WAVE-Based Short-Range Vehicular Localization Through Vector Match

Abstract

Precise and real-time position information are critical for an enforcement, autopilot, or navigation system at a toll station. However, global positioning system (GPS) has insufficient accuracy and low sampling rate. In this paper, a short-range vehicle localization algorithm is proposed to supplement the distance gap of the GPS. A normalized received signal strength indication (RSSI) sequence takes the place of RSSI at position evaluation in which problems of low signal noise ratio are alleviated. A position similarity degree algorithm is proposed to mitigate the low variability of a fingerprint, a seamless localization is achieved. A filter is designed to cut off the dataset of remote calibration points; the multipath effect is mitigated. Finally, the experiments verify the suitability of accuracy and execution time for enforcement and autopilot, and the insignificant effect of speed, traffic, and grid size.