Improper Gaussian signaling for low-resolution active RIS-assisted rate-splitting multiple access multiuser MIMO system

Abstract

This paper investigates a low-resolution active reconfigurable intelligent surface (aRIS)-assisted rate-splitting multiple access (RSMA) multiple-input multiple-output (MIMO) downlink system with improper Gaussian signaling (IGS). A max–min achievable log determinant (log-det) rate optimization problem is formulated by jointly optimizing the transmit beamforming, low-resolution aRIS’s power-amplified reconfigurable elements (aPREs), and rate-splitting (RS) vector. To address the nonsmooth optimization objective function, the formulated problem is decoupled into two subproblems, namely, transmit beamforming and RS vector optimization, and aPREs’ vector and RS vector optimization. For the large-scale mixed discrete-continuous problem imposed by the aPREs’ vector optimization, we introduce a novel penalized optimization framework that employs a cubic complexity quadratic solver for alternately optimizing the RS vector, transmit beamforming, and the aPREs’ vector. Simulation results show that the consideblack RSMA system outperforms the space-division multiple access (SDMA) scheme with respect to rate fairness, and the use of IGS is remarkably superior to the traditional proper Gaussian signaling (PGS) in improving the rate fairness among users.