Collaborative Joins in a Pervasive Computing Environment

We propose a collaborative query processing protocol based on the principles of Contract Nets. Locating and obtaining context-sensitive information in a mobile environment has always been a challenge. Pervasive computing environments only exacerbate this issue. In addition to operating on limited computing and battery resources, mobile devices cannot always rely on the help of a proxy-based wired infrastructure. Instead, a collaboration among peer mobile devices becomes necessary. Collaborating devices will be able to obtain data otherwise inaccessible due to the nature of the environment. Additionally, by reusing answers generated by peers, devices will be able to reduce their computation cost. In our model, a mobile device identifies information it needs via the help of user's profiles and constructs a query that requires one or more data streams. The device queries its vicinity and locates peer data sources. Our protocol then enables the device and the data sources to combine jointly their data streams in order to calculate a required answer. We show the effectiveness of our approach by performance evaluation of entities querying for data while moving in a city-like environment.