An Improved Adversarial Neural Network Encryption Algorithm Against the Chosen-Cipher Text Attack (CCA)

Abstract

Password attacks are classified into four types: known-plain text attack, cipher text attack, chosen-plain text attack, and chosen-cipher text attack. The chosen-cipher text attack is the strongest attacking mode among the four attack modes. In order to resist the most aggressive chosen-cipher text attack mode, this paper proposes an improved adversarial neural network encryption method. We call it CCA-ANC encryption algorithm which changes the attack mode of the attacker to use the chosen-cipher text attack (CCA) method that can give the attacker a stronger cracking capability so as to force both encryption and decryption to use a better encryption system generating a more secure encryption algorithm.