCA provides an UL timeslot with one SF16 resource unit that has the uplink Ec/Nt higher than
the required quality. If no cell with such a resource unit is available, the user is rejected.
392 AT271_TRG_E6 © Forsk 2009

HS-PDSCH
Scheduling is performed for active HSDPA users connected to A-DCH bearers on the downlink and having an HS-PDSCH
sub-connection status. The scheduling is performed as follows:
1. Each HS-PDSCH user is considered as the only served user. The scheduler allocates the best available HSDPA
bearer to each user. The best available HSDPA bearer is selected depending on the user’s Ec/Nt. If no bearer can
be allocated due to low Ec/Nt, the user is rejected for the reason "HSDPA Scheduler Saturation".
The required HS-PDSCH Ec/Nt value is read from receiver equipment properties. For each bearer, Atoll checks
that the Ec/Nt reaches the quality target. HS-PDSCH Ec/Nt is calculated by taking into account all intra and extra
cells interferences.
2. The scheduler sorts the HS-PDSCH users to whom bearers have been assigned in the order of decreasing RLC
peak rates. If two users have the same bearer, the user with the higher Ec/Nt has the higher rank.
3. The scheduler considers the group of HS-PDSCH users to whom bearers, HS-SCCH, and HS-SICH have been
assigned. The number of HS-PDSCH users cannot exceed the maximum number of HSDPA users ( n Max ) sup-
ported by the cell. If there are enough HSDPA power and resource units available in order to obtain a HSDPA
bearer, the users will be connected. Otherwise, they will be delayed and their connection status will be “HSDPA
Delayed”.
4. Other HS-PDSCH users will be rejected for the reason "HSDPA Scheduler Saturation".
For N-frequency mode compatible transmitters, the resource units available in the master and slave carriers can be
shared, i.e., a mobile can be connected to timeslots belonging more than one carrier.
8.3.2.3.4 HSDPA Dynamic Channel Allocation

For each mobile connected to the A-DPCH bearer:
1. Atoll selects the HSDPA bearers that match to the mobile terminal and UE category parameters.
2. For each bearer supported by a mobile:
a. The scheduler searches for the best collection of "n" ordered timeslots that can provide enough resource units
to support the service, and whose Ec/Nt is better than the minimum required and enough to reach the bearer’s
resource unit requirements. The best is determined by applying the R99 Dynamic Channel Allocation algo-
rithm.
b. The scheduler calculates the HS-PDSCH Ec/Nt for each timeslot of the best collection. The Ec/Nt value asso-
ciated with the mobile-bearer pair is the worst one of all selected timeslots.
c. If the scheduler is unable to find a satisfactory timeslot collection, the bearer is removed from the list of sup-
ported bearers.
3. The mobile is connected to the supported bearer having the highest RLC peak rate. If two bearers have the same
RLC peak rate, the best one is the one with the highest Ec/Nt.
8.3.2.3.5 Ressource Unit Saturation

For each time slot, a minimum and maximum number of resource units for HSDPA users are defined in the cell properties.
Atoll dynamically allocates the required number of codes respecting these limitations. The minimum number of HSDPA
codes is excluded from the set of codes available for R99 users. The scheduler checks if enough codes are available for
the selected HSDPA bearer (taking into account the maximum number of HSDPA codes). If not, the scheduler allocates
a lower HSDPA bearer which needs fewer codes. If there are no more resource units available for the lowest HSDPA
bearer, the user will be delayed or rejected.
8.3.2.4 Convergence Criteria

The convergence criteria are evaluated for each iteration and can be written as follows:
Max TX i  ic TS  M i  
 DL = Int  P   100
 All TX i Err 
 TX i  ic TS  M i   TX i  ic TS  M i   
 Max N Tot – UL – N Tot – UL 
 UL = Int  k k – 1  100
---------------------------------------------------------------------------------------------- 
 All TX i TX i  ic TS  M i   
 N Tot – UL 
k
TX i  ic TS  M i  
Where, P Err is given by:
TX i  ic TS  Mi   TX i  ic TS  M i  
TX i  ic TS  M i   Max P Rec – P Rec
 Step k  Step k – 1
P Err = - with smart antennas.
-----------------------------------------------------------------------------------------------------------------------
0   Step  360 TX i  ic  TS  Mi  
P Rec
 Step k
TX i  ic TS  Mi   TX i  ic TS  M i  
TX i  ic TS  M i   P Rec – P Rec
P Err k k – 1 without smart antennas.
= --------------------------------------------------------------------------------------------
-
P Rec
k
© Forsk 2009 AT271_TRG_E6 393

Atoll stops the simulations in the following cases:
• Convergence: Between two successive iterations,  DL and  UL are less than or equal to their respective thresh-
olds (defined when creating a simulation).
Example: Let us assume that the maximum number of iterations is 100, and the UL and DL convergence thresh-
olds are set to 5 %. If  DL  5 and  UL  5 between the 4th and the 5th iteration, Atoll stops the algorithm after
the 5th iteration. The simulation has converged.
• Divergence: After 30 iterations,  DL and/or  UL are still higher than their respective thresholds and from the 30th
iteration,  DL and/or  UL do not decrease during the next 15 successive iterations.
Examples: Let us assume that the maximum number of iterations is 100, and the UL and DL convergence thresh-
olds are set to 5 %.
a. After the 30th iteration,  DL and/or  DL equal 100 and do not decrease during the next 15 successive itera-
tions. Atoll stops the algorithm at the 46th iteration. The simulation has not converged.
b. After the 30th iteration,  DL and/or  UL equal 80, they start decreasing slowly until the 40th iteration (without
going under the thresholds) and then, do not change during 15 successive iterations. Atoll stops the algorithm
at the 56th iteration without converging.
• Last Iteration: If  DL and/or  UL are still much higher than their respective thresholds after the last iteration, the
simulation has not converged. If  DL and  UL are lower than their respective thresholds, the simulation has
reached convergence.
8.4 TD-SCDMA Prediction Studies

For each TBC transmitter, TXi, Atoll determines the value of the selected parameter on each studied pixel inside the TXi
calculation area. Each pixel within the TXi calculation area is considered a probe receiver.

• 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. Coverage prediction display resolution is independent of the path loss matrix and
geographic data resolutions, and can be different for each coverage prediction. Coverage predictions are calculated using
bilinear interpolation of multi-resolution path loss matrices (similar to the evaluation of site altitudes).
8.4.1 P-CCPCH Reception Analysis (Eb/Nt) or (C/I)

Eb C
These coverage predictions calculate and display the Eb/Nt or C/I on the P-CCPCH,  ------ or  ---- . The
 N t  P – CCPCH  I  P – CCPCH
coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The
best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with the high-
coverage predictions are calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there
will not be any pixels covered by this transmitter. If you perform these coverage predictions for all the carriers, Atoll calcu-
lates the Eb/Nt or C/I considering:
TX i  ic 
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and
E-----b-
C
 Q P – CCPCH or  ----
Req Req
 Q P – CCPCH are covered and coloured according to the selected display
option.
TX i TX i  ic  TX TX  ic 
E b TX i  ic    RSCP P – CCPCH C
TX i  ic  i
  RSCP P – CCPCH
i
Where  ------ -  G Proc

= ----------------------------------------------------------- and ---- = -----------------------------------------------------------
-
 N t  P – CCPCH TX i  ic  P – CCPCH  I  P – CCPCH TX i  ic 
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
394 AT271_TRG_E6 © Forsk 2009

Where
TX i  ic  TX i  ic  TX i TX i  ic  TX i
I Intra – DL = RSCP P – CCPCH     + RSCP OCCH – TS0  
TX i TX i 
 and  =  0 Without Useful Signal
Ortho Term
With  =    1 – F DL    1 – F JD
 1 Total Noise
TX i  ic  TX j  ic  TX j  ic 
I Extra – DL =   RSCPP – CCPCH + RSCPOCCH – TS0 
ji
  RSCPP – CCPCH + RSCPOCCH – TS0 
TX
I IC – DL  ic jc  = -----------------------------------------------------------------------------------------------------------
j -
TX i  ic 
TX i  ic  P OCCH – TS0
RSCP OCCH – TS0 = -------------------------------
LT
TX i Term Eb  Nt
L T = -------------------------------------------------------------------------------------------------------------------------------------
-
TX i Term
G G
TX i Term
 and N 0 are defined in "Definitions and Formulas" on page 365.
Coverage Display
• 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 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). For each
Eb TX  ic  TX  ic 
i
C i
layer, a pixel is covered if  ------  Threshold or  ----  Threshold . Each layer is assigned a
• 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. The pixel colour
depends on the Eb/Nt or C/I margin 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 prop-
TX i  ic 
erties). For each layer, a pixel is covered if E-----b- Req Eb  Nt
– Q P – CCPCH  M P – CCPCH or
 N t  P – CCPCH
TX i  ic 
C
---- Req CI
– Q P – CCPCH  M P – CCPCH . Each layer is assigned a colour and displayed with intersections
 I  P – CCPCH
between layers.
Atoll calculates the cell edge coverage probability on each pixel of the TX i  ic  best server coverage area. 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. There are as many layers as thresholds defined in the Display
E b TX i  ic 
tab (Prediction properties). For each layer, a pixel is covered if  ------
Req
 Q P – CCPCH or
 N t  P – CCPCH
CECP
TX i  ic 
C
---- Req
 Q P – CCPCH . Each layer is assigned a colour and displayed with intersections between
 I  P – CCPCH
CECP
layers.
8.4.2 DwPCH Reception Analysis (C/I)

C
This coverage prediction calculates and displays the C/I on the DwPCH,  ---- . The coverage prediction is calculated
 I  DwPCH
for a given set of a terminal type, a mobility type, a service, a carrier, and for DwPTS. The best server for the coverage
© Forsk 2009 AT271_TRG_E6 395

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. Afterwards, the coverage prediction is calculated for
the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this trans-
mitter. If you perform this coverage prediction for all the carriers, Atoll calculates the C/I considering:
C
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and  ----
Req
 Q DwPCH
 I  DwPCH
are covered and coloured according to the selected display option.
TX i  ic 
C   RSCP DwPCH
Where  ---- = ---------------------------------------------------
-
 I  DwPCH TX i  ic 
N Tot – DL
TX i  ic 
TX i  ic  P DwPCH
RSCP DwPCH = --------------------
-
LT
The downlink total noise is calculated as follows:

Where
TX i  ic  TX i  ic  TX i
I Intra – DL = RSCP DwPCH    
TX i TX i 
 and  =  0 Without Useful Signal
Ortho Term
With  =    1 – F DL    1 – F JD
 1 Total Noise
TX i  ic  TX j  ic 
I Extra – DL =   RSCPDwPCH 
ji
TX j  jc 
  RSCPDwPCH 
TX
I IC – DL  ic jc  = -----------------------------------------------
j -
TX i Term Eb  Nt
L T = -------------------------------------------------------------------------------------------------------------------------------------
-
TX i Term
G G
TX i Term
 and N 0 are defined in "Definitions and Formulas" on page 365.
Coverage Display
• C/I (dB)
Atoll calculates the C/I on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on the
C/I level. 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). For each layer, a pixel
TX i  ic 
C
is covered if  ----  Threshold . Each layer is assigned a colour and displayed with intersections between
 I  DwPCH
layers.
• C/I Margin (dB)

Atoll calculates the C/I margin on each pixel of the TX i  ic  best server coverage area. The pixel colour depends
on the C/I margin value. Coverage consists of several independent layers whose visibility in the workspace can
TX i  ic 
C
layer, a pixel is covered if  ----
Req CI
– Q DwPCH  M DwPCH . Each layer is assigned a colour and displayed with
 I  DwPCH
396 AT271_TRG_E6 © Forsk 2009

TX i  ic 
C
tab (Prediction properties). For each layer, a pixel is covered if  ----
Req
 Q DwPCH . Each layer is
 I  DwPCH
CECP
8.4.3 Downlink TCH RSCP Coverage

This coverage prediction calculates and displays the RSCP for the downlink traffic channel, RSCP TCH – DL . The coverage
prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for a downlink timeslot.
RSCP considering:
TX i  ic 
TX i  ic  Req
RSCP TCH – DL  RSCP TCH – DL  Service Mobility  are covered and coloured according to the selected display option.
TX i  ic 
Where RSCP TCH – DL is given by:
Max
TX i  ic  P TCH – DL  Service 
RSCP TCH – DL = ----------------------------------------------------
Model
LT
TX i Term Model
Model L Path  L  L  L Body  L Indoor  M Shadowing Max
LT - and P TCH – DL  Service  is the maximum downlink traffic
= -------------------------------------------------------------------------------------------------------------------------------------
TX i Term
G G
power defined for the selected service.
Max TX i TX i
P TCH – DL  Service  , if a smart antenna is available in the downlink. Otherwise, G and L are read from the main
antenna model.
Coverage Display
• DL TCH RSCP (dBm)
Atoll calculates the DL TCH RSCP on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the RSCP level. 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). For
TX i  ic 
each layer, a pixel is covered if RSCP TCH – DL  Threshold . Each layer is assigned a colour and displayed with

Atoll calculates the RSCP margin on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the RSCP margin value. Coverage consists of several independent layers whose visibility in the work-
space can be managed. There are as many layers as thresholds defined in the Display tab (Prediction properties).
For each layer, a pixel is covered if RSCP TCH – DL – RSCP TCH – DL  Service Mobility   M TCH – DL . Each layer

tab (Prediction properties). For each layer, a pixel is covered if
TX i  ic  Req
RSCP TCH – DL  RSCP TCH – DL  Service Mobility  . Each layer is assigned a colour and displayed with
CECP
© Forsk 2009 AT271_TRG_E6 397

8.4.4 Uplink TCH RSCP Coverage

This coverage prediction calculates and displays the RSCP for the uplink traffic channel, RSCP TCH – UL . The coverage
prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for an uplink timeslot.
RSCP considering:
TX i  ic 
TX i  ic  Req
RSCP TCH – UL  RSCP TCH – UL  Service Mobility  are covered and coloured according to the selected display option.
TX i  ic 
Where RSCP TCH – UL is given by:
Term
TX i  ic  P Max
RSCP TCH – UL = ----------------
-
Model
LT
TX i Term Model
Model L Path  L  L  L Body  L Indoor  M Shadowing Term
LT - and P Max is the maximum uplink traffic power defined for
= -------------------------------------------------------------------------------------------------------------------------------------
TX i Term
G G
the selected terminal.
Term TX i TX i
P Max , if a smart antenna is available in the uplink. Otherwise, G and L are read from the main antenna model.
Coverage Display
• UL TCH RSCP (dBm)
Atoll calculates the UL TCH RSCP on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the RSCP level. 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). For
TX i  ic 
each layer, a pixel is covered if RSCP TCH – UL  Threshold . Each layer is assigned a colour and displayed with

Atoll calculates the RSCP margin on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the RSCP margin value. Coverage consists of several independent layers whose visibility in the work-
For each layer, a pixel is covered if RSCP TCH – UL – RSCP TCH – UL  Service Mobility   M TCH – UL . Each layer

tab (Prediction properties). For each layer, a pixel is covered if
TX i  ic  Req
RSCP TCH – UL  RSCP TCH – UL  Service Mobility  . Each layer is assigned a colour and displayed with
CECP
8.4.5 Downlink Total Noise

This coverage prediction calculates and displays the total noise on the downlink, N Tot – DL . The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for a downlink timeslot. 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. Afterwards, the coverage predic-
tion is calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels
398 AT271_TRG_E6 © Forsk 2009

covered by this transmitter. If you perform this coverage prediction for all the carriers, Atoll calculates the downlink noise
for all the carriers but keeps the worst case value, i.e., the most interfered carrier. You can choose to display the minimum,
the maximum, or the average total noise values from among the values calculated for all the carriers. Pixels are covered
and coloured according to the total downlink noise thresholds defined in the display options.

Term
Total downlink noise is given by: N Tot – DL =  RSCP TCH – DL + RSCP OCCH  + N 0
All TX, c, and TS
P TCH – DL P OCCH
With RSCP TCH – DL = -----------------------
- and RSCP OCCH = -----------------
-
Model Model
LT LT
TX i Term Model
Model L Path  L  L  L Body  L Indoor  M Shadowing
LT - and P TCH – DL and P TCH – DL are respectively the down-
= -------------------------------------------------------------------------------------------------------------------------------------
TX i Term
G G
link traffic power and the other common control channel power for the selected timeslot.
Max TX i TX i
antenna model.
8.4.6 Downlink Service Area (Eb/Nt) or (C/I)

Eb
These coverage predictions calculate and display the Eb/Nt or C/I on the downlink traffic channel,  ------ or
C
---- . The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier,
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, or from the master carrier in case of N-frequency mode compat-
ible transmitters. Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier does
not exist on a transmitter, there will not be any pixels covered by this transmitter. If you perform these coverage predictions
for all the carriers, Atoll calculates the Eb/Nt or C/I considering:
TX i  ic 
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  ,
TX i  ic  E b TX i  ic  C
TX i  ic 
RSCP TCH – DL  RSCP TCH – DL  Service Mobility  , and  ------  Q TCH – DL or  ----
Req Req Req
 Q TCH – DL are
covered and coloured according to the selected display option.
TX i TX i  ic  TX i TX i  ic 
E b TX i  ic    RSCP TCH – DL  C-
TX i  ic    RSCP TCH – DL
Where  ------ -  G Proc Div Div
= ------------------------------------------------------
TX i  ic  DL  G DL and  --- 
I TCH – DL
= -------------------------------------------------------
TX i  ic 
 G DL
Max
TX i  ic  P TCH – DL  Service 
With RSCP TCH – DL = ----------------------------------------------------
LT
TX i Term  Eb  Nt  DL
L Path  L  L
 Eb  Nt  DL  L Body  L Indoor  M Shadowing Max
LT - and P TCH – DL  Service  is the maximum downlink traf-
= -------------------------------------------------------------------------------------------------------------------------------------
TX i Term
G G
fic power defined for the selected service.
 Eb  Nt  DL TX i SA TX i SA
Max TX i TX i
antenna model.
Where
TX i  ic   TX i Ortho Term TX i  TX i  ic  TX i  ic 
I Intra – DL =     1 – F DL    1 – F JD  +  1 –      RSCP TCH – DL + RSCP OCCH 
 
TX i  ic 
TX i  ic  P OCCH
With RSCP OCCH = -------------------------
-
LT
© Forsk 2009 AT271_TRG_E6 399

TX i  ic  TX j  ic  TX j  ic 
I Extra – DL =   RSCPTCH – DL + RSCPOCCH 
ji
TX
I IC – DL  ic jc  = ------------------------------------------------------------------------------------------
j -
Coverage Display
• Max Eb/Nt or Max C/I (dB)
E b TX i  ic  C
TX i  ic 
• 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. The pixel colour
depends on the effective Eb/Nt or C/I level. Coverage consists of several independent layers whose visibility in the
 E b TX i  ic  
Min   ------
Req
erties). For each layer, a pixel is covered if  Q TCH – DL  Threshold or
  N t  TCH – DL 
TX i  ic 
C
Min   ----   Threshold . Each layer is assigned a colour and displayed with intersections
Req
Q
  I  TCH – DL TCH – DL
between layers.

TX i  ic 
– Q TCH – DL  M TCH – DL or
TX i  ic 
C
---- Req CI
– Q TCH – DL  M TCH – DL . Each layer is assigned a colour and displayed with intersections between
layers.
• Required Power (dBm)

Atoll calculates the downlink required power on each pixel of the TX i  ic  best server coverage area. The pixel
colour depends on the required power level. 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
Req
properties). For each layer, a pixel is covered if P TCH – DL  Threshold , where
Req Req
Req Q TCH – DL Max Req Q TCH – DL Max
P TCH – DL = --------------------------------
TX i  ic 
-  P TCH – DL  Service  or P TCH – DL = ------------------------------
TX i  ic 
 P TCH – DL  Service  . Each layer
 E b
------ C
----
• Required Power Margin (dB)

Atoll calculates the downlink required power margin on each pixel of the TX i  ic  best server coverage area. The
pixel colour depends on the required power margin 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
Req Max
(Prediction properties). For each layer, a pixel is covered if P TCH – DL – P TCH – DL  Service   M arg in , where
Req Req
Req Q TCH – DL Q TCH – DL
-  P Max
P TCH – DL = --------------------------------
Req
TCH – DL  Service  or P TCH – DL = ------------------------------
Max
 P TCH – DL  Service  . Each layer
E TX i  ic  TX i  ic 
 -----b- C
 ----
400 AT271_TRG_E6 © Forsk 2009


E b TX i  ic 
Req
 Q TCH – DL or
CECP
TX i  ic 
C
---- Req
 Q TCH – DL . Each layer is assigned a colour and displayed with intersections between layers.
CECP
8.4.7 Uplink Service Area (Eb/Nt) or (C/I)

Eb
These coverage predictions calculate and display the Eb/Nt or C/I on the uplink traffic channel,  ------ or
C
---- . The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier,
and for an uplink 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, or from the master carrier in case of N-frequency mode compatible
transmitters. Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier does not
exist on a transmitter, there will not be any pixels covered by this transmitter. If you perform these coverage predictions
for all the carriers, Atoll calculates the Eb/Nt or C/I considering:
TX i  ic 
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  ,
TX i  ic  Eb TX  ic  TX  ic 
i
C i
RSCP TCH – UL  RSCP TCH – UL  Service Mobility  , and  ------  Q TCH – UL or  ----
Req Req Req
 Q TCH – UL are
Term TX i  ic  Term TX i  ic 
E b TX i  ic    RSCP TCH – UL  C-
Where  ------ -  G Proc Div
= ----------------------------------------------------------- UL  G UL and  --- -  G Div
= -----------------------------------------------------------
 N t  TCH – UL TX i  ic  I  TCH – UL TX i  ic  UL
N Tot – UL N Tot – UL
Term Req Req

TX i  ic  P Max Term Term Q TCH – UL Q TCH – UL
- or P Term
- and P Req = P Max  --------------------------------
With RSCP TCH – UL = -------------------------
Term
Req = P Max  ------------------------------
 Eb  Nt  UL E TX i  ic  TX i  ic 
LT  -----b- C ----
TX i Term  Eb  Nt  UL
L Path  L  L
 Eb  Nt  UL  L Body  L Indoor  M Shadowing Term
LT - and P Max is the maximum power defined for the
= -------------------------------------------------------------------------------------------------------------------------------------
TX i Term
G G
selected terminal.
 Eb  Nt  UL TX i SA TX i SA
Term TX i TX i
P Max , if a smart antenna is available in the uplink. Otherwise, G and L are read from the main antenna model.
Coverage Display
• Max Eb/Nt or Max C/I (dB)
E b TX i  ic  C
TX i  ic 
• 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. The pixel colour
depends on the effective Eb/Nt or C/I level. Coverage consists of several independent layers whose visibility in the
© Forsk 2009 AT271_TRG_E6 401

 E b TX i  ic  
Min   ------
Req
erties). For each layer, a pixel is covered if  Q TCH – UL  Threshold or
  N t  TCH – UL 
TX i  ic 
C
Min   ----   Threshold . Each layer is assigned a colour and displayed with intersections
Req
Q
  I  TCH – UL TCH – UL
between layers.

TX i  ic 
– Q TCH – UL  M TCH – UL or
TX i  ic 
C
---- Req CI
– Q TCH – UL  M TCH – UL . Each layer is assigned a colour and displayed with intersections between
layers.
• Required Power (dBm)

Atoll calculates the uplink required power on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the required power level. Coverage consists of several independent layers whose visibility in the work-
Req
Term Term Q TCH – UL
-  P Term
For each layer, a pixel is covered if P Req  Threshold , where P Req = -------------------------------- Max or
E TX i  ic 
 -----b-
Req
Term Q TCH – UL Term
P Req = ------------------------------
TX i  ic 
 P Max . Each layer is assigned a colour and displayed with intersections between layers.
C
 ----
• Required Power Margin (dB)

Atoll calculates the uplink required power margin on each pixel of the TX i  ic  best server coverage area. The
pixel colour depends on the required power margin 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
Term Term
(Prediction properties). For each layer, a pixel is covered if P Req – P Max  M arg in , where
Req Req
Term Q TCH – UL Term Term Q TCH – UL Term
P Req = --------------------------------
TX i  ic 
-  P Max or P Req = ------------------------------
TX i  ic 
 P Max . Each layer is assigned a colour and displayed
 E b
------ C
----

Eb TX i  ic 
Req
 Q TCH – UL or
CECP
TX i  ic 
C
---- Req
 Q TCH – UL . Each layer is assigned a colour and displayed with intersections between layers.
CECP
8.4.8 Effective Service Area (Eb/Nt) or (C/I)

These coverage predictions consist of pixels covered by the both the uplink and the downlink service areas. These cover-
Eb C
age predictions calculate the Eb/Nt or C/I on the downlink and uplink traffic channels,  ------ or  ---- and
E-----b- C
or  ---- , and display the pixels where both downlink and uplink Eb/Nt or C/I are above the required
quality thresholds.
The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for all
the 6 timeslots. 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.
Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier does not exist on a trans-
402 AT271_TRG_E6 © Forsk 2009

mitter, there will not be any pixels covered by this transmitter. If you perform these coverage predictions for all the carriers,
Atoll calculates the Eb/Nt or C/I considering:
The pixels in the TX i  ic  coverage area are covered and coloured according to the selected display option if all the follow-
ing conditions are satisfied:
TX i  ic 
• RSCP P – CCPCH  TAdd P – CCPCH  Mobility 
TX i  ic  Req
• RSCP TCH – DL  RSCP TCH – DL  Service Mobility 
TX i  ic  Req
• RSCP TCH – UL  RSCP TCH – UL  Service Mobility 
• E-----b- Req C
 Q TCH – DL or  ----
Req
 Q TCH – DL for any of the 6 timeslots
• E-----b- Req C
 Q TCH – UL or  ----
Req
 Q TCH – UL for any of the 6 timeslots
8.4.9 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. The timeslot configuration of each cell defines the direction of the link at any given
instance. During each subframe, the direction of the link changes twice (downlink to uplink, and then uplink to downlink).
These transitions are referred to as switching points.
The coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and a timeslot.
RSCP considering:
The mobility, service, and terminal are used to calculate the best server coverage of the interfered cell.
Assuming that a transmitter TX j is interfering a studied transmitter TX i on a timeslot, on the same carrier ic or on another
carrier jc, the cell to cell interference is given by:
TX j  ic  TX j  ic 
  RSCPTCH – DL + RSCPOCCH  + ------------------------------------------------------------------------------------------
TX j
I C2C  TX i TX j  = -
TX j
TX j  ic  TX j  jc 
TX j  ic  P TCH – DL    TX j  jc  P TCH – DL   
Where RSCP TCH – DL = --------------------------------
- and RSCP TCH – DL = --------------------------------
- using a smart antenna, and
LT LT
TX j  ic  TX j TX j  jc  TX j
TX j  ic  P TCH – DL G Ant TX j  jc  P TCH – DL G Ant
RSCP TCH – DL = ------------------------  -----------
- and RSCP TCH -  ------------ otherwise.
– DL = -----------------------
LT TX j LT TX j
L Ant L Ant
TX j  ic  TX j TX j  jc  TX j
TX j  ic  P OCCH G Ant TX j  jc  P OCCH G Ant
RSCP OCCH = ------------------  -----------
- and RSCP OCCH = ------------------  -----------
- otherwise.
LT TX j LT TX j
L Ant L Ant
ITU526 – 5 TX j TX i
L T = L Path  L TX  L RX
ITU526 – 5
L Path is the path loss calculated using the ITU526-5 propagation model without antenna loss.
 is the angle for the smart antenna pattern.
TX j
L Ant is the main antenna attenuation.
TX j
G Ant is the main antenna gain.
Atoll calculates the cell to cell interference on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the cell to cell interference level. Coverage consists of several independent layers whose visibility in the work-
space can be managed. There are as many layers as thresholds defined in the Display tab (Prediction properties). For
© Forsk 2009 AT271_TRG_E6 403

each layer, a pixel is covered if I C2C  TX i TX j   Threshold . Each layer is assigned a colour and displayed with intersec-
8.4.10 UpPCH Interference

UpPCH is usually carried by the UpPTS timeslot. However, if the interference on UpPTS is high, from unsynchronised
DwPTS or TS0 timeslots of other cells, it is possible to shift the UpPCH to TS1. This is called UpPCH shifting. If some cells
in a network use UpPCH shifting, you can use this coverage prediction to study the interference on the shifted UpPCH of
these cells from other cells. The interference from other cells is in this case generated by the traffic on the TS1 of interfering
cells.
This coverage prediction calculates and displays the uplink interference on the TS1, I TS1 – UL . The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS1. The best server for the cover-
age 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. Afterwards, the coverage prediction is calculated
for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this
transmitter. If you perform this coverage prediction for all the carriers, Atoll calculates the interference for all the carriers
but keeps the worst case value, i.e., the most interfered carrier. You can choose to display the minimum, the maximum,
or the average total noise. The coverage prediction is calculated using the main antenna.
Pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and I TS1 – UL  Threshold are
The uplink interference on TS1 is calculated from the uplink load calculated in the simulations or manually defiend for the
TS1.
TX i  ic 
TX i  ic  TX i X TS1 – UL
The uplink interference on TS1 is given by: I TS1 – UL = N 0  -------------------------------------
TX i  ic 
 1 – X TS1 – UL 
8.4.11 HSDPA Coverage

This coverage prediction calculates and displays the RLC peak rate or the MAC rate per pixel covered by HSDPA cells.
The coverage prediction is calculated for a given set of an HSDPA terminal type, a mobility type, an HSDPA service, a
carrier, and for all downlink timeslots. 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 compat-
ible transmitters. Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier does
not exist on a transmitter or if it does not support HSDPA, there will not be any pixels covered by this transmitter. If you
perform these coverage predictions for all the carriers, Atoll calculates the RLC or MAC rate considering:
The pixels in the TX i  ic  coverage area are covered and coloured if:
TX i  ic 
• RSCP P – CCPCH  TAdd P – CCPCH  Mobility  ,
TX i  ic 
• E
------C- Req
 Q HS – PDSCH , and
 N t  HS – PDSCH
TX i  ic 
• E
------C- is enough to select a bearer for the pixels.
 N t  HS – PDSCH
For more information on HSDPA bearer selection, see "HSDPA Part of the Algorithm" on page 390.
Coverage Display
• RLC Peak Rate
After selecting the bearer, Atoll reads the corresponding RLC peak rate. This is the highest rate that the bearer
can provide on each pixel. The pixel colour depends on the RLC peak rate. Coverage consists of several inde-
pendent 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.
• MAC Rate
MAC
Atoll displays the MAC rate ( R DL ) provided on each pixel. The MAC rate is calculated as follows:
MAC
R DL = S Block  500
404 AT271_TRG_E6 © Forsk 2009

Where, S Block is the transport block size (in kbits) of the selected HSDPA bearer; it is defined for each HSDPA
bearer in the related table. The value 500 corresponds to the number of blocks per second (there are 4 blocks per
2000
TTI and 2000 TTI in one second, i.e ------------- blocks per second).
4
The pixel colour depends on the MAC rate. Coverage consists of several independent layers whose visibility in the
erties). Each layer is assigned a colour and displayed with intersections between layers.
8.5 Smart Antenna Modelling

Atoll calculates the smart antenna gains and losses in the direction of a user during the simulations, and in the direction
of each pixel in coverage predictions. During simulations, Atoll determines the gains and losses using the smart antenna
models. In coverage predictions, 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.
If a smart antenna model is only downlink or only uplink, the other direction uses the main antenna gain and losses for
calculations. Therefore,
• If a smart antenna is available on the downlink and uplink:
TX SA TX SA TX SA TX SA
G UL = G UL , L UL = L UL and G DL = G DL , L DL = L DL
• If a smart antenna is available on the downlink only:

TX SA TX SA TX TX TX TX
G DL = G DL , L DL = L DL and G UL = G Ant , L UL = L = L Total – UL
• If a smart antenna is available on the uplink only:

TX SA TX SA TX TX TX TX
G UL = G UL , L UL = L UL and G DL = G Ant , L DL = L = L Total – DL
• If no smart antenna equipment is defined:

TX TX TX TX TX TX TX
G DL = G UL = G Ant , L UL = L = L Total – UL , and L DL = L = L Total – DL
8.5.1 Modelling in Simulations

8.5.1.1 Grid of Beams Modelling
A grid-of-beams smart antenna, called GOB, consists of more than one directional antenna pattern (beam) in different
directions. Each beam of a GOB has a different azimuth so that the GOB as a whole covers an entire sector. During the
simulations, Atoll determines the most suitable beam from the GOB for each user served by the smart antenna. The most
suitable beam (best beam) is the one which provides the highest gain towards the served user:
Beam Best = Beam H V

Max  G Beam – L Beam – L Beam 
H V
Where G Beam , L Beam , and L Beam are the gains, horizontal, and vertical attenuations of the beams of the GOB. In words,
the best beam is the one among all the beams of a GOB that has the highest difference between gain, and horizontal and
SA SA SA SA
vertical attenuations. The gains and losses of the GOB ( G DL , G UL , L DL , and L UL ) are determined from the selected
best beam.
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, and whose horizontal patterns are pointed
towards different directions as shown in the figure below:
© Forsk 2009 AT271_TRG_E6 405

Figure 8.3: Grid Of Beams Modelling
Let us assume that all the beams and the main antenna have the same 18 dBi gain, and the vertical attenuation at the
user location is 15 dB, which is also the same for all the beams because we assume that the vertical patterns are the same.
If the user is located at  = 70 azimuth, as shown in the figure below, Atoll determines the best beam, which has the
highest gain towards  , as follows:
Horizontal Vertical
H V Total Gain
Beam Gain (dBi) Attenuation Attenuation G Beam – L Beam – L Beam
(dB)
(dB) (dB)
0° 18 60 15 18 - 60 - 15 -57
30° 18 60 15 18 - 60 - 15 -57
60° 18 2.21 15 18 - 2.21- 15 0.79
-30° 18 60 15 18 - 60 - 15 -57
-60° 18 60 15 18 - 60 - 15 -57
Transmitter
Centre of the pixel where the served user is located

 Angle between the user and the transmitter azimuth
Figure 8.4: GOB Modelling - Determination of the Best Beam
In our example, the total gain of the beam at 60° is the highest. Therefore this beam is selected as the best beam.
If this beam has been selected in the downlink,
SA SA H V
G DL = 18 dB and L DL = L Beam + L Beam = 17.21 dB
If this beam has been selected in the uplink,

SA SA H V
G UL = 18 dB and L UL = L Beam + L Beam = 17.21 dB
8.5.1.2 Adaptive Beam Modelling

An adaptive beam smart antenna is capable of steering a given antenna pattern towards the direction of the served signal.
In Atoll, this is modelled using a single antenna pattern, called a beam because of its highly directional shape. During the
simulations, this adaptive beam is oriented in the direction of each served user in order to model the effect of the smart
antenna.
406 AT271_TRG_E6 © Forsk 2009

SA SA
The adaptive beam gains ( G DL and G UL ) are the antenna gains defined for the beam, and the adaptive beam losses
SA SA H V
( L DL and L UL ) are the horizontal and vertical pattern attenuations L Beam + L Beam towards the user direction.
The following example shows how Atoll calculates the adaptive beam gains and losses.
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, and the vertical attenuation at the user location is
15 dB.
If the user is located at  = 60 azimuth, 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 8.5: Adaptive Beam Modelling - Determination of the Best Beam
If the adaptive beam smart antenna is selected in the downlink, 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, 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, as the ideal beam steering algorithm steers the beam towards the served user, L Beam = 0 . These values are used
in interference calculation to determine the downlink interfering signal due to transmission towards the served user, as well
as for calculating the uplink interfering signals received at transmitter when decoding signal received from the served user.
8.5.1.3 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. 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,  Spread .
You can assign a spreading angle to each clutter class in your document. Atoll reads the clutter class in which the served
user is located to determine the spreading angle. Different clutter types have different spreading effects on the propagation
of radio waves. 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, you can enter the C/I gains and their cumulative proba-
bilities for each spreading angle, in the smart antenna equipment based on the statistical model. For each smart antenna
SA
equipment based on statistical modelling, you can set a probability threshold, TProb .
To find the smart antenna gain, Atoll determines the clutter class of the served user, it reads the spreading angle from the
clutter class properties, it reads the probability threshold from the smart antenna properties, and reads the smart antenna
SA
C/I gain defined for the Probability = 1 – TProb corresponding to the spreading angle.
The following example shows how Atoll calculates the statistical C/I gains and losses.
Example:
Let us assume that the served user is located at a an urban clutter class with  Spread = 10 . The smart antenna equip-
SA SA
ment has TProb = 80 % . Atoll will read the smart antenna C/I gain G for Prob = 20 % . If a gain for the exact
probability value of 20% is not defined, Atoll linearly interpolates the gain value from the two surrounding values.
SA SA SA
If G Prob = 19%
= 4.6298 dB and G Prob = 20.4%
= 4.7196 dB , then G Prob = 20%
= 4.6941 dB
The smart antenna gains are the same for uplink and downlink. Their are no losses for this type of smart antenna equip-
ment. Negative values of C/I gains are considered as losses.
© Forsk 2009 AT271_TRG_E6 407

8.5.1.4 Optimum Beamformer Model

The optimum beamformer performs simple beam forming in the downlink to steer the main beam towards the served user,
and uses the Minimum Mean Square Error algorithm in the uplink in order to cancel interference.
Adaptive antenna systems use more than one antenna element, along with smart signal processing, to locate and track
various types of signals, to dynamically minimize interference, and maximize useful signal reception. The signal processor
dynamically applies weights to each element of the adaptive antenna system to create array patterns in real-time.
Figure 8.6: Linear Adaptive Array System
The antenna patterns created for downlink transmission have a main beam pointed in the direction of the useful signal. In
the uplink, in addition to the main beam pointed in the direction of the useful signal, there can also be one or more nulls in
the direction of the interfering signals. If the adaptive antenna system is using E SA antenna elements, it is possible to
create E SA – 1 nulls to cancel E SA – 1 interfering signals. In a mobile environment where the interference is not station-
ary, antenna patterns are adjusted so that the nulls remain in the direction of the moving interference. A system using
adaptive antennas adjusts the weights on each antenna element to achieve such a pattern.
The optimum beamformer supports linear adaptive array systems, such as the one shown in Figure 8.6: on page 408.
In the following explanations, we assume:
• There are a total of E SA elements in the adaptive antenna system.

•  is the angle of arrival for the useful signal.
•  is the angle at which we want to calculate the smart antenna gain.
• d is the distance between two adjacent antenna elements.
8.5.1.4.1 Downlink
Figure 8.7: Downlink Beamforming
408 AT271_TRG_E6 © Forsk 2009

The downlink algorithm models a conventional beam former. The smart antenna processor applies complex weights, w n ,
to each antenna element in order to form a beam towards the served user. The magnitude of these complex weights is set
to 1. The beamforming is performed using only the phase of the complex weights. The steering vector, S  , representing
the complex weights for forming a beam towards the served user, i.e., at the angle of arrival  , is given by:
2 2 2 T
  
S  = 1 e e  ... e
Where the notation T represents the transpose of a matrix.
Therefore, the complex weight at any nth antenna element can be given by:
2
– j  -------  nd  sin 

wn = e
 – j    n  sin 
In Atoll, d = --- , therefore, w n = e .
2
The smart antenna gain in any direction  can be given by:
SA H
G    = gn     S  R  S
Where the notation H represents the Hilbert transform, which is the complex conjugate transpose of a matrix, g n    is the
gain of the nth antenna element in the direction  , and R  is the array correlation matrix for a given user direction  , given
by:
H
R = S  S
For the direction of the served user, i.e.,  , the smart antenna gain is calculated as follows:
SA H H H 2
G    = 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.
SA SA
The smart antenna gain in dB is G    = 10  Log  G  .
The smart antenna is able to form the beam only in the horizontal plane, therefore, the vertical pattern is assumed to
remain the same.
Power Combination Gain
Cell transmission power is fed to each antenna element of the smart antenna system. Since each element transmits the
same input power, this results in a gain due to power combination, i.e., the powers fed to each antenna element are
combined for transmission.
Additional Processing During Monte Carlo Simulations
During Monte Carlo simulations, as Atoll calculates the smart antenna gains for each victim mobile in a cell’s coverage
area, it averages the array correlation matrix R  over all the iterations in order to generate the angular distribution of the
downlink traffic power.
The average array correlation matrix is given by:
J
R Avg =  j  pj  Rj
j=1
Where R Avg is the average downlink array correlation matrix, J is the number of served mobiles during the simulation,  j
is the probability of presence of the mobile j, p j is the EIRP transmitted towards the mobile j, and R j is the array correlation
matrix for the mobile j.
8.5.1.4.2 Uplink
The uplink models the adaptive Minimum Mean Square Error algorithm which optimizes the useful signal as well as
cancels the interference from the most interfering E SA – 1 interfering mobiles. The optimal beam forming method used in
Atoll overcomes the limitations of a null steering beam former and maximizes the signal quality.
A simple null steering beam former requires the knowledge of the directions of interference sources, and the estimated
weights do not maximize the signal quality. Whereas, an optimal beam former does not require knowledge of directions
and power levels of interference to maximize the output. It only requires the direction of the useful signal in order to calcu-
late the optimum signal quality.
© Forsk 2009 AT271_TRG_E6 409

Figure 8.8: Uplink Adaptive Algorithm
Let ŵ represent the vector of ESA complex weights for the beam former. ŵ is given by:
–1
ŵ =    R N  S 
Where S  is the steering vector in the direction of the served user,  .   , which is a constant value for a given useful
signal, represents the smart antenna gain in the uplink given by the equation:
E SA
  = ----------------------------------
-
H –1
S  RN  S
–1
R N is the inverse of the total noise correlation matrix. The total noise correlation matrix is the sum of the thermal noise
correlation matrix R n , and the interference correlation matrix R I , given by:
J
H
2
RN = Rn + RI = n  I +
j=1
J
H
2
Where R n =  n  I and R I =
j=1
2
 n is the thermal noise power. I is the identity matrix. p j is the power received by one element of the smart antenna from
the jth interfering mobile. S j is the steering vector in the direction of the jth interfering mobile,  . J is the total number of
interfering mobiles.
The total noise power, including thermal noise and interference from all uplink interferers, received by a cell is given by:
ˆ = 2  S H  R–1  S
P N   N 
The total power received from the served user is given by:
ˆ = p  2   S H  R–1  S  2
P     N 
Where p  is the power received by one element of the smart antenna from the served user.
The uplink signal quality is calculated by:
2 H –1 2
SA Pˆ p     S  RN  S  H –1
Q UL = ------- = --------------------------------------------------------------- = p  S  RN  S
Pˆ 2 H –1
N   S  RN  S
From the above equation, we can determine the uplink smart antenna beam forming gain in the direction of the served
user. RSCP TCH – UL can be calculated from the above equation by considering the interference and noise to be null, i.e.,
–1
R N = I . This gives:
H
RSCP TCH – UL = p   S   I  S  = p   E SA
410 AT271_TRG_E6 © Forsk 2009

From the above equation, the uplink smart antenna beam forming gain equals the number of smart antenna elements, i.e.,
SA
G = E SA .
–1
The inverse noise correlation matrix R N for each iteration k includes the effect of the matrix calculated for the previous
iteration. Hence, Atoll is able to calculate an average of the smart antenna interference-cancellation effect. The result is
the angular distribution of the uplink load, which is calculated from the inverse of the noise correlation matrix obtained at
the end of the last iteration of a Monte Carlo simulation. This angular distribution of the uplink load can be stored in the
Cell Parameters per Timeslot table. The average of the inverse noise correlation matrices is calculated as follows:
K
1
 RN
–1 –1
RN = ---- 
Avg K k
k=1
–1 –1
Where R N is the average of the inverse noise correlation matrices of all the iterations from k = 1 to K, and R N is
Avg k
the inverse noise correlation matrix of the kth iteration. The uplink load is calculated from the average inverse noise corre-
lation matrix.
8.5.1.5 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. The smart antenna gains and losses are used during the simulations and the results are stored
in the Cell Parameters per Timeslot table, which can be used in coverage predictions.
8.5.2 Construction of the Geographic Distributions

During simulations, Atoll uses the smart antenna model selected for each transmitter to calculate the smart antenna gains
and losses. These values are calculated and stored for each user generated for the simulations. Therefore, these values
are calculated and are available for the given locations of the users, i.e., points, only. Atoll uses the Angular Step value
that you set when creating and running simulations to construct the geographic distributions of these results.
Once Atoll has calculated the downlink traffic power and the uplink load using the smart antenna gains and losses deter-
mined as explained in the previous section, at the location of a given user, it calculates the same for points located at the
angle equal to that of the Angular Step of the simulations.
At the end of the simulations, Atoll has a number of points, Angular Step apart, available with the values of these results.
The geographic distribution of these results, i.e., downlink traffic power and uplink loads, is constructed by connecting the
resulting value points.
The following example explains how the geographic distribution of downlink traffic power is created. The geographic distri-
bution of uplink loads is constructed in the same manner.
Example:
Let us assume a smart antenna equipment using adaptive beam modelling. The angular step defined for the simulations
is  Step = 30 . Therefore, the results are calculated for each point located at regular steps of 30 , i.e., 12 points. The
downlink traffic power at the served user (W) with the adaptive beam pointing in the user’s direction is P W . The downlink
traffic powers, using the same adaptive beam pointed towards the served user, at the 12 other points are also determined.
© Forsk 2009 AT271_TRG_E6 411

Figure 8.9: Construction of the Geographic Distribution of Downlink Traffic Power
The resulting geographic distribution is formed by linearly joining the obtained results.
Figure 8.10: Geographic Distribution of Downlink Traffic Power
The accuracy of the geographic distribution depends upon the value of the angular step. A radiation pattern created at a
1 step will be much more accurate than one created at 45 , for example. But, the latter will be computed 45 times faster
than the first. The value of the Angular Step should be the best possible compromise between calculation speed and accu-
racy.
8.5.3 Modelling in Coverage Predictions

The results of Monte Carlo simulations, including the smart antenna results, can be stored in the Cells and in the Cell
Parameters per Timeslot tables, and can be used to carry out coverage predictions. The main results of Monte Carlo simu-
lations used in coverage predictions are:
• If a smart antenna is used in both uplink and downlink:
UL –  DL – 
Geographic distribution of UL load X 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 DL
UL load X and DL traffic power P Traffic
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, and reading the uplink load and downlink traffic power from the geographic distribution
results. If an exact value for the angle is not available, the load and power are determined using linear interpolation for the
given angle between two available values.
412 AT271_TRG_E6 © Forsk 2009

For example, the figure below shows the distribution of downlink traffic power and uplink traffic load results from a simu-
DL – 315 UL – 315
lation. For a pixel located at  = 315 , the downlink traffic power P Traffic and the uplink load X are read
DL – 315 UL – 315
from these results. In this example, P Traffic  30 dBm , and X = 2.75 % .
For each pixel, Atoll determines the downlink traffic powers and the uplink loads from all the transmitters.
Figure 8.11: Geographic Distribution of downlink traffic power and uplink load
8.6 N-Frequency Mode and Carrier-Type Allocation

Transmitters that support N-frequency mode are multi carrier transmitters with a master and one or more slave carriers.
You can assign master and slave carriers to transmitters manually, or use the automatic frequency allocation in Atoll to
assign carrier types automatically.
8.6.1 Automatic Carrier-Type Allocation

For each transmitter, Atoll determines a list of "near" transmitters. For any transmitter TXi, its "near" transmitters would be
the ones that are geographically near the transmitter TXi, according to a sort function. Among the "near" transmitters, the
transmitters that are directed towards the transmitter TXi are considered before the transmitters that are directed in the
same direction as TXi. This is because transmitters directed towards each other interfere more compared to transmitters
pointing in the opposite direction.
Next, Atoll assigns one master carrier to each transmitter TXi, such that the master carrier of TXi is different from the
master carrier of TXj, where TXj belongs to the list of "near" transmitters. The master carrier is one of the cells defined in
the transmitter. All the other cells of the transmitter are assigned the carrier-type "slave".
For transmitters that support the N-frequency mode and have master carriers properly assigned, Atoll performs the neigh-
bour and scrambling code allocation for the master carrier only.
8.7 Neighbour Allocation

Atoll permits the automatic allocation of intra-technology neighbours in a TD-SCDMA document. Two allocation algorithms
are available, one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours. The intra-technology
neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all the cells of TBC trans-
mitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfil following conditions:
• They are active
• They satisfy the filter criteria applied to the Transmitters folder
• They are located inside the focus zone
Note:
© Forsk 2009 AT271_TRG_E6 413


We assume a reference cell A and a candidate neighbour, cell B. When automatic allocation starts, Atoll checks following
conditions:
1. The distance between both cells must be less than the user-defined maximum inter-site distance. If the distance
between the reference cell and the candidate neighbour is greater than this value, the candidate neighbour is
discarded.
2. 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: This option enables you to force cells located on the same site as reference cell
in the candidate neighbour list. This constraint can be weighted among the others and ranks the neighbours
through the importance field.
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. This constraint can be weighted
among the others and ranks the neighbours through the importance field.
Note:
• Adjacency criterion: Geographically adjacent cells are determined on the basis of their
best server coverages in TD-SCDMA projects. Let CellA be a candidate neighbour cell of
CellB. 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.e., P-CCPCH RSCP of CellA > P-
CCPCH RSCP T_Drop and P-CCPCH RSCP of CellA > P-CCPCH RSCP of CellB -
T_Comp.)
• When this option is selected, adjacent cells are sorted and listed from the most adjacent to
the least, depending on the above criterion. Adjacency is relative to the number of pixels
Force neighbour symmetry: This option enables you to force the reciprocity of a neighbourhood link. Therefore, if
the reference cell is a candidate neighbour of another cell, the later will be considered as candidate neighbour of
the reference cell.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you
may force/forbid a cell to be candidate neighbour of the reference 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, the existing neighbours are kept.
- Intra-carrier neighbours: intra-carrier handover is a baton handover.

