Achieving secure chaotic communication using EKF-based embedded-keys system

Abstract

In this paper, we describe methods that simultaneously transmit keys and messages in a three-port secure chaotic communications network. These messages are decoded using a matrix representation of linear multivariable systems. Our approach consists of an optimal extended Kalman filter (EKF)-based observer, which is a linearization with time. We consider the observer with messages and the input with keys to be the transmitting agents. The optimal linearization technique is utilized to obtain the exact linear models of a chaotic system for operating states of interest. An EKF algorithm is used to estimate the parameters and states in which the message is already embedded. By combining the EKF with our optimal linear model, the message can be recovered on the receiver end. We provide numerical examples and simulations to demonstrate the effectiveness of our proposed methodology.