Neural cryptography with queries

TL;DR

This paper introduces a neural cryptography protocol that uses state-dependent queries to enhance security, demonstrating through simulations that it reduces attack success probability without affecting synchronization speed.

Contribution

It extends neural cryptography protocols by incorporating queries based on network state, improving security against attackers while maintaining efficiency.

Findings

Queries restore security against cooperating attackers.

Success probability decreases with query-based approach.

Synchronization time remains unaffected.

Abstract

Neural cryptography is based on synchronization of tree parity machines by mutual learning. We extend previous key-exchange protocols by replacing random inputs with queries depending on the current state of the neural networks. The probability of a successful attack is calculated for different model parameters using numerical simulations. The results show that queries restore the security against cooperating attackers. The success probability can be reduced without increasing the average synchronization time.