On the effectiveness of using state-of-the-art machine learning techniques to launch cryptographic distinguishing attacks

Authors:
Jung-Wei Chou;Shou-De Lin;Chen-Mou Cheng
Affiliations:
National Taiwan University, Taipei, Taiwan Roc;National Taiwan University, Taipei, Taiwan Roc;National Taiwan University, Taipei, Taiwan Roc
Venue:
Proceedings of the 5th ACM workshop on Security and artificial intelligence
Year:
2012

Citing 7
Cited 0

Use of a genetic algorithm in the cryptanalysis of simple substitution ciphers

Cryptologia
Mersenne twister: a 623-dimensionally equidistributed uniform pseudo-random number generator

ACM Transactions on Modeling and Computer Simulation (TOMACS) - Special issue on uniform random number generation
A Practical Attack on Broadcast RC4

FSE '01 Revised Papers from the 8th International Workshop on Fast Software Encryption
Learning generative visual models from few training examples: An incremental Bayesian approach tested on 101 object categories

Computer Vision and Image Understanding
LIBLINEAR: A Library for Large Linear Classification

The Journal of Machine Learning Research
Plaintext Recovery Attacks against SSH

SP '09 Proceedings of the 2009 30th IEEE Symposium on Security and Privacy
LIBSVM: A library for support vector machines

ACM Transactions on Intelligent Systems and Technology (TIST)

Quantified Score

Hi-index	0.00

Visualization

Abstract

Cryptographic distinguishing attacks, in which the attacker is able to extract enough "information" from an encrypted message to distinguish it from a piece of random data, allow for powerful cryptanalysis both in theory and in practice. In this paper, we report our experience of applying state-of-the-art machine learning techniques to launch cryptographic distinguishing attacks on several public datasets. We try several kinds of existing and new features on these datasets and find that the ciphers' "modes of operation" dominate the performance of classification tasks. When CBC mode is used with a random initial vector for each plaintext, the performance is extremely bad, while the performance for certain datasets is relatively good when ECB mode is used. We conclude that, in contrary to the findings of several existing works, the state-of-the-art machine learning techniques cannot extract useful information from ciphertexts produced by modern ciphers operating in a reasonably secure mode such as CBC, let alone distinguish them from random data.