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).
SA is the area where the cell A is the best serving cell.
- The P-CCPCH RSCP from the cell A is greater than the P-CCPCH RSCP T_Add.
- The P-CCPCH RSCP from the cell A is greater than the P-CCPCH RSCP from all other cells.
SB is the area where the cell B can enter the handover set.
414 AT271_TRG_E6 © Forsk 2009

- The P-CCPCH RSCP from the cell B is greater than the P-CCPCH RSCP T_Drop.
- 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.
Figure 8.12: Intra-carrier Neighbours
- Inter-carrier neighbours: Inter-carrier handover is a hard handover. It is needed in a multi-carrier CDMA net-
work:
- To balance and share the load between carriers and layers.
- To make a coverage reason handover from micro cells to macro cells.
SA is the area where the received signal level from the transmitter A is greater than a minimum signal level.
SA is the coverage area of reference cell’s transmitter A restricted between an upper and a lower limit. The
upper limit represents the start of the handover area (best server area of A plus the handover margin named
"handover start"), and the lower limit represents the end of the handover area (best server area of A plus the
margin called "handover end").
SB is the coverage area where the candidate cell’s transmitter B is the best server.
Figure 8.13: Overlapping Coverages
© Forsk 2009 AT271_TRG_E6 415

SA  SB
Atoll calculates the percentage of covered area ( ----------------------  100 ), which it compares with the % minimum covered
SA
The coverage condition can be weighted among the others and ranks the neighbours through the importance field.
maximum number of neighbours to be allocated to each cell is exceeded. If 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
that have the most importance will be allocated to the reference cell. You can set different maximum numbers of neigh-
bours for intra-carrier and inter-carrier neighbours for each cell. If you have defined these values for each cell, Atoll uses
these limits instead of the value set in the allocation dialog.
As indicated in the table below, the neighbour importance depends on the neighbourhood cause; this value varies between
from 0 and 100%.
Importance
value
Existing
Existing neighbour If the Delete existing neighbours option is not selected
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
Adjacent transmitter If the Force adjacent cells as neighbours option is selected IF
If the % minimum covered area is exceeded IF
coverage conditions
If the Force neighbour symmetry option is selected IF
relationship
Except the case of forced neighbours (importance = 100%), priority assigned to each neighbourhood cause is determined
using the Importance Function (IF). The IF considers three factors for calculating the importance:
• The co-site factor (C) which is a Boolean
• The adjacency factor (A) which deals with the percentage of adjacency
The IF is user-definable using the Min importance and Max importance fields.
Factor Min importance Default value Max importance Default value

The IF evaluates importance as follows:
Neighbourhood Resulting IF using the

cause IF default values from the
Co-site Adjacent table above
No No Min  O  +   O   O  1% + 29%  O 
No Yes Min  A  +   A   Max  O   O  +  100% – Max  O    A   30% + 30%  30%  O  + 70%  A  
Yes Yes Min  C  +   C   Max  O   O  +  100% – Max  O    A   60% + 40%  30%  O  + 70%  A  
Where   X  = Max  X  – Min  X 
Notes:
fields, neighbours will be ranked in the following order:
i. Co-site neighbours
ii. Adjacent neighbours
iii. Neighbours based on coverage overlapping
• If the ranges of the importance factors overlap, the neighbours may not be ranked
according to the neighbourhood cause.
• The ranking between neighbours from the same category depends on the factors (A) and
(O).
416 AT271_TRG_E6 © Forsk 2009

• 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.
In the Results part, Atoll provides the list of neighbours, the number of neighbours, and the maximum number of neigh-
bours allowed for each cell. In addition, it indicates the importance (in %) of each neighbour and the allocation reason, i.e.,
a neighbour may be marked as exceptional pair, co-site, adjacent, coverage, or symmetric. For neighbours accepted for
co-site, adjacency, and coverage reasons, Atoll displays the percentage of area that satisfies the coverage conditions and
the corresponding surface area (km2), the percentage of area that satisfies the adjacency conditions and the correspond-
ing surface area (km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing.
Notes:
• You donot require simulations or coverage predictions for an automatic neighbour
allocation. For automatic neighbour allocation, Atoll automatically calculates the missing
path loss matrices.
• Although no specific terminal, mobility, or service is selected for automatic neighbour
allocation, 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. If this is the same
for all terminals, Atoll uses the terminal with the lowest noise figure.
- Mobility does not impact the allocation
• The percentage of area is calculated using the resolution specified in the Properties dialog
of the Predictions folder (Default Resolution).
relationship already exists and the Delete existing neighbours option is not selected. In this
case, 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.
• Symmetric neighbour relations are only added to the neighbour lists if the neighbour lists
are not already full. Thus, if the cell B is a neighbour of the cell A, but cell A is not a
neighbour of the cell B, there can be two possibilities:
i. There is space in the cell B neighbour list: cell A will be added to the list. It will be the
last one.
ii. 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.
• If you select Force exceptional pairs and Force symmetry options, Atoll considers the
constraints between exceptional pairs in both directions so as to respect the symmetric
relation. On the other hand, if a neighbour relation is forced in one direction and forbidden
in the other, symmetry cannot be respected. In this case, Atoll displays a warning message
in the Event Viewer.
• In the results, Atoll displays only the cells for which it finds new neighbours. Therefore, if a

• TBA cells
• Neighbours of TBA cells marked as exceptional pair, adjacent, or symmetric
• Neighbours of TBA cells that satisfy coverage conditions
8.8 Scrambling Code Allocation

Downlink scrambling codes enable mobile to distinguish one cell from another. In TD-SCDMA, there are 128 scrambling
codes (or P-CCPCH midamble codes) distributed in 32 clusters of 4 codes each. A different DL synchronisation code, or
SYNC_DL code, is assigned to each cluster. Scrambling codes are numbered from 0 to 127, and SYNC_DL codes from
0 to 31.
Depending on the options you select for automatic allocation of scrambling and SYNC_DL codes, Atoll takes into account
either all the cells of TBC transmitters, or only cells of active and filtered transmitters located inside the computation zone.
Atoll calculates a scrambling code and a SYNC_DL code to all these cells. But, it allocates scrambling codes and
SYNC_DL codes only to TBA cells (cells to be allocated). TBA cells are the cells that fulfill the following conditions:
• They are active
• They satisfy the filter criteria applied to the Transmitters folder
© Forsk 2009 AT271_TRG_E6 417


Furthermore, if there are transmitters that support the N-frequency mode among the TBC transmitters of your network, the
scrambling code allocation also considers the master and slave carrier allocations.
Note:

8.8.1.1 Allocation Constraints and Options
The scrambling code and SYNC_DL code allocation algorithm can take into account following constraints:
1. Neighbour relations between cells
You may consider:
- First order neighbours: The neighbours of TBA cells listed in the Intra-technology neighbours table,
- Second order neighbours: The neighbours of neighbours,
- Third order neighbours: The neighbour’s neighbour’s neighbours.
Notes:
• In the context of the scrambling code allocation, the term "neighbours" refers to intra-carrier
neighbours.
different scrambling codes to the TD-SCDMA neighbours of a GSM transmitter. In order to
consider inter-technology neighbour relations in the scrambling code allocation, you must
make the Transmitters folder of the GSM .atl document accessible in the TD-SCDMA .atl
document. For information on making links between GSM and TD-SCDMA .atl documents,
see the User Manual.
2. The scrambling code reuse distance

Reuse Distance: It is a constraint on the allocation of scrambling codes. The same scarmbling code or SYNC_DL
code cannot be allocated to two sites that are not farther apart than the reuse distance. Scrambling code reuse
distance can be defined for each cell in the cell properties. If this value is not defined, Atoll uses the default reuse
distance defined in the Automatic Scrambling Code and SYNC_DL code Allocation dialog. The reuse distance
constraint is used for clustered and distributed per cell allocation strategies.
3. The carrier for which you want to perform the automatic allocation
Carrier: You can select "All" or a specific carrier. If you select "All", Atoll allocates the same scrambling code to
each carrier of a transmitter.
4. 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. 3GPP spec-
ifications define 32 SYNC_DL codes with 4 corresponding scrambling codes each (SYNC_DL codes are
numbered from 0 to 31). However, it is possible to define a different value (e.g. if you set the number of scrambling
codes per SYNC_DL codes to 2, scrambling codes will be distributed among 64 SYNC_DL codes).
5. Atoll can use a maximum of codes
Use a Maximum of Codes: If you choose to use a maximum of codes, Atoll will try to spread the allocated spectrum
of scrambling codes as much as possible.
6. Existing allocation
Delete All Codes: If you select this option, Atoll will delete any existing scrambling code allocation and perform a
fresh allocation. Otherwise, Atoll keeps the existing allocation.
8.8.1.2 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, scrambling codes
among a minimum number of clusters. In this case, Atoll will preferentially allocate all the codes within the same
cluster.
• Distributed per Cell: This strategy consists in using as many clusters as possible. Atoll will preferentially allocate
codes from different clusters.
• One SYNC_DL code per site: This strategy allocates one cluster, i.e., one SYNC_DL code, per site, then one
scrambling code from the cluster to each cell of the site. When all the clusters have been allocated but there are
still sites remaining, Atoll reuses the clusters as far as possible at another site.
418 AT271_TRG_E6 © Forsk 2009

Notes:
• Same carriers must be assigned different scrambling codes. Different carriers of the same
site can be assigned the same scrambling code. Therefore, cells of a transmitter (i.e.
different carriers) are assigned the same scrambling code, if the scrambling code domains
associated with the carriers have a common cluster or enough codes in one cluster.
• Distributed per site: This strategy allocates a group of adjacent clusters, i.e., consecutive SYNC_DL codes, to
each site, then one cluster, or SYNC_DL code, to each transmitter on the site according to its azimuth, and finally
one scrambling code from each cluster to each cell of each transmitter. The number of adjacent clusters, or con-
secutive SYNC_DL codes, depends on the number of transmitters per site. When all the sites have been allocated
adjacent clusters, and there are still sites remaining to be allocated, Atoll reuses the adjacent clusters as far as
possible at another site.
In the Results table, Atoll only displays scrambling codes and SYNC_DL codes allocated to TBA cells.

and "First Order"),
Additional constraints are considered when:
• The neighbour cells cannot share the same cluster (for the "Distributed per site" allocation strategy only).
and the cost of the scrambling code plan. During the allocation, Atoll tries to assign different scrambling codes to the TBA
cell and its near cells. If it respects all the constraints, the cost of the scrambling code plan is 0. When a cell has too many
constraints and there are not anymore scrambling codes available, Atoll breaks the constraint with the lowest cost so as
to generate the scrambling code plan with the lowest cost. For information on the cost generated by each constraint, see

Strategies: Clustered and Distributed per Cell
Atoll processes TBA cells according to their priority. It allocates scrambling codes 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 calcu-
lating cell priority, see "Cell Priority" on page 420.
Strategy: One SYNC_DL Code per Site
Atoll assigns a cluster, i.e., a SYNC_DL code, 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
but there are still sites remaining, Atoll reuses the clusters at the other sites. When the Reuse Distance option is selected,
the algorithm reuses the clusters as soon as the 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 scrambling code from the cluster to each cell located on the sites (codes belong to the assigned
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.
Strategy: Distributed per Site
Atoll assigns a group of adjacent clusters, i.e., SYNC_DL codes, 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 sites have been
allocated adjacent clusters, and there are still sites remaining to be allocated, Atoll reuses the adjacent clusters at other
sites. When the Reuse Distance option is selected, the algorithm reuses the clusters as soon as the reuse distance is
exceeded. Otherwise, when the option is not selected, the algorithm tries to assign reused clusters as spaced out as possi-
ble.
Then, 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 Clus-
© Forsk 2009 AT271_TRG_E6 419

ters"). Then, Atoll allocates a scrambling code to each cell located on the transmitters (codes belong to the assigned clus-
ters). 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 423. or information on calculating cell priority, see
Determination of Groups of Adjacent Clusters
In order to determine the groups of adjacent clusters to be used, Atoll:

• Defines theoretical groups of adjacent clusters, independent of the defined domain, considering the 128 scram-
bling codes available and 4 codes per cluster.
• 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
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, the theoretical groups of adjacent clusters will be:
Group 1 Group 2 Group 3 Group 4 ... Group 11

Cluster 1 Cluster 4 Cluster 7 Cluster 10 ... Cluster 31
Cluster 2 Cluster 5 Cluster 8 Cluster 11
If no domain is assigned to cells, Atoll can use all these groups for the allocation. On the other hand, if a domain is used,
Atoll compares adjacent clusters actually available in the assigned domain with the theoretical groups and only keeps adja-
cent clusters common with the theoretical groups.
If we have a domain comprising 12 clusters: clusters 1 to 8 and clusters 12 to 15. In this case, Atoll will use the following
groups of adjacent clusters:
• Group 2 with cluster 3, 4 and 5
The clusters 1, 2 and 15 will not be used.
If a domain does not contain any adjacent clusters, Atoll displays a warning message in the Event Viewer.

In case you have a multi-carrier network and you run the scrambling code allocation on all the carriers, the 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 (Clustered and Distributed per cell), 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 same scrambling code is assigned to each cell of the transmitter.
In case of the "One SYNC_DL code per site" strategy, Atoll assigns a cluster, i.e., a SYNC_DL code, to each site and then,
allocates a scrambling code 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 scrambling code is
assigned to each cell of the transmitter.
In case of the "Distributed per site" strategy, Atoll assigns a group of adjacent clusters, i.e., SYNC_DL codes, to each site,
then a cluster to each transmitter and finally, allocates a scrambling code to each transmitter. It starts with the highest prior-
ity transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near
transmitters. The same scrambling code is assigned to each cell of the transmitter.
For information on calculating transmitter priority, see "Transmitter Priority" on page 423.
Note:
alphanumeric order.

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. 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 scrambling
code allocation process.
There are six criteria employed to determine the cell priority:
• Scrambling Code Domain Criterion
420 AT271_TRG_E6 © Forsk 2009

The cost due to the domain constraint, C i  Dom  , depends on the number of scrambling codes available for the allocation.
The domain constraint is mandatory and cannot be broken.
When no domain is assigned to cells, 128 scrambling codes are available and we have:
C i  Dom  = 0
When domains of scrambling codes are assigned to cells, each unavailable scrambling code 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  = 128 – Number of scrambling codes in the domain
C i  Dist  =  Cj  Dist  i  
j
Where

C i  EP  =  cEP  i – j 
j
Where
     
Ci  N  = 
C j  N1  i   = I j  c N1
Where
Because two first order neighbours must not have the same scrambling code, Atoll considers the cost created by two first
C j  N1  i   + C j  N1  i  
C j – j  N1  i   = ----------------------------------------------------------
-
2
© Forsk 2009 AT271_TRG_E6 421

Where
Because two second order neighbours must not have the same scrambling code, 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
C l  N3  i   = Max  j   c N3
Where
Because two third order neighbours must not have the same scrambling code, Atoll considers the cost created by two third
C l  N3  i   + C l  N3  i  
C l – l  N3  i   = ----------------------------------------------------------
-
2
Note:
And

C i  N 2G  =  cN 2G
 j – Tx 2G 
j
Where
2G
• Cluster Criterion
When the "Distributed per Site" allocation strategy is used, you can consider additional constraints on allocated clusters
(one cell, its first order neighbours and its second order neighbours must be assigned scrambling codes from different clus-
ters). In this case, the constraint level of any cell i depends on the number of first and second order neighbours, j and k.
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. This value can be defined in the Constraint Cost dialogue.
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 
422 AT271_TRG_E6 © Forsk 2009


In case you have a multi-carrier network and you run scrambling code allocation on "all" the carriers, 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 scrambling code allo-
cation process.
as:
With C Tx  U  = Max  C  U   and C  Dom  = 128 – Number of scrambling codes in the domain
i  Tx i Tx

In case of "Per Site" allocation strategies (One SYNC_DL code per Site and Distributed per Site), Atoll allots priorities to
sites. Priorities assigned to sites depend on how much constrained each site is and the cost defined for each constraint.
The higher the cost on a site, the higher the priority it has for the scrambling code allocation process.
Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. 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
Tx  S Tx S
Here, the domain considered for the site is the intersection of domains available for transmitters of the site. The domain
constraint is mandatory and cannot be broken.
8.8.2 IScrambling Code Allocation Example

8.8.2.1 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, let us consider the following sample scenario:
Figure 8.15: Scrambling Code Allocation Example
Let Site0, Site1, Site2, and Site3 be four sites, with 3 transmitters each using carrier 0, to whom scrambling codes have
to be allocated out of 6 clusters of 4 scrambling codes. This implies that the domain of scrambling codes for the four sites
is from 0 to 23 (cluster 0 to cluster 5). The reuse distance is supposed to be less than the inter-site distance. Only co-site
neighbours exist.
The following section shows the results of each combination of options with explanations where necessary.
© Forsk 2009 AT271_TRG_E6 423

8.8.2.1.1 Strategy: Clustered

Since the restrictions of neighbourhood only apply to co-sites and, in our case, the distances between sites are greater
than the reuse distance, every cell has the same priority. Allocation is performed in an alphanumeric order.
Without "Use a Maximum of Code" With "Use a Maximum of Code"
As it is possible to use a maximum of codes, Atoll starts

allocation at the start of a different cluster at each site.
Atoll starts allocating the codes from the start of cluster 0
When a cluster is reused, and there are non allocated
at each site.
codes left in the cluster, Atoll first allocates those codes
before reusing the already used ones.
8.8.2.1.2 Strategy: Distributed per Cell

Since the restrictions of neighbourhood only apply to co-sites and, in our case, the distances between sites are greater
than the reuse distance, every cell has the same priority. Allocation is performed in an alphanumeric order.
Atoll allocates codes from different clusters to each cell of Atoll allocates codes from different clusters to each site’s
the same site. Under given constraints of neighbourhood cells. As it is possible to use a maximum of codes, Atoll
and reuse distance, same codes can be allocated to each allocates the codes so that there is least repetition of
site’s cells. codes.
424 AT271_TRG_E6 © Forsk 2009

8.8.2.1.3 Strategy: One SYNC_DL Code per Site

Since the restrictions of neighbourhood only apply to co-sites, therefore, every site has the same priority. Cluster allocation
to sites is performed in an alphanumeric order.
In this strategy, a cluster of codes is limited to be used at

just one site at a time unless all codes and clusters have When it is possible to use a maximum of codes, Atoll can
been allocated and there are still sites remaining to be allocate different codes from a reused cluster at another
allocated. In this case, Atoll reuses the cluster as far as site.
possible at another site.
8.8.2.1.4 Strategy: Distributed per Site

Since the restrictions of neighbourhood only apply to co-sites, therefore, every site has the same priority. Cluster allocation
to sites is performed in an alphanumeric order.
A group of adjacent clusters is allocated to one site at a

time, unless all the codes and groups of adjacent clusters
When it is possible to use a maximum of codes, Atoll can
have been allocated but there are still sites remaining to
allocate different codes from a reused groups of adjacent
be allocated. In this case (here only one group of adjacent
clusters at another site.
clusters 0, 1, and 2 is available), Atoll reuses the group as
far as possible at another site.
8.8.2.2 Multi Carrier Network

If you have a multi carrier network, i.e., transmitters with more than one cells using different carriers, and you run scram-
bling code allocation on "all" the carriers, Atoll allocates the same scrambling code to each carrier of a transmitter.
Let Site0, Site1, Site2, and Site3 be four sites with 3 cells using carrier 0 and 3 cells using carrier 1. Scrambling codes
have to be allocated out of 6 clusters consisted of 4 scrambling codes. This implies that the domain of scrambling codes
for the four sites is from 0 to 23 (cluster 0 to cluster 5). The reuse distance is supposed to be less than the inter-site
© Forsk 2009 AT271_TRG_E6 425

distance. Only co-site neighbours exist. Every site has the same priority and the cluster allocation to sites is performed in
an alphanumeric order.
Atoll allocates one cluster at each site and then, one code to each transmitter. Then, the same code is given to each cell
of the transmitter.
Figure 8.16: Scrambling Code Allocation to All Carriers
8.9 Automatic GSM/TD-SCDMA Neighbour Allocation

It is possible to automatically calculate and allocate neighbours between GSM and TD-SCDMA networks. In Atoll, it is
called inter-technology neighbour allocation.
• When the TD-SCDMA coverage is not continuous. In this case, the TD-SCDMA coverage is extended by TD-
SCDMA to GSM handovers.
• In order to balance traffic and service distribution between both networks.
Atoll’s automatic inter-technology neighbour allocation algorithm takes into account both cases.
• An .atl document containing the GSM network, GSM.atl, and another one containing the TD-SCDMA network, TD-
SCDMA.atl,
• An existing link on the Transmitters folder of GSM.atl into TD-SCDMA.atl.
transmitters of GSM.atl are potential neighbours. The TD-SCDMA cells, in TD-SCDMA.atl, to be allocated neighbours are
called TBA cells which fulfill following conditions:
• They are active
• They satisfy the filter criteria applied to Transmitters folder
or one of its subfolders.
Only TD-SCDMA TBA cells can be assigned neighbours.

• The inter-transmitter distance
• The maximum number of neighbours
• Allocation options
• The selected allocation strategy
We assume we have a TD-SCDMA reference cell, A, and a GSM candidate neighbour transmitter, B.
426 AT271_TRG_E6 © Forsk 2009


1. The distance between the TD-SCDMA reference cell and the GSM neighbour must be less than the user-defined
maximum inter-site distance. If the distance between the TD-SCDMA reference cell and the GSM neighbour is
greater than this value, then the candidate neighbour is discarded.
Note:
further information on inter-transmitter distance calculation, please refer to "Calculation of
Inter-Transmitter Distance" on page 429.

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. This option is automatically selected.
may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.
neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
list if the maximum number of neighbours to be allocated to each cell is exceeded. The candidate neighbour pri-
ority depends on the neighbourhood cause. Priority assigned to each neighbourhood cause is listed in the table
below (1 is a higher than 2, and so on).

distance conditions
bours allowed for each cell. In addition, it indicates the reason of allocation for each neighbour. Therefore, a neighbour
may be marked as exceptional pair or co-site. If the neighbour is not forced but fulfills distance conditions, Atoll displays
the distance from the reference cell. Finally, if cells have previous allocations in the list, neighbours are marked as existing.

1. The distance between the TD-SCDMA reference cell and the GSM neighbour must be less than the user-defined
maximum inter-site distance. If the distance between the TD-SCDMA reference cell and the GSM neighbour is
greater than this value, then the candidate neighbour is discarded.
Note:
• The inter-transmitter distance is not effected by the azimuths. Only the geographical inter-
transmitter distance is considered.

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. This option is automatically selected.
may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.
neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
3. There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability.
© Forsk 2009 AT271_TRG_E6 427

- 1st case: SA is the area where the cell A is the best serving cell of the TD-SCDMA network.
- The pilot signal received from A is greater than the minimum pilot signal level and is the highest one.
- The margin is set to 0 dB.
- 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.
- 1st case: SB is the area where the cell B is the best serving transmitter of the GSM network.
In this case, the margin must be set to 0 dB.
is the highest one.
- 2nd case: The margin is different from 0 dB and SB is the area where:
SA  SB
SA
area. If this percentage is less than the minimum, the candidate neighbour B is discarded.
area.
list if the maximum number of neighbours to be allocated to each cell is exceeded. The candidate neighbour pri-
ority depends on the neighbourhood cause. Priority assigned to each neighbourhood cause is listed in the table
below (1 is a higher than 2, and so on).

coverage conditions
Therefore, among 15 candidate neighbours, only 8 (those with the highest priority) will be allocated to the reference cell.
be marked as exceptional pair or co-site. If the neighbour is not forced but fulfills coverage conditions, Atoll displays the
Notes:
• Coverage predictions are not required to perform an automatic neighbour allocation. When
starting an automatic neighbour allocation, Atoll automatically calculates the path loss
matrices, if not found.
dialog of the Predictions folder (Default Resolution).
relationship already exists and the Delete existing neighbours option is not selected when
you start a new allocation. In this case, Atoll displays a warning in the Event Viewer
indicating that the constraint on the forbidden neighbour will ignored by algorithm because
the neighbour already exists.
• In 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 a new
8.9.1.3 Appendices
8.9.1.3.1 Delete Existing Neighbours Option
not selected. If a new TBA cell i is created in TD-SCDMA.atl, you can run the automatic allocation with the Delete existing
neighbours option not selected, in order to allocate neighbours to the new cell i only.
If you change some allocation criteria (e.g., increase the maximum number of neighbours or create a new GSM TBC trans-
mitter) and start a new allocation without selecting the Delete existing neighbours option, Atoll examines the neighbour list
428 AT271_TRG_E6 © Forsk 2009

of the TBA cells and checks allocation criteria only if there is still space left in their neighbour lists. A new GSM TBC trans-
mitter can enter the TBA cell neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list.
8.9.1.3.2 Calculation of Inter-Transmitter Distance

When allocation algorithm is based on distance, Atoll takes into account the real distance and azimuths of antennas in
order to calculate the effective inter-transmitter distance.
Dist  CellA CellB  = D   1 + x  cos  – x  cos  

Where x = 0.5% so that the maximum variation in D does not to exceed 1%. D is stated in m.
Figure 8.17: Inter-Transmitter Distance Computation
The formula above implies that two cells facing each other have a smaller effective distance than the real physical
distance. It is this effective distance that is taken into account rather than the real distance.
This formula is not used when allocation algorithm is based on coverage overlapping. In this case, the actual inter-trans-
mitter distance is considered.
© Forsk 2009 AT271_TRG_E6 429

430 AT271_TRG_E6 © Forsk 2009

Chapter 9
WiMAX BWA Networks
simulations and coverage predictions available in WiMAX BWA projects.
Atoll
432 AT271_TRG_E6 © Forsk 2009

Chapter 9: WiMAX BWA Networks
9 WiMAX BWA Networks

This chapter describes all the calculations performed in Atoll WiMAX BWA documents, i.e, WiMAX 802.16d and WiMAX
802.16e. The titles of sections that present 16d- or 16e-specific algorithms include the name of the technology for you to
recognize them easily.
The first part of this chapter lists all the input parameters in the WiMAX BWA documents, their significance, location in the
Atoll GUI, and their usage. It also contains the lists of the formulas used during 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, Monte Carlo simulations, and traffic C/(I+N)-based
coverage predictions. 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 the effects of smart antennas, power control,
subchannelisation, MIMO etc., and the radio resource management algorithms used by the different available schedulers.
The third part also describes Forsk’s optimum beamformer model in detail.
If you are new to WiMAX, you can also see "Glossary of WiMAX Terms" on page 514 for information on WiMAX terms and
concepts, especially in the context of their user in Atoll.
Important:
• 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 74.
• 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 car-
rier 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).

The tables in the following subsections list the input and output parameters, and formulas used in simulations and other
computations.
9.1.1 Input
This table lists the input to computations, coverage predictions, and simulations.

Frame Duration
D Frame Global parameter ms Choice List: 2, 2.5, 4, 5, 8, 10, 12.5, 20
ms
Cyclic Prefix Ratio
r CP Global parameter None
Choice List: 1/4, 1/8, 1/16, 1/32
DL
O Fixed Global parameter SD Fixed time-domain overhead (DL)
UL
O Fixed Global parameter SD Fixed time-domain overhead (UL)
DL
O Variable Global parameter % Variable time-domain overhead (DL)
UL
O Variable Global parameter % Variable time-domain overhead (UL)
TDD Ratio of the DL subframe to the entire

r DL-Frame Global parameter %
frame (TDD only)
TDD
D TTG Global parameter ms Transmit Time Guard (TDD only)
© Forsk 2009 AT271_TRG_E6 433

TDD
D RTG Global parameter ms Receive Time Guard (TDD only)
M PC Global parameter dB Uplink power control margin
PZ Global parameter (WiMAX 802.16d) Number of subchannels per channel

N SC – UL or N SC – UL None
Permutation zone parameter (WiMAX 802.16e) in UL subframe
PZ Number of subchannels per channel
N SC – DL Permutation zone parameter (WiMAX 802.16e) None
in DL subframe
Global parameter (WiMAX 802.16d) Total number of subcarriers per
N SCa – Total None
Frame configuration parameter (WiMAX 802.16e) channel (FFT size)
Number of subcarriers used by the
N SCa – Preamble Frame configuration parameter (WiMAX 802.16e) None
preamble
N SCa – Used or
Global parameter (WiMAX 802.16d) Number of used subcarriers per
None
PZ
N SCa – Used Permutation zone parameter (WiMAX 802.16e) channel
N SCa – Data or
Global parameter (WiMAX 802.16d) Number of subcarriers per channel
None
PZ
N SCa – Data Permutation zone parameter (WiMAX 802.16e) used for data transfer
Number of DC subcarriers per

N SCa – DC Hard-coded parameter ( N SCa – DC = 1 ) None
channel
N SCa – Pilot or Calculation result ( N SCa – Pilot = N SCa – Used – N SCa – Data or
Number of pilot subcarriers per
None
PZ
N SCa – Pilot
PZ
N SCa – Pilot =
PZ
N SCa – Used –
PZ
N SCa – Data ) channel
Calculation result
N SCa – Guard or
( N SCa – Guard = N SCa – Total – N SCa – Used – N SCa – DC or Number of guard subcarriers per
None
PZ
N SCa – Guard PZ PZ
channel
N SCa – Guard = N SCa – Total – N SCa – Used – N SCa – DC )
PZ UL Permutation zone parameter (WiMAX 802.16e) None Uplink permutation zone
PZ DL Permutation zone parameter (WiMAX 802.16e) None Downlink permutation zone
Required signal quality for accessing

CNR Req – PZ Permutation zone parameter (WiMAX 802.16e) dB
a permutation zone
Speed limit for mobiles trying to
Speed Max – PZ Permutation zone parameter (WiMAX 802.16e) Km/hr
access a permutation zone
Maximum distance from the
d Max – PZ Permutation zone parameter (WiMAX 802.16e) m transmitter covered by a permutation
zone
p PZ Permutation zone parameter (WiMAX 802.16e) None Permutation zone priority
Segmented permutation zone parameter (WiMAX 802.16e) Ratio of channel bandwidth allocated
f Segment 3  PSG + 2  SSG None to each segment of a segmented
f Segment = ----------------------------------------------------
15 frame configuration
W Channel Frequency band parameter MHz Channel bandwidth
First First channel number of the frequency

N Channel Frequency band parameter None
band
Last Last channel number of the frequency
N Channel Frequency band parameter None
band
Start frequency of the TDD frequency
F Start – FB – TDD Frequency band parameter MHz
band
DL Start frequency of the FDD
F Start – FB – FDD – DL Frequency band parameter MHz
frequency band
UL Start frequency of the FDD
F Start – FB – FDD – UL Frequency band parameter MHz
frequency band
f Sampling Frequency band parameter None Sampling factor
f ACS – FB Frequency band parameter dB Adjacent Channel Suppression 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
bits/
B Bearer parameter Bearer Efficiency
symbol
434 AT271_TRG_E6 © Forsk 2009

TB Bearer parameter dB Bearer selection threshold
K 1.38 x 10-23 J/K Boltzmann’s constant

TX Transmitter parameter (user-defined or calculated from transmitter

nf dB Transmitter noise figure
Number of antennas used for MIMO in
N Ant – TX Transmitter parameter None
transmission
N Ant – RX Transmitter parameter None
reception
TX Antenna parameter dB Transmitter antenna gain
G
TX Transmitter parameter (user-defined or calculated from transmitter

L dB Transmitter loss
N Channel Cell parameter None Cell’s channel number
P Preamble Cell parameter dBm Preamble power
Cell parameter P Traffic = P Preamble – P Traffic in dB

P Traffic P Traffic dB Traffic power offset
------------------------
Ratio 10
P Traffic = 10 in %
Cell parameter P Pilot = P Preamble – P Pilot in dB

P Pilot P Pilot dB Pilot power offset
-------------------
Ratio 10
P Pilot = 10 in %
Cell parameter P Idle – Pilot = P Preamble – P Idle – Pilot in dB

P Idle – Pilot P Idle – Pilot dB Idle pilot power offset
---------------------------------
Ratio 10
P Idle – Pilot = 10 in %
TL DL Cell parameter % Downlink traffic load
TL UL Cell parameter % Uplink traffic load
NR UL Cell parameter dB Uplink noise rise
N Users – Max Cell parameter None Maximum number of users per cell
SU Cell parameter % Segmentation usage ratio
AU Cell parameter % AAS usage ratio
T AMS Cell parameter dB Adaptive MIMO switch threshold
PI Cell parameter None Preamble index
T Preamble Cell parameter dB Preamble quality threshold
Channel and preamble index reuse

D Reuse Cell parameter m
distance
QoS Service parameter None QoS class of the service
p Service parameter None Service priority

Highest bearer used by a service in
B DL – Highest Service parameter None
the downlink
Highest bearer used by a service in
B UL – Highest Service parameter None
the uplink
UL
f Act Service parameter % Uplink activity factor for voice services
DL Downlink activity factor for voice

f Act Service parameter %
services
Minimum throughput demand in the
TPD Min – UL Service parameter kbps
uplink
Minimum throughput demand in the
TPD Min – DL Service parameter kbps
downlink
Maximum throughput demand in the
TPD Max – UL Service parameter kbps
uplink
© Forsk 2009 AT271_TRG_E6 435

Maximum throughput demand in the

TPD Max – DL Service parameter kbps
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
transmission
N Ant – RX Terminal parameter None
reception
UL
G STTD Clutter parameter dB Uplink STTD/MRC gain offset
DL
G STTD Clutter parameter dB Downlink STTD/MRC gain offset
f SM – Gain Clutter parameter None Spatial Multiplexing gain factor
L Indoor Clutter parameter dB Indoor loss
L Path Propagation model result dB Path loss
M Shadowing – Model dB
standard deviation Only used in coverage predictions
Result calculated from cell edge coverage probability and C/I standard C/I Shadowing margin
M Shadowing – C  I dB
deviation Only used in coverage predictions
Rho factor
% of BTS signal correctly transmitted
Power spectral density of thermal
n0 Calculation result ( 10  Log 10  K  T  1000  = – 174 dBm/Hz ) dBm/Hz
noise
QoS
f Bias Scheduler parameter % QoS class bias factor
Maximum spatial-multiplexing gain for

Max
G SM MIMO configuration parameter None a given number of transmission and
reception antennas
Uplink STTD/MRC gain for a given
UL
G STTD MIMO configuration parameter dB number of transmission and reception
antennas
Downlink STTD/MRC gain for a given
DL
G STTD MIMO configuration parameter dB number of transmission and reception
antennas
9.1.2 Co- and Adjacent Channel Overlaps Calculation

TX i  ic  TX i  ic  TX i  ic  TX i  ic  Start frequency for the channel
F Start F Start – FB + N Channel  W Channel MHz
number assigned to a cell
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  End frequency for the channel number
F End F Start – FB + N Channel  W Channel + W Channel MHz
assigned to a cell
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX j  jc  TX i  ic 
W CCO Min  FEnd  F End  – Max  FStart  F Start  MHz Co-channel overlap bandwidth
TX i  ic  – TX j  jc 
TX i  ic  – TX j  jc  W CCO
r CCO --------------------------------------- None Co-channel overlap ratio
TX i  ic 
W Channel
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX j  jc  TX i  ic  TX i  ic  Bandwidth of the lower-frequency

W ACO Min  F End  F Start  – Max  F Start  F Start – W Channel  MHz
L adjacent channel overlap
436 AT271_TRG_E6 © Forsk 2009

TX i  ic  – TX j  jc  W ACO Lower-frequency adjacent channel
L
r ACO --------------------------------------- None
L TX i  ic  overlap ratio
W Channel
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX i  ic  TX j  jc  TX i  ic  Bandwidth of the higher-frequency

W ACO Min  F End  F End + W Channel  – Max  F Start  F End  MHz
H adjacent channel overlap
TX i  ic  – TX j  jc  W ACO Higher-frequency adjacent channel
H
r ACO --------------------------------------- None
H TX i  ic  overlap ratio
W Channel
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
r ACO r ACO + r ACO None Adjacent channel overlap ratio
L H
TDD
r DL – Frame
--------------------------
- if interferer uses a TDD frequency band and victim uses
100 None FDD – TDD overlap ratio
r FDD – TDD
an FDD frequency band
1 other wise
TX  ic 
i
 – f ACS – FB
 TXi  ic  – TXj  jc  TX i  ic  – TX j  jc  ---------------------------
10 TX i  ic  – TX j  jc 
 r CCO + r ACO  10   r FDD – TDD
 
 
if W Channel  W Channel
rO None Total overlap ratio
TX  ic 
i
 – f ACS – FB TX  ic 
 TXi  ic  – TXj  jc  TXi  ic  – TXj  jc  --------------------------- TX  ic  – TX  jc  W i
10 i j Channel
 r CCO + r ACO 10  r FDD – TDD -----------------------
TX j  jc 
-
  W Channel
 
if W Channel  W Channel
9.1.3 Preamble Signal Quality Calculations

9.1.3.1 Preamble Signal Level Calculation
TX i  ic  Mi
TX i  ic  EIRP Preamble – L Path – M Shadowing – Model – L Indoor + G
C Preamble Mi Mi Mi
dBm Received preamble signal level
–L – L Ant – L Body
TX i  ic  TX i TX i
Without smart antenna equipment: P Preamble + G –L
TX i  ic 
EIRP Preamble With smart antenna equipment: dBm Preamble EIRP of a cell
TX i  ic  TX i TX i TX i
P Preamble +G –L + 10  Log  E SA 
TX i
L Path L Model + L Ant dB Path loss
TX i TX i Mi Mi
L Path + L + L Indoor + M Shadowing – Model – G +L –G
L Total Mi Mi
dB Total losses
+ L Ant + L Body
© Forsk 2009 AT271_TRG_E6 437

9.1.3.2 Preamble Noise Calculation

TX i  ic 
 TX  ic  N SCa – Preamble
n 0 + 10  Log 10  F Sampling  ---------------------------------------
i
-
 TX i  ic  
 N SCa – Total 
TX i  ic 
n 0 – Preamble With Segmentation (WiMAX 802.16e): dBm Preamble thermal noise for a cell
TX i  ic 
 TX  ic  N SCa – Preamble 
n 0 + 10  Log 10  F Sampling   f Segment 
i Preamble
----------------------------------------
 TX i  ic  
Preamble 1
f Segment --- None Preamble segmenting factor
3
TX i  ic  TX i  ic  Mi
n Preamble n 0 – Preamble + nf dBm Preamble noise for a cell
9.1.3.3 Preamble Interference Calculation

TX j  jc  TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  Total interference generated by an
I Preamble C Preamble + f O + f Seg dBm
interfering cell
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  Interference reduction factor due to
fO 10  Log  r O  dB
the co- and adjacent channel overlap
f Seg 10  Log 10  p Collision  dB
segmentation (WiMAX 802.16e)
9.1.3.4 Preamble C/N Calculation

CNR Preamble C Preamble – n Preamble dB Preamble C/N for a cell
9.1.3.5 Preamble C/(I+N) Calculation

 TX  jc  
j
TX i  ic  TX i  ic   I Preamble TX i  ic 

------------------------- dB Preamble C/(I+N) for a cell
CINR Preamble C Preamble – 10  Log  10  –n Preamble – M Shadowing – C  I
 10 
 All TXj  jc  
TX  jc  TX  ic 
j i
 I Preamble n Preamble
TX i  ic   ------------------------- ---------------------------

10 10 dBm Preamble Total Noise (I+N) for a cell
I + N  Preamble 10  Log  10 + 10 
 
 All TX j  jc  
9.1.4 Traffic and Pilot Signal Quality Calculations

9.1.4.1 Traffic and Pilot Signal Level Calculation (DL)
TX i  ic  Mi
TX i  ic  EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G
C Traffic Mi Mi Mi
dBm Received traffic signal level
438 AT271_TRG_E6 © Forsk 2009

TX i  ic  Mi
TX i  ic  EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G
C Pilot Mi Mi Mi
dBm Received pilot signal level
EIRP Traffic P Traffic + G –L dBm Traffic EIRP of a cell
EIRP Pilot P Pilot +G –L dBm Pilot EIRP of a cell
P Traffic P Preamble – P Traffic dBm Traffic transmission power of a cell
P Pilot P Preamble – P Pilot dBm Pilot transmission power of a cell
9.1.4.2 Noise Calculation (DL)

TX i  ic  N SCa – Used
WiMAX 802.16d: n 0 + 10  Log 10  F Sampling  ------------------------------
 N SCa – Total
M
 PZ DL
i

 TXi  ic  N SCa – Used
WiMAX 802.16e: n 0 + 10  Log 10  F Sampling  -----------------------------
TX i  ic 
-
TX i  ic   N SCa – Total
n 0 – DL   dBm Thermal noise for a cell
With Segmentation (WiMAX 802.16e):
M
 PZ DL
i

