Based on the threat to user privacy posed by intelligent vehicle networks, this paper proposes an intelligent vehicle wireless communication attack detection algorithm based on air interface traffic identification. The technique determines whether there is a threat to user privacy by monitoring the vehicle network communication signals, analyzing the signals and the type of business being carried,and combining the type of business with whether it has the attribute of steganography attack, which includes location tracking, wireless eavesdropping and video echo of eavesdropping. Using the software radio platform, the wireless air interface electrical signals are collected and transmitted to the signal analysis platform, and the neural network model based on the CNN model achieves the identification of the three types of traffic: audio, image and video, and the accuracy of the simulation experiment results is 90.34%.