Robust Covert Communication in MISO-RSMA Systems with Internal Eavesdropper

Abstract

This paper investigates robust covert communication in a Multi-Input Single-Output (MISO) Rate-Splitting Multiple Access (RSMA) system facing both an internal eavesdropper and an external warden. To maximize the covert communication rate, we formulate a joint optimization problem that leverages the RSMA common stream for simultaneous transmission and jamming. To solve this highly coupled and nonconvex problem, we propose an iterative Robust Beamforming and Rate Allocation (RBRA) algorithm. This algorithm utilizes the S-Procedure, Semidefinite Relaxation (SDR), and Fractional Programming (FP) to effectively handle variable coupling and the warden’s Channel State Information (CSI) uncertainty. Simulation results demonstrate that the proposed scheme outperforms the Non-Orthogonal Multiple Access (NOMA) benchmark in both covertness and robustness. Furthermore, compared to traditional non-secure schemes, it drastically reduces the maximum internal eavesdropping rate by approximately 60% with only a 10% degradation in covert performance, effectively balancing external covertness and internal security.