A UAV-Aided Physical Layer Authentication Based on Channel Characteristics and Geographical Locations

In this article, we present a mobile unmanned aerial vehicle (UAV) aided physical layer authentication (PLA) framework to differentiate between a legitimate transmitter and a malicious adversary based on the physical layer channel characteristics and geographical locations of different transmitters. For a single mobile UAV, we derive new explicit expressions for the probability density function (PDF) of signal-to-noise ratio (SNR) difference, false alarm rate (FAR), and miss detection rate (MDR). Then, we optimize key system parameters including the detection threshold and UAV movement to minimize the MDR subject to a given FAR constraint. Next, we extend the theoretical analysis to consider the double mobile UAVs scenario and derive the PDF of averaged SNR difference, FAR and MDR in closed-form. Monte Carlo simulations verify the accuracy of our derived expressions. Moreover, simulation results demonstrate the effectiveness of our SNR-based solution and highlight the advantages of double UAVs on minimizing the MDR over single UAV.

I. Introduction

Security and reliability are the major considerations for fifth-generation (5G) wireless communications due to the critical demand of protecting against harmful attacks, such as eavesdropping, malicious jamming and spoofing [1]. Unlike eavesdropping or malicious jamming attacks which are launched to overhear or disrupt the legitimate transmissions [2], spoofing attacks which aim to replicate the identity number of a legitimate transmitter to send harmful messages result in a significant attack on the integrity of the communication infrastructure [3]. For example, when legitimate actors report traffic jam positioning information to the server in an intelligent transport system, a malicious spoofer is able to intercept this important information and modify it to falsely generate other locations [4].