Hybrid GA–SA optimization for eMBB-oriented spectrum allocation in 5G device-to-device communication

Abstract

The explosive growth of enhanced mobile broadband (eMBB) services in fifth generation (5G) networks presents new challenges in maintaining quality of service (QoS), particularly under dense deployments with device-to-device (D2D) communication. Interference caused by spectrum reuse among D2D pairs and cellular users can significantly degrade signal-to-interference-plus-noise ratio or signal-to-interference-plus-noise ratio (SINR), throughput, and fairness. This paper addresses the underexplored problem of optimizing spectrum allocation in eMBB-specific D2D scenarios by proposing a hybrid metaheuristic framework combining genetic algorithm (GA) and simulated annealing (SA). The proposed hybrid GA–SA algorithm leverages GA’s global exploration and SA’s local exploitation to improve allocation quality while ensuring robustness. Simulation results reveal that the hybrid approach achieves up to 25% improvement in SINR, an 18% increase in aggregate throughput, and a 22% reduction in interference compared to standalone GA and SA algorithms. Furthermore, the framework achieves improved fairness performance while maintaining competitive SINR and throughput under dense eMBB-oriented deployment scenarios. The algorithm demonstrates efficient convergence behavior and scalability to larger user populations, making it suitable for real-time or large-scale deployments. These results affirm the significance of tailored hybrid optimization in interference-aware spectrum management for future 5G networks.