Recurrent Neural Network with Backstepping controller based Fractional-Order Chaotic System based Solution for Secure Communication and Image Encryption

Synchronisation of Fractional-order Non-linear systems can be used by incorporating additional security attributes. In this work, a Recurrent Neural network with Backstepping controller is designed used for synchronizing two identical non-linear fractional-order chaotic systems in master-slave configuration. The controller is split into two parts. One is designed at the sender's end and the other at receiver section to obtain synchronisation between Master and Slave systems. Also, two signals are transmitted in which one has message signal embedded in it and the other is a complicated combination of master states which is used for synchronisation purpose at the receiver end so that information signal can be recovered. A complex key set is designed with system parameters, fractional orders, initial conditions, so that no intruder can decode the message signal and receiver system should also know the keyset. The efficacy of the proposed scheme is demonstrated with the help of results presented in the end.