C-Band Radio Signal Classification Method Based On Pcaand Wfcm

With the development of science and technology, radio is closely related with our Daily life. People’s demand for radio spectrum resources is increasing. Therefore, the limited spectrum resources become more valuable. Non-standard use of spectrum resources may lead to the waste of resources, affect the transmission of signals, hinder the development of communication industry and restrict the progress of social economy. In order to solve this problem, many experts and scholars devote themselves to this subject and a lot of theoretical achievements have been made. Recently, some uncertainty analysis tools, such as fuzzy set theory and neural network theory, are applied to abnormal signal analysis. In this article, principal component analysis (PCA) is applied to the feature extraction of radio signals so that the previous method for feature selection according to experts’experience can be improved. At the same time, the entropy weight method is draw into the fuzzy c-means algorithm to eliminate the asymmetry of the characteristics’contribution to classification. Experimental result shows that this algorithm is more effective and can determine the category of the abnormal signal quickly. The integrity of information can also be guaranteed to some extent. The following results have been achieved:1) Principal component analysis (PCA) approach is used to extract characteristic parameters of the four types signals of C band (radar, single carrier, jammer, single frequency point).2) The improved FCM method is presented and the weight of each characteristic parameter is calculated by the entropy weight method. C band signal classification method is proposed.3) C band signal classification algorithm is constructed.