Research On Digital Signal Processing Algorithms For Carrier Recovery In Coherent Optical Communication

With the increasing demand for communication rates,coherent optical communication technologies with high sensitivity and high spectrum utilization have attracted increasing attention.As an important part of the coherent optical communication system,digital signal processing needs to optimize the digital signal processing algorithm in order to reduce the hardware requirements.This paper mainly studies the carrier recovery algorithm of digital signal processing part.Firstly,this paper introduces the historical background of the development of coherent optical communication.Then,the modulation method,coherent detection principle and digital signal processing flow in coherent optical communication are briefly introduced.Then the paper introduces the carrier recovery algorithm,including the frequency offset estimation algorithm and the phase estimation algorithm.In this thesis,the traditional 4th power frequency offset estimation algorithm is optimized and improved,and an optimized frequency offset estimation algorithm based on remainder is obtained..The algorithm replace the traditional fourth-order operation with the remainder operation to remove the modulation phase which reduces the complexity of the algorithm.On this basis,a 2-stage frequency offset estimation algorithm based on redundancy operation for QPSK and 16QAM signal is proposed.when the signal-to-noise ratio is higher than about 10 dB,the redundancy algorithm is more accurate than the 4th power algorithm.Compared with the 4th power algorithm,the redundancy algorithm has higher accuracy.In the 16QAM signal,the utilization rate of the two-stage redundancy algorithm to the signal constellation points is four times that of the two-stage four-power algorithm.In this thesis,the traditional 4th power carrier phase estimation algorithm is optimized and improved,and a high-precision 4th power carrier phase estimation algorithm is obtained..The algorithm considers the phase offset caused by the residual frequency offset,additive Gaussian noise and phase noise as a whole,estimates the total phase offset caused by the these three factors for each symbol,and then recovers the modulated phase.This algorithm reduces the step of summing and averaging in the traditional 4th power phase estimation algorithm,and reduces the complexity of the algorithm.The simulation results show that the error between the signal phase and the original modulation phase is only 10^-16rad.However,when using the traditional M-th power carrier phase recovery algorithm the error can reach 0.3 rad.The algorithm in this paper can recover the modulation phase more accurately.When the frequency offset signal is not too large,even if the frequency offset is not compensated,we can also accomplish the phase recovery using the improved algorithm.However,when the phase descrambling is performed,a constraint condition is introduced,that is,the difference between the phase offsets of the adjacent two signals is less than a fixed value,it makes the accuracy of the phase estimation algorithm limited by residual frequency offset,additive Gaussian noise and phase noise.When the line width is 200KHz,for the QPSK signal,the lowest SNR that satisfies the phase estimation algorithm is 14.58dB while the frequency offset is compensated,the lowest SNR that satisfies the phase estimation algorithm is 16.8dB while the frequency offset is not compensated.For the 16QAM signal,the minimum SNR that satisfies the phase estimation algorithm is 24.66dB.