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 HiTech 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: MultiResolution Management...................................................................... 80
4.4 Propagation Models ............................................................................................................................... 82
4.4.1 OkumuraHata and CostHata 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 5293 Propagation Model .......................................................................................................... 84
4.4.2.1 ITU 5293 Path Loss Formula..................................................................................................... 84
4.4.2.2 Corrections to the ITU 5293 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 ITUR P.5265 Propagation Model ................................................................................................... 95
4.4.5.1 ITU 5265 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 TDSCDMA Networks....................................................................365
8.1 Definitions and Formulas ......................................................................................................................365
8.1.1 Inputs ..............................................................................................................................................365
8.1.2 PCCPCH 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 PCCPCH 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 PCCPCH Coverage Prediction.............................................................................374
8.2.2.4 PCCPCH 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 TDSCDMA Prediction Studies ............................................................................................................ 394
8.4.1 PCCPCH 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 NFrequency Mode and CarrierType Allocation.................................................................................. 413
8.6.1 Automatic CarrierType 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 MultiCarrier 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 Intercarrier Neighbours  1st Case ....................................................................... 256
Figure 6.13: Overlapping Zone for Intercarrier Neighbours  2nd Case ...................................................................... 256
Figure 6.14: Neighbourhood Constraints...................................................................................................................... 262
Figure 6.15: Primary Scrambling Codes Allocation ...................................................................................................... 264
Figure 6.16: InterTransmitter Distance Computation .................................................................................................. 270
Figure 7.1: IS95 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: TDSCDMA 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: Intracarrier 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: InterTransmitter 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: CoChannel 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 georeferencing 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 georeferencing 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 multiuser 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 standalone 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 multiuser 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 IS95CDMA, 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 packetswitched 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, FAXCCITT3 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 (multisectored 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 ‘transmitterTRX 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 multiband 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 realtime 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 .bmpfile.
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 = 4bit palletized. # of colours = 16
29 2 BitCount biBitCount
8 = 8bit palletized. # of colours = 256
16 = 16bit palletized. # of colours = 65536
24 = 24bit 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 8bit RLE encoding
2 = BI_RLE4 4bit 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 1bit 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 4bit 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 8bit mode every byte represents a pixel. The value points to an entry in the colour
table which contains 256 entries.
 In 24bit 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 RGBvalue, 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 (16bit or 32bit 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
1bit 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
32bit 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
4bit
4 0 to be the 16 MSWindows standard colours. Padding each line with
16 colour images
zeros up to a 32bit boundary will result in up to 28 zeros = 7 'wasted
pixels'.
Every byte holds 1 pixel. There are 256 colour table entries.
8bit
8 0 Padding each line with zeros up to a 32bit 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.
16bit
16 0 Padding each line with zeros up to a 16bit 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
24bit 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 2byte 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
4bit will have interleaved highorder 4bits and low order 4 bits
4 2
16 colour images (ABABA...). If the first byte is zero, the second defines an escape
code. The EndofBitmap is zero padded to end on a 32bit
boundary. Due to the 16bitness of this structure this will always be
either two zero bytes or none.
The pixel data is stored in 2byte chunks. The first byte specifies the
number of consecutive pixels with the same colour. The second byte
8bit defines their colour indices. If the first byte is zero, the second
8 1
256 colour images defines an escape code. The EndofBitmap is zero padded to end
on a 32bit boundary. Due to the 16bitness of this structure this will
always be either two zero bytes or none.
0 0 Endofline
0 1 EndofBitmap
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 16bit 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):
Australian1965
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:BDC6E3F06DA311d1A2A300AA00C14882'
xmlns:dt='uuid:C2F4101065B311d1A29F00AA00C14882'
xmlns:rs='urn:schemasmicrosoftcom: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:BDC6E3F06DA311d1A2A300AA00C14882'
xmlns:dt='uuid:C2F4101065B311d1A29F00AA00C14882'
xmlns:rs='urn:schemasmicrosoftcom: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 subcellinterferer 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 userspecified 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 subcellinterferer 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:
########
#TransmitterTransmitterBCCH duringBSIC during% of vic'% of int'
#Name Identifier calculationcalculationcoverage 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 subcellinterferer 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 transmitterreceiver profile and to get predictions for a userdefined
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 transmitterreceiver 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 userdefined 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 IS95CDMA documents, P Tx = P Pilot and L Tx = L total – DL .
• In UMTS, CDMA2000 and IS95CDMA 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 userdefined 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: CostHata 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 CostHata 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
100400 MHz 303000 MHz 3010000 MHz 3010000 MHz 1503500 MHz 3001500 MHz 19006000 MHz 1502000 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 EpsteinPeterson (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.
OkumuraHata CostHata
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 Hatabased 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 5293 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 5293 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 Hatabased 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 userdefined 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 xaxis and yaxis respectively represent positions and heights. We assume that xaxis 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 transmitterreceiver profile. They are detailed in the
Appendices.
Along the transmitterreceiver 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 transmitterreceiver 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 semideterministic (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 semideterministic clutter information, specify receiver clearance (m) per clutter class. Both ground altitude and
clutter height are considered along the whole transmitterreceiver 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 Hatabased 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 leastsquare 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: {659F0B9E28104e599F0DDA9E78E1E64B}
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 transmitterreceiver 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 semideter
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 transmitterreceiver 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 userdefined 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 3707 (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 ErcegGreenstein 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. 12051211.
