Research On Hybrid Precoding Technology For Millimeter Wave Massive MIMO Systems

As one of key technologies of the fifth-generation wireless communication(5G),massive multiple-input multiple-output(MIMO)technology is expected to gradually replace the traditional MIMO technology.Millimeter wave is favored by 5G communication technology because of its own vast spectrum resources that have not been developed.Millimeter wave massive MIMO can significant improve the spectral efficiency and channel capacity.However,its greatly increasing number of antennas has brought some challenges to system design,such as the increase of radio frequency chain,high cost and high-power consumption.Since the traditional full digital precoding requires the number of radio frequency chains to be equal to that of antennas at the base station,its implementation cost is too high.Therefore,the hybrid precoding technology that can effectively reduce the number of radio frequency chains and improve system performance,has become a crucial technique for massive antenna systems.In addition,because the traditional technology of orthogonal multiple access(OMA)is unable to meet the requirements of the growing users and device access,the combination of non-orthogonal multiple access(NOMA)and other 5G technologies is also a current focus of research.In this paper,the hybrid precoding in millimeter wave massive MIMO is analyzed and investigated in two different scenarios: A.massive antennas at base station,single-user with multiple-antennas at the user side;B.massive antennas combined with NOMA at the base station,multiple-users with single-antenna at the user side.The main work of this article is incorporated as follows:1.Based on the millimeter wave massive MIMO system,the impact of hybrid precoding on the system performance is analyzed in terms of two performance indicators.Firstly,the influence of digital precoding based on matrix decomposition on bit error rate is studied.Secondly,the research analyzes the effect of the hybrid precoding combined with multiple analog precoding schemes on the system spectrum efficiency.2.For millimeter wave multi-user NOMA systems,a user grouping scheme based on distance difference value is proposed by applying the hybrid precoding scheme with alternate minimization and iterative optimized power allocation method.The aim is to improve the spectrum efficiency of the system.The simulation shows that the proposed scheme is superiorto other hybrid precoding schemes,and the iterative optimization method also outperforms the random assignment method.3.In the millimeter wave multi-user NOMA systems,the dual-user grouping scheme including strong and weak users is proposed by using hybrid precoding combined with regularized zero-forcing and analog quantizer.The particle swarm-based power allocation scheme is utilized to optimize the system spectral efficiency.The simulation shows that the system spectrum efficiency with full-connected structure is better than that of partially-connected structure;the spectrum efficiency of the NOMA system is also better than the traditional OMA scheme.Moreover,the energy efficiency of the considered system is analyzed,and the performance of different power optimization algorithms is compared.