Research On Massive MIMO Precoding Algorithm Based On Channel Estimation

Massive Multiple Input Multiple Output?Massive MIMO?technology is one of the core technologies of 5G communication system research,which can greatly increase system's capacity by using a large number of antennas without additional bandwidth.However Massive MIMO systems will generate the pilot contamination when the number of antennas increases suddenly,which affects the accurate acquisition of CSI,resulting in the system capacity being limited.The focus of thesis is how to use effective measures or methods to reduce the influence of pilot contamination and improve the sum rat of users in cell.The major researches are as fellow:To begin with,this thesis introduces the Massive MIMO system model and makes a theoretical analysis of pilot contamination in massive MIMO system,which lays a theoretical foundation for the subsequent pilot contamination suppression.Secondly,aiming at the insufficient accuracy problem of least squares?LS?channel estimation algorithm,this thesis gives a channel estimation algorithm combining L₂ norm correction of variable step LMS algorithm based on S function and LS channel estimation in the uplink,which is used to improve the accuracy of the estimated channel matrix,so as to reduce the influence of pilot contamination.At first,in order to improve the comprehensive performance and anti-interference ability of S function variable step LMS algorithm,this thesis used the L₂ norm to control step update.Then,the improved LMS algorithm is applied to LS channel estimation,which get a channel matrix with higher accuracy by filtering.The simulation results show that this algorithm of the uplink channel estimation has lower normalized mean square error and normalized instantaneous error under the same pilot length,which can effectively reduce the influence of pilot contamination.Finally,in order to avoid the matrix inversion and reduce contamination in Massive MIMO,this thesis proposes the pre-coding strategy of Maximize Signal-to-Noise Ratio?Max-SNR?to reduce pilot contamination in downlink.In addition to estimating the channel,it is considered the error channel factors comprehensively in a real channel.The maximum SNR of the real channel is converted into a Rayleigh entropy optimization problem.The optimal matrix vector solution is obtained by searching for the optimal solution,which is used to judge the advantages and disadvantages of the channel and conduct pilot scheduling,so as to reduce pilot contamination.The simulation results show that the proposed algorithm has higher system capacity,lower bit error rate and better adaptability to changes in antenna number,and lower complexity,so it is more suitable for Massive MIMO systems with large data flow.