動態電壓調整是一種能有效降低微處理器能量消耗的技術，而由於製程的進步，加劇了漏電流的情況發生，也導致靜態功率大量地消耗，因此我們同時針對這兩種消耗功率來做改善，提出新的省電排程方式，應用在多處理器的即時系統上。在即時系統中，使用動態電壓調整技術可能會面臨到時間限制的問題，在系統工作量高時，無法有效地降低能量消耗；而動態降低電壓也會拉長執行時間，導致靜態功率消耗更為嚴重，所以如何達到一種動靜態功率消耗的平衡，是本篇論文所要探討的。
我們所建構的多處理器省電排程演算法大致可以分為兩部份：第一部分主要負責將系統中的任務分派到各顆處理器，使Slack Time能夠更有效地分配利用。第二部份則是著重在單顆處理器上的動靜態功率消耗排程機制上；在動態方面是採用最大省能逼近來算出最貼近的縮放比率，並藉由動態提升優先權來減少本文交換的耗電；在靜態方面，提出延遲機制使系統能夠切換到睡眠狀態，以更進一步來減少靜態功率消耗。根據實驗模擬結果顯示，在最佳情況下，系統在我們所提出的省電排程下，能量消耗可以減少達35%。

Dynamic Voltage Scaling (DVS) is a technique which can effectively reduce energy consumption of a processor. On the other hand, as the progress in process, occurrence of leakage current becomes severer which leads to static power consumption. Taking both dynamic and static power consumption into consdieration, this thesis proposed an energy-aware scheduling for multiprocessor real-time systems. In real-time systems, jobs in each task must meet their deadlines. Since lower execution voltage requires longer execution time, DVS might not be applicable when the systeme is in a heavy load. While execution time is prolonged by DVS, static power consumption increased as well. How to balance between dynamic and static power consumptions is the major issue of the thesis.
The proposed energy-aware scheduling algorithm contains two parts: The first part distributes tasks in the system to processors so that slack times could be better utilized. After each task is assigned to a processor, the second part of the algorithm takes care of dynamic/static energy-aware scheduling on each processor. We adopt the largest provincial energy to approach the closest scaling ratio. By adaptively changing priority of jobs, the number of context switches can be reduced. With the proposed delay scheme, the system has more chance to stay in sleep mode, which could better reduce the static power consumption. The simulation results show that the system with our energy-aware scheduling could save energy consumption up to 35%.

[1] I. Kadayif, M. Kandemir, and U. Sezer. ”An Integer Linear Programming Based Approach for Parallelizing Applications in On-Chip Multiprocessors” DAC, pp. 703-708，June，2002。

[2] B. Gonzalez and A. Horowitz. “Supply and Threshold Voltage Scaling for Low Power CMOS”IEEE Journal of Solid-State Circuits, Aug. 1997.

[3] Parth Malani, Prakash Mukre and Qing Wu. “Adaptive Scheduling and Voltage Scaling for Multiprocessor Real-time Applications with Non-deterministic Workload”DATE08, 2008 EDAA.

[4] Intel PXA27X Processor, Developer’s manual, Jan.2004.

[5] M. Vasic, O. Garcia, J.A. Oliver, P. Alou, J.A. Cobos.“A DVS System Based on the Trade-off Between Energy Savings and Execution Time”2008 IEEE.

[6] Dakai Zhu, Rami Melhem and Bruce Childers “Scheduling with Dynamic Voltage/Speed Adjustment Using Slack Reclamation in Multi-Processor Real-Time Systems” 2001 IEEE

[7] Y.-H. Lee, K.P. Reddy, and C.M. Krishna. “Scheduling techniques for reducing leakage power in hard real-time systems” In EuroMicroConf. On Real Time Systems, pages 105-112. IEEE Computer Society, 2003

[8] G. Quan, L. Niu, X.S. Hu, and B. Mochocki. “Fixed priority scheduling for reducing overall energy on variable voltage processors.” In 25th IEEE Real-Time System Symposium, pages 309-318. IEEE Computer Society, 2004.

[9] H. Aydin, V. Devadas, and D. Zhu. “System-level energy management for periodic real-time tasks.” In IEEE Real-Time Systems Symposium, pages 313-322, 2006.

[10] R. Jejurikar, C. Pereira, and R. Gupta. “Leakage aware dynamic voltage scaling for real-time embedded systems.” In DAC ’04：Proceedings of the 41st annual conference on Design automation, pages 275-280, New York, NY, USA, 2004. ACM Press.

[11] J.-J. Chen and T.-W. Kuo. “Procrastination for leakage-aware rate-monotonic scheduling on a dynamic voltage scaling processor.” In ACM SIGPLAN Conference on Language, Compiler, and Tool Support for Embedded Systems, pages （accepted）, June 2006.

[12] W. Zhang, M. Kandemir, N. Vijaykrishnan, M. Irwin, and V. De. “Compiler support for reducing leakage energy consumption” ,2003.

[13] R. Jejurikar and R. Gupta. “Dynamic slack reclamation with procrastination scheduling in real-time embedded systems.” In Design Automation Conference, June 2005.