Adaptive power management based on reinforcement learning for embedded system

Abstract

In this study, an adaptive power management method based on reinforcement learning is proposed to improve the energy utilization and battery endurance for resource-limited embedded systems. A simulator which traces battery endurance and device operations is developed to examine the proposed method. Experimental results show that, in terms of battery efficiency and endurance, the performance of our proposed method is better than the traditional power management techniques, such as static power management method. © 2008 Springer-Verlag Berlin Heidelberg.