Multiple-input multiple-output non-orthogonal multiple access (MIMO-NOMA) systems assisted by reconfigurable intelligent surfaces (RIS) have become a viable way to improve next-generation wireless networks. However, MIMO beamforming, antenna selection (AS), RIS phase configuration, and NOMA power allocation are governed by intricate and closely related relationships. Consequently, attaining optimal performance within these integrated systems remains very difficult. Especially in large-scale MIMO deployments, the resulting joint optimization problem is computationally demanding, multidimensional, and non-convex. Conventional convex optimization techniques and heuristic algorithms frequently struggle with slow convergence, local optima, and high computational burden. However, the practical applicability of these methods is significantly limited. This paper presents AQSA algorithm that simultaneously optimizes MIMO beamforming, NOMA power allocation, RIS phase shifts, and AS. It is performed in order to overcome these difficulties. An adaptive temperature control mechanism dynamically balances exploration and exploitation during the optimization process. In addition, the proposed AQSA framework integrates quantum computing concepts like superposition and probabilistic state transitions to improve search capability. Better convergence speed, resilience, and adaptability to various network configurations are guaranteed by this hybrid design. The performance of the suggested method is assessed using extensive numerical simulations in a range of system conditions. The findings show that, when compared to traditional approaches, the AQSA-based joint optimization considerably increases system sum-rate, energy efficiency (EE), and user fairness, while keeping the computational complexity manageable. Additionally, the framework can efficiently utilize the extra degrees of freedom offered by larger antenna arrays and RIS deployments. As well, it is scalable to large-scale MIMO–NOMA systems. These results demonstrate how the proposed work achieves noticeable enhancement in spectral effeciency (SE).
Farghaly et al. (Tue,) studied this question.