Spectral Component Analysis For Hardware Characterization In RF Applications

This contribution introduces a new method for the extraction of hardware-induced fingerprint features, referred to as spectral component analysis (SCA). SCA is based on the characterization of a signals state space in time-frequency domain. Accordingly, a novel partial differential equation (PDE) has been formulated, which establishes a mathematical framework for modeling the complex interactions among hardware components and their impact on the signal. This allows to extract a feature distribution in order to distinguish between hardware characteristics. Next to the extraction theory, digital hardware simulations are performed based on filter and IQ-offset properties. To proove the applicability in radio frequency (RF) applications, radar measurements in X-band are carried out and discussed in detail.