Uncovering missing symbols in communication with filtered chaotic signals

We investigate the physical modifications caused by a linear filter in a chaotic trajectory. We show that while the filter strongly modifies the topological characteristics of the chaotic signal, it might not alter its information content. We propose a chaos-based communication system that takes advantage of this fundamental characteristic of the dynamics. We devise procedures both to recover all the symbols at the receiver end and to decode the received higher-dimensional filtered chaotic signal by using techniques from the theory of pattern recognition. Our results show that a message bearing chaotic signal can be transmitted over a low bandwidth physical channel and be decoded with a low decoding bit error rate.