Research on trade data encryption of tobacco enterprises based on adversarial neural network.

In order to overcome the problems of long encryption process, low data security and poor anti-attack rate of traditional tobacco enterprise trade data encryption methods, a trade data encryption of tobacco enterprises based on adversarial neural network is proposed. This method optimizes the traditional neural network by generating countermeasure network, so as to form adversarial neural network. In the adversarial neural network, the encryption processing of tobacco enterprise trade data is completed through data feature classification, design of tobacco enterprise trade data encryption protocol and data encryption channel. The experimental results show that the encryption process of this method takes between 6 s and 20 s, and the encryption efficiency is high, and this method can effectively scramble the original arrangement of the data, so as to effectively hide the effective information in the tobacco enterprise trade data, improve the anti-attack rate of the data, and effectively improve the security of the data. [ABSTRACT FROM AUTHOR]