Polarization Mode Dispersion Compensation Based On Improved Artificial Fish Algorithm Logic Control Module Development

The research work in this dissertation is supported by Science and Technology KeyProjects of Shandong Province of China,"The research on polarization mode dispersionadaptive compensation integration module" and “The research on polarization modedispersion compensation control system based on artificial fish swarm algorithm”, undergrant No.2006GG2201002and2009GG10001026, respectively.With the single-channel rate of optical fiber communication system sharply increasingfrom10Gb/s to40Gb/s and above, polarization mode dispersion (PMD) has severelyrestricted the further growth of the transmission distance of optical fiber communicationsystem. Therefore, to compensate the PMD effect is not only conducive to the upgrading oflaid optical fiber, but also helps to the construction of the next generation of ultra-highspeed optical communication system and the intelligent optical network.This thesis mainly researches the working principles and technologies of optical fiberadaptive PMD compensation logical control module from practical point, including thebasic theories and detection methods of PMD, the compensation control algorithm anddetailed realization of adaptive PMD compensation control module. The innovations of thethesis are listed as follow: introducing an improved artificial fish swarm algorithm withgood performance which also can be transplanted to digital signal processor (DSP);applying the artificial fish swarm algorithm to adaptive PMD compensation for the firsttime at home and abroad so that the application fields of artificial fish swarm algorithmextend to the optical communications; successfully developing a novel adaptive PMDcompensation logical control module based on “DSP+field-programmable gate array(FPGA)” architecture.The main research work in the dissertation is summarized as follows: firstly the basicconcepts and generation mechanism of PMD are introduced, also the paper discusses theextraction methods of PMD monitoring signals; followed by it presents the design plan ofPMD adaptive compensation logic control module based on DSP and FPGA architecture;secondly the hardware implementation of the module is discussed in detail, including DSP reset circuit, clock circuit, the storage space allocation, the EMIF interface circuit, externalexpansion FLASH memory as well as emulator interface circuit design, and the globalcontrol module, FPGA and DSP interface module, DA output control module, AD inputcontrol module, the data temporary storage module, and the median filtering module withinFPGA, also the power module design and some encountered problems in hardwaredebugging process are briefly introduced; thirdly the dissertation mainly proposes a newadaptive PMD compensation control algorithm-local neighborhood artificial fish swarmalgorithm (LNAFSA) and uses three classical functions to prove that the algorithm can betransplanted to DSP for running successfully; fourthly we verify the hardware performanceof developed PMD compensation logic control module; finally LNAFSA and globalartificial fish swarm algorithm (GAFSA) are both used for40Gb/s Return-to-Zero (RZ)coding second-order PMD compensation experiments, and the performance comparisonsamong LNAFSA, Simplex algorithm (SA), Genetic algorithm (GA) and Particle SwarmOptimization(PSO) are also made. The experiment results show that when using LNAFSAas PMD compensation algorithm, DOP can increase above the threshold value of0.85afterat least1and at most12iterations in this module. After PMD compensation the openingsof eye diagrams increases significantly and the compensation effect of LNAFSA isobviously better than SA, GA and PSO. When considering the compensation time, onecompensation time unit of this module consumes0.611ms and the total hardware workingtime within it is110.7μs, which are only about1/3and1/8of the module using “DSP”,respectively.