Machine Learning-driven Energy-efficient Routing in Wireless Sensor Networks: Predicting Node Lifetime for Optimized Performance

This study introduces a hybrid machine learning framework for Wireless Sensor Networks (WSNs) designed to enhance energy efficiency and extend network longevity. The model integrates Q-learning for adaptive routing, hybrid clustering through Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), and decision tree regression for predictive energy depletion analysis. By dynamically balancing energy consumption and rerouting data to circumvent nodes approaching exhaustion, the framework improves reliability and operational stability. Simulation results demonstrate notable improvements over conventional protocols such as LEACH and PEGASIS, achieving a 40% reduction in energy consumption and a 37.76% extension of network lifespan. Statistical validation (t-test, p < 0.0001) confirms the significance of these results. The proposed approach holds promise for deployment in real-world WSN and IoT applications, where optimized energy utilization and extended network lifetime can reduce maintenance costs and ensure continuous, reliable data acquisition.