Research On Channel Reciprocity Error Of Massive MIMO Systems

After thirty years of rapid growth,mobile communication has become the basic information network of human society and has a significant impact on the national economy.By configuring a large number of antennas at the base station,large-scale Multiple-Input Multiple-Output(MIMO)can provide additional space gain,thereby effectively increasing the user's received SINR and improving the system spectrum efficiency and energy efficiency several times,therefor it has attracted wide attention in academia and industry.In large-scale MIMO system,time-division duplexing will become the main duplex mode.However,in actual communication,the complete communication channel includes not only the wireless transmission channel,but also the radio frequency circuit in the transceiver in the base station and the user terminal.The asymmetric gain of the radio frequency circuits at both ends and the mutual coupling effect caused by the densely placed antenna will undermine the reciprocity of the channel,so reciprocity calibration is needed during the downlink transmission process.This paper studies the reciprocity error theory of large-scale MIMO channel.First,this thesis studies the system performance in the presence of channel reciprocity error in large-scale MIMO systems.According to the coexistence of channel reciprocity error model caused by radio frequency gain error and mutual coupled error at the base station side,theoretical analysis is made based on regularization zero-forcing precoding,and the closed expression of system sum rate is given.The influence of channel reciprocity error on the system performance is analyzed,and the optimal regularization parameter is given according to the maximization rate criterion.The simulation results show that the channel reciprocity error in the system will lead to the rapid decline of the system sum rate,and with the increase of channel reciprocity error,the system performance will be further reduced.Therefore,it is necessary to correct the channel reciprocity error at the base station.In this case,by calculating the optimal regularization parameter,the system sum rate can be significantly improved in a high SNR region.Then,a reciprocal error calibration method for large-scale MIMO systems under non-reciprocal channels is proposed.The final correction matrix is obtained by additively linearly correcting the estimated reciprocity error matrix.Because of the influence of noise in the process of estimating the reciprocity error matrix,the resulting reciprocal error matrix inevitably has errors,and the correction can not achieve good performance.According to the rate maximization criterion,the additive correction matrix is introduced to get the final correction matrix.The simulation results show that by choosing the optimal additive correction matrix,the proposed reciprocity correction method can achieve a higher correction accuracy.After the base station side transmit signal is corrected,the downlink transmission and rate performance are improved.