Hilbert signal envelope-based multi-features methods for GNSS spoofing detection

The Global Navigation Satellite System (GNSS) provides precise Positioning, Navigation, and Timing (PNT) service for various industries such as military and national economy. However, civil GNSS is fragile and vulnerable to spoofing for the weak signal strength and the open architecture. To address the problem of limited performance of existing detection methods due to the strong concealment of spoofing, this paper proposes a novel scheme using Hilbert Signal Envelope-based multi-features. Firstly, multi-features are picked and calculated from the receiving signal, especially we propose a Tweaked Signal Quality Monitoring (TSQM) metric in this step. Secondly, an algorithm for extracting Hilbert Signal Envelope (HSE) is proposed for enhancing the features, which reduces features redundancy and pushes clean and spoofing samples away from each other. Thirdly, the enhanced features are sent to train and test the Machine Learning (ML) models. Finally, experimental validation is conducted on the Texas Spoofing Test Battery (TEXBAT) dataset and measurement data collected by low-cost Software-Defined Radio (SDR). The Area Under Curve (AUC) of the proposed method improves by 5.34 % and 3.07 % (reaching 97.34 % and 98.44 %) compared to the latest multi-parameters method on TEXBAT and measurement data, respectively, which provides an effective scheme for spoofing detection.