Physical Layer Authentication Utilizing Beam Pattern Features in Millimeter-Wave MIMO Systems

The super-directive arrays in millimeter wave (mmWave) multiple-input multiple-output (MIMO) systems possess some intrinsic beam pattern features like amplitude and phase errors of element excitations. We first show experiment results to illustrate the uniqueness and stability properties of the amplitude and phase error features, and apply the Rician probability density function to efficiently approximate the distributions of these features under hardware impairments. By utilizing such beam pattern features and the hypothesis testing theory, we then develop a new physical layer authentication protocol for transmitter authentication in the mmWave MIMO communication systems. Based on the theories of statistical signal processing and composite hypothesis testing, a related theoretical framework is also developed for the analytical performance modeling of the new authentication protocol under both fixed and random beam pattern scenarios. Finally, we provide simulation and theoretical results to evaluate the reliability and security of the new authentication protocol against the identity-based spoofing attack.