A New Algorithm for GPS Integrity Monitoring

Barry E. Griffiths, Sandra L. Berning, and Ann T. Orlando

Abstract:	In order for GPS to be suitable for sole-means navigation in the U.S. National Air Space, a GPS receiver must be able to detect and reject satellite signals that would lead to unacceptably large position and velocity errors. Requirements for detection are stringent. For example, in the non-precision approach phase, errors exceeding 0.3 mn must be detected within 10 set with probability 0.999 or greater, while false alarm rates must be less than one per 15,000 two-minute intervals. The methods widely discussed for GPS RAIM are based on instantaneous detection of anomalous events, and are typically set to a fixed false-alarm rate. As a result, the probability of correct detection of a given error may not meet the stated RAlM requirements. At the same time, the instantaneous decision structure means that past data are “thrown away” with respect to RAIM decisions. These difficulties might be avoided by the use of a new FDIR architecture that has been designed to take advantage of past as well as current data. This approach combines redundant parallel Kalman filters with a neural net that is responsible for decision-making. The parallel filters make it possible to design a decision rule (implemented in the neural net) that incorporates both correct-detection and false-alarm considerations. In addition, this algorithm includes identification and reconfiguration as intrinsic elements.
Published in:	Proceedings of the 6th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1993) September 22 - 24, 1993 Salt Palace Convention Center Salt Lake City, UT
Pages:	539 - 543
Cite this article:	Griffiths, Barry E., Berning, Sandra L., Orlando, Ann T., "A New Algorithm for GPS Integrity Monitoring," Proceedings of the 6th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1993), Salt Lake City, UT, September 1993, pp. 539-543.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In