Physical layer authentication based on channel response tracking using Gaussian processes

Physical-layer authentication techniques exploit the unique properties of the wireless medium to enhance traditional higher-level authentication procedures. We propose to reduce the higher-level authentication overhead by using a state-of-the-art multi-target tracking technique based on Gaussian processes. The proposed technique has the additional advantage that it is capable of automatically learning the dynamics of the trusted user's channel response and the time-frequency fingerprint of intruders. Numerical simulations show very low intrusion rates, and an experimental validation using a wireless test bed with programmable radios demonstrates the technique's effectiveness.