Professional Documents
Culture Documents
Version 2.7.1
Technical
Reference
Guide
AT271_TRG_E6
Technical Reference Guide
Contact Information
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
The product or brand names mentioned in this document are trademarks or registered trademarks of their respective
registering parties.
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.
Table of Contents
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
4.4.2.3 Calculations in Atoll .................................................................................................................... 85
4.4.3 Standard Propagation Model (SPM) ................................................................................................ 85
4.4.3.1 SPM Path Loss Formula............................................................................................................. 85
4.4.3.2 Calculations in Atoll .................................................................................................................... 86
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 Calculations in Atoll .................................................................................................................... 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
4.4.5.2 Calculations in Atoll .................................................................................................................... 95
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
5.2.1.2.1 Circuit Switched Services.................................................................................................... 136
5.2.1.2.2 Packet Switched Services................................................................................................... 136
5.2.1.3 HCS Layers .............................................................................................................................. 136
5.2.1.3.1 Circuit Switched Services.................................................................................................... 137
5.2.1.3.2 Packet Switched Services................................................................................................... 137
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
5.4.2.1.5 Blocking Probability............................................................................................................. 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
5.4.2.2.5 Blocking Probability............................................................................................................. 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
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
8.2.2.5.1 Coverage Condition .............................................................................................................375
8.2.2.5.2 Coverage Display ................................................................................................................375
8.2.2.6 UpPCH RSCP Coverage Prediction..........................................................................................376
8.2.2.6.1 Coverage Condition .............................................................................................................376
8.2.2.6.2 Coverage Display ................................................................................................................376
8.2.2.7 Baton Handover Coverage Prediction .......................................................................................376
8.2.2.7.1 Coverage Condition .............................................................................................................376
8.2.2.7.2 Coverage Display ................................................................................................................377
8.2.2.8 Scrambling Code Interference Analysis ....................................................................................377
8.3 Monte Carlo Simulations.......................................................................................................................377
8.3.1 Generating a Realistic User Distribution .........................................................................................377
8.3.1.1 Simulations Based on Raster and Vector Traffic Maps ............................................................ 378
8.3.1.1.1 Circuit Switched Service (i) ................................................................................................. 378
8.3.1.1.2 Packet Switched Service (j) ................................................................................................ 379
8.3.1.2 Simulations Based on Traffic Map Based on Transmitters and Services ................................. 381
8.3.1.2.1 Circuit Switched Service (i) ................................................................................................. 381
8.3.1.2.2 Packet Switched Service (j) ................................................................................................ 382
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
List of Figures
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 7.7: Neighbourhood Constraints...................................................................................................................... 354
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 8.14: Neighbourhood Constraints...................................................................................................................... 421
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
Atoll
Atoll
Microwave
RF PlanningLink
Microwave andPlanning
Optimisation Software
Software
Technical Reference Guide
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)
1.1.1.2 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.
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.
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.
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.
Code = "Name of the system"; Unit Code; Datum Code; Projection Method Code,
Projection Parameters; "Comments"
Examples:
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.
1.2 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.
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.
Atoll
Atoll
Microwave
RF PlanningLink
Microwave andPlanning
Optimisation Software
Software
Technical Reference Guide
Therefore, a n*n bin DTM file requires (n)2 points (altitude values).
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.
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.
2.1.1.6 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.
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
Image, Population, Other data
.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
Note:
• The .wld files may be used as georeferencement file for any type of binary raster file.
• Tiled .tif format is not supported.
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.
2.2.1 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.
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.
2.2.8.2 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.
Atoll
Atoll
Microwave
RF PlanningLink
Microwave andPlanning
Optimisation Software
Software
Technical Reference Guide
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)
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.
It can be:
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).
3.1.1.2 Samples
Here, the data is 20m.
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.
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
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.
Name
Start Size Description
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
8 = 8-bit palletized. # of colours = 256
16 = 16-bit palletized. # of colours = 65536
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.
Name
Start Size Description
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).
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.
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.
Atoll supports .bpw and .bmw header file extensions for Import, but exports headers with .bpw file extensions.
3.3.3 Sample
3.3.3.1 Clutter Classes File
100.00
0.00
0.00
-100.00
60000.00
2679900.00
3.4.2 Sample
3.4.2.1 Clutter Classes File
100.00
0.00
0.00
-100.00
60000.00
2679900.00
- 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.
3.12.2 Sample
ncols 303
nrows 321
xllcorner 585300.000000
yllcorner 5615700.000000
cellsize 100.000000
nodata_value 0
...
• 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):
Australian-1965
56
UTM
0 153 500000 10000000
• The index file gives clutter spatial references. The structure of clutter index file is the same as the structure of DTM
index file.
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
3.14.3.2 Sample
Index file associated with the vector files
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:
• The menu file, an ASCII text file, contains the text features. This file is optional.
1 Airport
2 Ferryport
3 Railway_Station
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
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")
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'>
FS = FlagShip D3 = dBaseIII+
Fb = FoxBase D4 = dBaseIV
Fp = FoxPro D5 = dBaseV
CL = Clipper
• Field descriptor array in the .dbf header (32 bytes for each field)
• Field type and size in the .dbf header, field descriptor (1 byte)
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.
FS = FlagShip D3 = dBaseIII+
Fb = FoxBase D4 = dBaseIV
Fp = FoxPro D5 = dBaseV
CL = Clipper
• Field descriptor array in the .dbf header (32 bytes for each field)
• Field type and size in the .dbf header, field descriptor (1 byte)
• 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:
<Column1><tab><Column2><tab><Column3><tab><Column4><newline>
3.19.1.1 Sample
# Calculation Results Data File.
# Version 1.1, Tab separated format. Commented lines start with #.
# 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.
• 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:
<Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline>
Note:
• The columns 1, 2, and 3 must be defined only in the first line of each histogram.
3.19.2.1.2 Sample
# Calculation Results Data File.
# Version 1.1, Tab separated format. Commented lines start with #.
# Remark: C/I results do not incorporate power offset values.
# Fields are:
##------------#------------#------------#-----------#------------------#
#| Interfered | Interfering| Interfered | C/I | Probability |
#| Transmitter| Transmitter| Trx type | Threshold | C/I >= Threshold |
##------------#------------#------------#-----------#------------------#
#
# 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
##---------------------------------------------------------------------#
1 2 TCH_INNER 8 1
9 0.944
10 0.904
11 0.892
14 0.844
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
• The second part provides information about transmitters taken into account in AFP.
The lines after the header are considered as comments if they start with the symbol "#". If not, they must have the following
format:
<Column1><tab><Column2><newline>
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.
# Version 2.1, Tab separated format. Commented lines start with #.
# Fields are:
##-----------#-----------#-----------#-----------#---------#---------#
#|Transmitter|Transmitter|BCCH during|BSIC during|% of vic'|% of int'|
#|Name |Identifier |calculation|calculation|coverage |coverage |
##-----------#-----------#-----------#-----------#---------#---------#
#
# 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 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
• 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:
<Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline>
3.19.3.1 Sample
# Calculation Results Data File.
# Version 1.1, Tab separated format. Commented lines start with #.
# Remark: C/I results do not incorporate power offset values.
# Fields are:
#------------------------------------------------------------------------
#Transmitter Interferer TRX type C/I Probability
#------------------------------------------------------------------------
#
# 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
Site0_2 Site0_1 BCCH,TCH -9 0.996
Site0_2 Site0_1 BCCH,TCH -6 0.976
Site0_2 Site0_1 BCCH,TCH -4 0.964
Site0_2 Site0_1 BCCH,TCH -1 0.936
Site0_2 Site0_1 BCCH,TCH 0 0.932
Site0_2 Site0_1 BCCH,TCH 1 0.924
Site0_2 Site0_1 BCCH,TCH 4 0.896
Site0_2 Site0_1 BCCH,TCH 7 0.864
Site0_2 Site0_1 BCCH,TCH 8 0.848
Site0_2 Site0_1 BCCH,TCH 9 0.832
Site0_2 Site0_1 BCCH,TCH 10 0.824
...
<Column1><SEP><Column2><SEP><Column3><SEP><Column4><newline>
C I req corresponds to the required C/I threshold. This parameter is defined for each subcell.
3.19.4.1 Sample
# Calculation Results Data File.
# Version 1.1, Tab separated format. Commented lines start with #.
# Remark: C/I results do not incorporate power offset values.
# Fields are:
#------------------------------------------------------------------------
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
Atoll
Atoll
Microwave
RF PlanningLink
Microwave andPlanning
Optimisation Software
Software
Technical Reference Guide
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:
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.
1st step: First of all, Atoll calculates the path loss ( L path ), using the selected propagation model.
L model is the loss on the transmitter-receiver path calculated through the propagation model. L model value depends on
the selected propagation model.
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.
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):
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,
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 .
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.
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.
Computation zone(s)
Rectangle containing the computation zone(s)
Calculation area defined (square)
Transmitter
Calculation area: real area for which Atoll calculates path losses
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)
Receiver antenna Valid because L ant = 0 Prediction study Sufficient Not necessary
Rx
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: 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.
3rd step: Atoll performs a second linear interpolation to evaluate the S altitude.
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.
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
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.
Transmitter
Geographic profiles
Receiver: it may be anywhere in point analysis or at the centre of each calculation bin in coverage studies
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.
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.
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
AT271_TRG_E6
Profile
Radial
extraction - - Radial Radial Radial Radial Radial
Systematic
mode
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
Chapter 4: Calculations
Notes:
• In formulas described above, L model is stated in dB.
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
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
Environment a(Hr)
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.
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
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).
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)
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
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
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
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
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
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)
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 = HTx
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.
The transmitter antenna height is calculated using the ground slope at receiver.
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.
Spot Ht
Absolute Spot Ht
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.
Atoll offers a new method called “Enhanced slope at receiver” to evaluate the effective transmitter antenna height.
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,
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 orig is the original height. Original terrain height is determined from extracted ground profile.
H orig i – H orig i – 1
1st case: If H orig i H orig i – 1 and ------------------------------------------------------ 0.05 ,
Res
H orig i – H orig i – 1
2nd case: If H orig i H orig i – 1 and ------------------------------------------------------ 0.05
Res
H orig i – H orig i + 1
1st case: If H orig i H orig i + 1 and ------------------------------------------------------- 0.05 ,
Res
H orig i – H orig i + 1
2nd case: If H orig i H orig i + 1 and ------------------------------------------------------- 0.05
Res
Then, for every point of profile, Atoll compares the two filtered heights and chooses the higher one.
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:
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.
d 20 1 – H Txeff – 20
where, K lowant = --------- – 0.3 H Txeff – 20 – ------------------------------------------------------------------------------
5 d - 6.93 + ------------ d -
10 9.63 + ------------
1000 1000
where,
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.
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.
If 0 h 20m , K h = 0
2
Else K h = 7.73 log h – 15.29 log h + 6.746
2 H 0Rx + H Rx – regr i Rx
Else K hf = – 2 – 1.616 log h + 14.75 log h – 11.21 ------------------------------------------------------------
h
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.
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.
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.
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).
• 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)
• Even with no clearance, the clutter height (extracted either from clutter class or clutter
height folders) is never considered at the last profile point.
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.
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:
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:
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.
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.
4.4.4.2.2 Diffraction
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.
• 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.
4.4.5.2.2 Diffraction
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.
where,
Cn is the field strength received in dBV/m,
Cn Calculation
First of all, Atoll evaluates the effective transmitter antenna height, H Txeff , as follows:
where,
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.
Therefore,
If HTxeff < 37.5
AHRxeff Calculation
c H Rx
AH = --- 20 log ----------
Rxeff 6 10
where,
HRx is the user-defined receiver height,
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.
d
PL = A + 10 a H BS Log 10 ------
d 0
4d 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
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
4d 0 d
PL = 20 Log 10 ------------- + 10 a H BS Log 10 ------ + a f – a H R (1)
d 0
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,
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.
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:
4d 4d
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,
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.
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.
• 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 meth-
od.
b. It determines the largest obstacle along the profile in accordance with the Deygout method and evaluates loss-
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.
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.
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.
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.
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.
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.
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.
• 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.
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:
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
- 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 ,
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
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, .
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 ak
tive Earth radius factor for mean refractivity conditions.
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 .
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.
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
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
10
If R' 10 m , C Receiver is reduced by 3.2 + 6.2 Log f Log ------ .
R'
h2
C Receiver = 3.2 + 6.2 Log f Log ------
10
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
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:
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 .
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
• : 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.
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
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:
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:
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:
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.
c0 n d1 d2
R = ---------------------------------
-
f d1 + d2
where,
n is the Fresnel zone index,
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.
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.
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
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
Therefore, we have
DiffractionLoss = J P + J t + J r
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
Therefore, we have
DiffractionLoss = J h
Let’s take M a measurement value and P i the path loss value at point i, before any correction.
1
E = --- ei
n
i
where:
Pi = Pi +E
new old
Ri = 1 – Di M – g – Pi so R i = 1 – D i M – g – P i + E
new old
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:
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).
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:
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:
We get,
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:
1
az = atan ------------------------------------------------------------------------------------------------
cos e Tx sin e Tx tan e Rx
-------------------------------------- + --------------------------------------------------
tan a Rx – a Tx sin a Rx – a Tx
and
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:
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.
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.
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.
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):
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.
A 2 – A 1
- i – 1
A i Smoothed = A i – F Smoothing A i – A + -----------------------
1
2 – 1
Where,
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.
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
[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
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.
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.
• 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+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.
M Shadowing – Eb Nt ) are added to Ec/I0 or Eb/Nt. This shadowing margin is calculated for a given cell edge
UL
• 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
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 or interference+noise
predictions as explained below.
• Signal Level-Based Predictions
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.
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+noise-Based Predictions
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.
M Shadowing – Eb Nt ) are added to Eb/Nt. This shadowing margin is calculated for a given cell edge coverage
UL
• 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.
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 or C/(I+N) 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+noise-Based Predictions
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.
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.
• 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.
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
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
2
x
– ------
1 z
dx = Q ---------
2
P L x z = ----------- e
2 dB
z-
---------
dB
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:
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.
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.
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.
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 .
2 2 2
model = L + P
2
L
= ----------------
-
2
model
Therefore,
2 2
P = model 1 –
2 2
L = model
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.
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.
2
2 = L + P for the link 2
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
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.
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
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
1 2signals 2 2signals 2
= P L P P M Shadowing – Eb Nt – L P P M Shadowing – Eb Nt – 1 – L
L P UL P UL
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
1 2signals 2 2signals 2
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
RL M Shadowing – Eb Nt
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
RL M Shadowing – Eb Nt
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
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
RL M Shadowing – Eb Nt
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.
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
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.
Therefore, in case of two links, we have:
1
1 = L + P for the link 1
2
2 = L + P for the link 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
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
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
1 2signals 2 2signals 2
= 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
1 2signals 2 2signals 2
= P L P P M Shadowing – Ec Io – L P P M Shadowing – Ec Io – 1 – L
L P P
noMRC 2signals
RL M Shadowing – Ec Io
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
RL M Shadowing – Ec Io
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
RL M Shadowing – Ec Io
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
2
1 =1 dB
2
1 =5 dB
2
1 =10 dB
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)
For further information about determination of the correlation coefficient, please see "Correlation Coefficient Determina-
tion" on page 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.
NF TX = NF BTS
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),
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,
Rp k
G amp is the repeater amplification gain (repeater property),
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
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.
And,
WithoutTMA
NF Composite = NF BTS + NF Feeder (in dB)
where,
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.
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).
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 loss),
DL
L Misc are the miscellaneous transmission losses,
DL
L BTS – Conf are the losses due to BTS configuration (BTS property).
NF TX = NF BTS
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),
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.
WithoutTMA
And NF Composite = NF BTS + NF Feeder (in dB)
where,
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.
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).
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),
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.
Atoll
RF Planning and Optimisation Software
Technical Reference Guide
Txi Txi
Path loss ( L path ) L path = L model + L ant
Tx
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,
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
“Indoor coverage” is selected,
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).
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.
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.
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.
