Joint Spoofing Detection Algorithm Based on Dual Control Charts and Robust Estimation

To address the issue that existing GNSS spoofing detection methods are not suitable for intermittent minor spoofing detection and spoofing duration identification, this paper theoretically analyzes the shortcomings of existing detection algorithms in terms of minor spoofing termination detection performance, and proposes comprehensively utilizing two types of control charts and robust estimation to detect the spoofing end moment, laying a foundation for spoofing duration identification and intermittent minor spoofing detection. The Shewhart control chart-based spoofing detection algorithm (M1) is proposed to achieve rapid spoofing termination detection, serving as one of the baseline algorithms for the joint algorithm. The strengths and weaknesses of the two baseline algorithms (M1 and existing EWMA control chart and robust estimation-based detection algorithm (M2)) in minor spoofing detection are analyzed. Under the robust estimation mechanism, a joint spoofing detection metric that can effectively indicate spoofing termination is constructed by combining their respective spoofing test statistics; then, anomaly detection on the joint detection metric is performed based on sample quantiles to identify the spoofing end moment. The experimental results under various typical abrupt spoofing and slowly varying spoofing scenarios demonstrate that the proposed joint spoofing detection algorithm based on dual control charts and robust estimation satisfies the spoofing alert time requirements specified by the International Civil Aviation Organization (ICAO) for the cruise phase. Compared with existing detection algorithms, the joint algorithm maintains excellent spoofing initiation detection performance while significantly improving both the speed and accuracy of spoofing termination detection. This effectively integrates the advantages of the two baseline algorithms and compensates for their individual limitations when operating independently. Upon timely and effective detection of the start and end moments of minor spoofing, it becomes possible to achieve spoofing duration identification and intermittent minor spoofing detection.