Specific Emitter Identification of ADS-B Signal Based on Deep Residual Network

Abstract

A specific emitter identification method based on deep residual network (DRN) is proposed to solve the problem that the traditional expert features relying on artificial extraction are difficult to characterize the subtle differences of specific emitters.The DRN is used to complete the identification task. The in-phase component (I-way) and the quadrature component (Q-way) of signal sample data are inputted into the DRN for training. The performance of the proposed method is evaluated by experiments on datasets containing actual collected ADS-B signals from different planes. The results show that the proposed DRN model achieves high classification accuracy without manual feature selection. Furthermore, data augmentation on signal-to-noise ratio can further improve the performance of the model.

Keywords

• |ads-b|specific emitter identification|deep residual network|deep learning