University of Maryland
Source Selection for Limited Bandwidth Environments

We consider the problem of remote data access for multiple users that share a limited-bandwidth link to a wide area network. One example of this occurs in mobile computing environments when multiple clients in a single cell connect to a base station. The clients send remote requests to the base station which accesses data over a fixed network and transmits it to the clients. There are several challenges in this type of environment. The first challenge is that accessing remote sources is slow and consumes bandwidth. A second challenge is that there may be alternate sources. By alternate sources we mean different servers that provide equivalent information that may vary with respect to recency and latency, and may have different access costs. An efficient algorithm for choosing which of these sources to access is needed. A third challenge is that a large number of users sharing limited bandwidth may request data from remote sources simultaneously. Different users may have different latency and recency requirements. It is important to develop solutions that satisfy as many users as possible while respecting money and bandwidth constraints. We study efficient source selection techniques that will find a set of sources well suited to a set of user requests and satisfy a set of constraints.

Recent papers

Please for a recent unpublished papers on Source Selection for limited bandwisth environments.

Return to CLIP home page