IEEE C802.

20-04/xx
IEEE P 802.20/PD<insert PD Number>/V<07>
Date: <March 15, 2005>

Draft 802.20 Permanent Document

Channel Models for IEEE 802.20 MBWA System

Simulations Rev 08

ThisdocumentisaDraftPermanentDocumentofIEEEWorkingGroup802.20.PermanentDocuments
(PD)areusedinfacilitatingtheworkoftheWGandcontaininformationthatprovidesguidanceforthe
developmentof802.20standards.Thisdocumentisworkinprogressandissubjecttochange.

{03/15/2004}

IEEEP802.20PD<number>/V<number>

Contents
1

Introduction..............................................................................................................................................5
1.1

Purpose.............................................................................................................................................5

1.2

Scope................................................................................................................................................5

1.3

Abbreviations...................................................................................................................................5

SISOChannelModels..............................................................................................................................5
2.1

LinkLevelSimulation......................................................................................................................6

2.2

SystemLevelSimulation..................................................................................................................7

MIMOChannelModels...........................................................................................................................7
3.1

Introduction......................................................................................................................................7

3.2

SpatialChannelCharacteristics........................................................................................................7

3.3

MIMOChannelModelClassification..............................................................................................8

3.4

MBWAChannelEnvironments.......................................................................................................8

3.5

MIMOCorrelationChannelMatrices..............................................................................................9

3.5.1

DefinitionofCorrelationChannelMatrices............................................................................9

3.5.2

ProceduretoGenerateCorrelationMatrixCoefficients........................................................10

3.6
3.6.1

LinkLevelSpatialChannelModelParameterSummarya...........................................................12

LinkLevelChannelModelParameterSummary..................................................................12

MIMOChannelModelforSystemLevelSimulations..........................................................................14
4.1

Introduction....................................................................................................................................14

Appendix................................................................................................................................................17
5.1

AnExampleofHowtheMIMOChannelModelsCollapsetoSISOModels...............................17

5.2

SpatialParametersfortheBaseStation.........................................................................................18

5.2.1

BSAntennaTopologies..........................................................................................................18

5.2.2

BSAnglesofDepartureandArrival......................................................................................20

5.2.3

BSAngleSpread....................................................................................................................21

ii

{03/15/2004}
5.2.4

BSPowerAzimuthSpectrum................................................................................................21

5.3

SpatialParametersfortheMobileStation.....................................................................................21

5.3.1

MSAntennaTopologies.........................................................................................................22

5.3.2

MSAngleSpread...................................................................................................................22

5.3.3

MSAngleofArrival...............................................................................................................22

5.3.4

MSPowerAzimuthSpectrum................................................................................................22

5.3.5

MSDirectionofTravel..........................................................................................................23

5.3.6

DopplerSpectrum...................................................................................................................23

5.4

Definitions,Parameters,andAssumptions....................................................................................23

5.5

MIMOChannelEnvironments.......................................................................................................25

5.6

GeneratingSCMParameters..........................................................................................................28

5.6.1

GeneratingModelParametersforUrbanandSuburbanMacrocellEnvironments...............28

5.6.2

GeneratingModelParametersforUrbanMicrocellEnvironments.......................................29

5.7
6

IEEEP802.20PD<number>/V<number>

GeneratingSCMCoefficients.......................................................................................................29

References..............................................................................................................................................30

iii

{03/15/2004}

IEEEP802.20PD<number>/V<number>

Introduction

ThisdocumentdescribestheSISOandMIMOradiochannelmodelsthataretobeusedforsimulating
proposalsforthefuture802.20standard.Inspecifyingthesemodels,wehavetriedtoaddressvarious
commentsandinputsfromallthe802.20participants,whohaveexpressedtheiropiniononthissubject.
Efforts have been made to makesure that theMIMO channel modelshave appropriate delay spread,
Dopplerspread,andspatialcharacteristicsthataretypicalofthelicensedbandsbelow3.5GHz.Theeffort
tokeepbackwardcompatibilitywithstandardizedITUSISOmodels[25]hasalsobeenmadeduringthe
selectionofchanneldelayprofiles.Proposalsareevaluatedandcomparedbasedonthechannelmodel
methodologies described in this document. This facilitates comparison with documented performance
resultsofexistingsystemsusingtheITUmodels.

1.1 Purpose
ThisdocumentspecifieschannelmodelsforsimulationsofMBWAAirInterfaceschemesatlinklevel,as
wellassystemlevel.

1.2 Scope
Thescopeofthisdocumentistodefinethespecificationsofmobilebroadbandwirelesschannelmodels.

1.3 Abbreviations
AoA
AoD
AS
BS
DoT
DS
MEA
MIMO
MISO
MS
PAS
PDP
PL
Rx
SCM
SISO
SIMO
TE
Tx
ULA

AngleofArrival
AngleofDeparture
AngularSpread
BaseStation
DirectionofTravel
DelaySpread
MultiElementArray
MultipleInputMultipleOutput
MultipleInputSingleOutput
MobileStation
PowerAzimuthSpectrum
PowerDelayProfile
PathLoss
Receiver
SpatialChannelModel
SingleInputSingleOutput
SingleInputMultipleOutput
TestEnvironment
Transmitter
UniformLinearArray

{03/15/2004}

IEEEP802.20PD<number>/V<number>

2 SISO Channel Models

SISO systems shall use the ITU model in simulations.

2.1 Link Level Simulation

Models

casei

caseii

caseiii

caseIv

PDP

PedestrianA

VehicularA

PedestrianB
(PhaseI)

VehicularB (Phase
I)

4
NumberofPaths

Delay(ns)

RelativePathpower(dB)

0
0

2.5

110

1.0

310

0.9

200

300

190

9.0

710

4.9

800

12.8

8900

410

10.0

1090

8.0

1200

10.0

12900

15.0

1730

7.8

2300

25.2

17100

20.0

2510

23.9

3700

16.0

20000

-9.7

-19.2
-22.8

2) 3, 30, 120
Speed(km/h)

30, 120, 250

3,

[30],120,250

[Ed.Note:
Subject

to
consistency with
EVdoc]

[Ed.Note: Subjectto
consistencywithEV
doc]

Table 2.1-1 Summary of SISO Channel Model Parameters

{03/15/2004}

IEEEP802.20PD<number>/V<number>

2.2 System Level Simulation

ChannelScenario

Suburban
Macro
(PhaseI)

Urban
Macro

UrbanMicro

IndoorPico

Numberofpaths(N)
Lognormal shadowing
standarddeviation

6
10dB

6,11
10dB

6,12
NLOS:12dB
LOS:4dB

Pathlossmodel(dB),
disinmeters

31.5

+
35log10(d)

34.5 +
35log10(d)

6,11
NLOS:10dB
LOS:4dB
NLOS:34.53+38log10(d)
LOS:30.18

+
26*log10(d)

n 2 0.46

n 1

37 30 log10 R 18.3n

Table2.21SISOChannelEnvironmentParameters
PleaseseeAppendixforanexampleofhowthecorrelationmatrixapproachtoMIMOchannelmodels
collapsestotheITURmodelforSISOsystems.