 TX i  ic  N SCa – Used 
n 0 + 10  Log 10  F Sampling  -----------------------------
TX i  ic 
-  f Segment
 N SCa – Total 
 
f Segment 3  PSG + 2  SSG

---------------------------------------------------- None Segmenting factor
15
n DL n 0 – DL + nf dBm Downlink noise for a cell
9.1.4.3 Traffic and Pilot Interference Calculation (DL)

TX  jc  TX  jc 
j j
 I Non – AAS I Idle 
 --------------------------
10
- ------------------
10
Monte Carlo Simulations: 10  Log  10 + 10  without
 
 
TX  jc 
j
 I AAS 
TX j  jc   -----------------
10 
- Total interference generated by an
I Total smart antennas, or 10  Log  10  with smart antennas dBm
interfering cell
 
 
TX  jc  TX  jc  TX  jc 
j j j
 I Non – AAS I Idle I AAS 
 --------------------------
10
- ------------------
10
------------------
10
Coverage Predictions: 10  Log  10 + 10 + 10 
 
 
TX j  jc  TX j  jc  Mi Mi Mi Mi Traffic interference power of an

I Traffic EIRP Traffic – L Path – L Indoor + G –L – L Ant – L Body dBm
interfering cell
TX j  jc  TX j  jc  Mi Mi Mi Mi Pilot interfering power of an interfering
I Pilot EIRP Pilot – L Path – L Indoor + G –L – L Ant – L Body dBm
cell
TX j  jc  TX j  jc  TX j TX j
EIRP Traffic P Traffic + G –L dBm Traffic EIRP of an interfering cell
EIRP Pilot P Pilot +G –L dBm Pilot EIRP of an interfering cell

 TXj  jc  TX j  jc 
10  Log  TL DL   1 – AU  Interference from the loaded part of

TX j  jc   the frame transmitted using the
I Non – AAS dBm
TX j  jc  TXj  jc  transmitter antenna of an interfering
 I Traffic TX j  jc  I Pilot TX  jc  TX j  jc  
 ----------------- ------------------ N j cell
SCa – Used – N SCa – Data 
- N SCa – Data
10
 10 - + 10 10  ----------------------------------------------------------------
 ----------------------------- - 
 N SCa – Used N SCa – Used 
 
© Forsk 2009 AT271_TRG_E6 439

TX j  jc  TX j  jc  Mi Mi Mi Mi Interference power of an interfering

I AAS EIRP AAS – L Path – L Indoor + G –L – L Ant – L Body dBm
cell transmitted using smart antenna
TX j  jc  TX j  jc  TX j TX j Traffic EIRP of an interfering cell using
EIRP AAS P Traffic + G –L dBm
smart antenna
Interference from empty part of the
TX j  jc  TX j  jc  Mi Mi Mi Mi frame transmitted using the
I Idle – Pilot EIRP Idle – Pilot – L Path – L Indoor + G –L – L Ant – L Body dBm
transmitter antenna of an interfering
cell
EIRP Idle – Pilot P Idle – Pilot + G –L dBm Idle pilot EIRP of an interfering cell
TX  jc 
  I Idle – Pilot
j
TX  jc  TX j  jc   Interference from the empty part of the
 - N j
 --------------------------
TX j  jc  TX j  jc  10 SCa – Used – N SCa – Data  dBm
frame transmitted using the
I Idle 10  Log   1 – TL DL    10  ----------------------------------------------------------------
TX j  jc 
-  transmitter antenna of an interfering
  N SCa – Used 
   cell

f Seg 10  Log 10  p Collision  dB
segmentation (WiMAX 802.16e)
9.1.4.4 Traffic and Pilot C/N Calculation (DL)

C Traffic – n DL
With MIMO (STTD/MRC):
TX i  ic  TX i  ic  DL DL
TX i  ic  CNR Traffic = CNR Traffic + G STTD + G STTD dB Traffic C/N for a cell
CNR Traffic
With MIMO (AMS) if CNR Preamble  T AMS :
CNR Traffic = CNR Traffic + G STTD + G STTD
C Pilot – n DL
With MIMO (STTD/MRC):
TX i  ic  CNR Pilot = CNR Pilot + G STTD + G STTD dB Pilot C/N for a cell
CNR Pilot
CNR Pilot = CNR Pilot + G STTD + G STTD
9.1.4.5 Traffic and Pilot C/(I+N) Calculation (DL)

 TX  jc 
j TX i  ic 
TX i  ic   I DL n DL
-

-----------------
- -------------------
C Traffic – 10  Log  10 + 10  – M Shadowing – C  I
 10 10 
TX i  ic  With MIMO (STTD/MRC):
CINR Traffic TX i  ic  TX i  ic 
dB Traffic C/(I+N) for a cell
DL DL
CINR Traffic = CINR Traffic + G STTD + G STTD
 TX  jc 
TX i  ic   I DL n DL
-

-----------------
- -------------------
C Pilot – 10  Log  10 + 10  – M Shadowing – C  I
 10 10 
TX i  ic  With MIMO (STTD/MRC):
CINR Pilot TX i  ic  TX i  ic 
dB Pilot C/(I+N) for a cell
DL DL
CINR Pilot = CINR Pilot + G STTD + G STTD
CINR Pilot = CINR Pilot + G STTD + G STTD
440 AT271_TRG_E6 © Forsk 2009

j i
 I DL n DL 
TX i  ic   ------------------ --------------------

10 10 dBm Traffic Total Noise (I+N) for a cell
 I + N  DL 10  Log  10 + 10 
 
9.1.4.6 Traffic Signal Level Calculation (UL)

Mi TX i
Mi EIRP UL – L Path – M Shadowing – Model – L Indoor + G
C UL TX i Mi Mi
dBm Received uplink signal level
Mi Mi Mi
P +G –L
Mi
EIRP UL Mi Mi Mi Mi dBm Uplink EIRP of a user equipment
With P = P Max without power control and P = P Eff after power
control
9.1.4.7 Noise Calculation (UL)

TX i  ic  N SCa – Used
WiMAX 802.16d: n 0 + 10  Log 10  F Sampling  ------------------------------
TX i  ic  M
n 0 – UL  PZ UL
i
 dBm Thermal noise for a cell
 TXi  ic  N SCa – Used
WiMAX 802.16e: n 0 + 10  Log 10  F Sampling  -----------------------------
TX i  ic 
-
 N SCa – Total
 
n UL n 0 – UL + nf dBm Uplink noise for a cell
9.1.4.8 Traffic Interference Calculation (UL)

Mj Mj TX i  ic  – TX j  jc  Mj
I UL C UL + f O + f TL – UL dBm Uplink interference received at a cell

Interference reduction factor due to
Mj Mj
f TL – UL 10  Log 10  TL UL  dB the interfering mobile’s uplink traffic
load
 Mj TX i  ic 
 I UL n UL 
 -------
-
10 
-------------------
- TX i  ic 
 Uplink noise at a cell without smart
TX i  ic  10
NR UL 10  Log 10  10 + 10  – n UL dB
  antenna
 All M j 
 All TX  jc  
j
2
I UL    +  n  I Uplink noise at a cell with smart
NR UL    ----------------------------------- dB
2
n  I antenna
 M
j
TX  ic 
i
 I UL n UL 
 -------
-
10 
-------------------
-

TX i  ic  10
 I + N  UL 10  Log 10  10 + 10  dBm Total Noise (I+N) for a cell
 
 All Mj 
j
9.1.4.9 Traffic C/N Calculation (UL)

© Forsk 2009 AT271_TRG_E6 441

Mi TX i  ic 
C UL – n UL
Mi Mi UL UL
Mi
With MIMO (STTD/MRC): CNR UL = CNR UL + G STTD + G STTD
CNR UL dB Uplink C/N at a cell
Mi Mi UL UL
CNR UL = CNR UL + G STTD + G STTD
9.1.4.10 Traffic C/(I+N) Calculation (UL)

Mi TX i  ic 
Without smart antenna: CNR UL – NR UL – M Shadowing – C  I
Mi TX i  ic 
With smart antenna: CNR UL – NR UL  – M Shadowing – C  I
Mi Mi Mi UL UL
CINR UL With MIMO (STTD/MRC): CINR UL = CINR UL + G STTD + G STTD dB Uplink C/(I+N) at a cell
Mi Mi UL UL
CINR UL = CINR UL + G STTD + G STTD
9.1.5 Channel Throughput Calculation

9.1.5.1 Calculation of Cell Capacity (Total Resources)
TX i  ic  6
TX i  ic   W Channel  10 
F Sampling Floor  f Sampling  ----------------------------------------  8000 Hz Sampling frequency
 8000 
TX i  ic  –3
TX i  ic  F Sampling  10
F ------------------------------------------- kHz Inter-subcarrier distance
TX i  ic 
N SCa – Total
TX i  ic  1
---------------------- ms Useful symbol duration
D Sym – Useful TX i  ic 
F
r CP
D CP -------- ms Cyclic prefix duration
F
D Symbol D Sym – Useful + D CP ms Symbol duration
Used TDD TDD

D Frame D Frame – D TTG – D RTG ms Used frame duration
 D Used 
TX i  ic 
Floor  --------------------
Frame 
- Frame duration in terms of symbol
N  SD – Used   Frame  TXi  ic   SD
durations
 D Symbol
TX i  ic  TX i  ic  TDD DL Downlink subframe duration in terms

N  SD – DL   Subframe RoundUp  N  SD – Used   Frame  r DL – Frame  – O Fixed SD
of symbol durations
WiMAX 802.16d:
DL
 TX i  ic  TX i  ic   O Variable 
TX i  ic  Floor  N  SD – DL   Subframe  N SCa – Data   1 – -----------------------
R DL =   100   Total downlink cell resources, i.e., the
Symbols number of symbols in the downlink
TX i  ic  WiMAX 802.16e:
N  Sym – DL   Subframe subframe
Mi DL
 TX i  ic  PZDL  O Variable 
Floor  N  SD – DL   Subframe  N SCa – Data   1 – -----------------------
  100  
TX i  ic  TX i  ic  TDD UL Uplink subframe duration in terms of

N  SD – UL   Subframe RoundDown  N  SD – Used   Frame   1 – r DL – Frame   – O Fixed SD
symbol durations
442 AT271_TRG_E6 © Forsk 2009

WiMAX 802.16d:
UL
 TXi  ic  TX i  ic   O Variable 
TX i  ic  Floor  N  SD – UL   Subframe  N SCa – Data   1 – -----------------------
R UL =   100   Total uplink cell resources, i.e., the
Symbols number of symbols in the uplink
TX i  ic  WiMAX 802.16e:
N  Sym – UL   Subframe subframe
M UL
i
 TXi  ic  PZUL  O Variable 
Floor  N  SD – UL   Subframe  N SCa – Data   1 – -----------------------
  100  
9.1.5.2 Bearer Determination

 Mi  Bearer assigned to the pixel,

Mi Min  Best Bearers  B DL – Highest  Service 
B DL M TX  ic  None subscriber, or mobile Mi in the
 TB 
i i
CINR Traffic 
downlink
Mi
 Mi  Bearer assigned to the pixel,
B UL Min  Best Bearers TX  ic  M
 B UL – Highest  Service  None
 TB
i

i
CINR UL  subscriber, or mobile Mi in the uplink
9.1.5.3 Channel Throughput Calculation

TX i  ic 
R DL  Mi
DL B
Without segmentation: ------------------------------------
-
D Frame
TX i  ic 
R DL  Mi
DL B
Mi -  f Segment
With segmentation: ------------------------------------ Downlink peak MAC channel
CTP P – DL D Frame kbps
throughput
Max
With MIMO (SM):  Mi =  Mi   1 + f SM – Gain  G SM – 1  
B DL B DL
With MIMO (AMS):

Max TX i  ic  TX i  ic 
 M =  M   1 + f SM – Gain  G SM – 1   if CNR Preamble  T AMS
i i
B DL B DL
Mi Mi Mi Downlink effective MAC channel

CTP E – DL CTP P – DL   1 – BLER  B DL   kbps
throughput
Mi
Mi Mi f TP – Scaling Mi Downlink application level channel
CTP A – DL CTP E – DL  ----------------------------- – TP Offset kbps
100 throughput
TX i  ic 
R UL  M
i
UL B
Without segmentation: ------------------------------------
-
D Frame
Mi Max
CTP P – UL With MIMO (SM):  M =  M   1 + f SM – Gain  G SM – 1   kbps Uplink peak MAC channel throughput
i i
B UL B UL
With MIMO (AMS):

 Mi =  Mi   1 + f SM – Gain  G SM – 1   if CNR Preamble  T AMS
B UL B UL
Mi Mi Mi Uplink effective MAC channel

CTP E – UL CTP P – UL   1 – BLER  B UL   kbps
throughput
Mi
Mi Mi f TP – Scaling Mi Uplink application level channel
CTP A – UL CTP E – UL  ----------------------------- – TP Offset kbps
100 throughput
9.1.6 Scheduling and Radio Resource Management

Sel
Mi
Sel TPD Min – DL Resources allocated to a mobile to
Mi ------------------------------
-
R Min – DL Sel
None satisfy its minimum throughput
Mi
CTP P – DL demand in downlink
© Forsk 2009 AT271_TRG_E6 443

Sel
Mi
Sel TPD Min – UL Resources allocated to a mobile to
Mi ------------------------------
-
R Min – UL Sel
None satisfy its minimum throughput
Mi
CTP P – UL demand in uplink
Sel
Mi Remaining downlink cell resources
TX i  ic 
R Rem – DL
100 % –  RMin – DL None after allocation for minimum
Mi
Sel throughput demands
Sel
Mi Remaining uplink cell resources after
TX i  ic 
R Rem – UL
100 % –  RMin – UL None allocation for minimum throughput
Mi
Sel demands
Mi
Sel
Mi
Sel
Mi
Sel Remaining throughput demand for a
TPD Rem – DL TPD Max – DL – TPD Min – DL kbps
mobile in downlink
Mi
Sel
Mi
Sel
Mi
Sel Remaining throughput demand for a
TPD Rem – UL TPD Max – UL – TPD Min – UL kbps
mobile in uplink
Sel
Mi
Mi
Sel TPD Rem – DL Remaining resource demand for a
RD Rem – DL --------------------------------- None
Mi
Sel mobile in downlink
CTP P – DL
Sel
Mi
Mi
Sel TPD Rem – UL Remaining resource demand for a
RD Rem – UL --------------------------------- None
Mi
Sel mobile in uplink
CTP P – UL
Sel
TX i  ic 
 Mi R Rem – DL
Proportional Fair: Min  RD Rem – DL -------------------------
 N 
Sel
Mi
TX i  ic  RD Rem – DL
Proportional Demand: R Eff – Rem – DL  ---------------------------------------
Sel
-
Mi
Sel
 RDRem – DL Resources allocated to a mobile to
Mi Sel
R Max – DL Mi None satisfy its maximum throughput
TX i  ic  demand in downlink
Sel
 Mi R QoS – DL
Biased (QoS Class): Min  RD Rem – DL ------------------------
 N QoS 
Sel
Mi
TPD Rem – DL
Max Aggregate Throughput: ---------------------------------
Sel
Mi
CTP P – DL
Sel
TX i  ic 
 Mi R Rem – UL
Proportional Fair: Min  RD Rem – UL -------------------------
 N 
Sel
Mi
TX i  ic  RD Rem – UL
Proportional Demand: R Eff – Rem – UL  ---------------------------------------
Sel
-
Mi
Sel
 RDRem – UL Resources allocated to a mobile to
Mi Sel
R Max – UL Mi None satisfy its maximum throughput
TX i  ic  demand in uplink
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  ic  Sel
Mi
 Effective remaining downlink
Min  R Rem – DL  RD Rem – DL
TX i  ic  i
R Eff – Rem – DL   None resources in a cell
 Sel  (Proportional Demand)
Mi
Mi
 Effective remaining uplink resources
Min  R Rem – UL  RD Rem – UL
TX i  ic  i
R Eff – Rem – UL   None in a cell
 Sel  (Proportional Demand)
Mi
Sel Sel Sel

QoS Mi Mi Mi
f Bias R Max – ErtPS R Max – rtPS R Max – nrtPS QoS class bias
 1 + ----------
- = ------------------------------
- = ------------------------------ = ------------------------------ None
100 Mi
Sel
Mi
Sel
Mi
Sel (Biased (QoS Class))
R Max – rtPS R Max – nrtPS R Max – BE
444 AT271_TRG_E6 © Forsk 2009

r
1 QoS
N QoS   --- Remaining downlink cell resources
TX i  ic   
TX i  ic  R Rem – DL  ------------------------------------------------------------ after allocation for minimum
R QoS – DL r None
1 QoS throughput demands for a QoS class
N QoS   ---
   (Biased (QoS Class))
All QoS
r
1 QoS
N QoS   --- Remaining downlink cell resources
TX i  ic   
TX i  ic  R Rem – UL  ------------------------------------------------------------ after allocation for minimum
R QoS – UL r None
1 QoS throughput demands for a QoS class
N QoS   ---
   (Biased (QoS Class))
All QoS
Sel Sel Sel Sel

Total resources assigned to a mobile
Mi Mi Mi Mi
TL DL = R DL R Min – DL + R Max – DL None in downlink
(Downlink traffic load of the mobile)
Sel Sel Sel Sel

Total resources assigned to a mobile
Mi Mi Mi Mi
TL UL = R UL R Min – UL + R Max – UL None in uplink
(Uplink traffic load of the mobile)
9.1.6.1 User Throughput Calculation

Sel Sel Sel
Mi Mi Mi
UTP P – DL R DL  CTP P – DL kbps Downlink peak MAC user throughput
Sel Sel
Sel Mi Mi Downlink effective MAC user
UTP P – DL   1 – BLER  B DL  
Mi
UTP E – DL kbps
   throughput
Sel
Mi
Mi
Sel
Mi
Sel
f TP – Scaling Sel
Mi
Downlink application level user
UTP A – DL kbps
UTP E – DL  ----------------------------- – TP Offset throughput
100
Sel Sel Sel
Mi Mi Mi
UTP P – UL R UL  CTP P – UL kbps Uplink peak MAC user throughput
Sel Sel Sel

Mi Mi
UTP P – UL   1 – BLER  B UL  
Mi
UTP E – UL kbps Uplink effective MAC user throughput
  
Sel
Mi
Mi
Sel
Mi
Sel
f TP – Scaling Sel
Mi
Uplink application level user
UTP A – UL kbps
UTP E – UL  ----------------------------- – TP Offset throughput
100
9.1.7 Optimum Beamformer Model

E SA Smart antenna model parameter None Number of smart antenna elements
 Calculation parameter Degrees Angle of arrival for the useful signal

Angle at which the smart antenna
 Calculation parameter Degrees
effect is calculated
 Distance between two adjacent
d --- , where  is the wavelength of the signal m
2 antenna elements
gn    Smart antenna model parameter None Gain of a single element
2 2 2 T
   None Steering vector for the direction of 
S 1 e e  ... e
2
– j  -------  nd  sin 

e
wn None Complex smart antenna weight
– j    n  sin  
e with d = ---
2
H Array correlation matrix for a given

R R = S  S None
user direction 
H H H
gn     S  R  S = gn     S  S  S  S
G SA    None Smart antenna gain in any direction 
10  Log  G SA     in dB
© Forsk 2009 AT271_TRG_E6 445

R Avg  j  pj  Rj None Average array correlation matrix

j=1
–1 Vector of ESA complex weights for the

ŵ   RN  S None
beam former
E SA
 ----------------------------------
- None Smart antenna gain in the uplink
H –1
S  RN  S
J
H
2
RN Rn + RI = n I+ None Total noise correlation matrix
j=1
2
Rn n  I None Thermal noise correlation matrix
J
H
RI
 pj  Sj  Sj None Interference correlation matrix
j=1
Total noise power, including thermal

Pˆ N
2 H –1
  S  RN  S W noise and interference from all uplink
interferers
2 Total power received from the served
Pˆ  2 H –1
p     S  RN  S  W
user
2 H –1 2
Pˆ p     S  RN  S  H –1
CINR UL ------- = --------------------------------------------------------------- = p  S  RN  S None C/(I+N) in the uplink
Pˆ 2 H –1
N   S  RN  S
H
S   I  S  = E SA Uplink smart antenna beam forming
G SA None
10  Log  E SA  in dB gain in the direction of the served user
K
–1 1 Average of the inverse noise
 RN
–1
RN ----  W
Avg K k correlation matrices
k=1
E SA 2
I UL    - – n
------------------------------------------ W Uplink interference
H –1
S  RN  S
Avg
2
I UL    +  n Angular distribution of uplink noise
NR UL    ----------------------------- None
n
2 rise
9.2 Calculation Processes

The following sections describe the processes of different calculations performed in Atoll and their results.

Point analysis profile tab and reception tab display the following calculation results based on the calculation algorithm
described in "Preamble Signal Level Calculation" on page 466.
TX i  ic 
• Preamble signal level C Preamble
• Path loss L Path
• Total losses L Total
Note:
Mi Mi Mi Mi
• L , G , L Ant , and L Body are not considered in the calculations performed for the point
analysis tool.
446 AT271_TRG_E6 © Forsk 2009

9.2.1.1 Profile Tab

TX i  ic 
Atoll displays the preamble signal level received from the studied cell, C Preamble , which you select on the map or in the
list of transmitters. You can also choose to display the path loss, L Path , or the total losses, L Total for a selected cell.

Analysis provided in the Reception tab is based on path loss matrices. So you can display received preamble signal levels
from the cells for which path loss matrices have been calculated over their calculation areas. For each cell, Atoll displays
TX i  ic 
the received preamble signal level, C Preamble .
Reception level bar graphs show the preamble signal levels in a decreasing order. The maximum number of bars in the
graph depends on the preamble signal level received from the best server. The bar graph displays cells whose received
preamble signal level is within a 30 dB margin from the highest preamble signal level.
Note:
• You can use a value other than 30 dB for the margin from the highest preamble signal
level, for example a smaller value for improving the calculation speed. For more
information on defining a different value for this margin, see the Administrator Manual.
9.2.2 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
These coverage predictions can be performed at any stage of network planning because they do not depend on the traffic
input. Therefore, these calculations are of special interest before and during the deployment stage of the network to study
the coverage footprint of the system.
TX i  ic 
For these calculations, Atoll calculates the received preamble signal level ( C Preamble ) as explained in "Preamble Signal
Level Calculation" on page 466. Then, Atoll determines the selected display criterion on each pixel inside the cell’s calcu-
lation area. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver.
Coverage prediction parameters to be set are:

• The coverage prediction conditions to determine the coverage area of each studied cell, and
• The display settings to colour the coverage areas.
Note:
Mi Mi Mi Mi
• L , G , L Ant , and L Body are not considered in the calculations performed for the
preamble signal level based coverage predictions.
The following sections describe the determination of coverage area of each cell ("Coverage Area Determination" on
page 447), and the display options ("Coverage Display" on page 448) of the coverage predictions.

Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine coverage
areas to display. There are three possibilities.

The coverage area of each cell TXi(ic) corresponds to the pixels where.
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). The Best function considers the highest value from a list of values.
© Forsk 2009 AT271_TRG_E6 447

Notes:
• If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is
the highest.
either the highest or within a 2 dB margin from the highest.
• If M = -2 dB, 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.

AND
TX  ic  nd TX  jc 
C Preamble  2 Best  C j
i
Preamble  – M
ji
Where M is the specified margin (dB). The 2nd Best function considers the second highest value from a list of values.
Notes:
the second highest.
either the second highest or within a 2 dB margin from the second highest.
• If M = -2 dB, 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.

9.2.2.2.1 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. Coverage predictions are generated using a bilinear
interpolation method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes).

It is possible to display the coverage predictions with colours depending on any transmitter or cell attribute, and other crite-
ria such as:
Signal Level (dBm, dBµV, dBµV/m)
Atoll calculates preamble signal levels received from cells on each pixel of the cells’ coverage areas. A pixel of a coverage
area is coloured if the preamble signal level exceeds (  ) the defined minimum thresholds (pixel colour depends on
received preamble signal level). Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as cell coverage areas. Each layer shows the different preamble signal levels
received in the cell coverage area.
Best Signal Level (dBm, dBµV, dBµV/m)
Atoll calculates preamble signal levels received from cells on each pixel of the cells’ coverage areas. Where other cover-
age areas overlap the studied one, Atoll chooses the highest value. A pixel of a coverage area is coloured if the preamble
signal level exceeds (  ) the defined thresholds (the pixel colour depends on the preamble signal level). Coverage consists
of several independent layers whose visibility in the workspace can be managed. There are as many layers as defined
thresholds. Each layer corresponds to an area where the preamble signal level from the best server exceeds a defined
threshold.
Path Loss (dB)
Atoll calculates path losses from cells on each pixel of the cells’ coverage areas. A pixel of a coverage area is coloured if
the path loss exceeds (  ) the defined minimum thresholds (pixel colour depends on path loss). Coverage consists of
several independent layers whose visibility in the workspace can be managed. There are as many layers as cell coverage
areas. Each layer shows different path loss levels in the cells’ coverage area.
448 AT271_TRG_E6 © Forsk 2009

Total Losses (dB)
Atoll calculates total losses from cells on each pixel of the cells’ coverage areas. A pixel of a coverage area is coloured if
total losses exceed (  ) 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 cell coverage
areas. Each layer shows different total losses levels in the cells’ coverage areas.
Atoll calculates preamble signal levels received from cells on each pixel of the cells’ coverage areas. Where other cover-
age areas overlap the studied one, Atoll determines the best cell (i.e., the cell with the highest preamble signal level) and
evaluates path loss from this cell. A pixel of a coverage area is coloured if the path loss exceeds (  ) the defined thresholds
(pixel colour depends on path loss). Coverage consists of several independent layers whose visibility in the workspace
can be managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the path loss
from the best server exceeds a defined threshold.
Atoll calculates preamble signal levels received from cells on each pixel of the cells’ coverage areas. Where coverage
areas overlap the studied one, Atoll determines the best cell (i.e., the cell with the highest preamble signal level) and eval-
uates total losses from this cell. A pixel of a coverage area is coloured if the total losses exceed (  ) the defined 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 defined thresholds. Each layer corresponds to an area where the total
losses from the best server exceed a defined threshold.
Number of Servers
Atoll evaluates the number of cells that cover a pixel (i.e., the pixel falls within the coverage areas of these cells). The pixel
colour depends on the number of servers. Coverage consists of several independent layers whose visibility in the work-
space can be managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the
number of servers exceeds (  ) a defined threshold.
On each pixel of a cell’s coverage area, the coverage corresponds to the pixels where the received preamble signal level
from this cell fulfils signal conditions defined in Conditions tab with different cell edge coverage probabilities. Coverage
the number of cells.
On each pixel of each cell coverage area, the coverage corresponds to the pixels where the best preamble signal level
received fulfils signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability.
9.2.3 Signal Quality Analysis Coverage Predictions

The following coverage predictions are based on the received preamble, traffic, or pilot signal levels and noise, and take
Mi Mi Mi Mi
into account the receiver characteristics ( L , G , L Ant , and L Body ) when calculating the required parameter.
• Signal Quality Analysis (DL)

• Signal Quality Analysis (UL)
For these calculations, Atoll calculates the received signal level and noise at each pixel for the signal type being studied,
i.e., preamble, traffic, or pilot. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. The
properties of the non-interfering probe receiver are set by selecting a terminal, mobility, and service.
These coverage predictions can be performed at any stage of network planning because they do not depend on the traffic
input. Therefore, these calculations are of special interest before and during the deployment stage of the network to study
the coverage footprint of the system.
• The coverage prediction conditions, and
page 449), the calculation of the coverage parameter ("Coverage Parameter Calculation" on page 450), and the display
options ("Coverage Display" on page 450) of the coverage predictions.

These coverage predictions are all best server coverage predictions, i.e., the coverage area of each cell comprises the
pixels where the cell is the best server. Best server for each pixel is calculated as explained in "Best Server Determination"
on page 471.
© Forsk 2009 AT271_TRG_E6 449

9.2.3.2 Coverage Parameter Calculation

The following parameters are calculated for the Signal Quality Analysis (DL) coverage prediction.
TX i  ic 
• Best Preamble Signal Level (DL) (dBm): C Preamble as explained in "Preamble Signal Level Calculation" on
page 466.
TX i  ic 
• Best Pilot Signal Level (DL) (dBm): C Pilot as explained in "Traffic and Pilot Signal Level Calculation (DL)" on
page 473.
TX i  ic 
• Best Traffic Signal Level (DL) (dBm): C Traffic as explained in "Traffic and Pilot Signal Level Calculation (DL)"
on page 473.
TX i  ic 
• Preamble C/N Level (DL) (dB): CNR Preamble as explained in "Preamble C/N Calculation" on page 470.
TX i  ic 
• Pilot C/N Level (DL) (dB): CNR Pilot as explained in "Traffic and Pilot C/N Calculation (DL)" on page 482.
TX i  ic 
• Traffic C/N Level (DL) (dB): CNR Traffic as explained in "Traffic and Pilot C/N Calculation (DL)" on page 482.
The following parameters are calculated for the Signal Quality Analysis (UL) coverage prediction.
Mi
• Signal Level (UL) (dBm): C UL as explained in "Traffic Signal Level Calculation (UL)" on page 484.
Mi
• C/N Level (UL) (dB): CNR UL as explained in "Traffic C/N Calculation (UL)" on page 487.

9.2.3.3.2 Signal Quality Analysis (DL) Display Types

It is possible to display the Signal Quality Analysis (DL) coverage prediction with colours depending on the following
display options.
Best Preamble Signal Level (DL) (dBm)
Atoll calculates preamble signal levels received from the best serving cells on each pixel of their coverage areas. A pixel
of a coverage area is coloured if the preamble signal level exceeds (  ) the defined thresholds (the pixel colour depends
on the preamble signal level). Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the preamble signal
level from the best server exceeds a defined threshold.
Best Pilot Signal Level (DL) (dBm)
Atoll calculates pilot signal levels received from the best serving cells on each pixel of their coverage areas. A pixel of a
coverage area is coloured if the pilot signal level exceeds (  ) the defined thresholds (the pixel colour depends on the pilot
signal level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There
are as many layers as defined thresholds. Each layer corresponds to an area where the pilot signal level from the best
server exceeds a defined threshold.
Best Traffic Signal Level (DL) (dBm)
Atoll calculates traffic signal levels received from the best serving cells on each pixel of their coverage areas. A pixel of a
coverage area is coloured if the traffic signal level exceeds (  ) the defined thresholds (the pixel colour depends on the
traffic signal level). Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the traffic signal level from the
best server exceeds a defined threshold.
Preamble C/N Level (DL) (dB)
Atoll calculates preamble C/N levels received from the best serving cells on each pixel of their coverage areas. A pixel of
a coverage area is coloured if the preamble C/N level exceeds (  ) the defined thresholds (the pixel colour depends on
the preamble C/N level). Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the preamble C/N
level from the best server exceeds a defined threshold.
450 AT271_TRG_E6 © Forsk 2009

Pilot C/N Level (DL) (dB)
Atoll calculates pilot C/N levels received from the best serving cells on each pixel of their coverage areas. A pixel of a
coverage area is coloured if the pilot C/N level exceeds (  ) the defined thresholds (the pixel colour depends on the pilot
C/N level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are
as many layers as defined thresholds. Each layer corresponds to an area where the pilot C/N level from the best server
exceeds a defined threshold.
Traffic C/N Level (DL) (dB)
Atoll calculates traffic C/N levels received from the best serving cells on each pixel of their coverage areas. A pixel of a
coverage area is coloured if the traffic C/N level exceeds (  ) the defined thresholds (the pixel colour depends on the traffic
C/N level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are
as many layers as defined thresholds. Each layer corresponds to an area where the traffic C/N level from the best server
exceeds a defined threshold.
9.2.3.3.3 Signal Quality Analysis (UL) Display Types

It is possible to display the Signal Quality Analysis (UL) coverage prediction with colours depending on the following
display options.
Signal Level (UL) (dBm)
Atoll calculates uplink signal levels received from each pixel, of the coverage areas of the best serving cells, at the cells.
A pixel of a coverage area is coloured if the uplink signal level exceeds (  ) the defined thresholds (the pixel colour
depends on the uplink signal level). Coverage consists of several independent layers whose visibility in the workspace can
be managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the uplink signal
level at the best server exceeds a defined threshold.
C/N Level (UL) (dB)
Atoll calculates uplink C/N levels received from each pixel, of the coverage areas of the best serving cells, at the cells. A
pixel of a coverage area is coloured if the uplink C/N level exceeds (  ) the defined thresholds (the pixel colour depends
on the uplink C/N level). Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the uplink C/N level
at the best server exceeds a defined threshold.
9.2.4 Calculations on Subscriber Lists

When calculations are performed on a list of subscribers by running the Automatic Server Allocation, Atoll calculates the
path loss again for the subscriber locations and heights because the subscriber heights can be different from the global
receiver height used for calculating the path loss matrices.
Atoll calculates the following parameters for each subscriber in the list whose Lock Status is set to None.
• Serving Base Station and Reference Cell as described in "Best Server Determination" on page 471.
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.
• Azimuth (  ): Angle with respect to the north for pointing the subscriber terminal towards its serving base station.
• Mechanical Downtilt (  ): Angle with respect to the horizontal for pointing the subscriber terminal towards its
serving base station.
Atoll calculates the following parameters for each subscriber in the list that has a serving base station assigned, using the
properties of the default terminal and service.
• Received Preamble Power (DL) (dBm) as described in "Preamble Signal Level Calculation" on page 466.
• Received Traffic Power (DL) (dBm) as described in "Traffic and Pilot Signal Level Calculation (DL)" on page 473.
• Traffic Total Noise (I+N) (DL) (dBm) as described in "Traffic C/(I+N) Calculation (DL)" on page 483.
• Traffic C/(I+N) (DL) (dB) as described in "Traffic C/(I+N) Calculation (DL)" on page 483.
• Bearer (DL) as described in "Bearer Determination" on page 496.
• Permutation Zone (DL) as described in "Permutation Zone Selection (WiMAX 802.16e)" on page 471.
TX i  ic 
• BLER (DL): Downlink block error rate read from the BLER vs. CINR Traffic graph available in the reception equip-
ment assigned to the terminal used by the subscriber.
• Diversity Mode (DL): Antenna diversity mode supported by the permutation zone assigned to the subscriber.
• Peak MAC Channel Throughput (DL) (kbps) as described in "Channel Throughput Calculation" on page 496.
• Effective MAC Channel Throughput (DL) (kbps) as described in "Channel Throughput Calculation" on
page 496.
• Received Power (UL) (dBm) as described in "Traffic Signal Level Calculation (UL)" on page 484.
• Total Noise (I+N) (UL) (dBm) as described in "Noise Rise Calculation (UL)" on page 487.
• C/(I+N) (UL) (dB) as described in "Traffic C/(I+N) Calculation (UL)" on page 490.
• Bearer (UL) as described in "Bearer Determination" on page 496.
• Permutation Zone (UL) as described in "Permutation Zone Selection (WiMAX 802.16e)" on page 471.
© Forsk 2009 AT271_TRG_E6 451

Mi
• BLER (UL): Uplink block error rate read from the BLER vs. CINR UL graph available in the reception equipment
assigned to the serving cell of the subscriber.
• Diversity Mode (UL): Antenna diversity mode supported by the permutation zone assigned to the subscriber.
• Transmission Power (UL) as described in "Traffic C/(I+N) Calculation (UL)" on page 490.
• Number of Used Subchannels (UL) as described in "Traffic C/(I+N) Calculation (UL)" on page 490.
• Peak MAC Channel Throughput (DL) (kbps) as described in "Channel Throughput Calculation" on page 496.
• Effective MAC Channel Throughput (DL) (kbps) as described in "Channel Throughput Calculation" on
page 496.
9.2.5 Monte Carlo Simulations

The simulation process is divided into two steps.
• Generating a realistic user distribution as explained in "Generating a Realistic User Distribution" on page 452.
Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data. 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 455.
9.2.5.1 Generating a Realistic User Distribution

During each simulation, Atoll performs two random trials. 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.
• "Simulations Based on Raster Traffic Maps, Vector Traffic Maps, and Subscriber Lists" on page 452.
• "Simulations Based on Traffic Map Based on Transmitters and Services" on page 454.
Once all the user characteristics have been determined, a second random trial is performed to obtain their geographical
locations weighted according to the clutter classes, and whether they are indoor or outdoor according to the percentage
of indoor users per clutter class defined for the traffic maps.
9.2.5.1.1 Simulations Based on Raster Traffic Maps, Vector Traffic Maps, and Subscriber Lists
a list of user profiles with an associated mobility type and a given density, i.e., number of users of a user profile per km².
Vector traffic maps are based on user profiles. 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, each point is assigned a number of users with given user
profile and mobility type.
Fixed subscribers listed in subscriber lists have a user profile assigned to each of them.
The user profile models the behaviour of the different user categories. Each user profile contains a list of services and their
associated parameters describing how these services are accessed by the user.
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).
N Users = S Env  D UP
Notes:
• In case of vector traffic map 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 an input when the vector traffic map is composed of points.
At any given instant, 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, i.e., the number of voice calls or data sessions, the average
duration of each voice call, or the volumes of the data exchanged in the uplink and the downlink in each data session.
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 number of calls per hour N Call .
- The average duration of a call (seconds) D Call .
N Call  D Call
Calculation of the service usage duration per hour ( p 0 : probability of a connection): 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 connection, i.e., the uplink and downlink
UL DL
activity factors defined for the voice type service v, f Act and f Act .
452 AT271_TRG_E6 © Forsk 2009

UL DL
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
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, a user can be either active on both links, inactive on both links, active on UL only, 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).
- The average number of data calls per hour N Call .
DL UL
- The average data volume (in kBytes) transferred in the downlink V and the uplink V during a session.
DL UL
- The average throughputs in the downlink TP Average and the uplink TP Average for the service d.
UL DL
UL N Call  V  8 DL N Call  V  8
Calculation of activity probabilities: f = ---------------------------------------------- and f = ----------------------------------------------
UL DL
TP Average  3600 TP Average  3600
UL DL
Probability of being inactive: p Inactive =  1 – f   1 – f 
UL UL DL
Probability of being active in the uplink: p Active = f  1 – f 
DL DL UL
Probability of being active in the downlink: p Active = f  1 – f 
UL + DL UL DL
Probability of being active in the uplink and downlink both: p Active = f f
Calculation of number of users:
Number of inactive users: n v – 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
Inactive users are not taken into account.
Note:
• The user distribution per service and the activity status distribution between the users are
average distributions. And the service and the activity status of each user are randomly
drawn in each simulation. Therefore, if you calculate several simulations at once, the
average number of users per service and average numbers of inactive, active on UL, active
on DL and active on UL and DL users, respectively, will correspond to calculated
distributions. But if you check each simulation, the user distribution between services as
well as the activity status distribution between users is different in each of them.
© Forsk 2009 AT271_TRG_E6 453

9.2.5.1.2 Simulations Based on Traffic Map Based on Transmitters and Services

Traffic maps based on transmitters and services are also referred to as live traffic maps. Live traffic data from the OMC is
spread over the best server coverage areas of the transmitters included in the traffic map. Either throughput demands per
service or the number of active users per service (uplink or downlink) are assigned to the coverage areas of each trans-
mitter.
For each transmitter TXi and each service s,
• Traffic Map Based on Transmitters and Services (Throughputs)

Atoll calculates the number of active users of each service s on UL and DL in the coverage area of TXi as follows:
UL DL
UL TP Cell DL TP Cell
N - 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
DL
the transmitter TXi, TP Cell is the total downlink throughput demand defined in the map for any service s for the
UL
coverage area of the transmitter TXi, TP Average is the average uplink throughput demand of the service s, and
DL
TP Average is the average downlink throughput demand of the service s.
• Traffic Map Based on Transmitters and Services (# Active Users)

UL DL
Atoll directly uses the defined N and N values, i.e., the number of active users on UL and DL in the TXi
coverage area using the service s.
At any given instant, Atoll calculates the probability for a user being active in the uplink and in the downlink as follows:
Voice Service (v)
UL DL
Users active in the uplink and downlink both are included in the N and N values. Therefore, it is necessary
UL DL
to accurately determine the number of active users in the uplink ( n v – Active ), in the downlink ( n v – Active ), and both
UL + DL
( n v – Active ). As for the other types of traffic maps, Atoll considers both active and inactive users for voice services.
The activity status of each user depends on the activity periods during the connection, i.e., the uplink and downlink
UL DL
activity factors defined for the voice type service v, f Act and f Act .
UL DL
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 voice service v:
UL UL UL + DL DL DL UL + DL
We have: N =  p Active + p Active   n v and N =  p Active + p Active   n v
Where, nv is the total number of active users in the TXi coverage area using the service v.
UL UL + DL DL UL + DL
UL + DL  N  p Active N  p Active 
- -------------------------------------------
Number of users active in the uplink and downlink both: n v – Active = Min  ------------------------------------------- -
 p UL UL + DL
Active + p Active p Active + p Active
DL UL + DL

UL UL UL + DL
Number of users active in the uplink: n v – Active = N – n v – Active
DL DL UL + DL
Number of users active in the downlink: n v – Active = N – n v – Active
UL DL UL + DL
And, n v = n v – Active + n v – Active + n v – Active
Calculation of the number of inactive users attempting to access the service v:
nv
Number of inactive users: n v – Inactive = ------------------------------  p Inactive
1 – p Inactive
454 AT271_TRG_E6 © Forsk 2009

Data Service (d)
Here, Atoll considers all the users as active. Activity probabilities are not calculated.
Calculation of the number of users attempting to access the service d:
UL DL
If N N
UL + DL UL
n d – Active = N
UL
n d – Active = 0
DL DL UL
n d – Active = N –N
UL DL
If N N
UL + DL DL
n d – Active = N
DL
n d – Active = 0
UL UL DL
n d – Active = N –N
nd is the total number of active users in the TXi coverage area using the service d:
UL DL UL + DL
n d = n d – Active + n d – Active + n d – Active
Note:
• The activity status distribution between users is an average distribution. In fact, in each
simulation, the activity status of each user is randomly drawn. Therefore, if you calculate
several simulations at once, average numbers of inactive, active on UL, active on DL and
active on UL and DL users correspond to the calculated distribution. But if you check each
simulation, the activity status distribution between users is different in each of them.
9.2.5.2 Simulation Process

WiMAX cells include intelligent schedulers and radio resource management features for regulating network traffic loads,
optimising spectral efficiency, and satisfying the QoS demands of the users. 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. The
number of WiMAX frames in 1 second depends on the selected frame duration, D Frame . The steps of this algorithm are
listed below.
The simulation process can be summed up into the following iterative steps.
For each simulation, the simulation process,
1. Generates mobiles according to the input traffic data as explained in "Generating a Realistic User Distribution" on
page 452.
2. Resets the cell loads to initial values, i.e., 100 % downlink traffic load, 0 % uplink traffic load, 0 dB uplink noise
rise, 0 % segmentation usage ratio, and 0 % AAS usage ratio.
- Cell transmission powers and offsets ( P Preamble , P Traffic , P Pilot , and P Idle – Pilot ) are set to the
values defined by the user.
Mi
- Mobile transmission power is set to the maximum mobile power ( P Max ).
- Cell loads ( TL DL , TL UL , NR UL , SU , and AU ) are set to their current values in the
Cells table.
3. Determines the best servers for all the mobiles generated for the simulation as explained in "Best Server Deter-
mination" on page 471.
© Forsk 2009 AT271_TRG_E6 455

Figure 9.1: WiMAX Simulation Algorithm
For each iteration k, the simulation process,

4. Determines the downlink and uplink traffic C/(I+N) for each of these mobiles as explained in "Traffic C/(I+N) Cal-
culation (DL)" on page 483 and "Traffic C/(I+N) Calculation (UL)" on page 490 respectively.
The C/(I+N) is calculated in different ways depending on whether a smart antenna has been assigned to a trans-
mitter or not.
- Without smart antennas:

The downlink traffic loads of cells are used to calculate the downlink interference, i.e., the interference re-
ceived from each interfering cell in the downlink is weighted by its downlink traffic load. The uplink traffic loads
of interfering mobiles are used to calculate the uplink interference, i.e., the interference received from each
interfering mobile in the uplink is weighted by its uplink traffic load.
- With smart antennas:

Victim and Interfering Mobiles:
In WiMAX simulations, the terms victim and interfering mobiles are used for mobiles served by the victim and
interfering cells respectively. In the downlink, victim mobiles receive interfering signals from interfering cells.
In the uplink, victim cells receive interfering signals from interfering mobiles.
456 AT271_TRG_E6 © Forsk 2009

Figure 9.2: Victim and Interfering Mobiles
Atoll assumes that a different beam is formed for each victim mobile.
Time-Slot Scenario:
For each victim mobile, instead of weighting the interference by traffic loads of the interfering mobiles or cells,
a Time-Slot Scenario (TSS) is generated. A time-slot scenario is a list of interfering mobiles per victim mobile.
In other words, a time-slot scenario generated for any victim mobile contains at most one (1) interfering mobile
in each interfering cell. Each time-slot scenario can generate either one (1) or no (0) interfering mobile in each
interfering cell for each victim mobile. One and only one time-slot scenario is generated per victim mobile. And,
there are as many time-slot scenarios generated as there are victim mobiles generated during the simulation.
For a given time-slot scenario, the probability that an interfering mobile is present in an interfering cell depends
on the traffic loads of the potential interfering mobiles in the interfering cell. All mobiles in an interfering cell
are potential interfering mobiles. Which one (1), if any, of these potential interfering mobiles is included in the
time-slot scenario is determined by weighting their probability of selection by the mobile traffic loads.
Interference Averaging Over Iterations:
The interferences calculated for each iteration are averaged for all the iterations in a simulation.
This process considers the probabilities of collision between slots used by victim and interfering mobiles,
which depend upon their transmission times, or in other words, their traffic loads. The averaging of interfer-
ences over all the iterations in a simulation gives a realistic average interference at the end of the simulation.
The above process is the same for uplink and downlink. In this way, Atoll simulates the simultaneous connec-
tions of victim and interfering mobiles, and considers the effect of the smart antenna, because beamforming
is performed for all victim and interfering mobiles for all time-slot scenarios.
Each iteration starts with the initial conditions which are the results from the previous iteration, i.e., interference
information are input to each iteration. The averaging of interferences over all the iterations in a simulation is
performed in a successive manner. This is done because each time-slot scenario may have a large number
of interfering mobiles for each victim mobile.
Successive averaging means that the interference information input to an iteration takes into account the out-
put interference information of the preceding iteration as well as the weighted input interference information of
all the previously carried out iterations. If I represents the interference information, the interference information
input to the kth iteration can be given by:
n k – 1 n k – 2
n k I out +    k – 2   I in
I in = -----------------------------------------------------------------------
1 +   k – 2
Where k is the iteration number, n is any given victim mobile, and  is a stability factor used to help simula-
tions converge quickly. The stability factor is currently set to 0.2. The effect of the stability factor can be un-
derstood by the following figure.
Figure 9.3: Simulation Convergence Stability Factor
© Forsk 2009 AT271_TRG_E6 457

5. Determines the mobiles which are within the service areas of their best serving cells as explained in "Service Area
Calculation" on page 471.
6. Determines the permutation zone assigned to each mobile as explained in "Permutation Zone Selection (WiMAX
802.16e)" on page 471.
7. Determines the best available WiMAX bearer at the mobile as explained in "Bearer Determination" on page 496.
8. Determines the channel throughputs at the mobile as explained in "Channel Throughput Calculation" on page 496.
9. Performs radio resource management and scheduling to determine the amount of resources to allocate to each
mobile according to the QoS and service demands of each mobile using the selected scheduler as explained in
"Scheduling and Radio Resource Management" on page 443.
10. Calculates the user throughputs after allocating resources to each mobile as explained in "User Throughput Cal-
culation" on page 504.
11. Updates the traffic loads, 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).
TX i  ic  Mi TX i  ic  Mi
TL DL =  %RDL and TL UL =  %RUL
Mi Mi
Calculation of Uplink Noise Rise:
For each victim cell TXi(ic), the uplink noise rise is calculated and updated by considering each interfering mobile
Mj as explained in "Noise Rise Calculation (UL)" on page 487.
Calculation of Segmentation Usage (WiMAX 802.16e):
Atoll calculates the segmentation usages for all the cells as follows:
Mi
 %R DL
Mi
PZDL = Seg
Mi
M
TX i  ic  i
PZ = Seg
SU = ------------------------------------------------------------------------
DL -
TX i  ic 
TL DL
Mi
Where  %R DL
M
i
PZ DL = Seg
is the sum of the percentages of the downlink cell resources allocated to
Mi
M
i
PZ DL = Seg
mobiles served by the segmented permutation zone.
Calculation of AAS Usage:
Atoll calculates the AAS usages for all the cells as follows:
Mi
 %R DL
AAS
TX i  ic  Mi
AU = -------------------------------------------
AAS
TX i  ic 
TL DL
Mi
Where  %R DL
AAS
is the sum of the percentages of the downlink cell resources allocated to mobiles served
Mi
AAS
by the smart antenna equipment.
12. Performs the convergence test to see whether the differences between the current and the new loads are within
the convergence thresholds.
The convergence criteria are evaluated at the end of each iteration k, and can be written as follows:
TL DL = Max  TL DL – TL DL 
k All TX  ic 
 k k – 1
i
TL UL = Max  TL UL – TL UL 
k 
All TX i  ic  k k – 1
NR UL = Max  NR UL – NR UL 
k All TX  ic 
 k k – 1
i
Let TL DL , TL UL , and NR UL be the simulation convergence thresholds defined when
Req Req Req
creating the simulation. Atoll stops the simulation in the following cases.
458 AT271_TRG_E6 © Forsk 2009

Convergence: Simulation has converged between iteration k - 1 and k if:
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic 
TL DL  TL DL AND TL UL  TL UL AND NR UL  NR UL
k Req k Req k Req
No convergence: Simulation has not converged even after the last iteration, i.e., k = Max Number of Iterations
defined when creating the simulation, if:
TL DL  TL DL OR TL UL  TL UL OR NR UL  NR UL
k Req k Req k Req
13. Repeats the above steps (from step 4) for the iteration k+1 using the new calculated loads as the current loads.
Simulation Results
At the end of the simulation process, the main results obtained are:
• Downlink Traffic Loads
• Uplink Traffic Loads
• 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
These results can be used as input for C/(I+N)-based coverage predictions.
In addition to the above parameters, the simulations also list the connection status of each mobile. Mobiles can be rejected
due to:
• No Service: If the mobile does not have any best serving cell (step 3.), if the mobile is not located in the service
area of its best server (step 5.), or if the mobile is not able to access a bearer in the direction of its activity (step 7.),
i.e., UL, DL, or UL+DL.
• Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 9.)
• Resource Saturation: If all the cell resources are used up before allocation to the mobile (step 9.)
Connected mobiles (step 9.) can be:
• Connected UL: If a mobile active in UL is allocated resources in UL.
• Connected DL: If a mobile active in DL is allocated resources in DL.
• Connected DL+UL: If a mobile active in UL+DL is allocated resources in UL+DL.
9.2.6 C/(I+N)-Based Coverage Predictions

The following coverage predictions are based on the received traffic or pilot signal levels, total noise, and the traffic and
pilot interference.
• Coverage by C/(I+N) Level (DL)
• Coverage by Best Bearer (DL)
• Coverage by Channel Throughput (DL)
• Coverage by C/(I+N) Level (UL)
• Coverage by Best Bearer (UL)
• Coverage by Channel Throughput (UL)
Mi Mi Mi Mi
These coverage predictions take into account the receiver characteristics ( L ,G , L Ant , and L Body ) when calculating
the required parameter. For these calculations, Atoll calculates the received signal level, noise, and interference at each
pixel. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. The properties of the non-
interfering probe receiver are set by selecting a terminal, mobility, and service.
The downlink coverage predictions are based on the downlink traffic loads of the cells, and the uplink coverage predictions
are based on the uplink noise rise values. These parameters can either be calculated by Atoll during the Monte Carlo simu-
lations, or set by the user manually for all the cells.
• The coverage prediction conditions, and
The minimum thresholds at the receiver are defined in the Display parameters.
page 459), the calculation of the coverage parameter ("Coverage Parameter Calculation" on page 460), and the display
options ("Coverage Display" on page 460) of the coverage predictions.

