Utilizing wireless positioning as a tracking data source

Abstract

Tracking data has become a valuable resource for establishing speed profiles for road networks, i.e., travel-time maps. While methods to derive travel time maps from GPS tracking data sources, such as floating car data (FCD), are available, the critical aspect in this process is to obtain amounts of data that fully cover all geographic areas of interest. In this work, we introduce Wireless Positioning Systems (WPS) based on 802.11 networks (WiFi), as an additional technology to extend the number of available tracking data sources. Featuring increased ubiquity but lower accuracy than GPS, this technology has the potential to produce travel time maps comparable to GPS data sources. Specifically, we adapt and apply readily available algorithms for (a) WPS (centroid and fingerprinting) to derive position estimates, and (b) map matching to derive travel times. Further, we introduce map matching as a means to improve WPS accuracy. We present an extensive experimental evaluation on real data comparing our approach to GPS-based techniques. We demonstrate that the exploitation of WPS tracking data sources is feasible with existing tools and techniques. © 2009 Springer Berlin Heidelberg.