The Application of RVM in GNSS Anti-Spoofing Field Based on the Hybrid Kernel Function

With the widespread application of global navigation satellite system (GNSS), spoofing attacks pose a threat to the security and reliability of GNSS. It is of great significance to design effective GNSS spoofing detection technology to ensure the security and reliability of GNSS system applications for receiver users. Traditional spoofing detection techniques generally only determine whether a spoofing attack has occurred by monitoring the feature changes of one or two data information in the receiver. However, some spoofing modes can cleverly make the monitored data very close to the real data, thus avoiding these detection methods and easily making them ineffective. In this study, a GNSS spoofing jamming detection method based on hybrid kernel relevance vector machine (RVM) is proposed. The improved signal quality monitoring (SQM) movement variance, carrier noise ratio movement variance, pseudo range Doppler consistency, pseudorange residual, Doppler frequency, clock offset and clock drift are used as detection characteristics. This technology can detect GNSS spoofing signals, effectively improving the safety and reliability of GNSS systems. The experimental results show that this technology has high detection accuracy and anti-interference ability and can effectively respond to various forms of spoofing attacks.