Vehicle Cooperative Localization Based on UWB Technology in GNSS-Denied Environments

Abstract

The accurate vehicle positioning is critical for many intelligent transportation systems (ITSs) applications. However, Global Navigation Satellite Systems (GNSS) often face significant challenges in various demanding scenarios. This article explores the cooperative vehicle localization in GNSS-denied environments by utilizing a fusion of relative distance and angle-of-arrival (AoA) information gathered from vehicles and roadside ultrawideband (UWB) base stations with known coordinates. In dense traffic situations, UWB base stations may experience nonline-of-sight (NLOS) effects when localizing vehicles. To tackle this issue, we propose a strategy that begins by classifying vehicles into line-of-sight (LOS) and NLOS categories based on distance and AoA measurements, as well as discrepancies observed between two different localization methods. Initially, we estimate the positions of LOS vehicles, which are then used to determine the positions of NLOS vehicles through the distance information connecting LOS and NLOS vehicles. Once the preliminary positions of all vehicles are established, we employ an iterative optimization algorithm to refine these estimates, ultimately yielding more accurate localization outcomes. Simulation experiments indicate that the proposed method demonstrates the excellent localization performance in both static and dynamic positioning scenarios.