Optimizing Wireless Sensor Networks: A Deep Reinforcement Learning-Assisted Butterfly Optimization Algorithm in MOD-LEACH Routing for Enhanced Energy Efficiency

Abstract

Wireless Sensor Networks (WSNs) play a crucial role in diverse applications, necessitating the development of energy-efficient routing protocols to extend network lifetime. This study proposes a novel Deep Reinforcement Learning-Assisted Butterfly Optimization Algorithm (DRL-BOA) integrated with the MOD-LEACH protocol to optimize routing in WSNs. The proposed hybrid approach leverages the exploration and exploitation capabilities of BOA and the adaptive decision-making power of DRL to dynamically select cluster heads and optimal routes based on network conditions. The DRL-BOA model was evaluated on various WSN scenarios with node densities ranging from 50 to 500, considering parameters such as energy consumption, packet delivery ratio (PDR), throughput, and network lifetime. Simulation results demonstrated that the proposed method achieved a 22% reduction in energy consumption compared to traditional MOD-LEACH, a 15% improvement in PDR, a 27% increase in throughput, and an 18% enhancement in network lifetime over the Hybrid PSO-GWO approach. These significant improvements highlight the effectiveness of the DRL-BOA model in overcoming the limitations of existing algorithms. The proposed framework demonstrates superior adaptability to dynamic network conditions, making it a promising solution for energy-efficient and reliable WSN operations. Future work will explore integrating this model with emerging technologies, such as edge computing and the Internet of Things (IoT), for further enhancements