Translator Disclaimer
3 November 2010 Estimating the presence of people from aggregate cell phone network activity
Author Affiliations +
We present a methodology for the analysis of aggregate cell phone network activity that allows estimating the presence of people in geographical areas at the scale of the borough. The methodology is based on hourly counts of GSM network data handled by a major Dutch telecom operator over the metropolitan area of Amsterdam. Our methodology is structured in four steps. First we build an allocation function that indicates how to redistribute the by-antenna statistics to each borough. Second we analyze the trends of the temporal signatures of the boroughs and we select the clusters with minimum inter-cluster variation. Then we analyze the correlation of two demographic measures against data from the network during the selected temporal clusters. Finally we show that location updates are a good proxy for estimating the presence people, and that their variations over time reflect major trends of social dynamics. We also discuss how the analysis of GSM network activity can provide unprecedented insights on how crowds move in the city. We finally discuss our future work, whose outcomes we believe have the potential to solve real-world problems like crowd management, the support for special events, and the analysis of urban dynamics and urban policies.
© (2010) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Andrea Vaccari and Filippo L. Scognamiglio Pasini "Estimating the presence of people from aggregate cell phone network activity", Proc. SPIE 7840, Sixth International Symposium on Digital Earth: Models, Algorithms, and Virtual Reality, 78400J (3 November 2010);


Back to Top