文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 人文社科 > 文学研究 > A channel state information feedback method for massive MIMO-OFDM
侵权投诉

A channel state information feedback method for massive MIMO-OFDM

352JOURNALOFCOMMUNICATIONSANDNETWORKS,VOL.15,NO.4,AUGUST2013

AChannelStateInformationFeedbackMethodfor

MassiveMIMO-OFDM

RiichiKudo,SimonM.D.Armour,JoeP.McGeehan,andMasatoMizoguchi

Abstract:Combiningmultiple-inputmultiple-outputorthogonalfrequencydivisionmultiplexing(MIMO-OFDM)withamassivenumberoftransmitantennas(massiveMIMO-OFDM)isanat-tractivewayofincreasingthespectrumef ciencyorreducingthetransmissionenergyperbit.TheeffectivenessofMassiveMIMO-OFDMisstronglyaffectedbythechannelstateinformation(CSI)estimationmethodused.Theoverheadsoftrainingframetrans-missionandCSIfeedbackdecreasemultipleaccesschannel(MAC)ef ciencyandincreasetheCSIestimationcostatauserstation(STA).ThispaperproposesaCSIestimationschemethatreducesthetrainingframelengthbyusinganovelpilotdesignandanovelunitarymatrixfeedbackmethod.Theproposedpilotdesignandunitarymatrixfeedbackenabletheaccesspoint(AP)toestimatetheCSIofthesignalspaceofalltransmitantennasusingasmallnumberoftrainingframes.SimulationsinanIEEE802.11nchan-nelverifytheattractivetransmissionperformanceoftheproposedmethods.

IndexTerms:Channelstateinformation(CSI)estimation,massivemultiple-inputmultiple-output(MIMO),MIMO-orthogonalfre-quencydivisionmultiplexing(MIMO-OFDM),multiuserMIMO,wirelesslocaareanetwork(LAN).

I.INTRODUCTION

Multiple-inputmultiple-output(MIMO)techniquesarewidelyemployedinwirelesscommunicationsystemssincespa-tialmultiplexingcanincreasespectralef ciency[1].Thenum-berofantennasincellularandwirelesslocalareanetwork(LAN)systemsisincreasinganditisexpectedthatthisin-creasewillcontinueinordertosatisfythedemandsforhigherthroughputs.Insystemswheretheaccesspoint(AP)hasamas-sivenumberofantennas(massiveMIMO)[2],[3],precod-ingtechniquescanincreasethethroughputand/orreducethetransmissionpowerbyusingchannelstateinformation(CSI).CombiningMIMOwithorthogonalfrequencydivisionmulti-plexing(MIMO-OFDM)isalsopromisingsinceOFDMyieldsfrequency- atMIMOchannels.

MultiuserMIMOtransmission(MU-MIMO)[3]–[5]isakeytechnologyforMassiveMIMOsystemsinallowtheAPbetteraccesstothelargesignalspacesincethenumberofantennasattheuserstation(STA)isrestrictedbythepowerconsumption,andthesmallantennaspaceavailable.MU-MIMOprecodingincreasesthespectrumef ciencyofcommunicationwithmulti-ManuscriptreceivedJanuary31,2013.

R.KudoandM.MizoguchiarewithNTTNetworkInnovationLaboratories,Yokosuka,Japan,239-0847,email:{kudo.riichi,mi-zoguchi.masato}@lab.ntt.co.jp.

S.M.D.ArmourandJ.P.McGeehanarewiththeCenterforCom-municationsResearch,UniversityofBristol,UK,email:{j.p.mcgeehan,Si-mon.Armour}@bristol.ac.uk.

DigitalObjectIdenti er10.1109/JCN.2013.000066

pleuserstations(STAs)thathavefewerantennas.However,theCSIattheAPmustbeaccurateinordertoreducetheinter-userinterference.Ef cientCSIestimationisthusrequiredinordertoimplementMU-MIMO-OFDMformassiveMIMOsystems.Closed-loopandopen-loopCSIestimationmethodshavebeenproposed.Theopen-loopapproachusesthechannelreci-procityandestimatesthedownlinkCSIusingtheuplinkCSIandcalibrationtechniques.Thisapproachcanbeusedonlyfortimedivisionduplex(TDD)systems.Ifthecalibrationisperfectlyconducted,theopen-loopapproachisveryattractiveformas-siveMIMOsystems.However,MU-MIMOtransmissionper-formancedegradesifthecalibrationerrorincreasesCSIerroratthetransmitter.Furthermore,itisdif culttoexploitthechannelreciprocityinfrequencydivisionduplex(FDD)systems.Thispaperfocusesontheclosed-loopapproachtodispensewithanyrequirementforchannelreciprocity.IntheclosedloopapproachformassiveMIMOsystems,theoverheadofthetrainingframesforCSIestimationandtheCSIfeedbackdegradetheMACef -ciency.

ToreducetheCSIfeedbackbits,timedomainfeedbackwasproposedin[6],[7].However,theaccuracyofCSIinthetimedomainisaffectedbythenulldatasubcarriersdemandedbyOFDMsymboldesignandacostlytransformationintothetimedomainisrequiredattheSTA.Thealternativeapproaches,CSIcompressionmethodsofchannelmatricesinfrequency atchannelswereproposedin[8],[9].ItisknownthatCSIiseffec-tivelycompressedbyutilizingitscharacteristicofunitaryma-tricesincomplexEuclideanspace[10].CSIcompressionbasedonaunitarymatrixinthefrequencydomainisusedinIEEE802.11n[11].NotethatIEEE802.11acsupportsonlyunitarymatrixCSIcompressionforCSIfeedback[12].CSIfeedbackmethodsformassiveMIMOsystemshavebeenproposedin[13].CSIestimationin[13]needstotransmitasmanyOFDMpilotsymbolsastherearetransmitantennastoobtainfullCSI.ThispaperproposesaCSIestimationmethodthatusesasmallnumberoflongtrainingframes(LTFs).ReducingthenumberofpilotsymbolsreducestheCSIfeedbackoverheadandthecomputationalloadsforCSIestimationandtransmis-sionweightcalculation.TheproposedCSIestimationschemeisbasedontheiterativeCSIestimationof[14].TheiterativeCSIestimationschemeobtainstheCSIforalltransmitantennasbyusingadditionalCSIestimationbeams.SincetheAPcalcu-latestheCSIestimationweightusingthepriorestimatedCSIanddifferentcodebookvectorsareusedastheCSIestimationweightsoftheadditionalbeams,theperformanceofMU-MIMOtransmissionusingiterativeCSIestimationapproachesthatofafullantennaestimation,whichtransmitsLTFsasmanyasthenumberoftransmitantennas.However,iterativeCSIestimationneedssuf cienttimedurationtoconvergeonthebesttransmis-

c2013KICS1229-2370/13/$10.00

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email