ARIS-NOMA for Cognitive Radio: A Joint Optimization Perspective

This paper investigates active reconfigurable intelligent surface (ARIS) assisted cognitive radio (CR) non-orthogonal multiple access (NOMA) network for joint beamforming and spectrum sensing (SS). Considering SS performance and transmission efficiency, our goal is to maximize the sum achievable rate through jointly designing the SS time, the secondary transmitter (ST) beamforming, and the ARIS reflecting coefficients. We propose a two-layer optimization framework to solve the proposed non-convex problem. In the outer layer, line search is adopted to determine the sensing time. In the inner layer, the problem is decoupled into two subproblems, the transmit beamforming subproblem and the ARIS reflecting coefficients subproblem. For the transmit beamforming subproblem, we resolve variable coupling and rank one constraint using the arithmetic-geometric mean (AGM) inequality and semidefinite relaxation (SDR). For the ARIS reflecting coefficients subproblem, we convexify detection probability viaQ-function approximation and handle non-convex constraints with successive convex approximation (SCA) and penalty method. The simulation results demonstrate that the proposed scheme outperforms the benchmark schemes and ARIS can significantly improve the spectrum efficiency of the CR Network.