Geomagnetic Sensor based Abnormal Parking Detection in Smart Roads

With the development of Internet of Things (IoT) and communication technology, abnormal parking detection based on geomagnetic sensors equipped with solar panels has become increasingly feasible and important. False parking detection is a main challenge affecting the widespread use of the aforementioned technology. Especially when ambient light change triggers changes in the current of solar panel, which may in turn cause local EM field changes and affect the detection of nearby geomagnetic sensor. To that end, we propose an abnormal parking detection scheme to distinguish between real parking and false detections. Specifically, we establish an equivalent model of the magnetic field around the solar panel, and design a method to compute the magnetic field at the periphery of the solar panel. Then, we analyze the difference between the magnetic field caused by the changes in ambient light and the magnetic field produced by vehicles, and extract features that distinguish abnormal parking from false detections. Based on the features, a state machine is designed to distinguish the false detection of light from the real parking based on the magnetic field waveform. Field tests show that the accuracy of the designed abnormal parking detection scheme can reach 97%.