摘要
在深入研究网格环境下任务调度算法的基础上,提出一种基于QoS的协作型任务调度遗传算法并通过引入协作型任务的形式化描述DAG图构造了QoS参数模型.该参数模型提出了任务完成时间、价格和可靠性三个QoS参数并将这些QoS参数引入遗传算法,实现了网格环境下协作型任务调度对服务质量的优化并保证了协作型任务之间的数据依赖.通过与DAG-MIN和DAG-GSA算法的对比实验表明,该算法能在保证较优调度性能的同时大幅度提高调度的服务质量.
On the basis of researching the resource scheduling strategy deeply in grid,a QoS-based cooperation-type task scheduling genetic algorithm is proposed.This article introduces formal description of cooperation-type task,DAG graph,to construct the QoS parameter model.This model proposes three QoS parameters including task finish time,price and reliability and bring them into genetic algorithm to achieve cooperation-type task scheduling optimization of quality of service in grid and assure the Data Dependency between cooperation-type tasks.Comparing to DAG-MIN and DAG-GSA Simulation results show that this algorithm can ensure a good schedule performance,and improves the quality of service.
出处
《小型微型计算机系统》
CSCD
北大核心
2011年第8期1643-1646,共4页
Journal of Chinese Computer Systems