MIMO Channel Models

3.1 Introduction
In this Chapter, a set of spatial channel model parameters are specified that have been developed to
characterizetheparticularfeaturesofMIMOradiochannels.SISOchannelmodelsprovideinformationon
thedistributionsofsignal powerlevel andDopplershiftsofreceivedsignals.MIMOchannelmodels,
which are based on the classical understanding of multipath fading and Doppler spread, incorporate
additionalconceptssuchasAngularSpread,AngleofArrival,PowerAzimuthSpectrum(PAS),andthe
antennaarraycorrelationmatricesforthetransmitter(Tx)andreceiver(Rx)combinations.

3.2 Spatial Channel Characteristics

Mobilebroadbandradiochannelisachallengingenvironment,inwhichthehighmobilitycausesrapid
variationsacrossthetimedimension,multipathdelayspreadcausesseverefrequencyselectivefading,
andangularspreadcausessignificantvariationsinthespatialchannelresponses.Forbestperformance,the
Rx& Tx algorithms must accurately track all dimensions ofthe channel responses (space, time, and
frequency). Therefore, a MIMO channel model must capture all the essential channel characteristics,
including

Spatialcharacteristics(AngleSpread,PowerAzimuthSpectrum,Spatialcorrelations),

Temporalcharacteristics(PowerDelayProfile),

Frequencydomaincharacteristics(Dopplerspectrum).

{03/15/2004}

IEEEP802.20PD<number>/V<number>

In MIMO systems, the spatial (or angular) distribution of the multipath components is important in
determiningsystemperformance.Systemcapacitycanbesignificantlyincreasedbyexploitingrichmulti
pathscatteringenvironments.

3.3 MIMO Channel Model Classification

There are three main approaches to MIMO channel modeling: the correlation model, the raytracing
model,andthescatteringmodel.Thepropertiesofthesemodelsarebrieflydescribedasfollows:

Correlation Model:Thismodel characterizesspatial correlationbyalinearcombiniationof

independent complex channel matrices at the transmitter and receiver. For multipath fading
channels,theITU(SISO)model[25]isusedtogeneratethepowerdelayprofileandDoppler
spectrum.SincethismodelisbasedonITUgeneralizedtapdelaylinechannelmodel,themodel
issimpletouseandbackwardcompatiblewithexistingITUchannelprofiles.

RayTracingModel:Inthisapproach,exactlocationsoftheprimaryscatterers,theirphysical
characteristics,aswellastheexactlocationofthetransmitterandreceiverareassumedknown.
Theresultingchannelcharacteristicsarethenpredictedbysummingthecontributionsfroma
largenumberofthepropagationpathsfromeachtransmitantennatoeachreceiveantenna.This
techniqueprovidesfairlyaccuratechannelpredictionbyusingsitespecificinformation,suchas
databaseofterrainandbuildings.Formodelingoutdoorenvironmentsthisapproachrequires
detailedterrainandbuildingdatabases.

ScatteringModel:Thismodelassumesaparticularstatisticaldistributionofscatterers.Using
thisdistribution,channelmodelsaregeneratedthroughsimulatedinteractionofscatterersand
planarwavefronts.Thismodelrequiresalargenumberofparameters.

3.4 MBWA Channel Environments

ThefollowingchannelenvironmentsshallbeconsideredforMBWAsystemsimulations:
1.

2.

Suburbanmacrocell
a.

Largecellradius(approximately16kmBStoBSdistance)

b.

HighBSantennapositions(aboverooftopheights,between1080m(typically32m))

c.

Moderatetohighdelayspreadsandlowanglespreads

d.

Highrangeofmobility(0250km/h)

Urbanmacrocell
a.

Largecellradius(approximately16kmBStoBSdistance)

b.

HighBSantennapositions(aboverooftopheights,between1080m(typically32m))

c.

Moderatetohighdelayandanglespread

{03/15/2004}
d.
3.

4.

IEEEP802.20PD<number>/V<number>
Highrangeofmobility(0250km/h)

Urbanmicrocell
a.

Smallcellradius(approximately0.30.5kmBStoBSdistance)

b.

BSantennapositions(atrooftopheightsorlower(typically12.5m))

c.

Highanglespreadandmoderatedelayspread

d.

Mediumrangeofmobility(0120km/h)

e.

The model is sensitive to antenna height and scattering environment (such as street
layout,LOS)

Indoorpicocell
a.

Verysmallcellradius(approximately100mBStoBSdistance)

b.

Bothbasestationsandpedestrianusersarelocatedindoor

c.

Highangularspreadandverylowdelayspread

d.

Lowrangeofmobility(03km/h)

e.

Themodelissensitivetoantennaheightsandscatteringenvironment(suchaswalls,
floors,andmetallicstructures)

3.5 MIMO Correlation Channel Matrices

Non-SISO systems will use a correlation matrix approach in simulations. The correlation matrices will be
only antenna system dependent. The correlation matrices [shall/may] be generated by using the SCM to
generate the correlation coefficients. The matrices used shall be submitted as part of the simulation report.
3.5.1 Definition of Correlation Channel Matrices
Bythecorrelationmatrixapproach,thechannelattheMreceiveantennaelementsisrelatedtotheN
independent channels that are associated with the N transmit antennae via a correlation matrix of
dimension M N . Forinstance,when N M 4 4 ,therelationbetweenthe channelatthe
receiverandtheindependentchannelsatthereceiveris

r Csn

where s [ s1. , s 2 , s 3 , s 4 ] referstothe4vectorofthe independent channelsassociatedwiththe4

transmit antenna , r [ r1 , r2 , r3 , r4 ] the 4vector of the received channel on each of the receive

{03/15/2004}

IEEEP802.20PD<number>/V<number>

antenna, n [ n1 , n 2 , n3 , n 4 ] the 4vector of Gaussian random variables modeling the noise/

interferenceoccurringoneachofthereceiveantenna,and

c11
c
C 21
c31

c 41

c12
c 22
c32
c 42

c13
c 23
c33
c 43

c14
c 24
c34

c 44

isthe correlation matrix between the incident independent channels and the resulting channels at the
receiver.

