Research On Key Technologies In High-speed Chaotic Optical Communication Systems

In recent years,with the rapid development of the Internet and communication technologies,various new Internet applications such as cloud computing,autonomous driving,and virtual reality are constantly to emerging,and optical fiber transmission is the core of Internet applications.At present,it has carried more than 90% of the traffic,and the security issue of optical fiber transmission has become more and more important.There is a risk of eavesdropping in submarines and terrestrial optical cables and users access optical fiber.Chaotic signals have noise-like and wide-spectrum characteristics,which makes chaotic optical communication technology,as a physical layer security encryption technology,possible to be the security guarantee of the physical layer of the optical fiber network.However,under this circumstance,chaotic optical communication systems are still not very compatible with existing broadband fiber-optic transmission systems.Studying the key technologies of chaotic optical communication systems has important scientific significance.Based on the aforementioned background,this thesis studies the key technologies of high-speed chaotic optical communication system,mainly involving generation of secure chaotic optical carriers,chaos-based encrypted transmission and decryption of high-speed signal,and the improvement of chaos synchronization,the purpose is to build a high-speed safe and practical chaotic optical communication system,the main research contents are as follows:1.Generation of secure chaotic optical carriersIn order to solve the security problem of the time-delay feature leakage in alloptical feedback intensity chaos,the tunable dispersion compensator(TDC)is introduced into the feedback loop of all-optical chaos generator.The time-delay feature can be concealed in time domain and frequency domain,and experimental verification is carried out.In order to solve the secure problem of the small controllable physical parameter space in electro-optic feedback phase chaos,the Mach-Zehnder interferometer is replaced by the TDC to generate phase chaos,and the dispersion curve of cascaded Gires-Tournois(G-T)etalons in TDC can be finely controlled by temperature,which effectively increase the controllable physical parameter space of chaotic optical generator,and experimental verification is carried out.2.30 Gbps chaotic optical communication system based on hardware synchronizationHigh-speed chaotic optical communication is limited by the generation and synchronization of broadband chaos.In this thesis,the 30-Gbps non-return-to-zero onoff-keying(NRZ-OOK)signal is compressed into the duo-binary signal with a bandwidth of 10 GHz,therefore the high-speed signal can be hidden by a chaotic optical carrier with limited bandwidth.Base on the experimental system of 10-GHz-wide chaos synchronization,the chaotic encryption and decryption of 30-Gbps NRZ-OOK signal is experimentally demonstrated,which breaks the previous record of the highest experimental bit rate of 10 Gbps in chaotic optical communication system.Dispersion compensation fiber(DCF)and tunable dispersion compensator(TDC)are combined to accurately compensate the fiber dispersion,and the injection power of fiber is controlled to suppress the nonlinear effect of fiber and the amplifier spontaneous emission noise in erbium doped fiber amplifier.Based on the above schemes,30-Gbps signal transmission of a NRZ-OOK message hidden in a chaotic optical carrier over100-km fiber is experimentally demonstrated.3.Chaotic synchronization error compensation based on digital signal processing technologyParameter mismatch of physical devices in the aforementioned electro-optic feedback chaotic optical communication system based on hardware synchronization will degrade the performance of chaos synchronization.In this thesis,feedforward equalizer(FFE),Volterra filter and neural network filter are proposed to compensate the chaotic synchronization error caused by different parameters mismatch,the results of simulation and experiment show that the correlation coefficient of chaos synchronization can be improved from 0 to 0.8 by the neural network filter in some occasions.Furtherly,three kinds of filters are used to improve the bit error rate(BER)performance of the chaotic optical communication system with parameter mismatch,and the BER of the system can be reduced from 0.1 to below the threshold(2e-2)of soft decision forward error correction(SD-FEC)by the neural network filter and Volterra filter.4.Chaos synchronization and decryption based on neural network modelingParameter matching of physical devices is a difficult problem for the practical application of chaotic optical communication systems.In this thesis,the hard problem can be solved by modeling based on neural network(NN).A 10-GHz chaos synchronization with a similarity of 97.57% is experimentally demonstrated.Feedforward equalizer(FFE)is proposed to compensate the degraded chaotic signal,then neural network is used to realize the chaos synchronization.Based on this,32-Gbps signal transmission of a quadrature amplitude modulation(16QAM)message hidden in a chaotic optical carrier is experimentally demonstrated.And the influence of digitalizing bit and sampling rate of analog-digital converter on the NN-based decryption performance under different feedback strengths is studied.The system with three common attacks and different feedback strengths is analyzed to prove the security.