TL;DR
This paper introduces a novel "slope-based" method for assessing Receiver Autonomous Integrity Monitoring (RAIM) availability, which offers the same accuracy as the ideal but computationally intensive "ideal-PL" method. By using a characteristic slope and its derived threshold, this approach significantly reduces the computational burden on onboard systems, making it more practical for real-world applications.
What Problem Does This Paper Address?
Ensuring the integrity of Global Navigation Satellite System (GNSS) solutions is critical for safety-critical applications like civil aviation. Receiver Autonomous Integrity Monitoring (RAIM) is a key technology for achieving this at the user level. A crucial step in RAIM is assessing its availability before a navigation solution is used, ensuring it can detect potential faults and meet stringent integrity risk requirements. Traditional methods for this assessment rely on calculating a "protection level" (PL), which represents an upper bound on the position error. However, existing PL-based methods have drawbacks: the "classic" and "enhanced" PLs can lead to either overly optimistic or overly conservative availability assessments, potentially compromising safety or availability. The "ideal PL" method, while accurate, is computationally too demanding for onboard systems. This paper addresses the need for a RAIM availability assessment method that is both accurate and computationally efficient.
Methods
The paper proposes an "improved slope-based RAIM availability assessment method." The core idea is to use a "characteristic slope" as the primary metric for assessing RAIM availability, rather than the protection level. This characteristic slope quantifies the relationship between positioning error and pseudo-range residuals. The authors derive an "ideal slope threshold" that, when used, yields the same RAIM availability assessment results as the ideal PL-based method. This ideal slope threshold can be pre-calculated offline and then efficiently searched online during operation, significantly reducing the onboard computational load. The method involves calculating the characteristic slope for each visible satellite and comparing the maximum slope against this ideal threshold. If the maximum slope is below the threshold, RAIM is considered available; otherwise, it is unavailable.
Datasets
The paper uses simulated data for its evaluation. The simulation covers a global area (latitude 60° S~60° N, longitude 180° W~180° E) for a full day, with a 1° x 1° grid and 300-second time steps. The simulation uses a 32-satellite GPS constellation and assumes dual-frequency (L1/L5) ionosphere-free measurements. The measurement noise standard deviation is set according to the ARAIM interim report. No specific real-world datasets are used for training or evaluation; the focus is on theoretical analysis and simulation-based performance comparison.
Key Insights and Ideas
- Slope as a Superior Metric: The characteristic slope offers a more direct and computationally tractable way to assess RAIM availability compared to complex PL calculations.
- Offline Threshold Calculation: The ability to pre-calculate the ideal slope threshold offline is a major advantage, allowing for efficient online searching and reducing onboard computational demands.
- Equivalence to Ideal PL: The proposed slope-based method achieves the same accuracy as the ideal PL method, which is considered the gold standard for RAIM availability assessment.
- Preventing Optimistic/Conservative Assessments: The ideal slope threshold effectively prevents both optimistic (leading to potential HMI) and conservative (reducing availability) assessment errors, a common issue with classic and enhanced PL methods.
Key Results
The simulation results demonstrate that the proposed slope-based method, along with the ideal-PL-based method, successfully prevents both optimistic and conservative RAIM availability assessments. In contrast, the classic-PL-based method shows significant rates of optimistic assessments (up to 100% in some grid points), and the enhanced-PL-based method exhibits a high rate of conservative assessments (98.53% of grid points). The slope-based method achieves a 0% conservative assessment rate due to the small step size used in offline Tslope calculation.
Limitations and Future Work
The paper acknowledges that both the ideal PL and the ideal slope threshold require numerical solutions, which can still be computationally intensive. However, the slope-based method significantly reduces this burden by separating the iterative search from the online RAIM availability assessment. The authors suggest that future work should include a semi-physical simulation experiment using common onboard processors to verify the real-time performance of the slope-based method. Additionally, they plan to validate the RAIM availability assessment performance using massive actual measurement data.
Summary updated