Sum-Rate Maximization for RSMA Systems With Pattern-Reconfigurable Fluid Antennas
TL;DR
Simulation results show that the proposed tri-hybrid RSMA framework with PRFA improves sum-rate performance compared with conventional hybrid precoding, where the CO-PRFA provides an idealized upper benchmark compared with practical DS-PRFA due to the more flexible reconfigurability.
Abstract
This paper investigates the sum-rate maximization problem for downlink rate-splitting multiple access (RSMA) systems equipped with pattern-reconfigurable fluid antennas (PRFA). Two PRFA models are developed: a deployable discrete-selection PRFA (DS-PRFA) model with finite predefined radiation modes, and an idealized continuous-optimization PRFA (CO-PRFA) model based on spherical harmonic expansion that serves as a performance upper bound. The sum-rate maximization problem is formulated by jointly optimizing digital, analog, and antenna-domain precoders along with RSMA power allocation. To solve this non-convex problem, we propose alternating optimization algorithms based on the weighted minimum mean square error (WMMSE) transformation and block coordinate descent with per-antenna decoupling. For DS-PRFA optimization, closed-form solutions are derived, while for CO-PRFA optimization, a preconditioned Riemannian conjugate gradient method is developed on the spherical manifold. Simulation results under the considered settings show that the proposed tri-hybrid RSMA framework with PRFA improves sum-rate performance compared with conventional hybrid precoding, where the CO-PRFA provides an idealized upper benchmark compared with practical DS-PRFA due to the more flexible reconfigurability.