Channel Estimation For Intelligent Reflecting Surface Assisted Millimeter-Wave Massive MIMO Systems

As a technology for reconfiguring wireless communication environment with low hardware cost and power consumption,Intelligent reflecting surface(IRS)has attracted extensive attention.However,the application of IRS in practice requires accurate channel state information(CSI),and in the IRS assisted large-scale multiple-input multiple-output(MIMO)system,it is a challenging task to get CSI: Firstly,all IRS elements are passive,which cannot transmit,receive,or process any pilot signals to realize channel estimation.Secondly,since an IRS usually consists of hundreds of elements,the dimension of channels to be estimated is much larger than that in conventional systems,which will result in a sharp increase of the pilot overhead for channel estimation.In this regard,this thesis has carried out the following analysis and research:Firstly,This thesis studies the channel estimation algorithm of IRS assisted mm Wave massive MIMO system.In view of the overwhelming training overhead and poor performance of the traditional algorithm,the Saleh-Valenzuela channel model is considered,and a new channel estimation method based on the double-structure compressed sensing theory is proposed.We construct the channel model of from User Terminal(UT)to IRS and IRS to Base Station(BS),and obtain the cascaded channel of UT-IRS-BS.The channel estimation problem is modeled as a Compressed Sensing(CS)recovery problem by using the double-structure sparse characteristics of the angular-domain cascaded channels in the IRS-assisted mm Wave massive MIMO system to reduce pilot overhead.Simulations demonstrate that,compared with the traditional algorithm,the proposed algorithm can obtain better channel estimation performance and lower pilot overhead.Secondly,channel estimation for user of multi-antenna scenarios is investigated.In this chapter,we address the receiver design for an IRS-assisted massive MIMO system via a tensor modeling approach.Considering a structured time-domain pattern of pilots and IRS phase shifts,we proposed a channel estimation method that relies on the parallel factor(PARAFAC)tensor modeling of the received signal and optimize it with the Enhanced Line Search(ELS)algorithm.The numerical results show that the proposed receiver The effectiveness of the proposed estimation method verifies the accuracy of the proposed estimation method and its superiority compared with traditional channel estimation,and the ELS algorithm greatly reduces the number of iterations of the method and reduces the convergence time by nearly one-half.