| | |
Summary: Global Reliability-Aware Power Management for Multiprocessor
Real-Time Systems
Xuan Qi, Dakai Zhu Hakan Aydin
Department of Computer Science Department of Computer Science
University of Texas at San Antonio George Mason University
San Antonio, TX, 78249 Fairfax, VA 22030
{xqi,dzhu}@cs.utsa.edu aydin@cs.gmu.edu
Abstract
Recently, the negative effect of the popular power
management technique Dynamic Voltage and Frequency
Scaling (DVFS) on the system reliability has been iden-
tified. As a result, various reliability-aware power
management (RAPM) schemes have been studied for
uniprocessor real-time systems. In this paper, we in-
vestigate global-scheduling-based RAPM (G-RAPM)
schemes for a set of frame-based real-time tasks run-
ning on a homogeneous multiprocessor system. An im-
portant dimension of the problem is how to select the
appropriate subset of tasks for energy and reliability
management (i.e., schedule a recovery for each selected
|
