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 nonstationary models can be derived.
|Number of pages||5|
|Journal||Proceedings -Design, Automation and Test in Europe, DATE|
|Publication status||Published - 1999|
|Event||Design, Automation and Test in Europe Conference and Exhibition 1999, DATE 1999 - Munich, Germany|
Duration: 1999 Mar 9 → 1999 Mar 12
All Science Journal Classification (ASJC) codes