Note:
• The minimum threshold is either globally defined or specifically for each subcell (subcell
reception threshold)
And
ji
And
Txi nd
P rec ic 2 Best P Txj ic – M
rec
ji
And
ji
And
ji
Txi
The received P rec tt exceeds the reception threshold defined per HCS layer
And
ji
And
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.
5.1.3.1.7 Second Best Signal Level per HCS Layer and a Margin
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
And
Txi nd
P rec BCCH 2 Best P Txj BCCH – M
rec
ji
C2 = C1 + CELL_RESELECT_OFFSET
where CELL_RESELECT_OFFSET is the reselection value (in dB) defined for at the transmitter level.
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
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.
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.
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.
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.
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).
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 minimum threshold.
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).
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.
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
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.
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.
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.
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.
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:
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.
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.
Where O max Txi,TCH_INNER is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell.
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.
Where O max Txi,TCH_INNER is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell.
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.
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 upm – 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 upm – 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.
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 .
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
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
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 .
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
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.
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).
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.
For each packet switched service, p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell (Txi, TCH) 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.
S Txi,TCH_INNER
D p t m Txi,TCH_INNER = ----------------------------------------------------- T p Txi TCH
S Txi TCH
and
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.
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
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.
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
S 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.
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 .
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
S Txi,TCH_INNER
The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S3 proportional to R3.
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
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_INNER) service area.
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
S 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 .
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
S Txi,TCH_INNER
The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S’3 proportional to R’3.
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.
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.
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.
5.3.1.2.1 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
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
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:
PX= n
RF X ----------------------
PX= 0
-
n=0
Or,
PX= n
RF = X ------------------------------
-
PX= 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:
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
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.
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.
PX= n for n = M N P + 1
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
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.
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.
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
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
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.
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:
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
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
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.
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
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.
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:
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.
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
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
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
ST C = TD C 1 – O C
TD T = TD C + TD P
TS C dedicated + TS P dedicated + TS S – ST C
- 100
O P = 1 – ----------------------------------------------------------------------------------------------------------------------------
TD P
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).
ST P = TD P 1 – O P
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.
TD T
Traffic Load = -----------------------------------------------------------------------------------------------
-
TS C dedicated + TS P dedicated + TS S
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.
ST P = TD P
• 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).
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.
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%
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.
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.
Note:
• For information on the common prediction studies (like coverage by transmitter, profile
study, …), please, refer to Common prediction studies part.
and
Txi Txj
P rec BCCH Best P rec BCCH – M
ji
where,
M is the specified margin (dB).
Best function: considers the highest value.
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
Minimum threshold P rec tt Maximum threshold
and
Txi Txj
P rec BCCH Best P rec BCCH – M
ji
and
Txi belongs to the HCS layer with the highest priority
where,
M is the specified margin (dB).
Best function: considers the highest value.
5.6.1.1.4 Second Best Signal Level per HCS Layer and a Margin
For each HCS layer, k, the service area of Txi corresponds to the bins where:
Txi
Minimum threshold P rec tt Maximum threshold
and
Txi nd Txj
P rec BCCH 2 Best P rec BCCH – M
ji
where,
M is the specified margin (dB).
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:
In case of BBH, assuming TRXs belong to the same TRX type, we have:
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.
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,
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
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
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.
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:
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
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
Note:
• Only the carrier power level is downgraded by the shadowing margin. The interference
level is not altered.
C C
Minimum threshold ---- Maximum threshold or Minimum threshold ------------ Maximum threshold
I v I + N v
TRX j TRX j
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.
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.
5.6.1.4.4 Transmitter
Atoll colours each bin of each transmitter coverage area. The bin colour corresponds to the transmitter colour. Coverage
consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as
interfered transmitters.
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.
5.7.1.3 Second 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 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:
• If the margin equals 0, the coverage area of Txi corresponds to the bins where
Txi – EGPRS
P rec BCCH is the second highest.
Notes:
• If the margin equals 0, the coverage area of Txi corresponds to the bins where
Txi – EGPRS
P rec BCCH is the highest.
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
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.
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
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.
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, TP = Highest TP=f C=P rec TRX - P Backoff 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:
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
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
And
Txi – EGPRS Txi – EGPRS
P rec TRX – P Backoff TRX C
- ---- Threshold
----------------------------------------------------------------------------------------------------------
I I CS
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
schemes from among the coding schemes available in the GPRS/EGPRS equipment, such that:
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
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, TP = Highest TP=f C=P rec TRX - P Backoff TRX
And
Txi – EGPRS Txi – EGPRS
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
schemes from among the coding schemes available in the GPRS/EGPRS equipment, such that:
For each TRX type, tt, cs C N = Lowest CS
Txi – EGPRS
P rec TRX
C-
----------------------------------------------------- ----------- Threshold
N I + N CS
And, cs C I + N = Lowest CS
Txi – EGPRS
P rec TRX
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
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
P rec TRX – P Backoff TRX C
For each TRX type, tt, ----------------------------------------------------------------------------------------------------------- ------------ Threshold
N I + N CS
And
Txi – EGPRS Txi – EGPRS
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.
TP = TP C N + 1 – TP C I + N
Where,
pN
= ---------------------
pI + 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
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
schemes from among the coding schemes available in the GPRS/EGPRS equipment, such that:
For each TRX type, tt, cs C N = Highest CS Txi – EGPRS
C P rec
TP = Highest TP=f ------------ = ----------------------------------- TRX
I + N N
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
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
P rec TRX – P Backoff TRX C
For each TRX type, tt, ----------------------------------------------------------------------------------------------------------- ------------ Threshold
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.
TP = TP C N + 1 – TP C I + N
Where,
pN
= ---------------------
pI + 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
SF
TP Application = TP RLC MAC ---------- – TP Offset
100
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.
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.
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)
• The maximum range option (available in the System tab of the Predictions property dialog) is inactive.
5.8.2.3 Second 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
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:
• If the margin equals 0, the coverage area of Txi corresponds to the bins where
Txi
P rec BCCH is the second highest.
Notes:
• If the margin equals 0, the coverage area of Txi corresponds to the bins where
Txi
P rec BCCH is the highest.
Note:
• 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.
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.
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 .
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:
• 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.
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
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
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
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
Atoll
RF Planning and Optimisation Software
Technical Reference Guide
Total losses ( L total ) in dBm L total = L path + L Tx + L term + L indoor + M Shadowing – model – G Tx + G term
where,
EIRP is the effective isotropic radiated power of the transmitter,
ic is a carrier number,
L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model,
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
“Indoor coverage” is selected,
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.
• For a selected transmitter, it is also possible to study the path loss, L path , or the total
losses, L total . Path loss and total losses are the same on any carrier.
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.
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.
Note:
• For a selected transmitter, it is also possible to study the path loss, L path , or the total
losses, L total . Path loss and total losses are the same on any carrier.
• 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.
And
ji
And
Txi nd
P rec ic 2 Best P Txj ic – M
rec
ji
Notes:
• If the margin equals 0 dB, Atoll will consider bins where the signal level received from Txi is
the second highest.
• If the margin is set to 2 dB, Atoll will consider bins where the signal level received from Txi
is either the second highest or 2dB lower than the second 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 3rd best servers.
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).
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.
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.
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
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.
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
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.
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where other service
areas overlap the studied one, Atoll determines the best transmitter and evaluates path loss from the best transmitter. A
pixel of a service 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 minimum threshold.
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.
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
fulfils signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability in the
explorer.
6.2.1 Inputs
This table lists simulation and prediction inputs (calculation options, quality targets, active set management conditions,
etc.).
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
req
Q pilot E
-----c- Mobility parameter None
Ec/Io target on downlink for the best
I 0 threshold server
E
DL
DL
Q req -----b- (Bearer, Mobility) parameter None Eb/Nt target on downlink
N t req
E
UL
UL
Q req -----b- (Bearer, Mobility) parameter None Eb/Nt target on uplink
N t req
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
DL
W -
---------------------
Gp DL None Service downlink processing gain
R nominal
W
Gp
UL ---------------------- None Service uplink processing gain
UL
R nominal
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
min
P term Terminal parameter W Minimum terminal power allowed
max
P term Terminal parameter W Maximum terminal power allowed
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
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
standard deviation Only used in prediction studies
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
UL npaths
G macro – diversity = M Shadowing – Eb Nt – M Shadowing – Eb Nt
UL UL UL UL quality gain due to signal diversity
G macro – diversity None
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
P pilot Txi ic
P c Txi ic ----------------------------------- W Chip power received at terminal
LT
UL
P term – R99 Bit power received at transmitter on
P b – R99 ic --------------------------- W
LT carrier ic used by terminal
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
the shadowing model ( M Shadowing – Eb Nt or M Shadowing – Ec Io ) while extra-cell interference level is not. Therefore,
DL
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
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
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.
Pb
UL
ic adj
UL Uplink inter-carrier interference at
I inter – carrier txi ic term W
txj j
terminal on carrier ic
--------------------------------------
-
RF ic ic adj
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
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
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.
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.
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
X = SD
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.
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.
N call d
p o = ---------------------
-
3600
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.
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.
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 ,
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.
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.
Note:
• 1kBytes = 1024Bytes.
UL DL
Where R average and R average are the uplink and downlink average requested rates defined for the service j.
UL N sess UL DL N sess DL
p Connection = -------------- D Connection and p Connection = -------------- D Connection
3600 3600
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:
• 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:
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
• 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
• 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
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
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.
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).
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
Probability of being inactive both on UL and DL: p inactive = 1 – f act 1 – 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.
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
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.
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.
If N UL N DL
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.
Initialisation
R99 part
HSDPA part
Mobile Scheduling
HSUPA part
Admission Control
Convergence Study
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
If a given carrier is specified for the service requested by Mb and if it is used by txi
EndFor
UL
BestCarrier k txi M b is the carrier with the lowest X k txi ic
Calculation of
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”
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
N tot txi ic
EndFor
UL
BestCarrier k txi M b is the carrier with the lowest X k txi ic
Calculation of
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, 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 .
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.
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
BestCell k Ni M b is the transmitter associated to the carrier specified for the service
EndFor
UL
BestCell k Ni M b is the cell with the lowest X k txi ic
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
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
EndFor
R99 – req
Calculation of the terminal power required by Mb to obtain the R99 radio bearer: P term M b ic k
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
End For
If (Mb is in not in handoff)
UL UL
Q k M b = Q tch txi M b ic k
UL UL UL
Q k M b = f rake efficiency Q tch txi M b ic k
txi ActiveSet
UL UL UL
Q k M b = Max Q tch txi M b ic k G macro – diversity 2 links
txi ActiveSet
UL UL UL
Q k M b = Max Q tch txi M b ic k G macro – diversity 3 links
txi ActiveSet
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
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 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
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 = ----------------------------------------------------
-
L T txi M b
DL
DL BTS P b txi M b ic k
- G DL
Q tch txi M b ic k = -----------------------------------------------------------------------------------------------------------------------------------
DL
p Service M b G div
DL DL
N tot ic – 1 – F ortho BTS P b txi M b ic k – 1
End For
DL DL DL
Q k M b = f rake efficiency Q tch txi M b ic k
txi ActiveSet
Do
For each cell (txi,ic) in Mb active set
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
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.
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 = ---------------------------------------------------
-
L T txi M b
DL
DL BTS P b txi M b ic
- G DL
Q tch txi M b ic k = --------------------------------------------------------------------------------------------------------------------------
DL
p Service M b G div
DL DL
N tot ic – 1 – F ortho BTS P b txi M b ic
DL DL DL
If the user is inactive, 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
Q k M b = f rake efficiency Q tch txi M b ic k
txi ActiveSet
DL DL
While Q k M b Q req Service M b Mobility M b and Mb active set is not empty
DL DL
If R nominal M b R max txi ic then Mb cannot be connected
Endif
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)
Rejection of the mobile with the lowest service priority starting from the last admitted
EndFor
For each cell (txi,ic)
Codes Codes
While N txi ic k N max txi ic
Rejection of the mobile with the lowest service priority starting from the last admitted
EndFor
For each site (Node B) Ni
CE – DL CE – DL
While N N i k N max Ni
Rejection of the mobile with the lowest service priority starting from the last admitted
CE – UL CE – UL
While N N i k N max Ni
Rejection of the mobile with the lowest service priority starting from the last admitted
EndFor
UL UL
For each cell (txi,ic) with X txi ic X max
Rejection of the mobile with the lowest service priority starting from the last admitted
EndFor
UL UL
While at least one cell with X txi ic X max exists
• 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.
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
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.
• 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".
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
I extra ic = P tot ic
txj j i
Ptot icadj
DL
DL txj j
I inter – carrier ic = --------------------------------------
-
RF ic ic adj
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
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:
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
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.
Therefore, we have:
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
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
N tot ic = I intra ic + I extra ic + I inter – carrier ic + N 0
DL DL
I extra ic = P tot ic
txj j i
Ptot icadj
DL
DL j
I inter – carrier ic = txj
--------------------------------------
-
RF ic ic adj
RF ic ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
P HS – SCCH ic
P c ic = --------------------------------------
-
i LT
i
and
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
req
Ec
------- ic DL
N tot ic
Nt HS – SCCH
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
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
= ------------------------------------ for the total noise option,
Nt HS – PDSCH DL
N tot ic
And
BTS P c ic
Ec
------- ic i
- for the without useful signal option.
= --------------------------------------------------------------------------------------------------------------------------------------
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
N tot ic = I intra ic + I extra ic + I inter – carrier ic + N 0
DL DL
I extra ic = P tot ic
txj j i
Ptot icadj
DL
DL txj j
I inter – carrier ic = --------------------------------------
-
RF ic ic adj
RF ic ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
P HS – PDSCH ic
P c ic = -----------------------------------------
-
i LT
i
And
term term
BTS , F ortho , F MUD 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:
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
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
• 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.
When the option “CQI based on HS-PDSCH quality” is selected, Atoll proceeds as follows.
1. HS-PDSCH Quality Calculation
Atoll proceeds as follows:
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.
Therefore, we have:
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
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
N tot ic = I intra ic + I extra ic + I inter – carrier ic + N 0
DL DL
I extra ic = P tot ic
txj j i
Ptot icadj
DL
DL txj j
I inter – carrier ic = --------------------------------------
-
RF ic ic adj
RF ic ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
P HS – SCCH ic
P c ic = --------------------------------------
-
i LT
i
And
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
req
Ec
------- ic DL
N tot ic
Nt HS – SCCH
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
Ec
Let us assume the following notation: ------- ic corresponds to the HS-PDSCH quality.
Nt HS – PDSCH
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
= ------------------------------------ for the total noise option,
Nt HS – PDSCH DL
N tot ic
And
BTS P c ic
Ec
------- ic i
- for the without useful signal option.
= --------------------------------------------------------------------------------------------------------------------------------------
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
N tot ic = I intra ic + I extra ic + I inter – carrier ic + N 0
DL DL
I extra ic = P tot ic
txj j i
Ptot icadj
DL
DL j
I inter – carrier ic = txj
--------------------------------------
-
RF ic ic adj
RF ic ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
P HS – PDSCH ic
P c ic = -----------------------------------------
-
i LT
i
And
term term
BTS , F ortho , F MUD 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:
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.
• Round Robin: HSDPA users are taken into account in the same order than the one in the simulation (random
order).
• 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.
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
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:
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
Nt E – DPDCH
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
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.
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.
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:
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
Nt E – DPDCH
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 ------- .
Nt E – DPDCH
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
Nt E – DPDCH
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
Nt E – DPDCH
Ec req
( ------- ). It is determined as follows:
Nt E – DPDCH
UL UL UL
X user unused = X user – X user req
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
).
Nt E – DPDCH
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.
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
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.
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.
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.
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
Pb
UL
ic adj
UL Uplink inter-carrier interference at
I inter – carrier txi ic term W
txj j
terminal on carrier ic
--------------------------------------
-
RF ic ic adj
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
I tot txi ic
F
UL
txi ic ------------------------------------------------------------------------------------------- None Cell uplink reuse factor on carrier ic
UL intra Tx
I tot txi ic 1 – F MUD term
DL
DL
I tot ic
F txi ic -------------------------------
- None Downlink reuse factor on a carrier ic
DL
I intra txi ic
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:
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.
