WSN Nodes for Real-Time Traffic Information Detection

With the rapid development of modern cities and fast increasing amount of vehicles, ITS has become crucial nowadays. Many elements of ITS are based on the real-time road condition represented mainly by the traffic volume, velocity and lane occupancy. Traditional traffic information detection systems can't meet the requirements of deployment convenience, detection accuracy and overall cost. WSN technology has potential wide prospect of applications. In this paper, the network architecture of a WSN-based wireless traffic information detection system, the hardware design of the detection nodes, and a series of detection algorithms are presented. Two kinds of algorithms are developed for traffic volume detection, which are a digital filter based algorithm (DiFiA) and a matched filter based algorithm (MaFiA) respectively. Two nodes for velocity detection are adopted and a velocity detection algorithm, namely VeDA is proposed. An efficient vehicle/non-vehicle classification algorithm is also designed to improve the accuracy of traffic volume, which is called ReSDiRA. And a lightweight synchronization method between two nodes is described for both velocity detection and vehicle/non-vehicle classification. On-road experiments are carried out to test and verify the accuracy and efficiency of the proposed methods, and the power consumption of the node is analytically evaluated. Experimental results indicate potential promises that the developed sensor node may be used in real-time traffic information detection as a new technique.