Validation of Urban Multipath Models with Record and Playback Simulation

Neil Bennett, Mark Hunter, Guy Buesnel

Abstract: GNSS Positioning and navigation solutions can be severely degraded by signal reflections, or multipath, in urban environments. With GNSS satellites orbiting in Medium Earth Orbits (MEO) – GPS satellites have a period of 11 hours 58 minutes – the effects of multipath are subject to sudden variation due to changes in satellite geometry. When this is combined with the dynamic trajectories of moving objects such as vehicles, it can be seen that urban multipath is a complex and often troublesome problem. Use of 3D ray-tracing techniques to model urban propagation of GNSS signals is a way of recreating multipath based on geometry and the physical material properties, so that it is representative of the GNSS signals arriving at an antenna located at any point in the urban environment at any instant of time. The degree of representation depends to a large extent on the resolution of the 3D model used to derive the obstruction, reflection and diffraction processes for a reflection incident on the antenna, and a small change in any of the main properties used in the modelling is likely to result in a large change to the outcome. Verifying or validating the results of such modelling is extremely important – there are trade-offs between the degree of modelling required to generate sufficient realism, and the array of other errors that contribute to the precision and accuracy of navigation measurements affecting the performance of equipment in the real world. In this paper the authors validate a 3D multipath model of an urban setting in London, UK using a record and playback simulator (RPS) to record actual GNSS signals taken on a repeated section of route in the city and compare them to the simulated environment created by the 3D ray-tracing model of obscuration and multipath. The two main issues affecting GNSS reception in an urban environment are obscuration (which leads to limitations in signal availability) and multipath (which generates interference that causes fading of the signal and positioning errors). Whilst these mechanisms are easily identified in the 3D model, it is harder to differentiate in the playback of recorded signals. The limitations of using record and playback systems to recreate signals in a laboratory environment are highlighted, as is the statistical relevance of the data being recorded for replay. The authors repeat the recording on the route several times to capture temporal changes to the obscuration and multipath environment and then replay and analyse the signals, highlighting similarities and variances of the derived position data with that generated in response to the model. The authors discuss the trade-off between improvements to the synthetic environment used in the 3D model and the fidelity of recorded data as well as the need to repeat recordings in order to determine the needed statistical confidence. Finally, the authors make conclusions about how to use measurements taken with a record/playback system to make effective adjustments or modifications to the 3D modelling parameters at locations of interest.
Published in: Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021)
September 20 - 24, 2021
Union Station Hotel
St. Louis, Missouri
Pages: 257 - 277
Cite this article: Bennett, Neil, Hunter, Mark, Buesnel, Guy, "Validation of Urban Multipath Models with Record and Playback Simulation," Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021), St. Louis, Missouri, September 2021, pp. 257-277.
Presentation Slides: ION Members/Non-Members: 1 Download Credit
Sign In