Q-learning-based Optimization of Smart Home's Wireless Sensors Network Lifetime
ISMAEL JRHILIFA, Hamid Ouadi, Abdelilah Jilbab
Abstract
— Wireless sensor networks (WSN) have witnessed increased utilization in recent years, particularly with the internet of things (IoT) trend in numerous sectors such as health, agriculture, marine, and smart buildings. The main challenge of these networks is energy efficiency; typically, sensor nodes are powered by tiny batteries with limited capacity. In this research, we suggest a Q-learning-based routing algorithm for optimizing WSN lifespan and sending huge amounts of data in smart home applications. The suggested routing approach takes advantage of the benefits of Q-learning to discover the ideal routing path to transmit data with the least amount of energy dissipation. To mimic the WSN in a smart home environment, the simulation is implemented in 3D. The presented routing method is evaluated in comparison to three different protocols: QLRP, EQL, and Dijkstra. The findings show that the created routing approach surpasses the other protocols in terms of prolonging WSN lifetime, total transferred data, and network energetic cost.