Multi-Domain Feature Fusion Based Radar Deception Jamming Recognition Method

Abstract

Considering the challenges posed by low recognition probability, high computational complexity, and difficult engineering implementation in radar deception jamming recognition under a low jamming-to-noise ratio (JNR), a method for recognizing radar deception jamming based on multi-domain feature fusion is proposed. This method analyzes the mathematical models of three types of deception jamming-intermittent sampling repeater jamming (ISRJ), dense false target jamming (DFTJ), and distance deception jamming (DDJ) -to extract two features: envelope fluctuation and waveform similarity in the time and frequency domains, respectively. Jamming classification and recognition are achieved using a multi-class support vector machine (multi-class SVM) model. Simulation results demonstrate that as JNR increases, there is a gradual improvement in the recognition rate which reaches 90%at ldB.