Leveraging social network theories to address data delivery issues in mobile ad hoc networks

点击翻译按钮获取中文摘要

We propose to investigate the problem of cross layer design of communication protocols that can reliably expedite end-to-end delivery of data in a mobile social network environment. Since the performance of such protocols is closely related to users' social relationships and movement patterns, our first task will be to develop a realistic mobility model for human interactions in a social networking context. As it turns out, previously proposed models are generally based on random walk mobility models which are contrary to human behavior, and as such their validity is suspect. To address this problem, we intend to leverage well known social networking concepts and previous work in modeling of human mobility to come up with a realistic model. We will test the validity of our model by using empirical data from large scale studies of human mobility patterns such as MIT's Reality Mining project. Next, we will proceed to design and develop a cross layer communication protocol where a novel social aspects layer interacts with all 7 layers of the standard OSI protocol stack in order to facilitate data delivery in a delay tolerant network. Our results will enable end-to-end delivery of data in a mobile social network environment where continuous end-to-end connectivity cannot be guaranteed. In the longer term, we propose to extend our investigations into social networking problems in pervasive computing settings. We intend to research the problem of social training of intelligent mobile devices carried by humans. We call these devices sociable guardians and we believe them to have important practical applications in low cost continuous medical monitoring, search and rescue operations and communications in hostile environments.