ADMM Based Channel Estimation for RISs Aided Millimeter Wave Communications

In reconfigurable intelligent surfaces (RISs) aided millimeter wave communication systems, accurate channel state information (CSI) is particularly crucial to achieve high passive beamforming gain. However, the complexity of channel estimation increases significantly due to the high-dimensional channel matrix and the additional RISs’ assisted channels. To tackle this challenge, an alternating direction method of multipliers (ADMM) based channel estimation method is proposed to improve the CSI estimation accuracy and reduce the training overhead. By converting the cascaded channel matrix into a sparse matrix recovery problem, both low rank and sparsity properties are exploited by the ADMM algorithm to solve the sparse matrix recovery problem. The simulation results verify that the proposed ADMM algorithm can significantly increase the accuracy of channel estimation and reduce the training overhead, compared with the well-known competing channel estimation algorithms.

I. Introduction

Recently, reconfigurable intelligent surface (RIS) is regarded as a key technology that can provide high spectral and energy efficiency for future wireless communication systems [1]–[3]. Specifically, RIS made of a planar array with low-cost passive reflective elements can smartly steer beams towards the desired directions by reconfiguring the reflective coefficients in real-time. Therefore, RIS can be exploited as an energy-efficient solution to add the virtual LoS path between transceivers for combating the blockage in mmWave communication.