CSR-DMSS, SM: DoC - Distributed Opportunistic Computing (0834493)
The objective of this project is to carry out preliminary, fundamental research work in the area of opportunistic computing. When pairs of devices come within each others? communication range, opportunistically, short-lived links (or opportunistic links) are created. Opportunistic computing exploits the opportunistic links created by pair-wise contacts, to share information content, resources and services, leading to a wide variety of applications. Groups of computing nodes and their associated pair-wise contacts in an opportunistic network give rise to a distributed opportunistic computing system. Essentially, opportunistic computing can be described as distributed computing with the caveats of intermittent connectivity and delay tolerance.
The novelty of the proposed work lies in the exploitation of opportunistic communication contacts to provide collaborative computing services to applications and users. The unique contributions of this project include the development of an adaptive protocol for opportunistic communication in support of computing, a middleware architecture for masking the disruptive nature of the underlying network from applications and users, and a mechanism for delay tolerant, remote execution of tasks.
The research will lead to tangible contributions to the improvement in human quality of life and make significant impact on Internet based applications, computer science and engineering. The project outcomes have relevance to applications in crisis management situations, entertainment, transportation, education and military. In general, the outcomes of this project have broad applicability in such areas as entertainment, health care, transportation and automobiles, military and education. Research findings will be incorporated into graduate and undergraduate courses.
When two mobile devices are within communication range, there is an opportunity to exchange message packets, execute remote services and share content. Such an encounter is called as an opportunistic contact. A series of opportunistic contacts in time and space leads to an opportunistic networks. Opportunistic networks are good for applications that do not demand hard-real time requirements. Prior to 2008, significant work had been done in routing in opportunistic networks. However, the area distributed opportunistic computing was not investigated in earlier work. Through funding from NSF, we began investigations in this exciting area. In particular, we developed novel algorithms for routing, opportunistic service executions, service compositions, content sharing, content consistency, and a collusion resistant incentive compatible scheme for incentivizing participation in opportunistic networks. Each novel algorithm has been validated through extensive analytical and simulation studies on synthetic as well as real traces. Experimental studies have been carried out on prototype systems. Furthermore, the project work has resulted in new ideas and concepts for further research in this area. Four PhD students and one MS have been involved in completing various components of the project. Moreover, the research work in this projected has motivated the work of two ongoing PhD dissertations. Research outcomes have been incorporated into graduate and undergraduate curriculum. Two undergrduate students were trained in research work related to the project. In summary, this was a very successful and satisfying project. The research work will continue for several years in the future as the PIs have made several interesting discoveries and contributions towards applications mobile resources.
