Interacting neural networks and cryptography

Wolfgang Kinzel, I. Kanter

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review


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.
Original languageAmerican English
Title of host publicationAdvances in Solid State Physics
EditorsProf. Bernhard Kramer
PublisherSpringer Berlin Heidelberg
ISBN (Print)978-3-540-45618-6
StatePublished - 2002

Publication series

NameAdvances in Solid State Physics


Dive into the research topics of 'Interacting neural networks and cryptography'. Together they form a unique fingerprint.

Cite this