QEnergy and SARSAEnergy Learning for Energy efficient Routing in Wireless Sensor Networks

Wireless sensor network nodes will be located at varying distances in the network, and the data travel between each node requires the shortest path to transmit data. In data transmission, when a particular node is active for a long time, it will lead to energy depletion and the node will remain active based on residual energy. Reinforcement learning with the Markov model assigns rewards for the best path taken to transmit data between nodes. The nodes are located randomly with the Erdos-Renyi random graph model, which is a stochastic graph. The energy consumption depends on the amount of data transmitted and how long the particular node is active. The proposed QEnergy-Learning and SARSAEnergy algorithms can be used to calculate energy consumption in the wireless sensor network with rewards. The proposed work efficiency is evaluated by taking into account various WSN infrastructure and calculating energy consumption. The agent learns its environment and transmits data packets, and its performance is compared to that of various algorithms in different network infrastructures.