Radar Active Jamming Signal Identification Method Based on Time-frequency Image Features

In this paper, the Choi-Williams method is used to describe five typical jamming signals, and the grayscale images of the time-frequency two-dimensional distribution of the jamming signals are obtained. The contour information is maximally enhanced by median filtering and image segmentation processing algorithms for the grayscale images. The effective time-frequency domain feature parameter set is extracted and the dimensionality reduction is done. SVM is used to identify typical radar active jamming signals and provide a priori information for radar anti-jamming. This paper verifies the adopted method by MATLAB simulation experiments and analyzes the performance and feasibility of the adopted method.