Automatic Modulation Recognition of Communication Signals Based on Instantaneous Statistical Characteristics and SVM Classifier

Abstract

This paper proposes an optimization algorithm based on instantaneous statistical characteristics of modulated signals and the Support Vector Machine (SVM) classifier. The proposed algorithm employs a novel characteristic parameter, which can effectively distinguish MSK signal and 2PSK signal. Moreover, two traditional characteristic parameters are revised which can be more effective in distinguishing 2ASK and 4ASK, and dividing the signal set {2PSK, MSK, 2FSK, FM} into two categories. Numerical simulation results indicate that our proposed algorithm can correctly classify the communication signals of 2ASK, 4ASK, 2FSK, 2PSK, MSK, FM, LSB and USB with more than 95% success rate at the signal-to-noise (SNR) of 6 dB, which shows its superiority compared with other existing algorithms.