网络流量可视化的新方法.pdfVIP

ISSN 1673．9418 CODEN JKYTA8 E—mail：fcst@vip．163．tom JournalofFrontiersofComputerScienceandTechnology http：／／www．ceaj．org 1673-94l8／2015／09(04)一0451-11 Tel：+86．．10． doi：10．3778~．issn．1673-9418．1405025 网络流量可视化的新方法 高丕红 ，徐明伟 清华大学 计算机科学与技术系，北京 100084 New M ethodofNetworkFlowVisualization GAOPihong，XU M ingwei DepartmentofComputerScienceandTechnology,TsinghuaUniversity,Beijing100084，China +Correspondingauthor：E-mail：danceintheair57@gmail．com GAO Pihong，XuM ingwei．New methodofnetworkflow visualization．JournalofFrontiersofComputerSci- enceandTechnology,2015，9(4)：451—461． Abstract：Flow visualizationiSofgreathelpfornetworkresearchersandsupervisorstodetectnewtorkanomalies， understandthecharacteristicsofnewtorktrafficandfindtimetrend，etc．Existingflow visualization technologies havefailedtofullymeetthegrowingneedsoftrafficvisualization．Aimingattime，dimensionandstructurecharac— teristicsofflow,thispaperadoptsspiralvisualizationtechnology,dimensionclassificationmethodandspacefilling technologytomeettheneedofdifferentcharacteristicsofflow．Byvisualizingthedatafrom4over6projectand compraedwithhtemethodscurrentlyusedintheproject，theresultsshowthatspiralvisualizationcanexcavatethe timechraacteristicsoftrafficdatamoreeffectivelythanlinemapping；dimensionclassificationismoreeffectiveofr specialdimensionsvisualization thanparallelcoordinatesandspacefillingCna expressnodeinformationwithno lossofstructurechraacteristicswhichnode—linkmappingcan’tachieve． Keywords：visualization；4over6；rtafficcharacteristics；deriveddata 摘 要：流量可视化对于网络研究和监管人员侦测网络异常，了解网络中流量的特征和趋势等有着重要的意 义。现有的流量可视化技术不能满足 日益广泛的流量可视化需求。针对流量数据的时间、维度、结构等特征， TheNationalNaturalScienceFoundationofChinaunderGrantNo．611611404