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
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."
UL
P term = P term – R99 f act – EDPCCH + P term – HSUPA for HSUPA users
And
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.).
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.
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.
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.
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:
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
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
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.
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.
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.
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
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
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
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:
Codes – HS – PDSCH
• 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
Codes – HS – PDSCH
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
Codes – HS – PDSCH
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.
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
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.
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
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.
We have:
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
nonOrtho term
F ortho BTS P CCH ic r + N 0 LT r
ortho nonOrtho
P CCH ic + P CCH ic + ----------------------------------------------------------------------------------------------------------------------------
-
1
tch ---------------------- + 1 – F ortho BTS
DL CI req r
P tx ic = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
I extra ic + I inter – carrier ic L T r
------------------------------------------------------------------------------------------------- + 1 – F ortho BTS r
DL
P tx ic
------------------------------------------------------------------------------------------------------------------------------------------------------------
1–
1
---------------------- + 1 – F ortho BTS
tch CI req r
The downlink load factor represents the signal degradation in relation to the reference interference (thermal noise plus
synchronisation channel power).
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).
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
UL W P b k req
Q req k = -----------------------------
UL
- -------------------------------------------------------------------------------
UL UL tx
-
R nominal k I intra F – P b k req + N 0
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
------
Nt E – DPDCH
I intra = ------------------------------------------------------------------------------------------------
1
UL
1–F --------------------------------------------------------------
K 1
---------------------------------------------
req
- + 1
Ec - k
------
Nt E – DPDCH
tx UL
N0 F
I intra = -----------------------------------------------------------------------------------------------
-
1
-------------------------------------------------------------------------------------- – 1
1
UL
F --------------------------------------------------------------
K 1
---------------------------------------------
req
- + 1
Ec - k
------
Nt E – DPDCH
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
------
Nt E – DPDCH
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
------
Nt E – DPDCH
UL
UL W P b k req
Q req k = -----------------------------
UL
- -----------------------------------------------------------------------------------
-
tx
R nominal k I intra + I extra + I inter – carrier + N 0
UL
UL W P b k req
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
Nt E – DPDCH
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
Nt E – DPDCH
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
Nt E – DPDCH
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.
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
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
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).
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.
P pilot i ic
P c i ic = ---------------------------
-
LT
I
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 = -------------------------------------------
RF ic ic adj
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.
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 .
Resulting DL
Q pilot ic = G macro – diversity max Q pilot i ic
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:
This is a user-specified input in the Terminal properties. It corresponds to the active set size.
Thermal Noise
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.
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.
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):
Resulting CM – activation
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 .
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
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.
Ptot j icadj
DL
DL txj j
I inter – carrier ic BS = -------------------------------------------
RF ic BS ic adj
RF ic BS ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
Qmax k icBS
DL DL DL
Q MAX ic BS = f rake efficiency
k
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 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):
Resulting CM – activation
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
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
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
Ptot j icBS
DL DL
I extra ic BS =
j j k
Ptot j icadj
DL
DL txj j
I inter – carrier ic BS = -------------------------------------------
RF ic BS ic adj
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.
Qmax k icBS
DL DL DL
Q MAX ic BS = f rake efficiency
k
Where
DL
f rake efficiency is the downlink rake efficiency factor defined in Terminal properties.
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).
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
= ------------------------------------ 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
N tot ic = I intra ic + I extra ic + I inter – carrier ic + N 0
DL DL
I extra ic = P tot ic
txj j i
Ptot icadj
DL
DL txj j
I inter – carrier ic = --------------------------------------
-
RF ic ic adj
RF ic ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
P HS – SCCH ic
P c ic = --------------------------------------
-
i LT
i
And
term term
BTS , F ortho , F MUD and N 0 are defined in "Inputs" on page 186.
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".
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.
max
Max terminal power ( P term ) is an input user-defined for each terminal. It corresponds to the terminal’s maximum 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
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
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
max Q max k ic BS corresponds to the highest Q max k icBS value.
UL
G macro – diversity 3 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.
For softer and softer-softer handoffs (1/2 and 1/3):
Qmax k icBS
UL UL UL
Q MAX ic BS = f rake efficiency
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
Q MAX ic BS = G macro – diversity 2 links max Q max k ic BS
req
3rd step: P term – R99 calculation
req
P term – R99 is the required terminal power.
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 is operated when:
- A mobile supporting compressed mode is connected to a cell located on a site with a
compressed-mode-capable equipment, and
- The received Ec/Io is lower than the Ec/Io activation threshold (Global parameters):
Resulting CM – activation
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
req Q req Q req
- 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
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 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
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
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
max Q max k ic BS corresponds to the highest Q max k ic BS value.
UL
G macro – diversity 3 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.
For softer and softer-softer handoffs (1/2 and 1/3):
Qmax k icBS
UL UL UL
Q MAX ic BS = f rake efficiency
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
Q MAX ic BS = G macro – diversity 2 links max Q max k ic BS
Effective Eb/Nt
UL
Q eff is the effective traffic channel quality at the transmitter on icBS.
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).
UL
G SHO corresponds to the uplink soft handover gain.
UL
UL Q MAX ic BS
G SHO = -----------------------------------------------------
-
UL
max Q max k ic BS
UL UL
max Q max k ic BS corresponds to the highest Q max k ic BS value.
• 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.
Ec req
It corresponds to the E-DPDCH Ec/Nt required to obtain the HSUPA bearer ( ------- ). This value is defined for a
Nt E – DPDCH
HSUPA bearer ( Index HSUPABearer ) and a certain number of retransmissions ( N Rtx ) in the HSUPA Bearer Selection
table.
Ec req
From -------
req
, Atoll calculates the terminal power required to obtain the HSUPA bearer, P term – HSUPA .
Nt E – DPDCH
Ec req
P term – HSUPA = -------
req UL
L T N tot
Nt E – DPDCH
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
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.
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.
Nt E – DPDCH
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:
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 finds 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.
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.
Resulting
bility, Q pilot ic BS .
Atoll displays the best pilot quality received with a fixed cell edge coverage probability.
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 .
Atoll displays the best pilot quality received with a fixed cell edge coverage probability.
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.
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.
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
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.
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.
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.
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
assigned a colour and displayed with intersections between layers.
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.
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.
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.
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
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.
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
user-defined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if
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.
Coverage consists of several layers with a layer per user-defined cell edge coverage probability, p, defined in the Display
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.
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
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.
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
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).
Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction prop-
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
compressed mode is activated). Each layer is assigned a colour and displayed with intersections between layers.
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.
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
req Q req Q req
- 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
displayed with intersections between layers.
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
intersections between layers.
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.
UL UL UL UL UL
Atoll displays a coverage if Q MAX ic Q req (or Q MAX ic Q req Q req if compressed mode is activated). Coverage
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
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 server transmitter.
In this case, receiver is not completely defined and no mobility is assigned. Coverage consists of several layers with a
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
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.
In this case, receiver is not completely defined and no service is assigned. Coverage consists of several layers with a layer
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
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.
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
user-defined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if
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.
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
UL
properties). For each layer, area is covered if Q MAX ic Threshold . Each layer is assigned a colour and displayed with
intersections between layers.
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
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).
Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction prop-
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
compressed mode is activated). Each layer is assigned a colour and displayed with intersections between layers.
Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction
req
properties). For each layer, area is covered if P term – R99 ic Threshold . Each layer is assigned a colour and displayed
with intersections between layers.
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 term – R99 ic – P term M arg in . Each layer is assigned a colour and displayed
with intersections between layers.
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-
sections between layers.
Ptot icadj
DL
Ptot ic + txj--------------------------------------
DL DL j term
N tot ic = - + N0
RF ic ic adj
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
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
intersections between layers.
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 maxN tot ic Threshold . Each layer is assigned a colour and displayed with
ic
intersections between layers.
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 averageN tot ic Threshold . Each layer is assigned a colour and displayed
ic
with intersections between layers.
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.
Atoll displays bins where maxNR 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.
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.
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 N tot ic Threshold . Each layer is assigned a colour and displayed with inter-
sections between layers.
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.
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-
Nt HS – PDSCH
sections between layers.
• 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.
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
T TTI is the TTI duration, i.e. 2 10 s (2000 TTI in one second). This value is specified by the 3GPP.
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.
TTI 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
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
with intersections between layers.
• 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,
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 finds the corresponding BLER.
TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal user
equipment category properties.
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
displayed with intersections between layers.
• 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:
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 finds 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.
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
with intersections between layers.
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,
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
assigned a colour and displayed with intersections between layers.
• 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
assigned a colour and displayed with intersections between layers.
• 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.
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
intersections between layers.
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
Nt E – DPDCH
assigned a colour and displayed with intersections between layers.
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.
UL
Atoll displays the MAC rate ( R MAC ) provided on each pixel. The MAC rate is calculated as follows:
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.
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.
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,
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 finds the corresponding BLER.
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
displayed with intersections between layers.
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:
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 finds 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.
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
with intersections between layers.
• 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.
Force neighbour symmetry: This option enables user to force the reciprocity of a neighbourhood link. Therefore, if the refer-
ence cell is a candidate neighbour of another cell, this one 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.
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.
• 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 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 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 signal received from the cell B is greater than the minimum pilot signal level.
- The pilot quality from B exceeds a user-definable minimum value (minimum Ec/I0).
- The pilot quality from B is the highest one.
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 signal received from the cell A is greater than the minimum pilot signal level.
- 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 signal received from the cell B is greater than the minimum pilot signal level.
- The pilot quality from B exceeds a user-definable minimum value (minimum Ec/I0).
- The pilot quality from B is the highest one.
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.
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.
The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see
after).
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 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.
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 %
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
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 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
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 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 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).
• 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 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.
• 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 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.
Note:
• If no focus zone exists in the .atl document, Atoll takes into account the computation zone.
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.
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.
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.
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.
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
"Cell Priority" on page 261.
All sites which have constraints with the studied site are referred to as near sites.
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
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.
For information on calculating site priority, see "Site Priority" on page 264. For information on calculating cell priority, see
"Cell Priority" on page 261.
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:
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'.
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:
• 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:
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.
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:
C j N1 i = I j c N1
Where
c N1 is the cost of the first order neighbour constraint. This value can be defined in the Constraint Cost dialogue.
Because two first order neighbours must not have the same 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
order neighbours to be each other.
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.
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
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:
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
C i U = C i Dist + C i EP + C i N + C i N 2G + C i Cluster
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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 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.
Note:
• Here, real inter-transmitter distance is considered.
3. There must be an overlapping zone ( S A S B ) with a given cell edge coverage probability.
- 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 signal received from A is greater than the minimum pilot signal level,
- The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/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 signal received from A is greater than the minimum pilot signal level,
- The pilot quality from A is within a margin from the best Ec/Io, where the best Ec/Io exceeds the minimum
Ec/Io.
- 4th case: SA represents the area where:
- The pilot signal received from A is greater than the minimum pilot signal level,
- The pilot quality from A is within a margin from the best Ec/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.
- 1st case: SB is the area where the cell B is the best serving cell of the GSM network.
In this case, the margin must be set to 0dB.
- The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and
is the highest one.
- 2nd case: The margin is different from 0dB and SB is the area where:
- The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and
is within a margin from the best BCCH signal level.
SA SB
Atoll calculates the percentage of covered area ( ---------------------- 100 ) and compares this value to the % minimum covered
SA
area. If this percentage is not exceeded, the candidate neighbour B is discarded.
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of covered
area.
- 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.
4. 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…).
If there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8.
Therefore, among 15 candidate neighbours, only 8 (those with the highest priority) will be allocated to the reference cell.
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 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 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:
• 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.
• The percentage of covered area is calculated with the resolution specified in the properties
dialog of the predictions folder (default resolution parameter).
• A forbidden neighbour must not be listed as neighbour except if the neighbourhood
relationship already exists and the Delete existing neighbours option is unchecked when
you start the new allocation. In this case, Atoll displays a warning in the Event viewer
indicating that the constraint on the forbidden neighbour will be ignored by algorithm
because the neighbour already exists.
• In the Results, Atoll displays only the cells for which it finds new neighbours. Therefore, if a
TBA cell has already reached its maximum number of neighbours before starting the new
allocation, it will not appear in the Results table.
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.
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.
Atoll
RF Planning and Optimisation Software
Technical Reference Guide
Total losses ( L total ) in dBm L total = L path + L Tx + L term + L indoor + M Shadowing – model – G Tx + G term
where,
EIRP is the effective isotropic radiated power of the transmitter,
ic is a carrier number,
L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model,
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
“Indoor coverage” is selected,
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).
Note:
• For a selected transmitter, it is also possible to study the path loss, L path , or the total
losses, L total . Path loss and total losses are the same on any carrier.
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.
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.
Note:
• For a selected transmitter, it is also possible to study the path loss, L path , or the total
losses, L total . Path loss and total losses are the same on any carrier.
• 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.
And
Txi Txj
P rec ic Best P rec ic – M
ji
And
Txi nd Txj
P rec ic 2 Best P rec ic – M
ji
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.
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.
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.
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.
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where 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).
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 minimum threshold.
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).
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.
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
fulfils signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability in the
explorer.
F ortho Clutter parameter or global parameter (default value) None Orthogonality factor
Tx
F MUD Site equipment parameter None MUD factor
E
FCH – DL
Eb/Nt target for FCH channel on
DL
Q req FCH -----b- (Service, Mobility) parameter None
N t req downlink
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
N max NI Site parameter None
for a site on downlink
CE – U L Number of channel elements of a site
N NI Simulation result None
consumed by users on uplink
CE – D L Number of channel elements of a site
N NI Simulation result None
consumed by users on downlink
Number of channel elements used by
Overhead – C E – UL Site equipment parameter None the cell for common channels on
N
uplink
Number of channel elements used by
Overhead – C E – DL Site equipment parameter None the cell for common channels on
N
downlink
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
N0 NF Term K T W W Thermal noise at terminal
UL
f rake efficiency Equipment parameter None Uplink rake receiver efficiency factor
DL DL DL
R SCH R FCH Frate SCH bps Downlink SCH bit rate
UL
R FCH Terminal parameter bps Uplink FCH nominal rate
UL UL UL
R SCH R FCH Frate SCH bps Uplink SCH bit rate
FCH – UL
W -
------------- Uplink service processing gain on
Gp UL None
R FCH FCH
DL
AF FCH Service parameter None Downlink activity factor on FCH
UL
AF FCH Service parameter None Uplink activity factor on 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
P FCH txi ic P FCH txi ic tch W Total FCH power on carrier ic
tch FCH 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
P term Terminal parameter W Minimum terminal power allowed
max
P term Terminal parameter W Maximum terminal power allowed
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
UL npaths
G macro – diversity = M Shadowing – Eb Nt – M Shadowing – Eb Nt
UL UL UL UL quality gain due to signal diversity
G macro – diversity None
n=2 or 3 in soft handoffc.
Global parameter (default value)
Random shadowing error drawn
E Shadowing Simulation result None during Monte-Carlo simulation
Only used in simulations
In prediction studiesd
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
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 UL
P c ic p P tot ic W Chip received power at transmitter
the shadowing model ( M Shadowing – Eb Nt or M Shadowing – Ec Io ) while extra-cell interference level is not. Therefore,
DL
DL Downlink extra-cell interference at
DL
I extra ic P tot txj ic W
terminal on carrier ic
txj j i
DL Downlink extra-cell interference at
DL
I extra ic P tot txj ic W
terminal on carrier ic
txj j i
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
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.
b. 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.
Pb
UL UL
ic adj + P c ic adj
UL Uplink inter-carrier interference at
I inter – carrier txi ic term W
txj j
terminal on carrier ic
-----------------------------------------------------------------------------
-
RF ic ic adj
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
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
tx ,tx ActiveSet
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
G macro – diversity
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
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.
b. 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.
