English

应用科学学报 ›› 2005, Vol. 23 ›› Issue (5): 483-488.

• 论文 • 上一篇    下一篇

基于AR模型的PCF动态电源管理预测策略

卜爱国, 刘昊, 胡晨, 王超   

  1. 东南大学国家专用集成电路系统工程技术研究中心, 江苏南京 210096
  • 收稿日期:2004-05-15 修回日期:2004-09-01 出版日期:2005-09-30 发布日期:2005-09-30
  • 作者简介:卜爱国(1978-),男,江苏兴化人,博士生,E-mail:patriotic@163.com

PCF Dynamic Power Management Predictive Policy Based on AR Model

BU Ai-guo, LIU Hao, HU Chen, WANG Chao   

  1. National ASIC System Engineering Research Center, Southeast University, Nanjing 210096, China
  • Received:2004-05-15 Revised:2004-09-01 Online:2005-09-30 Published:2005-09-30

PDF

83

被引次数

8

摘要： 基于AR(auto-regressive)模型提出了一种新颖的预测控制反馈(predictive-control-feedback,PCF)预测策略,该策略能够很好地解决某些文献所提出策略中存在的问题.实验结果表明,PCF策略具有稳定性;与其他预测性策略相比,该策略能够进一步降低系统的平均功耗.

关键词: 嵌入式系统, 预测策略, 动态电源管理, AR模型

Abstract: Energy consumption is an important design criterion for embedded and portable systems.Dynamic power management (DPM) is aimed at reducing power consumption of electronic systems by selectively shutting down idle components.In this paper, a novel PCF DPM policy based on an AR model is presented, which can solve the problems as mentioned in some publications.Experimental results show that the PCF policy is stable and the average power consumption of system can be further reduced as compared to other predictive policies.

Key words: embedded system, dynamic power management, AR model, predictive policy

中图分类号: