Signal Identification And Performance Optimization In Polarization-multiplexed Coherent Optical Communication System

Coherent optical communication technology is the key enabler for the optical backbone networks.Compared with the traditional direct detection,the intensity and phase information of the optical signals can be converted to the electrical signal by coherent detection,so that the high-order modulation formats(such as m PSK and m QAM)can be applied in the optical fiber communication system.In addition,in order to further improve the spectral efficiency and system capacity,PDM technology employs two orthogonal states of polarization to transmit signals on a single wavelength channel.Nowadays,PDM coherent optical communication system has been widely used in backbone networks with long distance and large capacity.In this system,the analog signals obtained by photoelectric conversion are sampled into digital signals through the high-speed analog-digital converter,and then sent to the DSP module of coherent receiver.In the DSP module,most of damages in the optical transmission link can be compensated by various equalization or recovery algorithms.Meanwhile,without any additional physical devices,DSP module can also monitor relevant parameters of optical network.It simplifies the management and improves the flexibility of optical network.Therefore,this thesis focuses on the signal identification and processing technology of PDM coherent optical communication system.The following three aspects have been extensively studied:polarization demultiplexing,modulation format identification and OSNR monitoring,respectively.1.Considering the influence of colored noise on the performance of polarization demultiplexing algorithm,an enhanced MMSE frequency-domain(FD)fractionally spaced equalization algorithm is proposed to achieve the polarization demultiplexing of coherent optical communication system.The 2-norm of estimated channel matrix and polynomial fitting of equalization coefficients are introduced to mitigate the colored noise without requiring knowledge of the optical and electrical filter.The proposed algorithm is demonstrated in an 8×112Gbit?s WDM PDM-QPSK transmission system over a 1200-km link.Simulation results show that?3-d B improvement of equalization performance is achieved compared to conventional MMSE FD fractionally spaced equalization algorithm at the BER of 10^-3.Meanwhile,the equalization performance of the proposed algorithm outperforms that of the conventional MMSE FD fractionally spaced equalization algorithm with 8-time averaging of channel estimation,and 87.5%of the overhead is reduced.Therefore,the effect on spectral efficiency is greatly reduced.2.A Stokes space modulation format identification(MFI)method using probabilistic neural network(PNN)is proposed.According to amplitude histograms obtained by the distribution of Stokes vectors on the s₁axis,the incoming PDM signals are first classified into PDM-m PSK,PDM-16QAM and PDM-64QAM signals based on PNN.To further identify PDM-m PSK signals,the constellation feature of Stokes vectors on plane(s₂,s₃)is extracted by image processing techniques and then processed by PNN.The high identification accuracy is demonstrated via numerical simulations with 28GBaud PDM-QPSK,PDM-8PSK,PDM-16QAM,and PDM-64QAM signals over wide OSNR ranges.Owing to the characteristic of PNN,the proposed MFI method has simple training process,small training data size and a small number of required symbols.Proof-of-concept experiments have been implemented to further verify the feasibility of the proposed MFI method among 28GBaud PDM-QPSK,PDM-8PSK,and PDM-16QAM signals.3.An OSNR monitoring method assisted by SVOR is proposed for digital coherent receivers.According to Godard's error of equalized signals obtained after polarization demultiplexing,the OSNR value could be estimated by using SVOR.The effectiveness of the proposed OSNR monitoring method is verified via numerical simulations with 28GBaud PDM-QPSK,PDM-8PSK,PDM-8QAM,PDM-16QAM and PDM-32QAM signals.The simulation results show that the high accuracy of OSNR monitoring over wide range can be achieved by the proposed method.The feasibility is further demonstrated by proof-of-concept experiments in 28GBaud PDM-QPSK,PDM-8PSK,PDM-8QAM and PDM-16QAM systems.For the four modulation formats,the error of each OSNR estimation is kept within 1 d B,and the average estimation errors of all modulation formats are less than0.1 d B.In conclusion,the polarization demultiplexing,modulation format identification and OSNR monitoring of the PDM coherent optical communication system are achieved based on DSP technology in this thesis,respectively.These technologies are able to improve the spectrum efficiency and system capacity,and can be used to enhance the transparency and reconfiguration of optical network.