Distance Estimation Methods in Vehicular Application: An Experimental Study

Passive Entry and Passive Start (PEPS) system allows drivers to enter the vehicle and start the engine without activating key fob. With growing demand for location-based service and rapid development of Bluetooth Low Energy (BLE), BLE PEPS with localization service becomes a promising choice for next generation of PEPS. This paper studies the critical problem of localizing the vehicle for user based on BLE received signal strength (RSS) provided by preinstalled BLE Access Points (APs) inside vehicle. Multiple APs are employed to reduce the error caused by RSS fluctuation on distance estimation. Fusion algorithms including Extended Kalman Filter, Monte Carlo Localization and Iterative Trilateration are compared based on experiments in vehicle. Experimental results show that distance estimation based on Extended Kalman Filter gets high accuracy in both long distance and short distance scenario.