基于嵌入式的海量光纤通信故障数据识别技术

Embedded data communication technology for massive optical fiber communication

针对当前方法在通信中存在的误报率较高,故障检测准确率较低,故障数据识别用时长的问题,提出一种基于嵌入式的海量光纤通信故障数据识别技术,该技术利用离散小波变换算子对光纤通信信号进行处理,根据临界阈值来判断小波系数的引发原因,通过小波系数完成对光纤通信信号的重构,采用基于小波变换的数据降噪处理方法对海量光纤通信数据的噪声进行去除,利用基于模糊灰关联的故障数据识别方法将待检测数据序列与标准故障数据序列进行关联分析,将分析结果依次排列,找到其中和标准故障数据序列最为接近的数据,该数据序列即为光纤通信中的故障数据。实验结果证明,本文方法可以实现对海量光纤通信故障数据的识别,并且具有误报率低,故障检测准确率高,故障数据识别用时短的特点。

Aiming at the problem that the current method has high false positive rate in communication,low fault detection accuracy and longtime of fault data identification,an embedded massive fiber-optic communication fault data recognition technology is proposed,which uses discrete wavelet transform. The operator processes the fiber-optic communication signal,judges the cause of the wavelet coefficient according to the critical threshold,reconstructs the fiber-optic communication signal through the wavelet coefficient,and uses the data-reduction method based on wavelet transform to denoise the noise of the massive fiber-optic communication data. By using the fault data identification method based on fuzzy gray correlation,the data sequence is correlated with the reference sequence,and the obtained correlation degrees are arranged in order from large to small,in which the sequence similar to the data sequence to be detected is found,and the similar sequence is associated. The largest sequence is the fault data in fiber-optic communication. The experimental results show that the proposed method can realize the identification of massive optical fiber communication fault data,and has the characteristics of low false alarm rate,high fault detection accuracy and short time for fault data identification.