Achievable Sum Rate of RIS-Aided Massive MIMO NOMA

Abstract

A reconfigurable intelligent surface (RIS)-aided massive multiple-input multiple-output (MIMO) non-orthogonal multiple-access (NOMA) system is investigated. A statistical channel state information (CSI) based RIS phase-shift optimization technique is adopted by considering the effective correlation matrices of the NOMA clusters pertaining to the composite channels. A linear minimum mean square error (LMMSE) estimation technique is used to estimate the uplink composite channel at the massive MIMO base-station via the user pilots for a fixed RIS phase-shift matrix. Thereby, for the downlink transmission, a maximum ratio transmission precoder is designed based on the LMMSE estimates of the uplink composite channels by virtue of channel reciprocity of time-division duplexing mode. The achievable sum rate is derived in closed-form for the spatially correlated Rayleigh fading channels in the presence of erroneously estimated CSI, intra-cluster pilot contamination, and imperfect successive interference cancellation by exploiting the statistical CSI-based optimal RIS phase-shift matrix, while adhering to the worst-case Gaussian technique. Monte-Carlo simulations are used to validate our achievable sum rate analysis and to depict the performance gains/degradations of the proposed RIS-aided massive MIMO NOMA system.