Automatic Recognition Of Commonly Used Digital Signal Modulations

Modulation recognition of communication signals is the research focus for many fieldssuch as the signal recognition and interference identification, and it is also the requiredfunction of software defined radio receiver. With the rapid development of communicationtechnology, the modulation recognition methods have become the research hotspot for thefield of software radio, and at the same time many new methods and ideas are also proposedin this area as well. Based on the existing analysis, the paper has studied the feature extractionbased on higher-order cumulants(HOC) first, which combines the decision tree and thesupport vector machine(SVM) in order to realize the recognition of the commonly useddigital signal modulation. And according to the advantages and disadvantages for the abovetwo methods, the paper presents a new method for the modulation recognition of digitalcommunication signals which is based on the higher order cumulants(HOC) and the algorithmof binary tree support vector machine. A set of feature parameters that extracted from HOCmethod are used for the construction of feature vector, and then the signals could berecognized by using binary tree support vector machine as the classifiers. The new methodcan fully perform the advantages of SVM for the two classification and it can also greatlyimprove the generalization ability of the algorithm.