A Wireless Signal Modulation Recognition Method Based on SVD Noise Reduction and CNN Classification
In order to solve the problem of low signal recognition accuracy caused by noise interference in low signal to noise ratio,a wireless signal modulation recognition method based on Singular Value Decomposition(SVD)noise reduction and Convolutional Neural Network(CNN)classification SVD-CNN is proposed.The signal denoising module based on SVD is proposed to denoise the input signal.The one-dimensional symbolic level CNN network architecture is designed to identify and classify signal features directly.For the modulation simulation data sets in Gaussian and Rayleigh channels,the proposed method is compared with typical modulation recognition methods such as CNN recognition method and Instantaneous Characteristic-Fully Connected Neural Network(IC-FCNN)recognition method.The experimental results show that the proposed method has higher recognition accuracy at low SNR,and the average recognition accuracy is improved by 38%～49%when the SNR is 0 dB.
NETL
NSTL
万方数据
