Study On The Optimization Problems In MIMO Systems Based On The Channel Time-varying Characteristics

With the continuous increase on the demand on high-speed network and the continual growth of data in wireless communication systems,the 5th generation wireless network and related technologies have been developed rapidly.MIMO systems,together with massive MIMO systems,which can provide considerable multiplexing gain and diversity gain to the communication network,have become an indispensable part in the embryonic form of 5G.In recent years,for a better quality of service in the communication systems,researches about MIMO systems are made continuously in academia,and some creative and practical MIMO system models has been proposed.The representative one is the time-varyingMIMO system.Considering the temporal correlation in the actual systems,the time-varying MIMO systems study on a variable channel and offer some solutions to optimal problems based on system evolution models and related algorithms.Different from the researches on invariable systems,the researches on time-varying systems consider the uncertainty in the actual systems and have a better practical significance.Another hot topic about MIMO systems is distributed MIMO systems,which can offer the system a better spectral efficiency and a larger coverage.The distributed MIMO systems are supposed to be a key technology to meet the increasing network demand.Compared with traditional centralized MIMO systems,the distributed MIMO systems can also provide better service for the cell?edge users and decrease the interference between the transmit antennas.Based on the temporal correlation in the actual systems,this paper will study the channel tracking scheme and the precoding matrix design in single-user MIMO system and multi-user MIMO system.respectively.Meanwhile,the system models in this paper will contain centralized MIMO systems and distributed MIMO systems.The main innovations in this paper include:Firstly,for the time-varying centralized MIMO system,this paper proposes a novel time-varving evolution model to study time-varying centralized MIMO channel.Fuirther,with the proposed channel evolution model,the paper piroposes a low overlhead channel tracking schenme,including the basic idea,the description and the analysis of the parameters in the proposed channel tracking scheme.Meanwhile,for a better performance analysis,this paper considers the performance of the spectral efficiency with different data stream number and different power allocation strategy.At last,the author determines the optimal channel tracking interval with the threshold on the receive power loss.Secondly,for the time-varying distributed single-user MIMO system,this paper introduces a system model of the distributed single-user MIMO system,together with the temporal channel evolution model.As the introduction of the distributed system,the channel evolution model of distributed single-user MIMO system is no longer a form of Mar-kov chain.Further,based on the system model,this paper proposes the iterative search algorithm to design the precoding matrix in the time-varying distributed single-user MIMO system.In addition,this paper gives the description,the choice of the parameters,the error analysis,and the power allocation strategy in the proposed algorithm.Meanwhile,compared with other precoding matrix design algorithms for distributed MIMO system,the iterative search algorithm has a better performance and a better adaptation for time-varying systems.Thirdly,for the time-varying distributed multi-user MIMO system,this paper introduces a system model of the distributed multi-user MIMO system,together with the temporal channel evolution model.The multi-user system makes the channel evolution model more complicated so that there is no explicit expression between the channel matrices of two adjacent channel indexes.The author has to introduce several auxiliary matrices to finish the channel evolution model.Further,with the proposed system model,the author proposes two optimal algorithm to improve the sum rate in the time-varying distributed multi-user MIMO system.One-dimension search algorithm can obtain a better performance of the sum rate,and cyclic MMSE search algorithm has a lower calculation complexity.From the simulation results,the two proposed algorithms can both improve the channel capacity in the time-varying distributed multi-user MIMO system effectively.