Characterizations of OCS Kalman Filter Errors

Kenneth R. Brown Jr.

Abstract:	The central source of all GPS navigation data is the OCS (Operational Control System) Kalman filter. Based only on pseudo-range measurements bctwccn satcllitcs and monitor stations, it determines the orbital and clock state estimates and predictions that define the navigation data, which is periodically uploaded to satcllitcs in order to bc broadcast to users. Although the filter formulation used is very general, the geometry and physics of the underlying estimation problem imply unique and systematic properties that characterize solution statistics. These characteristics are specific to the GPS environment: . The Periodic Steady-State property n A near steady-state property in RAC coordinates . A near decoupling property in RAC coordinates l Predictable time constants, variances, etc. l Approximate dccorrclation of distinct satcllitc orbits 9 Dominance of clock noises in prediction performance l Insensitivity of prediction to measurement quality m Periodic responses to modeling errors This paper discusses these and other useful properties that set the GPS estimation environment apart. One goal is to describe the properties in terms helpful to designers of other systems that combine GPS information with data from other sources. Performance dcpcndencics and tuning questions arc also discussed, and a new result is presented on the response of the GPS Kalman filter to modeling errors.
Published in:	Proceedings of the 4th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1991) September 11 - 13, 1991 Albuquerque, NM
Pages:	149 - 158
Cite this article:	Brown, Kenneth R., Jr., "Characterizations of OCS Kalman Filter Errors," Proceedings of the 4th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1991), Albuquerque, NM, September 1991, pp. 149-158.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In