DL Downlink extra-cell interference at
DL
I extra ic P tot txj ic W
terminal on carrier ic
txj j i
Pb
UL UL
ic adj + P c ic adj
UL Uplink inter-carrier interference at
I inter – carrier txi ic term W
txj j
terminal on carrier ic
-----------------------------------------------------------------------------
-
RF ic ic adj
UL
I tot txi ic
F
UL
txi ic ------------------------------------------------------------------------------------------- None Cell uplink reuse factor on carrier ic
UL intra Tx
I tot txi ic 1 – F MUD term
DL
P tx txi ic
-------------------------------- Percentage of max transmitter power
%Power txi ic - 100 None
P max txi ic used.
DL
DL
I tot ic
F txi ic -------------------------------
- None Downlink reuse factor on a carrier ic
DL
I intra txi ic
F ortho Clutter parameter or global parameter (default value) None Orthogonality factor
Tx
F MUD Site equipment parameter None MUD factor
UL
X max Simulation constraint (global parameter or cell parameter) % Maximum uplink load factor
Tx
N0 NF Tx K T W W Thermal noise at transmitter
Term
N0 NF Term K T W W Thermal noise at terminal
UL
f rake efficiency Equipment parameter None Uplink rake receiver efficiency factor
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
max
P term Terminal parameter W Maximum terminal power allowed
G RRI Terminal parameter (for 1xEV-DO Rev A terminals only) None Reverse Rate Indicator Channel gain
G Auxiliary – pilot Terminal parameter (for 1xEV-DO Rev A terminals only) None Auxiliary Pilot Channel gain
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/I0 Ec/I0 Shadowing margin
M Shadowing – Ec Io None
standard deviation Only used in prediction studies
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 b.
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
UL npaths
G macro – diversity = M Shadowing – Eb Nt – M Shadowing – Eb Nt
UL UL UL UL quality gain due to signal diversity
G macro – diversity None
n=2 or 3 in soft handoffc.
Global parameter (default value)
Random shadowing error drawn
E Shadowing Simulation result None during Monte-Carlo simulation
Only used in simulations
In prediction studiesd
For Ec/I0 and Ec/Nt calculations
L path L Tx L term L body L indoor M Shadowing – Ec Io
---------------------------------------------------------------------------------------------------------------------------------------------------
G Tx G term
For UL Eb/Nt calculation
LT None Transmitter-terminal total loss
L path L Tx L term L body L indoor M Shadowing – Eb Nt
UL
------------------------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx G term
In simulations
L path L Tx L term L body L indoor E Shadowing
--------------------------------------------------------------------------------------------------------------------------------
G Tx G term
UL
NR threshold txi ic Cell parameter dB Cell uplink noise rise threshold
d. 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 – Ec Io ) while extra-cell interference level is not. Therefore, M Shadowing – Ec Io is set to
1 in downlink extra-cell interference calculation.
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 Downlink extra-cell interference based
DL
I extra ic b traffic P tot txj ic b traffic W
on traffic at terminal on carrier ic
txj j i
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
Pb
UL
ic adj
UL Uplink inter-carrier interference at
I inter – carrier txi ic term W
txj j
terminal on carrier ic
--------------------------------------
-
RF ic ic adj
UL
No HO: Q 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
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
tx ,tx ActiveSet
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
G macro – diversity
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
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
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.
b. In simulations, the uplink Eb/Nt target is calculated whithout considering the aknoledgement signal.
DL Downlink extra-cell interference based
DL
I extra ic b traffic P tot txj ic b traffic W
on traffic at terminal on carrier ic
txj j i
Pb
UL
ic adj
UL Uplink inter-carrier interference at
I inter – carrier txi ic term W
txj j
terminal on carrier ic
--------------------------------------
-
RF ic ic adj
UL
UL
I tot txi ic
X txi ic -----------------------------
- None Cell uplink load factor on carrier ic
UL
N tot txi ic
UL
I tot txi ic
F
UL
txi ic ------------------------------------------------------------------------------------------- None Cell uplink reuse factor on carrier ic
UL intra Tx
I tot txi ic 1 – F MUD term
- 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.
X = SD
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.
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.
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…).
n j = n j UL + n j DL + n j UL + DL + n j inactive = n j UL + DL
UL DL
R FCH and R FCH are respectively the uplink and downlink FCH nominal rates.
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
4x R FCH AF FCH + 4 R FCH AF FCH + 4
Both FCH and SCH are used
UL UL DL DL
8x R FCH AF FCH + 8 R FCH AF FCH + 8
UL UL DL DL
16x R FCH AF FCH + 16 R FCH AF FCH + 16
UL DL
R FCH and R FCH are respectively the uplink and downlink FCH nominal rates.
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
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
Therefore, we have:
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
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
- 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,
N UL p UL + DL N DL p UL + DL
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:
• 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 , 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.
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 UL p UL + DL N DL p UL + DL
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
2x R FCH AF FCH + 2 R FCH AF FCH + 2
UL UL DL DL
4x R FCH AF FCH + 4 R FCH AF FCH + 4
Both FCH and SCH are used
UL UL DL DL
8x R FCH AF FCH + 8 R FCH AF FCH + 8
UL UL DL DL
16x R FCH AF FCH + 16 R FCH AF FCH + 16
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,
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
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:
• 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
- 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:
n j = n j UL + n j inactive
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
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.
ends once the network is balanced, i.e. when the convergence criteria (on UL and DL) are satisfied. Algorithm steps are
detailed below.
UL intra UL extra UL
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
N tot txi ic
If a given carrier is specified for the service requested by Mb and if it is used by txi
EndFor
UL
BestCarrier k txi M b is the carrier with the lowest X k txi ic
max
If Q pilot txi M b BestCarrier Q pilot M b
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
Q pilot M b = Q pilot txi M b BestCarrier
k k
tx BS M b = txi
Endif
EndFor
If no txBS has been selected, Mb cannot get a connection to the network
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
EndFor
req
Calculation of the required power for Mb, P term M b ic k
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 = -----------------------------------------------------
-
L T txi M b
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
End For
If (Mb is not in handoff)
UL UL
Q FCH M b = Q FCH txi M b ic k
k
UL UL UL
Q FCH M b = f rake efficiency Q FCH txi M b ic k
k
txi ActiveSet
UL UL UL
Q FCH M b = Max Q FCH txi M b ic k G macro – diversity 2 links
k
txi ActiveSet
UL UL UL
Q FCH M b = Max Q FCH txi M b ic k G macro – diversity 3 links
k
txi ActiveSet
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
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
min
FCH – D L P FCH Service M b
Pb txi M b ic = ------------------------------------------------------
-
L T txi M b
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
End For
DL DL DL
Q FCH M b = f rake efficiency Q FCH txi M b ic k
k
txi ActiveSet FCH
Do
For each cell (txi,ic) in Mb active set
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
EndFor
DL DL DL
Q FCH M b = f rake efficiency Q FCH txi M b ic k
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
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
Update of N tot ic
EndFor
Control of Radio Resource Limits (Walsh Codes, Cell Power and Site Channel Elements)
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
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
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
For each cell (txi,ic)
Codes Codes
While N txi ic k N max txi ic
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
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
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).
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 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.
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 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.
UL intra UL extra UL
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
Determination of BestCarrier k S j M i .
UL
UL I tot S j ic UL
X k S j ic = ----------------------------
- + X
UL
N tot S j ic
EndFor
UL
BestCarrier k S j M i is the carrier with the lowest X k S j ic
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
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
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
EndFor
req
Calculation of the required power for Mi, P term M i ic k
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
Gp Service
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
Gp Service
N tot ic – 1 – F MUD term P b M i S j ic + P b M i S j ic
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
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
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
= 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 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
req max
If P term M i ic k P term M i and Mi uses SCH then:
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
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
min
SCH – D L P SCH Service M i
Pb M i S j ic = -----------------------------------------------------
-
L T M i S j
SCH – D L
DL BTS P b M i S j
- 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
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
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
Q Service M Mobility M R
DL
Service M 2
req i i SCH i SCH
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
DL R SCH Service M i
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
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
Q Service M Mobility M R
DL
Service M 2
req i i SCH i SCH
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
DL R SCH Service M i
R SCH Service M i = -----------------------------------------------------
-
2
Codes
Codes N SCH M i k
N SCH M i k = --------------------------------
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
Q Service M Mobility M R
DL
Service M 2
req i i SCH i SCH
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
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
Update of N tot ic
EndFor
Control of Radio Resource Limits (Walsh Codes, Cell Power and Site Channel Elements)
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
EndWhile
EndFor
For each site Nl
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
req DL
If P tx S j ic k + P tch S j M b ic k %Power max P max + M Pooling S j ic
Mb is reconnected
EndIf
EndFor
EndIf
EndFor
For each cell (Sj,ic)
Codes Codes
While N S j ic k N max S j ic
req
Rejection of mobile with highest P tch M i S j k for the lowest service priority
CE – UL CE – UL
While N N I k N max NI
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
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).
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 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.
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 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.
UL
UL I tot S j ic
X k S j ic = ----------------------------
- = 0
UL
N tot S j ic
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 BestCarrier k S j M i .
UL
UL I tot S j ic UL
X k S j ic = ----------------------------
- + X
UL
N tot S j ic
EndFor
UL
BestCarrier k S j M i is the carrier with the lowest X k S j ic
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
S BS M i = S j
Endif
EndFor
If no SBS has been selected, Mi cannot get a connection to the network
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
EndFor
req
Calculation of the required power for Mi, P term M i ic k
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
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
UL UL UL
Q total M i = Max Q M i S j ic k G macro – diversity 2 links
k
I AS ActiveSet
UL UL UL
Q total M i = Max Q M i S j ic k G macro – diversity 3 links
k
I AS ActiveSet
= 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)
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
DL
Calculation of N tot ic b traffic
DL
Calculation of the maximum data rate supplied to Mi, R max M i S k
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:
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
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
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
MacIndexes MacIndexes
While N S j ic N max S j ic
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
Update of N tot S j ic
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
Update of N tot S j ic
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 downgraded
Downgrading the traffic data channel rate for the 1xEV-DO Rev. A mobile with the highest data rate
UL
Update of N tot S j ic
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
Update of N tot S j ic
Endwhile
UL UL
For each cell (Sj,ic) with X S j ic X max
DL
Calculation of N tot ic b traffic
EndFor
EndFor
DL
Calculation of the maximum data rate supplied to Mi, R max
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
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
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.
The simulation has not reached convergence (specific divergence symbol).
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).
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
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.
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).
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
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.
CE – UL
On uplink, Atoll consumes N j channel elements for each cell j on a site NI. This figure includes:
• 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
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
On downlink, only one user can be served by a cell at a time, so this resource is not limited.
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
term
P tch ic = CI req I extra ic + I intra ic + I inter – carrier ic + N 0 LT
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.
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:
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
The downlink load factor represents the signal degradation in relative to the reference interference (thermal noise).
cells. These parameters can be either outputs of a given simulation, average values calculated from a group of simulations,
or user-defined cell inputs.
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
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.
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
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 = -------------------------------------------------------
-
RF ic ic adj
Finally, we have,
DL
P tot j ic adj
DL j j
I inter – carrier ic = ---------------------------------------
RF ic ic adj
term
3rd step
N0 calculation
term
N0 = F term K T W
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)
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.
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.
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
DL BTS P b – max k ic BS
- 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 = -------------
- and N tot ic BS = I intra ic BS + I extra ic BS + I inter – carrier ic BS + N 0
LT
k
Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table.
DL
N tot ic BS is the total noise at the receiver on the best carrier of the best server.
With
DL
I intra ic BS = 1 – BTS F ortho P DL k ic
tot BS
Ptot j icBS
DL DL
I extra ic BS =
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
I inter – carrier ic BS is the inter-carrier interference at the receiver on the best carrier of the best server.
Ptot j icadj
DL
DL txj j
I inter – carrier ic BS = -------------------------------------------
RF ic BS ic adj
RF ic BS ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
Qmax k icBS
DL DL DL
Q MAX ic BS = f rake efficiency
k
Where
DL
f rake efficiency is the downlink rake efficiency factor defined in the Terminal properties.
req
3rd step: P FCH calculation
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
DL BTS P b – max k ic BS
- 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 = -------------
- and N tot ic BS = I intra ic BS + I extra ic BS + I inter – carrier ic BS + N 0
LT
k
Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table.
DL
N tot ic BS is the total noise at the receiver on the best carrier of the best server.
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
Ptot j icBS
DL DL
I extra ic BS =
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
I inter – carrier ic BS is the inter-carrier interference at the receiver on the best carrier of the best server.
Ptot j icadj
DL
DL txj j
I inter – carrier ic BS = -------------------------------------------
RF ic BS ic adj
RF ic BS ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
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).
Qmax k icBS
DL DL DL
Q MAX ic BS = f rake efficiency
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
DL
DL Q MAX ic BS
G SHO = --------------------------------------------------------
DL
max Q max k ic BS
k
DL DL
max Q max k ic BS corresponds to the highest Q max k ic BS value.
k
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
DL BTS P b –max k ic BS
- 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.
DL
N tot ic BS is the total noise at the receiver on the best carrier of the best server.
With
DL
I intra ic BS = 1 – BTS F ortho P DL k ic
tot BS
And
Ptot j icBS
DL DL
I extra ic BS =
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
I inter – carrier ic BS is the inter-carrier interference at the receiver on the best carrier of the best server.
Ptot j icadj
DL
DL txj j
I inter – carrier ic BS = -------------------------------------------
RF ic BS ic adj
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 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).
Where
DL
f rake efficiency is the downlink rake efficiency factor defined in Terminal properties.
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).
req req
3rd step: P FCH and P SCH calculation
DL
req Q req FCH
- 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
DL BTS P b –max k ic BS
- 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 = -------------
- and N tot ic BS = I intra ic BS + I extra ic BS + I inter – carrier ic BS + N 0
LT
k
Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table for a certain terminal.
DL
N tot ic BS is the total noise at the receiver on the best carrier of the best server.
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
Ptot j icBS
DL DL
I extra ic BS =
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
I inter – carrier ic BS is the inter-carrier interference at the receiver on the best carrier of the best server.
Ptot j icadj
DL
DL txj j
I inter – carrier ic BS = -------------------------------------------
RF ic BS ic adj
RF ic BS ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
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
DL BTS P b –max k ic BS
- 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 = -------------
- 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 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.
DL
N tot ic BS is the total noise at the receiver on the best carrier of the best server.
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
Ptot j icBS
DL DL
I extra ic BS =
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
I inter – carrier ic BS is the inter-carrier interference at the receiver on the best carrier of the best server.
Ptot j icadj
DL
DL txj j
I inter – carrier ic BS = -------------------------------------------
RF ic BS ic adj
RF ic BS ic adj is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
Where
DL
f rake efficiency is the downlink rake efficiency factor defined in Terminal properties.
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
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 .
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
max Q max k ic BS corresponds to the highest Q max k icBS value.
k
2. Required C/I
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
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 .
UL
1st step: Evaluation of the uplink traffic channel quality, Q max i ic BS , for each cell of active set
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
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
N 0 is the transmitter thermal noise.
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
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
max Q max i ic BS corresponds to the highest Q max i ic BS value.
i
UL
G macro – diversity 3 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.
For softer and softer-softer handovers, we have:
Qmax i icBS
UL UL UL
Q MAX ic BS = f rake efficiency
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
Q MAX ic BS = G macro – diversity 2 links max Q max i ic BS
i
req
3rd step: Calculation of P term
UL
req Q req FCH
- P max
P term = ------------------------------- term
UL
Q MAX ic BS
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.
req max
Therefore, the service on the uplink data traffic channel is available if P term P term .
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
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 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
N 0 is the transmitter thermal noise.
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
Q MAX ic BS = Q max i ic BS
UL
G macro – diversity 2 links is the uplink macro-diversity gain. This parameter is determined from the fixed cell edge coverage
probability and the uplink Eb/Nt standard deviation. When the option “Shadowing taken into account” is not selected
(Prediction properties), Atoll considers the uplink macro-diversity gain defined by the user in Global parameters.
