Research on Modulation Recognition of OFDM Signal Based on Hierarchical Iterative Support Vector Machine

Aiming at the problem that the modulation category of Orthogonal Frequency Division Multiplexing (OFDM) signal can’t be identified from the single-carrier signals and the Wavelet Packet Modulation (WPM) signals under the condition of low signal to noise ratio (SNR), a method of identifying OFDM signal based on hierarchical iterative support vector machine (SVM) is proposed. Firstly, based on the analysis of the high-order cumulant and double spectral envelope peak characteristics of wireless communication signals, three sets of two-dimensional feature vectors are constructed as input sample data of three-layer classifier. Secondly, a three-layer support vector machine classifier based on radial basis kernel function is designed. Thirdly, the hierarchical iterative method is used to train the classifier parameters, in which each layer is firstly trained and then the trained three-layer classifier is further trained by the overall iterative training to further optimize the parameters of the classifier. Finally, the classifier training process and OFDM signal recognition process are simulated and analyzed. The simulation results show that the proposed method can effectively identify OFDM signals and the recognition accuracy is significantly improved at low SNR compared with the traditional methods.