Identification of Multiple-Antenna Emitters Using Power Amplifier Nonlinearity

Power amplifier (PA) nonlinearity typically plays a crucial role in the application of specific emitter identification (SEI), due to its uniqueness at the radio frequency (RF) frontend for particular emitter. For a multiple-antenna emitter with multiple PAs inside, this paper first presents a PA parameter estimator based on a two-stage linear least square (LLS) method in the MIMO multipath communication scenario. Treated as the RF fingerprint, the extracted nonlinear coefficients are then used to classify the individual multipleantenna emitters based on the minimum error probability (MEP) criterion. In addition, we analyze the impacts of mismatch of PA model and rank-deficient condition of the LLS problem in depth. Finally, the proposed approach is validated to be effective through extensive numerical simulations.