STAR-RIS-Assisted MIMO-MAC Over Spatially Correlated Rician: Sum-Rate Analysis and Optimization

Abstract

Simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RIS) can intelligently control the electromagnetic waves in the whole space through its passive components, thereby enhancing the wireless propagation environment, which is considered to be a promising technology. This paper studies the achievable ergodic sum-rate maximization problem for the multi-users uplink multiple-input multiple-output (MIMO) multiple-access channels (MAC) assisted by STAR-RIS using full statistical channel state information (CSI). In addition to direct links, multiple users also communicate with the base station through cascaded links established by STAR-RIS, taking into account spatially correlated Rician channels. By leveraging large dimension random matrix theory, we derive an approximate expression of the achievable ergodic sum-rate in the large system regime. This closed-form expression facilitates the implementation of statistical CSI-enabled optimal design through an alternating optimization algorithm. Specifically, using the water-filling-like algorithm, the transmit covariance matrices for multiple users are designed, while the projection gradient ascent method is used to design the phase shift matrices of STAR-RIS. Simulation results demonstrate that even with a small number of antennas, the large system approximation results are closely aligned with Monte-Carlo average results. Moreover, the proposed optimization algorithm significantly enhances the achievable ergodic sum-rate.