Research On Signal Impairment Equalization Based On Extended Kalman Filter In Optical Fiber Communication System

Optical fiber communication technology is developing at an unbelievable speed.Ultra-high-speed,ultra-large-capacity and ultra-long-haul are the forever pursuit of optical fiber.communication.In the 1980s,the rate of optical fiber communication system was only 1 Gb/s.In the 1990s,the rate of optical fiber communication system reach 1 OGb/s.At the beginning of 21 st century,due to the application of various technologies enhancing the capacity of optical fiber communication system,such as polarization division multiplexing(PDM),high-order modulation format,space division multiplexing(SDM),and the integration of coherent detection and digital signal processing(DSP)module,the single wavelength rate reach 100Gb/s,and even more 400Gb/s.Around 2010,100Gb/s single wavelength PDM-QPSK coherent optical transmission system was widely put into commercial use in backbone network.Coherent detection technology and DSP digital signal processing technology have attracted more and more attention.The integration of DSP module in coherent receiver enables the various impairments experienced in optical fiber transmission to be equalized digitally in electric domain by proper DSP algorithms.Due to the benefits of coherent detection and DSP algorithms,the commercialization of Tb/s and Pb/s optical fiber communication systems will no longer be a dream.PDM coherent optical communication system based on DSP technology is the future trend of high-speed optical fiber communication.Related researches on coherent optical communication systems are of great significance.Optical signals in optical fibers inevitably suffered from all kinds of linear and nonlinear impairments.This thesis mainly focused on linear impairments,for example,loss,chromatic dispersion(CD),polarization effect impairment,and carrier mismatch.A lot of algorithms based on DSP technology have been proposed for corresponding impairments.With the rapid improvement of optical fiber communication rate,however,typical DSP equalization algorithms for low rate and low-order modulation format have become less and less applicable to the signal impairment equalization of high-speed optical fiber communication system.This dissertation focused on the topic of equalization of signal impairments in optical fiber communication system.We deeply analyzed the physical mechanism of signal impairments in optical fiber communication system,and presented the mathematical models of various impariments.We studied the typical signal impairments equalization algorithms based on DSP technology,and proposed a joint algorithm based on Extended Kalman Filter.The main research works are as follows:(1)The physical mechanism of signal damage in optical fiber communication and the mathematical model of various damage are introduced.The influence of various signal damage on optical fiber communication system is analyzed.It is pointed out that with the increase of system transmission rate,optical fiber communication system with direct detection has been impossible to resist the negative effects of these injuries,coherent optical communication systems have emerged.Then we introduced the structure principle of the coherent optical communication system,including the realization principle of transmitter and receiver.We established the coherent optical communication system utilizing Matlab simulation software,numerical simulations have been carried out.(2)The advantages and disadvantages of the typical equalization algorithms for various impairments in coherent optical communication systems are analyzed.An integrated signal damage equalization algorithm based on EKF filter is proposed,which can effectively balance a variety of signal impairments including polarization effect loss(RSOP:rotation of state of polarization,PMD:polarization mode dispersion,PDL:polarization dependent loss),carrier frequency offset and carrier phase noise.The principle is analyzed and its feasibility is verified by simulation.(3)Based on 112Gbit/s Nyquist-PDM-QPSK system,we verified the performance,convergence speed and complexity of the proposed two stage cascade EKF algorithm,and compared it with the traditional CMA+IMP+BPS algorithm and CMA+IMP+V-V algorithm.The results show that the PDL tolerance of the proposed EKF algorithm is up to 14dB,which is 3 times than the traditional algorithm.The PMD tolerance of the proposed EKF algorithm is up to 0.5T0,which is slightly improved based on the traditional algorithm.For RSOP,we analyzed the relationship between the three RSOP models mentioned in the literature,and given the equalization algorithms,which are applicable to three RSOP models.The RSOP tolerance of the proposed EKF algorithm reached 1800?rad/symbol,2000?rad/symbol,1200?rad/symbol,which is 10 times than the traditional algorithm.Finally,the CFO estimation tolerances of the proposed EKF algorithm are[±3GHz],and the CPN estimation tolerances measured with linewidth symbol duration product reache 1×10-3.Compared with the traditional algorithm,the proposed EKF algorithm has lower absolute mean estimation error and more stable performance for CFO.By comparison,it is found that the convergence speed of the proposed EKF algorithm is fast and the complexity is relatively low.