Lightweight Physical-Layer Authentication via IQ Sample Super-Resolution in LEO Satellite Networks

Satellite communication (SatCom) networks recently require more robust authentication techniques than before, driven by the increasing demands of applications such as commercial communication services and military operations. A physical-layer authentication (PLA) technique that exploits inherent wireless communication features, such as channel fading characteristics, has emerged as a promising solution against potential threats to the networks. Recent studies have shown that machine learning-based PLA methods can effectively compensate for the unexpected threats to traditional security protocols. However, previous research has not sufficiently considered practical feasibility, which is crucial, especially in SatCom networks (e.g., operating time limit). This paper presents enhanced physical-layer authentication with super-resolution for satellites (ePASS), integrating convolutional neural network models and super-resolution techniques for rapid authentication in SatCom networks. We evaluate the practical feasibility and detection performance of the proposed ePASS against spoofing attacks through extensive simulations, compared with the existing schemes.