Optimizing Smoothing Computation for Near Real-Time GPS Measurement Gap Filling in INS/GPS Systems

Eun-Hwan Shin and Naser El-Sheimy

Abstract:	The motivation for GPS/INS integration is to develop a navigation system that overcomes the shortcomings of each system using the complementary characteristics of both systems. In real-time environment, the INS/GPS integration Kalman filter works in prediction mode during GPS outages and the INS-predicted coordinates can be used to reset the position of the GPS antenna after loss of lock and signal re-acquisition. However, there are limitations in bridging by prediction. The bridging will result in correct ambiguities, only if all errors incurred during unaided INS operation remain below half a cycle. So, the length of the bridging interval is limited, depending upon the quality of the INS. For navigation-grade systems, typical intervals are 20-30 s. Since the stand-alone positioning accuracy of low cost inertial systems degrades rapidly with time, bridging by low cost inertial sensors can be applied for very short time or it is almost unavailable. The bridging intervals can be considerably extended if prediction is combined with smoothing after signal re-acquisition. Therefore, smoothing is clearly a post-mission method. To apply smoothing, the filtering solution has to be stored. Hence, the biggest problem, in using smoothing in near real-time, is the large storage requirement; especially the predicted covariance matrix which the takes largest portion because it has to be stored at high data rate. Therefore, the objective of the paper is to optimize smoothing computation for application areas that need accurate continuous trajectories in near real-time. The predicted covariance matrix is symmetric positive-definite and its variation is usually very smooth. This paper will introduce the piecewise-cubic Hermit polynomial approximation method to minimize the storage for the predicted covariance matrices. About 5.7 Mbytes of storage can be reduced to 115 Kbytes for 1 min GPS outages. The position errors caused by the approximation are discussed.
Published in:	Proceedings of the 15th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 2002) September 24 - 27, 2002 Oregon Convention Center Portland, OR
Pages:	1434 - 1441
Cite this article:	Shin, Eun-Hwan, El-Sheimy, Naser, "Optimizing Smoothing Computation for Near Real-Time GPS Measurement Gap Filling in INS/GPS Systems," Proceedings of the 15th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 2002), Portland, OR, September 2002, pp. 1434-1441.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In