Time-frequency analysis-based deep interference classification for frequency hopping system

It is known that, interference classification plays an important role in protecting the authorized communication system and avoiding its performance degradation hostile environment. In this paper, problem for frequency hopping discussed. Considering possibility of presence multiple interferences system, order to fully extract effective features from received signals, linear bilinear transform based composite time-frequency analysis method adopted. Then spectrograms obtained are constructed as matching pairs input deep neural network classification. particular, Siamese used classifier, where paired into two sub-networks networks, these type The simulation results confirm that proposed algorithm can obtain higher accuracy than both traditional single representation approach AlexNet transfer learning or convolutional methods.