Research On The Theory Of Millimeter Wave Extremely Large-scale Massive MIMO Channel Estimation

With the change of times,scientific and technological innovation is changing rapidly.The development of The 5th Generation Mobile Communication system has been quite mature,and the research of The 6th Generation Mobile Communication system(6G)is in full swing.Extremely large-scale Multiple Input Multiple Output(MIMO)has become one of the hot technologies in 6G system research due to its advantages of strong reliability,good robustness,simplified calculation,high security,high spectrum utilization,high energy efficiency and high transmission capacity.However,the application of extremely large-scale MIMO technology is bound to increase the difficulty of obtaining channel state information.With the increase of the number of antenna arrays,millimeter wave channels gradually show spherical wavefront propagation and spatial non-stationary characteristics.Therefore,in this paper,the Compressed Sensing(CS)theory and the Alternating Direction Method of Multipliers(ADMM)theory are used.Based on the channel propagation characteristics of extremely large-scale millimeterwave MIMO systems.In this paper,the estimation algorithm of millimeter wave ultra-large scale MIMO channel with accurate estimation performance and low complexity is studied.The specific research content is as follows:(1)Based on the propagation characteristics of the spherical wavefront of extremely largescale MIMO,a channel estimation scheme for extremely large-scale MIMO of millimeter wave based on partition array is proposed.A full MIMO channel matrix is reconstructed by dividing the large array with many antennas at the at the transmitting and receiving ends into a small number of sub-arrays,each of which is estimated independently.In a parallel approach,the combined estimation problem of Angle of Departure(AoD)at the transmitted end and Angle of Arrival(AoA)at the received end is transformed into an independent signal model of AoA and AoD.Then singular value decomposition method is used to obtain the initial sparse support set of signals.Then,Taylor expansion iterative method is used to optimize the off-grid errors and obtain the grid points closest to the real angle.The least square method is used to estimate the path gain of the channel,achieve the accurate estimation of the subarray,and finally reconstruct the complete channel matrix.The accuracy and complexity of the proposed algorithm are verified by simulation experiments.(2)Based on the abundant scatterers in the actual channel propagation environment,an extremely large-scale millimeter wave MIMO channel estimation scheme based on alternating direction multiplier method and double scattering cluster is proposed.The single scatterer model cannot fully satisfy the propagation of the channel in the time domain and the angle domain,so we use the double scatterer model.There are single-bounce scatterers near the receiving end and the transmitting end.The transmitting end reaches the receiving end through the double bounce scatterers.The propagation path between the double bounce scatterers is abstracted by virtual link delay.Then,the optimization model of the channel matrix and scattering gain matrix is constructed by using the alternating direction multiplier method,which takes advantage of the low-rank and sparsity of the millimeter-wave channel.Under the constraint condition of lagrange multiplier construction,the iteration model is updated continuously until the iteration result is close to the global optimal solution.Simulation results show the feasibility and fast convergence of the proposed algorithm.