Abstract
Neural cryptography
Highlights
Information security (IS) is a set of processes, methodologies and procedures for protecting the information and information systems from unauthorized access, use, modification, or destruction
Privacy ensures that communication between two parties remains secret; authentication is required to ensure that information is exchanged with a legitimate party; data integrity refers to overall completeness, accuracy and consistency of data while nonrepudiation assures that parties in communication cannot deny the authenticity of their signatures on a document or sending a message they generated (Rouse, 2015)
The parity machine (PM) is a neural network applied in cryptography to generate a secret key
Summary
Information security (IS) is a set of processes, methodologies and procedures for protecting the information and information systems from unauthorized access, use, modification, or destruction. It was shown that two randomly initialized feedforward neural networks (FNNs), with one layer of hidden units (Protić, 2015), which are trained on their mutual output, can synchronize identical time-dependent weight vectors. In comparison with traditional supervised learning, there is no fixed target function in mutual learning, because each of communicating parties acts as a “teacher” and a “student” simultaneously. Both of the two parties’ statuses are chaotic, driven by a random input (Zhou et al, 2004). Two parties synchronize their networks without transmitting inputs over a public channel.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
