Novel Two-Phase Channel Estimation for RIS-Aided MIMO mmWave Systems in Angle Domain

In most existing works focusing on channel estimation for reconfigurable intelligent surface (RIS)-Aided MU-MIMO mmWave systems, Simultaneous Orthogonal Matching Pursuit (SOMP) method is used to estimate the angle of departure (AoD) at the users to convert the overall MIMO cascaded channel into multiple MISO cascaded channels for further estimation. However, it has poor performance when the number of antennas is small due to the strong correlation of atoms in the dictionary. This paper proposes a novel uplink two-phase channel estimation scheme with high accuracy in angle domain. Specifically, in the first phase, by carefully designing the precoding matrix, the 1-th sub-cascaded channel related to the 1 -th antenna is separated and estimated. In the second phase, by utilizing the invariance of angles and the linear correlation of gains, AoDs can be estimated based on the one-dimensional search method with high accuracy. Thus, all the remaining sub-cascaded channels can be calculated and combined into the overall MIMO cascaded channel. Simulation results demonstrate the superiority of the proposed algorithm.