An individual emitter recognition method combining bispectrum with wavelet entropy

In order to research individual recognition of emitters with the same work and modulation mode, a new method combining bispectrum with wavelet entropy is proposed in this paper. The bispectrum and wavelet entropy are both used to extract the fingerprint features of radiation signals, and then, the neural network is used to complete the task of individual identification. Simulation results demonstrate that the recognition rate is above 85% with SNR of 5dB, achieving a better recognition performance than the conventional methods.