云环境下基于多目标的多科学工作流调度算法  

Multi-Scientific Workflow Scheduling Algorithm Based on Multi-Objective in Cloud Environment

作　　者：袁友伟[1,2] 鲍泽前 俞东进 李万清[1] YUAN You-Wei;BAO Ze-Qian;YU Dong-Jin;LI Wan-Qing(School of Computer Science and Technology,Hangzhou Dianzi University,Hangzhou 310018,China;Key Laboratory of Complex Systems Modeling and Simulation（Hangzhou Dianzi University）,Ministry of Education,Hangzhou 310018 China)

机构地区：[1]杭州电子科技大学计算机科学与技术学院,浙江杭州310018 [2]复杂系统建模与仿真教育部重点实验室(杭州电子科技大学),浙江杭州310018

出　　处：《软件学报》2018年第11期3326-3339,共14页Journal of Software

基　　金：国家自然科学基金(61370218);浙江省重点高校建设专项资金(GK158800205032)~~

摘　　要：针对现有云环境下的多科学工作流调度算法中存在的未考虑安全调度问题,提出了多科学工作流安全-时间约束费用优化算法MSW-SDCOA(multi-scientific workflows security-deadline constraint cost optimization algorithm).首先,MSW-SDCOA基于数据依赖关系压缩科学工作流,减少任务节点数从而节省了调度开销;并通过改进HEFT(heterogeneous earliest-finish-time)算法形成调度序列,以实现全局多目标优化调度;最后,通过优化ACO(ant colony optimization)中信息素更新策略和启发式信息,进一步改善费用优化效果.仿真实验表明,MSW-SDCOA算法在费用优化效果上比MW-DBS算法提高了约14%.To address the problem that safe scheduling is not taken into consideration in existing multi-scientific scheduling workflow algorithm in cloud environment, this paper proposes a multi-scientific workflows security-deadline constraint cost optimization algorithm（MSW-SDCOA）. First, based on data flow dependency, MSW-SDCOA compresses scientific workflow and reduces the number of task nodes to save scheduling cost. Secondly, through optimizing HEFT algorithm, a scheduling sequence is formed to realize overall multi-objective optimization scheduling. Lastly, by optimizing update strategies of pheromone and heuristic information in ant colony optimization（ACO）, cost optimization effect is further improved. The simulation experiment results show that the cost optimization effect of MSW-SDCOA algorithm is about 14% better than that of MW-DBS algorithm.

关 键 词：安全调度 费用优化 多科学工作流 压缩 分层计算 

分 类 号：TP306[自动化与计算机技术—计算机系统结构]