Optimizing Network Reliability and Fault Detection in WSN-Assisted Autonomous Vehicle Systems Using QHHO-GNN Model

Wireless Sensor Networks (WSNs) enable real-time environmental monitoring and unified, seamless communication within Autonomous Vehicle Systems (AVS).However, these WSN-assisted AVS often face reliability issues due to frequent sensor node failures, dynamic topology changes, and delayed fault detection, compromising vehicular safety and decisionmaking accuracy.To address these critical challenges, this research focuses on a novel hybrid framework called QHHO-GNN, which combines Quantum Enhanced Harris Hawks Optimization and Graph Neural Networks for efficient network optimization and intelligent automatic fault detection.Using exploration methods inspired by quantum mechanics, the QHHO optimizes the network characteristics, including energy consumption, routing paths, and connection stability, ensuring reliable communication even in highly mobile AV environments.Simultaneously, the GNN model effectively captures complex spatial and relational dependencies within WSN topologies, enabling accurate and real-time fault detection with minimal false alarms.Extensive network simulation is done using NS-3, and PyTorch is used for GNN fault detection with the data exchange of CSV/JSON files.The simulation results demonstrate that the proposed QHHO-GNN framework outperforms existing AVS routing and energy-efficient methods in terms of Packet Delivery Ratio, End-to-End Delay, Network Life Time, Energy Efficiency, Fault Detection Accuracy, and False Alarm Rate.With its adaptive optimization and AI-enabled predictive capabilities, this model presents a transformative solution for enhancing the resilience and operational safety of WSN-assisted AVS.