Enhancing Energy Efficiency in IoT-WSNs Through Optimized PSO Cluster Head Selection

Abstract

In IoT-based wireless sensor networks (WSNs), clustering is one of the most effective energy-saving approaches to optimizing the life cycle of networks. The primary challenge in the multi-hop technique arises when cluster heads (CHs) located near the base station quickly deplete their energy due to the high volume of inter-cluster relay traffic. Therefore, a clustering protocol needs to be both fault-tolerant and energy-efficient. This paper proposes a particle swarm optimization (PSO) based clustering algorithm that incorporates a double-exponential adaptive inertia weight update to dynamically balance global exploration and local exploitation. This adaptive mechanism mitigates premature convergence, improves the accuracy of CH selection, and enhances overall energy efficiency in the network. By classifying sensor nodes into cluster members, CHs, and free nodes, this intelligent swarm-based algorithm is used to maximize CH selection in IoT-WSNs. The residual energy of the nodes is utilized to dynamically determine the optimal number of clusters during cluster formation. Subsequently, the algorithm allocates nodes into clusters of uniform size, ensuring that high-energy nodes are efficiently assigned as CHs based on their location data. The proposed algorithm is evaluated across various network topologies and compared with four benchmark WSN routing algorithms: LEACH, LEACH-FL, LEACH-FC, and KM-PSO. The results indicate that the proposed algorithm outperforms LEACH by approximately 28%, LEACH-FL and LEACH-FC by around 20%, and KM-PSO by nearly 10% in terms of node survivability and network longevity. Besides that, the proposed algorithm enhances energy efficiency in IoT-based WSNs by selecting cluster heads with high neighborhood density and strategically using intermediate nodes, thereby reducing communication distances, balancing transmission load, and extending network lifetime. These findings highlight the effectiveness of the proposed algorithm in enhancing energy efficiency and prolonging the operational lifetime of IoT-enabled WSNs. © 2013 IEEE.