Improving Multi-GNSS Solutions with 3D Building Model and Tree Information

verfasst von
Anat Schaper, Qianwen Lin, Kim Sarah Janecki, Dennis Mußgnug, Max Leonard Heiken, Vimal Chawda, Lucy Ling Icking, Johannes Kröger, Steffen Schön

Positioning using multiple Global Navigation Satellite Systems (GNSS) offers a significant advantage, especially in dense urban areas. In particular in these areas, the combination of individual GNSS significantly increases the number of visible satellites and thus the GNSS availability for positioning.
However, signal interruptions, disturbances, and multipath effects due to buildings and trees still have an influence on positioning. To characterise this influence, we use a ray tracing algorithm to classify the observations into Line-of-Sight and Non-Line-of-Sight signals. For this purpose, a 3D building model of the city of Hanover and Open-Street-Map tree coordinates, the latter being supplemented by own measurements, are used.
In this paper, we focus on a multi-GNSS single point positioning algorithm that incorporates the environmental information. We perform adapted weighting models and compare the performance of these weightings with already established weighting schemes (considering satellite elevation, signal strength or unity weighting). In this way, we verify the effectiveness of these extended models. We show that incorporating environmental factors in the weighting models gives an improvement of up to 60% in terms of the 95% quantile of the 3D deviations to the ground truth. In fact, comparable accuracies to carrier-to-noise density dependent weighting can be achieved. Improving

Institut für Erdmessung

Details im Forschungsportal „Research@Leibniz University“