摘要
为提高拉曼光纤放大器(RFA)的放大带宽和光纤通信系统的传输容量,利用4个不同波长的泵浦设计了带宽为100nm的宽带RFA。在对拉曼耦合波微分方程数值求解的同时,结合粒子群优化算法优化了4个泵浦的波长及功率,得到了3组平均开关增益分别为18.690 9、19.638 7和21.764 1dB,所对应的增益平坦度分别为0.879 7、0.914 4和0.945 2dB的优化结果。基于3组优化后的泵浦功率和泵浦波长,对粒子群优化算法的边界条件进行改进,进一步提升了RFA的性能,也使算法的性能得到改善。最终实现了增益为22.480 4dB、增益平坦度为0.696 3dB的宽带RFA。同时,对算法中的惯性权重和学习因子等参数进行了适当的调整,用来分析对比不同参数选择时优化结果如何变化,分析对比的结果为算法参数的选择提供了参考。
In order to improve the amplification bandwidth of Raman Fiber Amplifier(RFA)and the transmission capacity of optical fiber communication systems,a broadband RFA with 100 nm bandwidth is designed by using four pumps with different wavelengths.Combined with the particle swarm optimization algorithm when solving the Raman coupled wave differential equation numerically,the wavelengths and powers of four pumps are optimized.The three sets of optimization results with average on-off gains of 18.690 9 dB,19.638 7 dB,21.764 1 dB respectively with corresponding gain flatness of 0.879 7 dB,0.914 4 dB,0.945 2 dB are obtained.Based on the three sets of optimized pump powers and pump wavelengths,the boundary conditions of particle swarm optimization algorithm are improved,which further improve the performance of RFA and particle swarm optimization algorithm.Finally,a broadband RFA with a gain of 22.480 4 dB and a gain flatness of 0.696 3 dB is realized.At the same time,the parameters of inertia weight and learning factors in the algorithm have been appropriately changed so as to analyze and compare how the optimization results change when different parameters are selected.The results of the analysis and comparison provide some references for the choice of algorithm parameters.
作者
巩稼民
丁哲
李思平
薛孟乐
侯玉洁
蔡庆
任帆
GONG Jia-min;DING Zhe;LI Si-ping;XUE Meng-le;HOU Yu-jie;CAI Qing;REN Fan(School of Electronic Engineering,Xi’an University of Posts and Telecommunications,Xi’an 710121,China)
出处
《光通信研究》
北大核心
2018年第5期46-51,68,共7页
Study on Optical Communications
基金
国家自然科学基金资助项目(61775180)
