Securing ADS-B Communication Systems Against Location Spoofing Attacks

Automatic Dependent Surveillance-Broadcast (ADS-B) is an integral part of the Next Generation Air Transport System, providing a more efficient and safer transport infrastructure by monitoring and managing congested airspace through advanced satellite surveillance. It is designed to replace traditional radar-based communication with a reliable system that periodically transmits real-time aircraft positions to air traffic control and nearby planes. However, despite being a relatively new standard, it lacks security measures. The absence of encryption and authentication leaves it vulnerable to various attacks, including message injection, modification, deletion, jamming, and eavesdropping. While eavesdropping may not have immediate consequences, attacks involving message modification and injection can result in significant damage, leading to severe traffic disruptions and potential aircraft collisions. Common cryptographic solutions offer enhanced security but face practical challenges due to their incompatibility with existing infrastructure, necessitating modifications to the current ADS-B protocol. In this thesis, we introduce an approach to identify location spoofing attacks within ADS-B communication that does not require protocol modification. Our method leverages the physical layer attributes of ADS-B signals, such as magnitude, phase, power and geospatial hexagonal tessellation. We employed two distinct anomaly detection models: One-class SVM and the Isolation Forest algorithm. The performance assessment was conducted using both the conventional scikit-learn library and the PyOD outlier detection library. Our experimental results consistently demonstrate that the Isolation Forest algorithm, implemented in both libraries performs better than OCSVM and achieves an average F1-score of 93% within geospatial hexagonal cells. However, we noticed that the selected physical layer properties may not provide a comprehensive representation of aircraft positions, as these characteristics can depend on the specific hardware of the aircraft, potentially leading to misclassification.