Traffic Flow Estimation Models Using Cellular Phone Data

Cáceres, Noelia; Romero, Luis; Benítez, Francisco G.; del Castillo, J. M. · 2012 · OpenAlex-citations

DOI: 10.1109/tits.2012.2189006

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of estimating traffic volume across road networks without relying on expensive and limited-coverage fixed sensors, such as inductive loops. The authors propose using anonymous cellular phone call data as a low-cost, large-scale alternative for traffic monitoring. The core problem is that while cellular networks can track phone movement via handovers and location updates, the number of active calls does not linearly correlate with vehicle counts due to varying user calling behaviors and time-dependent factors. The research aims to develop models that infer vehicle volumes from inter-cell phone mobility data, calibrated against traditional detector measurements. The study utilizes data from a Spanish cellular operator and the Spanish Traffic Management Centre (DGT). The experimental design focuses on six locations outside the metropolitan area of Seville, selected to ensure boundaries were crossed primarily by vehicular traffic, excluding pedestrian or railway interference. This setup created twelve monitored boundaries (virtual traffic counters) corresponding to inter-cell boundaries. Data collected over 18 weekdays included anonymized outgoing call records—specifically handover events and consecutive calls made within a 15-minute window in neighboring cells—and corresponding vehicle counts from loop detectors. The analysis focused on the period between 08:00 and 21:00, when call activity was statistically significant. The dataset was split into calibration and testing sets to derive and validate the estimation models. The authors developed models that correlate the number of in-motion calls ($X$) with vehicle volume ($Y$) by incorporating time-dependent intensity factors. Since the relationship between calls and vehicles is not constant throughout the day, the models include a vehicle intensity factor ($f_j$) and a call intensity factor ($g_j$). These factors account for the temporal variation in traffic flow and the probability of users making calls during specific hours, derived from aggregated historical data. The results demonstrate that reasonable estimates of traffic volume can be achieved by comparing the model outputs with detector measurements. The study confirms that while phone flow exhibits similar peak patterns to vehicle flow (morning and afternoon rush hours), accurate estimation requires correcting for the non-linear and time-varying nature of calling behavior. The significance of this work lies in providing a scalable method for traffic monitoring that leverages existing cellular infrastructure. By treating cellular systems as virtual traffic counters, the approach offers a fast and economical way to derive volume data for roads lacking physical sensors. The inclusion of temporal correction factors allows the models to be applied to various boundaries without requiring site-specific recalibration for every location, provided the general calling and traffic patterns of the region are known. This method serves as a complementary solution to traditional fixed sensors, enabling broader coverage and more frequent updates on traffic demand.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• exposure measurement

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: observational prevalence