Date of Award

Summer 2011

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Computer Science

Committee Director

Michele C. Weigle

Committee Member

Stephan Olariu

Committee Member

Michael Fontaine

Committee Member

Hussein Abdel-Wahab

Committee Member

Kurt J. Maly

Abstract

Traffic management centers (TMCs) need high-quality data regarding the status of roadways for monitoring and delivering up-to-date traffic conditions to the traveling public. Currently this data is measured at static points on the roadway using technologies that have significant maintenance requirements. To obtain an accurate picture of traffic on any road section at any time requires a real-time probe of vehicles traveling in that section. We envision a near-term future where network communication devices are commonly included in new vehicles. These devices will allow vehicles to form vehicular networks allowing communication among themselves, other vehicles, and roadside units (RSUs) to improve driver safety, provide enhanced monitoring to TMCs, and deliver real-time traffic conditions to drivers.

In this dissertation, we contribute and develop a framework for dynamic trafficmonitoring (DTMon) using vehicular networks. We introduce RSUs called task organizers (TOs) that can communicate with equipped vehicles and with a TMC. These TOs can be programmed by the TMC to task vehicles with performing traffic measurements over various sections of the roadway. Measurement points for TOs, or virtual strips, can be changed dynamically, placed anywhere within several kilometers of the TO, and used to measure wide areas of the roadway network. This is a vast improvement over current technology.

We analyze the ability of a TO, or multiple TOs, to monitor high-quality traffic datain various traffic conditions (e.g., free flow traffic, transient flow traffic, traffic with congestion, etc.). We show that DTMon can accurately monitor speed and travel times in both free-flow and traffic with transient congestion. For some types of data, the percentage of equipped vehicles, or the market penetration rate, affects the quality of data gathered. Thus, we investigate methods for mitigating the effects of low penetration rate as well as low traffic density on data quality using DTMon. This includes studying the deployment of multiple TOs in a region and the use of oncoming traffic to help bridge gaps in connectivity.

We show that DTMon can have a large impact on traffic monitoring. Traffic engineers can take advantage of the programmability of TOs, giving them the ability to measure traffic at any point within several km of a TO. Most real-time traffic maps measure traffic at midpoint of roads between interchanges and the use of this framework would allow for virtual strips to be placed at various locations in between interchanges, providing fine-grained measurements to TMCs. In addition, the measurement points can be adjusted as traffic conditions change. An important application of this is end-of-queue management. Traffic engineers are very interested in deliver timely information to drivers approaching congestion endpoints to improve safety. We show the ability of DTMon in detecting the end of the queue during congestion.

Share

COinS