IRS-aided NOMA in a Cell Free Massive MIMO System

Abstract

In this paper, intelligent reflecting surfaces (IRSs) are employed to improve spectral efficiency (SE) in downlink of a cell-free massive multiple input multiple output (CF-MaMIMO) network where, non-orthogonal multiple access (NOMA) is used to improve the SE and number of served users. High number of users, access points (APs) and IRSs in a CF-MaMIMO-NOMA system cause optimization complex for active and passive beamforming which is extremely time-consuming and infeasible in practice. Therefore, in this paper, we advocate conjugate beamforming for active and passive beamforming at the APs and IRSs, respectively. Each IRS performs passive beamforming to compensate the effect of channel to its assigned user. Consequently, two simple, practical and independent algorithms for IRS assignment to users and user clustering in NOMA are proposed. For performance analysis, complexity of the proposed algorithms are analyzed. Finally, numerical results corroborate the improved performance of the proposed system compared with traditional CF-MaMIMO-NOMA one.