云计算中多层次公平性QoS约束任务调度算法

Multiple-level fairness QoS constraint task scheduling algorithm in cloud computing

针对传统云任务调度算法只注重执行效率忽略分配公平性的问题,提出了一种满足多重公平性约束的任务调度QoS算法CTS_QFC.该算法利用社会资源分配的公平性理论模型,从用户任务与云资源提供方两个角度,将云任务调度问题建模为一种多重公平性QoS约束模型.第一层QoS按用户QoS偏好对任务分类,并按照任务分类建立一般期望效用函数.第二层QoS定义资源公平性评估函数,评估资源分配的公平性.结果表明,CTS_QFC算法不仅可以确保用户任务的高效执行,还可以提高资源分配与任务调度方案的公平性.

Aiming at the problem that the traditional cloud task scheduling algorithms only pay attention to the execution efficiency with ignoring the fairness of resource allocation,a cloud task scheduling QoS algorithm meeting multiple-level fairness constraints(CTS_QFC)was proposed.With the fairness theory model for the social resource allocation and from two aspects of user tasks and cloud resource providers,the cloud task scheduling problem was modeled as a multiple-level fairness QoS constraint model.The first level QoS classified the tasks according to the QoS preference of users,and established the general expected utility function according to the task classification.The second level QoS defined the resource fairness evaluation function to evaluate the fairness of resource allocation.The results showthat the CTS_QFC algorithm can not only ensure the efficient execution of user tasks,but can also improve the fairness of resource allocation and task scheduling.