France Telecom Propagation ChannelFrance Telecom Propagation Channel
ModelsModels
❖❖DigitalDigitalSignalSignalProcessingProcessingforforthetheGenericGenericModel:Model:TheTheUWB/WBUWB/WBpropagationpropagationchannelchannel
modelmodeldevelopeddevelopedbybyFranceFranceTelecomTelecomdiffersdiffersfromfromthetheclassicalclassicaltappedtappeddelaydelaylineline
modelmodelapproachapproachreferringreferringtotoaanon-uniformnon-uniformsamplingsamplingofofthethescatteringscatteringfunctionfunctionofof
thethechannelchannelPs(Ps(ν,τν,τ))withwithaalimitedlimitednumbernumberofoftapstapsandandananidenticalidenticalU-shapeU-shapeDopplerDoppler
for each tap.for each tap.
❖❖TheTheproposedproposedmodelmodelresortsresortsfromfromaatwo-dimensionaltwo-dimensionalmultiratemultiratefilterfilterappliedappliedonon
selectedselectedmeasuredmeasuredChannelChannelImpulseImpulseResponsesResponses(CIRs)(CIRs)ofofthethechannelchannelissuedissuedfromfromaa
statistical analysis of the selectivity parameters of the channelstatistical analysis of the selectivity parameters of the channel
ContCont……
❖❖TheThemodelmodelgeneratesgeneratesaatime-varianttime-variantfilterfilterfitfitwithwiththethesimulatedsimulatedsystemsystemusingusing
thethesystemsystembandwidthbandwidthtotofilterfilterthethechannelchannelandandcarriescarriesoutoutaaconversionconversionraterate
using an optimized interpolator-decimator filter.using an optimized interpolator-decimator filter.
❖❖ThisThismodelingmodelingapproachapproachhashasbeenbeenprimarilyprimarilyintroducedintroducedforforWLANsWLANsatat55GHzGHz
and extended to many other applications.and extended to many other applications.
❖❖ Actually, we use this concept to model the UWB channel. Actually, we use this concept to model the UWB channel.
❖❖TheThemainmainadvantageadvantageresultsresultsininaarealisticrealisticrepresentationrepresentationofofthethechannelchannelinin
connection with measurements in the face of stochastic models.connection with measurements in the face of stochastic models.
ContCont……
❖❖AsAsperperWSSUSWSSUSassumptions,assumptions,variablesvariablesttandand(τ(τ))ofofthetheimpulseimpulseresponseresponseh(t,h(t,ττ))ofof
thethechannelchannelshallshallbebeseparated,separated,hencehencethethefilteringfilteringprocessingprocessingisisprocessedprocessedininaa22××
1D assessment, reducing the complexity.1D assessment, reducing the complexity.
❖❖First,First,thetheinstantaneousinstantaneoustransversaltransversalfilterfilterhF(t,hF(t,ττ))describingdescribingthethechannelchannelatatthethetimetime
ttisisgeneratedgeneratedtotoananinterpolator-decimatorinterpolator-decimatorfilterfilterwithwithaaconversionconversionraterateRR==Tsig/TeTsig/Te
= po/qo expressed as a ratio of two integers.= po/qo expressed as a ratio of two integers.
❖❖TsigTsigisisthethesamplingsamplingperiodperiodofofthethecommunicationcommunicationsystemsystemandandTeTeisisthethesamplingsampling
rate of propagation measurements.rate of propagation measurements.
❖❖InInaasecondsecondstep,step,thetherenewalrenewalraterateofofeveryeverytransversaltransversalfilterfilterhF(t,hF(t,ττ))isisprocessedprocessedinin
performing a conversion rate of each transversal filter tap.performing a conversion rate of each transversal filter tap.