Inthespecialcasewherethereisonlyonetransmitantennaandonereceiveantenna, N M 1 1 ,
thematrix C reducestoascalarnumber.Then,therelationbetweentheresultingchannelatthereceiver
andtheindependentincidentchannelatthetransmitterbecomes

r cs n
wherethecorrespondence r r1 , s s1 , n n1 and c c11 canbeidentified.Thisisexactlythe
descriptionofthechannelrelationforsingleantennaconfigurationasdescribedbyITUmodel.

InspecialcasewhereN=M=1,thematrixreducestoaconstant.
3.5.2 Procedure to Generate Correlation Matrix Coefficients
TheprocedureforgeneratingMIMOcorrelationmatrixcoefficientsisshowninFigure3.5.21.

10

{03/15/2004}

IEEEP802.20PD<number>/V<number>

num. of antenna
antenna spacing
cluster
PAS
AO A, AS
BS and MS correlation
matrix

Phase 1
Spatial correaltion
matrix generation

R, Q , sigma

Spatial correaltion
matrix
R=kron(R BS,RMS) : down
R=kron(R MS,RBS) :up

Symmetrical mapping
matrix
C=chol(R)
x

Phase 2
Correlated fading
signal generation
Uncorrelated
fading signal
a

correlated signal matrix

A=sqrt(P)Ca

Figure3.5.21CorrelationChannelModelingProcedure

Theprocedureisdividedintotwomajorphases.Inthefirstphase,acorrelationmatrixisgeneratedfor
eachmobilestation(MS)andbasestation(BS)basedonthenumberofantennas,antennaspacing,number
ofclusters,powerazimuthspectrum(PAS),azimuthspread(AS),andangleofarrival(AoA).Thesetwo
correlationmatricesarecombinedtocreateaspatialcorrelationmatrixusingtheKroneckerproduct.In
thesecondphase,acorrelatedsignalmatrixiscreatedusingfadingsignalsderivedfromvariousDoppler
spectra and power delay profiles, and a symmetrical mapping matrix based on the spatial correlation
matrix.SomeoftheparametersthatcanbeusedinthecorrelationchannelmodelareshowninTable
3.5.21.

11

{03/15/2004}

IEEEP802.20PD<number>/V<number>

3.6 Link Level Spatial Channel Model Parameter Summary a

Thissectiondescribeslinklevelchannelmodelingparameters.Linklevelsimulationsalonewillnotbe
used for the comparison of MBWA technical proposals. Only system level simulations can achieve
accurateperformanceevaluationofdifferentMBWAAIproposals.

3.6.1 Link Level Channel Model Parameter Summary

Thefollowingtablesummarizesthephysicalparameterstobeusedforlinklevelmodeling.

12

{03/15/2004}

IEEEP802.20PD<number>/V<number>

casei

caseii

caseiii

caseIv

PDP

PedestrianA

VehicularA

PedestrianB
(PhaseI)

VehicularB
(PhaseI)

NumberofPaths

1) 4+1 (LOS
on, K = 6dB)
2) 4 (LOS off)

Delay(ns)

RelativePathpower(dB)

Models

2.5

1.0

310

0.9

200

300

110

9.0

710

4.9

800

12.8

8900

190

10.0

1090

8.0

1200

10.0

12900

410

15.0

1730

7.8

2300

25.2

17100

20.0

2510

23.9

3700

16.0

20000

0,

-9.7
19.2

-22.8

3, 30, 120
30, 120, 250

MobileStation

Speed(km/h)

Topology
PAS

0.5
1) LOS on: Fixed
AoA for LOS
component,
remaining power
has 360 degree
uniform PAS.
2) LOS off: PAS
with a Laplacian
distribution, RMS

0.5
RMS angle spread of
35 degrees per path
with a Laplacian
distribution

3,

[30,]120,250

[Ed.Note:
Subject

to
consistency with
EVdoc]

[Ed.Note: Subject
toconsistencywith
EVdoc]

0.5
RMS angle spread
of 35 degrees per
path with a
Laplacian
distribution

0.5
RMS angle spread
of 35 degrees per
path with a
Laplacian
distribution

Or 360 degree
uniformPAS

Or 360 degree
uniformPAS

13

{03/15/2004}

IEEEP802.20PD<number>/V<number>
anglespreadof35
degreesperpath

DoT
(degrees)

BaseStation

AoA
(degrees)

22.5

22.5

22.5

22.5 (LOS
component)
67.5 (all other
paths)

67.5(allpaths)

67.5(allpaths)

67.5(allpaths)

Reference: ULA with

Topology

0.5-spacing

or

4-spacing

or

10-spacing

Laplacian distribution with RMS angle spread of

PAS

2 degrees

or

5 degrees,

per path depending on AoA/AoD

AoD/AoA

50 for 2 RMS angle spread per path

20 for 5 RMS angle spread per path

(degrees)

Table3.6.11SummaryofLinkLevelChannelModelParameters

MIMO Channel Model for System Level Simulations

4.1 Introduction
ThespatialchannelmodelforMBWAsystemlevelsimulationsisdescribedinthischapter.Asinthelink
levelsimulations,thedescriptionisinthecontextofadownlinksystemwhereBStransmitstoaMS;
however the methodology described here can be applied to the uplink as well. [Note: Additional
informationmayberequiredtoapplythismethodologyfortheuplink.]Thegoalofthischapteristo
definethemethodologyandparametersforgeneratingthespatial andtemporalMIMOchannel model
coefficientsforMBWAsystemsimulations.
AsopposedtolinklevelsimulationswhereonlyconsideringthecaseofasingleBStransmittingtoa
single MS, the system level simulations typically consist of multiple cells/sectors, BSs, and MSs.
PerformancemetricssuchasdatathroughputsarecollectedoverDdrops,wherea"drop"isdefinedasa
simulationrunforagivennumberofcells/sectors,BSs,andMSs,overaspecifiednumberofframes.
Duringadrop,thechannelundergoesfadingaccordingtothespeedofMSs.Channelstateinformationis
fedbackfromtheMSstotheBSs,andtheBSsuseschedulerstodeterminewhichuser(s)totransmitto.
Typically,overaseriesofdrops,thecelllayoutisfixed,butthelocationsoftheMSsarestillrandom
variablesatthebeginningofeachdrop.

