Energy-driven K-means-based LEACH routing protocol for enhanced lifetime in wireless sensor networks

Abstract Wireless Sensor Networks (WSNs) are widely deployed in monitoring and automation applications; however, their performance is constrained by the limited battery capacity of sensor nodes. Traditional K-means-based LEACH routing protocol improves cluster formation using distance-based clustering, but it still relies on Euclidean distance, which does not accurately represent the communication energy cost. The main innovation of this paper is the introduction of an Energy-Driven K-Means-Based LEACH routing protocol in which Euclidean distance is replaced with a novel energy-proxy metric derived from the Radio Energy Dissipation Model. This allows the clustering process to favor low-energy-cost links and penalize long-range transmissions, improving cluster-head placement without increasing computational complexity. The proposed protocol is evaluated against both the traditional K-means-Based LEACH and DEEC-KM; a heterogeneous energy-aware clustering scheme. Simulation results on 100 randomly distributed nodes across deployment areas ranging from 100 × 100 m 2 to 500 × 500 m 2 show that the proposed approach achieves up to 16.98% improvement in network lifetime compared to K-means-Based LEACH, and consistently outperforms DEEC-KM across all densities and metrics, including lifetime, stability periods, delivered packets, and overall energy consumption. These gains are most evident in medium- and low-density networks, while maintaining comparable performance to K-means-Based LEACH in very dense deployments. Overall, the proposed protocol contributes a lightweight yet energy-aware enhancement to centralized LEACH clustering, improving scalability and extending network lifetime under energy-constrained WSN conditions.