∑∑∑∑
Thisarticlehasbeenacceptedforpublicationinafutureissueofthisjournal,buthasnotbeenfullyedited.Contentmaychangepriortofinalpublication.Citationinformation:DOI10.1109/TVT.2020.3006318,IEEE
TransactionsonVehicularTechnology
6
TableII:Computationalcomplexitycomparisonofdifferentalgorithms
16
Complex-valuedmultiplications(10)
AlgorithmsComputationalcomplexity
N=50N=100
rr
K+1
a2
ProposedStrOMPOf(K+1)JKNN+[JN(s+2s+9.617.6
atrr
s=1
233
AUD
2(sN))+J(s+(sN))]g
tt
2
AUDpartofTLSSCS[7]Of(K+1)[N(KN+J)+NJKN]+12.544.2
artrt
K+1
a223
[N+2N(sN)+(sN)]g
rrtt
s=1
K
a2
AUDlowerboundOfKJKNN+[JN(s+2s+7.113.2
atrr
s=1
233
2(sN))+J(s+(sN))]g
tt
K
a23
ProposedSIC-SSPOfJ[2sN(N+1)+14Ns+11s]g2.14.0
rtr
s=1
23
DatadetectionpartofO[JNKN+2N(KN)+(KN)]0.150.28
ratratat
Datadetection
TLSSCS[7]
23
GSP[8]OfJ[2sN(N+1)+14NK+11K]g0.651.2
rtraa
23
BERlowerboundO(JNK+2NK+K)0.010.02
raraa
23
Benchmark1O(JNK+2NK+K)0.010.02
raraa
1
Thenumberofthecomplex-valuedmultiplicationsiscalculatedundertheparametersJ=12,N=4,K=100,K=8.
ta
BERperformanceversusthenumberofreceiveantennasN,REFERENCES
r
respectively.ObservefromFig.4thatwhenNbecomeslarge,
r[1]C.Bockelmannetal.,“Massivemachine-typecommunicationsin5G:
PhysicalandMAC-layersolutions,”IEEECommun.Mag.,vol.54,no.
theAUDperformanceorBERoftheproposed“StrOMP+SIC-
9,pp.59-65,Sept.2016.
SSP”solutionisbetterthanthatoftheTLSSCSdetectorand
[2]B.K.Jeong,B.ShimandK.B.Lee,“MAP-basedactiveuseranddata
the“StrOMP+GSP”solution.Thisindicatesthesuperiorityof
detectionformassivemachine-typecommunications,”IEEETrans.Veh.
Technol.,vol.67,no.9,pp.8481-8494,Sept.2018.
theproposedsolutionforemploymentinmMIMO.
[3]B.Wang,L.Dai,T.MirandZ.Wang,“Jointuseractivityanddata
Thecomputationalcomplexityofdifferentsolutionsinour
detectionbasedonstructuredcompressivesensingforNOMA,”IEEE
simulationsarecomparedinTableII,wherethedifferent
Commun.Lett.,vol.20,no.7,pp.1473-1476,Jul.2016.
algorithmsaredividedintotwopartsbasedontheirfunctions[4]Y.Du,C.Cheng,B.Dong,Z.chen,X.Wang,J.FangandS.Li,“Block-
sparsity-basedmultiuserdetectionforuplinkgrant-freeNOMA,”IEEE
(i.e.,AUDordatadetection).Itisobviousthatthenumber
Trans.WirelessCommun.,vol.17,no.12,pp.7894-7909,Dec.2018.
ofcomplex-valuedmultiplicationsoftheproposedStrOMP
[5]B.Wang,L.Dai,Y.Zhang,T.MirandJ.Li,“Dynamiccompressive
algorithmisalittlelowerthanthatintheAUDpartofthesensing-basedmulti-userdetectionforuplinkgrant-freeNOMA,”IEEE
Commun.Lett.,vol.20,no.11,pp.2320-2323,Nov.2016.