These coverage predictions are all best server coverage predictions, i.e., the coverage area of each cell comprises the
pixels where the cell is the best server. Best server for each pixel is calculated as explained in "Best Server Determination"
on page 471.
© Forsk 2009 AT271_TRG_E6 459

9.2.6.2 Coverage Parameter Calculation

The following parameters are calculated for the Coverage by C/(I+N) Level (DL) coverage prediction.
TX i  ic 
• Traffic C/(I+N) Level (DL) (dB): CINR Traffic as explained in "Traffic C/(I+N) Calculation (DL)" on page 483.
TX i  ic 
• Traffic Total Noise (I+N) (DL) (dBm):  I + N DL as explained in "Traffic C/(I+N) Calculation (DL)" on
page 483.
TX i  ic 
• Preamble C/(I+N) Level (DL) (dB): CINR Preamble as explained in "Preamble C/(I+N) Calculation" on page 470.
TX i  ic 
• Preamble Total Noise (I+N) (DL) (dBm):  I + N  Preamble as explained in "Preamble C/(I+N) Calculation" on
page 470.
The following parameters are calculated for the Coverage by Best Bearer (DL) coverage prediction.
Mi
• Best Bearer (DL): B DL as explained in "Bearer Determination" on page 496.
The following parameters are calculated for the Coverage by Channel Throughput (DL) coverage prediction.
Mi
• Peak MAC Throughput (DL) (kbps): CTP P – DL as explained in "Channel Throughput Calculation" on page 496.
Mi
• Effective MAC Throughput (DL) (kbps): CTP E – DL as explained in "Channel Throughput Calculation" on
page 496.
Mi
• Application Throughput (DL) (kbps): CTP A – DL as explained in "Channel Throughput Calculation" on
page 496.
The following parameters are calculated for the Coverage by C/(I+N) Level (UL) coverage prediction.
Mi
• C/(I+N) Level (UL) (dB): CINR UL as explained in "Traffic C/(I+N) Calculation (UL)" on page 490.
TX i  ic 
• Total Noise (I+N) (UL) (dBm):  I + N  UL as explained in "Noise Rise Calculation (UL)" on page 487.
The following parameters are calculated for the Coverage by Best Bearer (UL) coverage prediction.
Mi
• Best Bearer (UL): B UL as explained in "Bearer Determination" on page 496.
The following parameters are calculated for the Coverage by Channel Throughput (UL) coverage prediction.
Mi
• Peak MAC Throughput (UL) (kbps): CTP P – UL as explained in "Channel Throughput Calculation" on page 496.
Mi
• Effective MAC Throughput (UL) (kbps): CTP E – UL as explained in "Channel Throughput Calculation" on
page 496.
Mi
• Application Throughput (UL) (kbps): CTP A – UL as explained in "Channel Throughput Calculation" on
page 496.

9.2.6.3.2 Coverage by C/(I+N) Level (DL) Display Types

It is possible to display the Coverage by C/(I+N) Level (DL) coverage prediction with colours depending on the following
display options.
Traffic C/(I+N) Level (DL) (dB)
Atoll calculates traffic C/(I+N) levels received from the best serving cells on each pixel of their coverage areas. A pixel of
a coverage area is coloured if the traffic C/(I+N) level exceeds (  ) the defined thresholds (the pixel colour depends on the
traffic C/(I+N) level). Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the traffic C/(I+N) level from
the best server exceeds a defined threshold.
Traffic Total Noise (I+N) (DL) (dBm)
Atoll calculates traffic total noise (I+N) levels received from the interfering cells on each pixel of the coverage areas of the
best serving cells. A pixel of a coverage area is coloured if the traffic total noise (I+N) level exceeds (  ) the defined thresh-
olds (the pixel colour depends on the traffic total noise (I+N) level). Coverage consists of several independent layers whose
460 AT271_TRG_E6 © Forsk 2009

visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer corresponds to
an area where the traffic total noise (I+N) level from the interfering cells exceeds a defined threshold.
Preamble C/(I+N) Level (DL) (dB)
Atoll calculates preamble C/(I+N) levels received from the best serving cells on each pixel of their coverage areas. A pixel
of a coverage area is coloured if the preamble C/(I+N) level exceeds (  ) the defined thresholds (the pixel colour depends
on the preamble C/(I+N) level). Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the preamble C/
(I+N) level from the best server exceeds a defined threshold.
Preamble Total Noise (I+N) (DL) (dBm)
Atoll calculates preamble total noise (I+N) levels received from the interfering cells on each pixel of the coverage areas of
the best serving cells. A pixel of a coverage area is coloured if the preamble total noise (I+N) level exceeds (  ) the defined
thresholds (the pixel colour depends on the preamble total noise (I+N) level). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer
corresponds to an area where the preamble total noise (I+N) level from the interfering cells exceeds a defined threshold.
9.2.6.3.3 Coverage by Best Bearer (DL) Display Types

It is possible to display the Coverage by Best Bearer (DL) coverage prediction with colours depending on the following
display options.
Best Bearer (DL)
Atoll determines the best bearer available on each pixel of the coverage areas of the best serving cells. A pixel of a cover-
age area is coloured if a bearer is available (the pixel colour depends on the available bearer). Coverage consists of
several independent layers whose visibility in the workspace can be managed. There are as many layers as available bear-
ers. Each layer corresponds to an area covered by an available bearer.
9.2.6.3.4 Coverage by Channel Throughput (DL) Display Types

It is possible to display the Coverage by Channel Throughput (DL) coverage prediction with colours depending on the
following display options.
Peak MAC Throughput (DL) (kbps)
Atoll calculates peak MAC channel throughputs available on each pixel of the coverage areas of the best serving cells. A
pixel of a coverage area is coloured if the peak MAC channel throughput exceeds (  ) the defined thresholds (the pixel
colour depends on the peak MAC channel throughput). Coverage consists of several independent layers whose visibility
in the workspace can be managed. There are as many layers as defined thresholds. Each layer corresponds to an area
where the peak MAC channel throughput exceeds a defined threshold.
Effective MAC Throughput (DL) (kbps)
Atoll calculates effective MAC channel throughputs available on each pixel of the coverage areas of the best serving cells.
A pixel of a coverage area is coloured if the effective MAC channel throughput exceeds (  ) the defined thresholds (the
pixel colour depends on the effective MAC channel throughput). Coverage consists of several independent layers whose
an area where the effective MAC channel throughput exceeds a defined threshold.
Application Throughput (DL) (kbps)
Atoll calculates application level channel throughputs available on each pixel of the coverage areas of the best serving
cells. A pixel of a coverage area is coloured if the application level channel throughput exceeds (  ) the defined thresholds
(the pixel colour depends on the application level channel throughput). Coverage consists of several independent layers
sponds to an area where the application level channel throughput exceeds a defined threshold.
9.2.6.3.5 Coverage by C/(I+N) Level (UL) Display Types

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)
Atoll calculates uplink C/(I+N) levels received at the best serving cells from each pixel of their coverage areas. A pixel of
a coverage area is coloured if the uplink C/(I+N) level exceeds (  ) the defined thresholds (the pixel colour depends on the
uplink C/(I+N) level). Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the uplink C/(I+N) level from
the pixels at the best serving cells exceeds a defined threshold.
© Forsk 2009 AT271_TRG_E6 461

Total Noise (I+N) (UL) (dBm)
Atoll calculates total noise (I+N) levels received at the best serving cells from each pixel of their coverage areas. A pixel
of a coverage area is coloured if the total noise (I+N) level exceeds (  ) the defined thresholds (the pixel colour depends
on the total noise (I+N) level). Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as defined thresholds. Each layer corresponds to an area where the total noise (I+N)
level from the pixels at the best serving cells exceeds a defined threshold.
9.2.6.3.6 Coverage by Best Bearer (UL) Display Types

It is possible to display the Coverage by Best Bearer (UL) coverage prediction with colours depending on the following
display options.
Best Bearer (UL)
Atoll determines the best bearer available on each pixel of the coverage areas of the best serving cells. A pixel of a cover-
age area is coloured if a bearer is available (the pixel colour depends on the available bearer). Coverage consists of
several independent layers whose visibility in the workspace can be managed. There are as many layers as available bear-
ers. Each layer corresponds to an area covered by an available bearer.
9.2.6.3.7 Coverage by Channel Throughput (UL) Display Types

It is possible to display the Coverage by Channel Throughput (UL) coverage prediction with colours depending on the
following display options.
Peak MAC Throughput (UL) (kbps)
Atoll calculates peak MAC channel throughputs available on each pixel of the coverage areas of the best serving cells. A
pixel of a coverage area is coloured if the peak MAC channel throughput exceeds (  ) the defined thresholds (the pixel
colour depends on the peak MAC channel throughput). Coverage consists of several independent layers whose visibility
in the workspace can be managed. There are as many layers as defined thresholds. Each layer corresponds to an area
where the peak MAC channel throughput exceeds a defined threshold.
Effective MAC Throughput (UL) (kbps)
Atoll calculates effective MAC channel throughputs available on each pixel of the coverage areas of the best serving cells.
A pixel of a coverage area is coloured if the effective MAC channel throughput exceeds (  ) the defined thresholds (the
pixel colour depends on the effective MAC channel throughput). Coverage consists of several independent layers whose
an area where the effective MAC channel throughput exceeds a defined threshold.
Application Throughput (UL) (kbps)
Atoll calculates application level channel throughputs available on each pixel of the coverage areas of the best serving
cells. A pixel of a coverage area is coloured if the application level channel throughput exceeds (  ) the defined thresholds
(the pixel colour depends on the application level channel throughput). Coverage consists of several independent layers
sponds to an area where the application level channel throughput exceeds a defined threshold.
9.3 Calculation Algorithms

The following sections describe all the calculation algorithms used in point analysis, calculation of coverage predictions,
calculations on subscriber lists, and Monte Carlo simulations.
9.3.1 Co- and Adjacent Channel Overlaps Calculation

A WiMAX network can consist of cells that use different channel bandwidths. Therefore, the start and the end of all the
channel bandwidths may not coincide exactly. Channel bandwidths of different cells can overlap channel bandwidths of
other cells with different ratios.
462 AT271_TRG_E6 © Forsk 2009

Figure 9.4: Co-Channel and Adjacent Channel Overlaps
The following sections describe how the co- and adjacent channel overlaps are calculated between the channel used by
any studied cell TXi(ic) and any other cell of the network TXj(jc). In terms of interference calculation, 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 
If the studied cell is assigned a channel number N Channel , it receives co-channel interference on channel number
N Channel , and adjacent channel interference on the adjacent channel numbers, i.e., N Channel – 1 and N Channel + 1 .
In order to calculate the co- and adjacent channel overlaps between two channels, 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 463). Once the start and end frequencies are known for the studied and other cells, the co- and adjacent overlaps,
and the total overlap ratio are calculated as respectively explained in:
• "Co-Channel Overlap Calculation" on page 464
• "Adjacent Channel Overlap Calculation" on page 464
• "Total Overlap Ratio Calculation" on page 465
9.3.1.1 Conversion From Channel Numbers to Start and End Frequencies

Input
• F Start – FB and F Start – FB : Start frequency of the frequency band assigned to the cells TXi(ic) and TXj(jc).
F Start – FB can be the start frequency of a TDD frequency band ( F Start – FB – TDD ), or the uplink or the downlink
start frequency of an FDD frequency band ( F Start – FB – FDD – UL or F Start – FB – FDD – DL ).
• N Channel and N Channel : Channel numbers assigned to cells TXi(ic) and TXj(jc).
For FDD networks, Atoll considers that the same channel number is assigned to a cell in the downlink and uplink,
i.e., the channel number you assign to a cell is considered for uplink and downlink both.
• W Channel and W Channel : Bandwidths of the channels assigned to cells TXi(ic) and TXj(jc).
Calculations
The following formulas are used to convert the channel numbers into start and end frequencies:
For cell TXi(ic):
F Start = F Start – FB + N Channel  W Channel
F End = F Start – FB + N Channel  W Channel + W Channel
For cell TXj(jc):
TX j  jc  TX j  jc  TX j  jc  TX j  jc 
F Start = F Start – FB + N Channel  W Channel
© Forsk 2009 AT271_TRG_E6 463

TX j  jc  TX j  jc  TX j  jc  TX j  jc  TX j  jc 
F End = F Start – FB + N Channel  W Channel + W Channel
Output
• F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc).
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc).
9.3.1.2 Co-Channel Overlap Calculation

Input
• 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 463.
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
Calculations
Atoll first verifies that co-channel overlap exists between the cells TXi(ic) and TXj(jc).
Co-channel overlap exists if:

TX i  ic  TX j  jc  TX i  ic  TX j  jc 
F Start  F End AND F End  F Start
Otherwise there is no co-channel overlap.

Atoll calculates the bandwidth of the co-channel overlap as follows:
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX j  jc  TX i  ic 
W CCO = Min  F End  F End  – Max  F Start  F Start 
The co-channel overlap ratio is given by:

TX i  ic  – TX j  jc  W CCO
r CCO = --------------------------------------
-
TX i  ic 
W Channel
Output
• r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc).
9.3.1.3 Adjacent Channel Overlap Calculation

Input
• F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
TX i  ic 
• W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic).
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).
Adjacent channel overlap exists on the lower-frequency adjacent channel if:

TX i  ic  TX i  ic  TX j  jc  TX i  ic  TX j  jc 
F Start – W Channel  F End AND F Start  F Start
Adjacent channel overlap exists on the higher-frequency adjacent channel if:

TX i  ic  TX j  jc  TX i  ic  TX i  ic  TX j  jc 
F End  F End AND F End + W Channel  F Start
Otherwise there is no adjacent channel overlap.

Atoll determines the adjacent channel overlap ratio as follows:
464 AT271_TRG_E6 © Forsk 2009

Bandwidth of the lower-frequency adjacent channel overlap:
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX j  jc  TX i  ic  TX i  ic 
W ACO = Min  F End  F Start  – Max  F Start  F Start – W Channel 
L
The lower-frequency adjacent channel overlap ratio is given by:
TX i  ic  – TX j  jc  W ACO
L
r ACO = --------------------------------------
-
L TX i  ic 
W Channel
Bandwidth of the higher-frequency adjacent channel overlap:
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX i  ic  TX j  jc  TX i  ic 
W ACO = Min  F End  F End + W Channel  – Max  F Start  F End 
H
The higher-frequency adjacent channel overlap ratio is given by:
TX i  ic  – TX j  jc  W ACO
H
r ACO = --------------------------------------
-
H TX i  ic 
W Channel
The adjacent channel overlap ratio is given by:

r ACO = r ACO + r ACO
L H
Output
• r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
9.3.1.4 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. The most common interference scenarios are FDD-only and TDD-only interferences. However,
co-existing FDD and TDD cells may also exist, and interfere each other. Atoll models the co-existence of FDD and TDD
cells in a network by determining the FDD – TDD overlap ratio as follows:
Input
TDD
• r DL – Frame : Downlink subframe ratio defined in the Global Parameters.
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  ic  – TX  jc 
i j
Overlap Ratio r FDD – TDD
TXi(ic) TXj(jc)
TDD TDD 1
TDD FDD 1
TDD
r DL – Frame
FDD TDD --------------------------
-
100
FDD FDD 1
Output
• r FDD – TDD : FDD – TDD overlap ratio between the cells TXi(ic) and TXj(jc).
9.3.1.5 Total Overlap Ratio Calculation

Input
• r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co-Channel
Overlap Calculation" on page 464.
• r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Adjacent
Channel Overlap Calculation" on page 464.
© Forsk 2009 AT271_TRG_E6 465

• 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 465.
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).
Calculations
The total overlap ratio is:
 TX i  ic 
  – f ACS – FB
  TXi  ic  – TXj  jc  TX i  ic  – TX j  jc  ---------------------------
10 TX i  ic  – TX j  jc  TX i  ic  TX j  jc 
  r CCO + r ACO  10   r FDD – TDD if W Channel  W Channel
  
TX i  ic  – TX j  jc    
rO = 
 TX i  ic 
  – f ACS – FB TX i  ic 
  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  ---------------------------
10 TX i  ic  – TX j  jc  W Channel TX i  ic  TX j  jc 
  r CCO + r ACO  10   r FDD – TDD  -----------------------
TX j  jc 
- if W Channel  W Channel
   W Channel
  
TX i  ic 
W Channel
The multiplicative factor -----------------------
TX j  jc 
- is used to normalise the transmission power of the interfering cell TXj(jc). This means
W Channel
TX j  jc 
that if the interfering cell transmits at X dBm over a bandwidth of W Channel , and it interferes over a bandwidth less than
TX i  ic 
TX j  jc  W Channel
W Channel , the interference from this cell should not be considered at X dBm but less than that. The factor -----------------------
TX j  jc 
-
W Channel
converts X dBm over W Channel to Y dBm (which is less than X dBm) over less than W Channel .
Output
• rO : Total co- and adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
9.3.2 Preamble Signal Quality Calculations

Calculation of preamble signal quality includes 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 are calculated over the preamble, and how preamble C/N and C/(I+N) ratios are calculated in Atoll. Refer to
the section corresponding to the details you need.
• "Preamble Signal Level Calculation" on page 466.
• "Preamble Noise Calculation" on page 467.
• "Preamble C/N Calculation" on page 470.
• "Preamble Interference Calculation" on page 469.
• "Preamble C/N Calculation" on page 470.
• "Preamble C/(I+N) Calculation" on page 470.
9.3.2.1 Preamble Signal Level Calculation

Input
TX i  ic 
• P Preamble : Preamble transmission power of the cell TXi(ic).
TX i
• E SA : Number of antenna elements defined for the smart antenna equipment used by the transmitter TXi.
TX i
• G : Transmitter antenna gain for the antenna used by the transmitter TXi.
TX i TX i
• L : Total transmitter losses for the transmitter TXi ( L = L Total – DL ).
• L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model.
TX i
• 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.
This parameter is taken into account when the option "Shadowing taken into account" is selected.
M Shadowing – Model = 0 when calculating the received preamble signal level from interfering cells TXj(jc).
466 AT271_TRG_E6 © Forsk 2009

• L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected.
These losses are defined for each clutter class.
Mi
• L : Receiver terminal losses for the pixel, subscriber, or mobile Mi.
Mi
• G : 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.
The antenna used for the calculations at the pixel, subscriber, or mobile Mi is pointed towards its best serving cell.
Mi Mi
This means that L Ant = 0 for calculating the signal level. Whereas, L Ant is calculated in the direction of the inter-
fering cells 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.
Note:
Mi Mi Mi Mi
• L , G , L Ant , and L Body are not considered in the calculations performed for the point
analysis tool and the basic, 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  Mi Mi Mi Mi
C Preamble = EIRP Preamble – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
Where EIRP is the effective isotropic radiated power of the cell calculated as follows:
• Without smart antenna equipment: EIRP Preamble = P Preamble + G –L
TX i  ic  TX i  ic  TX i TX i TX i
• With smart antenna equipment: EIRP Preamble = P Preamble + G –L + 10  Log  E SA 
L Path is the path loss (dB) calculated as follows:
TX i
Furthermore, the total losses between the cell and the pixel, subscriber, or mobile Mi can be calculated as follows:
TX i TX i Mi Mi Mi Mi
L Total = L Path + L + L Indoor + M Shadowing – Model – G +L –G + L Ant + L Body
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.
9.3.2.2 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. In WiMAX 802.16d, the number of
subcarriers used by the preamble is the same as the number of subcarriers used by the cell over the rest of the WiMAX
frame. But, in WiMAX 802.16e, 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
• K: Boltzmann’s constant.
• T: Temperature in Kelvin.
TX i  ic 
• N SCa – Preamble : Number of used subcarriers defined in the Global Parameters (WiMAX 802.16d) or the number
of subcarriers used by the preamble defined for the frame configuration of the cell TXi(ic) (WiMAX 802.16e).
TX i  ic 
• N SCa – Total : Total number of subcarriers defined in the Global Parameters (WiMAX 802.16d) or for the frame con-
figuration of the cell TXi(ic) (WiMAX 802.16e).
© Forsk 2009 AT271_TRG_E6 467

TX i  ic 
• F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on
page 493.
Mi
• nf : 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 10  K  T  1000  = – 174 dBm/Hz
The thermal noise over the preamble for a cell is calculated as:
TX i  ic 
TX i  ic 
 TX  ic  N SCa – Preamble
n 0 – Preamble  i
= n 0 + 10  Log 10 F Sampling  ---------------------------------------
-
 TX i  ic  
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.
n Preamble = n 0 – Preamble + nf
Effect of Segmentation (WiMAX 802.16e):
If you select the Segmentation Support check box in the Frame Configurations table for the frame configuration
used by the cell, it means that the preamble is segmented, and one of the three preamble carrier sets is used for
transmission.
Preamble 1
The preamble segmenting factor is f Segment = --- , which means that each preamble carrier set uses 1/3rd of the
3
1 -
total number of preamble subcarriers. The power transmitted over a segmented preamble has ----------------------
Preamble
times the
f Segment
spectral density of the power transmitted over the entire channel bandwidth. When calculating the C/N and C/(I+N)
1
ratios, the increase in power by ----------------------
Preamble
- due to this power concentration is equivalent to a reduction in the noise
f Segment
Preamble
level by f Segment . 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, if segmentation is used, the thermal noise at the pixel,
Preamble
subscriber, or mobile Mi covered by the preamble is reduced by a factor of f Segment . Which means that the ther-
mal noise over the preamble for a cell is calculated as:
TX i  ic 
TX i  ic 
 TX  ic  N SCa – Preamble Preamble
n 0 – Preamble  i
-  f Segment
= n 0 + 10  Log 10 F Sampling  ---------------------------------------
 TX i  ic  
The following table shows the different types of subcarriers and their numbers for preamble transmission in
WiMAX 802.16e.
Guard Subcarriers Preamble

N SCa – Total Segment DC Subcarrier N SCa – Preamble f Segment
Left Right Total
All 1 (54) 107 1
0 1 (54) 35 0.3271
128 10 10 20
1 None 36 0.3364
2 None 36 0.3364
All 1 (214) 428 1
0 None 143 0.3341
512 42 41 83
1 1 (214) 142 0.3318
2 None 143 0.3341
All 1 (426) 851 1
0 1 (426) 283 0.3325
1024 86 86 172
1 None 284 0.3337
2 None 284 0.3337
All 1 (852) 1703 1
0 1 (852) 567 0.3329
2048 172 172 344
1 None 568 0.3335
2 None 568 0.3335
468 AT271_TRG_E6 © Forsk 2009

Output
TX i  ic 
• n Preamble : Preamble noise for the cell TXi(ic).
9.3.2.3 Preamble Interference Calculation

The interference received by any cell TXi(ic) from other cells TXj(jc) for a pixel, subscriber, or mobile Mi can be defined as
the preamble signal level received from cells TXj(jc) depending on the overlap that exists between the channels used by
the cells TXi(ic) and TXj(jc), and (for WiMAX 802.16e) whether the preambles of the two cells are segmented or not.
Input
TX j  jc 
• C Preamble : Preamble signal level received from any interfering cell TXj(jc) (as calculated in "Preamble Signal
Level Calculation" on page 466) at the pixel, subscriber, or mobile Mi covered by the cell TXi(ic).
• rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent
Channel Overlaps Calculation" on page 462.
• N Seg and N Seg : Segment numbers assigned to the cells TXi(ic) and TXj(jc) calculated from their respective
preamble indexes ( n Preamble and n Preamble ) as follows:
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
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  TX i  ic  – TX j  jc 
I Preamble = C Preamble + f O + f Seg
Where f O is the interference reduction factor due to channel overlap between the cells TXi(ic) and TXj(jc),
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
fO = 10  Log  r O 
And, f Seg is the interference reduction factor due to preamble segmentation (WiMAX 802.16e), calculated as
follows:
f Seg = 10  Log  p Collision 
The probability of preamble subcarrier collision p Collision between the cells TXi(ic) and TXj(jc) is given by the follow-
ing table:
TX i  ic  TX j  jc  p Collision
Not Segmented Not Segmented 1

Not Segmented Segmented 1/3
Segmented Not Segmented 1
0 if N Seg  N Seg
Segmented Segmented
1 if N Seg = N Seg
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).
© Forsk 2009 AT271_TRG_E6 469

9.3.2.4 Preamble C/N Calculation

Input
TX i  ic 
• C Preamble : Received preamble signal level from the cell TXi(ic) as calculated in "Preamble Signal Level Calcula-
tion" on page 466.
TX i  ic 
• n Preamble : Preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on page 467.
Calculations
The preamble C/N for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
CNR Preamble = C Preamble – n Preamble
Output
TX i  ic 
• CNR Preamble : Preamble C/N from the cell TXi(ic) at any pixel, subscriber, or mobile Mi.
9.3.2.5 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 466) at the pixel, subscriber
or mobile under study. Secondly, 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 469). Interference from each cell
is weighted according to the co- and adjacent channel overlap between the studied and the interfering cells, and (in WiMAX
802.15e) the probabilities of subcarrier collision if segmentation is used. 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 Calcu-
lation" on page 467).
The receiver terminal is always considered to be pointed towards its best server. 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 466.
TX i  ic 
• n Preamble : Preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on page 467.
TX j  jc 
• I Preamble : Preamble interference received from any cell TXj(jc) at a pixel, subscriber, or mobile Mi covered by a
cell TXi(ic) as calculated in "Preamble Interference Calculation" on page 469.
• M Shadowing – C  I : Shadowing margin based on the C/I standard deviation.
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 470.
Calculations
The preamble C/(I+N) for a cell TXi(ic) is calculated as follows at any pixel, subscriber, or mobile Mi:
 TX  jc  
j
TX i  ic  TX i  ic   I Preamble TX i  ic 

-------------------------
CINR Preamble = C Preamble – 10  Log  10  – n Preamble – M Shadowing – C  I
 10 
The Preamble Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX j  jc  TX i  ic 
 I Preamble n Preamble
TX i  ic   ------------------------- -
--------------------------

10 10
 I + N  Preamble = 10  Log  10 + 10 
 
Output
TX i  ic 
• CINR Preamble : Preamble C/(I+N) from the cell TXi(ic) at a pixel, subscriber, or mobile Mi.
470 AT271_TRG_E6 © Forsk 2009

TX i  ic 
•  I + N Preamble : Preamble total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered
by a cell TXi(ic).
9.3.3 Best Server Determination

In WiMAX, best server refers to a cell ("serving transmitter"-"reference cell" pair) from which a pixel, subscriber, or mobile
TX j  jc 
Mi gets the best preamble signal level ( C Preamble ).
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
Mi Mi Mi
in "Preamble Signal Level Calculation" on page 466 using the terminal and service parameters ( L , G , L Ant ,
Mi
and L Body ) of Mi.
Calculations
The best server of any pixel, subscriber, or mobile Mi, BS M , is the cell from which the received preamble signal level is
i
the highest among the preamble signal levels received from all the cells. The best server is determined as follows:
BS M = TX i  ic 
i
TX  ic   TXi  ic  
i Best
C Preamble =  C Preamble 
All TX  ic 
i  
Output
• BS M : Best serving cell of the pixel, subscriber, or mobile Mi.

i
9.3.4 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 471) 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 quality 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
TX i  ic 
• T Preamble : Preamble quality 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:
CNR Preamble  T Preamble
Output
• True: If the calculation criterion is satisfied.

• False: Otherwise.
9.3.5 Permutation Zone Selection (WiMAX 802.16e)

In order to be able to calculate the traffic C/(I+N), and therefore 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 471) of
its best serving cell. The permutation zone assigned to Mi is one which covers Mi in terms of distance and preamble C/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.
© Forsk 2009 AT271_TRG_E6 471

Input
TX i  ic 
• d Max – PZ : Maximum distance covered by a permutation zone of a cell TXi(ic).
TX i  ic 
• CNR Req – PZ : Minimum preamble C/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).
M i – TX i  ic 
• d : Distance between the pixel, subscriber, or mobile Mi and a cell TXi(ic).
TX i  ic 
• CNR Preamble : Preamble C/N from the cell TXi(ic) as calculated in "Preamble C/N Calculation" on page 470.
• 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:
• The distance between Mi and TXi(ic) is less than or equal to the maximum distance covered by the permutation
zone:
M i – TX i  ic  TX i  ic 
d  d Max – PZ
• The preamble C/N at Mi is better than or equal to the minimum preamble C/N required by the permutation zone:
CNR Preamble  CNR Req – 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
 TX  ic  
PZ DL = Highest Priority  PZ DL 
i
 M – TX  ic  TX  ic  TX  ic  TX  ic  TX  ic  
 d i

i
 d Max – PZ AND  CNRPreamble  CNRReq – PZ
i i i
AND  Mobility  Mi   SpeedMax – PZ 
i
Mi
 TX  ic  
PZ UL = Highest Priority  PZ UL 
i
 M – TX  ic  TX  ic  TX  ic  TX  ic  TX  ic  
 d i

i
 d Max – PZ AND  CNRPreamble  CNRReq – PZ
i i i
AND  Mobility  Mi   SpeedMax – PZ 
i
If N (> 1) permutation zones satisfy the distance, speed, and signal quality 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 config-
uration) among these N zones.
Output
Mi Mi
• PZ DL and PZUL : Downlink and uplink permutation zones assigned to the pixel, subscriber, or mobile Mi.
9.3.6 Traffic and Pilot Signal Quality Calculations

Traffic and pilot subcarriers can be transmitted with different transmission powers than the preamble power of a cell. More-
over, traffic and pilot subcarriers do not undergo the same interference and noise as the preamble. The following sections
describe how traffic and pilot signal levels, noise and interference are calculated on the downlink and uplink.
• "Traffic and Pilot Signal Level Calculation (DL)" on page 473.
• "Noise Calculation (DL)" on page 474.
• "Traffic and Pilot Interference Calculation (DL)" on page 475.
• "Traffic and Pilot C/N Calculation (DL)" on page 482.
• "Traffic C/(I+N) Calculation (DL)" on page 483.
• "Traffic Signal Level Calculation (UL)" on page 484.
• "Noise Calculation (UL)" on page 485.
• "Traffic Interference Calculation (UL)" on page 486.
• "Traffic C/N Calculation (UL)" on page 487.
• "Traffic C/(I+N) Calculation (UL)" on page 490.
472 AT271_TRG_E6 © Forsk 2009

9.3.6.1 Traffic and Pilot Signal Level Calculation (DL)

Input
TX i  ic 
• P Preamble : Preamble transmission power of the cell TXi(ic).
TX i  ic 
• P Traffic : Traffic power offset of the cell TXi(ic).
TX i  ic 
• P Pilot : Pilot power offset of the cell TXi(ic).
TX i
TX i TX i TX i
- Without smart antenna equipment: G is the transmitter antenna gain, i.e., G = G Ant .
TX i
- With smart antenna equipment: G is the smart antenna gain in the direction of the pixel, subscriber, or
TX i
mobile Mi, i.e., G = G SA    . Where  is the direction in which Mi is located. For more information on the
calculation of G SA    , refer to section "Optimum Beamformer Model" on page 505.
TX i TX i
• L : Total transmitter losses for the transmitter TXi ( L = L Total – DL ).
TX i
• L Path : Path loss ( L Path = L Model + L Ant ).
TX i
M Shadowing – Model = 0 when calculating the received traffic and pilot signal levels for C/(I+N) calculations.
Mi
Mi
Mi
The antenna used for the calculations at the pixel, subscriber, or mobile Mi is pointed towards its best serving cell.
Mi
This means that L Ant = 0 for calculating the traffic and pilot signal level.
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:
C Traffic = EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body and
C Pilot = EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
TX i  ic  TX i  ic  TX i TX i TX i  ic  TX i  ic  TX i TX i
EIRP Traffic = P Traffic + G –L and EIRP Pilot = P Pilot +G –L
With P Traffic and P Pilot being the traffic and pilot transmission powers of the cell TXi(ic) calculated as follows:
P Traffic = P Preamble – P Traffic and P Pilot = P Preamble – P Pilot
Output
TX i  ic 
• C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
© Forsk 2009 AT271_TRG_E6 473

9.3.6.2 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. In WiMAX 802.16e, the numbers of subcar-
riers used by each permutation zone 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
• N SCa – Used : Number of used subcarriers defined in the Global Parameters (WiMAX 802.16d).
• N SCa – Total : Total number of subcarriers defined in the Global Parameters (WiMAX 802.16d).
M
i
PZ DL
• N SCa – Used : Number of subcarriers used by the downlink permutation zone of a WiMAX 802.16e cell TXi(ic)
assigned to Mi.
TX i  ic 
• N SCa – Total : Total number of subcarriers defined for the frame configuration of a WiMAX 802.16e cell TXi(ic).
TX i  ic 
page 493.
Mi
• nf : Noise figure of the terminal used for calculations by the pixel, subscriber, or mobile Mi.
Calculations
n 0 = 10  Log 10  K  T  1000  = – 174 dBm/Hz
The thermal noise for a cell is calculated as:
TX i  ic  TX i  ic  N SCa – Used
WiMAX 802.16d: n 0 – DL = n 0 + 10  Log 10  F Sampling  ------------------------------
M
 PZ DL
i

TX i  ic   TX i  ic  N SCa – Used
WiMAX 802.16e (DL): n 0 – DL = n 0 + 10  Log 10  F Sampling  -----------------------------
TX i  ic 
-
 
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.
n DL = n 0 – DL + nf
Effect of Segmentation (WiMAX 802.16e):
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 segmenting factor, f Segment , is calculated from the number of secondary subchannel groups assigned to the
permutation zone in the Permutation Zones table.
3  PSG + 2  SSG
f Segment = ----------------------------------------------------
15
Where, PSG is the number of primary subchannel groups, which is always 1, and SSG is the number of used
secondary subchannel groups, which can be 0, 1, 2, or 3. Segmenting factors for different numbers of secondary
subchannel groups can be.
Number of SSG f Segment
0 1/5
1 1/3
2 7/15
3 3/5
474 AT271_TRG_E6 © Forsk 2009

f Segment represents the fraction of the channel bandwidth used by a segment. The power transmitted over a
1
segment has ---------------------- times the spectral density of the power transmitted over the entire channel bandwidth.
f Segment
1
When calculating the downlink C/N and C/(I+N) ratios, the increase in power by ---------------------- due to this power
f Segment
concentration is equivalent to a reduction in the noise level by f Segment . Hence, if segmentation is used, the ther-
mal noise power at the pixel, subscriber, or mobile Mi covered by the segmented permutation zone is reduced by
a factor of f Segment . Which means that the thermal noise for the a segment of the channel used by a cell is calcu-
lated as:
M
 PZ DL
i

TX i  ic   TXi  ic  N SCa – Used 
n 0 – DL TX i  ic 
-  f Segment
= n 0 + 10  Log 10  F Sampling  -----------------------------
 N SCa – Total 
 
Output
TX i  ic 
• n DL : Downlink noise for the cell TXi(ic).
9.3.6.3 Traffic and Pilot Interference Calculation (DL)

The interference received by at 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 segmentation or not. Moreover, the interference can come from cells using classic as well as smart antennas.
The calculation can be divided into the two parts.
• "Traffic and Pilot Interference Signal Levels Calculation (DL)" on page 475.
• "Effective Traffic and Pilot Interference Calculation (DL)" on page 478.
9.3.6.3.1 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
TX j  jc 
• P Preamble : Preamble transmission power of the cell TXj(jc).
TX j  jc 
• P Pilot : Pilot power offset of the interfering cell TXj(jc).
TX j  jc 
• P Traffic : Traffic power offset of the interfering cell TXj(jc).
TX j  jc 
• P Idle – Pilot : Idle pilot power offset of the interfering cell TXj(jc).
TX j TX j
• L : Total transmitter losses for the transmitter TXj ( L = L Total – DL ).
TX j
TX j
• L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXj.
Mi
Mi
Mi
Mi Mi
When calculating L Ant , Mi’s antenna is pointed towards its best serving cell TXi(ic). L Ant is calculated in the direc-
tion of TXj(jc) from the antenna patterns of the antenna used by Mi while the antenna is pointed towards TXi(ic).
Mi
TX j  jc 
• TL DL : Downlink traffic load of the interfering cell TXj(jc).
© Forsk 2009 AT271_TRG_E6 475

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 455.
TX j  jc 
• AU : AAS usage ratio of the interfering cell TXj(jc).
AAS usage ratios are calculated using Monte Carlo simulations as explained in "Simulation Process" on page 455.
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 inter-
ference source. Therefore, the interference received from any cell TXj(jc) can be given by:
 I Non – AAS I Idle 
TX j  jc   --------------------------
10
- ------------------
10
Without smart antennas: I Total = 10  Log  10 + 10 
 
 
TXj  jc 
 I AAS 
TX j  jc   -----------------
10 
-
With smart antennas: I Total = 10  Log  10 
 
 
• Coverage Predictions:
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 j  jc  TX j  jc  TX j  jc 
 INon – AAS I Idle I AAS 
TX j  jc   --------------------------
10
- ------------------
10
------------------
10
I Total = 10  Log  10 + 10 + 10 
 
 
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 transmitter antenna,
TX j  jc 
• I AAS : interference from the loaded part of the frame transmitted using the smart antenna,
TX j  jc 
• I Idle : 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 transmitter 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:
TX j  jc  TX j  jc  Mi Mi Mi Mi
I Traffic = EIRP Traffic – L Path – L Indoor + G –L – L Ant – L Body
I Pilot = EIRP Pilot – L Path – L Indoor + G –L – L Ant – L Body
TX j  jc  TX j  jc  TX j TX j TX j  jc  TX j  jc  TX j TX j
EIRP Traffic = P Traffic + G –L and EIRP Pilot = P Pilot +G –L
With P Traffic and P Pilot being the traffic and pilot transmission powers of the cell TXj(jc) calculated as follows:
476 AT271_TRG_E6 © Forsk 2009

TX j  jc  TX j  jc  TX j  jc  TX j  jc  TX j  jc  TX j  jc 
P Traffic = P Preamble – P Pilot and P Pilot = P Preamble – P Pilot
TX j TX j
And, G = 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 transmitter antenna is given as:
  ITraffic TX j  jc  I Pilot TX  jc  TX j  jc  
 TX j  jc   ----------------- ------------------ N j
SCa – Used – N SCa – Data 
TX j  jc  TX j  jc  - N SCa – Data
10
I Non – AAS = 10  Log  TL DL   1 – AU    10 - + 10 10  ----------------------------------------------------------------
 ----------------------------- - 
  N SCa – Used N SCa – Used 
  
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:
I AAS = EIRP AAS – L Path – L Indoor + G –L – L Ant – L Body
EIRP AAS = P Traffic + G –L
TX j  jc 
With P Traffic being the traffic transmission power of the cell TXj(jc) calculated as follows:
P Traffic = P Preamble – P Pilot
TX j
And, G is the smart antenna gain in the direction of the victim mobile Mi, calculated as follows:
- Monte Carlo Simulations:

