Tracing Human Movement Through Mobile Signaling Big Data: New Possibilities for Mobility Analysis

Abstract

Mobile signaling data provides extensive population coverage and high temporal resolution for large-scale mobility analysis, but privacy regulations have restricted access to individual trajectories, resulting in anonymized grid-based datasets. Although such data lack continuous movement paths and contain aggregation noise, they still capture meaningful collective mobility patterns. This study applies DBSCAN to identify dense spatial clusters and activity cores, followed by Random Forest analysis to assess the influence of spatial and contextual factors. Results show that blurred grid-based signaling data reveals stable spatial aggregations, temporal rhythms, and shifts in movement intensity. Rather than replacing trajectory-based data, this approach complements conventional mobility datasets by offering population-level insights into large-scale movement dynamics and regional mobility structures.

Keywords: Mobile Phone Signaling Data, DBSCAN, Random Forest, 5G Networks

DOI: 10.54941/ahfe1007874