Radio individual identification via stable communication signals based on subordinate component analysis

In this paper, the wireless communication stable signal is decomposed into three components, i.e., communication waveforms, individual modulations and system noise. From the eigenvalue spectrum of the observed signal, it is found that the primary components, i.e., the several largest eigenvalues, correspond to the communication waveforms while those minimal components are caused by the noise. Meanwhile, the subordinate components can act as the features for individual identification. Accordingly, a novel method is proposed based on subordinate component analysis and real measuring data are provided to demonstrate its effectiveness.