STAR-RIS-Aided MISO Networks with Multi-User Rate Fairness

This letter considers a simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RIS)-assisted multiple-input single-output (MISO) system, investigating a computationally efficient approach to achieve rate fairness among users. To achieve fair rate distribution, a joint optimization scheme for the transmitting and reflecting coefficients (TARCs) of STAR-RIS and the transmit beamforming at the base station is proposed. Generally, rate fairness is achieved by solving max-min rate optimization, which is a computationally challenging non-smooth optimization objective. To address this, we adopt a pair of alternative objectives aiming at maximizing the geometric mean (GM) of the users’ rates and the soft minimum rate function. To address the energy conservation constraint for the TARCs and their joint optimization with the transmit beamformers, we propose computationally efficient alternating optimization algorithms. Our proposed algorithms leverage closed-form expressions of scalable complexity for efficient implementation, ensuring practicality in large-scale scenarios. Our simulation results show that the rate-fairness performance of the computationally efficient proposed algorithms is not very far from that of the computationally prohibitive non-smooth max-min rate (MMR) optimization. Moreover, they achieve better sum-rate performance than the conventional MMR optimization.