恒虚警下异常网络流量序列监测技术仿真分析

发布时间：2017-12-28 10:32

本文关键词：恒虚警下异常网络流量序列监测技术仿真分析　出处：《计算机仿真》2016年08期 　论文类型：期刊论文

【摘要】：在对恒虚警下的异常网络流量序列进行监测时,由于网络流量的多样性,使得流量序列特征发生动态波动。传统的监测技术,主要通过对网络流量序列特征进行分类,完成局部监测,忽略了动态波动会使流量序列特征发生变化的因素,导致监测误差大、精度低的问题。提出一种恒虚警下基于信息熵的异常网络流量序列监测技术,通过局部干扰功率均值评估待检测网络单元,对参考单元样本求均值,估计待检单元的背景杂波功率获取门限值,并过滤网络中的背景杂波。依据概率统计理论给出信源信息熵的定义,采用指数平滑法对下一周期信息熵进行预测,利用指数递减的权值对网络流量历史序列进行分配,观察熵和预测熵的偏差判断网络流量序列是否出现异常。仿真结果表明,所提方法具有很高的监测精度。
[Abstract]:When monitoring the abnormal network traffic sequence under the constant false alarm, the characteristics of the flow sequence are fluctuated dynamically because of the diversity of network traffic. Traditional monitoring technology mainly classifies network traffic sequence characteristics, completes partial monitoring, ignores the factors that cause dynamic sequence changes in traffic sequence characteristics, resulting in large monitoring error and low accuracy. A CFAR sequence of abnormal network traffic monitoring technology based on information entropy, the local mean to detect network interference power assessment unit, average unit reference samples, estimated to be detected unit background clutter power access threshold, and filter the background clutter in the network. According to the definition of the theory of probability and statistics are given information, carries on the forecast to the next cycle of information entropy by exponential smoothing method of network traffic history sequence distribution by using exponential weights, observation and prediction deviation entropy entropy judgment whether there is abnormal network traffic sequence. The simulation results show that the proposed method has high monitoring precision.
【作者单位】：江苏开放大学教务科研处;
【分类号】：TP393.06
【正文快照】： 1引言近年来,随着计算机和科技的逐渐发展,网络中的恶意攻击事件越来越多,给网络安全防护带来了新单挑战,人们对网络安全的关注程度也越来越多[1,2]。当网络被攻击,出现网络病毒和网络故障时,网络流量会出现异常的表现。由于传统检测方法容易受到恒虚警环境的干扰而产生杂波,

【参考文献】