Design of STAR-RIS Assisted Uplink NOMA for Maximum Fairness

To support massive connectivity with full service coverage, this paper integrates a simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) into uplink non-orthogonal multiple access (NOMA). In particular, the energy splitting (ES) protocol of STAR-RIS is considered, leading to two types of NOMA: ES/NOMA, which simultaneously supports all users in the transmitting and reflecting spaces of the STAR-RIS and ES/hybrid NOMA (H-NOMA), which supports multiple ES/NOMA pairs in time division multiple access (TDMA). To maximize rate fairness, we develop STAR-RIS beamforming and resource allocation algorithms for ES/NOMA and ES/H-NOMA, along with a globally optimal solution and a closed-form solution for two-user ES/NOMA. The proposed ES/NOMA algorithm achieves faster computation than conventional alternating optimization. The proposed ES/H-NOMA algorithm significantly reduces the complexity of ES/NOMA by using a closed-form solution for two-user ES/NOMA and designing a faster user assignment algorithm with optimal time allocation. The simulation results validate the algorithms and analysis for ES/NOMA and ES/H-NOMA and demonstrate their superiority over the time-switching NOMA and TDMA protocols.