Research On Hybrid Precoding Algorithm In Millimeter Wave Massive MIMO System

The arrival of the 5th Generation(5G)mobile communication indicates that mobile communication technology has entered an era of multi-technology and multi-service integration.In 5G,the combination of millimeter wave(mm Wave)and large-scale multiple input multiple output(MIMO)technology can effectively improve spectrum efficiency,but the large number of antennas and system costs limit the use of traditional all-digital precoding.As an effective solution,hybrid precoding reduces the number of RF chains and hardware costs.Therefore,this article mainly studies the hybrid precoding algorithm for single-user mm Wave massive MIMO systems,and presents two alternating optimization schemes,which have advantages in complexity and performance respectively.The main research contents of this paper are as follows:First,the channel model under mm Wave,the transmission model of massive MIMO and the classic algorithm of hybrid precoding are studied.After that,the maximization of spectrum efficiency is studied as the optimization objective function,and the matrix theory and optimization methods are used to analyze it in depth.Secondly,a hybrid precoding scheme based on alternating optimization is given.The core idea of the scheme is to use matrix operations to decompose the optimization problem and transform it into the sum of a series of sub-problems.Among them,the variable of each sub-problem includes the column of the analog domain matrix of the transmitter and receiver,so the alternating optimization method is selected to solve it.In addition,due to the limitation of the element amplitude in the analog domain matrix,the projected gradient ascent method is adopted for the sub-problem,so that the solution is always in the constraint set.After the analog domain matrix is solved,matrices in the digital domain are derived from the unitary matrix of the equivalent channel avoiding the iterative process.Simulation and calculation results show that the algorithm has low complexity and can achieve better performance.In addition,considering the influence of a single element in the analog precoding matrix on the optimization problem,another hybrid precoding scheme based on two-layer alternating optimization is given.The outer layer alternately optimizes the receiver and transmitter,and the inner layer alternately optimizes the digital and analog domain matrix at the receiver or transmitter.In the inner layer alternation,the optimization problem is decomposed by the equivalent channel and converted into the sum of a series of sub-problems.For each sub-problem,the analog domain matrix adopts the method of solving elements one by one,and the digital domain matrix is solved by the least square method.After that,perform multiple alternating iterations of the inner and outer layers to obtain hybrid precoding matrix and merge matrix.Simulation analysis shows that the proposed algorithm can achieve better performance.This paper presents two solutions to the non-convex constraint problem in hybrid precoding design.After verification,they have achieved better performance and lower complexity,so they have higher availability in mm Wave massive MIMO systems.