Computer Science and Engineering, Department of

 

Date of this Version

2004

Comments

University of Nebraska-Lincoln, Computer Science and Engineering
Technical Report # TR-UNL-CSE-2004-0001

Abstract

Dynamic voltage scaling (DVS) algorithms save energy by scaling down the processor frequency when the processor is not fully loaded. Many algorithms have been proposed for periodic and aperiodic task models but none support the periodic and sporadic task models when the deadlines are not equal to their periods. A DVS algorithm, called General Dynamic Voltage Scaling (GDVS), that can be used with sporadic or periodic tasks in conjunction with the preemptive EDF scheduling algorithm with no constraints on the deadlines is presented here. The algorithm is proven to guarantee each task meets its deadline while saving the maximum amount of energy possible with processor frequency scaling when tasks execute with their worst-case execution times. GDVS was implemented in the ¹C/OS-II real-time operating system for embedded systems. Though theoretically optimal, the actual power savings realized with GDVS depends on the type of the task set and the processor’s DVS support. GDVS is tested and evaluated with both a real-time application and a simulated task set. A difference exists between the theoretical power savings and the actual power savings which is due to the limited number of frequency levels the Rabbit 2000 processor supports.

Share

COinS