Max-Min Fair Beamforming for IRS-Aided Secure NOMA Systems

This letter investigates the secure beamforming (BF) of a two-user uplink non-orthogonal multiple access system aided by an intelligent reflecting surface (IRS). In this system, both users send messages to the base station while emitting artificial noise that confuses the eavesdropper. To improve the security with ensuring the max-min fairness, the joint BF optimization of the users and the IRS is formulated as a quadratic-fractional problem (QFP). By reformulating the QFP into a fraction-eliminated problem, we further obtain the active and passive BF sub-problems that can be iteratively solved by a proposed multi-beam cyclic optimization algorithm. Simulation results demonstrate the superior performance of the proposed algorithm.