Feature Extraction, Modulation and Recognition of Mixed Signal Based on SVM

This paper introduces likelihood-based and feature-based modulation recognition methods. In the feature-based modulation simulation part, instantaneous feature, cyclic spectrum, high-order cumulants, and wavelet transform features are used as the entry point, and six digital signals including 2ASK, 4ASK, BPSK, QPSK, 2FSK and 4FSK are simulated, showing the difference of signals in multiple dimensions