An Artificial Neural Network Technique of Modern Cryptography

Abstract

An enormous volume of data transmitted through the internet in every second nowadays, so data protection is still a big challenge as communication channels are mostly public. Several approaches are there for protection of data when transmitted through the network channel, cryptography is one of them. Cryptography is a technique to hide original and confidential information from intruder. Considering information security principles i.e. confidentiality, integrity, and availability (CIA-triad) an efficient, portable, dynamic, and simple, artificial neural network (ANN) based cryptography algorithm have been presented in this paper. A multi-substitution tree parity machine (TPM) approach is used here to generate dynamic single digit secret key. The secret key created by the tree parity machine will be different in each phase of execution. This single digit dynamic secret key is used in the proposed encryption and decryption techniques. The complexity and execution time taken by the proposed algorithm is improved comparing to the existing algorithms. The proposed algorithm provides higher reliability and robustness in terms of security because of it dynamic nature.