14

{03/15/2004}

IEEEP802.20PD<number>/V<number>

ForanSelementBSarrayandaUelementMSarray(SeeFigure3.1),thechannelcoefficientsforoneof
Nmultipathcomponentsaregivenbya U S complexmatrix.Wedenotethechannelmatrixforthenth
multipathcomponentas H n t ,wheren=1,,N.ThebroadbandMIMOradiochanneltransfermatrix

)
H t canbemodeledas

N
)
H (t ) Hn t (t n )
n 1

Where H t

U S

and

h11 n K

Hn t M O
n
hU 1 L

h1 Sn

M
n
hUS
U S

isacomplexmatrixwhichdescribesthelineartransformationbetweenthetwoconsideredantennaarrays
n

atdelay n ,where hu , s isthecomplextransmissioncoefficientfromantenna s attheBStoantenna u at

theMS.Noticethattheaboveequationisasimpletappeddelaylinemodelinamatrixformat,wherethe
channelcoefficientsatthe N delaysarerepresentedbymatrices.
ThesignalsattheMSantennaarrayaredenoted Y (t )

y1 (t ), y2 (t ),L

, yU (t ) ,where yu (t ) isthe
T

signalatthe u th antennaelement.Similarly,thesignalsattheBSantennaarrayarethecomponentsofthe
vertor X(t )

x1 (t ), x2 (t ),L

, xS (t ) .Therelationbetweenthevectors Y (t ) and X(t ) canbe

T

expressedas

)
Y(t ) H ( ) X(t )d t
(n)

whereitisassumedthat hu , s iszeromeancomplexGaussiandistributed,i.e.,

hu( n, s) isRayleigh

distributed, t isAWGNand

(t ) 1 (t ),2 (t ),...,U (t ) U 1
T

Theoverallprocedureforgeneratingthechannelmatricesconsistsofthreebasicsteps:
[EditorNote:Thisprocedurewillbeupdatedbaseduponcorrelationmatrixapproach]
1.

Specifyanenvironment,i.e.,suburbanmacro,urbanmacro,urbanmicro,orindoorpico.

2.

Obtaintheparameterstobeusedinsimulations,associatedwiththatenvironment.

3.

Generatethechannelcoefficientsbasedontheparameters.

15

{03/15/2004}

IEEEP802.20PD<number>/V<number>

Figure4.11MIMOModelwithSTransmitAntennasandUReceiveAntennas

Thefollowingsectionsdescribethedetailsofoverallprocedure.Thefigurebelowprovidesaflowchart
forgeneratingchannelcoefficients.

16

{03/15/2004}

IEEEP802.20PD<number>/V<number>

Figure4.12TheflowchartforthegenerationofMIMOchannelmodelcoefficients

Appendix

5.1 An Example of How the MIMO Channel Models Collapse to SISO Models
Bythecorrelationmatrixapproach,thechannelattheMreceiveantennaelementsisrelatedtotheN
independent channels that are associated with the N transmit antennae via a correlation matrix of
dimension M N . Forinstance,when N M 4 4 ,therelationbetweenthe channelatthe
receiverandtheindependentchannelsatthereceiveris

r Csn

where s [ s1. , s 2 , s 3 , s 4 ] referstothe4vectorofthe independent channelsassociatedwiththe4

transmit antenna , r [ r1 , r2 , r3 , r4 ] the 4vector of the received channel on each of the receive
antenna, n [ n1 , n 2 , n3 , n 4 ] the 4vector of Gaussian random variables modeling the noise/
interferenceoccurringoneachofthereceiveantenna,and

17

{03/15/2004}

c11
c
C 21
c31

c 41

c12
c 22
c32
c 42

IEEEP802.20PD<number>/V<number>

c13
c 23
c33
c 43

c14
c 24
c34

c 44

isthe correlation matrix between the incident independent channels and the resulting channels at the
receiver.
Inthespecialcasewherethereisonlyonetransmitantennaandonereceiveantenna, N M 1 1 ,
thematrix C reducestoascalarnumber.Then,therelationbetweentheresultingchannelatthereceiver
andtheindependentincidentchannelatthetransmitterbecomes

r cs n
wherethecorrespondence r r1 , s s1 , n n1 and c c11 canbeidentified.Thisisexactlythe
descriptionofthechannelrelationforsingleantennaconfigurationasdescribedbyITUmodel.

InspecialcasewhereN=M=1,thematrixreducestoaconstantwhile

5.2 Spatial Parameters for the Base Station

5.2.1 BS Antenna Topologies
AttheBS,alinearantennaarrayisassumed.Threevaluesforreferenceantennaelementspacingare
definedas:0.5 ,4 ,and10 .
The3sectorantennapatternisusedforBS,whichisplottedinError:Referencesourcenotfoundandis
specifiedby
2

A min 12
, Am
3 dB

where 180 180

isdefinedastheanglebetweenthedirectionofinterestandthebroadsideoftheantennaarray. 3dB is
the3dBbeamwidthindegrees,and Am isthemaximumattenuation.Fora3sectorscenario 3dB is70

18

{03/15/2004}

IEEEP802.20PD<number>/V<number>

degrees,and Am =20dB.Theterm broadside referstothedirectionfrom whichthesignal iscoming

perpendicularlytotheMEA.Anantennaarrayshowsthemaximumgainatitsbroadsidedirection.The
antennabroadsidepointingdirectionisillustratedbyFigure2.2fora3sectorscenario.Theantennagain,
asspecifiedby3GPP/3GPP2document[13]whichassumesthefrequencybandrangeis23GHz,is14
dBifora3sectorscenario.
3 Sector Antenna Pattern
0

Gain in dB.

-5

-10

-15

-20

-25
-120 -100 -80 -60 -40 -20

20

40

60

80 100 120

Azim uth in Degrees

Figure5.2.11Antennapatternfor3sectorcells

3-Sector Scenario
BS

Antenna Broadside in
direction of arrow

Figure5.2.12Illustrationofthebroadsidepointingdirectionofantennaarrayfor3sectorcells
Fora6sectorscenario, 3dB is35degree, Am =23dB,whichresultsintheantennapatternshowninthe
Figure2.3below,andthebroadsidepointingdirectionillustratedbyFigure2.4.Thegainspecifiedby
3GPPdocument[13]is17dBifora6sectorscenario.

