An Energy-Aware Cluster-Head Selection for Improving Network Lifetime in Wireless Sensor Networks Using Machine Learning Techniques

Energy efficiency and reliable communication continue to be critical challenges in Wireless Sensor Networks (WSNs), primarily due to the limited energy resources of battery-operated sensor nodes. To address this, the present study introduces a novel hybrid machine learning–driven clustering and routing framework that combines swarm intelligence with reinforcement learning for adaptive, energy-aware network management. Unlike traditional clustering schemes such as Low-Energy Adaptive Clustering Hierarchy (LEACH) and Hybrid Energy-Efficient Distributed (HEED), or single-heuristic approaches like Particle Swarm Optimization–based Cluster-Head selection (PSO-CH) and Artificial Bee Colony–based Cluster-Head selection (ABC-CH), the proposed design integrates PSO and ABC algorithms with a Firefly-based exploitation mechanism for improved CH selection. Additionally, a Q-learning-assisted adaptive decision layer enables each node to autonomously refine CH participation and routing choices, whereas a fuzzy logic and Genetic Algorithm (GA)–tuned reinforcement learning model dynamically optimizes inter-cluster routing paths. Extensive MATLAB simulations demonstrate that the proposed hybrid model achieves up to 25–30% improvement in network lifetime, with delays in First Node Death (FND), Half Node Death (HND), and Last Node Death (LND) compared to benchmark protocols. The framework also achieves a 20% reduction in energy consumption per delivered packet, a 15% higher Jain's fairness index, and near-unity Packet Delivery Ratio (PDR) across all test configurations. These gains confirm more balanced energy utilization, stable connectivity, and efficient data aggregation even in large-scale topologies. By unifying metaheuristic exploration, learning-based adaptation, and fuzzy optimization, the proposed approach closes the existing gap between static clustering and adaptive routing, offering a scalable and intelligent solution for real-world WSN and Internet of Things (IoT) applications.