Design of a Method for Automatic Digital Modulation Classification Based on Intelligent Clustering

This article proposes a method to classify digital modulation type from ASK, PSK and FSK groups, without prior information of received signal. This method has two steps and is called "clustering method". In the first step, the modulation group, and in the second step, modulation order is determined. Using clustering method, the least pre-processing, the most effective separation characteristics and appropriate classifiers was applied to each group and genetic algorithm has been used to improve system performance. The advantages of this method are high accuracy of identification, simplicity, blind identification and more modulation candidates compared to other similar methods.