Blind Modulation Classification via Combined Machine Learning and Signal Feature Extraction

In this study, an algorithm to blind and automatic modulation classification has been proposed. It investigates combined machine leaning and signal feature extraction in order to recognize diverse range of modulation in low signal to noise ratio (SNR). The presented algorithm includes four steps. First, it analyzes spectrum to branching modulated signal based on regular and irregular spectrum character. Second, a nonlinear soft margin support vector (NS SVM) problem is applied to received signal, and its symbols are classified as correct or incorrect (support vectors) symbols. The NS SVM employment leads to alleviation in physical layer noise effect on modulated signal. Then, a k-center clustering can find center of each class. Finally, estimation of scatter diagram is correlated with pre-saved ideal scatter diagram of modulations in correlation function which is classification outcome. For more evaluation, success rate, performance, and complexity in comparison to many published methods are provided. The simulation proves that the proposed algorithm can classify the modulated signal in low SNR. For example, it can recognize 4-QAM in SNR=-4.2 dB, and 4-FSK in SNR=2.1 dB with %99 success rate. Moreover, due to using of kernel function in dual problem of NS SVM and feature base function, the proposed algorithm has low complexity and simple implementation in practical issues.