Research On Chromatic Dispersion Estimation Algorithms In Coherent Optical Transmission Systems

With the rapid development of 5G,cloud computing,data center and other emerging network industries,the demand for large-capacity optical fiber communication is rising.As the main carrier of data traffic,the polarization multiplexing(PM)coherent detection with high spectral efficiency and the large-capacity and long-distance optical otransmission system combined with digital signal processing requires not only the support of more efficient modulation and demodulation technology,but also the accurate and reliable transmission damage monitoring/estimation technology to ensure the safe and stable operation of the system.However,in F aster-than-Nyquist/Nyquist Wavelength Division Multiplexing(FTN/Nyquist-WDM)systems,unsatisfactory and harsh shaping filter introduces serious Inter-symbol Inference in multiplexing,long distance optical transmission also leads to more complex link damage.Traditional Chromatic Dispersion(CD)estimation algorithms not only face the problems of contradictions in complexity and accuracy but also are sensitive to damage of Polarization Mode Dispersion(PMD).At the same time,CD and PMD joint estimation algorithm also faces the problem of high complexity.This paper carries out research on the above problems.The main research work and innovative research results are as follows.1.In order to solve the problem that complexity and accuracy of traditional CD estimation algorithms are very contradictory in FTN system,a low complexity CD estimation algorithm based on Fractional Fourier Transform(FrFT)and training sequence is proposed.The proposed algorithm uses a training sequence with a single symbol length of 256 to complete the coarse CD estimation,narrowing the search range of the fine CD estimation.Under the condition of certain overhead,its complexity is only 3%and 6%of that blind CD estimation algorithms based on FrFT and Peak-to-average Power Ratio(PAPR)with the same accuracy.When the acceleration factor is 0.95/0.9/0.85,the results of 64GBaud PM-16QAM FTN-WDM simulation and 60GBaud PM-QPSK/16QAM FTN offline experiment show that CD estimation can be achieved in the range of 15360ps/nm and 11520ps/nm,respectively.The estimated error is better than 65ps/nm and 70ps/nm.2.Aiming at the problem of large CD estimation error when traditional CD estimation algorithm has large PMD damage,a CD estimation algorithm with high PMD tolerance suitable for FTN system is proposed.The algorithm achieves coarse CD estimation by stacking training sequences insensitive to PMD damage,and completes fine CD estimation by using the consistency parameter of clock phase detection.64GBaud PM-16QAM FTN-WDM simulation results show that when the acceleration factor is 0.85 accumulated CD is 15360ps/nm and the CD estimation error is 200ps/nm,the PMD tolerance is about 100ps,and the complexity and PMD tolerance are 24%and 2.5 times of the PAPR blind CD estimation algorithm,respectively.3.Aiming at the problem of high complexity of CD and PMD joint estimation algorithm,a low complexity CD and PMD joint estimation algorithm suitable for Nyquist system is proposed.The algorithm takes advantage of the sensitivity of clock information to CD and PMD damage to complete the joint estimation.The algorithm complexity is only 2.1%of the joint estimation algorithm based on FrFT time-frequency reconstruction and multi-task deep neural network.The simulation results of 32GBaud PM-16QAM Nyquist-WDM show that the CD estimation can be achieved in the range of 20480ps/nm,and the estimation error is better than 150ps/nm.In addition,it can support accurate estimation of PMD in the range of 30ps,and the estimation error is better than 2ps.