Improved Joint Transmit and Receive Port Selection for Capacity Maximization in Fluid-MIMO Systems

One unique capability of fluid antenna system (FAS) is its position reconfigurability for enhancing wireless communication systems. This reconfigurability allows FASs to fully exploit spatial diversity, leading to significant improvements in metrics such as outage probability and capacity. Nevertheless, optimizing port selection remains a significant challenge, particularly in fluid multiple-input multiple-output (MIMO) systems where multiple fluid antennas are deployed at both ends. Existing joint convex relaxation-based methods, despite their performance benefits, are prohibitively complex, and scales badly with the number of ports. To overcome this, we propose a novel probability learning-based scheme within the cross-entropy optimization (CEO) framework to tackle the joint transmit and receive port selection problem. Unlike existing approaches, our proposed CEO-based approach avoids approximations to the original problem, hence mitigating performance degradation. Simulation results show that the proposed algorithm not only surpasses state-of-the-art algorithms but also significantly reduces computational complexity, providing an efficient solution for modern FAS applications.