An Integrated Approach for Applying Dynamic Voltage Scaling to Hard Real-Time Systems

  • Authors:
  • Yanbin Liu;Aloysius K. Mok

  • Affiliations:
  • -;-

  • Venue:
  • RTAS '03 Proceedings of the The 9th IEEE Real-Time and Embedded Technology and Applications Symposium
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

Wireless and portable devices depend on the limitedpower supplied by the battery. Dynamic Voltage Scaling(DVS) is an effective method to reduce CPU power consumption.For real-time systems, DVS algorithms must notonly provide enough CPU cycles, but also guarantee thatno job misses its deadline. In this paper, we propose an integratedapproach for applying DVS to real-time systems.We define two functions, the available cycle function (ACF)and the required cycle function (RCF), to capture the CPUworkload of the real-time tasks. We then formulate the DVSscheduling problem for real-time systems as a nonlinear optimizationproblem and propose an optimal off-line algorithmto solve this problem. We also propose a novel onlinealgorithm with time complexity O(1) to further reducepower consumption when a job uses fewer execution cyclesthan the worst-case budget. The algorithms in this paperare based solely on ACF and RCF, and may be applied todifferent scheduling policies. We illustrate the generalityof our approach over previous research by applying ourmethod to EDF and RM scheduling policies and derivingthe optimal off-line DVS algorithms for them. Our simulationresults show significant improvement over previouswork.