A Spiking Neural Network-based LEACH Protocol for Optimal Cluster Head Selection in Large-Scale Wireless Sensor Networks

Wireless Sensor Networks (WSNs) rely on efficient routing protocols to keep their energy consumption low.A major obstacle in the development of distributed clustering protocols is the ability to build clusters independently of centralized data collecting.To tackle these issues, a hybrid protocol called Low Energy Adaptive Clustering Hierarchy with the Analytic Hierarchy Process (LEACH-AHP) has been developed to optimize Cluster Head (CH) selection.However, the traditional AHP requires frequent recalculations of threshold values for CH selection due to energy fluctuations, leading to unnecessary overhead.In addition, using fixed thresholds based on AHP criteria may restrict the selection of appropriate CHs in large WSNs.Therefore, this study presents a novel machine-learning algorithm, namely Spiking Neural Network (SNN) to dynamically determine optimal threshold values for CH selection in large-scale WSNs.The SNN architecture takes into account inputs like levels of energy in the node, distances away from the sink, and network state factors to generate output neurons representing optimal threshold values.The SNN is trained with historical data to minimize the difference between its output and target threshold values, enabling it to select the best CH nodes efficiently.The proposed work, known as LEACH-SNN routing, reduces computational overhead while increasing packet throughput to the sink.Finally, extensive simulations demonstrate that the LEACH-SNN protocol achieves 0.21ms End-to-End (E2E) delay, 3.2 J energy consumption, 94.83% Packet Delivery Ratio (PDR), 257 kbps throughput, and 97.11% network lifetime for 200 nodes in a 1000500 m 2 area compared to the standard clustering protocols in WSNs, such as the LEACH, LEACH-Centralized routing protocol (LEACH-C), Hybrid Energy-Efficient Distributed (HEED) clustering protocol, Power-Efficient Gathering in Sensor Information Systems (PEGASIS), Threshold-sensitive Energy-Efficient Network (TEEN) protocol, and Stable Election Protocol (SEP).