An enhanced intelligent framework for 5G V2X communication using multi-objective optimization and mobility-aware transformer networks.
The 5G- enabled Vehicle -to-Everything (V2X) is a reliable communication technology that allows vehicles to communicate with other vehicles, networks, and infrastructure to enhance road safety and traffic proficiency. The challenges in 5G-V2X networks are high mobility, dynamic network topology, and strict quality-of-service (QoS) requirements. Especially in latency-sensitive applications such as collision avoidance and real-time traffic management, are degraded by frequent link failures, excessive routing overhead, and inefficient resource utilization. Recently, the machine learning-based routing algorithms integrated with fuzzy logic and metaheuristic optimization have achieved multi-objective performance with limited adaptability and slow convergence. To overcome these issues, a novel Multi-Objective Harris Hawks Optimization (MO-HHO) integrated with a Bayesian optimized Mobility-Aware Transformer Network (BMAT) is proposed to design an enhanced intelligent framework for 5G V2X Communication.MO-HHO optimizes cluster formation and routing paths with reduced latency and energy usage, and it re-clusters by improving throughput and stability using mobility-aware updates. The Bayesian optimized self-attention-based MAT transformer is used for typical long-range spatiotemporal dependencies for finding optimal cluster heads, routing stability, and probability of congestion with the Tree-structured Parzen Estimator (TPE) to produce better convergence. The proposed framework is assessed using realistic 5G V2X mobility scenarios under urban and highway conditions, and the obtained results achieve 20-30% reduced latency, a 15-35% reduction in energy consumption, and an 18-28% throughput enhancement compared to existing methods. Finally, the proposed framework ensures a wide improvement in reliability and adaptability in 5G V2X communication networks.