TLSSCSalgorithm(i.e.,lines1-14ofAlgorithm1in[7])
[6]Y.Du,B.Dong,Z.Chen,X.Wang,Z.Liu,P.GaoandS.Li,“Ef?cient
whenN=50.IfNisdoubled,thenumberofcomplex-
rr
multi-userdetectionforuplinkgrant-freeNOMA:Prior-information
aidedadaptivecompressivesensingperspective,”IEEEJ.Select.Areas
valuedmultiplicationsoftheproposedStrOMPalgorithm
Commun.,vol.35,no.12,pp.2812-2828,Jul.2017.
increaseslinearlywithN,whereasthecomplexityofthe
r
[7]X.Ma,J.Kim,D.YuanandH.Liu,“Two-levelsparsestructurebased
AUDpartoftheTLSSCSalgorithmisnearlyproportionalto
compressivesensingdetectorforuplinkspatialmodulationwithmassive
connectivity,”IEEECommun.Lett.,vol.23,no.9,pp.1594-1597,Sept.
thesquareofN.Hence,itisclearthatourStrOMPalgorithm
r
2019.
ismoresuitableformMIMOinconjunctionwithlargeantenna
[8]Z.Gao,L.Dai,Z.Wang,S.ChenandL.Hanzo,“Compressive-sensing
arrays.Furthermore,afterobtainingtheactiveMTDs,the
basedmultiuserdetectorforthelarge-scaleSM-MIMOuplink,”IEEE
datadetectionpartoftheTLSSCSalgorithmbecomesanLSTrans.Veh.Technol.,vol.65,no.10,pp.1860-1865,Feb.2017.
[9]L.Xiao,P.Yang,Y.Xiao,S.Fan,M.DiRenzo,W.Xiangand
operation(i.e.,line15ofAlgorithm1in[7])associatedwith
S.Li,“Ef?cientcompressivesensingdetectorsforgeneralizedspatial
alimitedBERperformanceforthemedia-modulatedsignal.
modulationsystems,”IEEETrans.Veh.Technol.,vol.66,no.2,pp.
Hence,ourproposedSIC-SSPalgorithmattainsabetterdata1284-1298,Feb.2017.
[10]L.Xiao,Y.Xiao,P.Yang,J.Liu,S.LiandW.Xiang,“Space-timeblock
detectionperformanceatthecostofahighercomputational
codeddifferentialspatialmodulation,”IEEETrans.Veh.Technol.,vol.
complexity.
66,no.10,pp.8821-8834,Oct.2017.
[11]L.Zhang,M.ZhaoandL.Li,“Low-complexitymulti-userdetectionfor
MBMinuplinklarge-scaleMIMOsystems,”IEEECommun.Lett.,vol.
22,no.8,pp.1568-1571,Aug.2018.
V.CONCLUSIONS
[12]B.Shamasundar,S.Jacob,L.N.TheagarajanandA.Chockalingam,
“Media-basedmodulationfortheuplinkinmassiveMIMOsystems,”
AmediamodulationbasedmMTCULschemerelyingon
IEEETrans.Veh.Technol.,vol.67,no.9,pp.8169-8183,Sept.2018.
mMIMOdetectionattheBSwasproposedforachieving
[13]J.W.Choi,B.Shim,Y.Ding,B.Rao,D.I.Kim,“Compressedsensing
reliablemassiveaccesswithanenhancedthroughput.Theforwirelesscommunications:Usefultipsandtricks”,Commun.Surveys
Tuts.,vol.19,no.3,pp.1527-1550,2017.
sparsenatureofthemMTCtraf?cmotivatedustopropose
[14]A.K.Khandani,“Media-basedmodulation:Anewapproachtowireless
aCS-basedsolution.First,anStrOMPalgorithmwaspro-
transmission,”inProc.IEEEInt.Symp.Inf.Theory,Jul.2013,pp.3050-
posedtodetecttheactiveMTDsexhibitingblock-sparsity3054.
[15]Y.NareshandA.Chockalingam,“Onmedia-basedmodulationusingRF
andstructuredsparsityoftheULsignals,whichimproved
mirrors,”IEEETrans.Veh.Tech.,vol.66,no.6,pp.4967-4983,June
theperformance.Then,anSIC-SSPalgorithmwasproposed
2017.
[16]E.Basar,“Media-basedmodulationforfuturewirelesssystems:A
fordetectingthedataofthedetectedMTDsbyexploitingthe
tutorial,”IEEEWirelessCommun.,vol.26,no.5,pp.160-166,Oct.
structuredsparsityofmedia-modulatedsymbolsforenhancing
2019.
theperformance.Furthermore,weanalysedthecomputational
[17]J.A.TroppandA.C.Gilbert,“Signalrecoveryfromrandommeasure-
mentsviaorthogonalmatchingpursuit,”inIEEETrans.Inform.Theory,
complexityoftheproposedalgorithms.Finally,oursimulation
vol.53,no.12,pp.4655-4666,Dec.2007.
quali?edthebene?tsoftheproposedsolution.
0018-9545(c)2020IEEE.Personaluseispermitted,butrepublication/redistributionrequiresIEEEpermission.Seehttp://www.ieee.org/publications_standards/publications/rights/index.htmlformoreinformation. |
|
