Estimating urban road traffic states using mobile network signaling data

It is intuitive that there is a causal relationship between human mobility and signaling events in mobile phone networks. Among these events, not only the initiation of calls and data sessions can be used in analyses, but also handovers between different locations that reflect mobility. In this work, we investigate if handovers can be used as a proxy metric for flows in the underlying road network, especially in urban environments. More precisely, we show that characteristic profiles of handovers within and between clusters of mobile network cells exist. We base these profiles on models from road traffic flow theory, and show that they can be used for traffic state estimation using floating-car data as ground truth. The presented model can be beneficial in areas with good mobile network coverage but low road traffic counting infrastructure, e.g. in developing countries, but also serve as an additional predictor for existing traffic state monitoring systems.