一种面向雾计算的任务调度算法研究

Research of a task scheduling algorithm in fog computing

在雾计算环境中,为提高雾计算效率和给用户请求的任务分配合适的资源,需要对任务调度问题进行研究。面向流水线独立任务进行调度研究,首先基于经典Apriori算法提出一种雾计算环境中任务的分类I-Apriori算法;将得到的分类规则以及加权最早完成时间作为任务优先级选择的依据;对于同时达到的任务,优先选择出现在调度关系中的任务进行调度,其他雾结点则依据最早完成时间和加权链接数的高低进行调度。通过仿真实验对ITPS(improved task priority scheduling)算法的性能进行了评估,结果表明,ITPS算法在makespan及AWT(average waiting time)方面具有较好的性能。

In the heterogeneous and distributed computing environment of fog computing,in order to improve the efficiency of fog computing and allocate appropriate resources to corresponding tasks,task scheduling problem needs to be studied.The scheduling of pipeline independent tasks is studied.Firstly,based on the traditional Apriori algorithm,a task classification algorithm I-Apriori algorithm in fog computing environment is proposed.Association rules generated by I-Apriori algorithm are combined with the weighted earliest completion time of tasks in the task set.Tasks appear in the association rules are selected to schedule first,other fog nodes are scheduled according to the earliest completion time and the number of weighted links.The performance of ITPS algorithm is evaluated by simulation experiments.The results show that ITPS algorithm has a good performance in makespan and AWT.