Channel Estimation for Reconfigurable Intelligent Surface-Aided mmWave Communications

Abstract

Reconfigurable intelligent surface (RIS) is a promising technology that can provide a virtual line-of-sight (LoS) link for the mmWave communications via intelligent signal reflection. In order to maximize the throughput of RIS-aided mmWave systems, acquisition of accurate downlink channel information at the base station (BS) is crucial. However, since the BS needs to acquire not only the conventional direct channel between the BS and user but also the channels reflected by RIS (i.e., BS to RIS and RIS to user channels), the pilot overhead is proportional to the number of RIS reflecting elements. In this paper, we propose an efficient channel estimation framework reducing the pilot overhead of RIS-aided mmWave systems. Key idea of proposed scheme is to decompose the RIS-aided channel into three major components, i.e., static BS- RIS angles, quasi-static RIS-user angles, and time-varying BS-RIS-user path gains, and then estimate them in different time scales. In doing so, the number of channel parameters to be estimated at each stage can be reduced significantly. From the simulation results, we demonstrate that the proposed scheme is very effective in reducing the pilot overhead of RIS-aided mmWave systems.