Model for Estimating Urban Mobility Based on the Records of User Activities in Public Mobile Networks

Urban mobility of a population is usually estimated within procedures that focus on specific domains, using limited datasets, indicators, and indices related to the targeted subsets of the urban population. This paper proposes a new approach to urban mobility estimation, based on the telecommunication activities within the public mobile telecommunication networks. The urban mobility indicators in this research are generated from a database of mobile phone users call data records and are integrated into the urban mobility index of the population based on the model defined through the adaptive neuro-fuzzy inference system (ANFIS). The following has been considered in the process: an initial fuzzy inference system, model learning, model quality control, limitations, errors, and deficiencies. The model is practically applied in the programming environment, on a set of real word data. The research results prove the following hypothesis set in this paper: the urban mobility of inhabitants in a specific timeframe, can be described with an urban mobility index based on the data on the recorded telecommunication activities of the public mobile communication network users.