Sistema de comunicação digital baseado em geração de órbitas periódicas instáveis de mapas caóticos

Abstract

In this work digital modulation techniques relying on numeric sequences generated by chaotic maps were investigated. Chaos based communication systems have attracted much attention due to an alleged security enhancement against unauthorized access. On the other hand, chaos based communication systems eciency in additive white Gaussian noise channels has been found to be inferior to that presented in communication systems based on conventional modulation schemes using periodic signals. A recent modulation and demodulation technique called ML-CSK (Maximum Likelihood Chaos Shift Keying) based on Viterbi algorithm, and therefore relying on maximum likelihood estimation, was investigated. From this technique, and consideringconventional modulation techniques, a variation was proposed, namely the use of high period Unstable Periodic Orbits (UPO) as symbols to be transmitted, instead of sequences of chaotic signals. This new system, called ML-UPOSK (Maximum Likelihood Unstable Periodic Orbits Shift Keying), was implemented using a new method to generated high period periodic orbits from the following chaotic maps: Bernoulli shift, Tent and Logistic. For a binary digital communication ML-UPOSK system, i.e. usingtwo UPO as symbols to be transmitted, it was observed, through numerical simulation, a significant enhancement in system eciency compared to that obtained in the MLCSK case. In addition, from the numerical results it was not observed linear correlation between UPO statistical characteristics and communication system eciency.