Submitted:
04 November 2026
Posted:
06 November 2026
You are already at the latest version
Abstract
Integrity of the Ground-Based Augmentation System (GBAS) relies on protection levels based on broadcast error models and on ground monitors comparing several reference receivers. Both were examined at an Istanbul reference station with real, independently labelled GPS code jumps. On 60,269 smoothed test epochs of a 23.8 km station pair, the ICAO Annex 10 fault-free vertical protection level bounded the error at every epoch when the error model was calibrated to the data tails; with standard models it was exceeded at 20 epochs for weighted least squares, but the error never exceeded the 10 m alert limit. With two to four simulated reference receivers, B-value monitoring detected almost all 20 m single-receiver jumps, so user-side exclusion and robust estimation mattered mainly for single-receiver stations and unmonitored faults. Adding Galileo reduced the median VDOP at Istanbul Airport from 1.48 to 1.10 and raised correct exclusion of two simultaneous faults injected into real smoothed code from 59% to 95%. Rare Galileo E1 code deviations of about one chip occurred at the reference station during signal acquisition; exclusion kept their vertical effect below 4 m, versus about 100 m without it. All findings held on 18 independent October days with frozen error models.
Keywords:
GBAS
; integrity
; protection level
; B-value monitoring
; Galileo
; fault detection and exclusion
; robust estimation
; multipath
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.