UAV-Aided High-Accuracy Relative Localization of Ground Vehicles

Accurate position information is critical for the emerging applications of autonomous vehicles. Compared with absolute position information, relative positions are often more relevant for neighboring vehicles to perform joint tasks such as fleeting and overtaking. In this paper, we propose a relative localization scheme for ground vehicles aided by the unmanned aerial vehicles (UAVs). Each UAV makes ranging measurements with ground base stations and other UAVs in the round-trip time mode. The UAVs then broadcast their measurements to the ground vehicles, and each vehicle estimates its position based on the received data and the time difference of arrival (TDOA) of the signals from the UAVs. The Craḿer-Rao lower bound for the vehicles' absolute position errors is derived using the information inequality, based on which the relative position error bound is obtained through subspace projection. Finally, a two- stage localization algorithm for the vehicles is developed, which estimates the UAVs' positions followed by a weighted TDOA method. Numerical results show that the relative position errors between vehicles can reach the decimeter level even with meter- level absolute position errors.