Documente Academic
Documente Profesional
Documente Cultură
20-04/xx
IEEE P 802.20/PD<insert PD Number>/V<07>
Date: <March 15, 2005>
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>
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)
3,
[30],120,250
[Ed.Note:
Subject
to
consistency with
EVdoc]
[Ed.Note: Subjectto
consistencywithEV
doc]
{03/15/2004}
IEEEP802.20PD<number>/V<number>
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.
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.
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.
RayTracingModel:Inthisapproach,exactlocationsoftheprimaryscatterers,theirphysical
characteristics,aswellastheexactlocationofthetransmitterandreceiverareassumedknown.
Theresultingchannelcharacteristicsarethenpredictedbysummingthecontributionsfroma
largenumberofthepropagationpathsfromeachtransmitantennatoeachreceiveantenna.This
techniqueprovidesfairlyaccuratechannelpredictionbyusingsitespecificinformation,suchas
databaseofterrainandbuildings.Formodelingoutdoorenvironmentsthisapproachrequires
detailedterrainandbuildingdatabases.
ScatteringModel:Thismodelassumesaparticularstatisticaldistributionofscatterers.Using
thisdistribution,channelmodelsaregeneratedthroughsimulatedinteractionofscatterersand
planarwavefronts.Thismodelrequiresalargenumberofparameters.
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)
r Csn
{03/15/2004}
IEEEP802.20PD<number>/V<number>
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
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>
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)
Topology
0.5-spacing
or
4-spacing
or
10-spacing
PAS
2 degrees
or
5 degrees,
(degrees)
Table3.6.11SummaryofLinkLevelChannelModelParameters
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
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
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
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
A min 12
, Am
3 dB
isdefinedastheanglebetweenthedirectionofinterestandthebroadsideoftheantennaarray. 3dB is
the3dBbeamwidthindegrees,and Am isthemaximumattenuation.Fora3sectorscenario 3dB is70
18
{03/15/2004}
IEEEP802.20PD<number>/V<number>
Gain in dB.
-5
-10
-15
-20
-25
-120 -100 -80 -60 -40 -20
20
40
60
80 100 120
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>
Gain in dB
-5
-10
-15
-20
-25
-60 -50 -40 -30 -20 -10
10
20
30
40
50
60
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 )
AS:2degreesatAoD/AoA= 50o
AS:5degreesatAoD/AoA= 20o
Attentionshouldbepaidwhencomparingthelinklevelperformancebetweenthetwoanglespreadvalues
sincetheBSantennagainsforthetwocorrespondingAoDsandAoAsaredifferent.
P ( , , ) N o exp
G ( ) ,
N o1
exp
G () d
Intheaboveequation, representspathcomponents(subrays)ofthepathpowerarrivingatanAoD .
21
{03/15/2004}
IEEEP802.20PD<number>/V<number>
AS:104o(resultsfromthePASwithauniformdistributionover360degree),
AS:35oforaLaplacianPASwithacertainpathspecificAngleofArrival(AoA).
AoA = 0
AoA < 0
AoA > 0
AoA:67.5degrees(associatedwithanRMSAngleSpreadof35 oand104o)
AoA:+67.5degrees(associatedwithanRMSAngleSpreadof35 oand104o)
AoA:+22.5degrees(associatedwithanRMSAngleSpreadof35 oorwithanLOScomponent)
22
{03/15/2004}
IEEEP802.20PD<number>/V<number>
P ( , , ) N o exp
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>
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.
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)
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)
U(40o,40o)
U(180o,180o)
e 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
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>
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.
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
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
0.0894
0.2826
0.4984
0.7431
1.0257
1.3594
1.7688
2.2961
3.0389
4.3101
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.
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.
hu( n, s) (t )
Pn SF
M
m 1
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.
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