Real-Time Spatiotemporal Analysis of Wireless User Activities

Yang Cai, Carnegie Mellon University
Rafael de M. Franco, Carnegie Mellon University
Xavier Boutonnier, Carnegie Mellon University


Abstract or Description

In this paper, we present a radio signal strength based positioning system that combines physical constraints and Bayesian model, which enable us to study wireless user group activities, such as wireless user footprints, periodicity and mobility. The user footprint maps show the spatiotemporal dynamics of wireless users in a building. By comparing the footprint maps in different buildings, we can discover interesting patterns. The periodicity transformation reveals the periodical user dynamics in a location, which provide clues for prediction. The mobility study shows the spatiotemporal anomaly patterns during social events in a building. Applications of this technology include: social network studies, privacy research, sensor networks, security systems, health care and intelligent houses.