Research On Precoding Technology For 5G Mobile Communication

In the past decade,the exponential growth of mobile data has brought unprecedented challenges to wireless service providers.To this end,the fifth generation(5G)mobile communication network has been proposed.In order to achieve the higher total data rate and lower latency requirement in5 G,key technologies including massive multiple input multiple output(MIMO)and millimeter wave(mm Wave)frequency bands are employed.In millimeter-wave massive MIMO systems,it is essential to investigate the signal processing on the transmitter and receiver,especially the precoding process-ing technology on the transmitter,which can not only effectively eliminate multi-user interference,thereby greatly improving system capacity,but also simplify the design of receiver to solve the power consumption and volume issues at the user terminal.This thesis studies the linear precoding technology in massive MIMO and the hybrid precod-ing technology in mm Wave MIMO.In massive MIMO systems,a linear precoding algorithm is put forward to reduce the computational complexity.For the mm Wave environment,two types of hy-brid precodings are proposed to reduce hardware complexity and maximize the spectral efficiency by using convex optimization theory.The specific contributions of this thesis are as follows:1)The system models of point-to-point MIMO,multi-user MIMO and massive MIMO are com-pared.The typical linear precoding technology,including: SVD(Singular Value Decomposition),MF(Matched Filter),ZF(Zero Forcing),BD(Block Diagonalization)and RZF(Regularized Zero Forcing)precoding,are analyzed in single-user MIMO and multi-stream multi-user MIMO.In addi-tion,the bit error rate and the system sum-rate performance of the common linear precoding algorithm are compared in simulation results.2)In order to avoid the high computational complexity caused by the inversion of the Gram matrix in ZF/RZF precoding,the precoding algorithms based on NSE(Neumann Series Expansion)and TPE(Truncated Polynomial Expansion)are analyzed.As well,the precoding algorithms based on JM(Jacobi Method),GS(Gauss-Seidel),and SOR(Successive Over Relaxation)are also analyzed for preprocessing the transmitted signals by a simple iterative method.Finally,a precoding algorithm based on CG(Conjugate Gradient)is proposed.The calculation complexity analysis shows that the proposed CG precoding algorithm is lower than other algorithms except for SOR,and the bit error rate and sum-rate performance of CG algorithm are also better than the NSE algorithm and close to the RZF precoding.3)Several recently proposed hybrid precoding algorithms for mm Wave single-user and multi-user MIMO are analyzed.In order to further reduce the hardware complexity of the mm Wave MIMO system,a alternate optimization hybrid precoding algorithm for sub-connected structure and an ED-based hybrid precoding algorithm with low-resolution DAC(Digital to Analog Converter)are pro-posed.Simulations demonstrate that the spectral efficiency performance of the proposed alternating optimization hybrid precoding algorithm is better than ICD(Iterative Coordinate Descent)algorithm and ED-based hybrid precoding algorithm are better than OMP(Orthogonal Matching Pursuit)algo-rithm,which indicates that they can be well applied to mm Wave channels.