English

Interacting neural networks and cryptography

Disordered Systems and Neural Networks 2007-05-23 v1

Abstract

Two neural networks which are trained on their mutual output bits are analysed using methods of statistical physics. The exact solution of the dynamics of the two weight vectors shows a novel phenomenon: The networks synchronize to a state with identical time dependent weights. Extending the models to multilayer networks with discrete weights, it is shown how synchronization by mutual learning can be applied to secret key exchange over a public channel.

Keywords

Cite

@article{arxiv.cond-mat/0203011,
  title  = {Interacting neural networks and cryptography},
  author = {Wolfgang Kinzel and Ido Kanter},
  journal= {arXiv preprint arXiv:cond-mat/0203011},
  year   = {2007}
}

Comments

Invited talk for the meeting of the German Physical Society