Double IRS Assisted Channel Estimation For Millimeter Wave MIMO Communication

Millimeter-Wave(Mm W)communication technology,is a key technology in future mobile communication.Despite its numerous benefits,the propagation characteristics of this technology,such as being easily blocked by obstacles and weak penetration,have posed issues cannot be ignored.Therefore,intelligent reflective surface(IRS)technology can be used to assist millimeter-wave communication.Currently,a lot of research has been done on IRS channel estimation,and a series of channel estimation protocols and solutions have been proposed.However,there are few studies on scene channel estimation based on multi block IRS assisted millimeter-wave communication,and due to the increase in channel dimensions in this communication scenario,the optimization problem becomes more complex.This thesis primarily focuses on channel estimation for millimeter-wave communication scenarios with double IRS assistance in order to address the issues mentioned.Firstly,considering the system model of double IRS assisted MMW single user MIMO uplink wireless communication,a compressed sensing channel estimation scheme is proposed based on the sparsity of MMW channels.At first,channel encoding technology is used to transform the original channel,and the autocorrelation features of the perception matrix are used to evaluate the channel estimation performance in double IRS communication scenarios.Then a new algorithm,the generalized approximate message passing compressed sensing algorithm(GAMP)based on Matching pursuit,is introduced.It is different from the traditional orthogonal matching pursuit(OMP)and GAMP methods,and can better use the prior information in the channel.By improving channel estimation methods,the efficiency of channel estimation can be improved to a certain extent.After simulation experiments,the performance of double IRS channel estimation is significantly better than that of single IRS within the constrained upper bound range,which strongly proves the theoretical derivation.At last,the variation of channel estimation performance with the distance between two IRS blocks was simulated,and this phenomenon was analyzed and explained.Moreover,considering the system model of double IRS assisted millimeter-wave multiuser MIMO uplink wireless communication,a low rank alternating manifold optimized channel estimation scheme is proposed based on millimeter wave channels.At first,the relevant definitions in manifold theory are derived based on the low rank characteristics of millimeter wave channels,and then an Alternating Least Squares theory(ALS)is utilized to create a constrained optimization problem with multi-objective variables.Then,the channel variable estimation value is obtained by updating and iterating the target variable according to the fixed rank popular theory.At last,the simulation of the performance of the relevant channel estimation is performed and the simulation results are analyzed.Furthermore,simulation shows that the algorithm proposed has better channel estimation performance compared to traditional optimization algorithms.Finally,for the system model of double IRS assisted millimeter-wave multi user single antenna uplink wireless communication,a channel estimation scheme based on optimized phase shift matrix and two-stage alternating channel estimation protocol is proposed.At first,a new channel estimation protocol for fully open intelligent reflector units is proposed.Then,based on the maximum likelihood posterior probability(MAP)method,a multi variable alternative optimization estimation is performed to obtain the desired channel variable optimization iteration process.In order to further improve the performance of channel estimation,a continuous convex approximation(SCA)algorithm is used to optimize the phase shift matrix of the IRS.Simulation has demonstrated that the estimation protocol and optimization algorithm proposed possess certain benefits over conventional channel estimation algorithms in enhancing channel estimation performance.Furthermore,a correlation between normalized mean square error and complementary cumulative distribution function(CCDF)has been simulated to validate the stability of the suggested estimation protocol and optimization algorithm.