Integrity Monitoring in the KC-IO's GPS Embedded Module

Redge Bartholomew and Robert Hoech

Abstract:	This paper presents the integrity monitoring capribility of the CDNU/G’s GPS Embedded Module (CGEM) designed and built by the Collins Avionics & Communication Division, Rockwell International, for the KC-10 aircraft. CGEM is a repackaged version of the GPS module from the DoD Stnndard Embedded GPS/INS (EGI). The CDNU is the Navy Standard Control Display Navigation Unit: the G version has been modified to accept the insertion of the CGEM. The paper will describe the standard module’s current integrity check (EGI’s) -- a four measurement, least squares residual comparison -- and the hardware and software architectures that sustain it: it will describe the CGEM’s RAIM capability as an increment. This will include a discussion not only of the detection and isolation schemes, but as well it will describe the annunciation capabilities. For DoD receivers, initially (ca 15X%)), integrity monitoring referred to real-time tests of hardware components, the reporting of anomalies found in data input from the aircraft, anomalies found in measured data, anomalies found in computed data, and the reporting of software execution errors. The accuracy of the receiver’s output navigation data was doubtful if, for example, some memory chip failed, or if the GDOP exceeded some value, or if the age of the ionospheric delay measurement increased beyond some threshold. As time passed (and experience taught a more subtle appreciation for system errors), test and analysis engineers felt it appropriate to make at least some kind of rudimentary consistency check of the satellite measurements used by the receiver. Recurring but subtle clock failures on some of the early satellites provided the motivation: measurements from sub-mzuginal satellites would corrupt the receiver’s output data, producing an out of tolerance position accuracy, or an unacceptably noisy velocity. It became difficult to segregate the satellite segment, control segment, and user segment contributions to system error ( as revealed in the receiver’s output data). As a consequence the receiver absorbed a simple least squares residu,?l check: if one measurement residual was grossly inconsistent with the others, the receiver would ignore it, and log the error in the fault log sent out over the 1553 interface. If it found itself ignoring measurements from the sczlne satellite over a period of several consecutive seconds, it would drop it in favor of some alternate satellite. This, however, was the extent of the scheme. To satisfy new requirements imposed on some of the DoD’s commercial aircraft (per RTCA-DO-208 via TSO- C129), the CGEM uses a five-channel architecture with eight satellite measurements processed by a modified least squares residual algorithm tu detect and isolate satellite failures. As a consequence, when the RAIM function is active, some form of sequential uacking must be implemented in what previously had been a continuous tracking receiver. Potentially, this could degrade the receiver’s perform‘ance -- in particular this could be true with respect to j‘amming resistance and weak signal tracking. Channel management strategies reduce this impact to a marginal falloff from the bieeline performance, and remove it to an environment of extreme dyrmmics.
Published in:	Proceedings of the 51st Annual Meeting of The Institute of Navigation (1995) June 5 - 7, 1995 Antlers Doubletree Hotel Colorado Springs, CO
Pages:	551 - 559
Cite this article:	Bartholomew, Redge, Hoech, Robert, "Integrity Monitoring in the KC-IO's GPS Embedded Module," Proceedings of the 51st Annual Meeting of The Institute of Navigation (1995), Colorado Springs, CO, June 1995, pp. 551-559.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In