Improving Quality of Service Using a Innovative Energy Effective Reinforcement Based Path Finding in Mobile Ad Hoc Network

RL stands for reinforcement learning, is a component of the broader field of machine learning. It provides a framework its features help to select an efficient action using previously learned information of interaction and its environment. Mobile nodes are grouped to form a network for communicating themselves termed autonomous network with mobile nodes (MANET). This is a dynamic and independent network requiring no administrative control. The mobile nodes in the network are to be part or apart at any time that causes the topology changes. The frequent topology changes make it hard to determine the destination route. The route discovery in the dynamic topology that consumes more node energy and, minimize the network lifetime. The energy usage and network lifetime are a vital part of the MANET. This paper proposes a new global routing protocol using a reinforcement learning algorithm called modified energy and signal strength Q-learning ad hoc on-demand distance vector algorithm (ESQAODV). This process conducts a result comparison to a local routing algorithm known as AODVSARSA. The proposed algorithm proved a better result with reference to average distribution ratio compared to the other, and complete delay metrics