摘要
移动网格环境具有高度的动态性,任意时刻都可能发生资源的变化,任务调度因此变得复杂,提出了一种面向服务质量(QoS)的移动网格任务调度策略.建立资源和任务模型,定义任务优先级,将任务分成若干子集,子集内调度采用改进遗传算法,选取时延和跨度为QoS指标,编码和遗传操作上体现资源对任务QoS的匹配调度,使最优解尽力满足任务QoS;同一资源上分配的任务,按优先级顺序执行.仿真结果表明,调度算法可以较好地满足用户的QoS,调度跨度也较优.
Mobile grid is a highly dynamic resource environment. Thusly,possible changes in resource anytime under such an environment will result in a complex task scheduling. In the paper, the authors put forward a QoS - oriented mobile grid task scheduling strategy. The strategy helps construct the resource and task models. After determining the priority of tasks, it further divides the tasks into several subsets. Besides, the scheduling of different subsets is based upon improving genetic algorithm,with time delay and makespan being the QoS in- dexes. Further, the matching and scheduling of QoS tasks will be performed in coding and genetic operation so that the optimal solution can satisfy QoS tasks. Finally,tasks on the same resource allocation can be performed according to the priority list. In conclusion, the simulation results indicate that scheduling algorithm can better satisfy users＇QoS, and system performance makespan can also be improved.
出处
《河南理工大学学报（自然科学版）》
CAS
北大核心
2014年第5期635-639,共5页
Journal of Henan Polytechnic University（Natural Science）
基金
国家自然科学基金资助项目(61340014)
河南省科技厅重点科技攻关项目(122102210116)
关键词
移动网格
任务调度
QOS
最优跨度
遗传算法
mobile grid
task scheduling
Quality of Service （QoS）
makespan
genetic algorithm
