Research On User Pairing And Precoding Methods For LTE Downlink MU-MIMO System

Multiple-Input Multiple-Output(MIMO),one of the critical techniques that can raise system throughput and spectral efficiency,is widely used in various communication systems such as 3GPP LTE/LTE-Advanced,IEEE802.11,IEEE802.16,et al...In recent years,Multi-user MIMO systems have won great research attention,since no extra cost are needed with user equipment to support multiple terminals communicate with BS simultaneously in a way of time and frequency sharing.As known to all,the number of users with transmission need is always larger than the supply level in practical cellular systems,and there exists inter-user interference in MU-MIMO systems.Consequently,to acquire multi-user diversity gain and spatial multiplexing gain,user pairing and precoding are of great significance to MU-MIMO system.In this paper,we study user pairing and precoding algorithms detailedly in MU-MIMO downlink systems.Besides,we improve the existing methods in aspects of throughput and computation.To sum up,major work and contribution are as follows:1.Due to the high cost of the current BD precoding methods,a novel solution for BD precoding is proposed.This novel BD precoding method introduce the orthogonal decomposition to the recursive derivation of precoding matrix.Simulation results shows that it can achieve the same performance as the existing BD precoding methods while with much lower complexity.2.Besides,in the complex case of multi-user pairing,two novel multi-user pairing algorithms are proposed.The first one defines a new user selection metric using the characteristics of BD and geometric angles of subspaces.Thanks to the consideration of determinant of the effective channel matrix of selective and new adding pairing users,this new method can achieve better tradeoff between system throughput performance and computation.Unlike existing pairing methods that using the greedy strategy,the second novel pairing scheme we proposed applies coordinate ascent method which consider the impact of the new user not only on the selective pairing users but also on the potential pairing users every iteration when selecting a new user to the pairing user set.Consequently,the novel pairing method surpasses the exiting pairing schemes in terms of ccapacity performance.