Carry Small Enjoy Large: a Mobile Cloud Computing Approach

移动的设备正在转变为无处不在的计算平台，对我们的生活、工作和娱乐方式产生了深远的影响。 具有高级功能的新移动的应用程序每天都在创建，并进入我们的生活。然而，这种向全能的移动的互联网发展的趋势受到以下事实的阻碍：与桌面设备相比，移动的设备由于有限的计算能力和电池寿命而固有地资源贫乏。因此，在计算密集型应用程序和资源贫乏的移动的设备之间存在着争斗。最近，一种移动的云计算范例正在出现，并且能够满足计算密集型移动的应用的需求，并且反映了主要研究者的“携带小享受大”的愿景：携带小型移动终端，同时享受云计算基础设施提供的大量资源。在这种模式下，一个小的移动终端可以提供丰富的经验，计算，电话，多媒体，娱乐，游戏和互联网。为了实现携带小享受大的愿景，本项目旨在开发新的理论和技术，显着推进当前的移动的云计算技术的承诺。 为了实现这一目标，本文的主要研究工作如下：1）提出一种基于马尔可夫决策过程和有效容量的计算任务调度方法，该方法能够在衰落信道下实现最优能耗，同时保证用户指定的任务完成延迟性能; 2）为移动的视频流设计数据预取调度器，其在满足观看者的体验质量的同时最小化未消耗的视频数据所引起的成本; 3）设计一个资源优化的视频编解码器，保证游戏视频的低延迟和高质量体验;以及4）开发一个实时无线测试平台。所提出的方法将填补一些重要的空白，在基本的理解移动的云计算和网络在时变衰落信道，并将提供系统设计和算法实现的理论基础。所提出的计算任务调度方法不仅将提供一种实现我们“以小博大”愿景的使能技术，而且还将推动绿色计算技术的前沿。 该项目将涉及研究生和本科生，并吸引来自代表性不足群体的学生。该移动的云计算应用程序将提供一个理想的平台，让本科生和K-12学生参与增强的教育和研究经验。该项目的变革方面是，拟议的移动的云计算研究将导致新的方法，用于设计移动的视频流系统和云游戏系统，具有前所未有的能力，实现最佳的资源利用和高质量的体验。

Mobile devices are being transformed into a ubiquitous computing platform, resulting in profound impact on the way we live, work and play. New mobile applications with advanced features are being created every day and finding their way into our lives. However, this trend toward omnipotent mobile Internet is hampered by the fact that mobile devices, compared to their desktop counterparts, are inherently resource-poor, due to limited computing power and battery lifetime. As a result, there exists a tussle between computation-intensive applications and resource-poor mobile devices. Recently, a mobile cloud computing paradigm is emerging and is capable of answering the needs of computation-intensive mobile applications, and reflects the principal investigator's vision of carry small enjoy large: carry a small mobile device while enjoying a large amount of resources offered by cloud computing infrastructure. Under this paradigm, a small mobile device can deliver a rich experience of computing, telephony, multimedia, entertainment, gaming, and Internet.To realize the vision of carry small enjoy large, this project is intended to develop new theories and techniques with the promise of significantly advancing the current technologies of mobile cloud computing. To accomplish this, the following research tasks will be conducted: 1) developing a computing-task scheduling approach based on Markov decision process and effective capacity, which is capable of achieving optimal energy consumption while guaranteeing user-specified task-completion delay performance over fading channels; 2) designing a data-prefetch scheduler for mobile video streaming, which minimizes the cost incurred by unconsumed video data while satisfying the viewer's quality of experience; 3) designing a resource-optimized video codec that guarantees low delay and high quality of experience for gaming video; and 4) developing a real-time wireless testbed. The proposed approaches will fill some important gaps in fundamental understanding of mobile cloud computing and networking over time-varying fading channels, and will provide the theoretical underpinning for system design and algorithm implementation. The proposed approach to computing-task scheduling will not only provide an enabling technology that realizes our vision of carry small enjoy large, but also push the frontier of green computing technologies. The project will involve graduate and undergraduate students, and attract students from underrepresented groups. The mobile cloud computing applications will offer an ideal platform to engage undergraduate and K-12 students with enhanced education and research experience.The transformative aspect of the project is that the proposed research on mobile cloud computing will result in new methodologies for the design of mobile video streaming systems and cloud gaming systems with unprecedented capabilities of achieving optimal resource usage and high quality of experience.