Unmasking Optical Chaotic Cryptosystems Based on Delayed Optoelectronic Feedback

The authors analyze the security of optical chaotic communication systems. The chaotic carrier is generated by a laser diode subject to delayed optoelectronic feedback. Transmitters with one and two fixed delay times are considered. A new type of neural networks, modular neural networks, is used to reconstruct the nonlinear dynamics of the transmitter from experimental time series in the single-delay case, and from numerical simulations in single and two-delay cases. The authors show that the complexity of the model does not increase when the delay time is increased, in spite of the very high dimension of the chaotic attractor. However, it is found that nonlinear dynamics reconstruction is more difficult when the feedback strength is increased. The extracted model is used as an unauthorized receiver to recover the message. Therefore, the authors conclude that optical chaotic cryptosystems based on optoelectronic feedback systems with several fixed time delays are vulnerable.