【Abstract】In order to resolve the problem of saving power in soft real-time embedded systems, this paper proposes a DVS algorithm which overcomes the uncertainty due to variable task execution times and pursues the optimal scheduling scheme during runtime. The experiment results show that this algorithm guarantees the CPU utilization and the stability of the soft real-time systems. Even in the situation of overload, the scheduling system can adjust itself automatically, and more than 99% tasks can be completed before their deadlines. The random tests to multiple task sets also demonstrate that more than 15% energy can be saved in average by this DVS algorithm. 【Key words】Real-time operating system; Real-time scheduling; Dynamic voltage scaling; Task execution time; Time prediction
[1]
Anantha P. Chandrakasan,et al.
Low-power CMOS digital design
,
1992
.
[2]
F. Frances Yao,et al.
A scheduling model for reduced CPU energy
,
1995,
Proceedings of IEEE 36th Annual Foundations of Computer Science.
[3]
Gang Quan,et al.
Energy efficient fixed-priority scheduling for real-time systems on variable voltage processors
,
2001,
DAC '01.
[4]
Eero Hyvönen,et al.
Constraint Reasoning Based on Interval Arithmetic: The Tolerance Propagation Approach
,
1992,
Artif. Intell..
[5]
Daniel Mossé,et al.
Adaptive scheduling server for power-aware real-time tasks
,
2004,
TECS.