A Fuzzy Q-learning Based Power Management for Energy Harvest Wireless Sensor Node

In this study, a fuzzy Q-learning is used for the power management of energy harvest wireless sensor node(EHWSN). The proposed method establishes a fuzzy inference system (FIS) for the power management of energy harvest wireless sensor node and employs Q-learning in evaluating and updating the rule base. By interacting with the environment of the energy harvest wireless sensor node, the learning agent of the fuzzy Q-Learning is able to obtain the strategy in updating the fuzzy rule base and exercising the decided sensing duty cycle of the wireless sensor node such that energy neutrality of the EHWSN can be achieved. Experiment results shows that the proposed fuzzy Q-learning based power management can obtain better battery charge status in comparing other existing methods and achieve sustainable operation for EHWSN.