Robust Beamforming Design for STAR-RIS-aided NOMA System under Short-packet Communication

Next-generation wireless applications are expected to enable extended ultra-reliable low latency communication (URLLC) to support high data rates along with ultra-high reliability and low end-to-end latency features beyond the capabilities of existing core services. This paper investigates simultaneous transmission and reflection re-configurable intelligent surface (STARRIS) aided non-orthogonal multiple access (NOMA) systems to enable spectral-efficient short-packet communication under imperfect channel state information. In particular, we focus on the problem of spectral-efficiency maximization via joint beamforming design at the base station and STAR-RIS subject to given URLLC requirements. Owing to the non-convexity of the formulated problem, we propose an alternating optimization framework that obtains sub-optimal solutions to the robust beamforming design, respectively by exploiting
$S$
-procedure and successive convex approximation. Simulation results confirm that the STAR-RIS-NOMA system can significantly boost the spectral efficiency by 10-15% compared to conventional reflecting-only RIS while guaranteeing the strict URLLC requirements.