UL UL
max Q max i ic BS corresponds to the highest Q max i icBS value.
i
UL
G macro – diversity 3 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.
For softer and softer-softer handovers, we have:
Qmax i icBS
UL UL UL
Q MAX ic BS = f rake efficiency
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
Q MAX ic BS = G macro – diversity 2 links max Q max i ic BS
i
4. Effective Eb/Nt
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
UL
UL Q MAX ic BS
G SHO = -----------------------------------------------------
UL
max Q max i ic BS
i
UL UL
max Q max i ic BS corresponds to the highest Q max i ic BS value.
i
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
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.
For each cell (i,icBS), we have:
UL – FCH
UL term P b –max i ic BS
- G UL
Q max i ic BS FCH = ---------------------------------------------------------------- p
– FCH
UL
N tot i ic BS
And
UL – SCH
UL term P b –max i ic BS UL – SCH
Q max i ic BS SCH = -----------------------------------------------------------------
UL
Gp
N tot i ic BS
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
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
N 0 is the transmitter thermal noise.
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.
And
UL UL UL
Q MAX ic BS SCH = G macro – diversity 2 links max Q max i ic BS SCH
i
UL
G macro – diversity 2 links is the uplink macro-diversity gain. 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
max Q max i ic BS corresponds to the highest Q max i icBS value.
i
And
UL UL UL
Q MAX ic BS SCH = G macro – diversity 3 links max Q max i ic BS SCH
i
UL
G macro – diversity 3 links is the uplink macro-diversity gain.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.
For softer and softer-softer handovers, we have:
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 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 other site
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 other site
i on the same site
otherwise,
UL UL UL
Q MAX ic BS FCH = G macro – diversity 2 links max Q max i ic BS FCH
i
And
UL UL UL
Q MAX ic BS SCH = G macro – diversity 2 links max Q max i ic BS SCH
i
req req
3rd step: Calculation of P term FCH and P term SCH
UL UL
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
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 .
UL – FCH
UL term P b –max i ic BS
- G UL
Q max i ic BS FCH = ---------------------------------------------------------------- p
– FCH
UL
N tot i ic BS
max
UL – FCH P term FCH
With P b –max i ic BS = ----------------------------
-
LT
i
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 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
N 0 is the transmitter thermal noise.
UL – SCH
UL term P b –max i ic BS UL – SCH
Q max i ic BS SCH = -----------------------------------------------------------------
UL
Gp
N tot i ic BS
max
UL – SCH P term SCH
With P b –max i ic BS = ----------------------------
-
LT
i
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 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
N 0 is the transmitter thermal noise.
And
UL UL UL
Q MAX ic BS SCH = G macro – diversity 2 links max Q max i ic BS SCH
i
UL
G macro – diversity 2 links is the uplink macro-diversity gain. 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
max Q max i ic BS corresponds to the highest Q max i icBS value.
i
And
UL UL UL
Q MAX ic BS SCH = G macro – diversity 3 links max Q max i ic BS SCH
i
UL
G macro – diversity 3 links is the uplink macro-diversity gain. 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.
For softer and softer-softer handovers, we have:
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 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 other site
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 other site
i on the same site
otherwise,
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
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
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
max Q max i ic BS corresponds to the highest Q max i ic BS value.
I
UL
1st step: Evaluation of the uplink quality, Q max i ic BS , for each cell of active set
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
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
N 0 is the transmitter thermal noise.
UL
2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status.
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
max Q max i ic BS corresponds to the highest Q max i icBS value.
i
UL
G macro – diversity 3 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.
For softer and softer-softer handovers, we have:
Qmax i icBS
UL UL UL
Q MAX ic BS = f rake efficiency
i
For softer-soft handover, there are two possibilities. If the MRC option is selected (option available in Global parameters),
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
Q MAX ic BS = G macro – diversity 2 links max Q max i ic BS
i
UL
3rd step: Evaluation of the required quality level on uplink, Q req
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 .
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
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
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 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
N 0 is the transmitter thermal noise.
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.
If there is no handoff, we have:
UL UL
Q MAX ic BS = Q max i ic BS
UL
G macro – diversity 2 links is the uplink macro-diversity gain. This parameter is determined from the fixed cell edge coverage
probability and the uplink Eb/Nt standard deviation. When the option “Shadowing taken into account” is not selected
(Prediction properties), Atoll considers the uplink macro-diversity gain defined by the user in Global parameters.
UL UL
max Q max i ic BS corresponds to the highest Q max i ic BS value.
i
UL
G macro – diversity 3 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.
Qmax i icBS
UL UL UL
Q MAX ic BS = f rake efficiency
i
For softer-soft handover, there are two possibilities. If the MRC option is selected (option available in Global parameters),
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
Q MAX ic BS = G macro – diversity 2 links max Q max i ic BS
i
UL
UL Q MAX ic BS
G SHO = -----------------------------------------------------
UL
max Q max i ic BS
i
Resulting
received with a fixed cell edge coverage probability, Q pilot ic BS .
Atoll displays the best pilot quality received with a fixed cell edge coverage probability.
Resulting
Q pilot ic given .
Atoll displays the best pilot quality received with a fixed cell edge coverage probability.
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
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
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 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
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.
6. Colour per quality level (Ec/I0)
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
Resulting
properties). 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.
7. Colour per quality margin (Ec/I0 margin)
Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction 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
assigned a colour and displayed with intersections between layers.
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.
2. Colour per transmitter
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.
3. 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, 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.
DL
req Q req FCH
- 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
displayed with intersections between layers.
11. 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-
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
intersections between layers.
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.
For further details on formulas, see "Definitions and Formulas" on page 276. For further details on calculations, see
"Downlink Sub-Menu" on page 321.
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
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
DL R RLC – peak Index DL – RevA – Bearer
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
DL R RLC – peak Index DL – RevA – Bearer
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
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.
2. Colour per transmitter
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.
3. 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, 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.
Note:
• Best server and active set determination is performed as in point prediction (AS analysis).
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
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 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.
6. 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
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
displayed with intersections between layers.
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
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
UL
properties). For each layer, area is covered if Q MAX ic TCH Threshold . Each layer is assigned a colour and displayed
with intersections between layers.
8. Colour per effective quality level (Effective Eb/Nt)
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
UL
properties). For each layer, area is covered if Q effective ic TCH Threshold . Each layer is assigned a colour and
displayed with intersections between layers.
UL UL UL
Q effective ic TCH = min Q MAX ic TCH Q req 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
with intersections between layers.
11. 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-
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
intersections between layers.
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
assigned a colour and displayed with intersections between layers.
UL UL UL
Q req R is the uplink quality required to obtain the data rate, R .
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 .
Each layer is assigned a colour and displayed with intersections between layers.
UL UL UL
Q req R av is the uplink quality required to obtain the average data rate, R av .
Ptot icadj
DL
Ptot ic + txj--------------------------------------
DL DL j term
N tot ic = - + N0
RF ic ic adj
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
Atoll displays bins where min NR 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.
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
noise rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
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.
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.
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:
• 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 cells may be assigned neighbours.
Note:
• If no focus zone exists in the .atl document, Atoll takes into account the computation zone.
3. There must be an overlapping zone ( S A S B ) with a given cell edge coverage probability:
- The pilot signal received from the cell A is greater than the minimum pilot signal level.
- 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 signal received from the cell B is greater than the minimum pilot signal level.
- 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.
- The pilot signal received from the cell A is greater than the minimum pilot signal level.
- 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 received from the cell B is greater than the minimum pilot signal level.
- The pilot signal from B is the highest one.
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 received from the cell A is greater than the minimum pilot signal level.
- 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 received from the cell B is greater than the minimum pilot signal level.
- The pilot signal from B is the highest one.
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.
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.
The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see
after).
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 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.
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 %
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
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 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
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 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 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).
• 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 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.
• 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 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.
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:
• If no focus zone exists in the .atl document, Atoll takes into account the computation zone.
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]).
Notes:
• In the context of the PN Offset allocation, the term "neighbours" refers to intra-carrier
neighbours.
• Atoll can take into account inter-technology neighbour relations as constraints to allocate
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.
• Atoll considers symmetry relationship between a cell, its first order neighbours, its second
order neighbours and its third order neighbours.
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.
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).
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.
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.
All sites which have constraints with the studied site are referred to as near sites.
Atoll assigns a PN-cluster of adjacent PN offsets to each site, starting with the highest priority site and its near sites, and
continuing with the lowest priority sites not allocated yet and their near sites. When all the clusters have been allocated
and there are still sites remaining to be allocated, Atoll reuses the clusters at another site. When the Co-PN Reuse
Distance option is selected, the algorithm reuses the clusters as soon as the Co-PN reuse distance is exceeded. 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.
For information on calculating site priority, see "Site Priority" on page 355. For information on calculating cell priority, see
"Cell Priority" on page 353.
All sites which have constraints with the studied site are referred to as near sites.
Atoll assigns one cluster to each site, starting with the highest priority site and its near sites, and continuing with the lowest
priority sites not allocated yet and their near sites. When all the clusters have been allocated and there are still sites
remaining to be allocated, Atoll reuses the clusters at another site. When the Co-PN Reuse Distance option is selected,
the algorithm reuses the clusters as soon as the Co-PN reuse distance is exceeded. Otherwise, when the option is not
selected, the algorithm tries to assign reused clusters as spaced out as possible.
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.
For information on calculating site priority, see "Site Priority" on page 355. For information on calculating cell priority, see
"Cell Priority" on page 353.
The cost due to the domain constraint, C i Dom , depends on the number of PN Offsets available for the allocation. The
domain constraint is mandatory and cannot be broken.
When no domain is assigned to cells, 512 PN Offsets are available and we have:
C i Dom = 0
When domains of PN Offsets are assigned to cells, each unavailable PN Offset generates a cost. The higher the number
of codes available in the domain, the less will be the cost due to this criterion. The cost is given as:
• 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:
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.
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:
C j N1 i = I j c N1
Where
c N1 is the cost of the first order neighbour constraint. This value can be defined in the Constraint Cost dialogue.
Because two first order neighbours must not have the same PN Offset, Atoll considers the cost created by two first order
neighbours to be each other.
C j N1 i + C j N1 i
C j – j N1 i = ----------------------------------------------------------
-
2
Each second order neighbour cell k generates a cost given as:
C k N2 i = Max ( C j N1 i C k N1 j , C j N1 i C k N1 j ) c N2
Where
c N2 is the cost of the second order neighbour constraint. This value can be defined in the Constraint Cost dialogue.
Because two second order neighbours must not have the same PN Offset, Atoll considers the cost created by two second
order neighbours to be each other.
C k N2 i + C k N2 i
C k – k N2 i = ------------------------------------------------------------
-
2
Each third order neighbour cell l generates a cost given as:
C N1 i C k N1 j C l N1 k C j N1 i C k N1 j C l N1 k
C l N3 i = Max j c N3
C j N1 i C k N1 j C l N1 k C j N1 i C k N1 j C l N1 k
Where
c N3 is the cost of the third order neighbour constraint. This value can be defined in the Constraint Cost dialogue.
Because two third order neighbours must not have the same PN Offset, Atoll considers the cost created by two third order
neighbours to be each other.
C l N3 i + C l N3 i
C l – l N3 i = ----------------------------------------------------------
-
2
Note:
• Atoll considers the highest cost of both links when a neighbour relation is symmetric and the
importance value is different.
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
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
Therefore, the total cost due to constraints on any cell i is defined as:
C i = C i Dom + C i U
With
C i U = C i Dist + C i EP + C i N + C i N 2G
C Tx = C Tx Dom + C Tx U
Here, the domain available for the transmitter is the intersection of domains assigned to cells of the transmitter. The
domain constraint is mandatory and cannot be broken.
C S = C S U + C S Dom
Here, the domain considered for the site is the intersection of domains available for transmitters of the site.
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.
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.
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.
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.
Note that the automatic inter-technology neighbour allocation algorithm takes into account both cases.
In order to be able to use the inter-technology neighbour allocation algorithm, you must have:
• An .atl document containing the GSM/TDMA network, GSM.atl, and another one containing the CDMA (i.e. IS-95
cdmaOne or CDMA2000) network, CDMA.atl,
• An existing link on the Transmitters folder of GSM.atl into CDMA.atl.
The external neighbour allocation algorithm takes into account all the GSM TBC transmitters. It means that all the TBC
transmitters of GSM.atl are potential neighbours. The cells to be allocated will be called TBA cells which, being cells of
CDMA.atl, fulfill following conditions:
• They are active,
• They satisfy the filter criteria applied to Transmitters folder,
• They are located inside the focus zone,
• They belong to the folder for which allocation has been executed. This folder can be either the Transmitters folder
or a group of transmitters subfolder.
Only CDMA TBA cells may be assigned neighbours.
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.
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 allocation cause for each neighbour. Therefore, a neighbour may
be marked as exceptional pair or co-site. If the neighbour is not forced but satisfies distance conditions, Atoll displays the
distance from the reference cell. Finally, if cells have previous allocations in the list, neighbours are marked as existing.
Note:
• Here, real inter-transmitter distance is considered.
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 signal received from A is greater than the minimum pilot signal level,
- The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0) and is the highest one.
In this case, the Ec/I0 margin must be equal to 0dB and the max Ec/I0 option disabled.
- 2nd case: SA represents the area where the pilot quality from the cell A strats decreasing but the cell A is still
the best serving cell of the CDMA network.
The Ec/I0 margin must be equal to 0dB, the max Ec/I0 option selected and a maximum Ec/I0 user-defined.
- The pilot signal received from A is greater than the minimum pilot signal level,
- The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/I0.
- The pilot quality from A is the highest one.
- 3rd case: SA represents the area where the cell A is not the best serving cell but can enter the active set.
Here, the Ec/I0 margin has to be different from 0dB and the max Ec/I0 option disabled.
- The pilot signal received from A is greater than the minimum pilot signal level,
- The pilot quality from A is within a margin from the best Ec/I0, where the best Ec/I0 exceeds the minimum
Ec/I0.
- 4th case: SA represents the area where:
- The pilot signal received from A is greater than the minimum pilot signal level,
- The pilot quality from A is within a margin from the best Ec/I0 (where the best Ec/I0 exceeds the minimum
Ec/I0) and lower than the maximum Ec/I0.
In this case, the margin must be different from 0dB, the max Ec/I0 option selected and a maximum Ec/I0
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.
- The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and
is the highest one.
- 2nd case: The margin is different from 0dB and SB is the area where:
- The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and
is within a margin from the best BCCH signal level.
SA SB
Atoll calculates the percentage of covered area ( ---------------------- 100 ) and compares this value to the % minimum covered
SA
area. If this percentage is not exceeded, the candidate neighbour B is discarded.
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of covered
area.
- 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.
4. 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).
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.
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 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 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:
• 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.
• The percentage of covered area is calculated with the resolution specified in the properties
dialog of the predictions folder (default resolution parameter).
• A forbidden neighbour must not be listed as neighbour except if the neighbourhood
relationship already exists and the Delete existing neighbours option is unchecked when
you start the new allocation. In this case, Atoll displays a warning in the Event viewer
indicating that the constraint on the forbidden neighbour will be ignored by algorithm
because the neighbour already exists.
• In the Results, Atoll displays only the cells for which it finds new neighbours. Therefore, if a
TBA cell has already reached its maximum number of neighbours before starting the new
allocation, it will not appear in the Results table.
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:
• They are active,
• They satisfy the filter criteria applied to Transmitters folder,
• They are located inside the focus zone,
• They belong to the folder for which allocation has been executed. This folder can be either the Transmitters folder
or a group of transmitters subfolder.
Only CDMA2000 TBA cells may be assigned neighbours.
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.
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 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.
Notes:
• A forbidden neighbour must not be listed as neighbour except if the neighbourhood
relationship already exists and the Delete existing neighbours option is unchecked when
you start the new allocation. In this case, Atoll displays a warning in the Event viewer
indicating that the constraint on the forbidden neighbour will be ignored by algorithm
because the neighbour already exists.
