Towards Integrity Monitoring of GNSS Velocity Estimates in Urban Environment

verfasst von

Dennis Kulemann, Steffen Schön

Abstract

Precise and reliable velocity estimation is mandatory not only in filtering or sensor fusion, but also in autonomous navigation, like e.g., traffic light free cities. Global Navigation Satellite Systems (GNSS) are able to estimate an absolute velocity up to a mm/s accuracy, depending on the used observation type. However, in urban environments the performance of GNSS-based velocity estimation might be degraded significantly, which is due to signal discontinuities, multipath effects and cycle slips, which are affecting the carrier phase observations. In this paper, velocity estimation with Doppler and time-differenced carrier phase (TDCP) observations is analyzed in different open-sky as well as urban environments, using a least-squares adjustment (LSA) approach. Pre-fit residuals of GPS L1 signals w.r.t a reference trajectory are analyzed. In terms of integrity monitoring, no clear relationship between the ray tracing classes and residuals size could be found. Similarly, no direct correlation between elevation nor C/N

₀ was visible, which makes the derivation of adequate observation noise models difficult. In addition, even for the static open sky conditions, the pre-fit residuals are not normal distributed. Furthermore, the behavior of more robust modern signals is investigated, as well as the benefits of a highly stable external clock. It can be seen, that the clock does not have a direct influence on the observations and velocity estimated. The more robust signals, i.e. GPS L5 and Galileo E5, are beneficial in terms of velocity estimation performance, but suffer from a significantly reduced number of velocity availability. Typically, in urban environment without dedicated algorithms the horizontal velocity deviation was smaller than 1 m/s in 95 % of the epochs.

Organisationseinheit(en)

Institut für Erdmessung

Typ

Aufsatz in Konferenzband

Seiten

245-261

Anzahl der Seiten

Publikationsdatum

05.10.2023

Publikationsstatus

Veröffentlicht

Elektronische Version(en)

https://doi.org/10.33012/2023.19419 (Zugang: Geschlossen)

Details im Forschungsportal „Research@Leibniz University“

BibTeX

@inproceedings{9fde222d8b4643789df995fc68eadf8f,
title = "Towards Integrity Monitoring of GNSS Velocity Estimates in Urban Environment",
abstract = "Precise and reliable velocity estimation is mandatory not only in filtering or sensor fusion, but also in autonomous navigation, like e.g., traffic light free cities. Global Navigation Satellite Systems (GNSS) are able to estimate an absolute velocity up to a mm/s accuracy, depending on the used observation type. However, in urban environments the performance of GNSS-based velocity estimation might be degraded significantly, which is due to signal discontinuities, multipath effects and cycle slips, which are affecting the carrier phase observations. In this paper, velocity estimation with Doppler and time-differenced carrier phase (TDCP) observations is analyzed in different open-sky as well as urban environments, using a least-squares adjustment (LSA) approach. Pre-fit residuals of GPS L1 signals w.r.t a reference trajectory are analyzed. In terms of integrity monitoring, no clear relationship between the ray tracing classes and residuals size could be found. Similarly, no direct correlation between elevation nor C/N 0 was visible, which makes the derivation of adequate observation noise models difficult. In addition, even for the static open sky conditions, the pre-fit residuals are not normal distributed. Furthermore, the behavior of more robust modern signals is investigated, as well as the benefits of a highly stable external clock. It can be seen, that the clock does not have a direct influence on the observations and velocity estimated. The more robust signals, i.e. GPS L5 and Galileo E5, are beneficial in terms of velocity estimation performance, but suffer from a significantly reduced number of velocity availability. Typically, in urban environment without dedicated algorithms the horizontal velocity deviation was smaller than 1 m/s in 95 % of the epochs.",
author = "Dennis Kulemann and Steffen Sch{\"o}n",
note = "Funding Information: This work has been funded in the framework of the i.c.sens research training group funded by DFG under grant GRK2159.",
year = "2023",
month = oct,
day = "5",
doi = "10.33012/2023.19419",
language = "English",
pages = "245--261",
booktitle = "Proceedings of the 36th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2023)",
}