摘要
针对蠕虫攻击对光纤无线网络稳定运行的严重影响问题，提出光纤无线网络相邻流量时间窗异常监测方法。该方法利用时间序列分析法对相邻流量时间窗进行预处理分析，计算其自相关函数，构建ARIMA模型确定相邻流量预测值动态阈值，通过该结果对两个相邻流量判断时间窗是否存在异常，实现光纤无线网络相邻流量时间窗异常监测。实验结果证明，所提方法在保证相邻流量时间窗异常检测率较高的基础上，排除了光纤无线网络相邻流量中的干扰项，降低了漏报率，具有较高实用性。
In view of the serious impact of worm attacks on the stable operation of optical fiber wireless networks, an anomaly monitoring method for adjacent traffic flow time windows is proposed. The method uses time series analysis to preprocess adjacent flow time window to calculate the autocorrelation function, construct the predictive value of dynamic threshold to determine the adjacent flow ARIMA model, through the results of the two adjacent traffic to determine the time window is abnormal, which realize optical wireless network traffic anomaly monitoring of adjacent time window. The experimental results show that the proposed method can ensure the adjacent flow time window anomaly detection based on high rate, exclude the interference of the adjacent fiber wireless network traffic, reduce the false negative rate, has high practicability.
作者
韦小铃
WEI Xiaoling(Qinzhou University, Qinzhou Guangxi 535011, Chin)
出处
《激光杂志》
北大核心
2018年第5期151-155,共5页
Laser Journal
基金
广西高校科研项目(No.KY2015YB314)
广西高校中青年教师基础能力提升项目(No.KY2016LX427)
关键词
蠕虫攻击
光纤无线网络
相邻流量
时间窗
自相关函数
异常监测
worm attack
optical fiber wireless network
adjacent traffic
time window
autocorrelation function
anomaly monitoring
