Robust Resource Allocation in RSMA-Based Star-RIS-Aided HAP Communication Networks with Imperfect SIC

The next generation of wireless communication networks must offer robust connectivity to serve users in remote or disaster-stricken regions where terrestrial infrastructure is unavailable or compromised. High-Altitude platforms (HAPs), functioning as non-Terrestrial network (NTN) nodes, can rapidly restore coverage and extend service reach by transmitting directly to ground users. To further enhance communication performance, simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RIS) can be deployed alongside HAPs, intelligently shaping the wireless channel to improve channel reliability. In this work, we investigate a HAP-Assisted NTN in which a STAR-RIS aids downlink transmission to multiple ground users under a rate-splitting multiple access (RSMA) protocol with imperfect successive interference cancellation (SIC). The joint design of power allocation at the HAP and STAR-RIS beamforming presents a challenging non-convex problem because of the coupled rate expressions and rank-one constraints on the STAR-RIS matrices. To address this, we introduce auxiliary variables and apply successive convex approximation (SCA) to convexify rate functions, while employing a difference-of-convex (DC) programming approach to handle the rank-one requirement. An alternating optimization framework is then developed to iteratively solve a convex power allocation subproblem and a penalized semi-definite program for STAR-RIS beamforming. Simulation results show the efficacy of the proposed framework, showing excellent performance even under imperfect SIC and with discretized phase shifts at the STAR-RIS.