❖❖TheTherenewalrenewalraterateisischosenchosenasasaamultiplemultipleofofthethelargestlargestDopplerDopplervaluevalueυmaxυmax==VitVitλλ
(λ(λisisthethewavelengthwavelengthassociatedassociatedtotothetheRFRFfrequency)frequency)andandfittedfittedtotothethedatadataframingframing
rate and time variations of the channelrate and time variations of the channel
CEPDCEPD
❖❖TheThelinklinklevel-focusedlevel-focusedpropagationpropagationchannelchannelmodel,model,alsoalsodenoteddenotedCEPDCEPD
model,model,isisa a multipathmultipathpropagationpropagationmodelmodelwhichwhichconjunctlyconjunctlyexploitsexploitsmulti-multi-
rate digital filter processingrate digital filter processing
The CEPD propagation channelThe CEPD propagation channel
modelmodel
The CEPD propagation channelThe CEPD propagation channel
modelmodel
The mathematical expression of the model is given byThe mathematical expression of the model is given by
AndAndthethecoefficientscoefficientsc(k,i)c(k,i)resultresultfromfromaaananexpansionexpansionraterateFF==q0q0××Fe,Fe,followedfollowedbybyaafilteringfilteringprocessingprocessingandandaa
decimation rate with a ratio p0. c(k,i) is expressed as follows:decimation rate with a ratio p0. c(k,i) is expressed as follows:
ContCont……
❖❖b(k,b(k,n)n)arearethethecoefficientscoefficientsofofthethebasebandbasebandequivalentequivalentmeasuredmeasuredimpulseimpulse
responseresponseofofthethechannelchannelsampledsampledtotoFeFeconvertedconvertedtotoTTacquisitionacquisitionsampling,sampling,
ggFFisisthethefilterfilterimpulseimpulseresponseresponseusedusedininthethemultiratemultirateprocessingprocessingandandTTisisaa
timetimeacquisitionacquisitionperiod.period.FigureFigure18.1318.13showsshowstypicaltypicalaverageaveragepowerpowerdelaydelay
spread profiles of a WPAN working in 60 GHz.spread profiles of a WPAN working in 60 GHz.
Typical average power delay spread profiles of aTypical average power delay spread profiles of a
WPAN working in 60 GHzWPAN working in 60 GHz..
Cont..Cont..
❖❖TheTheantennaantennaimpactimpactandandtransmissiontransmissionbandwidthbandwidthBwBwareareinvolvedinvolvedintointothethe
base-bandbase-bandfilterfilterrealizationsrealizationsthroughthroughmeasurementsmeasurementsandandinputinputfilesfilesofofthethe
model.model.
❖❖..TheThesub-carriersub-carrierspacingspacingandandSpectrumSpectrumEfficiencyEfficiency(SE)(SE)dependencydependencyisis
shown in Figureshown in Figure
ContCont……
TheTheinputinputfilesfilesofofthethemodelmodelareareissuedissuedfromfromselectedselectedmeasuredmeasuredCIRCIRfilesfiles
whosewhoseselectivityselectivityparameterparametervaluesvaluesareareincludedincludedininaaspecificspecificinterval.interval.
AssumingAssumingthetheselectivityselectivityparameterparameterfirst-orderfirst-orderstatisticsstatisticsfollowfollowaaGaussianGaussian
distribution,distribution,thethetypicaltypicalandandatypicalatypicalinputinputfilesfilesofofthethemodelmodelareareCIRCIRfilesfiles
verifying the equationverifying the equation::
ContCont……
❖❖mmii
jj
isisthetheaverageaverageofofthetheselectivityselectivityparameterparameterjjforforthetheii
thth
measurementmeasurement
position,position,jj=={delay{delayspread,spread,coherencecoherencebandwidth,bandwidth,delaydelayintervalintervaltoto66andand1515
dB,dB,PowerPowerstandardstandarddeviation}.deviation}.XX
jjisiseithereitherthethefirst-orderfirst-orderstatisticstatisticofofthetheMiMi
jj
variablevariableforforthethetypicaltypicalinputinputfilesfilesororthetheCumulativeCumulativedensitydensityfunctionfunctionsetsettoto0.80.8
for the atypical input files deduced from i = 1, N measurement points.for the atypical input files deduced from i = 1, N measurement points.
❖❖InputInputFilesFilesofofthetheCEPDCEPDModelModelininaaResidentialResidentialEnvironment:Environment:InputInputfilesfilesofof
thetheCEPDCEPDmodelmodelareareaameasurementmeasurementpointpointcomprisingcomprisingNriNriinstantaneousinstantaneous
impulseimpulseresponsesresponsesofofthethechannelchannelalongalongaalinearlinearpathpathofof30λ30λapproximately.approximately.
TypicalTypicalandandAtypicalAtypicalcasescaseshavehavebeenbeenextractedextractedfromfromaagivengivenscenario.scenario.InputInput
files are denoted CEPD_ xxfiles are denoted CEPD_ xx
ContCont……
FiguresFigures11andand22depictdepictthethecycliccyclicprefixprefixdurationdurationandandsubcarriersubcarrierspacingspacing
combined with SE, as well as typical phase noise values.combined with SE, as well as typical phase noise values.
Fig-1Fig-1 Fig-2Fig-2