Cell Towers as Urban Sensors: Understanding the Strengths and Limitations of Mobile Phone Location Data

Understanding urban dynamics and human mobility patterns not only benefits a wide range of real-world applications (e.g., business site selection, public transit planning), but also helps address many urgent issues caused by the rapid urbanization processes (e.g., population explosion, congestion, pollution). In the past few years, given the pervasive usage of mobile devices, call detail records collected by mobile network operators has been widely used in urban dynamics and human mobility studies. However, the derived knowledge might be strongly biased due to the uneven distribution of people’s phone communication activities in space and time.

This dissertation research applies different analytical methods to better understand human activity and urban environment, as well as their interactions, mainly based on a new type of data source: actively tracked mobile phone location data. In particular, this dissertation research achieves three main research objectives. First, this research develops visualization and analysis approaches to uncover hidden urban dynamics patterns from actively tracked mobile phone location data. Second, this research designs quantitative methods to evaluate the representativeness issue of call detail record data. Third, this research develops an appropriate approach to evaluate the performance of different types of tracking data in urban dynamics research.

The major contributions of this dissertation research include: 1) uncovering the dynamics of stay/move activities and distance decay effects, and the changing human mobility patterns based on several mobility indicators derived from actively tracked mobile phone location data; 2) taking the first step to evaluate the representativeness and effectiveness of call detail record and revealing its bias in human mobility research; and 3) extracting and comparing urban-level population movement patterns derived from three different types of tracking data as well as their pros and cons in urban population movement analysis.