19

{03/15/2004}

IEEEP802.20PD<number>/V<number>

6 Sector Antenna Pattern

0

Gain in dB

-5
-10
-15
-20
-25
-60 -50 -40 -30 -20 -10

10

20

30

40

50

60

Azim uth in Degrees

Figure5.2.13AntennaPatternfor6sectorcells

BS

Antenna Broadside in
direction of arrow
Figure5.2.14BroadsidePointingDirectionfor6sectorcells
5.2.2 BS Angles of Departure and Arrival
TheAngleofDeparture(AoD)isdefinedtobethemeananglewithwhichadepartingrayspoweris
transmittedbytheBSarraywithrespecttothebroadside.TheAngleofArrival(AoA)isdefinedtobethe
meananglewithwhichanarrivingrayspowerisreceivedbytheBSarraywithrespecttothebroadside.
AtthebasestationtheAoDandAoAhaveidenticalvalues.Thetwovalueschosentobeinsimulation
are:

AoD/AoA:50degrees(associatedwiththeRMSAngleSpreadof 2o )

20

{03/15/2004}

IEEEP802.20PD<number>/V<number>

AoD/AoA:20degrees(associatedwiththeRMSAngleSpreadof 5o )

5.2.3 BS Angle Spread

Thebasestationperpathanglespreadisdefinedastherootmeansquare(RMS)ofangleswithwhicha
rayspowerisreceivedortransmittedbythebasestationMEA.Theindividualpathpowersaredefinedin
the temporal ITU SISO channel models. Three values of BS angle spread (each associated with a
correspondingmeanAoD/AoA)areconsideredinthisdocument:

AS:2degreesatAoD/AoA= 50o

AS:5degreesatAoD/AoA= 20o

AS:25degreeatAoD/AoA= 20o forthecaseofindoorpicocell

Attentionshouldbepaidwhencomparingthelinklevelperformancebetweenthetwoanglespreadvalues
sincetheBSantennagainsforthetwocorrespondingAoDsandAoAsaredifferent.

5.2.4 BS Power Azimuth Spectrum

The Power Azimuth Spectrum (PAS) of a ray arriving at the base station MEA exhibits Laplacian
distribution.ForanAoD andRMSanglespread ,theBSperpathPASvalueatanangle isgiven
by:

P ( , , ) N o exp

G ( ) ,

wherebothangles and aregivenwithrespecttothebroadsideoftheMEA.Itisassumedthatall

antennaelementsorientationsarealigned.Also,PistheaveragereceivedpowerandGisthenumeric
basestationantennagaingivenby
G () 10 0.1A()
Finally,Noisthenormalizationconstant:

N o1

exp

G () d

Intheaboveequation, representspathcomponents(subrays)ofthepathpowerarrivingatanAoD .

21

{03/15/2004}

IEEEP802.20PD<number>/V<number>

5.3 Spatial Parameters for the Mobile Station

Calibrating MBWA simulators at link level requires the specification of a common set of spatial
parametersforthemobilestation.
5.3.1 MS Antenna Topologies
AttheMS,theMEAelementspacingis0.5 ,where isthewavelengthofthecarrierfrequency.
ForeachantennaelementattheMS,theantennapatternwillbeassumedomnidirectionalwithanantenna
gainof1dBi.
5.3.2 MS Angle Spread
TheMSperpathASisdefinedastherootmeansquare(RMS)ofanglesofanincidentpathspoweratthe
MSarray.Twovaluesofthepathsanglespreadareconsidered:

AS:104o(resultsfromthePASwithauniformdistributionover360degree),

AS:35oforaLaplacianPASwithacertainpathspecificAngleofArrival(AoA).

5.3.3 MS Angle of Arrival

TheperpathAngleofArrival(AoA)isdefinedasthemeanangleofanincidentrayattheMSMEAwith
respecttothebroadsideasshowninthefigurebelow,

AoA = 0

AoA < 0

AoA > 0

Figure 5.3.3-1 Angle of arrival orientation at the MS

TheAoAanalysiscanprovideanillustrationofthePASatMSMEA.ThreedifferentperpathAoAvalues
attheMSaresuggestedforthecasesofnonuniformPASanduniformPAS,

AoA:67.5degrees(associatedwithanRMSAngleSpreadof35 oand104o)

AoA:+67.5degrees(associatedwithanRMSAngleSpreadof35 oand104o)

AoA:+22.5degrees(associatedwithanRMSAngleSpreadof35 oorwithanLOScomponent)

22

{03/15/2004}

IEEEP802.20PD<number>/V<number>

5.3.4 MS Power Azimuth Spectrum

The Power Azimuth Spectrum (PAS) of a ray arriving at the MS is modeled as either Laplacian
distributionoruniformdistributionover360 o.SinceanomniantennaisassumedatMS,thereceivedper
pathPASwillremaineitherLaplacianoruniform.ForanincomingAoA andRMSanglespread ,the
MSperpathLaplacianPASvalueatanangle isgivenby:

P ( , , ) N o exp

wherebothangles and aregivenwithrespecttothebroadsideoftheMEA.Itisassumedthatall

antenna elements orientations are aligned. Also, P is the average received power and N o is the
normalizationconstant:

N o1

exp

Intheaboveequation, representspathcomponents(subrays)ofthepathpowerarrivingatanincoming
AoA .
5.3.5 MS Direction of Travel
Themobilestationdirectionoftravelisdefinedwithrespecttothebroadsideofthemobileantennaarray
asshowninthefigurebelow,
DoT = 0

DoT < 0

DoT > 0

Figure5.3.51DirectionofTravelforMS

23

{03/15/2004}

IEEEP802.20PD<number>/V<number>

5.3.6 Doppler Spectrum

TheperpathDopplerSpectrumisdefinedasafunctionofDoT,perpathPAS,andAoAatMS.Doppler
spectrum isaffectedbythePASandtheAngleofArrival.Dopplerspectrum affectsthetimedomain
behaviorofthechannel.

5.4 Definitions, Parameters, and Assumptions

ThereceivedsignalatMSconsistsofNtimedelayedmultipathreplicasofthetransmittedsignal.These
N pathsaredefinedbythechannelPDP,andarechosenrandomlyaccordingtothechannelgeneration
procedure.EachpathconsistsofMsubpaths.Figure3.3showstheangularparametersusedinthemodel.
Thefollowingdefinitionsareused:

BS

BS antenna array orientation, defined as the angle between the broadside of BS MEA and
theabsoluteNorth(N)referencedirection.