• In the Results, Atoll displays only the cells for which it finds new neighbours. Therefore, if a
TBA cell has already reached its maximum number of neighbours before starting the new
allocation, it will not appear in the Results table.
Atoll
RF Planning and Optimisation Software
Technical Reference Guide
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.
8.1.1 Inputs
This table lists the inputs to computations, coverage predictions, and simulations.
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)
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
Req
RU DL Cell parameter None Required resource units in downlink
Max
N HSDPA Cell parameter None Maximum number of HSDPA users
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
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
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
Req
RSCP DwPCH Mobility parameter W Required RSCP threshold for DwPCH
Req
RSCP UpPCH Mobility parameter W Required RSCP threshold for UpPCH
C Req
Req
Q DwPCH Mobility parameter ( ---- ) None Required quality threshold for DwPCH
I DwPCH
Ortho
F DL Clutter class parameter None Downlink orthogonality factor
Ortho
F UL Clutter class parameter None Uplink orthogonality factor
Term
N0 NF Term K T BW W Thermal noise at terminal
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
standard deviation coverage predictions
Eb Nt UL Result calculated from cell edge coverage probability and UL Eb/Nt UL Eb/Nt shadowing margin used in
M Shadowing None
standard deviation coverage predictions
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.
TX i TX i ic
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 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 jc TX j jc
I IC – DL ic jc
RSCPP – CCPCH + RSCPOCCH – TS0 W Inter-carrier interference
TX
-----------------------------------------------------------------------------------------------------------
j -
F IRF ic jc
TX i TX i ic
TX i ic RSCP DwPCH
C
---- ---------------------------------------------------- None DwPCH C/I for the cell TX i ic
I DwPCH TX i ic
N Tot – DL
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 jc
I IC – DL ic jc
RSCPDwPCH W Inter-carrier interference
TX
-----------------------------------------------
j -
F IRF ic jc
TX i TX i ic
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
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 jc TX j jc
I IC – DL ic jc
RSCPTCH – DL + RSCPOCCH W Inter-carrier interference
TX
------------------------------------------------------------------------------------------
j -
F IRF ic jc
Term TX i ic
TX i ic RSCP TCH – UL
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 ----
N t TCH – UL I TCH – UL
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
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
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
E SA
----------------------------------
- None Smart antenna gain in the uplink
H –1
S RN S
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
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
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,
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
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.
Where,
RSCP is the received signal code power. RSCP can be calculated for P-CCPCH, DwPCH, or the downlink TCH.
TX i ic TX i ic TX i TX i
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
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
Term
L is the terminal loss
Term
G is the receiver total gain
TX i
G is the transmitter antenna gain
TX i TX i
L is the transmitter loss ( L = L Total – DL )
• 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.
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 TX j jc
RSCP P – CCPCH Best RSCP P – CCPCH – M
ji
TX i ic TX j jc
RSCP P – CCPCH Best RSCP P – CCPCH – M
ji
• Dynamic channel allocation and power control as explained under "Power Control Simulation" on page 383.
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.
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.
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.
N Call D Call
Calculation of the service usage duration per hour ( p 0 : probability of a connection): p 0 = --------------------------------
-
3600
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 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
UL UL
Number of users active in the uplink: n i – Active = n i p Active
DL DL
Number of users active in the downlink: n i – Active = n i p Active
UL + DL UL + DL
Number of users active in the uplink and downlink both: n i – Active = n i p Active
Therefore, a connected user can be either active on both links, inactive on both links, active on UL only, or active on DL
only.
UL DL
In case of HSDPA services, f Eff and f Eff are the uplink and downlink A-DPCH activity factors, respectively.
Note:
• 1 kBytes = 1024 Bytes.
DL DL DL
D Inactivity Session = N PacketCall D Inactivity PacketCall
DL DL
DL DL N Packet S Packet 8
D Activity Session = N PacketCall ------------------------------------------------------
DL
-
R Nom 1000
UL UL UL DL DL DL
D Connection = D Activity Session + D Inactivity Session and D Connection = D Activity Session + D Inactivity Session
UL N Sess UL DL N Sess DL
p Connection = --------------- D Connection and p Connection = --------------- D Connection
3600 3600
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:
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:
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
UL UL
The probability of the user being inactive on both UL and DL: p2 Inactive = 1 – f p Connected
DL DL
The probability of the user being inactive on both UL and DL: p3 Inactive = 1 – f p Connected
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.
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.
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.
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
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
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
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
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
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.
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 ).
They are calculated as follows:
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:
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.
• 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
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 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.
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
L Path = L Model + L Ant
L Model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model
TX i
L Ant is the transmitter antenna attenuation (from antenna patterns)
Model
M Shadowing is the shadowing margin. 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
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
G is the transmitter antenna gain
TX i TX i
L is the transmitter loss ( L = L Total – DL )
A cell TX i ic is considered the best server of a mobile Mi if it satisfies the following conditions:
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.
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:
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:
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.
In other words, the direction of arrival for the served user Mi should not be the direction of arrival of an interfering
mobile.
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.
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
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 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 .
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
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 ):
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-
N t TCH – UL
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.
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.
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.
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
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
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 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 .
TX j jc TS Mi
RSCP Tot – DL Mi
All TX j
I IC – DL ic jc = ---------------------------------------------------------------------------
F IRF ic jc
TX i TX i Mi
1 – F JD and = 0 Without Useful Signal
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 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
TX i ic TS M i TX i ic TS M i
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
Model TX i SA TX i SA
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-
F IRF ic jc
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
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 ):
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-
N t TCH – DL
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
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.
Model TX i SA TX i SA
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 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.
For each mobile Mi interfered by Mj in the uplink by the connection between Mj and TX j ic , Atoll updates
TX i ic TS Mi
RSCP TCH – UL .
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.
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:
TX i ic TS M i TX i ic TS M i
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.
TX i ic TS M i TX i ic TS M i
If P R99 – DL P Max – DL – Eff the mobile with the lowest service priority is rejected for the reason "DL Load Saturation".
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.
TX i ic TS M i TX i ic TS M i
If X UL X Max – UL the mobile with the lowest service priority is rejected for the reason "UL Load Saturation".
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.
TX i ic TX i ic TX i ic TX i ic TX i ic
P Tot – DL = P R99 – DL + P HR + P HS – SCCH + P HS – PDSCH
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
TX i ic TX i ic
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:
TX i ic TX i ic TX i ic
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 TX i ic
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
1 – F JD and = 0 Without Useful Signal
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.
Model TX i SA TX i SA
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.
- Dynamic Allocation
HS-PDSCH power is calculated as follows:
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
TX i ic TX i ic TX i ic
Where P R99 – DL = P TCH – DL + P OCCH . The effective maximum cell traffic power per timeslot is calculated as:
TX i ic TX i ic TX i ic
P Max – DL – Eff = P Max – DL %P Max – DL . 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.
- Static Allocation
The static HS-SICH power is defined in the properties of the terminal used by the HSDPA mobile Mi.
- Dynamic Allocation
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
1 – F JD and = 0 Without Useful Signal
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.
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
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.
TX i ic TX i ic TX i ic TX i ic TX i ic
P Tot – DL = P R99 – DL + P HR + P HS – SCCH + P HS – PDSCH
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.
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.
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.
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.
= --------------------------------------------------------------------------------------------
-
TX i ic TS Mi
P Rec
k
• 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.
TX i ic
TX i ic P P – CCPCH
RSCP P – CCPCH = ----------------------------
-
LT
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
ji
TX j jc TX j jc
RSCPP – CCPCH + RSCPOCCH – TS0
TX
I IC – DL ic jc = -----------------------------------------------------------------------------------------------------------
j -
F IRF ic jc
I IC – DL ic jc is the inter-carrier interference from a carrier jc to another carrier ic on the downlink, which is reduced by
the interference reduction factor F IRF ic jc defined for the pair (ic, jc).
TX i ic
TX i ic P OCCH – TS0
RSCP OCCH – TS0 = -------------------------------
LT
TX i Term Eb Nt
L Path L L L Body L Indoor M Shadowing
L T = -------------------------------------------------------------------------------------------------------------------------------------
-
TX i Term
G G
TX i Term
and N 0 are defined in "Definitions and Formulas" on page 365.
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:
• 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
N t P – CCPCH I P – CCPCH
colour and displayed with intersections between layers.
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:
• 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.
TX i ic TX i ic
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 TX i ic
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
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
ji
TX j jc
RSCPDwPCH
TX
I IC – DL ic jc = -----------------------------------------------
j -
F IRF ic jc
I IC – DL ic jc is the inter-carrier interference from a carrier jc to another carrier ic on the downlink, which is reduced by
the interference reduction factor F IRF ic jc defined for the pair (ic, jc).
TX i Term Eb Nt
L Path L L L Body L Indoor M Shadowing
L T = -------------------------------------------------------------------------------------------------------------------------------------
-
TX i Term
G G
TX i Term
and N 0 are defined in "Definitions and Formulas" on page 365.
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:
• 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.
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
TX i ic
C
tab (Prediction properties). For each layer, a pixel is covered if ----
Req
Q DwPCH . Each layer is
I DwPCH
CECP
assigned a colour and displayed with intersections between layers.
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.
Model TX i SA TX i SA
In L T , G = G DL and L = L DL are calculated according to the smart antenna modelling method used, for
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
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:
• 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
intersections between layers.
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.
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
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
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:
• 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
intersections between layers.
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.
Model TX i SA TX i SA
In L T , G = G DL and L = L DL are calculated according to the smart antenna modelling method used, for
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.
C
---- . The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier,
I TCH – DL
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:
• 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.
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
N t TCH – DL I TCH – DL
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
N t TCH – DL
= ------------------------------------------------------
TX i ic DL G DL and ---
I TCH – DL
= -------------------------------------------------------
TX i ic
G DL
N Tot – DL N Tot – DL
Max
TX i ic P TCH – DL Service
With RSCP TCH – DL = ----------------------------------------------------
Eb Nt 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
In L T , G = G DL and L = L DL are calculated according to the smart antenna modelling method used, for
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.
TX i ic TX i ic TX i ic Term
N Tot – DL = I Intra – DL + I Extra – DL + I IC – DL ic jc + N 0
Where
TX i ic TX i 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 = -------------------------
-
Eb Nt DL
LT
TX i ic TX j ic TX j ic
I Extra – DL = RSCPTCH – DL + RSCPOCCH
ji
TX j jc TX j jc
RSCPTCH – DL + RSCPOCCH
TX
I IC – DL ic jc = ------------------------------------------------------------------------------------------
j -
F IRF ic jc
I IC – DL ic jc is the inter-carrier interference from a carrier jc to another carrier ic on the downlink, which is reduced by
the interference reduction factor F IRF ic jc defined for the pair (ic, jc).
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:
• Max Eb/Nt or Max C/I (dB)
Atoll calculates the Eb/Nt or C/I on each pixel of the TX i ic best server coverage area. 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
E b TX i ic C
TX i ic
layer, a pixel is covered if ------ Threshold or ---- Threshold . Each layer is assigned a
N t TCH – DL I TCH – DL
colour and displayed with intersections between layers.
C
---- . The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier,
I TCH – UL
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:
• 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.
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
N t TCH – UL I TCH – UL
covered and coloured according to the selected display option.
Term TX i ic Term TX i ic
E b TX i ic RSCP TCH – UL C-
TX i ic RSCP TCH – UL
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
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
In L T , G = G UL and L = L UL are calculated according to the smart antenna modelling method used, for
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
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:
• Max Eb/Nt or Max C/I (dB)
Atoll calculates the Eb/Nt or C/I on each pixel of the TX i ic best server coverage area. 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
E b TX i ic C
TX i ic
layer, a pixel is covered if ------ Threshold or ---- Threshold . Each layer is assigned a
N t TCH – UL I TCH – UL
colour and displayed with intersections between layers.
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.
E-----b- C
or ---- , and display the pixels where both downlink and uplink Eb/Nt or C/I are above the required
N t TCH – UL I TCH – UL
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-
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 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.
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
TX i ic TX i ic
• E-----b- Req C
Q TCH – DL or ----
Req
Q TCH – DL for any of the 6 timeslots
N t TCH – DL I TCH – DL
TX i ic TX i ic
• E-----b- Req C
Q TCH – UL or ----
Req
Q TCH – UL for any of the 6 timeslots
N t TCH – UL I TCH – UL
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 jc TX j jc
TX j ic TX j ic
RSCPTCH – DL + RSCPOCCH
RSCPTCH – DL + RSCPOCCH + ------------------------------------------------------------------------------------------
TX j
I C2C TX i TX j = -
F IRF ic jc
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.
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
each layer, a pixel is covered if I C2C TX i TX j Threshold . Each layer is assigned a colour and displayed with intersec-
tions between layers.
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.
TX i ic TX i ic
Pixels in the TX i ic coverage area where RSCP P – CCPCH TAdd P – CCPCH Mobility and I TS1 – UL Threshold are
covered and coloured according to the selected display option.
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
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
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:
• 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
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
workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction prop-
erties). Each layer is assigned a colour and displayed with intersections between layers.
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:
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
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
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.
Transmitter
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.
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.
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:
8.5.1.4.1 Downlink
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
j ------- d sin j ------- 2d sin j ------- E SA – 1 d sin
S = 1 e e ... e
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
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.
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, 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.
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
pj Sj Sj
2
RN = Rn + RI = n I +
j=1
J
H
pj Sj Sj
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.
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
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.
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.
The resulting geographic distribution is formed by linearly joining the obtained results.
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.
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.
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
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
satisfying the criterion.
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.
3. There must be an overlapping zone ( S A S B ) with a given cell edge coverage probability:
- 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.
- 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.
- 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.
SA SB
Atoll calculates the percentage of covered area ( ---------------------- 100 ), which it compares with the % minimum covered
SA
area. 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.
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 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
Neighbourhood cause When
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
Neighbourhood relationship that fulfils
If the % minimum covered area is exceeded IF
coverage conditions
Symmetric neighbourhood
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 overlapping factor (O) meaning the percentage of overlapping
The IF is user-definable using the Min importance and Max importance fields.
Yes Yes Min C + C Max O O + 100% – Max O A 60% + 40% 30% O + 70% A
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 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).
• 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).
• A forbidden neighbour must not be listed as neighbour except if the neighbourhood
relationship already exists and the Delete existing neighbours option is not selected. 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 cell has already reached its maximum number of neighbours before starting the new
allocation, it will not appear in the Results table.
- 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.
• Atoll can take into account inter-technology neighbour relations as constraints to allocate
different scrambling codes to the TD-SCDMA neighbours of a GSM transmitter. 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.
• Atoll considers symmetry relationship between a cell, its first order neighbours, its second
order neighbours and its third order neighbours.
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.
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.
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.
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.
All sites which have constraints with the studied site are referred to as near sites.
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.
For information on calculating site priority, see "Site Priority" on page 423. For information on calculating cell priority, see
"Cell Priority" on page 420.
All sites which have constraints with the studied site are referred to as near sites.
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-
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
"Cell Priority" on page 420.
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
• 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, Atoll displays a warning message in the Event Viewer.
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:
• 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:
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.
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:
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
C j N1 i = I j c N1
Where
c N1 is the cost of the first order neighbour constraint. This value can be defined in the Constraint Cost dialogue.
Because two first order neighbours must not have the same 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
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
order neighbours to be each other.
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.
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
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:
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
C i U = C i Dist + C i EP + C i N + C i N 2G + C i Cluster
C Tx = C Tx Dom + C Tx U
With C Tx U = Max C U and C Dom = 128 – Number of scrambling codes in the domain
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.
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.
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.
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.
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.
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 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.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you
may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.
Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current
neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
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 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).
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.
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 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.
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.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you
may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.
Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current
neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
3. There must be an overlapping zone ( S A S B ) with a given cell edge coverage probability.
- 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.
Two different cases may be considered for SB:
- 1st case: SB is the area where the cell B is the best serving transmitter of the GSM network.
In this case, the margin must be set to 0 dB.
