Enhancing secrecy performance for NOMA systems with intelligent reflecting surface: Analysis and optimization

Abstract

This paper studies the effectiveness of intelligent reflecting surface (IRS) in enabling robust secure communications for indoor non-orthogonal multiple access (NOMA) systems. Using an IRS-control and power-allocation (PA) strategy (IRSCaPAS), the system purposely allocates its reflecting tiles (RTs) with reasonable phase shifts to the receivers, including two users and a potential eavesdropper, as well as adjusts the PA factor of NOMA. The goal of the paper is to boost the average secrecy rates (ASRs) for the users via optimizing the IRSCaPAS. The closed-form expressions for the ASR are derived for the generalized case where the successive interference cancellation (SIC) receivers for NOMA are imperfect. Using these expressions, heuristic algorithms (HA) are designed to optimize the IRSCaPAS. Numerical results show a vital role of the IRS in enhancing the system secrecy performance and the efficiency of the proposed algorithms. Moreover, these results provide useful insights into the effects of various key parameters on the secrecy performance, such as the size of the IRS, the transmit power, the NOMA PA factors, the scattering-signal power, the $K$ factors of the Rice channels, the SIC imperfection factor, and the locations of devices.