Research On Equalization And Beamforming Algorithms Of MIMO Single Carrier Frequency Division Multiple Access Systems

As wireless communication systems evolve to higher bandwidth,various data traffic demands and more complicated application scenarios,multiple-input multiple-output(MIMO)and broadband multiple-access technologies have become the key technologies for future wireless communication systems.Single carrier frequency division multiple access(SC-FDMA)is one of the most important broadband multiple-access technologies.Therefore,its combination with MIMO technologies has become an essential trend.Nowadays,SC-FDMA has been selected as the uplink multiple-access of the 4th generation mobile communication systems(LTE-Advanced systems),and attracts more attentions from both academic and communication industry fields.In MIMO broadband communication systems,inter-symbol-interference caused by frequency selective channel fading and inter-stream-interference caused by multiple-antenna data streams are two important problems that degrade the performance of communication systems.This dissertation focuses on how to deal with the interference problems caused by frequency selective fading and multiple-antenna in terms of receiver equalization,transmitter beamforming and relay beamforming.Firstly,the performance of linear frequency domain equalizations in SC-FDMA systems,which include zero forcing(ZF)and linear minimum mean-squared error(MMSE)equalizations,are deeply investigated.The impacts of system bandwidth,channel lengths,number of transceiver antennas on the performance of different equalization algorithms are analyzed,respectively.In frequency selective Rayleigh fading channels,closed form expressions for the upper and lower bounds on the achievable sum rates of ZF and MMSE receivers are derived,which apply to scenarios with arbitrary numbers of subcarriers and arbitrary-length channels.By analyzing the bounds in the high SNR,low SNR,and large scale MIMO regimes,new results are obtained.For instance,when the number of transmitter and receiver antennas grows with a fixed ratio of less than one,the upper and lower bounds are approximately identical,therefore,the variation of system bandwidth and channel lengths has a neglectable effect on the performance.As for the nonlinear receivers,the per subcarrier maximum likelihood(PSML)receiver is proposed,and its performance is analyzed.The results show that the proposed PSML receiver is optimal in flat fading channels.Compared with the conventional ML receivers,the computational complexity of the proposed receiver algorithms is significantly reduced.Then,beamforming technologies of SC-FDMA systems are studied.Specifically,for SC-FDMA systems with frequency domain equalization,the optimal beamforming algorithms are derived,based on the sum of minimum mean-squared errors criterion.With imperfect channel state information(CSI)at the transmitter,CSI errors on each subcarrier channel matrix are modeled belong to an un certainty set with constrained spectral norm.Based on this CSI error model,the robust beamforming algorithms are derived by using theory of maj orizations and singular value inequalities.Simulation results show that,compared with equal power allocation algorithms,the proposed beamforming algorithms can provide better performance of bit error rate and sum rate,especially for systems with small CSI errors.Compared with the non robust beamforming algorithms,the performance is still superior,especially for systems with large CSI errors.Lastly,relay beamforming problem for cooperative SC-FDMA systems are investigated.With perfect CSI at the relay node,two beamforming optimization problems are considered,which are maximizing the SINR with constrained relay transmit power and minimizing the relay transmit power with constrained SINR.Results show that the relay beamforming can not meet arbitrary given SINR threshold when source node transmit power is fixed.To solve this,a joint optimization of source transmit power and relay beamforming algorithm is proposed,which can meet arbitrary SINR demand and minimize the sum transmit powers of source and relay nodes.Simulation results show that this algorithm can save more transmit powers of relay and source nodes,compared with algorithms derived for OFDM systems.Moreover,the relay beamforming algorithms under imperfect CSI are also derived,which can provide better performance than other beamforming algorithms.