Font Size: a A A

Research On Gain Flatness And Polarization Dependent Gain Of Fiber Raman Amplifier

Posted on:2010-08-25Degree:MasterType:Thesis
Country:ChinaCandidate:L WangFull Text:PDF
GTID:2178360278459302Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Fiber Raman amplifier (FRA), based on Stimulated Raman Scattering mechanism, is the amplifier which can work on the whole wave band from 1270nm to 1670nm. Moreover, hiring transmission fiber as the gain medium, FRA has some excellent features such as can amplify any wave band with suitable pumps, good noise performance, good ultra-broadband amplification characteristics etc. Particularly with the emergence of high-power pumping semi-conductor laser and development of fiber Bragg grating technology, FRA has been a research hotspot in recent years, which has a comprehensive and attractive potential application in the aspect of optical communication. Some numerical and experimental research work on FRA is discussed in this thesis, including numerical simulation method, optimization algorithm for FRA gain equalization, as well as polarization mode dispersion induced Raman gain fluctuations.First, in the paper we solve the propagation equation of FRA by using the classic Runge-Kutta arithmetic iteratively, adopted a revised calculation model in order to reduce the truncation error. Meanwhile, for two point boundary value problems in numerical solution of the multi-wavelength backward pumping FRA, we improved a novel weighted average algorithm with fast convergence which processed both signal and pump by weighted average. When selecting suitable weighting factor and dividing fiber into two part in reason, numerical solution results in MATLAB show that the weighted average algorithm has a better characteristic of convergence.Secondly, in order to obtain optimal flat gain spectrum, we improved a optimization algorithm based on neural network by adjusting the pump power. Compared with the original optimization algorithm, the improved algorithm obtained a better convergence characteristic by introducing learning factor; it also considered the Raman interaction between pump waves and got better optimal results in the case of selecting suitable pump wavelength and k factor.Finally, probability method is be used to analyze the statistics of polarization dependent gain (PDG) in fiber Raman amplifiers, compared the PDG influenced by the different pumping way and different fiber parameters, introduce the conception of degree of pumping directionality into bidirectional pumping way. We analyze the degree of pumping directionality resulting in effect on PDG on certain conditions.
Keywords/Search Tags:Fiber Raman Amplifier, Flat Raman Gain Spectrum, Optimization Algorithm, Polarization Dependent Gain
PDF Full Text Request
Related items