Advanced GNSS Spoofing Mitigation: A Gaussian Mixture Model Approach Based on Carrier Phase Double Differences

Global Navigation Satellite Systems (GNSS) are increasingly targeted by spoofing attacks, which manipulate the received signals of GNSS receivers to mislead positioning, velocity, and timing (PVT) solutions. Identifying and excluding spoofed satellites from PVT calculations is a crucial step in anti-spoofing techniques to ensure accurate PVT. This requires accurately identifying spoofed signals from received signals. In this paper, we propose a novel approach that combines Gaussian Mixture Model (GMM) clustering with double difference carrier phase measurements to enhance the detection and identification of spoofed satellites. Our method first identifies groups of satellites affected by spoofing using GMM clustering. We then refine the detection by computing the intersection of these flagged satellite sets, allowing us to pinpoint individual satellites consistently exhibiting spoofing characteristics. Experimental results demonstrate that our approach provides a precise identification of spoofed satellites. This research contributes to the development of more resilient GNSS anti-spoofing mechanisms and improves the security of satellite-based navigation.