According to the literature found related to the road trac surveillance using wireless sensors there are several prime strategies to recognize identify or detect traffic conditions. Detection with Magnetic Sensors, Video Sensors, Seismic Sensors, Acoustic (audio) Sensors are the main sensors used for those studies. Using the data collected from the sensors researches were analyzed the data using various ways. Then they came up with a mechanism for detect the vehicles/congestions. Most of the time they had used a threshold value for detection.
In the system which was designed in [1] comprised with two main components. Set of sensor nodes an the access point. Sensors are sensing the magnetic fields variation in the earth due to the vehicles and according to those measurements the system identifies the vehicle. For that the correlation between magnetic field and the vehicle level in the road was identified. Further more using a state machine and a threshold value system categorizes the identified vehicles. It is an enhancement for the system.
The solution suggested in the research [2] is to identify the trac status of the road using the GPS sensors. The GPS receiver records timestamp, altitude, speed, distance, heading, and coordinates once every 4 to 10 seconds. To find the correlation of traffic status and the sensing measurements this process was done for several days continuously and extracted some features of trac patterns on road segment. There are three parameters concern with the preliminary data. Those are speed, temporal data and spatial data. To obtain a trac pattern or feature each analysis should be concerned with those three. Without any of these will cause ambiguity in analysis. Finally using the analysis they found an association of trac pattern and the GPS data.
Refferences[1] Sinem Coleri, Sing Yiu Cheung and Pravin Varaiya, “Sensor Networks for Monitoring Trac”, University of California, Berkeley, August 5, 2004.
[2] Jungkeun Yoon, Brian Noble, Mingyan Liu, “Surface Street Trac Estimation”, Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, MI 48109-2122.