Temporally Correlated and Block-Sparse Channel Estimation for Millimeter-Wave Massive MIMO Systems

In this paper, we propose a multilayer channel prior model to better characterize the time-varying features and clustering properties of the channel based on the sparsity of millimeter-wave (mmWave) massive multiple-input multipleoutput (MIMO) channels. Specifically, we utilize a first-order Auto-Regressive (AR) process to simulate the correlation of channel gains over multiple measurements, as well as the angular spread within scattering clusters. Based on the Expectation Maximization (EM) algorithm and the sparse Bayesian learning (SBL) framework, we develop an estimation scheme that considers both the channel temporal correlation and the intra-block correlation. The simulation results show that the proposed scheme exhibits performance advantages with adequate consideration of the channel characteristics.