基于时间切片序列图像的数据流量检测技术  

Data flow anomaly detection method based on serial images by time⁃slice technology

作　　者：咸永强 王玉冉 高晓宇[1] 罗文源 谷伟[1] XIAN Yongqiang;WANG Yuran;GAO Xiaoyu;LUO Wenyuan;GU Wei(Unit 63612 of PLA,Dunhuang 736200,China;Xinjiang University of Science and Technology,Korla 841000,China)

机构地区：[1]中国人民解放军63612部队,甘肃敦煌736200 [2]新疆科技学院,新疆库尔勒841000

出　　处：《现代电子技术》2023年第23期164-166,共3页Modern Electronics Technique

基　　金：2023年新疆维吾尔自治区省级大学生创新创业训练计划项目(S202313561013)。

摘　　要：为实现I/O设备数据发送状态的智能快速检测,采用开源Wireshark软件对设备的时间序列数据流量进行采集,提出对I/O设备时间序列流量图像进行切片处理的方法。针对切片图像应用分形理论表征设备通信数据流量的发送状态,基于通信网络数据流量折线图利用Matlab编写计算程序计算数据流的分形维数,将I/O设备正常传输数据状态和停止传输数据状态两种情况下的分形维数进行对比分析。计算数据表明分形维数可以用来表征设备通信数据流量的变化,根据分形维数数值能够区分数据正常传输与异常传输两种状态,证实该方法可用于设备的网络数据流量检测。In order to realize the intelligent and rapid detection of data sending status of I/O devices,open source Wireshark is used to collect the time series data flow of the devices,and a method of slicing the time series flow image of I/O devices is proposed.In terms of slice images,fractal theory is applied to characterize the sending state of communication data flow of the devices.On the basis of the data flow line chart of communication network,the fractal dimension of data flow is calculated by using Matlab to compile a calculation program.By contrasting the fractal dimension between the normal data transmission state and the abnormal data transmission state of the I/O devices,the calculated data show that the fractal dimension can be used to describe the change of data flow in communication network.In addition,according to the data of fractal dimension,normal data transmission and abnormal data transmission can be distinguished,which proves that the proposed method can be used in network data flow detection of the devices.

关 键 词：时间切片 分形理论 通信网络 数据流量 流量检测 传输状态 

分 类 号：TN919-34[电子电信—通信与信息系统]