- 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 0 dB and SB is the area where:
- The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and
is within a margin from the best BCCH signal level.
SA SB
Atoll calculates the percentage of covered area ( ---------------------- 100 ) and compares this value to the % minimum covered
SA
area. If this percentage is less than the minimum, the candidate neighbour B is discarded.
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of covered
area.
4. 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 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).
If there are 15 candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8.
Therefore, among 15 candidate neighbours, only 8 (those with the highest priority) will be allocated to the reference cell.
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 allocation cause for each neighbour. Therefore, a neighbour may
be marked as exceptional pair or co-site. If the neighbour is not forced but fulfills 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:
• 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.
• The percentage of covered area is calculated with the resolution specified in the properties
dialog of the Predictions folder (Default Resolution).
• A forbidden neighbour must not be listed as neighbour except if the neighbourhood
relationship already exists and the Delete existing neighbours option is not selected 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
allocation, it will not appear in the Results table.
8.9.1.3 Appendices
8.9.1.3.1 Delete Existing Neighbours Option
As explained above, Atoll keeps the existing inter-technology neighbours when the Delete existing neighbours option is
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
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.
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.
Atoll
RF Planning and Optimisation Software
Technical Reference Guide
9.1.1 Input
This table lists the input to computations, coverage predictions, and simulations.
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
D RTG Global parameter ms Receive Time Guard (TDD only)
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
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 )
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
bits/
B Bearer parameter Bearer Efficiency
symbol
N Users – Max Cell parameter None Maximum number of users per cell
DL
G STTD Clutter parameter dB Downlink STTD/MRC gain offset
Result calculated from cell edge coverage probability and model Model Shadowing margin
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
BTS BTS parameter %
% 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
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 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 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
TX i ic – TX j jc
--------------------------
- 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
TX i ic TX j jc
TX i ic – TX j jc
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
TX i ic TX j jc
if W Channel W Channel
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
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
N SCa – Total
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
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
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
–L – L Ant – L Body
TX i ic TX i ic TX i TX i
EIRP Traffic P Traffic + G –L dBm Traffic EIRP of a cell
TX i ic TX i ic TX i TX i
EIRP Pilot P Pilot +G –L dBm Pilot EIRP of a cell
TX i ic TX i ic TX i ic
P Traffic P Preamble – P Traffic dBm Traffic transmission power of a cell
TX i ic TX i ic TX i ic
P Pilot P Preamble – P Pilot dBm Pilot transmission power of a cell
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
TX j jc TX j jc TX j TX j
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 ----------------------------------------------------------------
----------------------------- -
TX j jc TX j jc
N SCa – Used N SCa – Used
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
TX i ic TX i ic
C Pilot – n DL
With MIMO (STTD/MRC):
TX i ic TX i ic DL DL
TX i ic CNR Pilot = CNR Pilot + G STTD + G STTD dB Pilot C/N for a cell
CNR Pilot
TX i ic TX i ic
With MIMO (AMS) if CNR Preamble T AMS :
TX i ic TX i ic DL DL
CNR Pilot = CNR Pilot + G STTD + G STTD
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
All TX j jc
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 i ic TX i ic
With MIMO (AMS) if CNR Preamble T AMS :
TX i ic TX i ic DL DL
CINR Traffic = CINR Traffic + G STTD + G STTD
TX jc
j TX i ic
TX i ic I DL n DL
-
-----------------
- -------------------
C Pilot – 10 Log 10 + 10 – M Shadowing – C I
10 10
All TX j jc
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
TX i ic TX i ic
With MIMO (AMS) if CNR Preamble T AMS :
TX i ic TX i ic DL DL
CINR Pilot = CINR Pilot + G STTD + G STTD
TX jc TX ic
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
All TX j jc
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
TX i ic N SCa – Used
WiMAX 802.16d: n 0 + 10 Log 10 F Sampling ------------------------------
N SCa – Total
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
TX i ic TX i ic TX i ic
n UL n 0 – UL + nf dBm Uplink noise for a cell
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
All TX jc
j
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
TX i ic TX i ic
With MIMO (AMS) if CNR Preamble T AMS :
Mi Mi UL UL
CNR UL = CNR UL + G STTD + G STTD
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
TX i ic TX i ic
D Symbol D Sym – Useful + D CP ms Symbol duration
D Used
TX i ic
Floor --------------------
Frame
- Frame duration in terms of symbol
N SD – Used Frame TXi ic SD
durations
D Symbol
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
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
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
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
TX ic Sel
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
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 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
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
E SA
----------------------------------
- None Smart antenna gain in the uplink
H –1
S RN S
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
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
Note:
Mi Mi Mi Mi
• L , G , L Ant , and L Body are not considered in the calculations performed for the point
analysis tool.
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.
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.
TX i ic TX i ic TX i ic
MinimumThreshold C Preamble or L Total or L Path MaximumThreshold
TX i ic TX i ic TX i ic
MinimumThreshold C Preamble or L Total or L Path MaximumThreshold
AND
TX i ic TX j jc
C Preamble Best C Preamble – M
ji
Where M is the specified margin (dB). The Best function considers the highest value from a list of values.
Notes:
• If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is
the highest.
• If M = 2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is
either the highest or within a 2 dB margin from the highest.
• 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.
TX i ic TX i ic TX i ic
MinimumThreshold C Preamble or L Total or L Path MaximumThreshold
AND
TX ic nd TX jc
C Preamble 2 Best C j
i
Preamble – M
ji
Where M is the specified margin (dB). The 2nd Best function considers the second highest value from a list of values.
Notes:
• If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is
the second highest.
• If M = 2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is
either the second highest or within a 2 dB margin from the second highest.
• If M = -2 dB, 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.
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.
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.
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.
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
consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
• 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.
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.
• Scheduling and Radio Resource Management as explained under "Simulation Process" on page 455.
9.2.5.1.1 Simulations Based on Raster Traffic Maps, Vector Traffic Maps, and Subscriber Lists
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, 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.
- 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
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
Probability of being inactive: p Inactive = 1 – f Act 1 – f Act
UL UL DL
Probability of being active in the uplink: p Active = f Act 1 – f Act
DL DL UL
Probability of being active in the downlink: p Active = f Act 1 – f Act
UL + DL UL DL
Probability of being active in the uplink and downlink both: p Active = f Act f Act
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.
- 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
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
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.
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.
At any given instant, Atoll calculates the probability for a user being active in the uplink and in the downlink as follows:
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
Probability of being inactive: p Inactive = 1 – f Act 1 – f Act
UL UL DL
Probability of being active in the uplink: p Active = f Act 1 – f Act
DL DL UL
Probability of being active in the downlink: p Active = f Act 1 – f Act
UL + DL UL DL
Probability of being active in the uplink and downlink both: p Active = f Act f Act
Calculation of the number of active users trying to access the 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
nv
Number of inactive users: n v – Inactive = ------------------------------ p Inactive
1 – p Inactive
Here, Atoll considers all the users as active. Activity probabilities are not calculated.
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.
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.
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.
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.
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
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.
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
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:
TX i ic TX i ic TX i ic
TL DL = Max TL DL – TL DL
k All TX ic
k k – 1
i
TX i ic TX i ic TX i ic
TL UL = Max TL UL – TL UL
k
All TX i ic k k – 1
TX i ic TX i ic TX i ic
NR UL = Max NR UL – NR UL
k All TX ic
k k – 1
i
TX i ic TX i ic TX i ic
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.
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:
TX i ic TX i ic TX i ic TX i ic TX i ic TX i ic
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.
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.
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.
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
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.
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.
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.
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.
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.
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
visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer corresponds to
an area where the effective MAC channel throughput exceeds a defined threshold.
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
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 application level channel throughput exceeds a defined threshold.
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.
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.
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.
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.
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
visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer corresponds to
an area where the effective MAC channel throughput exceeds a defined threshold.
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
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 application level channel throughput exceeds a defined threshold.
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
TX i ic TX i ic TX i ic
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
TX i ic TX j jc
• F Start – FB and F Start – FB : Start frequency of the frequency band assigned to the cells TXi(ic) and TXj(jc).
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 ).
TX i ic TX j jc
• 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.
TX i ic TX j jc
• 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):
TX i ic TX i ic TX i ic TX i ic
F Start = F Start – FB + N Channel W Channel
TX i ic TX i ic TX i ic TX i ic TX i ic
F End = F Start – FB + N Channel W Channel + W Channel
TX j jc TX j jc TX j jc TX j jc
F Start = F Start – FB + N Channel W Channel
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
TX i ic TX j jc
• F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc).
TX i ic TX j jc
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc).
TX i ic TX j jc
• F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
Numbers to Start and End Frequencies" on page 463.
TX i ic TX j jc
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
Numbers to Start and End Frequencies" on page 463.
Calculations
Atoll first verifies that co-channel overlap exists between the cells TXi(ic) and TXj(jc).
Output
TX i ic – TX j jc
• r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc).
TX i ic TX j jc
• F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
Numbers to Start and End Frequencies" on page 463.
TX i ic TX j jc
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
Numbers to Start and End Frequencies" on page 463.
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).
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
TX i ic – TX j jc
TX i ic – TX j jc W ACO
L
r ACO = --------------------------------------
-
L TX i ic
W Channel
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
TX i ic – TX j jc
TX i ic – TX j jc W ACO
H
r ACO = --------------------------------------
-
H TX i ic
W Channel
Output
TX i ic – TX j jc
• r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
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
TX i ic – TX j jc
• r FDD – TDD : FDD – TDD overlap ratio between the cells TXi(ic) and TXj(jc).
TX i ic – TX j jc
• r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co-Channel
Overlap Calculation" on page 464.
TX i ic – TX j jc
• r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Adjacent
Channel Overlap Calculation" on page 464.
TX i ic – TX j jc
• r FDD – TDD : FDD – TDD overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "FDD – TDD
Overlap Ratio Calculation" on page 465.
TX i ic
• f ACS – FB : Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic).
TX i ic TX j jc
• W Channel and W Channel : Bandwidths of the channels assigned to the cells TXi(ic) and TXj(jc).
Calculations
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
TX j jc TX j jc
converts X dBm over W Channel to Y dBm (which is less than X dBm) over less than W Channel .
Output
TX i ic – TX j jc
• rO : Total co- and adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
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).
• L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected.
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:
TX i ic TX i ic TX i TX i
• 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
TX i
L Path = L Model + L Ant
Furthermore, the total losses between the cell and the pixel, subscriber, or mobile Mi can be calculated as follows:
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.
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).
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 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
N SCa – Total
The preamble noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the
pixel, subscriber, or mobile Mi.
TX i ic TX i ic Mi
n Preamble = n 0 – Preamble + nf
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
N SCa – Total
The following table shows the different types of subcarriers and their numbers for preamble transmission in
WiMAX 802.16e.
Output
TX i ic
• n Preamble : Preamble noise for the cell TXi(ic).
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).
TX i ic – TX j jc
• rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent
Channel Overlaps Calculation" on page 462.
TX i ic TX j jc
• N Seg and N Seg : Segment numbers assigned to the cells TXi(ic) and TXj(jc) calculated from their respective
TX i ic TX j jc
preamble indexes ( n Preamble and n Preamble ) as follows:
n Preamble N Seg
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
TX i ic – TX j jc
Where f O is the interference reduction factor due to channel overlap between the cells TXi(ic) and TXj(jc),
calculated as follows:
TX i ic – TX j jc TX i ic – TX j jc
fO = 10 Log r O
TX i ic – TX j jc
And, f Seg is the interference reduction factor due to preamble segmentation (WiMAX 802.16e), calculated as
follows:
TX i ic – TX j jc TX i ic – TX j jc
f Seg = 10 Log p Collision
TX i ic – TX j jc
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
TX i ic TX j jc p Collision
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).
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:
TX i ic TX i ic TX i ic
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.
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.
This parameter is taken into account when the option "Shadowing taken into account" is selected.
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
All TX j jc
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
All TX j jc
Output
TX i ic
• CINR Preamble : Preamble C/(I+N) from the cell TXi(ic) at a pixel, subscriber, or mobile Mi.
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).
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
Input
TX i ic
• CNR Preamble : Preamble C/N from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Preamble
C/N Calculation" on page 470.
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:
TX i ic TX i ic
CNR Preamble T Preamble
Output
A pixel, subscriber, or mobile Mi which is unable to get a permutation zone is considered to be outside the service area.
Input
TX i ic
• d Max – PZ : Maximum distance covered by a permutation zone of a cell TXi(ic).
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:
TX i ic TX i ic
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.
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
• G : Transmitter antenna gain for the antenna used by the transmitter TXi.
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 ).
• 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 traffic and pilot signal levels for C/(I+N) calculations.
• L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected.
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
This means that L Ant = 0 for calculating the traffic and pilot signal level.
Mi
• L Body : Body loss defined for the service used by the pixel, subscriber, or mobile Mi.
Calculations
The received traffic and pilot signal levels (dBm) from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi
as follows:
TX i ic TX i ic Mi Mi Mi Mi
C Traffic = EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body and
TX i ic TX i ic Mi Mi Mi Mi
C Pilot = EIRP Pilot – 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:
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
TX i ic TX i ic
With P Traffic and P Pilot being the traffic and pilot transmission powers of the cell TXi(ic) calculated as follows:
TX i ic TX i ic TX i ic TX i ic TX i ic TX i ic
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.
Input
• K: Boltzmann’s constant.
• T: Temperature in Kelvin.
• 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
• 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
TX i ic TX i ic N SCa – Used
WiMAX 802.16d: n 0 – DL = n 0 + 10 Log 10 F Sampling ------------------------------
N SCa – Total
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
-
N SCa – Total
The downlink noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the
pixel, subscriber, or mobile Mi.
TX i ic TX i ic Mi
n DL = n 0 – DL + nf
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 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.
0 1/5
1 1/3
2 7/15
3 3/5
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).
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
• L Path : Path loss ( L Path = L Model + L Ant ).
• L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model.
TX j
• L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXj.
• L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected.
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.
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
• L Body : Body loss defined for the service used by the pixel, subscriber, or mobile Mi.
TX j jc
• TL DL : Downlink traffic load of the interfering cell TXj(jc).
Traffic loads can either be calculated using Monte Carlo simulations, or entered manually for each cell. Calculation
of traffic loads is explained in "Simulation Process" on page 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:
TX j jc TX j jc
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
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
TX j jc TX j jc Mi Mi Mi Mi
I Pilot = EIRP Pilot – L Path – L Indoor + G –L – L Ant – L Body
Where EIRP is the effective isotropic radiated power of the cell calculated as follows:
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
TX j jc TX j jc
With P Traffic and P Pilot being the traffic and pilot transmission powers of the cell TXj(jc) calculated as follows:
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:
TX j jc TX j jc
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 ----------------------------------------------------------------
----------------------------- -
TX j jc TX j jc
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:
TX j jc TX j jc Mi Mi Mi Mi
I AAS = EIRP AAS – L Path – L Indoor + G –L – L Ant – L Body
Where EIRP is the effective isotropic radiated power of the cell calculated as follows:
TX j jc TX j jc TX j TX j
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:
TX j jc TX j jc TX j jc
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:
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.
- 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.
The average array correlation matrix (at the interfering cell) are given by:
R Avg = j pj Rj
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
calculated as follows:
The received interfering pilot signal level (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile
Mi as follows:
TX j jc TX j jc Mi Mi Mi Mi
I Idle – Pilot = EIRP Idle – Pilot – L Path – L Indoor + G –L – L Ant – L Body
Where EIRP is the effective isotropic radiated power of the cell calculated as follows:
TX j jc TX j jc TX j TX j
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:
TX j jc TX j jc TX j jc
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).
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.
TX i ic – TX j jc
• rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent
Channel Overlaps Calculation" on page 462.
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:
TX i ic – TX j jc TX i ic – TX j jc
fO = 10 Log r O
If you select the Segmentation Support check box in the Frame Configurations table for the frame configuration
that you are using, it means that the first zone in the downlink, i.e., the DL PUSC zone, is segmented. All other
zones are pooled together to form a group of non-segmented zones. There are two effects of segmentation:
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.
