Cooperative Localization Based on GPS Correction and EKF in Urban Environment

Abstract

In recent research, the data fusion of multi-sensor in vehicle and global positioning system (GPS) data can improve the accuracy of vehicle localization, however, the GPS data received by the vehicle has certain errors. In order to further improve the final vehicle localization accuracy, this paper proposes a cooperative localization method by using corrected GPS data, vehicle-to-infrastructure (V2I), and vehicle-to-vehicle (V2V) data. Firstly, taking advantage of the Gaussian mixtures, the GPS distance error of vehicles is calculated by the received GPS error of base stations (BSs). Secondly, combining V2I, V2V communication technology and distance measurement method based on received signal strength indication (RSSI) to establish the vehicle GPS position correction geometric model of each vehicle, and obtain the corrected GPS position data. Finally, according to the yaw rate data collected by sensors, the different model of extended Kalman filter (EKF) technology is selected to fuse the data of each sensor and the corrected GPS data to obtain the final estimated position of the vehicle.