Models for assessing accuracy and reliability of fibre-optic gyroscope-based navigation systems

Fibre-optic gyroscope (FOG) navigation systems are widely used in autonomous, aerospace, and terrestrial applications due to their high stability and independence from external navigation signals. While significant progress has been made in improving FOG performance at the sensor level, fewer studies have investigated the system-level impact of residual gyroscope errors, particularly during periods of GNSS unavailability. This study addresses that gap by analysing the accuracy and reliability of navigation in FOG-based inertial navigation systems under various operational modes: GNSS-aided, unaided, and intermittently aided. A mathematical model of navigation is implemented in a MATLAB/Simulink environment to assess error propagation. Performance is evaluated using practical metrics such as heading drift, position error accumulation, and recovery efficiency after GNSS signal restoration. Simulation results show that during unaided navigation, heading errors reach approximately 2–3 degrees over 600–900 seconds, while position errors grow to 40–80 meters during 300–600 seconds of GNSS outage. Upon GNSS reacquisition, error reduction of 80–90% is observed within 30–60 seconds. These results demonstrate that system-level modelling can significantly enhance navigation reliability without requiring modifications to the FOG hardware.