A Novel Sequential Reflection Optimisation Algorithm for RIS-Aided NOMA System

A reconfigurable intelligent surface (RIS) comprises a large number of low-cost reflecting elements that can collectively steer the incident signal through passive beamforming, dynamically adjusting the wireless environment to enhance system performance. In non-orthogonal multiple access (NOMA) systems, for cell edge users with severely obstructed direct links, users must transmit at maximum power, causing successive interference cancelation (SIC) to fail and generating many error bits. In this paper, we develop an uplink RIS-aided NOMA communication system, simultaneously considering the continuous coefficient constraint of the RIS and the weighted vector of antennas at the access point (AP). Since joint optimization of the RIS reflection magnitude, phase, and the weighted vector is challenging and has high complexity, minimizing the system's bit error rate (BER) is difficult. To handle this, we propose two sequential algorithms with low complexity to reduce the system's BER. The simulation results show that both algorithms can effectively reduce the system's BER.