An intelligent reinforcement learning enhanced improved LEACH protocol for prolonging wireless sensor network lifetime

This study suggests an intelligent clustering protocol in Wireless Sensor Networks, called RL-ILEACH, which incorporates Reinforcement Learning (RL) into the ILEACH (Improved Low-Energy Adaptive Clustering Hierarchy) framework for adaptive and energy-aware CH selection in order to overcome these drawbacks. An RL agent is used in the suggested RL-ILEACH protocol to learn the best CH selection strategies depending on current network conditions, such as residual energy, node density, and communication distance. RL-ILEACH reduces energy dissipation and improves load balancing during intra-cluster and inter-cluster communication phases by dynamically adjusting CH selection options. To assess RL-ILEACH’s performance against traditional procedures, such as LEACH, ILEACH, and other cutting-edge clustering techniques, extensive simulation tests are carried out. According to simulation studies, RL-ILEACH performs noticeably better than current protocols, resulting in reduced total energy consumption, a longer stability period, a greater packet delivery ratio, and a longer network lifetime. In particular, RL-ILEACH minimizes early node failures brought on by uneven energy depletion and sustains a larger number of active nodes throughout subsequent rounds. RL-ILEACH is a reliable and scalable solution for energy-efficient clustering in dynamic Wireless Sensor Network settings, as these enhancements demonstrate that the incorporation of Reinforcement Learning permits intelligent, adaptive decision-making. RL-ILEACH prolonged the network lifetime by $$\:784\:\:$$ % compare to LEACH. RL-ILEACH outperformed ILEACH by increasing the network lifetime by $$\:130$$ %. Moreover, relative to NNMH-LEACH, RL-ILEACH achieved a further $$\:108$$ % imporvment in network lifetime.