光纤网络中的离群异常数据检测方法研究  

Research on outlier anomaly data detection method in optical fiber network

作　　者：包科[1] 沈叶忠 

机构地区：[1]无锡职业技术学院,江苏无锡214121 [2]江南大学,江苏无锡214122

出　　处：《激光杂志》2017年第8期189-192,共4页Laser Journal

基　　金：苏州市教委科研计划项目(41132530)

摘　　要：对光纤网络中的离群异常数据准确检测可实现网络故障诊断排除。针对目前光纤网络离群异常数据散布性强,检测准确度不高的问题,提出一种基于鉴频响应谱峰搜索的光纤网络中离群异常数据检测算法。首先构建光纤网络离群异常数据片选信号的特征分布幅频响应曲线,分析离散异常数据的超线性混沌特性,其次提取离群异常数据的鉴频响应特征,采用自适应谱峰搜索方法实现离群异常数据的聚集性检测。最后进行仿真测试,结果表明,采用本文方法进行光纤网络中的离群异常数据检测的准确性好,抗干扰能力较强。Outlier abnormal data accurate detection of optical fiber network can realize the network fault diagnosis.For the strong distribution of the outlier data in fiber network and low detection accuracy at the present,a new algorithm of outlier anormaly data detection in optical network based on frequency response spectrum peak search is proposed in this paper. First construct the feature distributed frequency respond curve of the outlier anormaly data detection in optical network,analyze its superlinear chaotic feature and then extract its frequency respond feature,realize its clustering detection by using adaptive spectra peak research method. Finally implement the simulation test,and the results show that the method in this paper has good accuracy and strong anti-jamming capability.

关 键 词：光纤网络 数据 检测 异常 故障 

分 类 号：TN91[电子电信—通信与信息系统]