Research On Polarization Demultiplexing And Frequency Offset Estimation In 100G DP-QPSK Optical Signal Coherent Detection

With the rapid growth of traffic on fiber-optic transmission networks,the detection and processing of fiber-optic transmission backbone signals is the basis of cyberspace security and situational awareness.In recent years,100 G DP-QPSK coherent optical systems have been deployed on a large scale in fiber backbone networks.Therefore,research on 100 G DP-QPSK optical signal coherent detection technology is critical to cyberspace security.In this paper,the polarization demultiplexing and frequency offset estimation techniques in 100 G DP-QPSK optical signal coherent detection are studied.The research contents include:(1)The principle of 100 G DP-QPSK optical signal coherent detection is studied,including modulation emission and coherent reception principle,dispersion in fiber channel,polarization mode dispersion and polarization dependent loss,and the reasons for frequency offset and phase deviation.Aiming at these signal damages,the commonly used signal damage compensation algorithm is analyzed,which lays a theoretical foundation for the subsequent algorithm research.(2)An adaptive step size stokes space method for polarization demultiplexing is proposed,which solves the problem that the convergence of the constant step size of the iterable stokes space method is not good enough.The original iterative stokes space method has no singularity problem,fast convergence speed and good polarization tracking performance,but its iterative step size is fixed.When setting the step size,it needs to find a balance point between convergence speed and convergence stability.In this paper,the iterative stokes space method is improved.The proposed new algorithm retains the advantages of the iterable stokes space method and realizes the adaptive update of the step size.The simulation results show that the proposed new algorithm increases the number of operations compared with the iterative stokes space method,but the convergence speed is doubled,the convergence is more stable,and the polarization tracking performance is also improved.(3)A two-stage phase increment algorithm based on Kalman filter for frequency offset estimation is proposed,which solves the problem that the frequency offset estimation accuracy of the phase increment algorithm is not high.The original phase increment algorithm removes the spontaneous scattering noise in an averaging manner.The algorithm is simple and easy,but the frequency offset estimation accuracy is too low;and the required sequence is long,resulting in a large delay.In this paper,the phase increment algorithm is improved.The proposed new algorithm estimates the frequency offset in two stages.The first stage uses the phase increment algorithm to estimate the frequency offset of the short sequence,and the second stage uses the Kalman filter to estimate the exact frequency.Frequency offset.The simulation results show that compared with the phase increment algorithm,the proposed new algorithm has the same frequency offset estimation range,and the frequency offset estimation accuracy is improved by three orders of magnitude,and the computational complexity is only one-seventh of the previous one.