An Energy Efficient Routing Protocol based on Reinforcement Learning for WSN

Abstract

Wireless Sensor Network (WSN) seems to be critical because they are responsible for maintaining network routes, packet forwarding, and higher multi-hop connectivity. Clustering nodes is still a powerful technique for modelling routing protocol for WSNs, as it increases the range of communication services with energy efficiency. This paper focuses on the energy efficiency and improved lifetime of the network based on the reinforcement learning protocols. The system can adapt to network changes, such as energy efficiency, mobility and make better routing decisions attributable to Reinforcement Learning (RL). The legal restrictions on sensor nodes are taken into consideration and an energy-balancing routing model based upon reinforcement learning has been provided. The results show that the enhanced protocol outperforms the state of energy savings and network lifetime when compared to Q-learning and LARCMS energy-efficient routing protocols. The proposed protocol's effectiveness is analysed by end to end delivery and packet delivery.