New data types and operations to support geo-streams

Abstract

The volume of real-time streaming data produced by geo-referenced sensors and sensor networks is staggeringly large and growing rapidly. Queries on these geo-streams often require tracking spatio-temporal extent (e.g. evolving region) continuously in real time. The notion of real-time monitoring and notification requires support from a database capable of tracking and querying dynamic and transient spatio-temporal events as well as static spatial objects and sending out real-time notifications. In this paper, we leverage the work in data type based spatio-temporal databases and propose new data types called STREAM and their abstract semantics to support geo-stream applications. New operations on STREAM data types are defined and illustrated by embedding them into SQL. © 2008 Springer-Verlag Berlin Heidelberg.