Energy-Recognition Clustering Technique Based on Reinforcement Learning In WSN

Abstract

WSN (Wireless Sensor Network) technology has recently gained a lot of attention. This began with the deployment of small WSNs and moved to the deployment of big and IoT WSNs, all with an emphasis on energy conservation. Wireless sensor networks can benefit from network clustering to increase their energy efficiency. The practise of dividing nodes into clusters before picking multiple cluster heads is known as network clustering (CHs). Clustering in wireless sensor networks is known to save energy and extend the network's lifetime (WSNs). Energy efficiency is a hot topic in existing wireless sensor networks, although it's not generally discussed. In this research, we offer a reinforcement learning (RL) based energy-aware clustering approach, whereby peripheral cluster nodes monitor environmental factors like energy use and choose an optimal cluster leader (CH). Connect the CH (BS) to the base station. In the simulation (PDR), performance factors such as network lifetime, energy tax, network stability period, and packet delivery rate are all taken into account. The simulation results show that the proposed QL-ReLeC performs around 11% better than the reference protocol in terms of PDR and 11% better in terms of energy tax.