搜索
2015－06－055131507011987－、。101416。。MapＲeduceＲ28～40Ｒ+。ＲＲ+P208A1672－5867201510－0013－05SpatialIndexBuiltinCloudComputingEnvironmentLIZhen－juLIXue－junYANGShengLIUTaoDeptofInformationEquipmentAcademyEquipmentofPLABeijing101416ChinaAbstractSpatialdataindexissignificantlyimportantforspatialdatamanagementandanalysiswhichisrelatedwithdataprocessspeedandefficiency．Withthehumansocietyenteringintothebigdataeraremotesensingvectordataisakindofbigdata．Thetra-ditionalspatialindexconstructionmethodisinsufficientforthevectordataprocessing．Thispaperleveragedthecloudcomputingenvi-ronmentforspatialindexconstructionwiththeaimtoimprovetheconstructingspeed．ThepaperatthebeginninggaveareviewonthenormalspatialindexandspatialdatamanagementbasedonMapＲeducecloudingcomputingmodelthenanalyzedthespatialindexconstructionincloudcomputingtakingtheＲtreeindexconstructionasanexample．AtlastanexperimentresultsillustratedthatthespatialindexconstructingspeedissignificantlyimprovedincloudcomputingcomparedwithsinglePCenvironmentandinthesamecloudenvironmentgridfileconstructionspeedishigherthanＲtreeandＲ+treeindex．KeywordscloudcomputingspatialindexbuiltremotesensingvectordataＲtreeＲ+treegridfile0、、。。。、。、。GIS1－3。4GIS。、。3810201510GEOMATICS＆SPATIALINFOＲMATIONTECHNOLOGYVol．38No．10Oct．20151MapＲeduce。1．1、。。MBGB。TigerOpenStreetMapTigerTigerArealandmarkarealm、AreaHydrographyareawater、AllEdgesCombinededges、LinearHydrographylinearwater、PrimaryＲoadsprimaryroads55－DigitZIPCodeTabulationAreazcta510。OpenStreetMapAllPointsontheplanetAllNodes、AllextractedmapobjectAllObjects、BoundariesofallbuildingBuildings、BoundariesofcemeteryCemetery、BoundariesofparksorgreenareasParks、BoundariesofparksorgreenareasParks、BoundariesofpostalcodeareasPostalcodesＲoadnetwork、BoundariesofsportingareasSportsＲoadnetworkrepresentedasindi-vidualroadsegementsＲoads。12012Ti-gerOpenStreetMap。1Fig．1Ｒemotesensingvectordata114500G。2Ti-gerOpenStreetMap。2Fig．2Numberofremotesensingvectordata2。1．2。、ＲＲ+5。。ＲB。ＲB+。Ｒ、。Ｒ。6。Ｒ+ＲＲ“”Ｒ+。3。1．3MapＲeduceMapＲeduce7。、。MapＲeduceMap、Ｒeduce、Shuffle、Merge。44120153Fig．3Spatialdatadistributionisnotuniform．MapＲeduce。4MapＲeduceFig．4MapＲeducestagedivisionMapMapＲeduceＲeduceMap。MapkeyＲeducer。Mapkeyivalue()i→listkeymvalue()mＲeducelistkeymvalue()m→keyovalue()oMapＲeduceMapＲeduceMapＲeduce。、。8MapＲeduceＲ－k－。CG_Hadoop9MapＲeduce、、、。HydroCloud。SpatialHadoop10。Hadoop、、Ma-pＲeduce。HadoopGIS11Hadoop。MNTG12。100GB。Ｒ+。TAＲEEG13OpenStreetMapMNTGGoogleKMLESＲI。MapＲeduceOpenStreetMap。Pigeon。Ｒ+。SHAHED14NASANASA。SHA-HED。。。2。2．1ＲＲ。Ｒ。Hadoop。Ma-pＲeduceＲ3。1。Ｒ。