Research And Optimization Of Precoding Algorithm For Large Scale MIMO Systems

Massive multiple-input multiple-output(MIMO)technology allows to meet the needs of users of the configuration of a large number of antennas at the base station,which has a lot of high capacity,high link reliability,high spectrum efficiency and energy efficiency and other quality characteristics,has become an indispensable technology in 5G communication system.However,there are still some problems to be solved in practical applications,such as a sharp increase in the complexity of the precoding algorithm due to the increase in the number of antennas.How to reduce the computational complexity of the algorithm has become a difficult problem to be solved at the present time without consuming or even improving the performance of the precoding algorithm.This thesis focuses on the methods of reducing the computation of precoding algorithms and guaranteeing or improving performance in Massive MIMO systems.In this thesis,we first study the traditional Linear Precoding algorithms and previous algorithms based on the same theory and apply other optimization algorithms to explore the way to reduce the algorithm complexity.Then,based on the random matrix principle,we analyze the new characteristics of the channel matrix in the massive MIMO system,and try to improve the algorithm based on these characteristics.Then,the nonlinear precoding is studied and analyzed.Based on the reduction of inter user interference,how to optimize the performance of the algorithm and reduce the computational complexity of the algorithm.The main innovations and contributions of this article are as follows:(1)An improved precoding algorithm(RZF-SOR)for massive MIMO system is proposed.In order to improve the complexity of the algorithm caused by the large number of base station antennas,an improved algorithm to reduce the computation of the system is proposed.On the basis of the original RZF algorithm,this algorithm uses the super relaxation iterative method(SOR)to replace the inverse matrix.The optimal solution of the parameter relaxation factor is solved according to the principle of random matrix,and the convergence rate of the algorithm is improved.The simulation results show that RZF-SOR algorithm can achieve approximate performance with RZF algorithm with very small number of iterations,and is superior to Neumann algorithm,and reduces the complexity of algorithm of an order of magnitude.(2)An improved precoding algorithm ADMM for massive MIMO system based on interuser interference is proposed.A new ADMM algorithm is proposed,considering the mutual interference between the receivers and the high computational complexity of the algorithm.The algorithm takes the digital signal with limited character set as the object of study,and aims at reducing the interference among users.The alternating direction multiplier algorithm is used to transform the transmitted signal vector into vector iteration.In order to ensure the fast convergence of the algorithm,the parameter damping factor is put forward,and the iterative matrix is obtained by using the principle of random matrix,and the design steps of the algorithm are given.Finally,the computation and bit error rate of the ADMM algorithm are analyzed.The simulation results show that compared with the recently proposed SQUID algorithm,the computational complexity of the ADMM algorithm is reduced,and the performance of BER is better.