Conservative Sensor Error Modeling Using a Modified Paired Overbound Method and its Application in Satellite-Based...

Area of Science:

• Navigation Systems
• Aerospace Engineering
• Signal Processing

Background:

• Conservative sensor error modeling is crucial for safety-critical systems like satellite-based augmentation systems (SBASs) and ground-based augmentation systems (GBASs).
• The widely used overbound method effectively addresses non-Gaussian errors and user position error estimation but suffers from excessive conservatism, reducing system availability.
• Existing methods struggle to balance integrity assurance with the need for accurate error modeling, particularly for non-Gaussian error distributions.

Purpose of the Study:

• To improve the availability of SBASs by proposing an enhanced sensor error modeling technique.
• To develop a method that more accurately models real error characteristics without compromising system integrity.
• To assess the robustness of the proposed method against abnormal data loss and large deviations.

Main Methods:

• Introduction of an improved paired overbound method for sensor error modeling.
• Relaxation of the requirement for the overbound function to strictly adhere to probability distribution function characteristics.
• Analysis of the algorithm's robustness using maximum deviation under defined integrity risks.

Main Results:

• The improved paired overbound method demonstrated a significant increase in system availability, achieving approximately 99% probability.
• The new method allows for more accurate modeling and measurement of actual error characteristics.
• Experimental results validated the effectiveness of the improved method in handling large deviations and maintaining integrity.

Conclusions:

• The proposed improved paired overbound method effectively enhances SBAS availability by enabling more accurate sensor error modeling.
• This approach offers a better trade-off between system integrity and availability compared to traditional overbound methods.
• The method shows promising robustness, particularly in scenarios with potential data anomalies or large error deviations.