Evaluation Of Unintentional Modulation For Pulse Compression Signals Based On Spectrum Asymmetry

Specific emitter identification has attracted much attention of researchers in recent decades. This technology distinguishes different emitters by exploiting the unintentional modulation (UIM) contained in the signal. Heretofore, evaluation of UIM for pulse compression signals had not been properly solved, since previous evaluating results generally rely heavily on intentional modulation, and cannot solely and sufficiently describe the characteristics of UIM. In this study, the authors propose a novel feature extraction method which can naturally suppress intentional modulation, and meanwhile characterise the UIM. The proposed method utilises the facts that spectrum symmetry is an intrinsic and widespread property of modulated signals; and this symmetry will break when the UIM arises. By properly evaluating the spectrum asymmetry, the frequency-domain asymmetry distribution feature is derived. Experiments on real-world data demonstrate that the proposed feature outperforms conventional features in dealing with pulse compression signals. This study offers an effective way to extract UIM features from modulated signals, thus having broad applications to modern radar emitters.