Authors: Mihaljević, Miodrag J. 
Fossorier, Marc P. C.
Imai, Hideki
Affiliations: Mathematical Institute of the Serbian Academy of Sciences and Arts 
Title: A Low-Complexity and High-Performance algorithm for the fast correlation attack
Journal: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume: 1978
First page: 196
Last page: 212
Issue Date: 1-Jan-2001
Rank: M21
ISBN: 978-3-540-44706-1
ISSN: 0302-9743
DOI: 10.1007/3-540-44706-7_14
Abstract: 
An algorithm for cryptanalysis of certain keystream gene- rators is proposed. The developed algorithm has the following two ad- vantages over other reported ones: (i) it is more powerful and (ii) it provides a high-speed software implementation, as well as a simple hard- ware one, suitable for high parallel architectures. The novel algorithm is a method for the fast correlation attack with significantly better performance than other reported methods, assuming a lower complexity and the same inputs. The algorithm is based on decoding procedures of the corresponding binary block code with novel constructions of the parity- checks, and the following two decoding approaches are employed: the a posterior probability based threshold decoding and the belief propagation based bit-flipping iterative decoding. These decoding procedures offer good trade-offs between the required sample length, overall complexity and performance. The novel algorithm is compared with recently proposed improved fast correlation attacks based on convolutional codes and turbo decoding. The underlying principles, performance and complexity are compared, and the gain obtained with the novel approach is pointed out.
Keywords: Decoding | Fast correlation attack | Keystream generators | Linear feedback shift registers | Stream ciphers
Publisher: Springer Link

Show full item record

SCOPUSTM   
Citations

38
checked on Dec 6, 2022

Page view(s)

47
checked on Dec 6, 2022

Google ScholarTM

Check

Altmetric

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.