Characterizing and modeling the distance of mobile calls: A metropolitan case study

Abstract

During the past few years, with the democratization of open data, mobile network operators facilitated the sharing of mobile datasets. Such data are particularly interesting to better understand the way mobile users communicate, move and consummate services, but also to rethink the design or the evolution of current infrastructures with respects to new information and communication technologies and continuous increase of mobile usage. In this paper, we analyze a mobile dataset from a telecommunication provider in the city of Milan, Italy, and characterize mobile users' behavior. We show that, on the scale of the city, the number of short distance calls is massively higher than the number of long distance calls. Based on these observations, we propose a realistic yet simple model to describe the distance distribution of the calls in such a metropolitan environment. Finally, we focus on a specific area of the city, the soccer stadium area, and show that local events provoke regular deviation to the observed behaviors on the scale of the city. The macroscopic model we propose can be used to generate mobile datasets for any large city and thus support further simulations.