BS

TheanglebetweenLOSdirectionandthebroadsideofBSarray.

n , AoD AoDforthenthraywithrespecttotheLOS,where(n=1N).
n ,m, AoD Offsetforthemthsubpathofthenthraywithrespectto,where(m=1M).

n ,m , AoD AbsoluteAoDforthemthsubpathofthenthrayattheBSwithrespecttotheBSbroadside.
MS

MSantennaarrayorientation,definedastheanglebetweenthebroadsideoftheMSMEAandthe
absoluteNorthreferencedirection.

MS

AnglebetweentheBSMSLOSandtheMSbroadside

n , AoA AoAofthenthraywithrespecttoLOS
n ,m , AoA Offsetforthemthsubpathofthenthraywithrespectto n , AoA .

n ,m , AoA AbsoluteAoAforthemthsubpathofthenthrayattheMSw.r.t.theMSbroadside
V

MSvelocityvector

AngleofthevelocityvectorwithrespecttotheMSbroadside: v

arg V

Note:Theanglemeasuredinaclockwisedirectionisassumedtobenegativevalue.

24

{03/15/2004}

IEEEP802.20PD<number>/V<number>

Figure5.41TheMIMOchannelmodelangleparametersatbothBSandMS
Forthepurposeofsystemlevelsimulation,thefastfadingperpathwillbeevolvedintime,althoughbulk
parametersincludinganglespread,delayspread,lognormal shadowing,andMSlocationwillremain
fixedduringitsevaluationatonedrop.
Thefollowingassumptionsaremadeforallsimulations,independentofenvironment:
1.

UplinkDownlink Reciprocity: The AoD/AoA values are identical between the uplink and
downlink.

2.

ForFDDsystems,randomsubpathphasesbetweenUL,DLareuncorrelated.(ForTDDsystems,
thephaseswillbefullycorrelated.)

3.

Shadowingamongdifferentmobilesisuncorrelated.Inpractice,thisassumptionwouldnothold
ifmobilesareveryclosetoeachother,butwemakethisassumptionjusttosimplifythemodel.

4.

Thespatialchannelmodelshouldallowanytypeofantennaconfigurationtobeselected.Inorder
tocomparealgorithms,referenceantennaconfigurationsbasedonUniformLinearArray(ULA)
configurationswith0.5,4,and10wavelengthinterelementspacingwillbeused.

5.

ThecompositeAS,DS,andSFshadowfading,whichmaybecorrelatedparametersdepending
onthechannel scenario,areappliedtoallthesectorsorantennasofagivenbase. Subpath
phasesarerandombetweensectors.TheASiscomposedofNx20subpaths,andeachsubpath
hasapreciseAoD.TheSFisabulkparameterandiscommonamongalltheBSantennasor
sectors.

6.

Theelevationspreadisnotmodeledhere.

7.

Toallowcomparisonsofdifferentantennascenarios,thetransmitpowerofasingleantennacase
shallbethesameasthetotaltransmitpowerofamultipleantennacase.

25

{03/15/2004}
8.

IEEEP802.20PD<number>/V<number>

The generation of the channel coefficients assumes linear arrays. The procedure can be
generalizedforotherarrayconfigurations.

5.5 MIMO Channel Environments

Thefollowingchannelenvironmentswillbeconsideredforsystemlevelsimulations.

Suburbanmacrocell(PhaseIperevaluationdocument)

Urbanmacrocell

Urbanmicrocell

Indoorpicocell

Thetablebelowdescribestheparametersusedineachoftheenvironments.
ChannelScenario
SuburbanMacro
UrbanMacro
UrbanMicro
(PhaseI)
Numberofpaths(N)
Numberofsubpaths(M)per
path
MeanASatBS

6
20

6,11
20

6,11
20

E[ AS ]=50

E[ AS ]=80,150

ASatBSasalognormalRV

AS =0.69

80 AS =0.810

NLOS:
=190
N/A

AS =0.13

AS =0.34

AS 10 ^ AS x AS , x ~ (0,1)

Indoor Pico [ Note:

TBDs need to be
determined]
6,12
20

E[ AS ]

TBD
TBD

150 AS =1.18
AS =0.210

rAS AoD / AS

1.2

1.3

N/A

TBD

PerpathASatBS(Fixed)

20

20

50(LOSandNLOS)

25o,35o(LOS&NLOS)

BS perpath AoD Distribution

standarddistribution

(0, 2AoD ) wher

U(40o,40o)

U(180o,180o)

e AoD r AS AS

(0, 2AoD ) whe

re
AoD r AS AS

MeanASatMS

E[AS,MS]=680
350

E[AS,MS]=680
350

E[AS,MS]=680
350

E[AS,MS]=680
35o

(0, 2AoA (Pr))

(0, 2AoA (Pr))

(0, 2AoA (Pr))

(0, 2AoA (Pr))

DS=6.18
DS=0.18

N/A

TBD

E[ DS ] = 0.65
s
1.7

E[ DS ]=0.251s

0.035,0.1 s

N/A

TBD

U(0,1.2s)

U(0,031s)

PerpathASatMS(fixed)
MSPerpathAoADistribution

DelayspreadasalognormalRV DS=6.80

DS=0.288
DS 10 ^ DS x DS , x ~ (0,1)
MeantotalRMSDelaySpread

E[ DS ]=0.17s

rDS delays / DS

1.4

Distributionforpathdelays

26

{03/15/2004}

IEEEP802.20PD<number>/V<number>

Lognormal shadowing standard

deviation

10dB

10dB

Pathlossmodel(dB),
disinmeters

31.5+35log10(d)

34.5+35log10(d)

NLOS:10dB
LOS:4dB
NLOS:34.53+38log10
(d)
LOS:30.18

+
26*log10(d)

NLOS:12dB
LOS:4dB

Table5.51EnvironmentParameters
Thefollowingassumptionsaremadeforthesuburbanmacrocellandurbanmacrocellenvironments.
1.

The macrocell pathloss is based on the modified COST231 Hata urban propagation model:
d
) 45.5
1000
35.46 1.1hms log10 ( fc ) 13.82 log10 (hbs ) 0.7 hms C
PL[dB] 44.9 6.55log10 hbs log10 (

where hbs is the BS antenna height in meters, hms the MS antenna height in meters, f c is the
carrier frequency in MHz, d is the distance between the BS and MS in meters, and C is a
constant factor (C = 0dB for suburban macro and C = 3dB for urban macro). Setting these
parameters to hbs = 32m, hms = 1.5m, and f c =1900MHz, the path-losses for suburban and
PL 31.5 35log10 (d ) and
urban
macro
environments
become,
respectively,
PL 34.5 35log10 (d ) . The distance d is required to be at least 35m.
2.

