Dynamic Power Management is a design methodology aiming at reducing power consumption of electronic systems, by performing selective shutdown of the idle system resources. The effectiveness of a power management scheme depends critically on an accurate modeling of the environment, and on the computation of the control policy. This paper presents two methods for characterizing nonstationary service requests by means of a prediction scheme based on sliding windows. Moreover, it describes how control policies for non-stationary models can be derived.
Chung et al. (Fri,) studied this question.