Localization Based on Lidar and GNSS for Connected Vehicles

Abstract

Localization using GNSS alone may cause relatively large errors due to problems such as multipath and satellite visibility especially in urban areas compared to open areas. Many techniques for improving position accuracy using ADAS sensors such as cameras and Lidar were studied. In case of camera, performance decreases due to external environment such as illuminance is problematic, whereas the algorithm using Lidar is relatively robust to external factors, having less performance degradation than camera. V2X is ideally suited to obtain the navigation information of the surrounding objects. In this paper, we propose a localization method using the information of the surrounding objects acquired with V2X which can know the position in the navigation coordinate system and the Lidar which can know the relative position in the coordinate system of the center of the user's vehicle. Simulation results obtain by using a real-time driving simulator are included in the paper to show the performance of the proposed methods.