论文部分内容阅读
非凡的服务质量是网格的基本特征,基于QoS的网格任务调度问题已成为国内外研究的热点.由于网格环境的复杂性和用户主观判断的模糊性,传统基于QoS的网格调度算法不能处理具有多个模糊QoS需求的调度问题.用区间值模糊来描述用户的主观QoS需求信息,提出基于D-S理论的区间值模糊多QoS测量方法,对区间值模糊数表达的用户多QoS需求进行融合与处理,再将这些区间值模糊数QoS融合值作为任务的优先级加入到任务调度算法中,提出一种基于D-S理论的网格任务多匹配调度算法.研究结果和仿真实验表明,该算法不仅可以处理区间值模糊不确定性信息,在满足用户多QoS需求的情况下还可大大减小任务调度的时间跨度.
Due to the complexity of the grid environment and the vagueness of users’ subjective judgment, the traditional grid-based scheduling algorithm based on QoS (QoS) has become the focus of research at home and abroad. We can not deal with the scheduling problem with multiple fuzzy QoS requirements.Using interval fuzzy to describe the subjective QoS requirement information of users, we propose a fuzzy multi-QoS interval measurement method based on DS theory, Fusion and processing, and then these interval valued fuzzy number QoS fusion value is added as task priority to the task scheduling algorithm, a grid task multi-matching scheduling algorithm based on DS theory is proposed. The research results and simulation results show that the algorithm It can not only deal with fuzzy uncertainty information of interval value, but also greatly reduce the time span of task scheduling when satisfying the multi-QoS requirements of users.