On the Performance of STAR-RIS-Aided mmWave MIMO–NOMA Transmission Using Stochastic Geometry: Phase Shift Error Case

This work combines non-orthogonal multiple-access (NOMA) and millimeter-wave (mmWave) in a reconfigurable intelligent surface (RIS)-aided multiple-input multiple-output (MIMO) communication system, in which the RIS can simultaneously transmit and reflect signals (STAR-RIS). To adjust the STAR-RIS phase shift responses, we consider scenarios where the cascade channel phase information is available either perfectly or imperfectly. The stochastic geometric model is utilized to model the locations of the randomly deployed users. The users are divided into the cell-center users’ group and the cell-edge users’ group. To implement NOMA, we consider two user selection frameworks: 1) random user selection and 2) nearest user selection. For random users selection, one user from each group is randomly selected to be paired, while one user from each group with the shortest distance relative to STAR-RIS is selected in the nearest user selection strategy. To reduce system overhead and latency caused by the requirement to obtain channel state information (CSI) of all users, a beamforming approach is employed in the base station (BS). We derive the effective channel powers and provide the analytical expressions of the outage probability and outage sum rate for scenarios with and without error in the phase shift response of STAR-RIS. Besides, we provide an asymptotic analysis and derive the lower bound for outage probability when there are phase errors in the STAR-RIS phase shifts. We further analyze the impact of active STAR-RIS and imperfect CSI on the system performance. Simulation results are provided to validate our analyses and illustrate the effectiveness of establishing unobstructed transmission and reflection links in dead zones, thereby enhancing the reliability of communications.