计算机工程
計算機工程
계산궤공정
COMPUTER ENGINEERING
2015年
2期
12-16
,共5页
云计算%蚁群算法%任务调度%公平性%任务延迟时间
雲計算%蟻群算法%任務調度%公平性%任務延遲時間
운계산%의군산법%임무조도%공평성%임무연지시간
cloud computing%ant colony algorithm%task scheduling%fairness%task delay time
为解决云环境下的资源调度问题，提出一种能改善任务并行性与兼顾任务串行关系的调度模型，将用户提交的动态任务分割成具有制约关系的子任务，按运行次序放到具有不同优先级的调度队列中。针对同一调度队列中的子任务，采用基于最短任务延迟时间的改进蚁群算法( DSFACO )进行调度，在兼顾调度公平性与效率的前提下，最大化缩短任务延迟时间，从而提高用户满意度。实验结果表明，与任务调度增强蚁群算法相比， DSFACO算法在任务延迟时间、调度公平性及效率方面性能更好，能实现云计算环境下任务的最优调度。
為解決雲環境下的資源調度問題，提齣一種能改善任務併行性與兼顧任務串行關繫的調度模型，將用戶提交的動態任務分割成具有製約關繫的子任務，按運行次序放到具有不同優先級的調度隊列中。針對同一調度隊列中的子任務，採用基于最短任務延遲時間的改進蟻群算法( DSFACO )進行調度，在兼顧調度公平性與效率的前提下，最大化縮短任務延遲時間，從而提高用戶滿意度。實驗結果錶明，與任務調度增彊蟻群算法相比， DSFACO算法在任務延遲時間、調度公平性及效率方麵性能更好，能實現雲計算環境下任務的最優調度。
위해결운배경하적자원조도문제，제출일충능개선임무병행성여겸고임무천행관계적조도모형，장용호제교적동태임무분할성구유제약관계적자임무，안운행차서방도구유불동우선급적조도대렬중。침대동일조도대렬중적자임무，채용기우최단임무연지시간적개진의군산법( DSFACO )진행조도，재겸고조도공평성여효솔적전제하，최대화축단임무연지시간，종이제고용호만의도。실험결과표명，여임무조도증강의군산법상비， DSFACO산법재임무연지시간、조도공평성급효솔방면성능경호，능실현운계산배경하임무적최우조도。
To solve the problem of resource scheduling problem in cloud computing, a parallel scheduling model is proposed,which can improve the task parallelism while maintaining the serial relationships between tasks. Dynamic tasks submitted by users are divided into sub-tasks in some serial sequences,and it puts into scheduling queue with different priorities according to running order. For these tasks in the same priority scheduling queue, an improved Delay Time Shortest and Fairness Ant Colony Optimization(DSFACO) algorithm is applied to schedule. Considering both fairness and efficiency,DSFACO algorithm applies to subtask scheduling problem to realize shortest delay time,thus improves the user satisfaction. Experimental results show DSFACO algorithm is better than the TS-EACO algorithm in fairness, efficiency and task delay time,and it can realize the optimal scheduling in cloud computing.