TX j
G = G SA    is calculated in the direction of the victim mobile Mi, while the smart antenna at the interfering
cell has a beam formed in the direction of an interfering mobile Mj.  is the direction, with respect to the azi-
muth of the cell TXj(jc), in which the victim mobile Mi is located. For more information on the calculation of
G SA    , refer to section "Optimum Beamformer Model" on page 505. Victim and interfering mobiles are gen-
erated by a time-slot scenario as explained in "Simulation Process" on page 455.
1 TX i  ic  2
In the figure below, G SA    is calculated from the victim cell TXi(ic) to calculate C Traffic , and G SA    is cal-
TX j  jc 
culated from the interfering cell TXj(jc) to calculate I AAS . In other words, a beam is formed from the victim
cell towards the victim mobile, and another beam is formed by the interfering cell towards the interfering mo-
bile.
Figure 9.5: Downlink C/(I+N) calculation in Simulations
- Coverage Predictions:
TX j
G = G SA    is calculated in the direction of the victim mobile Mi from the angular distribution of the down-
link traffic power density available at the end of the simulations. The angular distribution of the downlink traffic
power density, which represents the average array correlation matrix, is calculated during Monte Carlo simu-
lations and can be stored in the Cells table.  is the direction in which the victim pixel or subscriber Mi is lo-
cated. For more information on the calculation of G SA    , refer to section "Optimum Beamformer Model" on
page 505.
1 TX i  ic  2
In the figure below, G SA    is calculated from the victim cell TXi(ic) to calculate C Traffic , and G SA    is cal-
TX j  jc 
culated from the interfering cell TXj(jc) to calculate I AAS . In other words, a beam is formed from the victim
cell towards the victim pixel or subscriber, and the interference is calculated from the simulation results.
© Forsk 2009 AT271_TRG_E6 477

Figure 9.6: Downlink C/(I+N) calculation in Coverage Predictions
The average array correlation matrix (at the interfering cell) are given by:
j=1
Where R Avg is the average array correlation matrix, J is the number of served mobiles during the simulation,
 j is the probability of presence of the mobile j, p j is the EIRP transmitted towards the mobile j, and R j is the
array correlation matrix for the mobile j.
The probability of presence of the mobile j is the ratio between the downlink resources provided to the mobile
j and the total amount of available downlink resources. For example, if a mobile has been granted 10% of the
number of available slots in the downlink subframe, it’s probability of presence is 10%.
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 transformation,
R Avg is the average array correlation matrix, and g n    is the gain of the nth antenna element in the direction
of  .
The interference from the empty, or idle, part of the frame transmitted using the transmitter antenna is
The received interfering pilot signal level (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile
Mi as follows:
I Idle – Pilot = EIRP Idle – Pilot – L Path – L Indoor + G –L – L Ant – L Body
EIRP Idle – Pilot = P Idle – Pilot + G –L
TX j  jc 
With P Idle – Pilot being the idle pilot transmission power of the cell TXj(jc) calculated as follows:
P Idle – Pilot = P Preamble – P Idle – Pilot
TX j TX j
And, G = 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 j  jc 
  I Idle – Pilot TX  jc  TX j  jc  
 - N j
 --------------------------
TX j  jc  TX j  jc  10 SCa – Used – N SCa – Data 
I Idle = 10  Log   1 – TL DL    10  ----------------------------------------------------------------
TX j  jc 
- 
  N SCa – Used 
  
Output
TX j  jc 
• I Total : Interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc).
9.3.6.3.2 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 segmentation parameters of the studied and interfering cells (WiMAX
802.16e). 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.
478 AT271_TRG_E6 © Forsk 2009

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 475.
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  TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
I DL = I Traffic + f O + f Seg
Calculations for the interference reduction factors due to channel overlapping, downlink traffic load, and 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 calcu-
lated as follows:
fO = 10  Log  r O 
Interference reduction due to segmentation (WiMAX 802.16e):
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:
a. 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 seg-
1
mented zone has ---------------------- times the spectral density of the power transmitted over the entire channel band-
f Segment
1
width. When calculating the C/(I+N) ratio, the increase in power by ---------------------- is equivalent to decreasing the
f Segment
noise and interference by f Segment . Hence, if segmentation is used, the interference received at the pixel, sub-
scriber, or mobile Mi covered by the segmented zone is reduced by a factor of f Segment .
b. Collision probability between the subcarriers used by the subchannels belonging to the segment of the studied
cell and the subcarriers used by other sectors, segmented or not. The following paragraphs explain how the
collision probability is calculated.
The Segmentation Usage (SU) ratio is the percentage of the total downlink traffic load present in the segment-
ed downlink PUSC zone. For example, if the downlink traffic load is 80 %, and the segmentation usage ratio
is 50 %, then this means that the downlink traffic load of the segmented zone is 40 % (i.e., 50 % of 80 %), and
the downlink traffic load of the non-segmented zones is 40 %.
In coverage predictions, Atoll uses the segmentation usage ratios stored in the cell properties for determining
the interference. In simulations, Atoll resets the segmentation usage ratios for all the cells to 0, and then cal-
culates the segmentation usage ratios according to the traffic loads of the mobiles allocated to the segmented
zone and in the non-segmented zones.
Figure 9.7: Segmentation
Atoll determines the switching point between the segmented and the non-segmented zones using the seg-
mentation usage ratio.
© Forsk 2009 AT271_TRG_E6 479

The switching points between the segmented and non-segmented zones of the victim and interfering cells,
TXi(ic) and TXj(jc) respectively, are calculated as follows:
TX i  ic 
TX i  ic  SU
SP = ------------------------------------------------------------------------------------------------ and
SU + f Segment   1 – SU 
TX j  jc 
TX j  jc  SU
SP = ------------------------------------------------------------------------------------------------
SU + f Segment   1 – SU 
Where, SP is the switching point between the segmented and the non-segmented zones, SU is the segmen-
tation usage ratios of the cells, and f Segment is segmenting factor, which gives the bandwidth used by a seg-
ment.
The segmenting factor, f Segment , is calculated from the number of secondary subchannel groups assigned to
the first downlink PUSC permutation zone in the Permutation Zones table.
3  PSG + 2  SSG
f Segment = ----------------------------------------------------
15
Where, PSG is the number of primary subchannel groups, which is always 1, and SSG is the number of sec-
ondary subchannel groups, which is 1 for FFT sizes 128 and 512, and can be 0, 1, 2, or 3 for FFT sizes 1024
and 2048. Segmenting factors for different numbers of secondary subchannel groups are:
SSG f Segment
0 1/5
1 1/3
2 7/15
3 3/5
Note:
• 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.
If the segmentation usage ratio is set to 0, it means that the segmented zone does not exist. Setting SU to 0
gives SP = 0, and setting SU to 1 gives SP = 1 (or 100%), which shows how the switching point varies with
the segmentation usage ratio.
Derivation of the switching point formula:

• The Segmentation Usage ratio is used to partition the total downlink traffic load into
segmented and non-segmented zones. Therefore, the switching point formula is derived
from the equation:
SU  TL DL  1 – SU   TL DL
---------------------------------------------------------------------------------
- = -------------------------------------------------------------------
-
SP  f Segment  W Channel  1 – SP   W Channel
With cells using segmentation, there can be four different interference scenarios.
- Between the segmented zone of the victim and the segmented zone of the interferer.
- Between the segmented zone of the victim and the non-segmented zone of the interferer.
- Between the non-segmented zone of the victim and the segmented zone of the interferer.
- Between the non-segmented zone of the victim and the non-segmented zone of the interferer.
480 AT271_TRG_E6 © Forsk 2009

Figure 9.8: Segmentation Interference Scenarios
Therefore, Atoll calculates the probabilities of collision for each scenario and weights the total interference ac-
cording to the total collision probability. The probability of collision p Coll for each scenario is given by the fol-
lowing formula:
3  PSG Com + 2  SSG Com

p Coll = ---------------------------------------------------------------------------------
-
3  PSG + 2  SSG
Where, PSGCom is the number of primary subchannel groups common in TXi(ic) and TXj(jc), which can be
either 0 or 1, SSGCom is the number of secondary subchannel groups common in TXi(ic) and TXj(jc),
PSG is the number of primary subchannel groups in the cell TXi(ic), and SSG is the number of
secondary subchannel groups in the cell TXi(ic). The numbers of primary and secondary subchannel groups
for different segments are:
Number of Primary Number of Secondary

Segment
Subchannel Groups (PSG) Subchannel Groups (SSG)
0 1 (Group 0) m
1 1 (Group 2) m
2 1 (Group 4) m
Where m = 0, 1, 2, or 3 for FFT sizes 1024 and 2048, and m = 1 for FFT sizes 128 and 512.
The numbers of primary and secondary subchannel groups for the different cases are given in the table below:
TX i  ic  TX i  ic  TX j  jc  TX j  jc 
p Coll TX i  ic  TX j  jc  PSG SSG PSG SSG PSG Com SSG Com
NN Not Not
p Coll 3 3 3 3 3 3
segmented segmented
NS Not
p Coll Segmented 3 3 1 m 1 m
segmented
SN Not
p Coll Segmented 1 m 3 3 1 m
segmented
1 if same m in TXi(ic)
SS
p Coll Segmented Segmented 1 m in TXi(ic) 1 m’ in TXj(jc) segment 
0 otherwise m’ in TXj(jc)
The segment numbers and the cell permutation base numbers (Cell PermBase) are determined from the cell’s
preamble index. The mapping between the preamble index, the segment number, and Cell PermBase is avail-
able in the IEEE specifications. This mapping is performed in Atoll as follows:
Preamble Index ( PI )
PI  96 96  PI  114
Range: 0 to 113
Cell PermBase ( PB )
PI Modulo 32 PI – 96
Range: 0 to 31
© Forsk 2009 AT271_TRG_E6 481

Segment Number ( N Seg ) PI

Floor  ------  PI – 96  Modulo 3
 32
Range: 0, 1, 2
There can be 2 cases for calculating the total probability of collision.
- Case 1: If the pixel, subscriber, or mobile Mi is covered by the segmented zone of TXi(ic), the total collision
probability for the pixel, subscriber, or mobile Mi is calculated as follows:
 SS TX j  jc  TX i  ic 
 p Coll If SP  SP
TX i  ic  – TX j  jc  
p Collision =  SS TX j  jc  SN TX i  ic  TX j  jc 
p Coll  SP + p Coll   SP – SP 
 --------------------------------------------------------------------------------------------------------------------------- TX  jc  TX i  ic 
- If SP j  SP
 TX i  ic 
 SP
- Case 2: If the pixel, subscriber, or mobile Mi is covered by the non-segmented zone of TXi(ic), the total
collision probability for the pixel, subscriber, or mobile Mi is calculated as follows:
 NN TX j  jc  TX i  ic 
 p Coll If SP  SP
TX i  ic  – TX j  jc   TX j  jc  TX j  jc  TX i  ic 
p Collision =  NN NS
 p Coll   1 – SP  + p Coll   SP – SP  TX j  jc  TX i  ic 
------------------------------------------------------------------------------------------------------------------------------------------ If SP  SP
 TX i  ic 
  1 – SP 
The interference reduction factor due to segmentation for the pixel, subscriber, or mobile Mi is calculated as
follows:
f Seg = 10  Log 10  p Collision 
Output
TX j  jc 
• I DL : Effective downlink traffic and pilot interference received at the pixel, subscriber, or mobile Mi from any
interfering cell TXj(jc).
9.3.6.4 Traffic and Pilot C/N Calculation (DL)

Input
TX i  ic 
• C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in
"Traffic and Pilot Signal Level Calculation (DL)" on page 473.
TX i  ic 
• C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in
TX i  ic 
• n DL : Downlink noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 474.
TX i  ic 
• CNR Preamble : Preamble C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Preamble
TX i  ic 
• T AMS : AMS threshold defined for the cell TXi(ic).
TX i  ic 
• N Ant – TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic).
Mi
• N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel, subscriber,
or mobile Mi.
• Mobility  M i  : Mobility used for the calculations.
DL
• G STTD : Downlink STTD/MRC gain from the MIMO Configurations table for the number of transmission antennas,
TX i  ic  Mi
N Ant – TX , the number of reception antennas, N Ant – RX , and user mobility, Mobility  M i  .
DL
• G STTD : Downlink STTD/MRC gain offset defined for the clutter class of the pixel, subscriber, or mobile Mi.
Calculations
The traffic and pilot C/N for a cell TXi(ic) are calculated as follows for any pixel, subscriber, or mobile Mi:
CNR Traffic = C Traffic – n DL
CNR Pilot = C Pilot – n DL
482 AT271_TRG_E6 © Forsk 2009

MIMO – STTD/MRC Gain:
If the permutation zone (WiMAX 802.16e) or the cell (WiMAX 802.16d) supports STTD/MRC or AMS, the downlink
STTD/MRC gain corresponding to the numbers of transmission and reception antennas, and the STTD/MRC gain
offset defined for the clutter class of the user are applied to the downlink C/N.
Therefore, the traffic and pilot C/N calculated above become:
In case of STTD/MRC:
CNR Traffic = CNR Traffic + G STTD + G STTD
CNR Pilot = CNR Pilot + G STTD + G STTD
In case of AMS:
TX i  ic  TX i  ic  DL DL TX i  ic  TX i  ic 
CNR Traffic = CNR Traffic + G STTD + G STTD if CNR Preamble  T AMS
CNR Pilot = CNR Pilot + G STTD + G STTD if CNR Preamble  T AMS
Output
TX i  ic 
• CNR Traffic : Traffic C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• CNR Pilot : Pilot C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
9.3.6.5 Traffic C/(I+N) Calculation (DL)

The carrier signal to interference and noise ratio is calculated in three steps. First Atoll calculates the received signal level
from the studied cell (as explained in "Traffic and Pilot Signal Level Calculation (DL)" on page 473) at the pixel, subscriber,
or mobile under study. Secondly, Atoll calculates the interference received at the same studied pixel, subscriber, or mobile
from all the interfering cells (as explained in "Traffic and Pilot Interference Calculation (DL)" on page 475). Interference
from each cell is weighted according to the co- and adjacent channel overlap between the studied and the interfering cells,
the traffic loads of the interfering cells, and (in WiMAX 802.16e) the probabilities of subcarrier collision if segmentation is
used. Finally, 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 474).
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 Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in
TX i  ic 
• n DL : Downlink noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 474.
TX j  jc 
• I DL : Effective downlink traffic and pilot interference from any cell TXj(jc) calculated for a pixel, subscriber, or
mobile Mi covered by a cell TXi(ic) as explained in "Traffic and Pilot Interference Calculation (DL)" on page 475.
• This parameter is taken into account when the option "Shadowing taken into account" is selected.
TX i  ic 
TX i  ic 
TX i  ic 
Mi
or mobile Mi.
DL
• G STTD : Downlink STTD/MRC gain from the MIMO Configurations table for the number of transmission antennas,
TX i  ic  Mi
DL
• G STTD : Downlink STTD/MRC gain offset defined for the clutter class of the pixel, subscriber, or mobile Mi.
© Forsk 2009 AT271_TRG_E6 483

Calculations
The traffic C/(I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
 TX  jc 
TX i  ic  TX i  ic   I DL n DL
-

------------------ -------------------
CINR Traffic = C Traffic – 10  Log  10 + 10  – M Shadowing – C  I
 10 10 
The Traffic Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
j i
 I DL n DL 
TX i  ic   ------------------ --------------------

10 10
I + N  DL = 10  Log  10 + 10 
 
If the permutation zone (WiMAX 802.16e) or the cell (WiMAX 802.16d) supports STTD/MRC or AMS, the downlink
offset defined for the clutter class of the user are applied to the downlink C/(I+N).
Therefore, the traffic C/(I+N) calculated above become:
In case of AMS:
CINR Traffic = CINR Traffic + G STTD + G STTD if CNR Preamble  T AMS
Output
TX i  ic 
• CINR Traffic : Traffic C/(I+N) from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
•  I + N  DL : Traffic Total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by
a cell TXi(ic).
9.3.6.6 Traffic Signal Level Calculation (UL)

Input
Mi
• P Max : Maximum transmission power of the terminal used by the pixel, subscriber, or mobile Mi without power con-
trol.
Mi
• P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi after power control
as calculated in "Traffic C/(I+N) Calculation (UL)" on page 490.
TX i
• E SA : Number of antenna elements defined for the smart antenna equipment used by the transmitter TXi.
TX i
TX i TX i TX i
- Without smart antenna equipment: G is the transmitter antenna gain, i.e., G = G Ant .
TX i
- With smart antenna equipment: G is the uplink smart antenna beam forming gain, i.e.,
TX i TX i
G = G SA = 10  Log  E SA  . For more information on the calculation of G SA , refer to section "Optimum
Beamformer Model" on page 505.
TX i TX i
• L : Total transmitter losses for the transmitter TXi ( L = L Total – UL ).
TX i
TX i
M Shadowing – Model = 0 when calculating the received signal level from interfering mobiles.
484 AT271_TRG_E6 © Forsk 2009

Mi
Mi
Mi
Mi
When calculating L Ant , Mi’s antenna is pointed towards its best serving cell. If TXi(ic) is the best serving cell for
Mi Mi Mi
Mi, then L Ant = 0 for calculating C UL . If TXi(ic) is not the best serving cell for Mi, then L Ant is calculated in the
direction of TXi(ic) from the antenna patterns of the antenna used by Mi while the antenna is pointed towards Mi’s
best serving cell.
Mi
Calculations
The received traffic signal level (dBm) from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated as
follows:
Mi Mi TX i TX i Mi Mi
C UL = EIRP UL – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
Where EIRP is the effective isotropic radiated power of the terminal calculated as follows:
Mi Mi Mi Mi
EIRP UL = P +G –L
Mi Mi Mi Mi
With P = P Max without power control at the start of the calculations, and is the P = P Eff after power control.
Output
Mi
• C UL : Received uplink signal level from the pixel, subscriber, or mobile Mi at a cell TXi(ic).
9.3.6.7 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 used bandwidth depends on the number of used subcarriers. In WiMAX 802.16e, the numbers of subcarriers used by
each permutation zone can be different.
The uplink 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
• N SCa – Used : Number of used subcarriers defined in the Global Parameters (WiMAX 802.16d).
• N SCa – Total : Total number of subcarriers defined in the Global Parameters (WiMAX 802.16d).
Mi
PZ UL
• N SCa – Used : Number of subcarriers used by the uplink permutation zone of a WiMAX 802.16e cell TXi(ic)
assigned to Mi.
TX i  ic 
TX i  ic 
page 493.
TX i  ic 
• nf : Noise figure of the cell TXi(ic).
Calculations
n 0 = 10  Log 10  K  T  1000  = – 174 dBm/Hz
The thermal noise for a cell is calculated as:
TX i  ic  TX i  ic  N SCa – Used
WiMAX 802.16d: n 0 – UL = n 0 + 10  Log 10  F Sampling  ------------------------------
© Forsk 2009 AT271_TRG_E6 485

M
 PZ UL
i

TX i  ic   TX i  ic  N SCa – Used
WiMAX 802.16e (UL): n 0 – UL = n 0 + 10  Log 10  F Sampling  -----------------------------
TX i  ic 
-
 
The uplink noise is the sum of the thermal noise and the noise figure of the cell TXi(ic).
n UL = n 0 – UL + nf
Output
TX i  ic 
• n UL : Uplink noise for the cell TXi(ic).
9.3.6.8 Traffic Interference Calculation (UL)

The uplink traffic interference is only calculated in Monte Carlo simulations. In coverage predictions, the uplink noise rise-
values already available in simulations on in the Cells table are used.
The interference received by a cell TXi(ic) from an interfering mobiles 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), on the traffic loads of the interfering mobiles Mj.
The calculation of uplink interference can be divided into two parts:

• Calculation of the uplink interference from an individual interfering mobile as calculated in "Traffic Interference
Signal Levels Calculation (UL)" on page 486.
• Calculation of the uplink noise rise which represents the total uplink interference from all the interfering mobiles as
calculated in "Noise Rise Calculation (UL)" on page 487.
9.3.6.8.1 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 calcu-
lated in "Traffic Signal Level Calculation (UL)" on page 484.
Mj
• TL UL : Uplink traffic load of the interfering mobile Mj.
Traffic loads are calculated during Monte Carlo simulations as explained in "Simulation Process" on page 455.
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  Mj
I UL = C UL + f O + 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- 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 calcu-
lated as follows:
fO = 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:
- Without smart antenna: All the mobiles present in other cells TXj(jc) that are transmitting in uplink contribute
to the interference received by TXi(ic) in uplink.
Mj Mj
f TL – UL = 10  Log 10  TL UL 
- With smart antenna: A set of interfering mobiles is generated for each mobile being served by the victim cell
TXi(ic). The probability of generating a mobile as an interferer depends on its uplink traffic load (see "Time-
Slot Scenario:" on page 457). As the traffic load has already been taken into account for generating the list of
interfering mobiles,
486 AT271_TRG_E6 © Forsk 2009

Mj
f TL – UL = 0
Output
Mj
• I UL : Uplink interference signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc).
9.3.6.8.2 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). More specifically, it is
the ratio (I+N)/N.
Input
Mj
• I UL : Uplink interference signal levels received at a cell TXi(ic) from interfering mobiles Mj covered by other cells
TXj(jc).
TX i  ic 
• n UL : Uplink noise for the cell TXi(ic).
Calculations
The uplink noise for the cell TXi(ic) is calculated as follows:
 M
j
TX  ic 
i
 I UL n UL 
TX i  ic   -------
-
10 
-------------------
- TX i  ic 

10
NR UL = 10  Log 10  10 + 10  – n UL
 
 All M j 
j
• With smart antenna:

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. The angular distribution of the uplink noise rise is given
by:
2
I UL    +  n  I
NR UL    = -----------------------------------
2
n  I
The Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
 M
j
TX  ic 
i
 I UL n UL 
TX i  ic   -------
-
10 
-------------------
-

10
I + N  UL = 10  Log 10  10 + 10 
 
 All M j 
j
Output
• NR UL or NR UL    : Uplink noise rise or the angular distribution of the uplink noise rise for the cell TXi(ic).
TX i  ic 
• I + N UL : Total Noise for a cell TXi(ic) calculated for any pixel, subscriber, or mobile Mi.
9.3.6.9 Traffic C/N Calculation (UL)

Input
Mi
• C UL : Received uplink signal level from the pixel, subscriber, or mobile Mi at its serving cell TXi(ic) as calculated
in "Traffic Signal Level Calculation (UL)" on page 484.
TX i  ic 
• n UL : Uplink noise for the cell TXi(ic) as calculated in "Noise Calculation (UL)" on page 485.
TX i  ic 
TX i  ic 
© Forsk 2009 AT271_TRG_E6 487

Mi
• N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel, subscriber,
or mobile Mi.
TX i  ic 
• N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic)..
UL
• G STTD : Uplink STTD/MRC gain from the MIMO Configurations table for the number of transmission antennas,
Mi TX i  ic 
UL
• G STTD : Uplink STTD/MRC gain offset defined for the clutter class of the pixel, subscriber, or mobile Mi.
• N SC – UL : Number of subchannels per channel (WiMAX 802.16d).
TX i  ic 
• T B – Lowest : Bearer selection threshold of the lowest bearer in the reception equipment assigned to the cell TXi(ic).
TX i  ic 
• T B – Highest : Bearer selection threshold of the highest bearer in the reception equipment assigned to the cell
TXi(ic) or defined in the properties of the service being accessed by the pixel, subscriber, or mobile Mi.
The bearer assigned to the pixel, subscriber, or mobile Mi is determined as explained in "Bearer Determination"
on page 496.
Mi
PZ UL
• N SC : Number of subchannels per channel defined for the uplink permutation zone assigned to the pixel, sub-
scriber, or mobile Mi as calculated in "Permutation Zone Selection (WiMAX 802.16e)" on page 471.
Mi
• P Max : Maximum transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
Mi
• P Min : Minimum transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
• M PC : Power control margin defined in the Global Parameters.
Calculations
The uplink C/N from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated as follows:
Mi Mi TX i  ic 
CNR UL = C UL – n UL
If the permutation zone (WiMAX 802.16e) or the cell (WiMAX 802.16d) supports STTD/MRC or AMS, the uplink
offset defined for the clutter class of the user are applied to the uplink C/N.
Therefore, the uplink C/N calculated above becomes:

Mi Mi UL UL
CNR UL = CNR UL + G STTD + G STTD
In case of AMS:
Mi Mi UL UL TX i  ic  TX i  ic 
CNR UL = CNR UL + G STTD + G STTD if CNR Preamble  T AMS
Uplink Subchannelisation (WiMAX 802.16d):
Subchannelisation decreases the bandwidth used by a mobile hence increasing the power spectral density for
transmission, and decreasing the noise and interference received. WiMAX terminals can perform subchannelisa-
tion in uplink to improve uplink coverage. In WiMAX 802.16d networks, if a terminal is unable to connect to its serv-
ing cell using all 16 subchannels, it can use less number of subchannels (8, 4, 2, or 1) in the uplink in order to
concentrate its transmission power on a smaller bandwidth, hence increasing its power spectral density and
increasing transmission range.
The uplink C/N is calculated above using the number of subchannels per channel set in the Global Parameters,
which is 16 by default. The number of subchannels in uplink is provided by the IEEE specifications.
N SC – UL = 16 8 4 2 or 1
If the uplink C/N from the pixel, subscriber, or mobile Mi is not enough to get the lowest bearer, i.e.,
Mi TX i  ic 
CNR UL  T B – Lowest , Atoll performs subchannelisation as follows:
Mi TX i  ic 
Until CNR UL  T B – Lowest OR N SC – UL  1 , Atoll improves the uplink C/N by adding the subchannelisation gain
to it.
488 AT271_TRG_E6 © Forsk 2009

Mi Mi
CNR UL = CNR UL + G SC – UL and reducing the number of subchannels used in the uplink by half.
The value of G SC – UL is calculated by determining the number of used subcarriers as follows:
Subchannelisation From Number of Used Subcarriers G SC – UL
200 to 100. 200

16 to 8 10  Log  ---------- = 3 dB
(192 Data + 8 Pilot to 96 Data + 4 Pilot)  100
100 to 50. 100

8 to 4 10  Log  ---------- = 3 dB
(96 Data + 4 Pilot to 48 Data + 2 Pilot)  50 
50 to 25. 50
4 to 2 10  Log  ------ = 3 dB
25 to 13. 25
2 to 1 10  Log  ------ = 2.84 dB
Even if after performing subchannelisation, the uplink C/N from the pixel, subscriber, or mobile Mi is not enough
Mi TX i  ic 
to get a bearer in the uplink, i.e., CNR UL  T B – Lowest , the pixel, subscriber, or mobile Mi is considered as covered
by the cell TXi(ic) in the uplink.
If you want to turn off subchannelisation in uplink, you can set the number of subchannels per channel to 1 in the
Global Parameters.
Uplink Subchannelisation (WiMAX 802.16e):
transmission, and decreasing the noise and interference received. The uplink C/N calculated above is given for
the number of subchannels defined for the permutation zone assigned to the pixel, subscriber, or mobile Mi in the
uplink.
In WiMAX 802.16e networks, subchannelisation is performed for all the pixels, subscribers, or mobiles in the
uplink. Slots are allocated to users first in the time domain and then in the frequency domain.
Figure 9.9: Uplink Subchannelization in WiMAX 802.16e Networks (Example)
The number of subchannels in uplink is provided by the IEEE specifications for each permutation zone. The uplink
C/N for any pixel, subscriber, or mobile Mi is determined for the entire channel bandwidth. To determine the
number of subchannels required to maintain the highest available bearer at the location of the pixel, subscriber,
or mobile, Atoll performs subchannelisation as follows:
M
 PZULi 
Mi Mi  N SC 
CNR UL = CNR UL + 10  Log 10  --------------------
-
 N Mi 
 SC – UL
Mi
 PZMi M 
Where N SC – UL = Min  N SC  N SC – UL  is the number of uplink subchannels after subchanneli-
UL i
 Mi Mi 
 B UL = B UL – Highest
sation with which the pixel, subscriber, or mobile Mi can get the highest available bearer. The highest available
Mi
bearer at the location of the pixel, subscriber, or mobile Mi ( B UL – Highest ) is the lowest one among:
© Forsk 2009 AT271_TRG_E6 489

- The highest bearer of the reception equipment of the serving cell of the pixel, subscriber, or mobile.
- The highest bearer of the service accessed by the pixel, subscriber, or mobile.
After subchannelisation, the number of uplink subchannels allocated to the pixel, subscriber, or mobile Mi is the
number of subchannels required to get the highest available bearer in the uplink.
Uplink Power Control:
Atoll performs power control in the uplink if the bearer available in the uplink for the pixel, subscriber, or mobile Mi
is the highest bearer. The highest bearer can be either the highest bearer defined in the properties of the service
used by Mi, or the highest bearer listed in the reception equipment of the serving cell TXi(ic). The pixel, subscriber,
or mobile Mi reduces its transmission power so that the uplink C/N from it at its cell is enough to get the highest
bearer.
Mi Mi Mi TX i  ic 
If with P = P Max AND CNR UL  T B – Highest + M PC , the transmission power of Mi is reduced to determine the
effective transmission power from the pixel, subscriber, or mobile Mi as follows:
Mi Mi Mi TX i  ic  Mi
P Eff = Max  P Max –  CNR UL –  T B – Highest + M PC   P Min 
Mi Mi
CNR UL is calculated again using P Eff .
Output
Mi
• CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
9.3.6.10 Traffic C/(I+N) Calculation (UL)

The carrier signal to interference and noise ratio is calculated in three steps. First, Atoll calculates the received signal level
from each pixel, subscriber, or mobile at its serving cell using the effective power of the terminal used by the pixel,
subscriber, or mobile as explained in "Traffic Signal Level Calculation (UL)" on page 484. Secondly, Atoll calculates the
uplink carrier to noise ratio as explained in "Traffic C/N Calculation (UL)" on page 487. And, finally, 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 487.
The uplink noise rise can be set by the user manually for each cell or calculated using the Monte Carlo simulations.
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
Mi
• CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
• NR UL or NR UL    : Uplink noise rise or the angular distribution of the uplink noise rise for the cell TXi(ic).
TX i  ic 
TX i  ic 
Mi
or mobile Mi.
TX i  ic 
• N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic)..
UL
• G STTD : Uplink STTD/MRC gain from the MIMO Configurations table for the number of transmission antennas,
Mi TX i  ic 
UL
• G STTD : Uplink STTD/MRC gain offset defined for the clutter class of the pixel, subscriber, or mobile Mi.
• N SC – UL : Number of subchannels per channel (WiMAX 802.16d).
TX i  ic 
• T B – Lowest : Bearer selection threshold of the lowest bearer in the reception equipment assigned to the cell TXi(ic).
490 AT271_TRG_E6 © Forsk 2009

TX i  ic 
• T B – Highest : Bearer selection threshold of the highest bearer in the reception equipment assigned to the cell
TXi(ic) or defined in the properties of the service being accessed by the pixel, subscriber, or mobile Mi.
The bearer assigned to the pixel, subscriber, or mobile Mi is determined as explained in "Bearer Determination"
on page 496.
Mi
PZ UL
• N SC : Number of subchannels per channel defined for the uplink permutation zone assigned to the pixel, sub-
scriber, or mobile Mi as calculated in "Permutation Zone Selection (WiMAX 802.16e)" on page 471.
Mi
• P Max : Maximum transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
Mi
• P Min : Minimum transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
• M PC : Power control margin defined in the Global Parameters.
Calculations
The uplink C/(I+N) for any pixel, subscriber, or mobile Mi at a cell TXi(ic) is calculated as follows:

CINR UL = CNR UL – NR UL – M Shadowing – C  I
• With smart antenna:

- Monte Carlo simulations: The uplink C/(I+N) is calculated as described in the section "Optimum Beamformer
Model" on page 505. Victim and interfering mobiles are generated by a time-slot scenario as explained in
"Simulation Process" on page 455.
- Coverage predictions: CINR UL    = CNR UL – NR UL    – M Shadowing – C  I

If the permutation zone (WiMAX 802.16e) or the cell (WiMAX 802.16d) supports STTD/MRC or AMS, the uplink
offset defined for the clutter class of the user are applied to the uplink C/(I+N).
Therefore, the uplink C/(I+N) calculated above becomes:
Mi Mi UL UL Mi Mi UL UL
CINR UL = CINR UL + G STTD + G STTD and CINR UL    = CINR UL    + G STTD + G STTD
In case of AMS:
CINR UL = CINR UL + G STTD + G STTD if CNR Preamble  T AMS and
CINR UL    = CINR UL    + G STTD + G STTD if CNR Preamble  T AMS
Uplink Subchannelisation (WiMAX 802.16d):
transmission, and decreasing the noise and interference received. WiMAX terminals can perform subchannelisa-
tion in uplink to improve uplink coverage. In WiMAX 802.16d networks, if a terminal is unable to connect to its serv-
ing cell using all 16 subchannels, it can use less number of subchannels (8, 4, 2, or 1) in the uplink in order to
concentrate its transmission power on a smaller bandwidth, hence increasing its power spectral density and
increasing transmission range.
The uplink C/(I+N) is calculated above using the number of subchannels per channel set in the Global Parameters,
which is 16 by default. The number of subchannels in uplink is provided by the IEEE specifications.
N SC – UL = 16 8 4 2 or 1
If the uplink C/(I+N) from the pixel, subscriber, or mobile Mi is not enough to get the lowest bearer, i.e.,
Mi TX i  ic 
CINR UL  T B – Lowest , Atoll performs subchannelisation as follows:
Mi TX i  ic 
Until CINR UL  T B – Lowest OR N SC – UL  1 , Atoll improves the uplink C/(I+N) by adding the subchannelisation
gain to it.
Mi Mi
CINR UL = CINR UL + G SC – UL and reducing the number of subchannels used in the uplink by half.
The value of G SC – UL is calculated by determining the number of used subcarriers as follows:
© Forsk 2009 AT271_TRG_E6 491

Subchannelisation From Number of Used Subcarriers G SC – UL
200 to 100. 200

16 to 8 10  Log  ---------- = 3 dB
100 to 50. 100

8 to 4 10  Log  ---------- = 3 dB
(96 Data + 4 Pilot to 48 Data + 2 Pilot)  50 
50 to 25. 50
4 to 2 10  Log  ------ = 3 dB
25 to 13. 25
2 to 1 10  Log  ------ = 2.84 dB
Even if after performing subchannelisation, the uplink C/(I+N) from the pixel, subscriber, or mobile Mi is not enough
Mi TX i  ic 
to get a bearer in the uplink, i.e., CINR UL  T B – Lowest , the pixel, subscriber, or mobile Mi is considered as
covered by the cell TXi(ic) in the uplink.
If you want to turn off subchannelisation in uplink, you can set the number of subchannels per channel to 1 in the
Global Parameters.
Uplink Subchannelisation (WiMAX 802.16e):
transmission, and decreasing the noise and interference received. The uplink C/(I+N) calculated above is given
for the number of subchannels defined for the permutation zone assigned to the pixel, subscriber, or mobile Mi in
the uplink.
In WiMAX 802.16e networks, subchannelisation is performed for all the pixels, subscribers, or mobiles in the
uplink. Slots are allocated to users first in the time domain and then in the frequency domain.
Figure 9.10: Uplink Subchannelization in WiMAX 802.16e Networks (Example)
The number of subchannels in uplink is provided by the IEEE specifications for each permutation zone. The uplink
C/(I+N) for any pixel, subscriber, or mobile Mi is determined for the entire channel bandwidth. To determine the
number of subchannels required to maintain the highest available bearer at the location of the pixel, subscriber,
or mobile, Atoll performs subchannelisation as follows:
M
 PZULi 
Mi Mi  N SC 
CINR UL = CINR UL + 10  Log 10  --------------------
-
 N Mi 
 SC – UL
Mi
 PZMi M 
Where N SC – UL = Min  N SC  N SC – UL  is the number of uplink subchannels after subchanneli-
UL i
 i
M M
i

 B UL = B UL – Highest
sation with which the pixel, subscriber, or mobile Mi can get the highest available bearer. The highest available
Mi
bearer at the location of the pixel, subscriber, or mobile Mi ( B UL – Highest ) is the lowest one among:
- The highest bearer of the reception equipment of the serving cell of the pixel, subscriber, or mobile.
- The highest bearer of the service accessed by the pixel, subscriber, or mobile.
492 AT271_TRG_E6 © Forsk 2009

After subchannelisation, the number of uplink subchannels allocated to the pixel, subscriber, or mobile Mi is the
number of subchannels required to get the highest available bearer in the uplink.
Uplink Power Control:
Atoll performs power control in the uplink if the bearer available in the uplink for the pixel, subscriber, or mobile Mi
is the highest bearer. The highest bearer can be either the highest bearer defined in the properties of the service
used by Mi, or the highest bearer listed in the reception equipment of the serving cell TXi(ic). The pixel, subscriber,
or mobile Mi reduces its transmission power so that the uplink C/(I+N) from it at its cell is enough to get the highest
bearer.
Mi Mi Mi TX i  ic 
If with P = P Max AND CINR UL  T B – Highest + M PC , the transmission power of Mi is reduced to determine the
effective transmission power from the pixel, subscriber, or mobile Mi as follows:
Mi Mi Mi TX i  ic  Mi
P Eff = Max  P Max –  CINR UL –  T B – Highest + M PC   P Min 
Mi Mi
CINR UL is calculated again using P Eff .
Output
Mi Mi
• CINR UL or CINR UL    : Uplink C/(I+N) from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
Mi
• N SC – UL : Number of subchannels used by the pixel, subscriber, or mobile Mi in the uplink after subchannelisation.
Mi
• P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
9.3.7 Channel Throughput Calculation

Channel throughputs are calculated in three steps.
• Calculation of uplink and downlink total resources in a cell as explained in "Calculation of Cell Capacity (Total
Resources)" on page 493.
• Determination of the uplink and downlink bearers as explained in "Bearer Determination" on page 496.
• Calculation of channel throughputs as explained in "Channel Throughput Calculation" on page 496.
9.3.7.1 Calculation of Cell Capacity (Total Resources)

Cell capacity is the total amount of resources available in a cell. The total amount of resources in a cell is the number of
modulation symbols that can be used for data transfer in each frame. The cell capacity can be calculated separately for
the downlink and the uplink subframes. The following sections describe how the cell capacities are calculated for TDD and
FDD networks.
9.3.7.1.1 Calculation of Sampling Frequency

Input
TX i  ic 
• f Sampling : Sampling factor defined for the frequency band of the cell TXi(ic).
TX i  ic 
• W Channel : Channel bandwidth of the cell TXi(ic).
Calculations
Atoll determines the sampling frequency as follows:

TX i  ic  6
TX i  ic   W Channel  10 
F Sampling = Floor  f Sampling  ----------------------------------------  8000
 8000 
Output
TX i  ic 
• F Sampling : Sampling frequency for the cell TXi(ic).
9.3.7.1.2 Calculation of Symbol Duration

Input
TX i  ic 
page 493.
© Forsk 2009 AT271_TRG_E6 493

TX i  ic 
• r CP : Cyclic prefix ratio defined for the network in the Global Parameters.
Calculations
From the sampling frequency, Atoll determines the inter-subcarrier spacing.

TX i  ic  –3
TX i  ic  F Sampling  10
F = ------------------------------------------
-
TX i  ic 
N SCa – Total
Atoll calculates the useful symbol duration.

TX i  ic  1
D Sym – Useful = ----------------------
TX i  ic 
F
And, the duration of the cyclic prefix.
r CP
D CP = --------
F
Adding the Cyclic prefix ratio to the useful symbol duration, Atoll determines the total symbol duration.
D Symbol = D Sym – Useful + D CP
Output
TX i  ic 
• D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic).
9.3.7.1.3 Calculation of Cell Capacity - TDD Networks

Input
• D Frame : Frame duration.

TDD
• D TTG : TTG duration.
TDD
• D RTG : RTG duration.
TX i  ic 
• D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic) as calculated in "Calculation of Symbol
Duration" on page 493.
TDD
• r DL – Frame : DL ratio.
DL
• O Fixed : Downlink fixed overhead.
DL
• O Variable : Downlink variable overhead.
UL
• O Fixed : Uplink fixed overhead.
UL
• O Variable : Uplink variable overhead.
TX i  ic 
• N SCa – Data : Number of data subcarriers for a WiMAX 802.16d cell TXi(ic).
M
i
PZ DL
• N SCa – Data : Number of data subcarriers of the downlink permutation zone of a WiMAX 802.16e cell TXi(ic)
assigned to Mi.
Mi
PZ UL
• N SCa – Data : Number of data subcarriers of the uplink permutation zone of a WiMAX 802.16e cell TXi(ic) assigned
to Mi.
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:
TX i  ic 
 D Used 
N  SD – Used   Frame = Floor  --------------------
Frame 
-
 TXi  ic  
 D Symbol
494 AT271_TRG_E6 © Forsk 2009

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 Parameters:
TX i  ic  TX i  ic  TDD DL
N  SD – DL   Subframe = RoundUp  N  SD – Used   Frame  r DL – Frame  – O Fixed
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:
DL
TX i  ic  TX i  ic   TX i  ic  TX i  ic   O Variable 
WiMAX 802.16d: R DL = N  Sym – DL   Subframe = Floor  N  SD – DL   Subframe  N SCa – Data   1 – -----------------------
  100  
M DL
i
TX i  ic  TX i  ic   TX i  ic  PZ DL  O Variable 
WiMAX 802.16e: R DL = N  Sym – DL   Subframe = 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
Parameters:
TX i  ic  TX i  ic  TDD UL
N  SD – UL   Subframe = RoundDown  N  SD – Used   Frame   1 – r DL – Frame   – O Fixed
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:
UL
TX i  ic  TX i  ic   TX i  ic  TX i  ic   O Variable 
WiMAX 802.16d: R UL = N  Sym – UL   Subframe = Floor  N  SD – UL   Subframe  N SCa – Data   1 – -----------------------
  100  
M UL
i
TX i  ic  TX i  ic   TX i  ic  PZ UL  O Variable 
WiMAX 802.16e: R UL = N  Sym – UL   Subframe = Floor  N  SD – UL   Subframe  N SCa – Data   1 – -----------------------
  100  
Output
• R DL = N  Sym – DL   Subframe : Amount of downlink resources in the cell TXi(ic).
• R UL = N  Sym – UL   Subframe : Amount of uplink resources in the cell TXi(ic).
9.3.7.1.4 Calculation of Cell Capacity - FDD Networks

The cell capacity 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.
Input

TX i  ic 
• D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic) as calculated in "Calculation of Symbol
Duration" on page 493.
X
• O Fixed : Downlink or uplink fixed overhead.
X
• O Variable : Downlink or uplink variable overhead.
TX i  ic 
• N SCa – Data : Number of data subcarriers for a WiMAX 802.16d cell TXi(ic).
M
i
PZ X
• N SCa – Data : Number of data subcarriers of the downlink or uplink permutation zone of a WiMAX 802.16e 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:
© Forsk 2009 AT271_TRG_E6 495

TX i  ic 
 DX 
N  SD – X   Subframe = Floor  -------------------------
Subframe
- – OX
 TXi  ic   Fixed
 D Symbol 
The total numbers of symbols in the downlink or uplink subframes after removing the variable overheads are:
X
TX i  ic  TX i  ic   TXi  ic  TX i  ic   O Variable 
WiMAX 802.16d: R X = N  Sym – X   Subframe = Floor  N  SD – X   Subframe  N SCa – Data   1 – -----------------------
  100  
i M X
TX i  ic  TX i  ic   TXi  ic  PZ X  O Variable 
WiMAX 802.16e: R X = N  Sym – X   Subframe = Floor  N  SD – X   Subframe  N SCa – Data   1 – -----------------------
  100  
Output
• RX = N  Sym – X   Subframe : Amount of downlink or uplink resources in the cell TXi(ic).
9.3.7.2 Bearer Determination

The second step in calculating the channel throughput for a cell at any pixel, subscriber, or mobile Mi within its service area
is to determine which bearer it is able to get in the downlink and in the uplink.
Input
TX i  ic 
• CINR Traffic : Traffic C/(I+N) from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Traffic C/(I+N)
Calculation (DL)" on page 483.
Mi
• CINR UL : Uplink C/(I+N) from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic) as calculated in "Traffic C/
(I+N) Calculation (UL)" on page 490.
Mi
• T B : Bearer selection thresholds of the bearers defined in the reception equipment used by Mi’s terminal.
TX i  ic 
• TB : Bearer selection thresholds of the bearers defined in the reception equipment used by the cell TXi(ic).
Mi
• B DL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel, sub-
scriber, or mobile Mi.
Mi
• B UL – Highest  Service  : Highest uplink bearer defined in the properties of the service used by the pixel, sub-
scriber, or mobile Mi.
Calculations
The bearer available for data transfer for the pixel, subscriber, or mobile Mi is either the highest bearer from among the
bearers whose selection thresholds are less than the traffic C/(I+N) at Mi, or the highest bearer defined for the service
being accessed by Mi, whichever is lower.
Mi  Mi 
B DL = Min  Best Bearers Mi TX i  ic 
 B DL – Highest  Service  and
 T B  CINRTraffic 
Mi  Mi 
B UL = Min  Best Bearers TX  ic  M
 B UL – Highest  Service 
 TB
i
 CINR UL
i 
Output
Mi
• B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink.
Mi
• B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink.
9.3.7.3 Channel Throughput Calculation

Channel throughputs are calculated for the entire channel resources allocated to the pixel, subscriber, or mobile Mi.
Input
TX i  ic 
• R DL : Amount of downlink resources in the cell TXi(ic) as calculated in "Calculation of Cell Capacity (Total
496 AT271_TRG_E6 © Forsk 2009

TX i  ic 
• R UL : Amount of uplink resources in the cell TXi(ic) as calculated in "Calculation of Cell Capacity (Total
•  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the downlink
i
B DL
in "Bearer Determination" on page 496.

•  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the uplink in
i
B UL
"Bearer Determination" on page 496.

Mi
• PZDL : Permutation zone assigned to the pixel, subscriber, or mobile Mi in the downlink.
• f Segment : Segmenting factor for the first downlink PUSC zone as calculated in "Effective Traffic and Pilot Interfer-
ence Calculation (DL)" on page 478.
TX i  ic 
TX i  ic 
• N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic).
Mi
or mobile Mi.
Mi
or mobile Mi.
• f SM – Gain : SM Gain Factor defined for the clutter class where the pixel, subscriber, or mobile Mi is located.
Max
• G SM : Max MIMO gain for a given number of transmission and reception antennas read from the Max MIMO Gain
vs. C/(I+N) graph from the MIMO Configurations table.
TX i  ic 
• CNR Preamble : Preamble C/N the cell TXi(ic) as calculated in "Preamble C/N Calculation" on page 470.
TX i  ic 
• T AMS : Adaptive MIMO switch threshold defined for the cell TXi(ic).
Mi TX i  ic 
• BLER  B DL  : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the reception
equipment assigned to the terminal used by the pixel, subscriber, or mobile Mi.
Mi Mi
• BLER  B UL  : Uplink block error rate read from the BLER vs. CINR UL graph available in the reception 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.
Mi
• TP Offset : Throughput offset defined in the properties of the service used by the pixel, subscriber, or mobile Mi.
Calculations
Downlink:
TX i  ic 
R DL  Mi
Mi 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 predic-
tions, while D Frame = 1 sec for Monte Carlo simulations.
Segmentation (WiMAX 802.16e):
Mi
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:
TX i  ic 
R DL  M
i
Mi B DL
CTP P – DL -  f Segment
= ------------------------------------
D Frame
MIMO - Spatial Multiplexing Gain:
If the permutation zone assigned to the pixel, subscriber, or mobile Mi (WiMAX 802.16e) or the cell (WiMAX
802.16d) supports SM or AMS, spatial multiplexing gain is applied to the bearer efficiency.
© Forsk 2009 AT271_TRG_E6 497

Max
Atoll reads the Max MIMO Gain, G SM , from the Max MIMO Gain graphs for the number of transmission antennas,
TX i  ic  Mi
N Ant – TX , the number of reception antennas, N Ant – RX , and user mobility, Mobility  M i  . It then applies this
spatial multiplexing gain to the bearer efficiency taking into account the SM Gain Factor defined for the clutter class
of the user Mi as follows:
Max
In case of SM:  M =  M   1 + f SM – Gain  G SM – 1  
i i
B DL B DL
In case of AMS:  Mi =  Mi   1 + f SM – Gain  G SM – 1   if CNR Preamble  T AMS
B DL B DL
Mi Mi Mi
• Effective MAC Channel Throughput: CTP E – DL = CTP P – DL   1 – BLER  B DL  
Mi
Mi Mi f TP – Scaling Mi
• Application Level Channel Throughput: CTP A – DL = CTP E – DL  ----------------------------- – TP Offset
100
Uplink:
TX i  ic 
R UL  Mi
Mi B UL
• Peak MAC Channel Throughput: CTP P – UL = ------------------------------------
-
D Frame
In the above formula, the actual value of D Frame is used to calculate the channel throughput for coverage predic-
tions, while D Frame = 1 sec for Monte Carlo simulations.
MIMO - Spatial Multiplexing Gain:
If the permutation zone assigned to the pixel, subscriber, or mobile Mi (WiMAX 802.16e) or the cell (WiMAX
802.16d) supports SM or AMS, spatial multiplexing gain is applied to the bearer efficiency.
Max
Atoll reads the Max MIMO Gain, G SM , from the Max MIMO Gain graphs for the number of transmission antennas,
Mi TX i  ic 
N Ant – TX , the number of reception antennas, N Ant – RX , and user mobility, Mobility  M i  . It then applies this
spatial multiplexing gain to the bearer efficiency taking into account the SM Gain Factor defined for the clutter class
of the user Mi as follows:
Max
In case of SM:  M =  M   1 + f SM – Gain  G SM – 1  
i i
B UL B UL
In case of AMS:  M =  M   1 + f SM – Gain  G SM – 1   if CNR Preamble  T AMS
i i
B UL B UL
Mi Mi Mi
• Effective MAC Channel Throughput: CTP E – UL = CTP P – UL   1 – BLER  B UL  
Mi
• Application Level Channel Throughput: CTP A – UL = CTP E – UL  ----------------------------- – TP Offset
100
Output
Mi
• CTP P – DL : Downlink peak MAC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP E – DL : Downlink effective MAC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP A – DL : Downlink application level channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP P – UL : Uplink peak MAC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP E – UL : Uplink effective MAC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP A – UL : Uplink application level channel throughput at the pixel, subscriber, or mobile Mi.
9.3.8 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 499 and the calculation of user throughputs is explained in "User Throughput Calculation"
on page 504.
498 AT271_TRG_E6 © Forsk 2009

9.3.8.1 Scheduling and Radio Resource Allocation

Input
TX i  ic 
• N Users – Max : Maximum number of users defined for the cell TXi(ic).
Mi
• QoS : QoS class of the service (UGS, ErtPS, rtPS, nrtPS, or Best Effort) accessed by a mobile Mi.
Mi
• p : Priority of the service accessed by a mobile Mi.
Mi
• TPD Min – DL : Downlink minimum throughput demand for the service accessed by a mobile Mi.
Mi
• TPD Min – UL : Uplink minimum throughput demand for the service accessed by a mobile Mi.
Mi
• TPD Max – DL : Downlink maximum throughput demand for the service accessed by a mobile Mi.
Mi
• TPD Max – UL : Uplink maximum throughput demand for the service accessed by a mobile Mi.
Mi TX i  ic 
• BLER  B DL  : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the reception
equipment assigned to the terminal used by the mobile Mi.
Mi Mi
• BLER  B UL  : Uplink block error rate read from the BLER vs. CINR UL graph available in the reception 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.
Mi
• TP Offset : Throughput offset defined in the properties of the service used by the mobile Mi.
Mi
• CTP P – DL : Downlink peak MAC channel throughput at the mobile Mi as calculated in "Channel Throughput Cal-
Mi
• CTP P – UL : Uplink peak MAC channel throughput at the mobile Mi as calculated in "Channel Throughput Calcu-
lation" on page 493.
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 i  ic 
The scheduler selects N Users mobiles for the scheduling and RRM process. If the Monte Carlo user distribution has
TX i  ic 
generated a number of users which is less than N Users – Max , the scheduler keeps all the mobiles generated for the cell
TXi(ic).
N Users = Min  N Users – Max N Users – Generated 
Sel TX i  ic 
For a cell, mobiles M i  N Users are selected for RRM by the scheduler.
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 maxi-
mum throughput demands can be considered as the peak MAC, effective MAC, or application throughput. Therefore:
• Target Throughput = Peak MAC Throughput
Sel Sel Sel Sel
Mi Mi Mi Mi
TPD Min – DL , TPD Min – UL , TPD Max – DL , and TPD Max – UL are used for the RRM process as defined in the
service properties.
• Target Throughput = Effective MAC Throughput

Sel Sel
Sel Mi Sel Mi
Mi TPD Min – DL Mi TPD Max – DL
Downlink: TPD Min – DL = ---------------------------------------------------
Sel
, TPD Max – DL = ---------------------------------------------------
Sel
Mi
 1 – BLER  B   1 – BLER  B Mi  
  DL    DL 
Sel Sel
Sel Mi Sel Mi
Mi TPD Min – UL Mi TPD Max – UL
Uplink: TPD Min – UL = ---------------------------------------------------
Sel
TPD Max – UL = ---------------------------------------------------
Sel
Mi
 1 – BLER  B   1 – BLER  B Mi  
  UL     UL  
• Target Throughput = Application Level Throughput
© Forsk 2009 AT271_TRG_E6 499

Sel Sel
Sel Mi Mi Sel Mi Mi
Mi TPD Min – DL + TP Offset Mi TPD Max – DL + TP Offset
Downlink: TPD Min – DL - , TPD Max – DL = --------------------------------------------------------------------------------------
= --------------------------------------------------------------------------------------
Sel Sel
-
Mi Mi
 1 – BLER  B   f  1 – BLER  B Mi    f Mi
  DL   TP – Scaling   DL   TP – Scaling
Sel Sel
Sel Mi Mi Sel Mi Mi
Mi TPD Min – UL + TP Offset Mi TPD Max – UL + TP Offset
Uplink: TPD Min – UL - , TPD Max – UL = --------------------------------------------------------------------------------------
= --------------------------------------------------------------------------------------
Sel Sel
-
Mi Mi
 1 – BLER  B   f  1 – BLER  B Mi    f Mi
  UL   TP – Scaling   UL   TP – Scaling
Resource Allocation for Minimum Throughput Demands:
Sel TX i  ic 
1. For the QoS classes UGS, ErtPS, rtPS, and nrtPS, Atoll sorts the M i  N Users in order of decreasing service
Sel
Mi
priority, p :
Sel Sel Sel

Mi Mi
Mi QoS p
Sel
1 UGS Mi
p =n
Sel
2 Mi
... n > p > 0 ...
Sel
: Mi
p =0
Sel
: ErtPS Mi
p =n
Sel
: Mi
... n > p > 0 ...
Sel
: Mi
p =0
Sel
: rtPS Mi
p =n
Sel
: Mi
... n > p > 0 ...
Sel
: Mi
p =0
Sel
: nrtPS Mi
p =n
Sel
N–1 Mi
... n > p > 0 ...
Sel
N Mi
p =0
Where N  N Users , if there are some Best Effort users, or N = N Users if there are no Best Effort users selected.
Sel Sel
2. Starting with M i = 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 Sel
Sel Mi Sel Mi
Mi TPD Min – DL Mi TPD Min – UL
R Min – DL Sel
- and R Min – UL = ------------------------------
= ------------------------------ Sel
-
Mi Mi
CTP P – DL CTP P – UL
3. Atoll stops the resource allocation in downlink or uplink,

Sel
Mi
- When/If in downlink  RMin – DL = 100 % , i.e., the resources available in downlink have been used up for
Sel
Mi
satisfying the minimum throughput demands of the mobiles.

Sel
Mi
- When/If in uplink  R Min – UL = 100 % , i.e., the resources available in uplink have been used up for satis-
Sel
Mi
fying the minimum throughput demands of the mobiles.
4. Mobiles which are active UL+DL must be able to get their minimum throughput demands in both UL and DL in
order to be considered connected UL+DL. If an active UL+DL 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.
500 AT271_TRG_E6 © Forsk 2009

Sel Sel
Mi Mi
5. If  R Min – DL  100 % or  R Min – UL  100 % , and all the minimum throughput resources demanded by the
Sel Sel
Mi Mi
mobiles have been allocated, Atoll goes to the next step for allocating resources to satisfy the maximum
throughput demands.
The remaining cell resources available for the next step are:
Sel
TX i  ic  Mi
Downlink: R Rem – DL = 100 % –  R Min – DL
Sel
Mi
Sel
TX i  ic  Mi
Uplink: R Rem – UL = 100 % –  RMin – UL
Sel
Mi
Resource Allocation for Maximum Throughput Demands:

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:
Sel Sel Sel
Mi Mi Mi
Downlink: TPD Rem – DL = TPD Max – DL – TPD Min – DL
Sel Sel Sel

Mi Mi 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:
1. 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.
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:
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 de-
mands:
Sel Sel
Sel Mi Sel Mi
Mi TPD Rem – DL Mi TPD Rem – UL
RD Rem – DL = ---------------------------------
Sel
and RD Rem – UL = ---------------------------------
Sel
Mi Mi
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 Proportional Fair scheduling method for satisfying its maximum
throughput demands are:
Sel Sel Sel

TX i  ic  Sel
TX i  ic 
Mi  Mi R Rem – DL Mi  Mi R Rem – UL
R Max – DL = 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, whichever is smaller.
d. Atoll stops the resource allocation in downlink or uplink,

Sel
Mi TX i  ic 
- When/If in downlink  RMax – DL = RRem – DL , i.e., the resources available in downlink have been used
Sel
Mi
up for satisfying the maximum throughput demands of the mobiles.
Sel
Mi TX i  ic 
- When/If in uplink  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.
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:
© Forsk 2009 AT271_TRG_E6 501

Sel Sel Sel Sel

TX i  ic  Mi Mi TX i  ic  Mi Mi
R Rem – DL = 100 % –  R Min – DL –  R Max – DL and R Rem – UL = 100 % –  R Min – UL –  R Max – UL
Sel Sel Sel Sel
Mi Mi Mi Mi
g. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satis-
fied until either R Rem – DL = 0 and R Rem – UL = 0 , or all the maximum throughput demands are satisfied.
2. Proportional Demand:
The goal of this scheduling method is to allocate resources to users weighted according to their remaining through-
put 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 propor-
tionally to their demands.
a. Atoll converts the remaining throughput demands of all the users to their respective remaining resource de-
mands:
Sel Sel
Sel Mi Sel Mi
RD Rem – DL = ---------------------------------
Sel
and RD Rem – UL = ---------------------------------
Sel
Mi Mi
b. Atoll calculates the amount effective remaining resources for the cell of each user to distribute among the us-
ers as follows:
TX i  ic 
Mi
 TX i  ic 
Mi

R Eff – Rem – DL = Min  R Rem – DL  RD Rem – DL and R Eff – Rem – UL = Min  R Rem – UL  RD Rem – UL
i i
   
 Sel   Sel 
Mi Mi
c. The resources allocated to each user by the Proportional Demand scheduling method for satisfying its maxi-
mum throughput demands are:
Sel Sel
Sel
Mi Sel
Mi
Mi TX i  ic  RD Rem – DL Mi TX i  ic  RD Rem – UL
R Max – DL = R Eff – Rem – DL  ---------------------------------------
Sel
- and R Max – UL = R Eff – Rem – UL  ---------------------------------------
Sel
-
Mi Mi
 RDRem – DL  RDRem – UL
Sel Sel
Mi Mi
3. 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:
Sel Sel Sel

QoS Mi Mi Mi
f Bias R Max – ErtPS R Max – rtPS R Max – nrtPS
 = 1 + ----------
- = ------------------------------
- = ------------------------------ = ------------------------------
100 Mi
Sel
Mi
Sel
Mi
Sel
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:
r r
1 QoS 1 QoS
N QoS   --- N QoS   ---
TX i  ic  TX i  ic    TX i  ic  TX i  ic   
R QoS – DL = R Rem – DL  ------------------------------------------------------------
r
and R QoS – UL = R Rem – UL  ------------------------------------------------------------
r
1 QoS 1 QoS
 N QoS   --- N QoS   ---
   
All QoS All QoS
502 AT271_TRG_E6 © Forsk 2009

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:
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 de-
mands:
Sel Sel
Sel Mi Sel Mi
RD Rem – DL = ---------------------------------
Sel
and RD Rem – UL = ---------------------------------
Sel
Mi Mi
c. The resources allocated to each user by the Biased scheduling method for satisfying its maximum throughput
demands are:
Sel Sel Sel

TX i  ic  Sel
TX i  ic 
Mi  Mi R QoS – DL Mi  Mi R QoS – UL
R Max – DL = Min  RD Rem – DL ------------------------ and R Max – UL = Min  RD Rem – UL ------------------------
 N QoS   N 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 TX i  ic 
- When/If in downlink  R Max – DL = R QoS – DL , i.e., the resources available in downlink for the QoS class
Sel
Mi
have been used up for satisfying the maximum throughput demands of the mobiles.
Sel
Mi TX i  ic 
- When/If in uplink  RMax – UL = RQoS – UL , i.e., the resources available in uplink for the QoS class have
Sel
Mi
been 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:
Sel Sel Sel Sel
TX i  ic  Mi Mi TX i  ic  Mi Mi
R QoS – DL = 100 % –  R Min – DL –  R Max – DL and R QoS – UL = 100 % –  R Min – UL –  R Max – UL
Sel Sel Sel Sel
Mi Mi Mi Mi
g. Atoll repeats the all the above steps for the users of the QoS class whose maximum throughput demands have
not been satisfied until either R QoS – DL = 0 and R QoS – UL = 0 , or all the maximum throughput demands are
satisfied.
4. Max Aggregate Throughput:
The goal of this scheduling method is to achieve the maximum aggregate throughput for the cells. This is done by
allocating the 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 TX i  ic 
a. Atoll sorts the M 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 Sel
Sel Mi Sel Mi
R Max – DL = ---------------------------------
Sel
and R Max – UL = ---------------------------------
Sel
Mi Mi
c. Atoll stops the resource allocation in downlink or uplink,
© Forsk 2009 AT271_TRG_E6 503

Sel
Mi TX i  ic 
- When/If in downlink  R Max – DL = R Rem – DL , i.e., the resources available in downlink have been used
Sel
Mi
up for satisfying the maximum throughput demands of the mobiles.
Sel
Mi TX i  ic 
- When/If in uplink  RMax – UL = RRem – UL , i.e., the resources available in uplink have been used up for
Sel
Mi
satisfying the maximum throughput demands of the mobiles.

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 Sel Sel
Mi Mi Mi Mi
Downlink: TL DL = R DL = R Min – DL + R Max – DL
Sel Sel Sel Sel

Mi Mi Mi Mi
Uplink: TL UL = R UL = R Min – UL + R Max – UL
Output
Sel Sel
Mi Mi Sel
• TL DL = R DL : Downlink traffic load or the amount of downlink resources allocated to the mobile M i .
Sel Sel
Mi Mi Sel
• TL UL = R UL : Uplink traffic load or the amount of uplink resources allocated to the mobile M i .
9.3.8.2 User Throughput Calculation

User throughputs are calculated for the percentage of resources allocated to each mobile selected by the scheduling for
Sel
RRM during the Monte Carlo simulations, M i .
Input
Sel
Mi Sel
• R DL : Amount of downlink resources allocated to the mobile M i as calculated in "Scheduling and Radio
Resource Allocation" on page 499.
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 499.
Sel
Mi Sel
• CTP P – DL : Downlink peak MAC channel throughput at the mobile M i as calculated in "Channel Throughput
Calculation" on page 493.
Sel
Mi Sel
• CTP P – UL : Uplink peak MAC channel throughput at the mobile M i as calculated in "Channel Throughput Cal-
Sel
Mi TX i  ic 
• BLER  B DL  : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the reception
 
Sel
equipment assigned to the terminal used by the mobile M i .
Sel
Mi Mi
• BLER  B UL  : Uplink block error rate read from the BLER vs. CINR UL graph available in the reception equip-
 
ment 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 .
Sel
Mi Sel
• TP Offset : Throughput offset defined in the properties of the service used by the mobile M i .
Calculations
Downlink:
Sel Sel Sel
Mi Mi Mi
• Peak MAC User Throughput: UTP P – DL = R DL  CTP P – DL
Sel Sel Sel
Mi Mi Mi
• Effective MAC User Throughput: UTP E – DL = UTP P – DL   1 – BLER  B DL  
  
Sel
Sel Sel
Mi Sel
• Application Level User Throughput: UTP A – DL = UTP E – DL  ----------------------------- – TP Offset
100
504 AT271_TRG_E6 © Forsk 2009

Uplink:
Sel Sel Sel
Mi Mi Mi
• Peak MAC User Throughput: UTP P – UL = R UL  CTP P – UL
Sel Sel Sel
Mi Mi Mi
• Effective MAC User Throughput: UTP E – UL = UTP P – UL   1 – BLER  B UL  
  
Sel
Sel Sel
Mi Sel
• Application Level User Throughput: UTP A – UL = UTP E – UL  ----------------------------- – TP Offset
100
Output
Sel
Mi Sel
• UTP P – DL : Downlink peak MAC user throughput at the pixel, subscriber, or mobile M i .
Sel
Mi Sel
• UTP E – DL : Downlink effective MAC user throughput at the pixel, subscriber, or mobile M i .
Sel
Mi Sel
• UTP A – DL : Downlink application level user throughput at the pixel, subscriber, or mobile M i .
Sel
Mi Sel
• UTP P – UL : Uplink peak MAC user throughput at the pixel, subscriber, or mobile M i .
Sel
Mi Sel
• UTP E – UL : Uplink effective MAC user throughput at the pixel, subscriber, or mobile M i .
Sel
Mi Sel
• UTP A – UL : Uplink application level user throughput at the pixel, subscriber, or mobile M i .
9.3.9 Optimum Beamformer Model

This section describes how the optimum beamformer smart antenna model works. The smart antenna performs simple
beam forming in the downlink to steer the main beam towards the served user, and uses the Minimum Mean Square Error
algorithm in the uplink in order to cancel interference.
9.3.9.1 Introduction
Adaptive antenna systems use more than one antenna elements, along with smart signal processing, to locate and track
various types of signals, to dynamically minimize interference, and maximize useful signal reception. The signal processor
dynamically applies weights to each element of the adaptive antenna system to create array patterns in real-time to maxi-
mize the output C/(I+N).
Figure 9.11: Linear Adaptive Array System
The antenna patterns created for downlink transmission have a main beam pointed in the direction of the useful signal. In
the uplink, in addition to the main beam pointed in the direction of the useful signal, there can also be one or more nulls in
the direction of the interfering signals. If the adaptive antenna system is using E SA antenna elements, it is possible to
create E SA – 1 nulls to cancel E SA – 1 interfering signals. In a mobile environment where the interference is not station-
ary, the antenna patterns are adjusted so that the nulls remain in the direction of the moving interference. A system using
adaptive antennas adjusts the weights on each antenna element to achieve such a pattern.
The optimum beamformer supports linear adaptive array systems, such as the one shown in Figure 9.11: on page 505.
© Forsk 2009 AT271_TRG_E6 505

In the following explanations, we assume:
• There are a total of E SA elements in the adaptive antenna system.

•  is the angle of arrival for the useful signal.
•  is the angle at which we want to calculate the smart antenna gain.
• d is the distance between two adjacent antenna elements.
9.3.9.2 Downlink
Figure 9.12: Downlink Beamforming
The downlink algorithm models a conventional beam former. The smart antenna processor applies complex weights, w n ,
to each antenna element in order to form a beam towards the served user. The magnitude of these complex weights is set
to 1. The beamforming is performed using only the phase of the complex weights. The steering vector, S  , representing
the complex weights for forming a beam towards the served user, i.e., at the angle of arrival  is given by:
2 2 2 T
  
S  = 1 e e  ... e
Where the notation T represents the transpose of a matrix.
Therefore, the complex weight at any nth antenna element can be given by:
2
– j  -------  nd  sin 

wn = e
 – j    n  sin 
In Atoll, d = --- , therefore, w n = e .
2
The smart antenna gain in any direction  can be given by:
H
G SA    = g n     S   R   S 
Where the notation H represents the Hilbert transform, which is the complex conjugate transpose of a matrix, g n    is the
gain of the nth antenna element in the direction  , and R  is the array correlation matrix for a given user direction  , given
by:
H
R = S  S
For the direction of the served user, i.e.,  , 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.
The smart antenna gain in dB will be G SA    = 10  Log  G SA     .
The smart antenna is able to form the beam only in the horizontal plane, therefore, the vertical pattern is assumed to
remain the same.
506 AT271_TRG_E6 © Forsk 2009

Power Combination Gain
Cell transmission power is fed to each antenna element of the smart antenna system. Since each element transmits the
same input power, this results in a gain due to power combination, i.e., the powers fed to each antenna element are
combined for transmission.
During Monte Carlo simulations, Atoll generates a time-slot scenario for each victim mobile and calculates the downlink
C/(I+N) as described in the section "Traffic and Pilot Interference Signal Levels Calculation (DL)" on page 475. However,
as it calculates the smart antenna gains for each victim mobile in a cell’s coverage area, it averages the array correlation
matrix R  over all the iterations in order to generate an angular distribution of the downlink traffic power density, which is
a combination of signal power and angles.
The average array correlation matrix is given by:
J
j=1
Where R Avg is the average downlink array correlation matrix, J is the number of served mobiles during the simulation,  j
is the probability of presence of the mobile j, p j is the EIRP transmitted towards the mobile j, and R j is the array correlation
matrix for the mobile j.
The probability of presence of the mobile j is the ratio between the downlink resources provided to the mobile j and the
total amount of available downlink resources. For example, if a mobile has been granted 10 % of the number of available
slots in the downlink subframe, it’s probability of presence is 10 %.
9.3.9.3 Uplink
The uplink models the adaptive Minimum Mean Square Error algorithm which optimizes the useful signal as well as
cancels the interference from the most interfering E SA – 1 interfering mobiles. The optimal beam forming method used in
Atoll overcomes the limitations of a null steering beam former and maximizes the output C/(I+N).
A simple null steering beam former requires the knowledge of the directions of interference sources, and the estimated
weights do not maximize the output C/(I+N). Whereas, an optimal beam former does not require knowledge of directions
and power levels of interference to maximize the output C/(I+N). It only requires the direction of the useful signal in order
to calculate the optimum C/(I+N).
Figure 9.13: Uplink Adaptive Algorithm
Let ŵ represent the vector of ESA complex weights for the beam former. ŵ is given by:
–1
ŵ =    R N  S 
Where S  is the steering vector in the direction of the served user,  .   , which is a constant value for a given useful
signal, represents the smart antenna gain in the uplink given by the equation:
E SA
  = ----------------------------------
-
H –1
S  RN  S
© Forsk 2009 AT271_TRG_E6 507

–1
R N is the inverse of the total noise correlation matrix. The total noise correlation matrix is the sum of the thermal noise
correlation matrix R n , and the interference correlation matrix R I , given by:
J
H
2
RN = Rn + RI = n  I +
j=1
J
H
2
Where R n = n  I and R I =
j=1
2
 n is the thermal noise power. I is the identity matrix. p j is the power received by one element of the smart antenna from
the jth interfering mobile. S j is the steering vector in the direction of the jth interfering mobile,  . J is the total number of
interfering mobiles.
The total noise power, including thermal noise and interference from all uplink interferers, received by a cell is given by:
ˆ =  2  S H  R–1  S
P N   N 
And, the total power received from the served user is given by:
ˆ = p  2   S H  R –1  S 2
P     N 
Where p  is the power received by one element of the smart antenna from the served user.
The C/(I+N) in the uplink is then calculated by:
2 H –1 2
Pˆ p     S  RN  S  H –1
CINR UL = ------- = --------------------------------------------------------------- = p  S  RN  S
Pˆ 2 H –1
N   S  RN  S
From the above equation, we can determine the uplink smart antenna beam forming gain in the direction of the served
–1
user. C UL can be calculated from the above equation by considering the interference and noise to be null, i.e., R N = I .
This gives:
H
C UL = p   S   I  S  = p   E SA
From the above equation, the uplink smart antenna beam forming gain equals the number of smart antenna elements, i.e.,
G SA = E SA .
During Monte Carlo simulations, Atoll generates a time-slot scenario for each victim mobile and calculates the uplink C/
(I+N) as described above.
–1
The inverse noise correlation matrix R N for each iteration k includes the effect of the matrix calculated for the previous
iteration. The interference power and its direction is stored at the end of each simulation. Hence, Atoll is able to calculate
an average of the smart antenna interference-cancellation effect. The result is the angular distribution of the uplink noise
rise, which is calculated from the inverse of the noise correlation matrix obtained at the end of the last iteration of a Monte
Carlo simulation. This angular distribution of the uplink noise rise can be stored in the Cells table. The average of the
inverse noise correlation matrices is calculated as follows:
K
1
 RN
–1 –1
RN = ---- 
Avg K k
k=1
–1 –1
Where R N is the average of the inverse noise correlation matrices of all the iterations from k = 1 to K, and R N is
Avg k
the inverse noise correlation matrix of the kth iteration.

The interference can be isolated from the thermal noise and can be calculated for any direction using the formula.
E SA 2
I UL    = ------------------------------------------
- – n
H –1
S  RN  S
Avg
Where I UL    is the interfering signal in the direction  , ESA is the number of smart antenna elements, S  is the steering
2
vector in the direction  , and  n is the thermal noise power, with I being an identity matrix.
The angular distribution of the uplink noise rise is given by:
508 AT271_TRG_E6 © Forsk 2009

2
I UL    +  n
NR UL    = ----------------------------
-
2
n
9.4 Automatic Allocation Algorithms

The following sections describe the algorithms for automatic neighbour and preamble index planning.
9.4.1 Automatic Neighbour Allocation

The intra-technology neighbour allocation 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:
Only TBA cells are assigned neighbours.
Note:
We assume a reference cell TXi(ic) and a candidate neighbour cell TXj(jc). When automatic allocation starts, Atoll checks
the following conditions:
is discarded.
- Force Co-site Cells as Neighbours: If selected, Atoll adds all the cells located on the same site as the ref-
erence 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 ref-
erence 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.
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 ranking.
Figure 9.14: Determination of Adjacent Cells
- Force Neighbour 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).
- 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.
© Forsk 2009 AT271_TRG_E6 509

If you select "Force exceptional pairs" and "Force symmetry", Atoll considers the constraints between excep-
tional pairs in both directions so as to respect symmetry condition. On the other hand, if neighbourhood rela-
tionship is forced in one direction and forbidden in the other, symmetry cannot be respected. In this case, Atoll
- Delete Existing Neighbours: If selected, Atoll deletes all the current neighbours and carries out a new neigh-
bour allocation. If not selected, the existing neighbours are kept in the list.
3. The coverage areas of TXi(ic) and TXj(jc) must have an overlap ( S TX  ic   S TX  jc  ).

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
TX i  ic 
received preamble signal level ( C Preamble ) and the preamble signal level threshold are calculated from
CNR Preamble and T Preamble , respectively, by adding the value of the noise to them.
- S TX  ic  is the surface area covered by TXi(ic) within C Preamble + HO Start and C Preamble + HO End .
i
HO Start is the margin with respect to the best preamble signal level at which the handover starts, and
HO End is the margin with respect to the best preamble signal level at which the handover ends.
- S TX  jc  is the coverage area where the candidate cell TXj(jc) is the best server.
j
Note:
• 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 coverage areas for the cells.
dialogue of the Predictions folder.
When the above conditions are met, Atoll calculates the percentage of the coverage area overlap
S TX  ic   S TX  jc 
i j
-  100 ), and compares this value with the % Min Covered Area. TXj(jc) is considered a neighbour of
( --------------------------------------------
S TX  ic 
i
S TX  ic   S TX  jc 
i j
-  100  % Min Coverage Area .
TXi(ic) if --------------------------------------------
S TX  ic 
i
Figure 9.15: Overlapping Zones
Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours in decreasing
order of 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 reason of allocation:
Importance
Neighbour Cause When
Value
Only if the Delete Existing Neighbours option is not selected Existing
Existing neighbour
Exceptional pair Only if the Force Exceptional Pairs option is selected 100 %
510 AT271_TRG_E6 © Forsk 2009

Only if the Force Co-site Cells as Neighbours option is Importance

Co-site cell
selected Function (IF)
Only if the Force Adjacent Cells as Neighbours option is Importance
Adjacent cell
selected Function (IF)
Neighbourhood relationship that fulfils Importance
Only if the % Min Covered Area is exceeded
coverage conditions Function (IF)
Symmetric neighbourhood Importance
Only if the Force Neighbour Symmetry option is selected
relationship Function (IF)
The importance is evaluated using an Importance Function (IF), which takes into account the following factors:
• Co-site factor (C): a Boolean,
• Adjacency factor (A): the percentage of adjacency,
• Overlapping factor (O): the percentage of overlapping
The minimum and maximum importance assigned to each of the above factors can be defined.
Min Max
Factor Default Value Default Value
Importance Importance
Overlapping factor (O) Min(O) 1% 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:
Neighbour Cause
Importance Function IF with With Default Values
Co-site Adjacent
no no Min(O) + Delta(O)(O) 1 % + 29 %(O)
no yes Min(A)+Delta(A){Max(O)(O)+(100 %-Max(O))(A)} 30 % + 30 %{30 %(O) + 70 %(A)}
yes yes Min(C)+Delta(C){Max(O)(O)+(100 %-Max(O))(A)} 60 % + 40 %{30 %(O )+ 70 %(A)}
Where Delta(x) = Max(x) - Min(x)

Notes:
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 ranges of the IF factors overlap each other, the neighbours may not be ranked by
neighbour cause.
• The ranking between neighbours from the same category depends on (A) and (O) factors.
• 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 greater than 0 % only if there is some overlapping.
In the results, Atoll lists only the cells for which it finds new neighbours. Therefore, if a transmitter has already reached its
maximum number of neighbours before starting the allocation, it will not appear in the results table.
9.4.2 Automatic Preamble Index Allocation

IEEE 802.16e defines 114 preamble indexes. Each preamble index, from 0 to 113, contains the following information:
• Segment number (0, 1, or 2),
• DL PermBase (0 to 31) for the obligatory first DL PUSC zone, and
• A pseudo-noise sequence transmitted using the subcarriers corresponding to the preamble carrier set.
The downlink subframe can be divided into a 3-segment structure, and includes a preamble which begins the transmission
(the first symbol of the downlink transmission). The preamble subcarriers are divided into 3 carrier sets. There are three
possible groups consisting of a carrier set each which may be used by any segment. These are defined by allocation of
different subcarriers to each one of them. The subcarriers are modulated using a BPSK modulation with a specific Pseudo-
Noise (PN) sequence.
Preamble carrier sets are defined using equation below:
PreambleCarrierSet n = n + k  3
Where PreambleCarrierSetn gives the subcarriers used by the preamble, n is the number of the preamble carrier set
indexed 0, 1, or 2, k is a running index from 0 to 567 for FFT 2048, from 0 to 283 for FFT 1024, from 0 to 142 for FFT 512,
and from 0 to 35 for FFT 128.
In a WiMAX 802.16e network, each base station transmits a different PN sequence, out of the 114 available, on the pream-
ble carrier set. A mobile trying to connect to the network scans all the preamble subcarriers, listens to all the preambles
© Forsk 2009 AT271_TRG_E6 511

(i.e., PN sequences) from all the base stations it can receive, 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.
It selects the base station as its server whose preamble it receives with either the highest signal level or the highest C/
(I+N). Once the best server is known, its PN sequence is used to identify its transmission. The PN sequence of the best
server gives the preamble index, which in turn gives the segment number, and the IDCell (DL PermBase of the first DL
PUSC zone, referred to as Cell PermBase in Atoll). Therefore, the mobile knows which subcarriers to listen to for the FCH,
DCD, UCD, DL-MAP, and UL-MAP.
As can be understood from the above description, if all the cells in the network transmit the same preamble index, the
network will have 100% interference on downlink preambles, and it will be impossible for a mobile to identify different cells.
Cell search and selection will be impossible. Therefore, it is important to intelligently allocate preamble indexes to cells so
as to reduce preamble interference, and allow easy recognition of cells by mobiles.
The following describes the automatic allocation method for preamble indexes in a WiMAX 802.16e network, which takes
into account the distance between transmitters, the frequency plan of the network (i.e., co- and adjacent channel interfer-
ence probabilities), and the neighbour relations.
The preamble index allocation algorithm takes into account the cells of all the TBC transmitters. The cells to be allocated
will be called TBA cells. They must fulfil the following conditions:
• Their status is not set to locked,
Note:
9.4.2.1 Constraint and Relationship Weights

The automatic preamble index allocation algorithm is based on a cost-based function which takes into account the follow-
ing constraints, in the order of priority:
1. Same preamble index,
Assigned weight  PI = 0.6
2. Same segment number,

Assigned weight  Seg = 0.38
3. Same cell permbase,

Assigned weight  PB = 0.02
Note that the sum of the weights assigned to the above constraints is 1.
The above constraints are studied between each TBA cell and its related cells. Atoll calculates the cost between each indi-
vidual TBA and related cell, and then the global cost for the TBA cell.
Related cells of a TBA cell are:
• Its neighbours, if the check box "Take Neighbours into Account" is selected,
Assigned weight  Neighbour = 0.35
• Cells within the cell’s (or the default) minimum reuse distance,
Assigned weight  Dis tan ce = 0.35
9.4.2.2 Calculation of Cost Between TBA and Related Cells

Atoll calculates the constraint violation level between the TBA cell TXi(ic) and its related cell TXj(jc) as follows:
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  PI Seg PB
VL = rO    PI  p Coll +  Seg  p Coll +  PB  p Penalty 
Where r O is the total channel overlap ratio between the TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent
Channel Overlaps Calculation" on page 462,  PI ,  Seg , and  PB are the weights assigned to the preamble index,
segment number, and cell permbase constraints, and,
 TX i  ic  TX j  jc 

the preamble index collision probability is given by p Coll =  1 if PI = PI
PI
• ,
 TX i  ic  TX j  jc 
 0 if PI  PI
 TX i  ic  TX j  jc 
Seg  1 if N Seg = N Seg
• the segment number collision probability is given by p Coll =  , and
 TX i  ic  TX j  jc 
 0 if N Seg  N Seg
512 AT271_TRG_E6 © Forsk 2009

• the cell permbase penalty is given by


 1 if PB TXi  ic   PB TXj  jc  AND Site
TX i  ic 
= Site
TX j  jc 
PB 
p Penalty =  TX i  ic  TX j  jc  TX i  ic  TX j  jc  if the cell permbase allocation
 0.001 if PB  PB AND Site  Site
 0 Otherwise

PB
strategy is set to "Same Cell PermBase per Site", and by p Penalty = 0 if the cell permbase allocation strategy is
set to "Free".
The cell permbase penalty models the cell permbase constraint.
Next, Atoll calculates the importance of the relation between the TBA cell and its related cell.
 Total =  Neighbour   Neighbour +  Dis tan ce   Dis tan ce
Where  Neighbour is the importance of the relationship between the TBA and its related neighbour cell, and
 Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance
between them.
 Neighbour is calculated during automatic neighbour allocation by Atoll as explained in "Automatic Neighbour Alloca-
tion" on page 509. For manual neighbour allocation, this value is equal to 1.
 Dis tan ce is calculated by the preamble allocation algorithm as follows:
 TX i  ic  – TX j  jc 
 1 if D 1

TX i  ic  – TX j  jc    D Reuse  2
 Dis tan ce =  Log   ------------------------------------- - 
   D TXi  ic  – TXj  jc  
 --------------------------------------------------------------- Otherwise
 Log  D Reuse 
2

Where D Reuse is the minimum reuse distance, either defined for the TBA cell individually or set for all the cells in the auto-
matic allocation dialogue, and D is the weighted distance between the TBA cell TXi(ic) and its related cell
TXj(jc) calculated as follows:
D = d   1 + x   cos    – cos    – 2  
D is weighted according to the orientations of the TBA and its related cell with respect to the straight line
joining them. d is the distance between the two cells considering any offsets with respect to the site locations.
x is set to 15 % so that the maximum variation in D due to the azimuths does not exceed 60 %.  and 
are, of course, calculated from the azimuths of the two cells as shown in Figure 9.16: on page 513.
Figure 9.16: 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, and two cells pointing in opposite directions will have a greater effective distance.
The importance of the distance relation is explained in Figure 9.17: on page 514. 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,
which is interpreted as a high cost, 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.
© Forsk 2009 AT271_TRG_E6 513

Figure 9.17: Importance Based on Distance Relation
From the constraint violation level and the total importance of the relation between the TBA and its related cell, Atoll calcu-
lates the quality reduction factor for the pair as follows:
 TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
TX i  ic  – TX j  jc   1 if r O =0 OR VL =0
QRF = 
 TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
 1 – VL   Total Otherwise
The quality reduction factor is a measure of the cost of an individual relation.

