A New Method for Sorting Radar Signal Based on Entropy Features

摘要

In order to figure out shortcomings of the existing method, such as the accuracy is not high and the method is sensitive to noise, we firstly Obtain the Wigner distribution space and bispectrum space for the received radar signal, then divide the space into subspace. Under the subspace distribution probability, the information entropy features of radar signals are constructed into two-dimensional eigenvectors, which reflects the energy distribution of the signal in these two spaces. Through the simulation results, it shows that extracted characteristic parameters have good separation and stability in the range of wide signal to noise ratio, verifying the validity of the method in this passage.