Analysis Of Hybrid Precoding Schemes For Massive MIMO Systems

With the advent of 5G era,the technology of massive MIMO replaces the conventional MIMO gradually,which becomes one of the key technologies to the next wireless communication system.With the increasing number of antennas,the energy efficiency and the channel capacity improve greatly.However,the technology also brings some challenges,such as pilot contamination and the hardware cost problem.The number of required RF chain in the conventional fully-digital precoding is the same with that of antennas,which causes the cost problem.Hybrid precoding is proposed under this background to reduce the number of required RF chains and becomes an important research area.Hybrid precoding in massive MIMO systems has been researched in the paper.From the perspective of antenna structure,two kinds of hybrid precoding schemes based on different antenna structures are proposed innovatively.The methods in convex approximation and stochastic optimization are adopted to solve the optimization problems.The complexity and convergence are analyzed and the performance is verified through simulation results.The main contributions of the paper are as follows:1)A mixed-timescale per-group hybrid precoding scheme is proposed.The hybrid precoding optimization problem is designed to max the minimal statistical SLNR among all the groups and based on the adaptive partially-connected antenna structure,which considers the hardware constraints on antenna structure.Specifically,the RF precoder is adaptive to the long-term channel statistics and the baseband precoder is adaptive to instantaneous CSI,which reduces the CSI signaling overhead.The proposed algorithm is compared with the existing algorithms based on the fully-connected and partially-connected antenna structures,which shows that the proposed algorithm can achieve a better balance between performance,complexity and CSI signaling overhead.2)Considering the total power(wireless transmit power and hardware power consumption)minimization problem,an adaptive fully-connected antenna structure is proposed to adjust the active number of APSs and RP chains according to the objective of the optimization problem.A sparse hybrid precoding(SHP)scheme is also proposed based on the proposed antenna structure.Considering the average data rate of users,the optimization problem is aiming at minimizing the total power consumption,including the wireless transmit power and hardware consumption power.A constrained stochastic successive convex approximation for SHP(CSSCA-SHP)is proposed to solve the optimization problem.It is proved that the proposed algorithm has lower power consumption and higher energy efficiency with a good convergence speed compared to the other existing algorithms.