基于静态表驱动的实时调度能量优化算法  

A Real-Time Scheduling Energy Optimization Algorithm Based on Static Table Driven Approaches

作　　者：刘云生[1] 张锐[1] 许贵平[1] 

机构地区：[1]华中科技大学计算机科学与技术学院,湖北武汉430074

出　　处：《计算机工程与科学》2007年第2期73-75,88,共4页Computer Engineering & Science

基　　金：国家自然科学基金资助项目(60073045);"十五"预研项目基金资助项目(J15.3.3)

摘　　要：传统的嵌入式实时调度策略为了保证系统的实时性,一般使处理器工作在最大频率(最高电压)下,但面对受到能量约束的嵌入式应用显然是不适宜的。本文提出了一种实时调度静态能量优化算法RT-SEO。若一个任务集可以被静态表驱动方法调度,该算法采用动态电压调节技术对系统实现静态的能量优化。RT-SEO在不影响任务可调度性的前提下,使系统的能耗得到显著的改善。Traditional embedded real-time scheduling strategies make the CPU work under its maximum voltage to satisfy the deadline constraints of real-time tasks.But this is obviously unsuitable for the embedded application with energy constraints.In this paper,a real-time energy optimization algorithm named RT-SEO is presented.If a task set is schedulable under static table driven approaches,this algorithm optimizes the energy characteristics of the system statically by dynamic voltage scaling.RT-SEO can reduce the energy cost of a real-time system obviously without deferring any deadline of the task set.

关 键 词：静态表驱动 动态电压调节 任务段 优化周期 

分 类 号：TP316.2[自动化与计算机技术—计算机软件与理论]