摘要
针对光纤网络链路故障在诊断与定位过程中受到冗余干扰的影响,导致其诊断和定位精度不高,以提高光纤网络链路故障检测的精度为目的,提出了基于深度学习的光纤网络链路故障诊断与定位方法.在建立链路故障诊断网络的基础上,结合深度学习的联合概率分布,对光纤网络链路故障进行重构,通过计算光纤网络输出层与目标输出层之间的均方误差和梯度值,完成了光纤网络链路故障的诊断,通过筛选光纤网络链路故障,设计了光纤网络链路故障定位算法,实现了光纤网络链路故障的定位.实验结果表明,所提方法不仅可以降低光纤网络链路故障的误检率,还可以提高光纤网络链路故障的检测率.
In order to improve the accuracy of optical fiber network link fault detection,a method of optical fiber network link fault diagnosis and location based on deep learning is proposed.Based on the establishment of link fault diagnosis network,the energy function of deep learning is defined.Combined with the joint probability distribution of deep learning,the link fault of optical network is reconstructed.By calculating the mean square error and gradient value between the output layer of optical network and the target output layer,the link fault diagnosis of optical network is completed,The algorithm of link fault location in optical fiber network is designed,and the link fault location in optical fiber network is realized.Experimental results show that the proposed method can not only reduce the error detection rate of optical fiber network link fault,but also improve the detection rate of optical fiber network link fault.
作者
丁智
DING Zhi(School of Computer Engineering,Bengbu University,Bengbu Anhui 233030)
出处
《宁夏师范学院学报》
2021年第4期85-91,共7页
Journal of Ningxia Normal University
基金
蚌埠学院星网锐捷网络有限公司实践教育基地(2017sjjd1).
关键词
深度学习
光纤网络
链路故障
诊断模型
定位算法
Deep learning
Optical fiber network
Link fault
Diagnosis model
Location algorithm
