Integration of CHIMERA GPS Anti-Spoofing Methods with Inertial Navigation Systems
Abstract
Tightly-coupled INS/GPS navigation algorithms have been developed to use GPS data that have latency due to authentication using the CHIMERA system. Such an algorithm is needed to produce non-delayed navigation results that are guaranteed, via CHIMERA, not to have been spoofed. The developed algorithms are modified versions of the standard tightly-coupled INS/GPS method in which the INS data define a Kalman filters position, velocity, and attitude dynamics via the technique known as model replacement. The Kalman filters measurements are the GPS observables pseudo range and carrier Doppler shift. The CHIMERA authentication latency is handled by running multiple filters or partial filters. One filter uses only the authenticated data up to a past authentication epoch and INS data that run up to the present. One or more other filters process unauthenticated data in preparation for future use. Some filters consider the possibility that authentication times are staggered in hopes of improving performance. The methods have been evaluated using real data collected from a general-aviation aircraft and using simulations of what INS of varying quality would have output had they been taken on the same flight. Results for the fast CHIMERA channel, which entails latencies of only 2 seconds, have accuracies commensurate with zero-latency standalone GPS regardless of the IMU quality, i.e., 2 m RMS position errors. Results for the slow CHIMERA channel are problematic. The slow channel entails latencies of 180 sec. For non-staggered slow channel CHIMERA with a navigation-grade INS, RMS position error is 9.5 m, and peak position error is more than 20 m. Lower-grade INS produce much larger RMS position errors, on the order of 100 to 2400 m, when paired with slow-channel CHIMERA. Even the navigation-grade INS performance with slow CHIMERA is significantly worse than zero-latency stand-alone GPS.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 31, 2021
- Accession Number
- AD1153887
Entities
People
- Mark L. Psiaki