摘要
提出一种基于人工神经网络(ANN)的方法用以解决数字背向传输、Volterra级数等数字光纤通信系统均衡方法需要预先了解信道参数、计算复杂的问题。该方法利用神经网络对畸变信号的特征提取能力,通过将其进行0、1二分类来实现对畸变信号的正确识别。通过仿真分析了传输距离、采样率、样本量等参数的影响,验证了算法在消除信号短距离传输过程中的色散、非线性等效应影响的有效性。
Due to the complex and prior knowledge of the optical path’s parameters of the equalization methods based on digital backpropagation,Volterra Series Transfer Function,a method based on artificial neural network(ANN)is proposed for signal recognition at the receiver of optical fiber communication system.ANN is used to learn and extract features of distorted signals at the receiving end,which can recover and distinguish the original data from the distorted signals by binary-classification with0-1.The parameters such as transmission distance,sampling rate,and numbers of samples are analyzed,and the method is verified by the simulation experiments in short SMF transmission.
作者
梁猛
杨斯淇
Liang Meng;Yang Siqi(College of Electronic Engineering,Xi'an University of Posts and Telecommunications Xi'an,Xi'an 710121,China)
出处
《信息通信》
2020年第6期15-18,共4页
Information & Communications
关键词
光纤通信系统
人工神经网络
特征
分类
畸变
识别
误码率
短距离传输
optical fiber communication system
artificial neural network
features
classification
distortion
distinguishment
bit error rate
short SMF transmission
