Sum-Rate Maximization of IRS-Aided SCMA System

We study an intelligent reflecting surface (IRS)-aided downlink sparse code multiple access (SCMA) system for massive connectivity in future machine -type communication networks. Our objective is to maximize the system sum-rate subject to the constraint of minimum user data rate, the total power of base station, SCMA codebook structure, and IRS channel coefficients. To this end, a joint optimization problem involving IRS phase vector, factor graph matrix assignment, and power allocation problem is formulated, which is non-convex in nature. This problem is solved by developing an alternating optimization (AO) algorithm. A key idea is to first divide the formulated non-convex problem into three subproblems (i.e., factor graph matrix assignment, power allocation, and phase vector of IRS) and then tackle them iteratively. The validity of the proposed schemes is shown using the simulation results. Moreover, compared to the SCMA system without IRS, a significant performance improvement in the IRS-aided SCMA system is shown in terms of achievable sum-rate.