Neural Networks Algorithms Prototyping to Mitigate GNSS Multipath for LEO Positioning Applications

W. Vigneau, O. Nouvel, M. Manzano-Jurado, C. Carrascosa Sanz, H. Abdulkader, D. Roviras, J.M. Juan, P. Holsters

Abstract:	Multipath is one of the most serious threats to the accuracy of GNSS measurements on a generic LEO spacecraft as GNSS antennas are surrounded by large reflecting structures (payload, solar arrays, antenna accommodation). Therefore, multipath will strongly degrade the performance in GNSS applications: Orbit/Attitude determination, clock synchronisation and relative positioning between spacecraft (docking, rendezvous, formation flying). Much effort has been put into solving the multipath problem in GNSS navigation, drawing a number of solutions based on different concepts (e.g. multipath mitigation by sitting, signaling techniques, mitigation techniques at antenna, receiver or software level). In the LEO context introduced above, this paper focuses on the use of innovative neural network (NN) algorithms both at receiver level (signal processing stage) and at observable level (GNSS measurements before position computation) to mitigate the impact of adverse multipath environments. The findings described in this paper are part of the work performed under a project initiated by the European Space Agency (ESA) (contract number 18824/05/NL/AG) which aims at designing, implementing and validating a fully functional neural network embedded into a GNSS navigation receiver prototype for multipath mitigation in a LEO satellite. This paper describes the selected neural network algorithms as well as their simulation environment when using real data and simulated data. It concludes on the most important findings when validating/testing the selected techniques and obtained performances with the neural networks, showing good improvements in correcting the carrier phase and code delay tracking errors due to multipath for LEO applications.
Published in:	Proceedings of the 19th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2006) September 26 - 29, 2006 Fort Worth Convention Center Fort Worth, TX
Pages:	1752 - 1762
Cite this article:	Vigneau, W., Nouvel, O., Manzano-Jurado, M., Sanz, C. Carrascosa, Abdulkader, H., Roviras, D., Juan, J.M., Holsters, P., "Neural Networks Algorithms Prototyping to Mitigate GNSS Multipath for LEO Positioning Applications," Proceedings of the 19th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2006), Fort Worth, TX, September 2006, pp. 1752-1762.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In