• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2010, Vol. 32 ›› Issue (12): 145-148.doi: 10.3969/j.issn.1007130X.2010.

• 论文 • 上一篇    下一篇

一种动态可重构系统的实时任务调度算法

焦铬1,2,李仁发2,彭日光2,李浪1,2,邹祎1   

  1. (1.衡阳师范学院计算机科学系,湖南 衡阳 421008;2.湖南大学计算机与通信学院,湖南 长沙 410082)
  • 收稿日期:2009-08-28 修回日期:2009-12-05 出版日期:2010-12-25 发布日期:2010-12-25
  • 通讯作者: 焦铬
  • 作者简介:焦铬(1979),男,湖南岳阳人,硕士生,讲师,CCF会员(E200013163G),研究方向为嵌入式计算;李仁发,教授,博士生导师,研究方向为嵌入式计算、无线网络等;彭日光,硕士生,研究方向为可重构计算;李浪,博士生,研究方向为嵌入式系统。
  • 基金资助:

    国家863计划资助项目(2007AA01Z104);湖南省教育厅科学研究项目(10C0488)

A RealTime Task Scheduling Algorithm for Dynamic Reconfigurable Systems

JIAO Ge1,2,LI Renfa2,PENG Riguang2,LI Lang1,2,ZOU Yi1   

  1. (1.Department of Computer Science,Hengyang Normal University,Hengyang 421008;
    2.School of Computer and Communications,Hunan University,Changsha 410082,China)
  • Received:2009-08-28 Revised:2009-12-05 Online:2010-12-25 Published:2010-12-25

摘要:

硬件任务的软实时调度是影响动态可重构系统性能的关键因素之一。本文提出了一种基于顶点链表的硬件任务间最小空间调度算法MSSA,该算法将硬件任务按照长、宽及调度时间构成一个三维资源模型,以到达任务与已放置任务在三维空间的邻接度来构建代价函数,获取具有最大代价函数值的放置位置和启动时间,可使任务安排得更紧凑,减小对系统资源的浪费,提高并行度。仿真实验表明,与MSG4V和Stuffing算法相比,本文算法具有更高的芯片利用率和任务接受率。

关键词: 动态可重构, 调度算法, 代价函数

Abstract:

Soft realtime scheduling of the hardware tasks is one of the key factors which affect the performance of dynamic reconfigurable systems. This paper presents a scheduling algorithm MSSA (minimun space scheduling algorithm) based on the minimum space between the hardware tasks of the vertex list. The algorithm constitutes a threedimensional resources model according to the length,width and the activation time to reach the adjacent degree of the task and the task placed in the threedimensional space to construct the cost function, which obtains the place position and the start time of the greatest value of the cost function and enables the task scheduling even more compact, reducing the waste of the system resources to improve the degree of parallelism.The simulation results show that this algorithm has a higher chip utilization and acceptance rate of tasks compared with the MSG4V algorithm and the Stuffing algorithm.

Key words: dynamic reconfiguration;scheduling algorithm;cost function