Traffic Information Estimation Methods From Handover Events

Che-I Wu, Chi-Hua Chen, Bon-Yeh Lin, Chi-Chun Lo

Research output: Contribution to journalArticlepeer-review

5 Scopus citations


Fast growth of the economy and technology upgrades have led to improvements in the quality of traditional transport systems. As such, the use of intelligent transportation systems (ITS) has become more and more popular. The implementation and improvement of real-time traffic information systems are an important parts of ITS. Compared with other traditional methods, traffic information estimations from cellular network data are now readily available, more cost-effective, and easier to deploy and maintain. This study assumed that nonvehicle calls could be filtered out and vehicles could be tracked on road segments. A novel ITS model was proposed to indicate the relationship between call arrival rate and traffic density. Moreover, the vehicle speed and traffic flow were estimated by using cellular floating vehicle data (CFVD) and the proposed novel ITS model. In experiments, this study used a VISSIM traffic simulator and adopted the average call inter-arrival time and call holding time to simulate communication behavior on road segments. The estimated traffic information was compared with the simulated traffic information from stationary vehicle detectors (VD). The results indicated that the average accuracies for vehicle speed estimation, traffic flow estimation, and traffic density estimation in the congested flow case were 97.63, 89.72, and 90.45 %, respectively. Therefore, this approach was feasible to estimate traffic information for ITS improvement.
Original languageEnglish
Pages (from-to)656-664
Number of pages9
JournalJournal of Testing and Evaluation
Issue number1
StatePublished - Jan 2016


  • intelligent transportation system; cellular network; speed estimation; traffic flow estimation; traffic density estimation

Fingerprint Dive into the research topics of 'Traffic Information Estimation Methods From Handover Events'. Together they form a unique fingerprint.

Cite this