Research On Channel Estimation For Intelligent Reflecting Surface Assisted MIMO Systems

The commercial use of the 5th Generation Mobile communication Technology(5G)makes the number of intelligent wireless devices grow rapidly,and the requirements of data transmission rate are increasing.Although 5G is in the initial stage of large-scale commercial use,its relevant technical features need to be further improved.To satisfy the needs of users and adapt to new application scenarios as well as the trends of network in the future,researchers have proposed and developed the 6th generation Mobile communication Technology(6G).Unlike previous wireless communication technologies,the Intelligent reflecting Surface(IRS)can intelligently control the transmission environment to improve system spectrum efficiency and energy efficiency and has the advantages of low cost and easy deployment.Therefore,it is expected to play a particularly important role in future 6G communication networks.However,the ideal performance of IRS-assisted communication system relies on accurate Channel State Information(CSI),and accurate CSI acquisition depends on the channel estimation.Due to the passive characteristics of IRS reflection element and the large number of reflection elements,it’s a major challenge to solve the channel estimation problem of IRS-assisted MIMO communication system.Therefore,this paper mainly studies the efficient channel estimation algorithm of IRS-assisted MIMO communication system in Rayleigh channel and millimeter wave channel.In the first part,the channel estimation algorithm based on IRS-assisted MIMO multi-user system in Rayleigh channel is studied.Firstly,the model of the system is introduced and a phased channel estimation scheme is outlined,which uses channel correlation and the fact that different users share the same IRS link to the base station to jointly estimate all users.Based on this idea,the length of minimum pilot sequence for three stages is analyzed,and the advantages of the phased channel estimation scheme in comparison to the pilot cost of different users are analyzed by simulation.Then the detailed flow of the three-phase channel estimation scheme is given,and the estimation performance of the three-phase channel estimation algorithm is analyzed by simulation.The results of simulation show that the phased channel estimation scheme can achieve reliable channel estimation with the relatively low length of pilot sequence,and the estimation performance under the same pilot cost is obviously better than the scheme with different users.Finally,based on this channel estimation scheme,the fairness performance of IRS-assisted MIMO multi-user downlink system under imperfect CSI is studied.By alternately optimizing power distribution and phase shift matrix,the system makes the users with the weakest spectrum efficiency maximize under the premise of satisfying the fairness of users.The results of simulation show that the spectrum efficiency of the weakest users is significantly improved by the reliable CSI obtained by the phased channel estimation scheme.In the second part,the channel estimation algorithm based on IRS-assisted MIMO multi-user system in millimeter wave channel is studied.Firstly,the architecture of the system is introduced,and the sparsity of millimeter-wave channel is analyzed,and the row sparsity and column sparsity of the channel matrix are described.Then,according to the same IRS-to-base station link shared by different users and part of the IRS-to-user link shared by different users,the channel estimation is divided into three stages.After the detailed steps of three stages are introduced,the complexity of channel estimation scheme is given.The results of simulation show that the phased channel estimation scheme of millimeter wave channel still has obvious advantages in pilot cost under the premise of satisfying the reliable channel estimation.Finally,the channel estimation algorithm is used to maximize the downlink weighted sum rate by alternately iterating and jointly optimizing the active and passive beamforming matrix in an imperfect channel.The results of simulation show that the reliable CSI obtained by the phased channel estimation scheme has good weighted and rate performance after joint beamforming.