Modulation Recognition Of Communication Signals Based On High Order Cumulants And Support Vector Machine

With the increasing development of science and computer technology, the modulation modes of wireless signal have gained great achievements. But the trouble is that in many fields, we could not obtain abundant knowledge about the modulated signals the receiver received, thus how to tell the difference about the modulation scheme and get the concrete contents of the signal is a problem remaining to be solved.The recognition of modulation modes can be attributed to a question with several unknown parameters. The recognition of modulation modes is always based on statistical pattern recognition and decision theory, we can make a study about the characteristic pattern of all kinds of modulated signals, and extract the characteristic values of the signals. The signal to be recognized is operated in the same way, and we can get the characteristic value of the specific signal and determine the modulation type of the signal through the decision theory. Thus the question is attributed to how to extract the characteristic values of the signal.The extraction of characteristic values is based on mathematical analysis of the modulated signal. The paper introduces a new method how to extract characteristic values through the instantaneous characters and high order cumulants.More than ten kinds of modulated signals including digital signals and analog signals as well can be well differentiated at a rower signal to noise ratio. The first step is to construct fourth order and sixth order cumulants, then the analog signals and part of the digital signals can be distinguished to two kinds. Then we can extract instantaneous characters of the two kinds of signals, thus we can achieve the recognition of single signal. The modulation mode is much simpler and the recognition type is more complex, and the recognition accuracy is much higher.LabVIEW is a virtual instrument development tool and is suitable for communicate system construct. The paper constructs a system of modulation recognition based on LabVIEW, and discusses the application of the method of characteristic value extraction in biomedical engineering.