Robust cooperative positioning using hybrid GNSS and wireless-ranging signals

Abstract Accurate vehicle localization is essential for ensuring operational safety in cooperative vehicular systems. Nevertheless, maintaining positioning accuracy in dynamic and signal-blocked environments is a critical technical challenge. While cooperative positioning methods exploit spatial correlations among vehicular nodes and swarm data fusion to enhance positioning reliability, their performance remains susceptible to outliers. Under the cooperative positioning framework, this study presents the robust centralized and distributed positioning (DP) using hybrid Global Navigation Satellite System (GNSS) and wireless-ranging measurements. The proposed centralized positioning (CP) establishes geometric constraints through inter-vehicle wireless ranging measurements while incorporating GNSS pseudo-range observations for swarm position estimation. In contrast, the DP implements a two-step estimation where the results from GNSS single point positioning are taken as a prior information to achieve refined solutions together with the ranging measurements. To address outlier susceptibility, both methods further integrate robust estimation with the IGG III weight function, enabling effective outlier detection and mitigation. To demonstrate the performance of robust cooperative positioning, semi-simulation experiments are conducted with and without robust estimation. The results reveal dm-level’s relative positioning accuracy is achieved using both cooperative positioning methods. The robust DP represents 56.3% accuracy improvement compared to robust CP while reducing computational time cost by 22.5% and achieves 48.8% enhancement in accuracy against DP without robust estimation. Besides, optimal hybrid GNSS/wireless-ranging performance requires wireless-ranging noise below 0.5 m, and comparative analyses under varying noise conditions confirm the robust DP sustains positioning accuracy better than 0.5 m across noise scenarios below 0.5 m.