TY - GEN

T1 - Contributions to Velocity Integrity for Autonomous Navigation in Urban Environments

AU - Kulemann, Dennis

AU - Schön, Steffen

N1 - Publisher Copyright: © 2025 IEEE.

PY - 2025/4/28

Y1 - 2025/4/28

N2 - Precise and reliable velocity estimation is required for determining highly dynamic trajectories, as well as for autonomous navigation and collision avoidance. Using Doppler observations of Global Navigation Satellite Systems, the instantaneous and absolute velocity can be estimated at cm/s-level. Since the Doppler observations are less affected by multipath and non line-of-sight signal propagation effects than the code and carrier phase observations, they are especially useful for navigation in urban environments. To assess the trustworthiness of velocity information, the concept of integrity can be used. First approaches to bound the velocity error showed promising results even under challenging conditions, however up to now, no velocity alert limits have been defined.In this paper, we develop new alert limits for position, attitude and, especially for velocity, based on traffic statistics, the geometry of German roads and dimension of European vehicles. Due to the coupling of heading and velocity, the velocity alert limit is dependent on the speed itself, leading to a smaller influence of the velocity errors at higher speeds. In addition, we evaluate the capability of Kalman filter-derived protection levels to bound the position and velocity errors for a trajectory in an urban environment. The estimated position errors are bounded by the protection level in up to 70% of the epochs, while the velocity is bounded in more than 98% of the epochs. Furthermore, the velocity can still be estimated accurately, i.e. with an RMSE of 0.16 m/s even if the position errors are large with an RMSE of 2.93 m, which is due to its comparatively high robustness again multipath and non line-of-sight effects.

AB - Precise and reliable velocity estimation is required for determining highly dynamic trajectories, as well as for autonomous navigation and collision avoidance. Using Doppler observations of Global Navigation Satellite Systems, the instantaneous and absolute velocity can be estimated at cm/s-level. Since the Doppler observations are less affected by multipath and non line-of-sight signal propagation effects than the code and carrier phase observations, they are especially useful for navigation in urban environments. To assess the trustworthiness of velocity information, the concept of integrity can be used. First approaches to bound the velocity error showed promising results even under challenging conditions, however up to now, no velocity alert limits have been defined.In this paper, we develop new alert limits for position, attitude and, especially for velocity, based on traffic statistics, the geometry of German roads and dimension of European vehicles. Due to the coupling of heading and velocity, the velocity alert limit is dependent on the speed itself, leading to a smaller influence of the velocity errors at higher speeds. In addition, we evaluate the capability of Kalman filter-derived protection levels to bound the position and velocity errors for a trajectory in an urban environment. The estimated position errors are bounded by the protection level in up to 70% of the epochs, while the velocity is bounded in more than 98% of the epochs. Furthermore, the velocity can still be estimated accurately, i.e. with an RMSE of 0.16 m/s even if the position errors are large with an RMSE of 2.93 m, which is due to its comparatively high robustness again multipath and non line-of-sight effects.

KW - GNSS

KW - integrity

KW - urban navigation

KW - velocity alert limit

UR - http://www.scopus.com/inward/record.url?scp=105009207146&partnerID=8YFLogxK

U2 - 10.1109/plans61210.2025.11028388

DO - 10.1109/plans61210.2025.11028388

M3 - Conference contribution

SN - 979-8-3315-2318-3

SP - 1455

EP - 1466

BT - 2025 IEEE/ION Position, Location and Navigation Symposium (PLANS)

ER -