基于WoT平台的数据存储与处理策略翟天铎1，纪阳21北京邮电大学移动生活与新媒体实验室，北京1008762北京邮电大学泛网无线通信教育部重点实验室，北京100876摘要：Web物联网（WOT）包含传感器，RFID设备，手机，摄像头等各种设备。这决定了WoT数据的两个特性：异种和海量，为数据的存储和处理带来了巨大的挑战。根据研究工作，对WoT数据进行分析，本文提出了一种数据存储处理方案，设计了WoT的数据存储结构，介绍了异种数据的存储方案，同时做了数据处理工作来应对数据重复和数据过快增长的问题。关键词：Web物联网，数据冗余，数据存储，数据处理中图分类号：TP311DataStorageAndProcessPolicyBasedOnWoTPlatformZHAITian-Duo1,JIYang21BeijingUniversityofPostsandTelecommunicationsMobileLifeandNewMediaLab,Beijing1008762BeijingUniversityofPostsandTelecommunicationsKeyLaboratoryofUniversalWirelessCommunications,MinistryofEducation,Beijing100876Abstract:WebofThings(WoT)consistsofvariousofdevicessuchassensors,RFIDdevices,cellphones,camerasandsoon.ThisdeterminesthetwocharacteristicsofWoTdata:heterogeneousandmassive,andbringsdatastoragehugechallenges.BasedonanalysisofWoTdataandrecentwork,thispaperproposedadatastoragemanagementsolution.AdatamodelisdevisedforWoTdatastorageandasolutionforheterogeneousdatastorageisintroduced.Dataprocessingtodealwiththedataduplicationandtheexcessivegrowthofdatahasbeendone.Keywords:WoT,dataredundancy,datastorage,dataprocess0IntroductionTheWebofThings(WoT)isthenetworkconnectsobjectssuchassensors,RFIDdevices,cellphonesandcamerastotheInternetthroughWeb.ThegoalofWoTismonitoring,manage-基金项目：TheWirelessUbiquitousBusinessEnvironmentArchitecture,KeyTechnologyResearchandDemon-strationBasedonWeb(2012ZX03005008-001)作者简介：ZhaiTianduo（1990-），male，graduatestudentofBeijingUniversityofPostsandTelecommunica-tions，majorresearchdirection：WebofThings.Email:zhaitianduo@126.comCorrespondenceauthor：JiYang（1972-），male，professor，majorresearchdirection：InternetofThings,WebofThings.Email:jiyang@bupt.edu.cn-1-ment,analysisandintelligence.DataisthemostvaluableaspectinWoT.Dataiscentralizedinplatformforstorage,management,andanalysis.WoTdataconsistsofservicedatawhichisgeneratedbydifferentdevicesandattributewhichdescribestheservicedata.WoTservicedatahasfollowingfeatures:1.Hugescale.Theplatformisreceivingdatafromdevicesorgatewaysallthetime.Devicesorgatewayssenddataperiodicallyoratacertainpatterntotheplatformwhichleadsarapidexpansionofdatascale.2.Heterogeneity.DifferentdevicesexistinWoTlikecameras,sensors,RFIDdevicesbringsmul-tipledataformats.Scalardatamaycontainvaluesofsensors,RFIDinformation.Multimediadatamaycontainpictures,audioorvideo.3.Temporal-spatialcorrelation.Datacollectedfromrealworldreflectthechangesintherealworld.Twopiecesofadjacentdatamaybethesamebecausenothingchangesduringthistime.Anddatainthesamepiecefromgatewayarerelatedsincetwodevicesmaybeinthesamelocation.Thecorrelationleadstoanotherproblem,dataredundancy.Sodataprocessingisnecessary.4.DataisretrievedviaWebleveragingRESTfulAPI.Soplatformprovidestwokindsofdataformats:XMLandJSON.Xmlformatisunified,independentfromplatformandprogramminglanguage,andhasbecomeanindustry-acceptedstandard.It’saheavyweightdataexchangeformat.JSONisalightweightdataexchangeformatwhichiseasytoparseandgenerateformachines.5.Highfrequencyofdataretrieval.Thereal-timedataisretrievedmostfrequently.Thisbringstwochallengesfortheplatform.a)Anefficientandrationaldatacachingsolutionisnee...