Travel Time Estimation Techniques for Traffic Information Systems Based on Intervehicle Communications

Abstract

As traffic congestion continues to grow on our roadway systems, trip travel times are becoming less consistent and less predictable. To help travelers plan trips better, traffic information systems are becoming increasingly valuable. These traffic information systems can be used off board (e.g., on the Internet before trip departure) or on board; several navigation systems exist that can provide real-time traffic information. Most traffic information systems are based on a centralized architecture focused on a traffic management center (TMC) that collects, processes, and disseminates traffic data. As an alternative approach, there has been recent interest in decentralized traffic information systems, that is, those that are based on intervehicle communications (IVC). As IVC-equipped vehicles travel the roadways, they can share information on network traffic conditions, and regional traffic information can soon be established. Decentralized systems avoid potential single-point failures that a TMC-based system might have and are capable of covering roadways that do not have embedded loop detectors. Several techniques are investigated on the way traffic information can be collected, processed, and shared in a decentralized IVC-based traffic information system. These techniques vary from simple blind averaging between all participating vehicles, to more sophisticated techniques using decay factors or filtered estimation. Simulation experiments have been carried out with the use of a unique integrated vehicle-traffic–network-communication tool to analyze the efficacy of decentralized IVC-based traffic information systems, on which each estimation technique has been rigorously evaluated.