改进的NetLinX开放网络动态入侵检测方法

Improved method of dynamic intrusion detection in NetLinX open network

针对当前时频特征检测算法的检测准确概率较低,网络入侵信息的检测准确性不高,提出基于经验模态特征分解和功率谱密度特征提取的NetLinX开放网络动态入侵检测算法.通过对NetLinX开放网络模型分析和动态入侵信号模型的构建,实现网络动态入侵信号的经验模态特征分解,将网络入侵的动态信号分解成若干个IMF分量之和.对分解的NetLinX开放网络动态入侵信号进行功率谱密度特征提取,完成对入侵特征的波束聚焦检测,实现NetLinX开放网络动态入侵信号的检测过程.仿真结果表明,采用该算法进行网络入侵检测的准确概率较高,抗干扰性能较好,实时性较强.

Due to the inaccuracy of frequency feature detection algorithm and network intrusion information, NetLinX open dynamic network intrusion detection algorithm based on feature extraction of power spectral density is proposed.Through analyzing the NetLinX open network model and constructimg the dynamic intrusion signal model,the empirical mode decomposition of dynamic network intrusion signal is realized, and the signal is decomposed into several IMF components.The power spectral density feature of NetLinX open network intrusion dynamic signal of decomposition is extracted and beam focusing detection for intrusion detection feature is completed, and the detection process is thus realized. Simulation results show that the algorthm has high accuracy, good anti-interference and strong real-time effect in network intrusion detection