This project’s focus is on providing basic support for the class of tracking services, where there is a need for continuously monitoring of the current positions of a population of moving objects. Given a certain required accuracy with which the positions of mobile objects need to be known, the problem is how to most efficiently capture on a central server the positions of a large population of mobile objects within that accuracy. The techniques developed in the project are based on movement prediction, and it is an underlying assumption that exact locations of moving objects are not needed. Rather, we assume that tracking with a certain minimum accuracy is needed. Location-based games may need high accuracy, while a localized weather information service needs only low accuracy. Specifically, the project has designed, prototyped, and evaluated so-called point-based, vector-based, and segment-based tracking techniques that reduce the number of updates. Further, the project has designed, prototyped, and evaluated techniques that aim to improve the basic segment-based approach: Some of these are based on re-segmentation of the underlying road network. Others use anticipated routes and, possibly, acceleration profiles associated with these routes to obtain more efficient tracking.
|Effektiv start/slut dato||01/06/2003 → 31/12/2003|
- Electronics and Telecommunications Research Institute, South Korea