Innovative Method of Secure Communication

Author: Salome Shekiladze
Co-authors: Gigia Aptsiauri, Davit Khanjaladze
Keywords: Artificial Neural Network, Machine Learning, Secure Communication, Messaging Software Application, Cryptography
Annotation:

Today, when social networks and virtual messaging occupy most of our life, the most important thing is the privacy of our private messages, photos or important documents, which we send it to each other. Consequently, the method of secure communication that cryptography offers is important. With using of special algorithms, encrypted information is transmitted in to the network, that can only be read by the receiver. Such methods can protect information from third parties. Despite many ways of protecting the information, many messaging applications do not use end-to-end encryption, So there is a requirement for applications where the conversation is completely secure. Our goal is to create secure messaging app. Well-known cryptography algorithms can be used to transfer information securely, But we will use one of the types of neural networks Generative Adversarial Networks (GAN), that are capable of detecting encryption and decryption methods without learning special algorithms. Accordingly, the transmitted information will be protected and practically impossible to use it it is captured.