Neural Network-Based Cryptography: A Primary Study on the Performances and Techniques

Abstract

Cybersecurity is getting more and more important in this big data era. Neural Network-based cryptosystem which is a complex system and deserve more research study for improvement. This article will discuss and analyse existing neural network-based cryptography performance and techniques. The in-depth analysis will examine the effectiveness of older structure as well as the effects initiated against them. Several advanced techniques will be discussed that reliant on generative adversarial networks (GANs) to contend with one another in pursuit of a common goal such as efficient encryption for communication. At the end of the paper, it will review a few decent neural network methods best applied on cryptography.