基于平行因子分解的大规模MIMO盲信道估计
|
摘要
针对单小区大规模多输入多输出(Multiple-Input Multiple-Output,MIMO)系统上行链路,提出了一种基于平行因子(Parallel Factor,PARAFAC)模型的信道估计方法。在基站端,将接收信号构造成PARAFAC模型,利用大规模MIMO系统中信道的渐近正交的性质,提出了一种基于约束二线性迭代最小二乘算法(Constrained Blinear Alternating Least Squares,CBALS),从而实现了盲信道估计。理论分析及仿真结果表明,所提方法与传统最小二乘方法相比,不仅提高了频带利用率而且具有更高的估计精度;与已有的二线性交替最小二乘方法(BALS)相比,所提算法有更快的收敛速度。
A novel tensor-based channel estimation algorithm is proposed for single-cell massive multipleinput multiple-output(MIMO) uplink systems.At the base station(BS),the received signal is modeled using parallel factor(PARAFAC). A constrained bilinear alternating least squares(CBALS) blind channel estimation scheme is proposed,which utilizes asymptotic orthogonality characteristic of massive MIMO system.Numerical simulation results illustrate that the proposed scheme not only has a superior estimation accuracy than traditional least square method,but also improves the spectral efficiency.Furthermore,compared with the bilinear alternating least squares(BALS) algorithm,the proposed algorithm has a faster convergence speed.
A novel tensor-based channel estimation algorithm is proposed for single-cell massive multipleinput multiple-output(MIMO) uplink systems.At the base station(BS),the received signal is modeled using parallel factor(PARAFAC). A constrained bilinear alternating least squares(CBALS) blind channel estimation scheme is proposed,which utilizes asymptotic orthogonality characteristic of massive MIMO system.Numerical simulation results illustrate that the proposed scheme not only has a superior estimation accuracy than traditional least square method,but also improves the spectral efficiency.Furthermore,compared with the bilinear alternating least squares(BALS) algorithm,the proposed algorithm has a faster convergence speed.
引文
[1] MARZETTA T L. Noncooperative cellular wireless withunlimited numbers of base station antennas[J]. IEEETransactions on Wireless Communications,2010,9(11):3590-3600.
[2] BAI Z D,SILVERSTEIN J W.No eigen value outside thesupport of the limiting spectral distribution of large-di-mensional sample covariance matrices[J].Annals of Prob-ability,1998,26(1):316-345
[3] TAOKA H,HIGUCHI K.Experiments on peak spectral ef-ficiency of 50bps/Hz with 12-by-12 MIMO multiplexingfor future broadband packet radio access[C]//Proceed-ings of 2010 4th International Symposium on Communica-tions,Control and Signal Processing. Limassol:IEEE,2010:1-6.
[4]尤力,高西奇.大规模MIMO无线通信关键技术[J].中兴通讯技术,2014,20(2):26-28.
[5] BARHUMI I,LEUS G,MOONEN M.Optimal training de-sign for MIMO OFDM systems in mobile wireless channels[J]. IEEE Transactionson Signal Processing,2003,51(6):1615-1624.
[6] CHANG M X,SU Y T. Model-based channel estimationfor OFDM signals in Rayleigh fading[J]. IEEE Transac-tions on Communications,2002,50(4):540-544.
[7] ZHANG J,ZHANG B,CHEN S,et al.Pilot contaminationelimination for large-scale multiple antenna aided OFDMsystems[J]. IEEE Journal of Selected Topics in SignalProcessing,2014,8(5):759-772.
[8] NGO H Q,LARSSON E G.EVD-based channel estima-tion in multicell multiuser MIMO systems with very largeantenna arrays[C]//Proceedings of 2012 IEEE Interna-tional Conference on Acoustics,Speech and Signal Pro-cessing.Kyoto:IEEE,2012:3249-3252.
[9] LI S,MU X,GUO X,et al.Tensor-based channel estima-tion approach for one-way multi-hop relaying communi-cations[J]. KSII Transaction on Internet&InformationSystems,2015,9(12):90-99.
