Investigation of Zonal Sampling Strategies in Probe Vehicle-Based Traffic Monitoring Systems

Probe-based traffic monitoring systems offer a promising way to collect traffic data from large portions of the transportation network, but there remain a number of fundamental questions about how these systems should be designed and operated to ensure that the data obtained is accurate and reliable. Field deployments and simulation studies conducted to date have treated the entire roadway network as a single entity with uniform sampling characteristics. Recent research has shown that this “area wide” method has inherent limitations, and that sampling parameters may need to be adjusted based on localized conditions to ensure accurate speed estimates. Specifically, the level of granularity permitted by the system in defining sampling parameters must be explored. In this paper, the problem is examined by focusing on wireless location technology (WLT) systems. A strategy to determine sampling parameters at the level of individual wireless cells (referred to as “zones”) is discussed. By examining WLT-based monitoring from a zonal perspective, it is possible to set parameters, such as the number of vehicles tracked or the time between each position reading, based on the traffic and geometric conditions observed in a specific zone. This zonal strategy is tested using a simulated network, which has complex geometric and traffic flow conditions. The results indicate that the zonal sampling strategy overcomes many of the limitations of the area wide systems. Specifically, it helps ensure probe penetration across the entire network, making accurate speed estimates more likely in uncongested or low volume areas.