Spoofing Attack Detection in ADS-B Data Through Frequency Enhanced Patch Attention Mechanism
Linfeng Zhong, Hao Yang, Lei Zhang, Qinwei Zhong, Jin Huang, Fei Hu · 2024
In the face of growing air traffic complexity and the escalating sophistication of cybersecurity threats, the centrality of Automatic Dependent Surveillance-Broadcast (ADS-B) in air traffic control systems is undeniable, necessitating rigorous defense measures for its data due to the elevated security risks that can compromise the safety and efficiency of the airspace. To address these challenges, this paper proposed the Frequency Enhanced Patch Attention Network (FEPAN)-an advanced method built on the Transformer architecture and enriched with a novel frequency-based contextual enhancement specifically designed to expose anomalous patterns and activities within ADS-B data that conventional analyses might miss. Thoroughly tested against expansive ADS-B datasets featuring simulated deceptive maneuvers across diverse flight phases, our FEPAN methodology exhibits superior ability to pinpoint such intricacies with high accuracy, offering an indispensable tool for reinforcing the safeguarding of airspace operations.