远距离多信道光纤通信网络异常数据检测研究

Research on abnormal data detection in long distancemulti channel optical fiber communication network

异常数据的准确检测是远距离多信道光纤通信网络稳定运行的前提,研究远距离多信道光纤通信网络异常数据检测方法,提升远距离多信道光纤通信网络异常数据检测精度与实时性。使用基于小波变换的光纤通信信号去噪方法,去除远距离多信道光纤通信网络中噪声干扰;将去噪后远距离多信道光纤通信信号,作为基于多元增量分析的光纤通信网络异常数据检测方法的输入样本,设置异常检测阈值,按照异常检测阈值判断标准,实现远距离多信道光纤通信网络异常数据检测。研究结果验证:该方法对四类光纤通信网络异常数据的检测结果与预设的异常数据出现次数一致,平均检测时延低于0.3 s,可高精度、实时检测远距离多信道光纤通信网络异常数据。

The accurate detection of abnormal data is the premise of the stable operation of long-distance multichannel optical fiber communication network. This paper studies the detection method of abnormal data in long-distance multi-channel optical fiber communication network,and improves the detection accuracy and real-time performance of abnormal data in long-distance multi-channel optical fiber communication network. The method of optical fiber communication signal de-noising based on wavelet transform is used to remove the noise interference in the long-distance multi-channel optical fiber communication network. The de-noising long-distance multi-channel optical fiber communication signal is used as the input sample of abnormal data detection method based on multivariate incremental analysis,and the abnormal detection threshold is set. According to the judgment standard of abnormal detection threshold,the long-distance multi-channel optical fiber communication network is realized. Channel optical fiber communication network abnormal data detection. The results show that the detection results of four types of abnormal data in optical fiber communication network are consistent with the preset number of abnormal data,and the average detection delay is less than 0. 3 s,which can detect the abnormal data in long-distance multi-channel optical fiber communication network with high precision and real-time.