Deciphering Secure Chaotic Communication

A simple technique for decoding an unknown modulated chaotic time-series is presented. We point out that, by fitting a polynomial model to the modulated chaotic signal, the error in the fit gives sufficient information to decode the modulating signal. For analog implementation, a lowpass filter can be used for fitting. This method is simple and easy to implement in hardware.