Black box Neuro-Identifier for RSA and AES Cryptosystems

Abstract

Black box identification system is a feature of neural networks; hence such system could be utilized for cryptanalysis approach to construct a neuro-identifier. The most widely used and hard to break cryptosystems on the web nowadays are AES and RSA cryptosystems. This paper reports some efforts to incorporate the black box techniques and neuro-identifier for cryptanalysis of these two systems. Theoretical study of a black box model has manifested considerable success in the identification of encryption/decryption keys search.