Map15ＲeduceMapＲ－1Ｒ5。2。Ｒ16Ｒ。MapＲeduce1Map()fxＲeduceＲ。MBＲ。3。2ＲHadoop6。2．2Ｒ+。MapＲeduceＲ51105Fig．5Spaceobjectblockcalculationstep6ＲFig．6MergeandgeneratethefinalＲtree31。Ｒ+。Ｒ+Ｒ。Ｒ+。Ｒ+3Ｒ1Ｒ+。33．1HadoopVirualBox3Hadoop。Ubuntu10．04LTS1024MB64MB。。csv。MapＲeduce。3．2Hadoop。Hadoop31232、62464、93696、124928、156160。。1Ｒ78。7HadoopＲFig．7ＲindexconstructiontimecomparisonbetweensinglemachineandHadoopcluster8ＲFig．8ConstructionofＲindexfordifferentspaceobjectsHadoop28～45Ｒ。2Hadoop、ＲＲ+9。9HadoopFig．9IndexconstructiontimeofdifferentspaceobjectsinHadoopcluster9612015GridＲＲ+。4、。。HadoopMapＲeduce。ＲＲ+。。1．J．201329425－29．2MadduriKBaderDABerryJWetal．Parallelshor-testpathalgorithmsforsolvinglarge－scaleinstancesC．9thDIMACSImplementationChallenge———TheShortestPathProblemNewJerseyUSAＲutgersUni-versity20061－39．3WaughTCHopkinsS．AnalgorithmforpolygonoverlayusingcooperativeparallelprocessingJ．InternationalJournalofGeographicalInformationScience199266457－467．4．J．2013362209－225．5．J．2003201212－14．6．M．2009．7DeanJGhemawatS．MapＲeducesimplifieddatapro-cessingonlargeclustersJ．CommunicationsoftheACM2008511107－113．8．MapＲeduceＲ－k－J．20132481836－1851．9EldawyALiYMokbelMFetal．CG_Hadoopcom-putationalgeometryinMapＲeduceC．Proceedingsofthe21stACMSIGSPATIALInternationalConferenceonAdvancesinGeographicInformationSystemsOrlandoFLUSAACM2013294－303．10EldawyAMokbelMF．SpatialHadoopAMapＲeduceFrameworkforSpatialDataC．IEEEInternationalCon-ferenceonDataEngineeringICDE2015SeoulSouthKoreaIEEE2015．11AjiAWangFVoHetal．Hadoopgisahighper-formancespatialdatawarehousingsystemovermapreduceC．ProceedingsoftheVLDBEndowmentTrentoIta-lyVeryLargeDatabaseEndowment20131009－1020．12MokbelMFAlarabiLBaoJetal．MNTGanextensi-bleweb－basedtrafficgeneratorM．Springer2013．13AlarabiLEldawyAAlghamdiＲetal．TAＲEEGaMapＲeduce－basedwebserviceforextractingspatialdatafromOpenStreetMapC．Proceedingsofthe2014ACMSIGMODinternationalconferenceonManagementofdataNewYorkNYUSAACM897－900．14EldawyAMokbelMFAlharthiSetal．SHAHEDAMapＲeduce－basedSystemforQueryingandVisualizingSpatio－temporalSatelliteDataC．InternationalConfer-enceonDataEngineeringICDE2015SeoulSouthKo-reaIEEE2015．15CaryASunZHristidisVetal．Experiencesonpro-cessingspatialdatawithmapreduceC．ScientificandStatisticalDatabaseManagementNewOrleansLAUSASpringer2009302－319．16LiuYJingNChenLetal．Parallelbulk－loadingofspatialdatawithMapＲeduceAnＲ－treecaseJ．Wu-hanUniversityJournalofNaturalSciences2011166513－519．121．ArcGISEngineJ．2010296866－869．2．ArcGISJ．2010458－60．3．J．2009313－15．4．ArcGISJ．2011203145－150．5．GISJ．201447448－53．6．SketchUpArcGISJ