Design of Intelligent Reflecting Surface Phase Shifts for Uplink SCMA

In this letter, we consider an intelligent reflecting surface (IRS) assisted uplink sparse code multiple access (SCMA) system. We first derive an upper bound on word error rate (WER), based on which we design IRS phase shifts to minimize WER. Specifically, we propose two IRS design algorithms, semidefinite relaxation (SDR)-based approach and gradient ascent approach. We also provide computational complexity analysis. Then we derive an upper bound on average WER and analyse error performance. Finally, simulations demonstrate that both algorithms have performance similar to the known approach, while the proposed gradient ascent approach has the least computational complexity.