Atoll determines the switching point between the segmented and the non-segmented zones using the seg-
mentation usage ratio.
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
TX i ic TX i ic TX i ic
SU + f Segment 1 – SU
TX j jc
TX j jc SU
SP = ------------------------------------------------------------------------------------------------
TX j jc TX j jc TX j jc
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.
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.
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:
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),
TX i ic TX i ic
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:
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
- 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:
TX i ic – TX j jc TX i ic – TX j jc
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).
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
"Traffic and Pilot Signal Level Calculation (DL)" on page 473.
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
C/N Calculation" on page 470.
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:
TX i ic TX i ic TX i ic
CNR Traffic = C Traffic – n DL
TX i ic TX i ic TX i ic
CNR Pilot = C Pilot – n DL
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.
In case of STTD/MRC:
TX i ic TX i ic DL DL
CNR Traffic = CNR Traffic + G STTD + G STTD
TX i ic TX i ic DL DL
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
TX i ic TX i ic DL DL TX i ic TX i ic
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.
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
• 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.
• M Shadowing – C I : Shadowing margin based on the C/I standard deviation.
• This parameter is taken into account when the option "Shadowing taken into account" is selected.
TX i ic
• CNR Preamble : Preamble C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Preamble
C/N Calculation" on page 470.
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 C/(I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc
j TX i ic
TX i ic TX i ic I DL n DL
-
------------------ -------------------
CINR Traffic = C Traffic – 10 Log 10 + 10 – M Shadowing – C I
10 10
All TX j jc
The Traffic Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX jc TX ic
j i
I DL n DL
TX i ic ------------------ --------------------
10 10
I + N DL = 10 Log 10 + 10
All TX j jc
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/(I+N).
In case of STTD/MRC:
TX i ic TX i ic DL DL
CINR Traffic = CINR Traffic + G STTD + G STTD
In case of AMS:
TX i ic TX i ic DL DL TX i ic TX i ic
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).
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
• G : Transmitter antenna gain for the antenna used by the transmitter TXi.
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
• L Path : Path loss ( L Path = L Model + L Ant ).
• L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model.
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 signal level from interfering mobiles.
• L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected.
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.
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
• L Body : Body loss defined for the service used by the pixel, subscriber, or mobile 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).
Input
• K: Boltzmann’s constant.
• T: Temperature in Kelvin.
• 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
• N SCa – Total : Total number of subcarriers defined for the frame configuration of a WiMAX 802.16e cell TXi(ic).
TX i ic
• F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on
page 493.
TX i ic
• nf : Noise figure of the cell TXi(ic).
Calculations
TX i ic TX i ic N SCa – Used
WiMAX 802.16d: n 0 – UL = n 0 + 10 Log 10 F Sampling ------------------------------
N SCa – Total
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
-
N SCa – Total
The uplink noise is the sum of the thermal noise and the noise figure of the cell TXi(ic).
TX i ic TX i ic TX i ic
n UL = n 0 – UL + nf
Output
TX i ic
• n UL : Uplink noise for the cell TXi(ic).
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.
TX i ic – TX j jc
• rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent
Channel Overlaps Calculation" on page 462.
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:
TX i ic – TX j jc TX i ic – TX j jc
fO = 10 Log r O
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,
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).
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
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
All TX jc
j
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
All TX jc
j
Output
TX i ic TX i 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
• I + N UL : Total Noise for a cell TXi(ic) calculated for any pixel, subscriber, or mobile Mi.
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
• CNR Preamble : Preamble C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Preamble
C/N Calculation" on page 470.
TX i ic
• T AMS : AMS threshold defined for the cell TXi(ic).
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)..
• Mobility M i : Mobility used for the calculations.
UL
• G STTD : Uplink STTD/MRC gain from the MIMO Configurations table 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 .
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
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 uplink C/N.
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
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.
Mi Mi
CNR UL = CNR UL + G SC – UL and reducing the number of subchannels used in the uplink by half.
50 to 25. 50
4 to 2 10 Log ------ = 3 dB
(48 Data + 2 Pilot to 25 Data + 1 Pilot) 25
25 to 13. 25
2 to 1 10 Log ------ = 2.84 dB
(24 Data + 1 Pilot to 12 Data + 1 Pilot) 13
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.
Subchannelisation decreases the bandwidth used by a mobile hence increasing the power spectral density for
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.
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:
- 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.
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).
Input
Mi
• CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
• M Shadowing – C I : Shadowing margin based on the C/I standard deviation.
This parameter is taken into account when the option "Shadowing taken into account" is selected.
TX i ic TX i 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
• CNR Preamble : Preamble C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Preamble
C/N Calculation" on page 470.
TX i ic
• T AMS : AMS threshold defined for the cell TXi(ic).
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)..
• Mobility M i : Mobility used for the calculations.
UL
• G STTD : Uplink STTD/MRC gain from the MIMO Configurations table 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 .
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/(I+N) for any pixel, subscriber, or mobile Mi at a cell TXi(ic) is calculated as follows:
In case of STTD/MRC:
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:
Mi Mi UL UL TX i ic TX i ic
CINR UL = CINR UL + G STTD + G STTD if CNR Preamble T AMS and
Mi Mi UL UL TX i ic TX i ic
CINR UL = CINR UL + G STTD + G STTD if CNR Preamble T AMS
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/(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.
50 to 25. 50
4 to 2 10 Log ------ = 3 dB
(48 Data + 2 Pilot to 25 Data + 1 Pilot) 25
25 to 13. 25
2 to 1 10 Log ------ = 2.84 dB
(24 Data + 1 Pilot to 12 Data + 1 Pilot) 13
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.
Subchannelisation decreases the bandwidth used by a mobile hence increasing the power spectral density for
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.
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.
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.
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.
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
Output
TX i ic
• F Sampling : Sampling frequency for the cell TXi(ic).
TX i ic
• F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on
page 493.
TX i ic
• N SCa – Total : Total number of subcarriers defined for the frame configuration of a WiMAX 802.16e cell TXi(ic).
• r CP : Cyclic prefix ratio defined for the network in the Global Parameters.
Calculations
r CP
D CP = --------
F
Adding the Cyclic prefix ratio to the useful symbol duration, Atoll determines the total symbol duration.
TX i ic TX i ic
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).
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
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
TX i ic TX i ic
• R DL = N Sym – DL Subframe : Amount of downlink resources in the cell TXi(ic).
TX i ic TX i ic
• R UL = N Sym – UL Subframe : Amount of uplink resources in the cell TXi(ic).
Input
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:
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
TX i ic TX i ic
• RX = N Sym – X Subframe : Amount of downlink or uplink resources in the cell TXi(ic).
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.
Input
TX i ic
• R DL : Amount of downlink resources in the cell TXi(ic) as calculated in "Calculation of Cell Capacity (Total
Resources)" on page 493.
TX i ic
• R UL : Amount of uplink resources in the cell TXi(ic) as calculated in "Calculation of Cell Capacity (Total
Resources)" on page 493.
• M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the downlink
i
B DL
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.
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
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,
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
Max TX i ic TX i ic
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.
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
Max TX i ic TX i ic
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
Mi Mi f TP – Scaling 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.
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-
culation" on page 493.
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).
TX i ic TX i ic TX i 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.
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
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
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 :
TX i ic TX i ic
Where N N Users , if there are some Best Effort users, or N = N Users if there are no Best Effort users selected.
Sel 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
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
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:
TX i ic TX i ic
R Rem – DL R Rem – UL
------------------------
- and ------------------------
-
N N
b. Atoll converts the remaining throughput demands of all the users to their respective remaining resource 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
CTP P – DL CTP P – UL
Remaining resource demands of a user are given by the ratio between its remaining throughput demands and
the peak channel throughputs at the user’s location.
c. The resources allocated to each user by the Proportional Fair scheduling method for satisfying its maximum
throughput demands are:
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.
g. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satis-
TX i ic TX i ic
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
Mi TPD Rem – DL Mi TPD Rem – UL
RD Rem – DL = ---------------------------------
Sel
and RD Rem – UL = ---------------------------------
Sel
Mi Mi
CTP P – DL CTP P – UL
Remaining resource demands of a user are given by the ratio between its remaining throughput demands and
the peak channel throughputs at the user’s location.
b. Atoll calculates the amount effective remaining resources for the cell of each user to distribute among the us-
ers as follows:
TX i ic
TX ic Sel
Mi
TX i ic
TX ic Sel
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
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:
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
Resource Allocation:
Once the remaining resources available for the users of each QoS class have been determined, the allocation of
resources within each QoS class is performed as for the proportional fair scheduler.
Sel
Let the number of users belonging to a QoS class N QoS M i .
a. Atoll divides the remaining resources of the QoS class into equal parts for each user:
TX i ic TX i ic
R QoS – DL R QoS – UL
------------------------ and ------------------------
N QoS N QoS
b. Atoll converts the remaining throughput demands of all the users to their respective remaining resource 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
CTP P – DL CTP P – UL
Remaining resource demands of a user are given by the ratio between its remaining throughput demands and
the peak channel throughputs at the user’s location.
c. The resources allocated to each user by the Biased scheduling method for satisfying its maximum throughput
demands are:
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
TX i ic TX i ic
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
Mi TPD Rem – DL Mi TPD Rem – UL
R Max – DL = ---------------------------------
Sel
and R Max – UL = ---------------------------------
Sel
Mi Mi
CTP P – DL CTP P – UL
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
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 .
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-
culation" on page 493.
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
Mi Mi f TP – Scaling Mi
• Application Level User Throughput: UTP A – DL = UTP E – DL ----------------------------- – TP Offset
100
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
Mi Mi f TP – Scaling Mi
• 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 .
9.3.9.1 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).
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.
9.3.9.2 Downlink
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
j ------- d sin j ------- 2d sin j ------- E SA – 1 d sin
S = 1 e e ... e
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
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 is able to form the beam only in the horizontal plane, therefore, the vertical pattern is assumed to
remain the same.
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
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.
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).
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
pj Sj Sj
2
RN = Rn + RI = n I +
j=1
J
H
pj Sj Sj
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.
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
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.
2
I UL + n
NR UL = ----------------------------
-
2
n
We assume a reference cell TXi(ic) and a candidate neighbour cell TXj(jc). When automatic allocation starts, Atoll checks
the following conditions:
1. The distance between both cells must be less than the user-definable maximum inter-site distance. If the distance
between the reference cell and the candidate neighbour is greater than this value, then the candidate neighbour
is discarded.
2. The calculation options,
- 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.
- 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.
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
displays a warning in the Event viewer.
- 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.
- 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
TX i ic TX i ic
CNR Preamble and T Preamble , respectively, by adding the value of the noise to them.
TX i ic TX i ic
- 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.
• The percentage of covered area is calculated with the resolution specified in the properties
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
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
and in case of a new allocation importance
Exceptional pair Only if the Force Exceptional Pairs option is selected 100 %
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 %
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)}
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.
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
(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:
• They are active,
• Their status is not set to locked,
• 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:
• If no focus zone exists in the .atl document, Atoll takes into account the computation zone.
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
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
TX i ic – TX j jc
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
Next, Atoll calculates the importance of the relation between the TBA cell and its related cell.
TX i ic – TX j jc TX i ic – TX j jc TX i ic – TX j jc
Total = Neighbour Neighbour + Dis tan ce Dis tan ce
TX i ic – TX j jc
Where Neighbour is the importance of the relationship between the TBA and its related neighbour cell, and
TX i ic – TX j jc
Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance
between them.
TX i ic – TX j jc
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.
TX i ic – TX j jc
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-
TX i ic – TX j jc
matic allocation dialogue, and D is the weighted distance between the TBA cell TXi(ic) and its related cell
TXj(jc) calculated as follows:
TX i ic – TX j jc TX i ic – TX j jc
D = d 1 + x cos – cos – 2
TX i ic – TX j jc
D is weighted according to the orientations of the TBA and its related cell with respect to the straight line
TX i ic – TX j jc
joining them. d is the distance between the two cells considering any offsets with respect to the site locations.
TX i ic – TX j jc
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.
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.
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
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
• 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
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.
Atoll
RF Planning and Optimisation Software
Technical Reference Guide
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,
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 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)
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
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
G amp is the amplification gain of the repeater,
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)
Rpk Rpk – coverage – side Rpk – coverage – side
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.
Where,
Txd
G ant is the gain of the donor transmitter antenna,
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,
Rpk
G amp is the amplification gain of the repeater,
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.
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.
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:
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,
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.
Where,
Txd
P pilot ic is the pilot power of the donor transmitter on the carrier ic,
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),
Rpk
L path is the path loss between the repeater Rpk and the receiver,
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 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)
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)
Where,
Txd – Rpk
L MW is the user-defined microwave link loss between the donor transmitter and the repeater,
Rpk
G amp is the amplification gain of the repeater,
Rpk Txd – Rpk Rpk Rpk – coverage – side Rpk – coverage – side
G total – MW – UL = L MW + G amp + G ant – L feeder – Rec (in dB)
Where,
Txd – Rpk
L MW is the user-defined microwave link loss between the donor transmitter and the repeater,
Rpk
G amp is the amplification gain of the repeater,
Where,
Txd
P pilot ic is the pilot power of the donor transmitter on the carrier ic,
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,
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 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)
Rpk Txd – Rpk Rpk Rpk – coverage – side Rpk – coverage – side
G total – Fibre – DL = L Fibre + G amp + G ant – L feeder – Tr (in dB)
Where,
Txd – Rpk
L Fibre is the user-defined fibre link loss between the donor transmitter and the repeater,
Rpk
G amp is the amplification gain of the repeater,
Rpk Txd – Rpk Rpk Rpk – coverage – side Rpk – coverage – side
G total – Fibre – UL = L Fibre + G amp + G ant – L feeder – Rec (in dB)
Where,
Txd – Rpk
L Fibre is the user-defined fibre link loss between the donor transmitter and the repeater,
Rpk
G amp is the amplification gain of the repeater,
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:
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:
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.
Rpk – coverage – side
G ant is the gain of the repeater coverage side antenna
max
P is the maximum downlink power allowed by the equipment.
Rpk – coverage – side
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.
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)
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
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
Where,
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
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)
And,
Txd
N 0 = NF KTW
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 KTW
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),
Rpk
L path is the path loss between the repeater Rpk and the receiver,
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 in point analysis or coverage predictions.
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)
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)
Rpk Rpk – coverage – side Rpk – coverage – side
G amp + G ant – L feeder – Tr
Where,
Txd
P is the power of the donor transmitter,
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 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
G amp is the amplification gain of the repeater,
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.
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.
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:
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,
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.
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),
Rpk
L path is the path loss between the repeater Rpk and the receiver,
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 in point analysis or coverage predictions.
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)
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
P is the power of the donor transmitter,
Txd – Rpk
L MW is the user-defined microwave link loss between the donor transmitter and the repeater,
Rpk
G amp is the amplification gain of the repeater,
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),
Rpk
L path is the path loss between the repeater Rpk and the receiver,
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 in point analysis or coverage predictions.
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)
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)
Where,
Txd
P is the power of the donor transmitter,
Txd – Rpk
L Fibre is the user-defined fibre link loss between the donor transmitter and the repeater,
Rpk
G amp is the amplification gain of the repeater,
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:
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.
max
P is the maximum downlink power allowed by the equipment.
Where,
Txd
P pilot ic is the pilot power of the donor transmitter on the carrier ic,
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,
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 in point analysis or coverage predictions.
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)
Where,
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,
Ra – coverage – side
L feeder – Tr corresponds to the losses of the remote antenna 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
Where,
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,
Ra – coverage – side
L feeder – Rec corresponds to the losses of the remote antenna 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.
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),
Rpk
L path is the path loss between the remote antenna Ra and the receiver,
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 in point analysis or coverage predictions.
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)
Where,
Txd
P is the power of the donor transmitter,
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,
Ra – coverage – side
L feeder – Tr corresponds to the losses of the remote antenna 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.
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
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
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
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
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
Release 2.7.1