PUF-Enhanced Physical-Layer Key Generation for Secure Drone Communication with Untrusted Relays

Physical-layer key generation (PLKG) has attracted significant attention due to its lightweight and strong randomness, making it highly suitable for drones with energy and computational constraints. However, when drone communications rely on relay nodes, untrusted relay nodes may launch attacks such as eavesdropping, tampering, replay and man-in-the-middle, leading to key leakage. To address this problem, we propose a drone -physical unclonable functions (PUFs) - PLKG (DPPLKG) scheme that enhances resistance to untrusted relay attacks. In DPPLKG, legitimate drones are equipped with paired PUF hardware, and artificial noise is injected during the PLKG process to reduce the accuracy of untrusted relay channel estimation. The generated physical-layer key serves as the PUFs challenge, and the unique hardware response of PUFs generates the final session key, ensuring consistent and secure key generation among legitimate drones. After analysis, DPPLKG can effectively resist various threats such as relay eavesdropping, tampering, replay, and impersonation attacks. The simulation results show that DPPLKG outperforms traditional PLKG in terms of key generation rate, bit error rate, and key entropy. It can also resist Doppler frequency changes caused by drone mobility and hardware noise caused by temperature changes in PUF devices. In addition, the proposed scheme has an acceptable delay time, making it a practical and efficient solution for achieving drone secure communications in the presence of untrusted relays.