The total cost of the current preamble index allocation for any TBA cell is given as follows, considering all the cells with
which the TBA cell has relations:
TX i  ic  TX i  ic  – TX j  jc 
$ Total = 1–  QRF
TX j  jc 
And, the total cost of the current preamble index allocation for the entire network is simply the sum of the total TBA cell
costs calculated above, i.e.,
TX i  ic 
$ Total =  $ Total
TX i  ic 
9.4.2.3 Automatic Allocation Algorithm

The automatic preamble index allocation algorithm is an iterative algorithm which:
• Calculates the cost (as described above) of the current preamble index allocation,
• Allocates new preamble indexes to cells in order to reduce the costs, and calculates the cost again,
• Memorises the different allocation plans in order to determine the best allocation, i.e., which provides the lowest
total cost.
9.5 Glossary of WiMAX Terms

Figure 9.18: on page 515 shows a TDD WiMAX frame along with its different components and explains the significance
of different terms in the Atoll WiMAX module.
Atoll allows you to set network level parameters which are common to all the transmitters and cells in the entire network.
These parameters are used in coverage predictions as well as during Monte Carlo simulations by the radio resource
management and scheduling algorithms. The global WiMAX parameters include.
• Symbol: A symbol is the smallest resource unit that can be allocated to a user in WiMAX networks.
• Symbol Duration (SD): The symbol duration is the length of each symbol in the frame. The length of a frame, i.e.,
the frame duration, can be expressed in terms of the number of symbol durations in the frame. It is referred to as
OFDM symbol in the IEEE 802.16 specifications.
• Frame duration: It is the length of a WiMAX frame in milliseconds. You can choose from a list of frame durations
defined in the IEEE 802.16 specifications.
• Cyclic prefix ratio: The total symbol duration in WiMAX comprises the useful part of the symbol, carrying the data
bits, and a CRC part, which is a portion of the useful data part repeated at the beginning of each symbol. The cyclic
prefix is the method used by WiMAX to counter inter-symbol interference (ISI). The cyclic prefix and the orthogo-
nality of subcarriers ensure that there is negligible intra-cell interference in WiMAX.
514 AT271_TRG_E6 © Forsk 2009

Figure 9.18: WiMAX Frame
• Fixed and variable overheads: The fixed and variable overheads in the uplink and downlink subframes are used
to model the preamble and other time domain overheads such as broadcast messages including DL-MAP, UL-
MAP, UCD, and DCD, and the FCH, in downlink, and Ranging and Bandwidth Request messages in the uplink.
The preamble is always one symbol duration long and can be modelled using the fixed overhead, while other mes-
sages whose lengths vary according to either the frame duration or the channel width can be modelled using the
variable overhead. This is the reason why the fixed overheads are available in terms of symbol durations (SD) and
the variable overheads in terms of percentages of the uplink and downlink subframes. Variable overheads are per-
centages of the DL and the UL subframe sizes, after the fixed overheads have been removed from the subframe
size.
• DL subframe ratio (TDD only): It is the percentage of the entire frame duration which corresponds to the downlink
subframe. In FDD networks, the downlink and uplink subframes cover entire frame durations in different frequency
bands.
• Transmit and receive time guards (TDD only): Transmit and receive time guards are also time domain over-
heads, i.e., these are portions of the frame which cannot be used for data transfer. You can enter TTG and RTG
times in milliseconds. These time guards allow the receiver and the transmitter to switch between transmit and
receive modes.
• Subchannels: A subchannel is a group of subcarriers. A channel can be divided into a number of subchannels.
You can set the number of these subchannels at the network level in Atoll.
• Subcarriers (or tone): The entire channel contains a number of subcarriers which compose the upper and lower
guard bands, the pilot subcarriers, and the data subcarriers. The guards, pilots, and the DC subcarrier can not be
used for data transfer. The total thermal noise over the entire channel bandwidth is calculated according to the
number of used subcarriers out of the total number of subcarriers. The used subcarriers are the data and the pilot
subcarriers. The data transfer capacity of a channel is calculated by considering the data subcarriers only.
• User: A general term used interchangeably with subscribers, mobiles, and receivers.
• Subscriber: Term used for users with fixed geographical coordinates.
• Mobile: Term used to indicate the users generated and distributed during simulations. These users have, among
other parameters, definite services, terminal types, and mobility types assigned for the duration of the simulations.
• Receiver: A probe mobile, with the minimum required parameters needed for the computation of path loss, used
for propagation loss and raster coverage predictions.
• Bearer: Defines a Modulation and Coding Scheme (MCS) used to carry data over the channel.
• Peak MAC Throughput: The maximum MAC layer throughput (user or channel) that can be achieved at a given
location using the best possible WiMAX bearer. This throughput is the raw data rate without considering the effects
of retransmission due to errors and higher layer coding and encryption.
• Effective MAC Throughput: The net MAC layer throughput (user or channel) that can be achieved at a given
location using the best possible WiMAX bearer calculated taking into account the reduction of throughput due to
retransmissions due to errors.
• Application Throughput: The application layer throughput (user or channel) that can be achieved at a given loca-
tion using the best possible WiMAX bearer calculated taking into account the reduction of throughput due to PDU/
SDU header information, padding, encryption, coding, and other types of overhead.
• Channel Throughputs: Peak MAC, effective MAC or application level throughputs achieved at a given location
using the best possible WiMAX bearer with the entire channel resources.
• User Throughputs: Peak MAC, effective MAC or application level throughputs achieved at a given location using
the best possible WiMAX bearer with the amount of resources allocated to a user by the scheduler.
• Uplink Noise Rise: The uplink noise rise is the engineering metric that provides an idea of the uplink interference.
Uplink noise rise is defined as the ratio of the total uplink interference power to the background noise power.
• Smart Antenna: Smart antenna systems refers to a system of antenna arrays with smart signal processing algo-
rithms that are used to identify the direction of arrival (DOA) of the signal, and use it to calculate beamforming
vectors, to track and locate the antenna beam on the mobile.
• Adaptive Antenna Systems (AAS): AAS is a type of smart antenna systems. Adaptive antenna systems use
more than one antenna elements along with smart signal processing in order to locate and track various types of
© Forsk 2009 AT271_TRG_E6 515

signals to dynamically minimize interference and maximize useful signal reception. The signal processor applies
dynamic weighting to each element of the adaptive antenna system hence creating dynamic array patterns in real-
time to maximise the output C/(I+N).
• Multiple Input Multiple Output (MIMO): Multiple Input Multiple Output systems use more than one transmission
and reception antennas for increasing system capacity by either using space-time transmit diversity or spatial mul-
tiplexing.
• Spatial Multiplexing (SM): Uses more than one transmission antenna to send different signals (data streams) on
each antenna. The receiver can also have more than one antenna for receiving different signals. Using spatial
multiplexing with M transmission and N reception antennas, the throughput over the transmitter-receiver link can
be theoretically increased M or N times, depending on which is smaller, M or N. SM improves the throughput
(channel capacity) for a given C/(I+N), and is used for the regions of a cell that have sufficient C/(I+N) conditions.
SM is often referred to as MIMO.
• Space-Time Transmit Diversity (STTD): Uses more than one transmission antenna to send the same signal on
all antennas. The signals are constructively combined (using optimum selection or maximum ratio combining,
MRC) at the receiver to extract the useful signal. As the receiver gets more than one copy of the useful signal, the
signal level at the receiver after combination of all the copies is much more resistant to interference than a single
signal would be. Therefore, STTD/MRC improves the C/(I+N) at the receiver. It is often used for the regions of a
cell that have bad C/(I+N) conditions. STTD is also known as STC (Space Time Coding) and STBC (Space-Time
Block Codes).
• Adaptive MIMO Switch (AMS): AMS is a technique for switching from spatial multiplexing to space-time transmit
diversity as the C/(I+N) conditions get worse than a given threshold. AMS can be used in cells to provide spatial
multiplexing gains to users that have better C/(I+N) conditions than a given AMS threshold, and space-time
transmit diversity to users that have worse C/(I+N) conditions than the threshold. AMS provides the optimum solu-
tion using both MIMO features to their best.
516 AT271_TRG_E6 © Forsk 2009

Chapter 10
Repeaters and Remote Antennas
This chapter describes how repeaters and remote antennas are modelled in Atoll. It also provides information
on the differences in modelling for different project types.
Atoll
518 AT271_TRG_E6 © Forsk 2009

Chapter 10: Repeaters and Remote Antennas
10 Repeaters and Remote Antennas

A repeater receives, amplifies, and re-transmits the radiated or conducted RF carrier both in downlink and uplink. It has a
donor side and a server side. The donor side receives the signal from a donor transmitter or repeater and the server side
re-transmits the received signal.
The link between the donor transmitter/repeater and the repeater may be chosen between three types:
• Over the Air
• Fibre link
• Microwave link
For the air link, a user-defined or computed propagation loss is involved whereas user-defined link losses must be defined
for the other two cases.
Repeater modelling focuses on the additional coverage they provide to transmitters or repeaters. A repeater is modelled
in the same manner for GSM (GSM GPRS EGPRS) and CDMA (UMTS HSPA, TD-SCDMA, IS-95 cdmaOne, and
CDMA2000) networks, except that in the GSM networks Atoll deals with EIRP while in the CDMA it is concerned with a
total gain. We assume that all the TRXs of GSM donor transmitters and all the carriers of CDMA donor transmitters are
amplified.
Remote antennas enable you to place antennas at locations that would normally require very long feeder cables. A remote
antenna is connected to the base station via optic fibre. The main difference from a repeater is that a remote antenna
generates its own cell whereas a repeater extends the coverage of an existing cell.
10.1 Modelling Repeaters

We assume in this part that the repeater receives a signal from a donor transmitter.
10.1.1 CDMA Documents

10.1.1.1 Over the Air
10.1.1.1.1 Signal Level Received From Repeaters
Atoll calculates the signal level received from a repeater Rpk on a carrier ic as follows:
Rpk Txd Rpk Rpk
P rec  ic  = P pilot  ic  + G total – Air – DL – L path – M Shadowing – L indoor (in dB)
Figure 10.1: CDMA Documents - Over the Air Repeater
Where,
Txd
P pilot  ic  is the pilot power of the donor transmitter on the carrier ic,
Rpk
G total – Air – DL is the output downlink total gain of repeater linked to a donor transmitter with an air link. This gain will be
evaluated considering the path loss between the donor transmitter and the donor side of the repeater (see after),
Rpk
L path is the path loss between the repeater Rpk and the receiver,
M Shadowing is the shadowing margin,
“Indoor coverage” is selected in point analysis or coverage predictions.
The total gain can be either user-specified or directly calculated by Atoll from the link budget.
On each pixel, the resulting received signal on a carrier ic is calculated as follows:
Txd – Rpk Txd Rpk
P rec  ic  = P rec  ic  + P rec  ic  (not in dB6)
© Forsk 2009 AT271_TRG_E6 519

10.1.1.1.2 Gain Automatic Calculation

Two types of repeater total gains can be calculated: Total Uplink and Downlink gains. The gains are evaluated from the
basis of the donor BTS to the output of the repeater coverage antenna.
• Downlink total gain
The Downlink total gain is calculated as follows:
Txd Txd Txd – Rpk Rpk – donor – side Rpk – donor – side
Rpk G ant – L total – DL – L model + G ant – L feeder – Rec +
G total – Air – DL = (in dB)
Rpk Rpk – coverage – side Rpk – coverage – side
G amp + G ant – L feeder – Tr
Figure 10.2: Over the Air Repeater - Downlink Total Gain
Where,
Txd
G ant is the gain of the donor transmitter antenna,
Txd
L total – DL corresponds to the total downlink losses of the donor transmitter (user-defined or calculated considering
transmitter equipment characteristics - see Transmitter radio equipment part),
Txd – Rpk
L model corresponds to path losses between the donor transmitter and the repeater. They are either user-
defined or calculated using the selected propagation model. If you do not select a propagation model, the propa-
gation losses between the donor transmitter and the repeater are calculated using the ITU 526-5 propagation
model,
Rpk – donor – side

G ant is the gain of the repeater’s donor side antenna,

L feeder – Rec refers to the losses of the repeater donor side due to feeders (see Transmitter radio equipment
part). They involve the cable length at reception on the donor side and the loss per metre,
Rpk
G amp is the amplification gain of the repeater,
Rpk – coverage – side

G ant is the gain of the repeater coverage side antenna,

L feeder – Tr corresponds to the losses of the repeater coverage side due to feeders (see Transmitter radio
equipment part). They involve the cable length at transmission on the coverage side and the loss per metre.
• Uplink total gain

The Uplink total gain is calculated as follows:
Txd Txd Txd – Rpk Rpk – donor – side Rpk – donor – side
Rpk G ant – L total – UL – L model + G ant – L feeder – Tr +
G total – Air – UL = (in dB)
G amp + G ant – L feeder – Rec
6. Formula cannot be directly calculated from components stated in dB and, therefore, must be converted in linear
values.
520 AT271_TRG_E6 © Forsk 2009

Figure 10.3: Over the Air Repeater - Uplink Total Gain
Where,
Txd
Txd
L total – UL corresponds to the total uplink losses of the donor transmitter (user-defined or calculated considering
transmitter equipment characteristics - see Transmitter radio equipment part),
Txd – Rpk
L model corresponds to path losses between the donor transmitter and the repeater. They are either user-
defined or calculated using the selected propagation model. If you do not select a propagation model, the propa-
gation losses between the donor transmitter and the repeater are calculated using the ITU 526-5 propagation
model,

G ant is the gain of the repeater’s donor side antenna,

L feeder – Tr refers to the losses of the repeater donor side due to feeders (see Transmitter radio equipment
part). They involve the cable length at transmission on the donor side and the loss per metre,
Rpk


L feeder – Rec corresponds to the losses of the repeater coverage side due to feeders (see Transmitter radio
equipment part). They involve the cable length at reception on the coverage side and the loss per metre.
10.1.1.1.3 Donor Side Parameter Automatic Calculation

Azimuth
This is the angle at which the donor antenna is situated with respect to the North at the repeater side. This angle is meas-
ured clock-wise as shown in the figure below.
This parameter can be recalculated and applied using the Calculate button on the Donor Side tab. This 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.
Figure 10.4: Angle from North (Azimuth)
Mechanical Downtilt
This is the tilt angle for the repeater’s donor-side antenna, which ensures that it points towards the donor antenna in the
vertical plane. As a general rule, downtilt angles are considered positive and uptilt angles negative.
© Forsk 2009 AT271_TRG_E6 521

Figure 10.5: Positive/Negative Mechanical Downtilt
Since this parameter depends on the difference of heights/altitudes between the donor transmitter and the repeater, it can
be automatically calculated in the repeater’s Donor side properties. If the height/altitude of the antenna is modified, the
corresponding tilt angle can be found out and applied using the Calculate button. The following example may further clarify
the viewpoint:
Figure 10.6: Tilt Angle Computation
The tilt angle in the above figure would be:

Rpk – donor – side Txd
Rpk – donor – side  H ant – H ant 
T ant = atan  --------------------------------------------------------------
Txd – Rpk
-
 D 
As obvious, this angle will be negative for uptilts and positive for downtilts of the antenna. Here,
Txd – Rpk
D is the distance between the donor transmitter antenna and the repeater antenna,
Txd
H ant is the height/altitude of the donor transmitter antenna,

H ant is the height/altitude of the repeater’s donor-side antenna, and

T ant is the calculated tilt angle for the repeater’s donor-side antenna.
In the case of the above figure, the donor-side antenna at the repeater should have a tilt angle of X to point directly at the
transmitter antenna.
10.1.1.2 Microwave Link

Rpk Txd Rpk Rpk
P rec  ic  = P pilot  ic  + G total – MW – DL – L path – M Shadowing – L indoor (in dB)
Figure 10.7: CDMA Documents - Microwave Link Repeater
Where,
Txd
Rpk
G total – MW – DL is the output downlink total gain of repeater linked to a donor transmitter with an microwave link. This gain
will be evaluated considering the link loss between the donor transmitter and the donor side of the repeater (see after),
522 AT271_TRG_E6 © Forsk 2009

Rpk
Txd – Rpk Txd Rpk

In that case, the total UL and DL gains are evaluated without considering the donor antenna parameters (gain and total
TX losses) and the repeater donor-side antenna parameters (gain and feeder losses).
Rpk Txd – Rpk Rpk Rpk – coverage – side Rpk – coverage – side
G total – MW – DL = L MW + G amp + G ant – L feeder – Tr (in dB)
Figure 10.8: Microwave Link Repeater - Downlink Total Gain
Where,
Txd – Rpk
L MW is the user-defined microwave link loss between the donor transmitter and the repeater,
Rpk



G total – MW – UL = L MW + G amp + G ant – L feeder – Rec (in dB)
Figure 10.9: Microwave Link Repeater - Uplink Total Gain
Where,
Txd – Rpk
© Forsk 2009 AT271_TRG_E6 523

Rpk


10.1.1.3 Fibre Link

Rpk Txd Rpk Rpk
P rec  ic  = P pilot  ic  + G total – Fibre – DL – L path – M Shadowing – L indoor (in dB)
Figure 10.10: CDMA Documents - Fibre Link Repeater
Where,
Txd
Rpk
G total – Fibre – DL is the output downlink total gain of repeater linked to a donor transmitter with an optical fibre link. This
gain will be evaluated considering the fibre loss between the donor transmitter and the donor side of the repeater (see
after),
Rpk
L path is the link loss between the repeater Rpk and the receiver,
Txd – Rpk Txd Rpk

In that case, the total UL and DL gains are evaluated without considering the donor antenna parameters (gain and total
TX losses) and the repeater donor-side antenna parameters (gain and feeder losses).
G total – Fibre – DL = L Fibre + G amp + G ant – L feeder – Tr (in dB)
524 AT271_TRG_E6 © Forsk 2009

Figure 10.11: Fibre Link Repeater - Downlink Total Gain
Where,
Txd – Rpk
L Fibre is the user-defined fibre link loss between the donor transmitter and the repeater,
Rpk



G total – Fibre – UL = L Fibre + G amp + G ant – L feeder – Rec (in dB)
Figure 10.12: Fibre Link Repeater - Uplink Total Gain
Where,
Txd – Rpk
Rpk


10.1.1.4 Appendices
10.1.1.4.1 Automatic Controls
Atoll performs systematic controls when creating a new repeater:
• On the amplification gain,
• On the downlink power.
Amplification Gain
The amplification gain of the repeater must be consistent with the limits of its equipment. Atoll checks that:
min Rpk max

G amp  G amp  G amp
Where
© Forsk 2009 AT271_TRG_E6 525

Rpk
G amp is the amplification gain of the repeater.
min
G amp is the minimum amplification gain allowed by the equipment.
max
G amp is the maximum amplification gain allowed by the equipment.
Downlink Power
The downlink power after amplification must be consistent with the limit of the equipment installed on the repeater. Atoll
checks that:
Txd Rpk max Rpk – coverage – side Rpk – coverage – side
P pilot  ic  + G total – DL  P + G ant – L feeder – Tr
Where
Txd
P pilot  ic  is the pilot power of the donor transmitter on the carrier ic. When there are several carriers on a transmitter, Atoll
takes the highest pilot power on all the carriers.
G ant is the gain of the repeater coverage side antenna
max
P is the maximum downlink power allowed by the equipment.
L feeder – Tr corresponds to the transmission losses due to feeders installed on the repeater coverage side (see
Transmitter radio equipment part). They involve the cable length at transmission on the coverage side and the loss per
metre.
10.1.1.4.2 Carrier Power and Interference Calculation

This section explains how Atoll calculates the received carrier power and interference when a transmitter has a connected
repeater.
A mobile receiver receives signal from the donor transmitter as well as its repeater. Similarly, the signal from the mobile
is received at the donor transmitter as well as its repeater. In practice, when a mobile receiver is in the vicinity of the donor
transmitter, the signal to/from the repeater would be very weak due to high pathloss between the repeater and the mobile
receiver. Similarly, when the mobile receiver is located in the vicinity of the repeater, the signal to/from the donor transmit-
ter would be very weak due to the same reason.
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.
Calculation of Total Path Loss
The total pathloss, L total , is calculated by computing a downlink budget. If we take the case of a CDMA project, without
considering any shadowing margin or indoor loss, the power received from the donor transmitter, Txd on a carrier ic, at
the mobile receiver can be stated as (for a link over the air):
Txd Txd
Txd  P pilot  ic   G ant 
P rec  ic  = -----------------------------------------------------
Txd Txd – Rx
 L total –DL  L path 
Where,
Txd
P rec  ic  is the carrier power received at the receiver from the donor transmitter on a carrier ic (in W)
Txd
P pilot  ic  is the pilot power of the donor transmitter on the carrier ic (in W)
Txd
G ant is the donor transmitter antenna gain.
Txd
L total – DL is the transmission feeder loss of the donor transmitter.
Txd – Rx
L path is the path loss between the donor transmitter and the mobile receiver.
Similarly, the power received at the mobile receiver from the repeater, Rpk, is:
Txd Rpk
Rpk  P pilot  ic   G total – Air – DL 
P rec  ic  = --------------------------------------------------------------------
-
Rpk – Rx
L path
Where,
Rpk
P rec  ic  is the carrier power received at the mobile receiver from the repeater on a carrier ic (in W)
Txd
P pilot  ic  is the pilot power of the donor transmitter on the carrier ic (in W)
526 AT271_TRG_E6 © Forsk 2009

Rpk
G total – Air – DL is the output downlink total gain of repeater linked to a donor transmitter with an air link.
Rpk – Rx
L path is the path loss between the repeater and the mobile receiver
So, the total carrier power received at the mobile receiver is:
Txd Rpk
Txd – Rpk Txd Rpk Txd  G ant G total – Air – DL
P rec  ic  = P rec  ic  + P rec  ic  = P pilot  ic    ------------------------------------------------------ -
- + -----------------------------------
  L Txd
total – DL  L
Txd – Rx
path 
Rpk – Rx
L path 
Since,
Txd Txd
P pilot  ic   G ant
L total = ---------------------------------------------------------------
-
Txd Txd – Rpk
L total – DL  P rec  ic 
Therefore,
Txd Txd
P pilot  ic   G ant
L total = ----------------------------------------------------------------------------------------------------------------------------------------------------------------
-
Txd Rpk
Txd Txd  G ant G total – Air – DL
L total – DL  P pilot  ic    ------------------------------------------------------- + ------------------------------------
  L Txdtotal – DL  L path
Txd – Rx

Rpk – Rx
L path 
Hence,
Txd
G ant
L total = -----------------------------------------------------------------------------------------------------------------------------------
-
Txd Rpk
Txd  G ant G total – Air – DL
L total – DL   ------------------------------------------------------ - + ----------------------------------- -
  L Txd
total – DL  L path
Txd – Rx

Rpk – Rx
L path 
This total path loss depends on the location of the mobile receiver in realistic network scenarios. As a mobile in the donor
transmitter/repeater coverage area is likely to be far from the repeater/donor transmitter coverage area, the respective
pathloss value will be very large. This implies that we can study the two cases separately without influencing the results
much.
• Case 1: Receiver in Donor Transmitter Coverage Area
Rpk
Rpk – Rx G total – Air – DL
L path is likely to be very high, so the term -----------------------------------
Rpk – Rx
- can be ignored. This implies that:
L path
Txd – Rx
L total = L path
Considering this total pathloss value, the total received power in the uplink and in the downlink can be stated as:
Txd Txd Txd Txd
Txd  P pilot  ic   G ant   P pilot  ic   G ant 
P rec – DL  ic  = ---------------------------------------------- = ------------------------------------------------------
-
Txd Txd Txd – Rx
 L total – DL  L total   L total – DL  L path 
Rx Txd Rx Txd
Txd  P output  ic   G ant   P output  ic   G ant 
P rec – UL  ic  = --------------------------------------------------
- = ------------------------------------------------------
-
Txd Txd Txd – Rx
 L total – UL  L total   L total – UL  L path 
Where,
Rx
P output  ic  is the transmitted power from the mobile terminal on the carrier ic (in W)
Txd
L total – UL is the reception feeder loss of the transmitter
• Case 2: Receiver in Repeater Coverage Area

Txd
Txd – Rx G ant
L path is likely to be very high, so the term ------------------------------------------------------
Txd Txd – Rx
- can be ignored. This implies that:
 L total – DL  L path 
Txd Txd
G ant G ant
L total = --------------------------------------------------------------------------- = ----------------------------------------------------------------
Rpk Rpk Txd
 Txd  total – Air – DL 
G G total – Air – DL  L total – DL
 L total – DL   ----------------------------------- -  ----------------------------------------------------------------
  L Rpk path
– Rx
 Rpk – Rx
L path
Txd Txd Txd Rpk

Txd  P pilot  ic   G ant   P pilot  ic   G total 
P rec – DL  ic  = ---------------------------------------------- = -----------------------------------------------
-
Txd Rpk – Rx
 L total – DL  L total   L path 
Rx Txd Rx Rpk Txd

Txd  P output  ic   G ant   P output  ic   G total  L total – DL
-  ------------------------
- = ---------------------------------------------------
P rec – UL = --------------------------------------------------
Txd Rpk – Rx Txd
 L total – UL  L total   L path  L total – UL
Where,
© Forsk 2009 AT271_TRG_E6 527

Rx
P output  ic  is the transmitted power from the mobile terminal (in W)
Txd
L total – UL is the reception feeder loss of the transmitter
Calculation of Eb/Nt Uplink
In the uplink, the quality level at the transmitter on a traffic channel is:
E-----b-
C W
= ----  -----
 N t  UL I R
Where,
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 data rate 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.
Rx Txd
P output  G ant
C = P total – UL = -----------------------------------------
-
Txd
L total – UL  L total
I = I total + N 0
Where,
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, for each mobile terminal Rxi,

Rxi Txd
 P output  G ant 
I total =   -----------------------------------------
 L Txd
total – UL  L total
Rxi
-

Rxi
And,
Txd
N 0 = NF KTW
Where,
Txd
NF is the noise figure of the transmitter equipment at the reference point, i.e. the entry of the BTS
K is Boltzman constant
T is the ambient temperature (in K)
Hence
BTS
N 0 = NF KTW
10.1.1.4.3 Consideration of Repeater Noise Figure

You can define and assign a repeater equipment to each repeater. In addition to the allowed ranges of gains and powers
allowed to each repeater, these equipment contain a noise figure which is applied to the repeater they are assigned to.
This noise figure has an impact on the donor transmitter total reception losses. For information, see "Transmitter Radio
Equipment" on page 124.
10.1.2 GSM Documents

10.1.2.1 Over the Air
Atoll evaluates the signal level received from a repeater Rpk on a TRX type tt as follows:
Rpk Rpk Rpk
P rec  tt  = EIRP Air  tt  – P  tt  – L path – M Shadowing – L indoor +  G ant – L Rx  (in dB)
Rx
528 AT271_TRG_E6 © Forsk 2009

Figure 10.13: GSM Documents - Over the Air Repeater
Where,
Rpk
EIRP Air  tt  is the effective isotropic radiated power of the repeater on the TRX type tt linked to a donor transmitter over
the air. This value will be evaluated considering the path loss between the donor transmitter and the donor side of the
repeater (see after),
P  tt  is the power offset defined for the selected TRX type,
L Rx is the receiver loss,

Rx
Rpk
Total gain and EIRP can be either user-specified or directly calculated by Atoll from the link budget.
On each pixel, the resulting received signal on a TRX type tt is calculated as follows:
Txd – Rpk Txd Rpk
P rec  tt  = P rec  tt  + P rec  tt  (not in dB6)
10.1.2.1.2 EIRP Automatic Calculation

In GSM projects, an EIRP is calculated and repeated by the repeater. This EIRP is evaluated from the basis of the donor
BTS to the output of the repeater coverage antenna. In this technology, the UL part of the repeater is not modelled.
EIRP is calculated as follows:
Txd Txd Txd Txd – Rpk Rpk – donor – side Rpk – donor – side
Rpk P + G ant – L total – DL – L model + G ant – L feeder – Rec +
EIRP Air = (in dB)
G amp + G ant – L feeder – Tr
Figure 10.14: Over the Air Repeater - EIRP
Where,
Txd
P is the power of the donor transmitter,
Txd
© Forsk 2009 AT271_TRG_E6 529

Txd
L total – DL corresponds to the total downlink losses of the donor transmitter (user-defined or calculated considering trans-
mitter equipment characteristics),
Txd – Rpk
L model corresponds to path losses between the donor transmitter and the repeater. They are either user-defined or
calculated using the selected propagation model. If you do not select a propagation model, the propagation losses
between the donor transmitter and the repeater are calculated using the ITU 526-5 propagation model,
Rpk


L feeder – Tr corresponds to the losses of the repeater coverage side due to feeders (see Transmitter radio equip-
ment part). They involve the cable length at transmission on the coverage side and the loss per metre.
10.1.2.1.3 Donor Side Parameter Automatic Calculation

Azimuth
This is the angle at which the donor antenna is situated with respect to the North at the repeater side. This angle is meas-
ured clock-wise as shown in the figure below.
This parameter can be recalculated and applied using the Calculate button on the Donor Side tab. This 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.
Figure 10.15: Angle from North (Azimuth)
Mechanical Downtilt
This is the tilt angle for the repeater’s donor-side antenna, which ensures that it points towards the donor antenna in the
vertical plane. As a general rule, downtilt angles are considered positive and uptilt angles negative.
Figure 10.16: Positive/Negative Mechanical Downtilt
Since this parameter depends on the difference of heights/altitudes between the donor transmitter and the repeater, it can
be automatically calculated in the repeater’s Donor side properties. If the height/altitude of the antenna is modified, the
corresponding tilt angle can be found out and applied using the Calculate button. The following example may further clarify
the viewpoint:
Figure 10.17: Tilt Angle Computation
The tilt angle in the above figure would be:

Rpk – donor – side Txd
Rpk – donor – side  H ant – H ant 
T ant = atan  --------------------------------------------------------------
Txd – Rpk
-
 D 
530 AT271_TRG_E6 © Forsk 2009

As obvious, this angle will be negative for uptilts and positive for downtilts of the antenna. Here,
Txd – Rpk
D is the distance between the donor transmitter antenna and the repeater antenna,
Txd
H ant is the height/altitude of the donor transmitter antenna,

H ant is the height/altitude of the repeater’s donor-side antenna, and

T ant is the calculated tilt angle for the repeater’s donor-side antenna.
In the case of the above figure, the donor-side antenna at the repeater should have a tilt angle of X to point directly at the
transmitter antenna.
10.1.2.2 Microwave Link

Rpk Rpk Rpk
P rec  tt  = EIRP MW  tt  – P  tt  – L path – M Shadowing – L indoor +  G ant – L Rx  (in dB)
Rx
Figure 10.18: GSM Documents - Microwave Link Repeater
Where,
Rpk
EIRP MW  tt  is the effective isotropic radiated power of the repeater on the TRX type tt linked to a donor transmitter
through a microwave link. This value will be evaluated considering the link loss between the donor transmitter and the
donor side of the repeater (see after),

Rx
Rpk
Txd – Rpk Txd Rpk

Rpk Txd Txd – Rpk Rpk Rpk – coverage – side Rpk – coverage – side
EIRP MW = P – L MW + G amp + G ant – L feeder – Tr (in dB)
Where,
Txd
Txd – Rpk
© Forsk 2009 AT271_TRG_E6 531

Rpk


Figure 10.19: Microwave Link Repeater - EIRP
10.1.2.3 Fibre Link

Rpk Rpk Rpk
P rec  tt  = EIRP Fibre  tt  – P  tt  – L path – M Shadowing – L indoor +  G ant – L Rx  (in dB)
Rx
Where,
Rpk
EIRP Fibre  tt  is the effective isotropic radiated power of the repeater on the TRX type tt linked to a donor transmitter
through an optical fibre link. This value will be evaluated considering the fibre loss between the donor transmitter and the
donor side of the repeater (see after),

Rx
Rpk
Figure 10.20: GSM Documents - Fibre Link Repeater
Txd – Rpk Txd Rpk

532 AT271_TRG_E6 © Forsk 2009

Rpk Txd Txd – Rpk Rpk Rpk – coverage – side Rpk – coverage – side
EIRP Fibre = P – L Fibre + G amp + G ant – L feeder – Tr (in dB)
Figure 10.21: Fibre Link Repeater - EIRP
Where,
Txd
Txd – Rpk
Rpk


10.1.2.4 Appendices
10.1.2.4.1 Automatic Controls
Atoll performs systematic controls when creating a new repeater:
• On the amplification gain,
• On the downlink power.
Amplification Gain
The amplification gain of the repeater must be consistent with the limits of its equipment. Atoll checks that:
min Rpk max

G amp  G amp  G amp
Where
Rpk
G amp is the amplification gain of the repeater.
min
G amp is the minimum amplification gain allowed by the equipment.
max
G amp is the maximum amplification gain allowed by the equipment.
Downlink Power
The downlink power after amplification must be consistent with the limit of the equipment installed on the repeater. Atoll
checks that:
Rpk max Rpk – coverage – side Rpk – coverage – side
EIRP  tt   P + G ant – L feeder – Tr
Where
Rpk
EIRP  tt  is the effective isotropic radiated power of the repeater on the TRX type tt.

G ant is the gain of the repeater coverage side antenna.
max
P is the maximum downlink power allowed by the equipment.

L feeder – Tr corresponds to the transmission losses due to feeders installed on the repeater coverage side (see
Transmitter radio equipment part). They involve the cable length at transmission on the coverage side and the loss per
metre.
© Forsk 2009 AT271_TRG_E6 533

10.2 Modelling Remote Antennas

Atoll models remote antennas in a similar way to repeaters. A remote antenna is connected to the base station via an optic
fibre. The main difference from a repeater is that a remote antenna covers its own cell whereas a repeater extends the
coverage of an existing cell. In the case of a remote antenna, the donor transmitter does not have any main antenna. In
addition, it is assumed that a remote antenna, as opposed to a repeater, does not have any equipment and generates no
amplification gain nor noise.
10.2.1 CDMA Documents

10.2.1.1 Signal Level Received From Repeaters
Atoll calculates the signal level received from a remote antenna Ra on a carrier ic as follows:
Ra Txd Ra Ra
P rec  ic  = P pilot  ic  + G total – Fibre – DL – L path – M Shadowing – L indoor (in dB)
Figure 10.22: CDMA Documents - Remote Antenna Signal Level
Where,
Txd
Ra
G total – Fibre – DL is the output downlink total gain of the remote antenna of a donor transmitter with an optical fibre link.
This gain will be evaluated considering the fibre loss between the donor transmitter and the donor side of the repeater (see
after),
Ra
L path is the link loss between the repeater Ra and the receiver,
Since the donor transmitter antenna is not defined, the resulting received signal on a carrier ic is calculated as follows:
Txd – Ra Ra
P rec  ic  = P rec  ic  (not in dB6)
10.2.1.2 Gain Automatic Calculation

The downlink and uplink total gain are calculated as follows:
Ra Txd – Ra Ra – coverage – side Ra – coverage – side

G total – Fibre – DL = L Fibre + G ant – L feeder – Tr (in dB)
Figure 10.23: Remote Antennas - Downlink Total Gain
Where,
534 AT271_TRG_E6 © Forsk 2009

Txd – Ra
L Fibre is the user-defined fibre link loss between the donor transmitter and the remote antenna,
Ra – coverage – side
G ant is the gain of the remote antenna coverage side,
L feeder – Tr corresponds to the losses of the remote antenna coverage side due to feeders (see Transmitter radio
Ra Txd – Ra Ra – coverage – side Ra – coverage – side

G total – Fibre – UL = L Fibre + G ant – L feeder – Rec (in dB)
Figure 10.24: Remote Antennas - Uplink Total Gain
Where,
Txd – Ra
L feeder – Rec corresponds to the losses of the remote antenna coverage side due to feeders (see Transmitter radio
10.2.2 GSM Documents

10.2.2.1 Signal Level Received From Repeaters
Atoll calculates the signal level received from a remote antenna Ra on a TRX type tt as follows:
Ra Ra Ra
P rec  tt  = EIRP Fibre  tt  – P  tt  – L path – M Shadowing – L indoor +  G ant – L Rx  (in dB)
Rx
Figure 10.25: GSM Documents - Remote Antenna Signal Level
Where,
Ra
EIRP Fibre  tt  is the effective isotropic radiated power of the remote antenna on the TRX type tt linked to a donor trans-
mitter through an optical fibre link. This value will be evaluated considering the fibre loss between the donor transmitter
and the donor side of the remote antenna (see after),

Rx
Rpk
L path is the path loss between the remote antenna Ra and the receiver,
© Forsk 2009 AT271_TRG_E6 535

Since the donor transmitter antenna is not defined, the resulting received signal on a TRX type tt is calculated as follows:
Txd – Ra Ra
P rec  tt  = P rec  tt  (not in dB6)
10.2.2.2 EIRP Automatic Calculation

Ra Txd Txd Txd – Ra Ra – coverage – side Ra – coverage – side

EIRP Fibre = P + L total – DL – L Fibre + G ant – L feeder – Tr (in dB)
Where,
Txd
Txd – Ra
L feeder – Tr corresponds to the losses of the remote antenna coverage side due to feeders (see Transmitter radio
Figure 10.26: Remote Antennas - EIRP
536 AT271_TRG_E6 © Forsk 2009

Index
Index
Symbols Adaptive antenna systems 408, 505

Adaptive MIMO Switch 498, 516
Inter-carrier neighbours 255, 347
Adaptive Modulation and Coding 209
.agd format 36, 38
Additional Ec/Io conditions 259, 351
.bil file 43, 44
Additional electrical downtilt 111
.bil format 35, 36, 37, 43
Adjacence criterion 157, 254, 347
.bmp file 46
Adjacency factor 159, 511
.bmp format 35, 36, 37, 46
Adjacent channel suppression factor 434
.bmp raster data encoding 48
Adjacent PN-Clusters per site allocation 352
.bmw file 37, 46
Adjacent PN-Clusters Per Site Strategy 357
.bpw file 37, 49
Adjacent transmitter 159, 257, 349
.bpw sample 49
Admission Control 217
.clc file 66
Admission control 202, 299, 304, 311, 316
.clc format 65
Admission control in the R99 part 226
.clc sample 66
A-DPCH Eb/Nt DL 248
.cs file 27
A-DPCH Eb/Nt UL 248
.dbf file 58, 60, 61, 63
AEDT 111
.dbf format 50, 58, 61
AHRxeff calculation 96
.dbf header 58, 62
Algorithm based on coverage overlapping 268, 359, 427
.dct file 67
Algorithm based on distance 268, 358, 427
.dct sample 67
Algorithm initialization 200, 298, 303, 310
.dxf format 37, 50
All servers 161, 166, 176, 184, 274, 447
.ecw format 37, 51
Allocate Carriers Identically 266
.grc format 51
Allocate carriers identically 259, 352
.grd format 51
Allocation examples 264
.hdr file 43
Allocation for a group of cells 258, 350
.im0 format 64
Allocation for a group of transmitters 160
.im0 Sample 65
Allocation strategies 260, 264, 266, 352, 418
.im1 format 68
Allocation strategy 267, 426
.im1 sample 68
Altitude 77, 81
.im2 format 69
Ambient temperature 187, 277, 284, 368, 435, 528
.im2 sample 69
AMS 497, 498, 516
.los file 60, 61
AMS threshold 497
.mid format 37
Analysis based on a specific carrier 242, 338
.mif format 37, 50
Analysis based on all carriers 338
.mnu format 56
Analysis on a specific carrier 247, 346
.mnu sample 56
Analysis on all carriers 247, 346
.pts file 63, 64
Angular distribution of the uplink noise rise 508
.shp format 37, 50
Antenna 38
.shx format 50
Antenna attenuation 73, 110, 131, 183, 273, 372, 373, 384, 475
.tab format 50
Antenna attenuation calculation 109
.tfw file 45
Antenna coordinates 76
.tfw sample 46
Antenna diversity gain 124, 126
.tif format 35, 36, 37, 45
Antenna height 76
.wld file 49
Antenna pattern 76
.wld sample 49
Antenna pattern 3-D interpolation 110
Antenna pattern smoothing 111
Application level channel throughput 498
Numerics Application level user throughput 505
3-D interpolation of horizontal and vertical patterns 109 Application throughput 250, 460, 515
8PSK modulation 166, 168, 169, 170, 171, 172 Application throughput/timeslot study display type 173
ArcView Grid file description 52
ArcView Grid file sample 52
A ArcView Grid format 52
AS 192, 289
AAS 515
AS analysis tab 232, 319
AAS usage 435, 458
Atoll.ini file 45, 111, 123, 127, 193, 201, 290, 377
Absolute Spot Ht 87
Automatic allocation description 259, 267, 351, 358, 361, 418, 426
Acknowledgement gain 286
Automatic cdmaOne-CDMA2000 neighbour allocation 360
Acl calculation 96
Automatic GSM/TD-SCDMA neighbour allocation 426
Active set 192, 199, 234, 258, 321, 335, 339, 342, 343, 350
Automatic GSM-CDMA neighbour allocation 357
Active set determination 204
Automatic GSM-UMTS neighbour allocation 267
Active set management 192, 289
Automatic neighbour allocation 160, 254, 346
Activity factor 163
Automatic repeater calculation 520, 523, 524, 529, 531, 532, 534, 536
Activity factor for voice services 435
Automatic SPM Calibration 91
Activity status 193, 194, 226, 291
Average application throughput per mobile 251
Adaptive Antenna Systems 515
© Forsk 2009 AT271_TRG_E6 537

Average array correlation matrix 409, 507 C

Average call duration 291
C/(I+N) 451, 461, 462
Average MAC throughput per mobile 251
C/(I+N) level 170, 171, 460
Average noise level 247
C/I level 165, 167, 169, 170, 341
Average noise rise 247, 346
C/I ratio 160, 310
Average packet switched traffic demand 152
C/I threshold 66, 68
Average RLC throughput 250
C/N 170, 171
Average RLC throughput per mobile 251
C/N level 450
Average throughput/timeslot 173
Calculate 76
Azimuth 76, 109, 127, 268, 358, 361, 427, 451
Calculate / Force Calculation comparison 76
Calculate All 76
Calculate or Force Calculation? 77
B Calculation area 75
Bar graph and pilot sub-menu 232, 320 Calculation area determination 75
Base band hopping 162 Calculation area management 77
BBH 162 Calculation areas 76
BCCH 39 Calculation bin 74, 83, 86
Beam forming 408, 505 Calculation criteria 131, 183, 273, 373
Bearer 211, 434, 451 Calculation of azimuth and tilt angles 109
Bearer efficiency 434, 497 Calculation of Eb/Nt uplink 528
Bearer selection threshold 435 Calculation of inter-transmitter distance 270, 429
BER 174, 179 Calculation of the required power for DL traffic channel 300, 307
Best bearer 460, 461, 462 Calculation of the service usage duration per hour 194, 291, 378, 452
Best cell edge coverage probability (%) 186, 276, 449 Calculation of total path loss 526
Best coding schemes 173 Calculation options 167, 176
Best Effort 501 Calculation radius 75
Best Idle Mode Reselection Criterion (C2) 134 Calculations based on C 167
Best server 193, 289 Calculations based on C/(I+N) 177
Best server path loss (dB) 185, 275, 449 Calculations based on C/I considering thermal noise 168
Best server total losses (dB) 185, 276, 449 Calculations Based on C/I Without Considering Thermal Noise 167
Best signal level 273 Calculations based on C/N 177
Best signal level (dBm) 374 Candidate neighbours 158, 257, 349, 358, 360, 361
Best signal level (in dBm, dBµV, dBµV/m) 185, 275, 448 Carrier power 147, 153, 162, 166
Best signal level and a margin 132, 184, 274, 447 Carrier power received at the mobile receiver from the repeater 526
Best signal level of the highest priority HCS layer and a margin 161 Carrier power received from the mobile terminal 528
Best signal level of the highest priority layer and a margin 167, 176 Carrier selection mode 201, 202, 203, 299, 304, 311
Best signal level per HCS layer and a margin 161, 166, 176 carrier selection mode 201, 202, 299, 304, 311
Best throughput/timeslot 173 Carrier to interference and noise ratio 168, 177
Best traffic signal level 450 Carrier to interference ratio 162, 167
Biased (QoS Class) scheduler 502 Carrier to interference ratio calculation 161
BIL samples 44 Carrier-to-interference distribution 147
Bilinear interpolation 77, 80, 96, 275 Carrier-to-interference ratio 153
Bin 74 CDMA2000 82
Bin size 113 CDMA2000 1xEV-DO 327, 335, 340
Bin vertices 35 CDMA2000 1xEV-DO power/rate control simulation algorithm 310
Binary tree 317 CDMA2000 1xEV-DO services 292, 296
Bit Error Rate 174 CDMA2000 1xEV-DO users 320
Bit power received at terminal on carrier ic 190 CDMA2000 1xRTT 324, 331
Bit power received at transmitter on carrier ic used by terminal 190 CDMA2000 1xRTT power control simulation algorithm 302
Bit received power at terminal for FCH on carrier ic 279 Cell 39
Bit received power at terminal for FCH+SCH on carrier ic 279 Cell edge coverage probability 73, 77, 114, 115, 116, 117, 158, 233, 242, 243,
Bit received power at terminal for SCH on carrier ic 279 244, 255, 321, 339, 340, 342, 344, 347, 359
Bit received power at transmitter for FCH on carrier ic 280 Cell edge coverage probability (%) 186, 276, 374, 376, 449
Bit received power at transmitter for SCH on carrier ic 280 Cell FCH power for a traffic channel on carrier ic 278
Bit received power at transmitter for SCH+FCH on carrier ic 280 Cell other common channel power (except CPICH and SCH) 278
Bit received power at transmitter on carrier ic 286 Cell pilot power 188, 273, 278
BLER 451, 497, 499, 504 Cell priority 261, 353
BLER percentage 173 Cell size 82
Block error rate 497, 499, 504 Cell synchro channel power 188, 189
Block error rate computation 173 Cell synchronisation channel power 278
Blocking Probability 152 Cell type 39
Blocking probability 147, 149, 150, 154, 156 Central meridian 53
Body loss 189, 279, 286, 367, 436 Channel bandwidth 434
Boltzman constant 187, 277, 284, 368, 435, 528 Channel element management 317
Break distance 86 Channel elements management 227
Broadcast 82 Channel throughput 515
BSIC format 31 Chip power received at terminal 190, 279
BTS 124 Chip rate 187, 277, 285
BTS Noise Factor 366 Chip received power at transmitter 280
BTS noise figure 124, 125, 126 Circuit quality indicators 174
Buddy algorithm 227, 317 Circuit switched service 194, 197, 378, 381
Circuit switched traffic 155
Circuit switched traffic demand 150, 151, 155
Circuit switched traffic overflow rate 155
Cirquit quality indicators studies 174
538 AT271_TRG_E6 © Forsk 2009

Index
Cirquit quality indicators study display types 179 Colour per required power margin 245, 246, 340, 343, 344
Clearance angle 96 Colour per Required Terminal Power 252
Clustered allocation 260, 418, 419, 424 Colour per RLC Peak Rate 253
Clustered strategy 265 Colour per RLC peak rate 250
Clutter 36, 76, 80, 83, 113 Colour per RLC peak throughput 250
Clutter class 78, 83 Colour per RLC peak throughput per mobile 251
Clutter classes 36, 43, 45, 46, 113 Colour per service 244, 246, 339, 342, 343
Clutter classes file 44, 46, 49, 53 Colour per transmitter 242, 243, 245, 338, 339, 342, 343
Clutter classes file sample 54 Colour per UL soft handover gain 246
Clutter determination 78 Common channels power (except CPICH and SCH) 188
Clutter height 78, 80, 90 Complex smart antenna weight 371, 445
Clutter height file 78 Complex weights 410, 507
Clutter heights 36, 43, 45 Compressed mode 205, 206, 235, 236, 239
Clutter maps 36 Computation zone 75, 157, 254, 259, 351, 418, 509, 512
Cn calculation 95 Connector transmission loss 125, 126, 127
Coding rate 434 Control of radio resource limits (MAC indexes and site channel elements) 314
Coding scheme based on C with ILA 169 Control of radio resource limits (OVSF codes, cell power, channel elements)
Coding Scheme Based on C Without ILA 168 206
Coding scheme based on C/I with ILA and thermal noise 171 Control of radio resource limits (Walsh codes, cell power and site channel ele-
ments) 301, 308
Coding scheme based on C/I with ILA and without thermal noise 170
Convergence 220, 302, 310, 316
Coding scheme based on C/I without ILA and thermal noise 169
Convergence criteria 220
Coding scheme based on C/I without ILA and with thermal noise 170
Convergence criterion 302, 309, 316
Coding schemes 172
Coordinate system 127
Collision 164
Coordinate system file formats 27
Collision probability for BBH and SFH modes 164
Coordinate systems 25
Collision probability for non hopping mode 164
Coordinate systems in Atoll 26
Colour by application throughput 461, 462
Corrected standard loss 95
Colour by best bearer 461, 462
Correction for hilly regions in case of LOS 89
Colour by C/(I+N) level 461
Corrections in the Hata formula 83
Colour by effective MAC throughput 461, 462
Correlated MSAs m and n 164
Colour by peak MAC throughput 461, 462
Correlation coefficient 122
Colour by preamble signal level 450
Correlation coefficient determination 120, 123
Colour by total noise 462
Co-site cell 268, 358, 360, 361, 427
Colour by traffic C/N level 451
Co-site factor 159, 511
Colour by traffic signal level 450
Co-site transmitter 159, 257, 349
Colour by uplink C/N level 451
Cost-Hata 83
Colour by uplink signal level 451
Cost-Hata model 80
Colour per Application Throughput 253
Coverage area determination 165, 166, 175, 447
Colour per application throughput 250
Coverage area display 165
Colour per application throughput per mobile 251
Coverage by best bearer 459, 461, 462
Colour per average noise level 247, 346
Coverage by C/(I+N) level 459, 460, 461
Colour per average noise rise 247, 346
Coverage by C/I level study 160
Colour per average RLC throughput 250
Coverage by channel throughput 459, 461, 462
Colour per C/I 341
Coverage display 172, 179, 185, 275, 448, 450, 460
Colour per Cell Edge Coverage Probability 252
Coverage plot reports 37
Colour per cell edge coverage probability 243, 244, 339, 340, 342, 344
Coverage resolution 448, 450, 460
Colour per CQI 249
Coverage studies 131, 160, 184, 241, 274, 338
Colour per data rate 340, 343, 344
Coverage study conditions 77
Colour per effective quality level (Effective Eb/Nt) 244, 246, 340, 342, 344
Coverage study display options 77
Colour per HS-PDSCH Ec/Nt 249
Coverage study resolution 77
Colour per HS-SCCH Ec/Nt 249
Coverage Study Scenarios 168
Colour per HS-SCCH power 249
CPICH quality 209
Colour per MAC Rate 252
CQI 209, 238, 249
Colour per MAC rate 249
CQI based on CPICH quality 209
Colour per MAC throughput 249
CQI based on HS-PDSCH quality 213
Colour per MAC throughput per mobile 251
CQI study based on C/(I+N) with ILA 179
Colour per max A-DPCH Eb/Nt DL 248
CQI study based on C/(I+N) without ILA 178
Colour per max A-DPCH Eb/Nt UL 248
CQI study based on C/N with ILA 178
Colour per maximum noise level 247, 346
CQI study based on C/N without ILA 178
Colour per maximum noise rise 247, 346
Cyclic prefix ratio 433, 514
Colour per maximum quality level (Max Eb/Nt) 246, 342, 344
Colour per maximum quality level (max Eb/Nt) 244, 340
Colour per minimum noise level 247, 346
Colour per minimum noise rise 247, 346 D
Colour per Minimum RLC Throughput 253 Data Erlangs 147
Colour per mobility 242, 243, 246, 339, 342, 343 Data rate 340, 343
Colour per noise level 247, 346 Data Rate Control gain 286
Colour per noise rise 248, 346 Data service users 291, 294
Colour per probability 242, 244, 246, 339, 340, 342, 343 Data services 453, 455
Colour per quality level (Ec/Io) 243, 339 Datum 25
Colour per quality margin (Eb/Nt margin) 244, 246, 340, 342, 344 Datum codes 28
Colour per quality margin (Ec/Io margin) 243, 339 DC subcarrier 434
Colour per rate 341 Dedicated packet switched timeslots 151
Colour per Required E-DPDCH Ec/Nt 252 Delay 149, 150, 152, 154, 156
Colour per required power 245, 246, 340, 343, 344 Delete existing neighbours 158, 160, 255, 268, 347, 358, 359, 361, 427
© Forsk 2009 AT271_TRG_E6 539

