A dynamic model for urban population density estimation using mobile phone location data

Based on the analysis of existing mobile phone positioning and detection methods, a dynamic distribution model has been proposed to detect urban population density based on the phone location data using an improved K-means clustering algorithm. The purpose of this real-time model analysis is to first detect the density and flow characteristics of urban population mobility according to the phone location data and then to dynamically monitor the real-time state and density of road traffic and urban population with low cost, all-weather detection capability and so on. The experimental study shows that better results can be achieved using our improved K-means clustering algorithm. We believe that such a model is helpful to the automatic detection of population density and road traffic.