Research On Equalization Algorithms For Ultrawide Band Fiber Channel

To cope with the explosive growth of global information,various technologies have been implied to realize ultra-large transmission capacity,ultra-longer transmission distance and ultra-high transmission rate in optical communication system.Multifarious multiplexing methods such as polarization division multiplexing(PDM),wavelength division multiplexing(WDM),and advanced modulation formats such as quadrature phase shift keying(QPSK),quadrature amplitude modulation(QAM)are applied for higher spectral efficiency and larger transmission capacity of the system.However,the transmitted signals suffer different types of distortion including linear and nonlinear effects.And the polarization effect with random characteristics is the most important time varying impairment in the optical fiber channel.With the continuous improvements of transmission rate and bandwidth of optical fiber communication system,the influence of polarization effect on optical signals is becoming more and more serious.In ultrawide band optical fiber channel,the distribution of polarization effect and its evolution in time and frequency domain are complex and the mathematical model of polarization effect originally designed for narrowband channel is no longer applicable.Moreover,the original classical polarization equalization algorithms own high computational complexity and power consumption in the digital signal processing(DSP)module,whose equalization performance is greatly reduced in the ultrawide band fiber channel,which seriously affects the transmission performance of the optical communication system.Therefore,it is significant to remodel the polarization impairment and propose the corresponding low complexity equalization scheme to ensure the efficient operation of the next generation 800 Gbit/s high-speed optical fiber communication system.The paper focuses on the modeling and equalization of optical fiber polarization effect in ultrawide band channel,constructs the ultrawide band channel polarization effect model,proposes low complexity equalization algorithms,and analyzes the distribution and frequency domain evolution law of polarization effect in ultrawide band channel.Besides,combined with Digital subcarrier multiplexing(DSCM)technology,the equalization scheme to compensate the polarization impairments of ultrawide band fiber channel is proposed.The main research work of this paper is as follows:(1)Research on modeling of polarization effects in ultrawide band channelThe modeling of polarization effects in ultrawide band channel is proposed,which including static cascaded model and dynamic cascaded model.The former prefers to show the statistical characteristics of polarization effect,and the impairments induced by high order polarization effect can be reproduced,which is convenient to test the impact on the communication system.The latter prefers to show the time-varying characteristics of state of polarization(SOP),which puts higher requirements on polarization tracking in DSP.Different model can be selected according to needs in application.(2)Research on polarization de-multiplexing based on KF and MIMO structured algorithmsKalman filter(KF),and multiple-input and multiple-output(MIMO)structured algorithms can be applied for de-multiplexing in PDM coherent optical communication system.On the basis of the existing algorithms,the new versions with lower calculational complexity are proposed because of the needs of practical application.For the KF,DKF based on the diagonalization matrix processing is proposed and its effectiveness is verified on the 64 GBaud-PDM-QPSK/16QAM coherent optical simulation platform.When jointly compensating rotation of state of polarization(RSOP),which is set to a maximum of 10 Mrad/s for QPSK and 5 Mrad/s for 16QAM,combined with the differential group delay(DGD)set as 75 ps and residual chromatic dispersion(RCD)set up to 100 ps/nm,compared with KF,DKF reduces the computational complexity by more than 30%at the cost of 10%convergence delay without significant performance degradation.For MIMO structured algorithms,after setting the initial taps to avoid the singularity problem and comparing the performance in the time domain and frequency domain on 28GBaud-PDM16QAM coherent optical simulation platform.It can be seen that when RSOP is set of 5 Mrad/s with DGD of 20 ps and RCD of 50 ps/nm,the performance of time and frequency domain MIMO structured algorithms is similar,but the calculational complexity of the latter algorithm is 40%lower than that of the former algorithm.(3)Research on the equalization of polarization effect in the ultrawide band channel applying combined of DSCM technique and the demultiplexing algorithmsAn equalization scheme for all-order PMD impairment in the ultrawide band channel based on DSCM combined with time domain MIMO structured algorithm is proposed in this paper,which is verified on 200 GBaud-PDM-QPSK/16QAM coherent optical communication simulation platform.When the number of birefringence segment N=12,for QPSK and 16QAM,the OSNR is set as 22 dB and 27 dB,the average DGD is set to 20 ps and 10 ps,and each RSOP hinge is set to 10 Mrad/s and 1 Mrad/s(which make the total RSOP is(?)×10=36 Mrad/s and 3.6 Mrad/s),respectively.The modified constant module algorithm(CMA)and multiple modulus algorithm(MMA)are used to process the single carrier signals and 8 subcarriers DSCM signals,respectively.It is found that the failure probability of the former sigle carrier situation is close to 30%according to the Monte Carlo simulation results,and that of the latter is close to 0.