Delete existing neighbours option 270, 360, 362, 428 Downlink power control 205, 300, 306
DEM 35 Downlink rake efficiency factor 236, 322, 323, 325, 327
Description of coordinate systems 25 Downlink rake receiver efficiency factor 187, 277
Determination of Mi’s Best server 298, 303, 311 Downlink rate control 315
Determination of Mi’s best server 201 Downlink reuse factor on a carrier ic 221, 283
Determination of the active set 299, 304, 312 Downlink SCH bit rate 277
Deterministic clutter 91 Downlink service area analysis 243, 339
Deygout construction 106, 107 Downlink service processing gain on FCH 278
Deygout method 84, 94, 98, 105 Downlink service processing gain on SCH 278
Deygout method with correction 107 Downlink soft handover gain 237, 324
Diffraction 85, 87, 90, 95 Downlink soft handover gain for FCH channel on carrier ic 281
Diffraction calculation method 82 Downlink soft handover gain for SCH channel on carrier ic 281
Diffraction loss 83, 90, 98, 104 Downlink soft handover gain on FCH 327
Diffraction losses 78 Downlink soft handover gain on SCH 327
Digital elevation model 35 Downlink sub-menu 234, 321
Digital Terrain Model 35, 43, 45, 46, 77 Downlink total noise analysis 246, 345
Digital Terrain Model file 44 Downlink total power 232
Dimensioning 150 Downlink traffic channel quality 339
Dimensioning model 150, 154 Downlink traffic data quality 324, 340
Display BSIC format 31 Downtilt 76, 109
Display coordinate system 26 DTM 35, 76, 77, 80, 83, 89, 91, 98
Display length units 31 DTM file 53
Display power units 30 DTM sample 53
Display types 185, 275, 448 DTX 163, 166
Distance criterion 262, 353, 420, 423
Distributed per cell 418, 419, 424
Distributed per cell allocation 260 E
Distributed per site 419, 425
Earth curvature 104
Distributed per site allocation 260
Eb/Io target on downlink 186
Distributed per site strategy 266
Eb/Io target on downlink FCH 276
Distributed PN-clusters per site allocation 352
Eb/Io target on downlink SCH 277
Distributed PN-Clusters Per Site Strategy 357
Eb/Nt 115, 119, 244, 246
Distributed strategy 265
Eb/Nt margin 244, 246, 340, 342, 344
Diversity mode 451
Eb/Nt max 236, 323
DL Ec/Io 115
Eb/Nt max for each cell of active set 236, 239
DL gain due to availability of several pilot signals at the mobile 189, 279, 286
Eb/Nt max for the first f (number of fingers) cells of active set 322, 323, 324
DL macro-diversity gain 115, 123
Eb/Nt max on FCH and Eb/Nt max on SCH 327
DL macro-diversity gain evaluation 121
Eb/Nt max on FCH for the first f (number of fingers) cells of active set 325
DL min power 201, 202, 203, 298, 303, 311
Eb/Nt max on SCH for the first f (number of fingers) cells of active set 326
DL minimum power 242, 320
Eb/Nt target 228, 234, 322
DL noise rise 346
Eb/Nt target for FCH channel on uplink 277
DL subframe ratio 515
Eb/Nt target for SCH channel on uplink 277
DL total noise 346
Eb/Nt target on FCH 324
Donor side parameters 521, 530
Eb/Nt target on SCH 324
Donor transmitter antenna gain 520, 521, 526, 529
Eb/Nt target on uplink 186, 288, 367
Donor transmitter antenna height/altitude 522, 531
Ec/Io 123, 193, 243, 255, 259, 320, 339
Donor transmitter antenna-to-repeater antenna distance 522, 531
Ec/Io activation threshold 235, 236, 239
Donor transmitter feeder losses 526
Ec/Io evaluation 232
Donor transmitter pilot power 519, 522, 526, 534
Ec/Io margin 243, 255, 339
Donor transmitter power 529, 531, 533, 536
Ec/Io target on downlink for active set members 276, 284
Donor transmitter total downlink losses 520, 521
Ec/Io target on downlink for the best server 186, 276, 284
Donor transmitter’s total downlink losses 530
Effective C/I 328
Donor transmitters 519
Effective Eb/Nt 237, 240, 244, 246, 323, 330, 340, 342, 344
Donor transmitter-to-mobile receiver path loss 526
Effective Eb/Nt FCH and SCH 334
Donor transmitter-to-repeater path loss 520, 521, 523, 525, 530, 531, 533, 535,
536 Effective Eb/Nt on FCH and Eb/Nt on SCH 327
Downlink activity factor on FCH 278 Effective MAC channel throughput 451, 498
Downlink average cell rate 285 Effective MAC throughput 460, 515
Downlink Average Rate 341 Effective MAC user throughput 505
Downlink budget 526 Effective pilot quality level 327
Downlink Data Rate 341 Effective quality level 340, 342, 344
Downlink extra-cellular interference at terminal on carrier ic 191, 192, 221, 280, Effective rate 328
281, 283, 287, 288, 289 Effective rate of traffic overflow 151
Downlink extra-cellular interference based on pilot at terminal on carrier ic 287 Effective receiver antenna height 89
Downlink extra-cellular interference based on traffic at terminal on carrier ic Effective service data rate in the uplink 528
287, 289 Effective traffic channel 335
Downlink FCH nominal rate 277 Effective traffic channel quality 240, 327
Downlink intra-cellular interference at terminal on carrier ic 191, 280, 287, 288 Effective traffic overflow rate 152, 155
Downlink load factor 228 Effective transmitter antenna height 86, 95
Downlink load factor calculation 228, 318 EIRP 74, 117, 131, 183, 273, 372, 373, 467, 473
Downlink load factor on carrier ic 221, 283, 289 Electrical tilt 110
Downlink load factor per cell 228, 318 Ellipsoid 25
Downlink load factor per mobile 229, 319 Ellipsoid codes 29
Downlink macro-diversity gain 234, 321 Enhanced slope at receiver 87
Downlink maximum rate supplied to the terminal 285 Environment class 36
Downlink noise rise 346
540 AT271_TRG_E6 © Forsk 2009

Index
Environment traffic maps 36, 43, 45 Geographic data formats 37

Epstein-Peterson method 107 Geographic file resolution 76
Erceg-Greenstein (SUI) path loss formula 97 Geographic layer order 76
Erceg-Greenstein (SUI) propagation model 96, 98 Geographic profile extraction 78
Erdas Imagine 35, 36, 37, 38 Georeferenced formats 37
Erlang 147 Global allocation for all cells 254, 347
Erlang B 147, 151, 155 GPRS EGPRS coverage studies 166
Erlang C 147, 151, 155 GPRS/EGPRS coding schemes study display types 172
Erlangs 37, 151 GPRS/EGPRS Studies Based on C With ILA 169
ErtPS 500, 501 GPRS/EGPRS Studies Based on C Without ILA 168
Exceptional pair 159, 257, 268, 349, 358, 360, 361, 427 GPRS/EGPRS studies based on C/I with ILA and thermal noise 171
Exceptional pair criterion 263, 355, 422, 423 GPRS/EGPRS studies based on C/I with ILA and without thermal noise 170
Exceptional pairs 259, 351 GPRS/EGPRS studies based on C/I without ILA and thermal noise 169
Existing neighbour 159, 257, 268, 349, 358, 360, 361, 427 GPRS/EGPRS studies based on C/I without ILA and with thermal noise 170
Exponential weighting function 90 Grade of Service 151
External neighbour allocation 358 Greenwich 25
External neighbour allocation algorithm 361 Grid resolution 76
Externalised propagation results format 58 Ground altitude determination 77
Ground elevation 86, 89, 96
Ground height 96
F Ground slope at receiver 86
GSM GPRS EGPRS documents 39, 131
Fading 121
GSM1800 82
Far from transmitter 86
GSM900 82
Fast fading 113
Guard subcarriers 434
Fast link adaptation 209, 248
Fast link adaptation modelling 248
FAX-CCITT3 45
FCH 322 H
FCH activity factor 291, 293 Half-rate circuit switched traffic 151
FCH nominal rate 293 Half-rate traffic 151
FCH nominal rates 294 Half-rate traffic ratio 151
FCH rates 291, 294 Handoff 199, 204, 235, 236, 238, 240, 305, 312, 322, 323, 325, 327, 332, 334,
FDD 434 335, 337, 383
Feeder 124 Handover 329, 330
Feeder length 125, 127 Handover end 158
Feeder loss per metre 125, 127 Handover start 158
Feeder noise figure 125, 126 HARQ 341, 345
Feeder reception loss 124, 125, 126 Hata calculations 83
Feeder transmission loss 125, 126, 127 Hata formula 83
Feeder UL gain 125, 126 HCS 161, 166, 175
FER 174, 179 Height above average profile 86
FFT size 434 Height above ground 86
Fibre repeater link 524, 532 Hexagonal design 39
Field strength 95 Hierarchical cell structure layer 161
File formats 43 Hierarchical knife-edge sorting 105
Focus zone 157, 254, 259, 351, 418, 509, 512 Histogram 66
Force adjacent cells as neighbours 254, 347 Hopping sequence 164
Force adjacent transmitters as neighbours 157 Horizontal pattern 110
Force co-site cells as neighbours 254, 268, 347, 359, 361, 427 HSDPA 209, 212
Force co-site transmitters as neighbours 157 HSDPA application throughput 223
Force exceptional pairs 158, 254, 268, 347, 350, 358, 359, 361, 427 HSDPA Bearer 209
Force neighbour symmetry 158, 254, 347, 350 HSDPA bearer allocation process 208
Force symmetry 350 HSDPA bearer selection 211, 215
Frame duration 433, 497, 514 HSDPA Part of the Algorithm 207
Frame Erasure Rate 174 HSDPA power allocation 207, 390
Free space loss 95, 104 HSDPA power dynamic allocation 224
Frequency 76, 83 HSDPA prediction studies 248
Frequency band 60, 82 HS-DSCH 209
Frequency hopping mode 162, 163 HSN 164
Fresnel zone 104 HSPA Related Simulation Results 222
Fresnel zone index 105 HS-PDSCH 212
Friis' equation 125, 126 HS-PDSCH CQI 211, 215
HS-PDSCH CQI Determination 215
HS-PDSCH Ec/Nt 237, 249
HS-PDSCH quality calculation 210, 213
G HS-PDSCH quality update 213
Gaussian 118, 120, 122 HS-SCCH Ec/Nt 237, 249
Gaussian distribution 113 HS-SCCH power 249
General prediction studies 131, 183, 273 HSUPA application throughput 223
Generating a realistic user distribution 193, 290, 377, 452 HSUPA Bearer 217
Generic geographic data 43, 45, 46 HSUPA bearer allocation process 218
Generic raster header file (.wld) 49 HSUPA Bearer Selection 218
Geodetic datum 25 HSUPA Part of the Algorithm 216
Geographic coordinate system 25 HSUPA Prediction Study 252
Geographic data 35 HSUPA User Equipment Categories 217
© Forsk 2009 AT271_TRG_E6 541

Hybrid Automatic Repeat Request 341, 345 L

Lambert Conformal-Conic method 25
Land use 36
I Large city 83
Ideal Link Adaptation (ILA) 168, 169, 170, 171 Length units 31
Ideal link adaptation (ILA) 178 Line of sight 86, 87, 94
Identity matrix 508 Linear regression 88
Idle power gain 285 Link budget 117
IEEE working group 802.16 97 Live traffic maps 36
Image files 55 Load 147
Inter-carrier neighbour 254 Load factor 207, 299, 318
Inter-carrier neighbours 346 Load rise 316
Inter-carrier power sharing 231 Load saturation 226, 316
Interfered areas study 160 Logarithmic weighting function 90
Interfered transmitter 65, 66, 68 Log-normal distribution 113
Interfered TRX type 65, 66, 68 LOS 87, 88
Interference 74, 163, 166, 302, 308, 314, 383 Losses due to clutter 90
Interference calculation 163 LZW 45
Interference correlation matrix 372, 446
Interference histograms formats 64
Interference prediction studies 160 M
Interference studies 74
MAC rate 249
Interference tab 166
MAC throughput 249
Interfering transmitter 65, 66, 68
Macro diversity 186
Internal BSIC format 31
Macrocell 82
Internal coordinate system 26
Macro-diversity gain 115, 233, 239, 240
Internal length units 31
Main antenna 127
Internal power units 30
MAIO 162, 164
Inter-technology neighbour allocation 357
MAL 162, 164
Inter-technology neighbours 362
MAS 166
Inter-transmitter distance 267, 358, 426
Matrix validity 76
Intra-carrier neighbour 254
Max Aggregate Throughput scheduler 503
Intra-carrier neighbours 255, 346, 347
Max BER 179
Intra-technology neighbour allocation 157
Max C/I level 165
Intra-technology neighbour allocation algorithm 254, 346
Max Cell power 188
Invalid path loss matrices 76
Max cell power 285
Inverse noise correlation matrix 411, 508
Max FER 179
Io (Best server) 234, 321
Max MOS 179
IS-95 cdmaOne 321, 328
Max terminal power 238, 328, 335
IS-95 cdmaOne and CDMA2000 1xRTT 339
Max terminal power on FCH and SCH 331
IS-95 cdmaOne and CDMA2000 1xRTT services 291, 293
Maximum allowable blocking probability 151
IS-95 cdmaOne and CDMA2000 1xRTT users 320
Maximum BLER percentage 174
IS-95 cdmaOne and CDMA2000 documents 273
Maximum cell power 273, 278
IS-95 cdmaOne and CDMA2000 prediction studies 319
Maximum Eb/Nt 342, 344
IS-95 cdmaOne power control simulation algorithm 297
Maximum inter-site distance 157, 347
Iteration 199, 383
Maximum loading factor 187, 277, 285
Iterations 220, 302, 303, 310, 316
Maximum noise level 247, 346
ITU 1546 82
Maximum noise rise 247, 346
ITU 370-7 82
Maximum number of HSDPA users 207
ITU 370-7 calculations 95
Maximum number of MAC indexes available per cell (59) 284
ITU 370-7 formula 95
Maximum number of neighbours 267, 358, 360, 362, 426
ITU 526-5 82
maximum number of neighbours allowed for each cell 159, 258, 350
ITU 526-5 calculations 95
Maximum number of OVSF codes available per cell (512) 187, 277
ITU 526-5 formula 95
Maximum number of TRXs per transmitter 151
ITU 529-3 82
Maximum permissible delay 151
ITU Rec. 370-7 96
Maximum power allowed on FCH 324, 326
ITU Rec. 526-5 104
Maximum power allowed on traffic channel 235
ITU-R P.1546-2 98
Maximum quality level 340, 342, 344
ITU-R P.370-7 95
Maximum reduction factor 148
ITU-R P.526-5 model 95
Maximum terminal power 436
Maximum terminal power allowed 328
Maximum throughput demand 436
J Maximum traffic channel power per transmitter allowed on FCH 321
JD factor 366 Mean Opinion Score 174
Mechanical downtilt 451
Meridian 25
K Microwave repeater link 522, 531
Millington method 107
Key Performance Indicators calculation 154
MIMO 435, 497, 498, 516
Knife-edge diffraction 104
Min C/I level 165
KPI 150
Minicell 82
KPIs 152
Minimum covered area 158
Minimum dedicated packet switched timeslots per transmitter 151
Minimum Mean Square Error 408, 505
542 AT271_TRG_E6 © Forsk 2009

Index
Minimum noise level 247, 346 Number of inactive users 194, 198, 291, 292, 296, 378, 382, 453, 454
Minimum noise rise 247, 346 Number of MAC indexes used by the cell 284
Minimum percentage of covered traffic 158 Number of mobiles connected to transmitter Sj on carrier ic 289
Minimum pilot quality level on uplink 284 Number of OVSF codes used by the cell 187, 277, 284
Minimum power allowed on FCH 300, 306 Number of rake fingers 279, 321
Minimum power allowed on SCH 306 Number of required TRXs per transmitter 151
Minimum power allowed on traffic channel 205, 299, 304, 312 Number of servers 186, 276, 449
Minimum required throughput per user 151 Number of shared timeslots 154
Minimum signal level 158 Number of subchannels per channel 434
Minimum terminal power 436 Number of used subchannels 452
Minimum throughput demand 435 Number of users active on DL 194, 198, 291, 292, 296, 378, 382, 453, 454
Minimum throughput reduction factor 147 Number of users active on UL 194, 198, 291, 292, 296, 378, 382, 453, 454
Minimum throughput reduction factor calculation 153 Number of users active on UL and DL 194, 198, 291, 292, 293, 294, 296, 378,
MMSE Smart Antenna Model 445 382, 453, 454
Mobile allocation list 162
Mobile receiver 166, 174
Mobile Station Allocation 162 O
Mobility 200, 234, 239, 242, 243, 246, 298, 303, 336, 338, 339, 342, 343, 345, Okumura-Hata 83
449, 459
One cluster per site allocation 260
Model standard deviation 73, 113, 233, 321
One cluster per site strategy 266
Modelling the transmitter-repeater link 534
One histogram per line format 64
Modulation 434
One SYNC_CL code per site 418
Monte-Carlo algorithm 193, 290
One SYNC_DL code per site 419, 425
Monte-Carlo simulation 117
One value per line with dictionary file format 65
Monte-Carlo simulator 37
One value per line with transmitter name repeated format 68
MOS 174, 179
Only co-channel and adjacent values format 69
MSA 162, 164
Optimum Beamformer 408
MUD factor 186, 276, 284
Orthogonality factor 186, 276, 284, 368
Multi user gain 285
Other geographic data 37
Multipath interference 113
Over the air repeater link 519, 528
Multiple Input Multiple Output 516
Overhead 515
Multiple knife-edges 105
Overlapping factor 159, 511
Multi-resolution profile management 80
OVSF code management 226
OVSF codes 227
N
Near cells 260, 352, 419 P
Near transmitter 86
Packbit 45
Neighbour allocation 157
Packet queue length 149
Neighbour priority 159, 257, 349, 361
Packet switched service 194, 198, 379, 382
Neighbourhood cause 159, 257, 268, 349, 358, 360, 361, 427
Packet switched traffic 155
Neighbourhood criterion 262, 354, 421, 423
Packet switched traffic demand 150, 152, 157
Neighbout importance 158
Packet switched traffic load 147, 152
Neighbout Importance Function 159
Packet switched traffic overflow 156
Network dimensioning 37
Packet traffic overflow 154
Network dimensioning engine 150
Parameters used for CDMA2000 1xEV-DO modelling 283
Network dimensioning process 150
Parameters used for IS-95 cdmaOne and CDMA2000 1xRTT modelling 276
Network dimensioning steps 151
Path loss 43, 73, 83, 89, 113, 127, 131, 183, 189, 273, 279, 286, 369, 372, 436
Network regulation mechanism 199, 290, 297, 302, 383
Path loss (dB) 185, 275, 448
NH 162
Path loss calculations 77
Noise 177
Path loss matrices 74, 76, 183, 274, 350, 360, 373, 447
Noise correlation matrix 372, 446
path loss matrix 60
Noise figure 125, 126
Path loss matrix resolution 275
Noise figure of the transmitter equipment 528
Peak MAC channel throughput 451, 498, 499, 504
Noise level 247, 346
Peak MAC throughput 460, 515
Noise rise 247, 248, 346
Peak MAC user throughput 505
Noise rise on downlink 221, 283, 289
Percentage of covered area 160
Noise rise on uplink 221, 283, 286, 289
Percentage of delayed circuit switched traffic 155
Noise Rise Scheduling 218
Percentage of max transmitter power used 283
Noise rise scheduling 218
Percentage of maximum transmitter power used. 187, 277
Non correlated MSAs m and n 164
Percentage of traffic covered in the overlapping area 160
Non hopping 162
Permutation zone 434, 451, 497
nrtPS 500, 501
Pilot burst transmitted by the transmitter on carrier ic 285
Number of active users on UL and DL 197, 381, 454
Pilot bust received at terminal from a transmitter on carrier ic 286
Number of calls per hour 291
Pilot power 183, 233
Number of cells in active set 234, 321
Pilot power percentage 279
Number of channel elements available for a site on downlink 186, 277
Pilot quality 193, 234, 242, 320, 321
Number of channel elements available for a site on uplink 186, 277
Pilot quality level 315
Number of channel elements of a site consumed by users on downlink 186, 277
Pilot quality level at terminal on carrier ic 287
Number of channel elements of a site consumed by users on uplink 186, 277
Pilot quality level on uplink 311, 336, 345
Number of dedicated circuit switched timeslots 154
Pilot reception analysis 232, 242, 338
Number of dedicated packet switched timeslots 154
Pilot RSCP activation threshold 235, 236, 239
Number of EVDO channel elements available for a site on uplink and downlink
284 Pilot signal 347
Number of HS-SCCH channels 207 Pilot subcarriers 434
© Forsk 2009 AT271_TRG_E6 543

Pixel size 35 191

Planet 35, 36, 37 Quality level at terminal on pilot for carrier ic 191, 280
Planet format 52 Quality level at transmitter on a traffic channel for a FCH channel on carrier ic
Plot resolution 185, 275 282, 288
PN offset allocation 350 Quality level at transmitter on a traffic channel for a SCH channel on carrier ic
282
PN offset per cell allocation 352
Quality level at transmitter on a traffic channel for carrier ic 192
PN Offset per Cell Strategy 356
Quality margin 339, 340, 342, 344
PN sequence 350
PN-cluster size 351
Point analysis 131, 166, 183, 232, 273, 319, 372
Point index 87 R
Population 37 R99 200
Population density vector maps 36 R99 related simulation results 220
Population maps 43, 45, 46 Radial calculation mode 84, 98
Power Backoff 168, 169, 170, 171, 172 Radial profile extraction 78
Power combination gain 409, 507 Radio data 38
Power control 310, 320 Radio equipment 124
Power control simulation 193, 199, 302, 383 Radio resource management 317
Power control simulations 258, 350 Random carrier selection mode 201, 202, 203, 299, 304, 311
Power pooling between transmitters 301, 309 Random shadowing error drawn during Monte-Carlo simulation 190, 279, 286
Power received at the receiver from the donor transmitter 526 Raster images 43, 45, 46
Power spectral density of thermal noise 436 Raster maps 36
Power units 30 Rayleigh fading 113
Preamble C/N 497 Receive Time Guard 434
Preamble index 435 Received preamble power 451
Predicted path loss 116 Received signal level 158
Prediction studies 76 Received traffic power 451
Presentation of the algorithm 298, 303, 311 Receiver 82
Primary scrambling code allocation 258, 417 Receiver antenna 76
Prime meridian 25 Receiver antenna gain 74, 117, 183, 273, 374, 384, 529, 531, 535
Principal knife-edge 106 Receiver antenna height 83
Priority determination 261, 353 Receiver clearance 91, 94
Probability density function (pdf) 116 Receiver gain 76
Probability of being active on DL 194, 197, 198, 291, 292, 293, 295, 296, 378, Receiver height 76, 94, 96
381, 382, 453, 454 Receiver height in metre 61
Probability of being active on UL 194, 197, 198, 291, 292, 293, 295, 296, 378, Receiver in donor transmitter coverage area 527
381, 382, 453, 454
Receiver in repeater coverage area 527
Probability of being active on UL and DL 194, 197, 198, 291, 292, 293, 295,
296, 378, 381, 382, 453, 454 Receiver location 82
Probability of being connected 194, 291, 378, 452 Receiver loss 529, 531, 535
Probability of being inactive 194, 198, 291, 292, 293, 295, 296, 378, 381, 382, Receiver losses 74, 76, 117, 131, 183, 273, 374, 384
453, 454 Reception tab 183, 274, 373, 447
Probability of Having a Certain RLC Peak Rate 252 REDT 111
Profile extraction 73 Reduction Factor 149, 150, 152, 153
Profile extraction methods 78 Regression line 88, 89
Profile extraction mode 82 Relation between RLC/MAC and application throughputs 173
Profile resolution 87 Reliability level 153
Profile tab 183, 273, 372 Remote Antennas 39
Projection 26, 53 Remote electrical downtilt 109, 111
Projection coordinate system 25, 26 Repeater 38
Projection method codes 29 repeater
Projection parameter indices 30 definition (UMTS) 519
Propagation model 73, 76, 113 Repeater amplifier gain 520, 521, 523, 524, 525, 526, 530, 532, 533
propagation model 60 Repeater coverage side antenna gain 520, 521, 523, 524, 525, 526, 533, 535
Propagation model appendices 104 Repeater EIRP 529, 531, 535
Propagation model parameters 76 Repeater noise figure 528
Propagation models 82 Repeater output global amplification gain 519, 522, 527, 534
Proportional Demand scheduler 502 Repeater’s coverage side antenna gain 530, 532, 533
Proportional Fair scheduler 501 Repeater’s coverage side feeder losses 520, 521, 523, 524, 525, 530, 532,
Pseudo Noise sequence 350 533, 535, 536
Repeater’s donor side antenna azimuth 521, 530
Repeater’s donor side antenna gain 520, 521
Repeater’s donor side antenna’s mechanical Downtilt 521, 530
Q Repeater’s donor side feeder losses 520, 521
QoS class 435 Repeater’s donor-side antenna height/altitude 522, 531
QoS class bias factor 436, 502 Repeater’s donor-side antenna tilt 522, 531
QoS class rank 502 Repeaters and Remote Antennas 519
Quality level 205, 299, 300, 306, 339 Repeater-to-mobile receiver path loss 527
Quality level at terminal for FCH using carrier ic due to combination of all trans- Repeater-to-receiver path loss 519, 523, 529
mitters of the active set 281
Required C/I 327
Quality level at terminal for SCH using carrier ic due to combination of all trans-
Required power 245, 246, 299, 318, 340, 343, 344
mitters of the active set 281
Required power margin 245, 246, 340, 343, 344
Quality level at terminal on a traffic channel from one transmitter for a FCH
channel on carrier ic 281 Required quality 318
Quality level at terminal on a traffic channel from one transmitter for a SCH Required quality level on uplink 336
channel on carrier ic 281 Required rate 327
Quality level at terminal on a traffic channel from one transmitter on carrier ic Required terminal power 236, 238, 239, 328
544 AT271_TRG_E6 © Forsk 2009

Index
Required terminal power on FCH and SCH 331 Shadowing margin calculation in predictions 116
Required terminal power to achieve Eb/Nt target at transmitter for FCH on car- Shadowing model 113
rier ic 282 Signal level 131, 167, 183, 273, 372, 373, 450
Required terminal power to achieve Eb/Nt target at transmitter for SCH on car- Signal level (in dBm, dBµV, dBµV/m) 185, 275, 448
rier ic 282
Signal level at the receiver 74
Required terminal power to achieve Eb/Nt target at transmitter on carrier ic 288
Signal level received from a repeater on a carrier (CDMA) 519, 522, 534
Required terminal power with ACK 335
Signal level received from a repeater on a TRX type (GSM) 528, 532, 535
Required terminal power without ACK 337
Signal level received on a carrier 274
Required transmitter FCH traffic channel power to achieve Eb/Nt target at ter-
Signal level study 74
minal on carrier ic 281
Signal Quality Analysis 449
Required transmitter SCH traffic channel power to achieve Eb/Nt target at ter-
minal on carrier ic 281 Signal strength 113
Required transmitter traffic channel power to achieve Eb/Nt target at terminal Signal to noise ratio 119
on carrier ic 191, 281 Simulation 455
Resolution 74, 80, 81 Simulations 193, 226, 243, 248, 290, 377
Resource control 317 Simulations based on raster traffic and vector traffic maps 193, 290, 378, 452
Restricted to neighbours 290 Simulations based on traffic map per service and per transmitter 197, 293, 381,
Reuse distance 259, 351, 418 454
Rho factor 436 Single colour 242, 243, 245, 338, 339, 342, 343
RLC Peak Rate 238 Site 38
RLC peak rate 250 Site altitude 77
RLC peak throughput 250 Site position 76
RLC/MAC throughput/timeslot study display type 173 Site priority 264, 355
Round trip time 149 Slope at receiver between 0 and distance min 86
rtPS 500, 501 Slow fading 113
Rural area 83 SM 497, 498, 516
Smart Antenna 515
Smart antenna 371, 445
Smoothing angle 111
S Smoothing factor 111
Sample Values for SPM Formula Parameters 92 Soft handoff 118, 121, 205, 238, 240, 312, 318
Sampling factor 434 Soft handover 329, 330, 332, 334, 335, 337
Scanned images 37 Soft handover gain on downlink 191
Scarmbling code allocation examples 423 Soft handover gain on uplink 192
SCH rate 293, 306 Soft/soft handoff 205, 300, 305, 312
SCH rate factor 277, 292, 295 Softer handoff 120, 204, 239, 240, 300, 305, 312, 317
Scheduling algorithms 215 Softer handover 332, 334, 336, 338
Scrambling code allocation order 420 Softer/soft handoff 205, 300, 305, 312
Scrambling code allocation process 260, 352 Softer-soft handoff 239, 240
Scrambling code domain criterion 263, 422, 423 Softer-soft handover 329, 330, 332, 334, 336, 338
Scrambling codes 258, 417 Softer-softer handoff 239, 240
Scrambling codes domains 259, 351 Softer-softer handover 332, 334, 336, 338
Second best signal level and a margin 132, 184, 274, 448 Soft-soft handoff 239, 240
Second best signal level per HCS layer and a margin 161, 166, 176 Soft-soft handover 329, 330, 332, 334, 336, 337
Secondary antenna 127 Space-Time Transmit Diversity 516
Secondary antenna gain 127 Spatial Multiplexing 497, 498, 516
Secondary antenna index 127 Spatial Multiplexing gain factor 497
Secondary antennas 127 Speed of light 105
Secondary knife-edge 106 Spheroid 53
Segment 434 SPM 82, 85, 94, 111
Segmentation 478 SPM calculations 86
Segmentation usage 435, 458 SPM formula 85
Segmenting factor 497 Spot Ht 86
Semi-deterministic clutter 91 Spreading angle 368
Sequential carrier selection mode 201, 202, 299, 304, 311 Spreading Bandwidth 187, 277, 284
Served circuit switched traffic 151, 152, 154, 155 Spreading bandwidth 528
Served packet switched traffic 151, 154, 156 Stability factor 457
Service 200, 234, 239, 244, 246, 298, 303, 338, 339, 342, 343, 449, 459 Standard deviation 116, 117, 118
Service area determination 161, 184, 274 Standard meridian 25
Service coding factor on downlink 187 Standard Propagation Model 80, 85, 111
Service downlink effective bit rate 187, 188 Stanford University Interim model 97
Service downlink nominal bit rate 187, 367 Start frequency 434
Service downlink process gain 188 Station 39
Service priority 301, 302, 308, 315 Steering vector 371, 409, 410, 445, 506, 508
Service processing gain 528 STTD 516
Service uplink effective bit rate 187, 188 Subcarrier 515
Service uplink nominal bit rate 187, 188, 367 Subcell 39
Service uplink process gain 188 Subchannel 515
Services 324, 326, 333 Subchannels 434
SFH 162 Suburban area 83
Shadow fading 98 SUI terrain types 97
Shadowing 113 Supported geographic data formats 37
Shadowing error 118 Symbol 514
Shadowing error pdf (n signals) 118, 121 Symmetric neighbourhood relationship 159, 257, 349
Shadowing error pdf (one signal) 116 Synthesised frequency hopping 162
Shadowing margin 73, 121, 123, 131, 165, 183, 189, 233, 273, 279, 286, 321, Systematic profile extraction 79
369, 372, 374, 384, 436, 519, 523, 529, 531, 534, 536
© Forsk 2009 AT271_TRG_E6 545

T Total power received at transmitter from intra-cellular terminals using carrier ic

192, 221
Take into account Covered Area 158 Total power transmitted by the terminal on carrier ic 280
Take into account Covered Traffic 158 Total received interference at transmitter on carrier ic 192, 221, 281, 283, 288,
Target throughput 499 289
Target traffic overflow rate 151 Total received noise at terminal on carrier ic 191, 280
TBA cells 417 Total received power at terminal from a transmitter on carrier ic 279
TBA transmitters 157 Total received power at terminal from traffic channels of a transmitter on carrier
TBC transmitter 74 ic 279
TBC transmitters 433 Total received power at transmitter from extracell terminals using carrier ic 281,
TBF 149 283, 287, 289
TBF blocking 149 Total received power at transmitter from intracell terminals using carrier ic 281,
283, 287, 289
TCH 39
Total SCH power on carrier ic 278
TCH_INNER 39
Total traffic demand 155
TDD 433, 515
Traffic analysis 37, 147, 151
Terminal 153, 236, 241, 325, 327, 331, 338, 449, 459
Traffic burst received at terminal from a transmitter on carrier ic 286
Terminal FCH power transmitted in carrier ic 278, 285
Traffic burst transmitted by the transmitter on carrier ic 285
Terminal gain 189, 279, 286, 368, 436
Traffic C/N level 450
Terminal loss 189, 279, 286, 368, 436
Traffic channel availibility on downlink 321
Terminal Noise Factor 187, 277, 284, 368
Traffic channel quality 235, 236, 238, 243, 245, 322, 323, 329
Terminal noise figure 436
Traffic channel quality on FCH 324, 325, 327, 328
Terminal power transmitted 189
Traffic channel quality on SCH 324, 325, 327
Terminal power transmitted on carrier ic 285
Traffic data 36
Terminal SCH power transmitted on carrier ic 278
Traffic data channel gain 286
Terrain clearance 95
Traffic demand 151
Terrain profile 87, 89
Traffic density maps 43, 45, 46
Text data 52
Traffic load 147, 149, 151, 155, 156, 163, 166, 435
Text data files 55
Traffic loads 458
Thermal noise 170, 171, 232, 234, 238, 240, 319, 321, 329, 330, 332, 333, 334,
335, 337, 508 Traffic Map Based on Transmitters and Services (# Active Users) 381, 454
Thermal noise at terminal 187, 277, 285, 368 Traffic Map Based on Transmitters and Services (Throughputs) 381, 454
Thermal noise at transmitter 187, 277, 285, 368 Traffic maps 160
Thermal noise ratio 168 Traffic overflow 152
Throughput 171, 172 Traffic quality level at terminal on carrier ic 287
Throughput based on C with ILA 169 Traffic quality target on FCH 330
Throughput Based on C Without ILA 168 Transmit Time Guard 433
Throughput based on interpolation between C/N and C/(I+N) with ILA 172 Transmitter 38, 165
Throughput based on interpolation between C/N and C/(I+N) without ILA 171 Transmitter antenna azimuth 60
Throughput based on worst case between C and C/I with ILA 170 Transmitter antenna gain 74, 183, 273, 372, 374, 384
Throughput based on worst case between C and C/I without ILA 170 Transmitter antenna height 83
Throughput offset 436, 497, 499, 504 Transmitter antenna mechanical tilt 60
Throughput reduction factor 154, 156 Transmitter equipment thermal noise 528
Throughput scaling factor 436, 497, 499, 504 Transmitter gain 189, 279, 285, 368, 435
Throughput/timeslot 173 Transmitter height in metre 61
Tilt 127 Transmitter Identifier 67
Tilt angle 109 Transmitter loss 183, 189, 273, 279, 286, 366, 373, 374, 384, 435
Time-domain overhead 433 Transmitter losses 127
Timeslot capacity 149 Transmitter Noise Factor 187, 277, 284, 435
Timeslot configuration 151 Transmitter power 74, 127
Time-slot scenario 457 Transmitter priority 264, 355
TMA 124 Transmitter radio equipment 124
TMA noise figure 125, 126 Transmitter SCH power for a traffic channel on carrier ic 278
TMA reception gain 125, 126 Transmitter total transmitted power on carrier ic 189, 278
TMA transmission loss 125, 126, 127 Transmitter traffic channel power on carrier ic 188, 189
Total effective interference based on traffic at terminal on carrier ic 289 Transmitter-terminal total loss 190, 279, 286, 369
Total FCH power on carrier ic 278 Transverse Mercator coordinate system 25
Total interference 528 Triangular weighting function 90
Total loss between transmitter and receiver 233 TRX 39
Total losses 131, 183, 273, 372 TRX power offset 529, 531, 535
Total losses (dB) 185, 275, 449 Typical values for losses per clutter class (SPM) 93
Total noise 204, 205, 232, 299, 300, 306, 312, 329, 330, 332, 333, 334, 335,
337
Total noise (I+N) 460 U
Total noise at transmitter on carrier ic 192, 221, 282, 283, 288, 289
UGS 500
Total noise based on pilot received at terminal on carrier ic 287
UL load factor 316
Total noise based on traffic received at terminal on carrier ic 287, 289
UL load percentage 232, 338
Total number of EVDO channel elements of a site consumed by users on uplink
UL macro-diversity gain 115, 121
and downlink 284
UL macro-diversity gain evaluation 118
Total power on a carrier ic 200
UL min noise 201, 202, 203, 298, 303, 311
Total power on carrier ic 298, 303
UL minimum noise 242, 320
Total power received 233
UL quality gain due to signal diversity in soft handoff 190, 279, 286
Total power received at terminal from a transmitter on carrier ic 190
UL SHO gain 337, 338
Total power received at terminal from traffic channels of a transmitter on carrier
ic 190 UL soft handover gain 335
Total power received at transmitter from extra-cellular terminals using carrier ic UMTS 82
192, 221 UMTS HSPA documents 183
546 AT271_TRG_E6 © Forsk 2009

Index
UMTS HSPA Prediction studies 232 WLL calculations 94

Uniform weighting function 90 WLL formula 94
Unit codes 27
Units 30
Units for distances 31 Z
Units for heights 31
Zero mean gaussian random variable 117, 118, 121
Units for offsets 31
Zero-mean unit-variance Gaussian 116
Units for reception power 30
Units for transmission power 30
Universal Transverse Mercator method 25
Update of interference on active mobiles 301
Uplink activity factor on FCH 278
Uplink and downlink FCH nominal rates 291
Uplink and downlink interference update 206
Uplink and downlink interference updates 301, 308
Uplink Average Rate 345
Uplink Data Rate 344
Uplink data traffic quality 333
Uplink effective traffic channel quality 331, 335
Uplink FCH nominal rate 278
Uplink interference updates 314
Uplink load factor 226, 238, 240
Uplink load factor control 207, 302, 309, 314
Uplink load factor on carrier ic 221, 283, 289
Uplink macro-diversity gain 329, 330, 332, 334
Uplink noise rise 435, 458, 515
Uplink power control 204, 299, 304, 312
Uplink power control margin 434
Uplink quality 335, 337
Uplink rake receiver efficiency factor 187, 277, 285
Uplink received powers on carrier ic 200, 298, 303
Uplink reuse efficiency factor on carrier ic 221, 283, 289
Uplink reuse factor on carrier ic 221, 283, 289
Uplink SCH bit rate 278
Uplink service area analysis 245, 342
Uplink service effective bit rate 285
Uplink service processing gain on FCH 278, 285
Uplink service processing gain on SCH 278
Uplink service rate 285
Uplink soft handover gain 240, 331
Uplink soft handover gain FCH and SCH 335
Uplink soft handover gain for FCH channel on carrier ic 282, 288
Uplink soft handover gain for SCH channel on carrier ic 282
Uplink sub-menu 238, 328
Uplink traffic channel quality 238, 329, 331
Urban environment 83
Use a maximum of codes 259, 264, 418
Use a maximum of PN offsets 352
User density traffic maps 37
User multiplexing factor 149
User profile 291
User profile traffic maps 36
User profiles 36
User throughput 515
V
Vector data 37, 52
Vector file sample 55
Vector files 54
Vertical Mapper 51
Vertical pattern 110
Victim and interfering mobiles 456
Visibility and distance between the transmitter and the receiver 86
Voice service users 291, 293
Voice services 452, 454
W
Waiting queue 149
Walsh code management 317
WGS84 ellipsoid 25
WiMAX 82
WLL 82, 94, 106
© Forsk 2009 AT271_TRG_E6 547

548 AT271_TRG_E6 © Forsk 2009

© Forsk 2009 AT271_TRG_E6 549

Release 2.7.1
7 rue des briquetiers – 31700 – Blagnac – France AT271_TRG_E6

Tel: +33 (0)5 62 74 72 10 – Fax: +33 (0)5 62 74 72 11 February 2009
http://www.forsk.com

Technical Reference Guide E6 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Technical Reference Guide E6 PDF

Uploaded by

Copyright:

Available Formats

Atoll

RF Planning and Optimisation Software

Forsk (Head Office) 7 rue des Briquetiers  www.forsk.com Web

Atoll 2.7.1 Technical Reference Guide Release AT271_TRG_E6

© Copyright 1997 - 2009 by Forsk

About Technical Reference Guide

© Forsk 2009 AT271_TRG_E6 iii

iv AT271_TRG_E6 © Forsk 2009

1 Coordinate Systems and Units ....................................................... 25

2 Geographic and Radio Data ........................................................... 35

3 File Formats .................................................................................... 43

© Forsk 2009 AT271_TRG_E6 v

3.1.1 HDR Header File...............................................................................................................................43

vi AT271_TRG_E6 © Forsk 2009

3.19.1 One Histogram per Line (.im0) Format............................................................................................. 64

© Forsk 2009 AT271_TRG_E6 vii

4.4.5.2.1 Free Space Loss....................................................................................................................95

5 GSM GPRS EGPRS Networks......................................................131

viii AT271_TRG_E6 © Forsk 2009

© Forsk 2009 AT271_TRG_E6 ix

5.6.1.1 Service Area Determination ......................................................................................................161

x AT271_TRG_E6 © Forsk 2009

5.8.2.4 HCS Servers and a Margin....................................................................................................... 176

6 UMTS HSPA Networks ................................................................. 183

© Forsk 2009 AT271_TRG_E6 xi

6.4.4.2.1 OVSF Codes Management..................................................................................................226

7 IS-95 cdmaOne and CDMA2000 Networks ...................................273

xii AT271_TRG_E6 © Forsk 2009

7.1.2.2 Reception Tab .......................................................................................................................... 274

© Forsk 2009 AT271_TRG_E6 xiii

7.5.2.3 Uplink Service Area Analysis ....................................................................................................342

xiv AT271_TRG_E6 © Forsk 2009

© Forsk 2009 AT271_TRG_E6 xv

8.8.2.2 Multi Carrier Network.................................................................................................................425

9 WiMAX BWA Networks..................................................................433

xvi AT271_TRG_E6 © Forsk 2009

9.2.6.3.3 Coverage by Best Bearer (DL) Display Types .................................................................... 461

10 Repeaters and Remote Antennas................................................. 519

© Forsk 2009 AT271_TRG_E6 xvii

10.1.1.2.2 Gain Automatic Calculation .................................................................................................523

xviii AT271_TRG_E6 © Forsk 2009

Figure 2.1: Digital Terrain Model.................................................................................................................................. 35

© Forsk 2009 AT271_TRG_E6 xix

xx AT271_TRG_E6 © Forsk 2009

Figure 10.15: Angle from North (Azimuth) ..................................................................................................................... 530

© Forsk 2009 AT271_TRG_E6 xxi

xxii AT271_TRG_E6 © Forsk 2009

24 AT271_TRG_E6 © Forsk 2009

1 Coordinate Systems and Units

1.1.1 Description of Coordinate Systems

1.1.1.1 Geographic Coordinate System

© Forsk 2009 AT271_TRG_E6 25

1.1.1.6 Projection Coordinate System

1.1.2 Coordinate Systems in Atoll

1.1.2.1 Projection Coordinate System

1.1.2.2 Display Coordinate System

1.1.2.3 Internal Coordinate Systems

26 AT271_TRG_E6 © Forsk 2009

1.1.3 File Formats

4230 = "ED50"; 101; 230; 1; "Europe - west"

32045 = "NAD27 / Vermont"; 2; 267; 6, -72.5, 42.5, 500000, 0, 0.9999643; "United

1.1.3.1 Unit Codes

© Forsk 2009 AT271_TRG_E6 27