Cryptanalysis of SDES Using Genetic and Memetic Algorithms

In this paper, we exploit evolutionary algorithms for cryptanalysis and we focus on a chosen-plaintext attack model, in which the attacker is able to access both the ciphertext and the plaintext. The aim of this attack is to determine the decryption key for the Simplified Data Encryption Standard, so that other encrypted texts can be easily deciphered. We propose to extract the key using genetic and memetic algorithms (the latter being a hybrid of the evolutionary techniques and some refinement procedures). An extensive experimental study, coupled with the sensitivity analysis on method components and statistical tests, show the convergence capabilities of our approaches and prove they are very competitive compared with other state-of-the-art algorithms.