项目名称: 面向多云块并行移动计算迁移的环境自适应程序分割技术研究
项目编号: No.61502050
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: 范文浩
作者单位: 北京邮电大学
项目金额: 20万元
中文摘要: 移动计算迁移是移动云计算中的核心技术之一,它将移动应用程序的计算任务从资源受限的移动终端迁移至外部资源相对丰富的计算设备中执行,可有效增强移动终端的应用程序承载能力。本项目重点聚焦于移动计算迁移技术中的移动应用程序分割技术,以降低终端能耗为目标,在多云块并行迁移模式下,从程序分割技术对无线与计算环境变化的适应性角度出发,深入研究面向多云块并行移动计算迁移的环境自适应程序分割技术,主要包括:研究基于多元属性、随机关联与可变粒度的程序分析方法,建立组件级与任务级二维抽象,实现程序抽象粒度的细化和粗化转换机制;研究基于异构云块多重映射的并行迁移机制,以图映射方式分别探索在服务运营者和用户角度下的多约束优化问题;研究具有环境变化容忍性的动态程序分割算法,实现计算复杂度与求解精确度的动态调整机制。研究工作对于推动移动计算迁移技术的发展,实现“端-管-云”资源的优化利用,具有重要的科学意义和应用价值。
中文关键词: 计算迁移;移动云计算;移动应用程序分割;环境自适应;组合优化
英文摘要: Mobile computation offloading is the core in clouding computing technologies, it migrates computing tasks from resource-constrained mobile terminals to external relative resourceful computing devices, which can effectively enhance mobile terminals’ capacities for supporting diverse mobile applications. This project puts the emphasis on mobile application partitioning technologies in mobile computation offloading, and aims at decreasing energy consumptions of mobile terminals. From the view of adaptability of partitioning in wireless and computing environments, the project studies multi-cloudlets parallel mobile computation offloading-oriented environment self-adaptive application partitioning technologies, which mainly includes: we explore application analysis methods using multiple attributes, random associations and variable granularities, complete component-level and task-level abstractions, and realize conversion mechanisms between fine and coarse granularities; we explore parallel offloading strategies based on multiple mappings of heterogeneous cloudlets, and investigate multi-constrained optimization problem for service providers and terminal users, respectively; we explore environment change-tolerant application dynamic partitioning algorithms, and design methods with configurable algorithm complexities and result accuracies. Our research work has important scientific significance and application value on promoting the development of mobile computation offloading technologies, and achieving optimal utilization of the resources in the ‘terminal – pipeline – cloud’ architecture.
英文关键词: Computation Offloading;Mobile Cloud Computing;Mobile Application Partitioning;Environment Self-adaption;Combinatorial Optimization