No abstract

Radio frequency fingerprinting (RFF) is used to uniquely identify individual emitters by exploiting the radio frequency characteristics, which are originated from transmitter imperfections. The theoretical performance is analyzed to evaluate the feasibility of RFF with channel noise and receiver imperfections. The distortions from oscillators and analog-to-digital converter (ADC) are considered, and a RFF signal model describing the discrepancies of transmitter imperfections, channel noise, and receiver distortions are constructed. A likelihood ratio test algorithm for RFF is proposed to analyze the theoretical performance. The upper and lower bound of theoretic performance is derived through the analysis of the likelihood ratio test statistic, which enlightens how to design a "just enough" receiver to meet the demands of RFF. The simulation results are in agreement with the theoretical evaluation.

No abstract

No abstract

No abstract