NeTS: Small: Exploring Theoretical Foundation of Mobile Cloud: From One-Hop Neighbors to the Internet

2012年第三季度，全球智能手机使用数量达到10.38亿部，预计到2015年，智能手机用户将超过20亿人。随着平板电脑和智能手机这一快速增长的巨大资产，无线设备的绝对数量和密度已经超过了网络资源，无法支持用户期望的质量水平。利用云计算，移动云计算(MCC)最近被引入，以帮助移动用户充分利用云计算的优势，即使在可用网络容量减少的情况下。该项目的目标是探索基本问题，以促进我们对使用移动云传输无线数据流量的理解。所提出的工作旨在：(I)通过一跳通信来探索移动云小程序的充分必要条件，以确保短延迟和容易管理的任务调度；(Ii)通过利用对链路级动力学的最新理解来识别移动小云程序的理论极限，重点关注移动小云程序随时间的演变中同质和异质移动节点的差异；(Iii)研究新的算法，该算法不仅基于移动小云程序的空间位置，而且基于其他因素，例如节能、连接接入点、非指数接触间时间分布以及移动设备之间的社会联系；(Iv)在混合网络模型中，找出数据到达云之前可能需要多长时间和多远。

The number of smartphones in use worldwide reached 1.038 billion units during the third quarter of 2012, and smartphone users are expected to be over 2 billion by 2015. With this fast-growing and a large asset of tablets and smartphones, the sheer number and density of wireless devices has overwhelmed network resources to support the level of quality expected by users. Utilizing cloud computing, mobile cloud computing (MCC) has been recently introduced to facilitate mobile users to take full advantages of cloud computing, even with the reduced available network capacity. The objective of this project is to explore fundamental issues that advance our understanding of using mobile clouds in delivering wireless data traffic. The proposed work aims to (i) explore sufficient and necessary conditions of mobile cloudlets through one-hop communications to ensure short delay and easy management of task dispatch; (ii) identify theoretical limits of mobile cloudlets by harnessing the latest understanding of link-level dynamics with the focus on the discrepancy in homogeneous and heterogeneous mobile nodes in the evolution of mobile cloudlets over time; (iii) investigate new algorithms that detect mobile clouds that are based on not only their spatial locations, but other factors, such as power-saving, tethering access points, non-exponential inter-contact time distributions, as well as social connections among mobile devices; (iv) find how long and how far it could take before data arrives at the cloud in a hybrid network model.