Managing trajectories of moving objects as data streams

The advent of modern monitoring applications, such as location-based services, presents several new challenges when dealing with continuously evolving spatiotemporal information. Frequent updates in the positions of moving objects, unexpected fluctuations in data volume and the requirement for real-time responses to continuous spatiotemporal queries indicate the limitations of traditional database systems. We attempt to model management of moving objects with the underlying assumption that their trajectories are essentially continuous, time- varying and possibly unbounded data streams. We propose a basic framework for managing trajectory streams, and suggest the introduction of enhanced constructs for advanced query capabilities. Our first experience with querying moving objects in two data stream prototype systems is promising for the feasibility and extensibility of this approach.