Abstract
The various aspects of neural cryptography with feedback are discussed. Neural cryptography is based on a competition between attractive and repulsive stochastic forces. It is noted that when a system of interacting neurons adjusts its couplings to a set of externally produced examples, this network is able to estimate the rule which produced the examples. It is observed that a network with feedback generates a pseudorandom bit sequence which can be used to encrypt and decrypt a secret message.
Original language | English |
---|---|
Article number | 046110 |
Pages (from-to) | 046110-1-046110-7 |
Journal | Physical Review E |
Volume | 69 |
Issue number | 4 2 |
State | Published - Apr 2004 |