Individual Identification of OFDM Signal Radiation Source Based on Depth Adaptive Wavelet Network

Abstract

To solve the problem of information security hidden danger when using consumer electronic products for wireless communication, an individual identification method of OFDM signal radiation source based on a deep adaptive wavelet network is proposed. Firstly, a mathematical model is established to generate different fine features between subcarriers of OFDM signal transmitter. Then, I/Q channel signals are input into the network. In the process of network training, the lifting wavelet transform process is implemented adaptively to generate multiple signal segments with different time-frequency domain resolutions. Finally, multiple OFDM radiation sources are classified and identified by a classifier. The experiment results show that, under the condition of a signal-to-noise ratio of 30dB, the accuracy of end-to-end identification of five OFDM transmitters by using the deep adaptive wavelet network can reach 95.6%, and the anti-noise performance and parameter number of the network are better than the traditional convolutional neural network.