Research On High Spectrum Efficient Hybrid Precoding Algorithm In Mm Wave Massive MIMO Systems

Different from the traditional digital precoding,the RF chains cost and energy consumption are too high to realize,the hybrid precoding which combines digital precoding with analog precoding becomes a more practical scheme by introducing phase shifters in mm Wave massive MIMO system.However,the amplitude of the signal cannot be adjusted by phase shifters in the hybrid precoding,so that the analog precoding matrix is limited by the constant modulus constraint.Under such non-convex constrants,the optimization problem of hybrid precoding is difficult to be solved directly.We analyze and study the hybrid precoding design scheme in the mm Wave MIMO communication system,aiming at maximizing the spectrum efficiency of the system and taking into account the computational complexity in this paper.The details are as follows:An efficient and energy-saving hybrid precoding algorithm based on matrix decomposition optimization is proposed for the single-user massive MIMO communication system with sub-connection structure.The main idea of the algorithm is to divide the overall optimization problem into two stages,that is,first fix the digital precoding matrix,solve the analog precoding matrix,and then optimize the digital precoding matrix according to another one.In the first stage,the optimization problem is decomposed into a series of sub-problems according to the independent sub-matrix structure of the analog precoding matrix under the sub-connection structure.After the optimized analog precoding matrix is obtained,it is easy to solve the digital precoding matrix by using the minimum mean error(MMSE)method.Finally,the digital precoding matrix is normalized to satisfy the previously ignored power constraints.The simulation results demonstrate that the spectrum efficiency of the our algorithm is about 96% of that of the full digital precoding based on sub-connection structure,and it is obviously better than the existing algorithms.In addition,this paper also presents the simulation results of the algorithm performance when the channel state information(CSI)cannot be obtained completely and accurately in practice.Simulation results show that when the estimation accuracy of channel state information is 0.8,the spectrum efficiency of the proposed algorithm can reach about 93%of the optimal digital precoding performance under perfect CSI.Even if the accuracy of channel estimation is reduced to 0.6,the spectrum efficiency achieved by the algorithm is still about 92% of that of the perfect CSI..Considering the more widely used multi-user massive MIMO communication system,we propose a hybrid precoding scheme based on the idea of singular value decomposition and zero-forcing(ZF)algorithm which can effectively eliminate the interference between users in this paper.In this scheme,the full-digital zero-forcing precoding matrix is used as the criterion of the unconstrained optimal precoding matrix,and the singular value decomposition is performed to obtain the constrained analog precoding matrix with unconstrained analog phase information effectively.Then,according to the approximate optimal analog precoding matrix,the corresponding digital precoding matrix is obtained by using the classical ZF precoding algorithm with the idea of equivalent channel.Through theoretical analysis,the algorithm can be extended to the sub-connection structure and multi-user multi-antenna massive MIMO system.The simulation results demonstrate that the spectral efficiency of the our algorithm is approximately equal to that of the full digital ZF precoding algorithm,and is significantly better than the existing singular value decomposition hybrid precoding scheme.In addition,the performance of the algorithm is improved by about 15% when the number of RF chains is greater than the data stream.The performance of the algorithm is still robust to a certain extent when the channel estimation error exists in practice,that is,the perfect channel state information cannot be obtained.