[10]刘亚丽,李双志,段海鹏,等.面向能效优化的MIMO系统参数配置[J].电讯技术,2017,57(9):1035-1040.
[11] ZHOU Z,FANG J,YANG L,et al.Low-rank tensor decom-position-aided channel estimation for millimeter wave MI-MO-OFDM systems[J].IEEE Journal on Selected Areas inCommunications,2017,35(7):1524-1538.
[12] RAJIH M,COMON P,HARSHMAN R A.Enhanced linesearch:a novel method to accelerate PARAFAC[J]. SI-AM Journal on Matrix Analysis and Applicatons,2008,30(3):1128-1147.
[13] SIDIROPOULOS N D,GIANNAKIS G B,BRO R. Blindparafac receivers for DS-CDMA systems[J].IEEE Transac-tions on Signal Processing,2000,48(3):810-823.
[14] LIU X,XU Z. Structure constrained PARAFAC modelwith application to signal processing[C]//Proceedingsof 2008 IEEE International Symposium on KnowledgeAcquisition and Modeling Workshop. Wuhan:IEEE,2009:651-654.
[2] BAI Z D,SILVERSTEIN J W.No eigen value outside thesupport of the limiting spectral distribution of large-di-mensional sample covariance matrices[J].Annals of Prob-ability,1998,26(1):316-345
[3] TAOKA H,HIGUCHI K.Experiments on peak spectral ef-ficiency of 50bps/Hz with 12-by-12 MIMO multiplexingfor future broadband packet radio access[C]//Proceed-ings of 2010 4th International Symposium on Communica-tions,Control and Signal Processing. Limassol:IEEE,2010:1-6.
[4]尤力,高西奇.大规模MIMO无线通信关键技术[J].中兴通讯技术,2014,20(2):26-28.
[5] BARHUMI I,LEUS G,MOONEN M.Optimal training de-sign for MIMO OFDM systems in mobile wireless channels[J]. IEEE Transactionson Signal Processing,2003,51(6):1615-1624.
[6] CHANG M X,SU Y T. Model-based channel estimationfor OFDM signals in Rayleigh fading[J]. IEEE Transac-tions on Communications,2002,50(4):540-544.
[7] ZHANG J,ZHANG B,CHEN S,et al.Pilot contaminationelimination for large-scale multiple antenna aided OFDMsystems[J]. IEEE Journal of Selected Topics in SignalProcessing,2014,8(5):759-772.
[8] NGO H Q,LARSSON E G.EVD-based channel estima-tion in multicell multiuser MIMO systems with very largeantenna arrays[C]//Proceedings of 2012 IEEE Interna-tional Conference on Acoustics,Speech and Signal Pro-cessing.Kyoto:IEEE,2012:3249-3252.
[9] LI S,MU X,GUO X,et al.Tensor-based channel estima-tion approach for one-way multi-hop relaying communi-cations[J]. KSII Transaction on Internet&InformationSystems,2015,9(12):90-99.
[10]刘亚丽,李双志,段海鹏,等.面向能效优化的MIMO系统参数配置[J].电讯技术,2017,57(9):1035-1040.
[11] ZHOU Z,FANG J,YANG L,et al.Low-rank tensor decom-position-aided channel estimation for millimeter wave MI-MO-OFDM systems[J].IEEE Journal on Selected Areas inCommunications,2017,35(7):1524-1538.
[12] RAJIH M,COMON P,HARSHMAN R A.Enhanced linesearch:a novel method to accelerate PARAFAC[J]. SI-AM Journal on Matrix Analysis and Applicatons,2008,30(3):1128-1147.
[13] SIDIROPOULOS N D,GIANNAKIS G B,BRO R. Blindparafac receivers for DS-CDMA systems[J].IEEE Transac-tions on Signal Processing,2000,48(3):810-823.
[14] LIU X,XU Z. Structure constrained PARAFAC modelwith application to signal processing[C]//Proceedingsof 2008 IEEE International Symposium on KnowledgeAcquisition and Modeling Workshop. Wuhan:IEEE,2009:651-654.