Support vector machine assisted genetic programming for MQAM classification

Automatic modulation classification is used to identify automatically the modulation type of an incoming signal with limited or no prior knowledge to it. Various classifier systems have been developed to solve this problem. However, for certain types of modulations such as 16 QAM and 64 QAM, the classification performance under noisy condition still needs to be improved. In this paper, we propose a new AMC scheme by combining genetic programing (GP) with support vector machine (SVM) for the classification of 16 QAM and 64 QAM signals. The benchmark result shows that SVM assisted GP can produce better accuracy than some other existing methods.