AntennapatternsattheBSarethesameasthoseusedinthelinksimulations.

3.

SitetositeSFcorrelationis 0.5 .

4.

The hexagonal cell repeats will be the assumed layout.

Thefollowingassumptionsaremadeforthemicrocellenvironment.
1.

The microcell NLOS pathloss is based on the COST 231 Walfish-Ikegami NLOS model with the
following parameters: BS antenna height 12.5m, building height 12m, building to building distance
50m, street width 25m, MS antenna height 1.5m, orientation 30deg for all paths, and selection of
metropolitan center. With these parameters, the equation simplifies to:
PL(dB) = -55.9 + 38*log10(d) + (24.5 + 1.5*fc/925)*log10(fc).
The resulting pathloss at 1900 MHz is: PL(dB) = 34.53 + 38*log10(d), where d is in meters. The
distance d is at least 20m. A bulk log normal shadowing applying to all sub-paths has a standard
deviation of 10dB.
The microcell LOS pathloss is based on the COST 231 Walfish-Ikegami street canyon model with the
same parameters as in the NLOS case. The pathloss is
PL(dB) = -35.4 + 26*log10(d) + 20*log10(fc)
The resulting pathloss at 1900 MHz is PL(dB) = 30.18 + 26*log10(d), where d is in meters. The
distance d is at least 20m. A bulk log normal shadowing applying to all sub-paths has a standard
deviation of 4dB.

27

n 2 0.46

n 1

37 30 log10 R 18.3n

{03/15/2004}

IEEEP802.20PD<number>/V<number>

2.

Antenna patterns at the BS are the same as those used in the link simulations.

3.

Sitetositecorrelationis 0.5 .

4.

Thehexagonalcellrepeatswillbetheassumedlayout.

Thefollowingassumptionsaremadefortheindoorpicocellenvironment.
1.

TheindoorpathlossisbasedontheCOST231model:
n2

0.46

n 1

PL(dB ) 37 30 log10 R 18.3n

whereRisthedistancebetweenBSandMSinmeters,nisthenumberofpenetratedfloors(n=4
isanaverageforindoorofficeenvironment).
2.

Alognormalshadowfadingstandarddeviationof12dBcanbeexpectedfromindoorpicocell

3.

FadingrangesfromRiceantoRayleighwithDopplerfrequencyoffsetssetbywalkingspeed.

Sub-path #
(m)
1, 2
3, 4
5, 6
7, 8
9, 10
11, 12
13, 14
15, 16
17, 18
19, 20

Offset for a 2 deg AS

at BS (Macrocell)
n ,m ,AoD (degrees)

0.0894
0.2826
0.4984
0.7431
1.0257
1.3594
1.7688
2.2961
3.0389
4.3101

Offset for a 5 deg AS

at BS (Microcell)
n ,m ,AoD (degrees)

Offset for a 35 deg AS

at MS
n ,m ,AoA (degrees)

0.2236
0.7064
1.2461
1.8578
2.5642
3.3986
4.4220
5.7403
7.5974
10.7753

1.5649
4.9447
8.7224
13.0045
17.9492
23.7899
30.9538
40.1824
53.1816
75.4274

Table5.52SubpathAoDandAoAoffsets
Thevaluesinabovetableareselectedtoproduceabiasedstandarddeviationequalto2,5,and35degrees,
whichisequivalenttotheperpathpowerweightedazimuthspreadforequalpowersubpaths.

5.6 Generating SCM Parameters

Foragivenscenarioandasetofparameters,realizationsofeachchannelmodelparameterssuchasthe
pathdelays,powers,andsubpathAoDandAoAcanbederivedbyusingtheproceduredescribedinthis
section.Inparticular,Section3.4.1describesthestepsfortheurbanmacrocellandsuburbanmacrocell

28

{03/15/2004}

IEEEP802.20PD<number>/V<number>

environments,andSection3.4.2describesthestepsfortheurbanmicrocell environment.Fordetailed
information,pleaseseereference[13].
5.6.1 Generating Model Parameters for Urban and Suburban Macrocell Environments
Step1:Chooseeitheranurbanmacrocellorsuburbanmacrocellenvironment.
Step2:Determinevariousdistanceandorientationparameters.
Step3:DeterminetheDS,AS,andSF.
Step4:DeterminerandomdelaysforeachoftheNmultipathcomponents.
Step5:DeterminerandomaveragepowersforeachoftheNmultipathcomponents.
Step6:DetermineAoDsforeachoftheNmultipathcomponents.
Step7:AssociatethemultipathdelayswithAoDs.
Step8:Determinethepowers,phasesandoffsetAoDsoftheM=20subpathsforeachoftheNpathsat
theBS.
Step9:DeterminetheAoAsforeachofthemultipathcomponents.
Step10:DeterminetheoffsetAoAsattheUEoftheM=20subpathsforeachoftheNpathsattheMS.
Step11:AssociatetheBSandMSpathsandsubpaths.
Step12:DeterminetheantennagainsoftheBSandMSsubpathsasafunctionoftheirrespectivesub
pathAoDsandAoAs.
Step13: ApplythepathlossbasedontheBStoMSdistancefromStep2,andthelognormalshadow
fadingdeterminedinstep3asbulkparameterstoeachofthesubpathpowersofthechannelmodel.
5.6.2 Generating Model Parameters for Urban Microcell Environments
Urbanmicrocellenvironmentsdifferfromthemacrocellenvironmentsinthattheindividualmultipathsare
independentlyshadowed.]
Step1:Choosetheurbanmicrocellenvironment.
Step2:Determinevariousdistanceandorientationparameters.
Step3:Determinethebulkpathlossandlognormalshadowfadingparameters.
Step4:DeterminetherandomdelaysforeachoftheNmultipathcomponents.
Step5:DeterminerandomaveragepowersforeachoftheNmultipathcomponents.
Step6:DetermineAoDsforeachoftheNmultipathcomponents.
Step7:RandomlyassociatethemultipathdelayswithAoDs.

