Physical Layer Authentication for UAV Communications Under Rayleigh and Rician Channels

Abstract

In this paper, aimed to against spoofing attack, we propose a novel physical layer authentication (PLA) framework for unmanned aerial vehicle (UAV) communication networks under Rayleigh and Rician channels. A new PLA metric, called authentication distance (AD), is defined by jointly considering the geographical locations, elevation angles, and channel randomness between a legitimate sensor and a malicious spoofer. For Rayleigh channel in dense urban environment, the closed-form expressions for the false alarm probability (FAP) and miss detection probability (MDP) are obtained by adopting method of convolution and integration by parts. Next, the PLA hypothesis test model with Rician channel is established in suburban environment where both the Rician factor and the path loss exponent are functions of UAV altitude. To proceed, the expressions for the FAP and MDP are derived based on the doubly non-central F distribution. In addition, MDP minimization solutions subject to certain FAP requirement are developed in both Rayleigh and Rician channels by optimizing the detection threshold and UAV altitude jointly. Simulation results show that our derived analytical expressions of FAP and MDP match the Monte Carlo simulations well. Moreover, simulation results also imply the effectiveness of the proposed PLA framework for UAV communication networks.