[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
May1 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 usermodifiable 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 ErcegGreenstein (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 KnifeEdge 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 transmitterreceiver 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 transmitterreceiver 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 transmitterreceiver distance is less than the smoothEarth 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 transmitterreceiver distance is greater than or equal to the smoothEarth 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 transmitterreceiver 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 (frequencydependent 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 transmitterreceiver 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 transmitterreceiver 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 transmitterreceiver 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 transmitterreceiver
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 SakagamiKuboi 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 SakagamiKuboi 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 SakagamiKuboi 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 SakagamiKuboi 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 SakagamiKuboi 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 PerezArriaga, "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 TxRx axis).
Hence,
2
For 1 knifeedge method, if – 0.7 , J = 6.9 + 20 log – 0.1 + 1 + – 0.1
Else, J = 0
Note:
• In case of multipleknife 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 TxRx 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 knifeedge section, another half, constituted by the knifeedge
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 5265 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 5265 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 transmitterreceiver 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 crosssections of the 3D
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 3D
antenna pattern.
Let’s take an example of an antenna pattern to be smoothed, as shown in Figure 4.21: on page 112. Let DPeaktoNull 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 DPeaktoNull 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 uptodate 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. 180198
[3] Jhong S., Leonard M. “CDMA systems engineering handbook” pp. 309315, 10511053”
[4] Remy J.G., Cueugnet J., Siben C. “Systèmes de radiocommunications avec les mobiles” pp. 309310
[5] Laiho J., Wacker A., Novosad T. “Radio network planning and optimisation for UMTS” pp. 8081
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 LevelBased Predictions
Signal levelbased 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.
• InterferenceBased Predictions
Interferencebased 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 MonteCarlo Simulations" on page 117, and applied to signal level, Ec/I0, or Eb/Nt as
explained below.
• Signal LevelBased Predictions
Signal levelbased 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+noiseBased Predictions
Interference+noisebased 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 intracellular 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
• MacroDiversity Gains
UL DL
Atoll calculates the uplink and downlink macrodiversity 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 "MacroDiversity Gains Calculation" on page 118.
• MonteCarlo Simulations
Random values for shadowing margins are calculated for each transmitterreceiver link and added to the predicted
path loss. A shadowing margin for each transmitterreceiver 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.
TDSCDMA Documents
The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 116 and
"Shadowing Margin Calculation in MonteCarlo Simulations" on page 117, and applied to signal level or interference+noise
predictions as explained below.
• Signal LevelBased Predictions
Signal levelbased predictions include coverage predictions (Best Server and RSCP PCCPCH 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+noiseBased Predictions
Interference+noisebased predictions include coverage predictions (PCCPCH 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
• MonteCarlo Simulations
Random values for shadowing margins are calculated for each transmitterreceiver link and added to the predicted
path loss. A shadowing margin for each transmitterreceiver 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 MonteCarlo Simulations" on page 117 , and applied to signal level or C/(I+N) as
explained below.
• Signal LevelBased Predictions
Signal levelbased 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+noiseBased Predictions
Interferencebased 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.
• MonteCarlo Simulations
Random values for shadowing margins are calculated for each transmitterreceiver link and added to the predicted
path loss. A shadowing margin for each transmitterreceiver 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 userdefined standard deviation of the error,
• G(0,1) is a zeromean unitvariance 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 interferencebased 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 CDMAbased 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 zeromean 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 transmitterreceiver pair. This values represents the shadowing error
Path
not related to the location of the receiver ( E Shadowing – model ). These values also have a zeromean 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 transmitterreceiver links are uncorrelated, the method
influences the calculated macrodiversity 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, IS95CDMA), 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 cosite 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 IS95CDMA), 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 1xEVDO, IS95 cdmaOne, TDSCDMA, WiMAX 802.16d, and WiMAX 802.16e documents,
• Transmitter total losses L Total in GSM GPRS EGPRS documents.
In Atoll, the transmitterequipment 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 IS95CDMA 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 transmitterreceiver 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, carriertointerference 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 enduser 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 24timeslot 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, Xaxis)
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, Xaxis)
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 halfrate circuit switched traffic by taking into account a userdefined percentage of halfrate
traffic. Atoll computes the effective equivalent number of fullrate timeslots that will be required to carry the total traffic with
the defined percentage of halfrate traffic.
If the number of timeslots required to accommodate the fullrate circuit switched traffic is TSreq. FR, and the percentage of
halfrate traffic within the subcell is defined by HR, then the effective number of equivalent fullrate 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 halfrate circuit switched traffic, which provides approximately the
same results as obtained by using the halfrate 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
infinitelength 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 carriertointerference ratio. According to the measured carrier power or carriertointerference 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
carriertointerference ratio C/I, depending on the userdefined 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 usermanageable 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 cosite transmitters as neighbours option is
Cosite 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:
• Cosite 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 userdefinable 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%
Cosite 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, cosite, adjacent, coverage or symmetric. For neighbours
accepted for cosite, 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 cochannels, 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 cochannels, 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,