Deep fusion of vector tracking GNSS receivers and a 3D city model for robust positioning in urban canyons with NLOS signals

Abstract

In urban canyons, the GNSS satellite signals may travel an additional distance due to reflection and diffraction before reaching the receiver antenna. Where no direct path is available, this is called a non-line of-sight (NLOS) propagation and adds a positive bias to the geometric measured pseudorange. In this paper, we address the issue of GNSS positioning in harsh environments using constructively the NLOS signals. To exploit these biased signals, we compensate for the NLOS bias using a 3D GNSS simulation model of the environment. We use the 3D model as a priori information to characterize the additional NLOS bias. In this work, we propose a deep integration of the 3D model-based signal characterization into an advanced GNSS receiver architecture called vector tracking loops. The proposed architecture is a combination of a vector tracking receiver and a co-processor of external local information to correct the GNSS measurements. Obtained results with real GPS signals illustrates the effectiveness of this approach to improve the accuracy and reliability of the navigation solution in urban canyons.

Fisheye-Based Method for GPS Localization Improvement in Unknown Semi-Obstructed Areas

“…Authors of [ 10 , 11 ] use a recorded 3D model to detect NLOS satellites. They use a Kalman filter to smooth the localization estimation, taking into account LOS or NLOS status of the signals.…”