Positioning Security in 5G and Beyond: Model and Detection of Physical Layer Threats

Accurate localization is an essential functionality of 5G and beyond systems to enable location-based applications, such as autonomous vehicles and emergency response. Nevertheless, the integrity of location data faces challenges not only from unintentional sources of error, such as wireless propagation impairments and synchronization failures but also from malicious and intentional threats, such as spoofing attacks. This paper specifically addresses the risk to localization integrity posed by malicious attacks. It provides a framework for modeling security threats at the physical layer of cellular positioning, with a focus on 5G and beyond systems. Two detection methods are proposed to mitigate the impact of spoofing attacks, by leveraging cross-correlation analysis and Gaussian Mixture Models (GMMs). These methods leverage standard metrics already defined in the localization procedure, thus eliminating the need for additional signal processing steps. Simulation results in 3GPP standard-compliant scenarios demonstrate the effectiveness of these methods in significantly reducing the integrity risk under attack conditions, thus providing a foundation for developing resilient mobile network location-based services.