云计算下网络入侵检测频率智能调整算法仿真

Simulation of Network Intrusion Detection Frequency Intelligent Adjustment Algorithm in Cloud Computing

针对当前入侵检测算法捕获异常数据包并对其进行分析,存在较低的检出率和检测率等问题,提出基于概率统计的云计算下网络入侵检测频率智能调整算法。采用概率统计方法提取网络攻击高、低危数据频谱特征,计算相应频谱特征幅值在全网中出现的总概率。利用经验模态分解方法根据概率计算结果分离出高、低危信号特征谐波分量,给出两类信号特征谐波分量在时域分布上的特征差异,建立频率调整阈值,根据该阈值确定入侵检测频率,完成入侵检测。实验结果表明,所提算法通过对入侵检测频率的调整,在一定程度上提高了异常数据检出率和检测正确率。

Currently, the intrusion detection algorithm has low detected rate and detection rate in capturing and analyzing abnormal data packet. Therefore, this article focuses on an intelligent adjustment algorithm for frequency of network intrusion detection in cloud computing based on probability statistics. This research used probabilistic statistical method to extract high-risk and low-risk data spectrum characteristics of network attacks and calculate the total probability that the corresponding spectrum characteristic amplitude appeared in whole network. Then, this research used the method of empirical mode decomposition and probabilistie calculation results to separate the characteristic harmonic components of high-risk and low-risk signals. Meanwhile, the research gave characteristic differences between two kinds of characteristic harmonic components in the time domain distribution and established the frequency adjustment threshold value. Thus, we determined frequency of intrusion detection through this threshold value and completed the intrusion detection. Simulation results prove that the proposed algorithm improves the detected rate and detection accuracy of abnormal data in a certain degree by adjusting the frequency of intrusion detection.