The feature extraction and classification for signals based on the S-transform

Abstract

The S-transform is an very efficient time-frequency representation method. According to the contrast of the traditional methods, the S-transform has a progressive anti-noise property and time-frequency resolution. Making use of the advantage of the S-transform, the characteristic variables of the modulation signal can be gotten, which can be used to distinguish the different type of modulation signals. The simulation experiment results show that the S-transform can express the instantaneous frequency of modulation signals clearly. Through extracting the linear feature, this paper goes deep into scooping out the new characteristic and extracts the character vectors, which can be used to set up a classification and recognition model. It can be used to correctly classify the different type of modulation signals.