Dual-frequency Collaborative Positioning for Minimization of GNSS Errors in Urban Canyons

Abstract

Global Navigation Satellite Systems (GNSS) provide precise positioning under open-sky conditions, such as highways. However, in urban canyons, buildings block and reflect the signals, causing multipath positioning errors. Multi-frequency transmission and collaborative positioning are two technologies that have been proposed to reduce multipath errors. Still, the magnitude of their individual and combined advantage in reducing multipath errors is unknown. To fill this gap, we simulated dual-frequency collaborative positioning with four vehicles in an open-sky environment and in an urban environment. We compared two solution algorithms for position estimation: the Gauss-Newton solver (GN) and the extended Kalman filter (EKF). This paper presents the performance of these two algorithms under the previously mentioned assumptions. Furthermore, we show how the information from dual-frequency reception can be used to select the most relevant satellites. In the urban environment, the GN and the EKF using dual-frequency reception and collaborative positioning are the solutions with the smallest RMS positioning error (under 2.5 m), Additionally, in the simulated urban environment, dual-frequency reception contributes more to reducing multipath errors than collaborative positioning. As a consequence, when developing automotive positioning systems, multi-frequency reception and collaborative positioning should ideally be combined, but with higher priority on multi-frequency reception.