29

{03/15/2004}

IEEEP802.20PD<number>/V<number>

Step8:Determinethepowers,phases,andoffsetAoDsoftheM=20subpathsforeachoftheNpathsat
theBS.
Step9:DeterminetheAoAsforeachofthemultipathcomponents.
Step10:DeterminetheoffsetAoAsoftheM=20subpathsforeachoftheNpathsattheMS.
Step11:AssociatetheBSandMSpathsandsubpaths.Subpathsarerandomlypairedforeachpath,
andthesubpathphasesdefinedattheBSandMSaremaintained.
Step12:DeterminetheantennagainsoftheBSandMSsubpathsasafunctionoftheirrespectivesub
pathAoDsandAoAs.
Step 13: Apply the path loss based on the BS to MS distance and the log normal shadow fading
determinedinStep3asbulkparameterstoeachofthesubpathpowersofthechannelmodel.

5.7 Generating SCM Coefficients

GiventheuserparametersgeneratedinSection3.4,weusethemtogeneratethechannelcoefficients.For
an S elementlinearBSarrayanda U elementlinearMSarray,thechannelcoefficientsforoneof N
multipathcomponentsaregivenbya U S matrixofcomplexamplitudes.Wedenotethechannelmatrix
forthenthmultipathcomponent(n=1,,N)as H n (t ) .The(u,s)thcomponent(s=1,,S;u=1,,U)
of H n (t ) isgivenby

GBS n ,m , AoD exp j kd s sin n ,m , AoD n ,m

hu( n, s) (t )

Pn SF
M

GMS n ,m, AoA exp jkdu sin n ,m , AoA

m 1

exp jk v cos n ,m, AoA v t

where

Pn

isthepowerofthenthpath(Step5).

SF

isthelognormalshadowfading(Step3),appliedasabulkparametertothenpathsfora
givendrop.

isthenumberofsubpathsperpath.

n , m , AoD

istheAoDforthemthsubpathofthenthpath(Step12).

n, m, AoA

istheAoAforthemthsubpathofthenthpath(Step12).

G BS ( n ,m, AoD )

istheBSantennagainofeacharrayelement(Step12).

G MS ( n, m, AoA )

istheMSantennagainofeacharrayelement(Step12).

30

{03/15/2004}

IEEEP802.20PD<number>/V<number>

isthesquarerootof1.

isthewavenumber 2 / where isthecarrierwavelengthinmeters.

ds

isthedistanceinmetersfromBSantennaelementsfromthereference(s=1)antenna.
Forthereferenceantennas=1, d1 =0.

du

isthedistanceinmetersfromMSantennaelementufromthereference(u=1)antenna.
Forthereferenceantennau=1, d1 =0.

n,m
v

isthephaseofthemthsubpathofthenthpath(Step8)
isthemagnitudeoftheMSvelocityvector(Step2)
istheangleoftheMSvelocityvector(Step2)

The path loss and the log normal shadowing is applied as bulk parameters to each of the subpath
componentsofthenpathcomponentsofthechannel.

References

[1]C802.2003/50,OverviewofMETRAModelforMBWAMIMOChannel,IEEE802.20Session#2
[2]C802.2003/49,ComparisonofSFBCandSTBCforTransmitDiversityinOFDMSystem,IEEE
802.20Session#2
[3] C802.2003/48,Channel Models and Performance ImplicationsforOFDMbasedMBWA, IEEE
802.20Session#2
[4]C802.2003/46r1,ChannelRequirementsForMBWA(Rev1),IEEE802.20Session#2.
[5]C802.2003/43,802.20EvaluationMethodologyStrawman,IEEE802.20Session#2
[6]C802.2003/42,ChannelModelingforMBWA,IEEE802.20Session#2.
[7]C802.2003/35,EvaluationMethodologyforMBWA,IEEE802.20Session#2.
[8]C802.2003/18,MIMOChannelModelforMBWA,IEEE802.20Session#1.
[9]C802.2003/15r1,ChannelModelsandPerformanceImplicationsforOFDMbasedMBWA,IEEE
802.20Session#1.
[10] C802.2003/12, Antenna Arrays for MBWA: Overview and Field Experiments, IEEE 802.20
Session#1.
[11]C802.2003/09,ChannelModelingSuitableforMBWA,IEEE802.20Session#0.
[12]3GPPTS45.005V5.9.0,3GPP:TechnicalSpecificationGroupGSM/EDGERadioAccessNetwork;
Radiotransmissionandreception(Release1999).

31

{03/15/2004}

IEEEP802.20PD<number>/V<number>

[13]3GPP&3GPP2SpatialChannelModelAHG,SpatialChannelModelTextDescription,SCMText
V7.0.
[14] ETSITR101112,UMTS30.03,V3.2.0,AnnexB,Sections1.2.3,1.3,1.4.
[15]IEEE802.16.3c01/29r4,ChannelModelsforFixedWirelessApplications,20010717.
[16]IEEE802.16.3c00/49r2,Interim Channel ModelsforG2MMDS FixedWirelessApplications,
20001115.
[17]C802.2003/77,SummaryofDelayProfilesforMBWA,IEEE802.20Session#4.
[18]I.E.Telatar,CapacityofMultiAntennaGaussianChannels,TechnicalReport,AT&TBellLabs.,
1995.
[19]K.I.Pedersen,J.B.Andersen,J.P.Kermoal,andP.E.Mogensen,AStochasticMultipleInput
MultipleOutput RadioChannel Model forEvaluationofSpaceTimeCodingAlgorithms,VTC 2000
Fall,Boston,MA,Sept.2000,pp.893897.
[20] J.P. Kermoal, L. Schumacher, K. I. Pedersen, P.E. Mogensen, and F. Frederiksen,A Stochastic
MIMORadioChannel Model withExperimentalValidation,IEEEJSAC,V.20,No.6,pp.12111226,
August2002.
[21]3GPP2TSGCR1002,1xEV_DVEvaluationMethodology(V13),2003.
[22] K.I. Pedersen, P.E. Mogensen, and B.H. Fleury, Spatial Channel Characteristics in Outdoor
EnvironmentsandtheirImpactonBSAntennaSystemPerformance,VTC98,Ottawa,Canada,pp.719
723,May1998.
[23]L.Greenstein,V.Erceg,Y.S.Yeh,M.V.Clark,ANewPathGain/DelaySpreadPropagationModel
forDigitalCellularChannels,IEEETransactionsonVehicularTechnology,Vol.46,No.2,May1997,
pp.477485.
[24]IEEEC802.200479,OverviewoftheSpatialChannelModeldevelopedin3GPP3GPP2.
[25] ITUR RECOMMENDATION M.1225, "GUIDELINES FOR EVALUATION OF RADIO
TRANSMISSIONTECHNOLOGIESFORIMT2000"

32