摘要
为了提高对网络不稳定节点定位和检测精度,提出基于经验模态分解和功率谱密度特征提取的网络不稳定节点的动态特征挖掘模型。首先对网络不稳定节点输出信号进行经验模态分解,将一个复杂的网络不稳定节点的动态信号分解成若干个IMF分量之和,对分解信号进行功率谱密度特征提取,实现对网络不稳定节点的动态特征挖掘。仿真结果表明,该挖掘模型能准确实现对不稳定节点输出信号的参量估计和动态特征提取,特征挖掘精度较高,较好地实现了对不稳定节点的定位识别。
In order to improve the location and detection accuracy of the network unstable nodes,a network unstable nodes′ dynamic feature mining model based on empirical mode decomposition and power spectral density feature extraction is proposed. The empirical mode decomposition is performed for output signals of the network unstable nodes to decompose the dynamic signal of a complex network unstable node into the sum of several IMF components. The power spectral density feature ofthe decomposed signal is extracted to mine the dynamic features of the network unstable nodes. The simulation results show that the mining model can accurately realize the output signal parameter estimation and dynamic feature extraction of the unstable nodes,has high feature mining accuracy,and can locate and recognize the unstable nodes better.
作者
刘菲
LIU Fei(Modern Educational Technology Center,The National Police University for Criminal Justice,Baoding 071000,C)
出处
《现代电子技术》
北大核心
2017年第3期19-22,共4页
Modern Electronics Technique
关键词
网络不稳定节点
输出信号
动态特征挖掘
经验模态分解
network unstable node
output signal
dynamic feature mining
empirical mode decomposition
