Jamming Pattern Recognition Using Spectrum Waterfall: A Deep Learning Method

Abstract

It is very important for communication equipment to realize the recognition and prediction of jamming signals in a short time. In the communication environment with variable jamming mode and noise, it is very difficult to achieve rapidly jamming recognition. To solve this problem, this paper constructs a spectrum waterfall containing white gaussian noise signals, designs a jamming pattern recognition method based on deep learning, and realizes rapid recognition of jamming patterns and important parameters. Simulation results show that the training process of this method has considerable accuracy and the test process has high recognition rate.