SIGIL: A Novel GNSS/INS Integration for Challenging Environment

J-B. Lacambre, M-L. Duplaquet, J-M. Louge, Y. Paturel, R. Deurloo, F. Boon, K. Smolders, B. Bougard

Abstract:	Mobile mapping applications require high performance positioning and attitude determination systems, typically a cm-level accuracy for the position and a 0.01°-level accuracy for the attitude. Such systems usually use RTK GNSS receivers coupled with Inertial Navigation Systems (INS) to achieve such performance. Nevertheless, those requirements are very challenging in difficult environments, such as urban canyons, where buildings may hide or reflect satellite signals, impeding the GNSS receiver to accurately compute its position. RTK relies on the correct estimation of the carrier phase ambiguities to achieve its high accuracy. An inherent weakness of RTK however is that when correct ambiguities are inadvertently not found, only high precision is achieved without the required accuracy and the reported RTK position is biased. This phenomenon is commonly known as a wrong fix. As the typical INS Kalman filter makes Gaussian assumptions on the GNSS positions the wrong fixes will propagate the biased position into the INS solution - and worse - into the estimation of the IMU error components. Consequently, it is preferable not to receive a position from the GNSS receiver rather than to receive a position with a wrong associated standard deviation. GNSS-INS integration is typically implemented according to one of the following approaches: - The simplest integration method, so-called 'Loose Coupling' (LC), has the advantage of being very versatile. The GNSS receiver and the INS effectively operate as separate units. The GNSS receiver provides position and possibly velocity solutions to the INS to correct the integrated position, velocity and attitude of the INS using a Kalman filter. - In the more advanced ‘Tight Coupling’ (TC) approach the GNSS receiver provides pseudorange, carrier phase and Doppler measurements to the INS. This has the advantage of allowing better fault detection in the GNSS measurements and, in addition, allowing an INS solution even when there are insufficient satellites for a GNSS only solution. - Various algorithms can be found in the literature to extend this approach by using the corrected INS solution to aid the tracking loop of the GNSS receiver (deep coupling or ultra-tight coupling). Despite the obvious advantages of these more advanced coupling approaches, they are not as versatile as the loose coupling approach. The tight coupling approach requires the complete integration of INS and GNSS algorithms. In this paper, a new GNSS-INS integration algorithm developed jointly by a GNSS receiver manufacturer (Septentrio) and an INS manufacturer (iXBlue) is proposed. This algorithm, called Smart Close Coupling (SCC) allows the GNSS receiver to improve its overall performance. It remains however as versatile as a traditional loose coupling approach since both GNSS and INS Kalman filters still work conjointly, allowing both companies to focus on their core competence, and the algorithm to benefit from advances made by each of them as soon as they are available in their R&D scheme. The idea behind the SCC is to take advantage of the position variation computed by the INS in between the GNSS measurement updates. Since the position variation is precisely computed by the INS, the GNSS receiver can forego its own motion assumption and use the significantly more accurate INS state covariance prediction. This way, the problems described above are tackled: - Since the INS position variation has a centimeter level accuracy for a few seconds, the search space of the carrier phase ambiguity is significantly smaller, and more accurate. This results in: . a shorter time to obtain RTK fix . a more reliable and robust RTK fix position under adverse conditions - GNSS measurement outlier detection is significantly more efficient as the INS-based predicted state covariance leads to smaller residual covariance. This results in: . the GNSS receiver transmits a more accurate RTK covariance to the INS Kalman filter in the presence of measurement outliers. . the GNSS receiver keeps the fixed ambiguity solution for longer time. The paper will present results obtained on real data to show how the SCC has been able to enhance the integration of a Septentrio GNSS receiver and an iXBlue inertial navigation system. The paper will focus on the increase of RTK fix available, the reduction of false RTK fix, and the overall accuracy of the system in challenging environment. Various environments will be tested: foliage and canopy, standard urban environment, difficult urban environment (including tunnels and deep urban canyons).
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	2094 - 2101
Cite this article:	Lacambre, J-B., Duplaquet, M-L., Louge, J-M., Paturel, Y., Deurloo, R., Boon, F., Smolders, K., Bougard, B., "SIGIL: A Novel GNSS/INS Integration for Challenging Environment," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 2094-2101.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In