Data encryption technology based on adversarial neural network

With the popularization of modern networks and the rapid development of communication technology, digital technology has been widely used in all walks of life. This paper deeply studies the security of the encryption algorithm of adversarial neural passwords. By analyzing the idea of adversarial neural networks and introducing the adversarial attack game theory of cryptography, an adversarial neural network encryption method based on chosen ciphertext attack is proposed. By analyzing the algorithm simulation results, it is found that the algorithm can learn a secure password generation algorithm without the need for artificial knowledge, and its speed has obvious advantages compared with algorithms such as DES, 3DES, and AES.