Channel Estimation Optimization for RIS-Aided Multiuser MISO Structures

Abstract

Achieving the promised gain of intelligent reflecting surface (IRS) relies on acquiring channel state information (CSI) in communication systems, a primary constraint in practice. Specifically, a significant pilot overhead is needed to estimate the total of KMN + KM channel coefficients, where K denotes the number of users, N the number of IRS reflecting elements, and M the number of antennas at the base station (BS). Therefore, the increasing number of RIS elements in Terahertz (Thz) communication systems facing significant signal degradation makes the implementation cost very high and the time duration too long. Herein, a new approach is proposed in this paper founded on a three-step pilot-driven channel state estimation model for Uplink multiuser communication with IRS support where the direct links between users and the base station are estimated under the “ON” IRS element status in level I and The second phase involves estimating the combined IRS-related channels of a selected reference user by only emitting pilots in a reduced-dimensional subspace identified through basis vectors, consequently, the pilot overhead will be minimized. The remaining user-IRS-BS reflected channels are predicted in level III by leveraging their correlation with the channels of the typical user.