Journal of Systems Engineering and Electronics ›› 2009, Vol. 31 ›› Issue (2): 384-389.

• 系统工程 • 上一篇    下一篇

行为综合功能流水线中的资源约束LB-ACO算法

宋德强, 段成华, 黄庆明   

  1. 中国科学院研究生院信息科学与工程学院, 北京, 100049
  • 收稿日期:2007-11-27 修回日期:2008-03-24 出版日期:2009-02-20 发布日期:2010-01-03
  • 作者简介:宋德强(1981- ),男,博士研究生,主要研究方向为专用集成电路行为综合调度算法.E-mail:songdeq@mails.gucas.ac.cn
  • 基金资助:
    国家“863计划”项目资助课题(2002AA141041)

Resource-constrained LB-ACO algorithm for functional pipelines in behavioral synthesis

SONG De-qiang, DUAN Cheng-hua, HUANG Qing-ming   

  1. School of Information Science and Engineering Graduate Univ. of Chinese Academy of Sciences, Beijing 100049, China
  • Received:2007-11-27 Revised:2008-03-24 Online:2009-02-20 Published:2010-01-03

摘要: 针对行为综合功能流水线中带资源约束的调度问题,提出了一种新的功能流水线调度算法负载平衡蚁群调度算法(load-balanced ant colony scheduling,LB-ACO).该算法将蚁群算法和力向算法的受力公式相结合,在保证较低运行时间复杂度O(cn2)的前提下,可获得近似最优的流水线调度结果.通过确定蚁群算法的局部试探因子和全局试探因子的计算方法,描述了该算法的实现步骤.通过仿真实验证明了该算法的流水线调度性能.

Abstract: To study the problems of resource-constrained functional pipeline scheduling in behavioral synthesis,proposes a new pipeline scheduling algorithm—Load-Balanced Ant Colony Scheduling(LB-ACO).This algorithm combines Ant Colony Optimization evolutionary algorithm with force equation in Force-Directed Scheduling,it can get near optimal pipeline scheduling result while ensuring low time complexity O(cn2).We describe this algorithm by introducing the compute methods of the global heuristic and local heuristic in ACO al- gorithm.To use simulate experiment to prove this algorithm's pipeline scheduling performance.

中图分类号: