SARSA based method for WSN transmission power management

Alexander Alexandrov, Vladimir Monov
The scope of this research is to propose an adaptive machine learning approach which can help the WSN’s nodes to manage their transmission power and to improve the internode wireless communications. The optimized transmission power has benefits in terms of WSN energy consumption and RF interlink interference. The paper proposes an adaptive method of a wireless sensor node based on Multi-Layer Perceptron (MLP) network representation and machine learning. The presented in the paper approach uses the SARSA (State-Action-Reward-State-Action) algorithm which is a form of reinforcement machine learning. The aim of the new method is to improve the sensor nodes Transmission Power Management (TPM) process. This inspires many practical solutions that maximize resource utilization and prolong the shelf life of the battery-powered wireless sensor networks.