Research On Channel Estimation Methods For B5G Ultra Massive MIMO System

In order to meet the increasing transmission rate demand of users and solve the problem of scarcity of spectrum resources,the combination of massive multiple-input-multiple-output(MIMO)and millimeter wave technology is the evolution direction of future mobile communications.With the development of emerging technologies,the fifth generation(5G)and beyond 5G(B5G)mobile communication systems have greatly improved in terms of peak rate,spectrum efficiency,and mobility management.Accurate and efficient estimation of channel state information(CSI)is the key to achieving the expected performance indicators of B5 G system.Therefore,the research on the channel estimation method of the ultra massive MIMO system is very important to the design of the B5 G mobile communication system.The main contributions of this paper are as follows:(1)Combining the sparse characteristics of millimeter-wave channels and the low-rank nature of large-scale arrays,the problem of millimeter-wave massive MIMO channel estimation is studied.Firstly,in the case of the uniform linear array,the received signal is decomposed by Kronecker dimensionality reduction,and the angles of the receiving and transmitting ends are separated.The two-dimensional angle domain optimization models are established for the receiving and transmitting ends.Secondly,in order to solve the off-grid error problem of the traditional orthogonal matching pursuit(OMP)algorithm,an off-grid error estimation algorithm based on compressed sensing is proposed.The Taylor formula expansion and the orthogonality of the signal and noise subspaces are used to construct an optimized formula for the off-grid error,and the angle information is obtained through loop iteration to achieve accurate channel estimation.Finally,the accuracy and advancement of the proposed algorithm are verified by numerical simulation.(2)We study the channel estimation problem of ultra massive MIMO non-stationary channel based on spherical wavefront.The non-stationary characteristics of the channel are described based on the birth and death process of the scattering cluster.Aiming at the problem of non-stationary channel estimation,the Newton-OMP algorithm is proposed,and the channel estimation is divided into two steps.In the first step,the OMP algorithm is used to roughly estimate the position of the scattering clusters.In the second step,the rough estimated value is used as the initial condition,and the optimal value is finally obtained through Newton iterative optimization.Finally,the improved algorithm is compared with other algorithms and numerical simulation is carried out through Matlab.The simulation results can prove the superiority and accuracy of the Newton-OMP algorithm in the ultra massive MIMO system.The millimeter-wave massive MIMO channel estimation algorithm proposed in this paper provides a theoretical basis and practical basis for subsequent work such as transmission scheme design,link-level simulation,and interference analysis in B5 G wireless communication systems.