Channel Estimation for RIS Assisted Millimeter Wave Systems via OMP with Optimization

Reconfigurable intelligence surface (RIS) can be deployed to assist the communications in millimeter wave (mmWave) systems. Employing the sparsity of the mmWave channel, the compressive sensing (CS) techniques can be leveraged to reduce the pilot overhead of the channel estimation (CE). However, conventional CS-based algorithms are based on the discrete grids and the difference between the real continuous angle and its nearest grid point is called off-grid error. Off-grid errors seriously deteriorate the CE performance. In this paper, we propose the orthogonal matching pursuit with discrete-continuous optimization (DC-OMP) method for RIS assisted mmWave systems to mitigate the impact of the off-grid errors. Simulation results show that compared with existing works, the proposed DC-OMP can efficiently mitigate the impact of the off-grid errors without adding much complexity.