Timing Phase Recovery And Adaptive Channel Equalization Algorithm For Coherent Optical Communication System

Coherent optical communication which is the key technology for long-haul and large-capacity transmission, has attracted a lot of attention. With the implementation of high-order modulation format and the polarization division multiplexing technology, coherent optical communication gains its advantage of higher spectrum efficiency and larger communication capacity than that of the intensity detection technology. At the same time, the digital signal processing (DSP) algorithms have also been widely studied.At the receiver,the DSP algorithms can compensate the interference of dispersion, optical fiber nonlinear and phase noise in electrical domain to avoid the cost of physical compensation device. This article summarizes the implementation principle of the coherent optical communication system, and mainly focuses on the digital signal processing, such as the timing phase recovery and adaptive channel equalization algorithms.1) The implementation principle of coherent optical communication system is summarized in this article, from the transmitter, optical fiber channel to the receiving part. At the transmitter, we analyse the theoretical principle of the modulator and a generation method of the 16 Quadrature Amplitude Modulation (QAM) is proposed.2) This article summarizes the three mainstream timing phase recovery algorithm. We study the three algorithms via theoretical formula and prove that the equivalence of the three kinds of algorithms in mathematical formula. Two kinds of double sampling rate algorithms are equivalent in the frequency domain expression. Finally, through the simulation data, we compare the performance of the three algorithms, and validate our theoretical results.3) On the optical platform, we conduct the experiment of 10G Baud/s Quadrature Phase Shift Keying (QPSK) coherent optical system. Through the study of experimental data, we compare the three timing phase recovery algorithms in such aspects as jitter variance and bit error rate (BER). Our theoretical and simulation results are well proved by the experiment.4) This article summarizes the classical polarization-demultiplexing and adaptive equalization algorithms. Theoretical formula of constant modulus algorithm (CMA) is deduced. Our proposed algorithms can accomplish blind equalization and carrier phase recovery simultaneously.Morever, our new improved algorithm has faster convergence speed and lower steady-state error, more suitable for high-order QAM signals than CMA.