Channel Estimation for RIS-Aided mmWave MIMO System from 1-Sparse Recovery Perspective

In this paper, we develop a two-phase based uplink channel estimation strategy with reduced pilot overhead for an reconfigurable intelligent surface (RIS)-aided millimeter wave (mmWave) multiple-input multiple-output (MIMO) communication system. Specifically, in Phase I, an OMP-based method is adopted to estimate the AoDs at the users. The remaining parameters including the common AoAs at the BS, the cascaded AoDs at the RIS, and the cascaded channel gains are estimated in Phase II. In particular, the estimation of cascaded AoDs and channel gains can be formulated as 1-sparse recovery problems by decomposing the estimation of a multi-antenna channel with J scatterers into estimating J single-scatterer channels for virtual single-antenna users. Finally, the theoretical number of pilots required for the proposed method are analyzed and the simulation results are presented to demonstrate the high channel estimation accuracy.