On large distributions for linear cryptanalysis

  • Authors:
  • Alexander Maximov

  • Affiliations:
  • LACS, University of Luxembourg, Luxembourg

  • Venue:
  • ICISC'07 Proceedings of the 10th international conference on Information security and cryptology
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Calculating the distribution of certain functions during the linear cryptanalysis of stream ciphers is a frequently encountered problem. Let a function N (or a noise variable) be expressed via k mutually independent and uniformly distributed n-bit random variables X1,X2, . . . , Xk. The possibility to construct its distribution depends on the form of the expression N, and sometimes it becomes a bottleneck of the cryptanalysis. In this paper we propose several new techniques to construct such distributions and widen the class of functions for which its distribution can efficiently be calculated.