An Efficient Solution to Phase-Shift Optimization for RIS Enabled Joint Communication and Sensing

In reconfigurable intelligent surface (RIS) enabled joint communication and sensing (JCAS) systems, conventional phase-shift optimizations mostly suffer from high computational complexity in solving non-convex problems, which hinders their practical applications. To address this issue, we propose a novel and efficient method, namely, principal angle based penalty successive convex approximation (PAB-PSCA), to quickly yield locally optimal phase-shift solutions. To this end, we first determine the optimal direction of the global maximizer of the objective function by means of subspace analysis. Then, we project the global maximizer onto the feasible domain of the phase-shift vector by designing a penalty successive convex approximation procedure. Our results show that compared to the existing minorization-maximization (MM) and semidefinite relaxation (SDR) based method, the proposed PAB-PSCA is much more computationally efficient while achieving almost the same optimization performance.