基于效益函数的网格任务调度算法

在动态、异构、分布广泛的网格环境中，对资源的调度是一个非常复杂而重要且具有挑战性的问题。本文针对网格环境中的动态性特点，特别是用户QoS要求的动态变化性，提出了一种基于效益函数的网格任务调度算法，并采用GridSim模拟器分别对该调度算法和模拟器自带的代价最优和时间最优的网格任务调度算法进行模拟。实验的结果表明：该调度算法更能体现用户对QoS要求的动态变化；在系统完成相同数量的网格任务时，消耗相同时间的情况下，该调度算法在代价上优于基于时间优化的调度算法；而花费相同预算的情况下，在时间上优于基于代价优化的调度算法。

Scheduling Algorithm in Grids Based on Benefit

Scheduling grid tasks onto the dynamic, heterogeneous and widely distributed grid environments is a complex and challenging task. According to dynamic characteristic in grids, especial for the dynamic QoS request by the users, a scheduling algorithm based on benefit is present in this paper. Running the algorithm onto the simulator GridSim and comparing it to the algorithm based on deadline and the algorithm based on budget implies that the algorithm based on benefit take advantageous over the two algorithms of the simulator in some degree. Because of adapting to the dynamic request of QoS by the users, the algorithm not only achieves better performance in budget than the algorithm based on deadline when they finished the number of gridlets is same and spent equal deadline, but also in deadline than the algorithm based on budget on the assumption that they finished the same set of gridlets and consumed the equal quantity of budget.