改进cs算法结合决策树的云工作流调度-电子科技大学学报.pdfVIP

第45 卷 第6 期 电 子 科 技 大 学 学 报 Vol.45 No.6 2016年11月 Journal of University of Electronic Science and Technology of China Nov. 2016 改进CS算法结合决策树的云工作流调度 陈 超 ( 四川理工学院计算机学院 四川 自贡 643000) 【摘要】对云计算环境下工作流任务调度的现有方案进行分析，针对存在运行时间长、资源利用率低等不足，提出一种 结合改进型布谷鸟搜索算法和决策树的工作流任务调度方案。首先，根据工作流任务属性分配截止期限；其次，利用改进型 布谷鸟搜索算法将工作流分割成多个子工作流，最小化数据依赖性，再利用决策树选择出满足任务QoS约束的资源；最后， 根据任务的计算时间、排队时间和通信延迟的总和来判断是否满足截止期限约束，以此配置相应的资源。实验结果表明，该 方案具有较短的总运行时间和较高的任务完成率。 关 键 词 云计算; 布谷鸟搜索; 决策树; 工作流划分; 工作流调度 中图分类号 TP311 文献标志码 A doi:10.3969/j.issn.1001-0548.2016.06.017 Workflow Task Scheduling in Cloud Computing Based on Hybrid Improved CS Algorithm and Decision Tree CHEN Chao (School of Computer Science, Sichuan University of Technology Engineering Zigong Sichuan 643000) Abstract The existing workflow task scheduling schemes in cloud computing environment are analyzed, For the issues of the long operation time and low resource utilization, a workflow task scheduling scheme base on hybrid improved cuckoo search and decision tree in cloud computing is proposed. First, the deadline is assigned according to the work-flow task attribute; then, the improved cuckoo search algorithm is used to split the workflow into several sub workflow, minimizing data dependent; then, the decision tree is used to choose the resources which meet the QoS constraints of tasks; finally, the deadline constraints to be satisfied is judged according to satisfy according to the sum of task computing time, queuing time and communication delay, so as to configure the appropriate resources. Experimental results show that the proposed scheme has shor