Robust Beamforming for RIS-Aided ISAC With Joint Hardware Impairments and Imperfect CSI

Abstract

To address imperfect channel state information (CSI) and reconfigurable intelligent surface (RIS) hardware impairments in multi-user multi-target RIS-aided integrated sensing and communication (ISAC) systems, this letter proposes a robust beamforming design that maximizes the worst-case sum rate subject to a radar beampattern gain constraint and per-user quality-of-service (QoS) constraints. Specifically, the considered RIS hardware impairments are modeled as residual phase errors, which are incorporated via a deterministic-equivalent approximation based on the sinc function, while CSI errors are handled via worst-case robust optimization. The resulting non-convex problem is reformulated into a tractable form and solved by the generalized S-procedure and an alternating optimization (AO) algorithm. Numerical results show that the proposed design is robust against CSI uncertainty and residual RIS phase errors while maintaining a favorable communication-sensing trade-off.