Intelligent reflecting surface assisted untrusted NOMA transmissions: a secrecy perspective

In this paper, we investigate the intelligent reflecting surface (IRS) assisted secure transmissions for the untrusted non-orthogonal multiple access (NOMA) network threatened by an external eavesdropper, where the near user and two untrusted far users receive the privacy information from the transmitter assisted by the IRS’s passive reflection. To protect the privacy information against internal and external eavesdropping, we first provide a comprehensive analysis of the secrecy rates, which can offer suitable methods for separately handling the internal and external secrecy performance. Following the above results, we formulate a secrecy rate maximization problem by optimizing the transmission power allocation and phase shifts. To achieve this goal, we decouple the non-convex optimization problem into the external security problem and the internal security problem. Then, we propose a heuristic simulated annealing algorithm to iteratively solve the above problems, where the semi-definite relaxation (SDR) technique, the singular value decomposition (SVD) technique, and the Lagrange multiplier method are utilized to relax the problems and optimize the power allocation and phase shifts. Finally, numerical results show that the proposed scheme can improve the secrecy rate compared with the existing IRS-aided secure NOMA transmission schemes.