350 rub
Journal Neurocomputers №12 for 2012 г.
Article in number:
Features of the neural network of public key-cryptography
Authors:
N.I. Chervyakov, A.A. Yevdokimov, E.V. Maslennikova
Abstract:
In this article, the possibility of using a multilayer neural network in public key cryptography. Neural networks have two interesting properties: the ability to program the values of weighting factors and the presence of non-linear sigmoidal functions. In the paper the properties of multi-layer neural network, which can be useful in public key cryptography. Particular attention is given to change the values of weights of the neural network when the network is trained on the different sequences of training sample, the types of training sample at different frequencies and learning. Overall performance of the neural network approach for implementing the algorithm of public key cryptography. This follows from the analysis of the performance of the algorithm proposed in the paper. For the experiments considered neural network structure 64×64, which is used for the implementation of a cryptographic algorithm with a public key: public neural networks and sensitive neural networks. Private key are the weights of neural networks. Studies have shown that you can not get the final weights of the two open neural networks, since the number of combinations for testing will be very large because of the number of weights in the neural network.
Pages: 18-22
References
  1. Diffie, W. and Hellman, M. E., Multiuser Cryptographic Techniques Proceedings of AFIPS National Computer Conference. 1976. P. 109-112.
  2. Diffie, W. and Hellman, M. E., New Directions in Cryptography // IEEE Transactions on Information Theory. 1976. V. IT-22. № 6. P. 644-654.
  3. Rivest, R. L., Shamir, A. and Adleman, L., On Digital Signatures and Public Key Cryptosystems // Communications of the ACM. February 1978. V. 21. № 2. P. 120-126.
  4. Schneier, B., Applied Cryptography // Chap 19. John Wiley, 1996. P 476-480.
  5. Blake, I., Seroussi, G. and Smart, N., Elliptic Curves in Cryptography. London Mathematical Society Lecture Note Series. V. 256. Cambridge University Press. 2000.
  6. Bruss, D., Optimal Eavesdropping in Quantum Cryptography with Six States // Phys. Rev. Lett., 1998. V. 81. № 14. P. 3018-3021.
  7. Sarle W.S., Neural Network Faq, np://ftp.sas.com/pub/neural/FAQ.html, 2001.
  8. McCulloch, W. and Pitts,W., A Logical Calculus of the Ideas Immanent in Nervous Activity // Bulletin of Mathematical Biophysics, 1943. Chap 7. P. 115-133.