噪声环境下的网络异常信号检测方法研究

发布时间：2018-01-16 20:26

本文关键词：噪声环境下的网络异常信号检测方法研究　出处：《计算机仿真》2016年08期 　论文类型：期刊论文

【摘要】：存有噪声环境下的网络异常信号检测,可保证网络安全稳定的运行。网络在有噪声的干扰下容易导致真实的异常信号特征与振动信号的频率极其相似,传统方法针对噪声干扰,无法提取稳定的异常网络信号特征,不能够准确完成网络信号特征的分类,使得获取的网络异常信号检测结果存在较大的偏差。提出梯度自适应网络异常信号检测方法。以离散的网络信号梯度信息为先验条件进行自适应去噪,并融合于混沌相空间重构理论对去噪后的离散网络信号进行重构,得到的重构时间序列和ELMAN网络算法相结合,构建ELMAN网络异常信号检测函数模型,依据网络异常信号观测的序列值和实际值间的对比建立网络异常信号检测模型,进而完成对网络异常信号精确检测。仿真结果表明,梯度自适应网络异常信号检测方法检测精确度高。
[Abstract]:The detection of abnormal signals in noisy environment can ensure the safe and stable operation of the network. The characteristics of the real abnormal signals are very similar to the frequency of vibration signals when the network is disturbed by noise. The traditional method can not extract the stable abnormal network signal features and can not accurately complete the classification of the network signal features for noise interference. A gradient adaptive network anomaly signal detection method is proposed. The discrete network signal gradient information is used as a priori condition for adaptive denoising. The chaotic phase space reconstruction theory is used to reconstruct the signal of the de-noised discrete network. The reconstruction time series is combined with the ELMAN network algorithm. The ELMAN network anomaly signal detection function model is constructed, and the network anomaly signal detection model is established according to the comparison between the sequence value and the actual value of the network abnormal signal observation. The simulation results show that the gradient adaptive network anomaly detection method has high accuracy.
【作者单位】：石家庄铁道大学;
【分类号】：TP393.08
【正文快照】： 1引言由于网络环境较为复杂,致使产生的异常信号会与一些毫无规律的干扰或噪声混淆在一起,给网络稳定的运行带来了极大的隐患[1-3]。而网络异常信号检测技术是解决这一隐患的有效途径,引起了很多专家与学者的重视[4-6]。由于网络异常信号检测技术具有广阔的发展空间,因此也成

【参考文献】