基于分层和分组的移动网格资源选择算法

Hierarchical and group-based resource selection algorithm

针对移动网格的特殊环境,利用统计学中多次线性回归函数提出一种基于分层和分组的移动网格资源选择算法,该算法考虑了移动设备电池能量低、移动性及其处理能力弱等特征。算法中将移动资源分成不同级别层次并分成K组,将任务同时分配到符合资源请求的n组(n<K)中,且同时执行。实例分析结果表明,该算法不但避免了由于资源的移动性或者电池能量的不足而导致任务执行结果无法反馈的情况,而且能加快任务与资源的匹配速度,支持异构环境中的容错性。

As to the special environment of the mobile grid, this paper proposed a hierarchical and group-based resource selection algorithm that supported fault-tolerance in mobile grid and considered various characteristics of mobile devices. The proposed algorithm considered remaining battery power, mobility, and performance of mobile devices, calculated rank of mobile devices and classified to K groups, and then assigned a job to top-N groups which could follow the resource request concurrently for fault-tolerance. Result manifests that ranking and grouping of mobile devices prevent there is no results due to the powerless or mobility of mobile device whenever a job is submitted and speed up the matching between the task and resource.