Impact of machine learning algorithms on analysis of stream ciphers

Stream ciphers are widely used for information security. The keystream produced by a cipher must be unpredictable. Attacks on stream ciphers typically exploit some underlying patterns existing in the keystream. The objective of this paper is to develop such an attack with the help of machine learning algorithms. The Linear Feedback Shift Register (LFSR) has been solved for several test cases using machine learning algorithms. We also study some variants of LFSR and Geffe Generator and propose a model for predicting the future bits of a keystream generator. The results for Geffe Generator using this model have been presented. However the approach did not yield encouraging results when confronted with the keystream generators of the eSTREAM project.

[1]  Manuel Blum,et al.  How to generate cryptographically strong sequences of pseudo random bits , 1982, 23rd Annual Symposium on Foundations of Computer Science (sfcs 1982).

[2]  Andrew Chi-Chih Yao,et al.  Theory and application of trapdoor functions , 1982, 23rd Annual Symposium on Foundations of Computer Science (sfcs 1982).

[3]  J. Ross Quinlan,et al.  Induction of Decision Trees , 1986, Machine Learning.

[4]  Alfred Menezes,et al.  Handbook of Applied Cryptography , 2018 .

[5]  Geoffrey I. Webb,et al.  Averaged One-Dependence Estimators: Preliminary Results , 2002, AusDM.

[6]  Nir Friedman,et al.  Bayesian Network Classifiers , 1997, Machine Learning.

[7]  Shehroz S. Khan,et al.  Analyzing a class of pseudo-random bit generator through inductive machine learning paradigm , 2006, Intell. Data Anal..

[8]  Nir Friedman,et al.  Building Classifiers Using Bayesian Networks , 1996, AAAI/IAAI, Vol. 2.

[9]  Jacob Ziv,et al.  An efficient universal prediction algorithm for unknown sources with limited training data , 2002, IEEE Trans. Inf. Theory.

[10]  James L. Massey,et al.  Shift-register synthesis and BCH decoding , 1969, IEEE Trans. Inf. Theory.

[11]  Geoffrey I. Webb,et al.  Lazy Bayesian Rules: A Lazy Semi-Naive Bayesian Learning Technique Competitive to Boosting Decision Trees , 1999, ICML.

[12]  A. Biryukov.  A New 128-bit Key Stream Cipher LEX , 2005 .

[13]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[14]  Hu Chuan-Gan,et al.  On The Shift Register Sequences , 2004 .

[15]  Mark A. Pitt,et al.  Advances in Minimum Description Length: Theory and Applications , 2005 .

[16]  Markus Dichtl,et al.  Document NES / DOC / SAG / WP 3 / 019 / 2 y About the NESSIE Submission \ Using the general next bit predictor like an evaluation criteria " , 2001 .

[17]  J.M. Sierra,et al.  Using classifiers to predict linear feedback shift registers , 2001, Proceedings IEEE 35th Annual 2001 International Carnahan Conference on Security Technology (Cat. No.01CH37186).

[18]  J. Ross Quinlan,et al.  Improved Use of Continuous Attributes in C4.5 , 1996, J. Artif. Intell. Res..

[19]  William Millan,et al.  Dragon: A Fast Word Based Stream Cipher , 2004, ICISC.

[20]  Philip Hawkes,et al.  Specification for NLSv2 , 2008, The eSTREAM Finalists.

[21]  Sang Joon Kim,et al.  A Mathematical Theory of Communication , 2006 .

[22]  Martin Boesgaard,et al.  The Stream Cipher Rabbit , 2005 .

[23]  T. Bayes,et al.  Studies in the History of Probability and Statistics: IX. Thomas Bayes's Essay Towards Solving a Problem in the Doctrine of Chances , 1958 .