Classification of Encryption Algorithms Based on Ciphertext Using Pattern Recognition Techniques

Abstract

In digital era security in data communication is a challenging task. For secure data communication between the parties, encryption algorithms are being used. Recently many attacks have been reported for breaking ciphertext using various cryptanalytic techniques. Cryptanalysts are trying to break the cipher without knowing the key. Present day scenario an attacker may not know an encryption algorithm being used by communication entities. Identifying the algorithm itself is a challenging task. Once the encryption algorithm has identified, the cryptanalysts can analyzes the weakness of the encryption algorithm and be able to retrieve the plain text without knowledge of key. Here we present various pattern recognition techniques for identifying encryption algorithm which are used between two parties for communication of data using ciphertext. Thus, we consider the block cipher algorithms DES, IDEA, AES and RC operating in Electronic Code Book (ECB) mode. The classification techniques used are Support Vector Machine (SVM), Bagging (Ba), AdaBoostM1, Neural Network, Naïve Bayesian (NB), Instance Based Learning (IBL), Rotation Forest (RoFo) and Decision Trees to identify the right algorithm for the given ciphertext.