Nonlinear dynamics reconstruction of chaotic cryptosystems based on delayed optoelectronic feedback

In this work we show that the nonlinear dynamics of the chaotic carrier can be reconstructed from experimental data. Then it is possible to extract the message by using the nonlinear model as a receiver. The transmitter consists of a DBR laser with a feedback loop formed by a delay line and an optical device with a nonlinearity in wavelength. The first step to extract the nonlinear dynamics is the estimation of the delay time T. Then a new type of modular neural network (NN) is used to obtain the transmitter nonlinear dynamics. It is checked using chaotic synchronization that the NN describes correctly the transmitter nonlinear dynamics, although it is a high dimensional system.