云端融合计算环境下QoS敏感的任务调度机制研究

项目名称： 云端融合计算环境下QoS敏感的任务调度机制研究

项目编号： No.61502103

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 自动化技术、计算机技术

项目作者： 林立

作者单位： 福建师范大学

项目金额： 21万元

中文摘要： 云计算和移动计算是当前的两大技术趋势。一方面是云计算强大的计算和存储能力，另一方面是智能移动设备无处不在的计算特性，结合云端/终端计算的云端融合计算正成为工业界和学术界广泛关注的焦点。云端融合计算环境是包含云端，终端以及两者连接网络的复杂环境，在云端和终端之间进行任务迁移具有很大的不确定性。当前的云端融合系统缺乏任务迁移执行时的QoS保证，对于不同移动应用程序无法进行程序特征识别的自适应调度。因此，本项目研究云端融合计算环境下应用程序特征自识别的QoS敏感的任务调度机制。具体而言，研究QoS目标度量和应用程序特征描述策略以及策略的规则表达，并基于这种规则表达进行任务划分和优化，实现运行时资源感知的自适应任务调度机制。该机制的引入将有助于解决云端融合中执行性能保证和任务调度的关键问题，使云端融合更好的应用于实际。

中文关键词： 云端融合；QoS；云计算；移动计算；任务调度

英文摘要： Cloud computing and mobile computing are currently two major technology trends. On the one hand, cloud computing has a powerful computing and storage capacity. On the other hand, smart mobile devices are characteristics of ubiquitous computing. The fashion of cloud mobile fusion which combines cloud computing and mobile computing has caught much attention of industry and academia. The runtime of cloud mobile fusion is a complicate environment that contains cloud, mobile and network, in which it is an uncertain thing to migrate tasks between cloud and mobile. However, the current cloud mobile fusion systems are lack of QoS guarantee when executing migrated task, and can not adjust their task scheduling according to the characteristic of applications. This project studies a task scheduling mechanism which is QoS-Sensitive and application self-identification in cloud mobile fusion. Specifically, we make researches on the description of QoS metrics and application characteristic, and then the rule expression of this description. Further, we study task partitioning and optimization based on the rule expression in order to implement an adaptive and resource-aware task scheduling mechanism. The introduction of this mechanism will help address the key issue of performance guarantee and task scheduling in cloud mobile fusion, and make it better applied.

英文关键词： Cloud Mobile Fusion;QoS;Cloud Computing;Mobile Computing;Task Scheduling