Project title: Extracting mobility paths from passive network location data

Supervisor name: Konstantina Papagiannaki, Rade Stanojevic

PhD enrolment: Y

Objectives: Analysis of the location data from the users of a cellular operator. With billions of people carrying cell phones non-stop 24/7, they leave valuable information on their location to the cellular providers in the form of call data records (CDRs) and network events (cell handovers, signaling, etc.). However such data has rather low resolution both spatially (a typical cell diameter ranges from few 100 meters in urban environments to several kms in rural areas) and temporally. Consequently, studying cellular network location data requires developing novel tools for refining such sparse information.

Tasks and methodology:

• Trajectory mining using the CDRs and/or network event data
• Validation of results against ground truth observed by detailed tracking of a small subset of users
• Understanding the limitations of cellular network data for studying mobility/location.

Results:

• A trajectory mining tool for queering individual user mobility pattern, inferring her: points of interests (PoI), regular routes (e. g. commute routes), means of transportation, etc.
• Inferring the common route patterns on a large population of cellular users.

This task will contribute to evaluate and quantify the measurements results to visualize the user experience in mobile networks based on location and the content that is consumed.

Dissemination: Scientific publication and communication internal to Telefonica

Planned secondment: UCL

The ESR will benefit from the training given by UCL on design of API for exposing measurements results.