Intelligent Traffic Information Collecting System

With the rapid development of modern cities and fast increasing amount of vehicles, intelligent transportation system ( ITS) has become crucial nowadays. The major elements of ITS are from the real-time road condition represented mainly by the traffic volume, velocity and lane occupancy. To date, the traditional traffic information detection systems cannot meet the requirements of deployment convenience, detection accuracy and overall cost.

Hence, we develop a traffic information collecting system based on WSN. In this system, the sensor nodes can collect traffic information by detecting the magnetic disturbance caused by vehicles. And by using the technologies of signal processing and information aggregation, sensor node can extract information from original magnetic signals. Experimental results indicated potential promises that the developed sensor node can be used in real-time traffic information detection as a new technique.

System Architecture

The WSN-based traffic information collecting system is comprised of wireless traffic information collection nodes (TICoN), wireless data aggregation node (DAN) and remote traffic monitoring center (RTMC). TICoN nodes collect and process real-time information in a distributed and coordinated fashion. They transmit the results to the DAN node wirelessly. Then, the DAN node transfers the information to the RTMC or uses the information to control the traffic light. RTMC archives this traffic information and takes it as direction when making decisions.

Approaches/Experiments

A kind of traffic information collecting node (TICoN) is designed and produced based on the EZ210 (one product of the EZ-series WSN-platform of ICT WSN laboratory), which can collect the strength of magnetic field in real time. A series of novel traffic information detection algorithms are proposed. By processing the original magnetic signals, the traffic information, such as traffic volume, velocity and object types of bicycles or vehicles, can be extracted. We also do some work on the embedded implementation of signal processing method and energy saving mechanisms.

All the original traffic data we use is collected from real road and experiments. TICoN nodes are set by the road side. The experiments we carried out contain traffic volume detection, velocity detection and object classification. Experimental results showed that our system can provide traffic information with high accuracy in real-time.

Accomplishments

Yuhe Zhang presented a paper of “WSN Nodes for Real-time Traffic Information Detection” in the 1 st conference of Chinese Wireless Sensor Networks in Harbin ( China ) in August 2007 and this paper was awarded one of the Best Papers.

In the 14 th World Congress on Intelligent Transportation Systems (ITSWC2007), Yuhe Zhang presented a paper named “Design and Evaluation of a Wireless Sensor Network for Monitoring Traffic”, and this paper was awarded the First Prize for Best Student Paper in an Interactive Session.

In October 2007, Haiming Chen attended the 7 th International Symposium on Communications and Information Technologies (ISCIT'07) at Sydney, Australia, and in this symposium, he presented a paper named “Lightweight Signal Processing in Sensor Node for Real-time Traffic Monitoring”, which is selected as the best paper candidate.

Future Directions

In future, we plan to complete the integration of traffic monitoring system will be completed by the end of 2008. Till then, the system will be able to collect traffic information and report it to traffic monitoring center automatically. In this summer, we will focus on the efficient classification algorithms of different kinds of vehicles and do some work on the embedded implementation. In this fall and winter, we will research on multi-node information aggregation and prediction, to provide more accurate and more applicable information monitoring methods.