A Systematic Literature Review of Current Machine Learning Approaches for Detecting GNSS Spoofing Attacks
Susan Bertram, Luca Eisentraut, Ricardo Buettner · IEEE Access · 2025
Global Navigation Satellite System (GNSS) spoofing attacks become increasingly important due to their impact on navigation security as well as their significant risks to privacy and infrastructure in civilian and military domains. Those spoofing attacks manipulate GNSS signals to create false positions, which can lead to dangerous situations. In recent years, the use of machine learning approaches to detect such attacks has gained relevance. Although numerous technical studies on GNSS spoofing exist, there is no systematic review that applies a theory-driven, socio-technical perspective. This paper addresses that gap by structuring the existing literature along two dimensions: machine learning approaches and Bell/Whaley’s Theory of Deception. Most detection methods are based on supervised learning methods or neural networks and have mainly been tested through simulations. By introducing the Deception Theory into this technical domain, the review explores which types of deception GNSS spoofing involves and what defenses exist. The results show that most of the methods are aimed at ‘mimicking’ attacks.