Loading...

A new analysis of RC4: A data mining approach (J48)

HajSalehi Sichani, M ; Sharif University of Technology

601 Viewed
  1. Type of Document: Article
  2. Abstract:
  3. This paper combines the cryptanalysis of RC4 and Data mining algorithm. It analyzes RC4 by Data mining algorithm (J48) for the first time and discloses more vulnerabilities of RC4. The motivation for this paper is combining Artificial Intelligence and Machine learning with cryptography to decrypt cyphertext in the shortest possible time. This analysis shows that lots of numbers in RC4 during different permutations and substitutions do not change their positions and are fixed in their places. This means KSA and PRGA are bad shuffle algorithms. In this method, the information theory and Decision trees are used which are very powerful for solving hard problems and extracting information from data. The results of this Data mining approach could be used to improve the existing methods of breaking WEP (or other encryption algorithms) in less time with fewer packets
  4. Keywords:
  5. J48 (C4.5) ; RC4 ; WEP ; Data mining algorithm ; Encryption algorithms ; Existing method ; Extracting information ; Hard problems ; Machine-learning ; Algorithms ; Artificial intelligence ; Cryptography ; Decision trees ; Information theory ; Network security ; Data mining
  6. Source: SECRYPT 2009 - International Conference on Security and Cryptography, Proceedings, 7 July 2009 through 7 October 2009, Milan ; 2009 , Pages 213-218 ; 9789896740054 (ISBN)
  7. URL: http://www.scitepress.org/Papers/2009/21836/index.html