A hybrid GA-SA resource allocation scheme enhanced with SINR optimization for NOMA-MIMO systems in 5G networks

Efficient resource allocation in Non-Orthogonal Multiple Access (NOMA) systems is a critical challenge, especially in optimizing throughput and ensuring fairness in multi-user environments. This study introduces a hybrid Genetic Algorithm-Simulated Annealing (GA-SA) approach to enhance user pairing and power allocation in a NOMA-MIMO downlink system. Genetic Algorithm (GA) optimizes user pairing by evolving high-throughput configurations, while Simulated Annealing (SA) refines power allocation to maximize Signal-to-Interference-plus-Noise Ratio (SINR) and minimize interference. Unlike traditional methods, which are limited by computational inefficiency or local convergence, the hybrid algorithm integrates dynamic SINR updates to adaptively optimize resources in real-time. Simulation results demonstrate that the hybrid GA-SA algorithm achieves up to 15% higher throughput compared to standard GA and SA methods, with a 20% improvement in average SINR and reduced SINR variance, ensuring fairness across users. While the hybrid approach incurs a 5-10% increase in power consumption compared to heuristic methods, it delivers a 25% enhancement in throughput efficiency, affirming its practicality in high-performance applications. The results align with the hypothesis that combining GA and SA overcomes the limitations of standalone approaches, making it a promising solution for next-generation wireless networks like 5 G.