Using DTMon to monitor transient flow traffic

Abstract

We evaluate the performance of the DTMon dynamic traffic monitoring system to measure travel times and speeds in transient flow traffic caused by non-recurring congestion. DTMon uses vehicular networks and roadside infrastructure to collect data from passing vehicles. We show DTMon's ability to gather high-quality real-time traffic data such as travel time and speed. These metrics can be used to detect transitions in traffic flow (e.g., caused by congestion) especially where accurate flow rate information is not available. We evaluate the accuracy and latency of DTMon in providing traffic measurements using two different methods of message delivery. We show the advantages of using dynamically-defined measurement points for monitoring transient flow traffic. We compare DTMon with currently in-use probe-based systems (e.g., AVL) and fixed-point sensors and detectors (e.g., ILD).