Massive MIMO Precoding Algorithms Research

Massive multiple-input multiple-output(MIMO) is a novel cellular network architecture,that has emerged recently as a promising solution to meet the demanding spectral efficiency requirements of 5G systems. Precoding technology is utilized to eliminate interference, such as channel interference, cell fading and noise in Massive MIMO system by deploying a large number of antennas at the base station. Several challenges must be addressed in Massive MIMO systems, one of which is the design of low complexity precoding.In this paper, firstly,the channle characteristics of wireless communication systems are analyzed including fading and expansion. The Rayleigh fading channel, the related Rayleigh fading channel, and the Rice fading chanel is introduced. The system sum rate and precoding process of large-scale MIMO is analyzed. Secondly,The performance of zero-force algorithm(ZF) and the minimum mean square error precoding (MMSE) is compared by simulation.Finally, three kinds of low complexity algorithms are studied, that is, the successive over reaxiation(SOR) algorithm, the conjugate gradient algorithm(CG), the conjugate gradient and jacobi algorithm(CGJC) for solving the problem of high complexity of MMSE precoding algotithm in the matrix inversion process. The complexity, the working condition and the convergence of three kinds of low complexity algorithms is analyzed. The basic idea of the three kinds of low complexity algorithms is to avoid the matrix inversion process by solving the solution of of linear equations. The bit error rate and system sum rate performance of SOR, CG, CGJC algorithm is compared with MMSE algorithm under different number of transmitting antennas and different channnels models by constructing simulation platform.