Service Oriented Resource Allocation in Mobile Cloud Computing

无线网络和云计算技术的成功带来了技术进步，为不断增长的移动无线设备提供完整、透明和灵活的服务。移动云计算是一种新的模式，它承诺通过将计算资源从移动设备转移到云服务来提供处理和存储分流能力。挑战来自这样一个事实，即移动云计算功能组件的设计原理一直只关注非常有限的关键性能因素。高级移动云场景需要更复杂的机制，因为部署在更分布式和异类的基础设施中，从而推动了对新设计理念的需求。 该提案的长期目标是开发创新贡献(例如，技术、算法、方法等)，以支持与分布式资源分配系统相结合的卸载决策的高效建模和设计，该分布式资源分配系统能够动态地组成和自组织端到端的最优云基础设施。组合的基础设施针对移动应用工作负载的执行进行了优化。研究工作将包括以下四个短期目标： (1)为分散处理卸载决策开发分析模型，其中战略性软件组件和算法及时决定启动处理卸载过程的内容、时间和方式。 (2)设计具有各自功能的移动云计算体系结构的构建块，以满足移动应用的可扩展性要求；(Ii)为移动设备提供分散的决策机制，以减轻云计算提供商的管理负担；(Iii)最小化移动设备与分布式云计算基础设施之间的控制信令开销；以及(Iv)实现决策更新过程负载的动态计算(监控、弹性)。 (3)根据云资源的可用性、性能优化目标以及云和移动设备的拓扑感知等因素，提出了一种动态的资源配置和分配方法。 (4)开发工作原型，展示新型移动云计算框架的有用性、能力和性能。 该项目的成果将极大地推动移动云计算资源分配的最先进水平。这也将对工业移动云计算资源管理系统的发展产生深远的影响。这项研究对于移动应用的持续增长尤为关键。最后，它将为培养高素质的人才提供极好的机会。

The success of wireless networks and Cloud Computing technologies is carrying the technological advances to offer complete transparent and elastic services to the ever-growing number of mobile wireless devices. Mobile Cloud Computing is a novel paradigm that promises to provide processing and storage offloading capabilities by transferring computing resources from the mobile devices to cloud services. Challenges come from the fact that design rationale for Mobile Cloud Computing functional components has been focusing on only a very limited scope of key performance factors. Advanced Mobile Cloud scenarios require more complex mechanisms as result of being deployed in more distributed and heterogeneous infrastructures, thus driving the need for a novel design rationale. The long-term objective of this proposal is to develop innovative contributions (e.g., techniques, algorithms, methodologies, and the like) to support efficient modeling and design of offloading decision making coupled with a distributed resource allocation system capable of dynamically composing and self-organizing an end-to-end optimal cloud Infrastructure. The composed infrastructure is optimized for the execution of a mobile application workload. Research effort will include the following four short-term objectives: (1) Develop analytical modelling for decentralized processing offloading decision making where strategic software components and algorithms take timely decisions on what, when, and how to start the processing offloading procedures. (2) Design the building blocks of a Mobile Cloud Computing architecture with their respective functionalities, aiming at (i) accommodating scalability requirements for mobile applications; (ii) providing decentralized decision making mechanisms to mobile devices in order to ease the management load at the cloud computing providers; (iii) minimizing control signaling overhead between mobile devices and the distributed cloud computing infrastructure; and (iv) realizing the dynamic computation of?oading decision update procedures (monitoring, elasticity). (3) Develop a dynamic resource provisioning and allocation based on factors such as the availability of cloud resources, the performance optimization objective and the topology awareness of clouds and mobile devices. (4) Develop a working prototype to demonstrate usefulness, capabilities, and performance of the novel Mobile Cloud Computing framework. The outcome of the project will significantly advance the state-of- the-art on resource allocation in Mobile Cloud Computing. It will also have profound impact on the development of industrial Mobile Cloud Computing resource management systems. The research is particularly critical for the sustained growth of mobile applications. Finally, it will provide excellent opportunities for training highly qualified personal.