本是基于EMD細(xì)分網(wǎng)絡(luò)的編碼，我們?cè)O(shè)計(jì)了一個(gè)細(xì)分網(wǎng)絡(luò)的網(wǎng)絡(luò)地址編碼。在這篇文章中，編碼EMD細(xì)分網(wǎng)絡(luò)集合的方法建議為IPV6。通過(guò)這種方法我們能記住每個(gè)服務(wù)器代表一個(gè)地理位置的空間地力模型來(lái)組織空間的數(shù)將GEOIPIPV6的地址中，再然后，IPV6網(wǎng)絡(luò)被VLSM根據(jù)EMD細(xì)分網(wǎng) IPV6地 細(xì)分網(wǎng)絡(luò)編緒AAA里的數(shù)據(jù)系的方法。這種方法是通過(guò)GEOIP去實(shí)現(xiàn)地圖關(guān)系。GEOIP的設(shè)計(jì)規(guī)則和EMD細(xì)分網(wǎng)絡(luò)編碼與IPV6地址結(jié)合的方法介紹如下。EMDEMD模型通過(guò)精度和緯度間隔組成分級(jí)網(wǎng)絡(luò)，基于傳統(tǒng)的測(cè)繪部門的方式22部分：每一個(gè)部分是第一級(jí)細(xì)88088°圈北極是W1803620個(gè)細(xì)分網(wǎng)格，編碼20細(xì)分網(wǎng)格從南到北，從西到東的順序。第一層的細(xì)分網(wǎng)絡(luò)和他們的編碼設(shè)為133部分。前一層的細(xì)分網(wǎng)絡(luò)均等的分成四份。細(xì)分網(wǎng)絡(luò)的編碼在GIG中展示。編碼是0和21的兩個(gè)第一層的細(xì)分網(wǎng)絡(luò)根據(jù)正多邊形準(zhǔn)則被分成正三角形。DUTTON-ZOT提出來(lái)的方法，將正多邊形投影在全球。用直線把各個(gè)路口相連。路口穿過(guò)北緯88度和十條第二層細(xì)分網(wǎng)絡(luò)的經(jīng)線。然后連接路口和北極。細(xì)分網(wǎng)絡(luò)的編GeoIP的是基于細(xì)分網(wǎng)格的編碼。細(xì)分網(wǎng)格由二進(jìn)制編碼編碼。22占用2位。274cm275+2+6+22+42×21=6364位的長(zhǎng)整數(shù)，以確定的細(xì)分代碼的長(zhǎng)度是不同的，與現(xiàn)有的長(zhǎng)度固定IP地址的。最長(zhǎng)的細(xì)分代碼為64位，比IPv4地址的長(zhǎng)度。細(xì)分代碼有這么多的地方和不同的部分有不同的意義，這是從現(xiàn)有的IP地址不同。GEOIP設(shè)計(jì)策略細(xì)分網(wǎng)格代碼。GeoIP5EMD細(xì)分網(wǎng)格代碼來(lái)識(shí)別細(xì)分級(jí)別前，我們稱這部分EMD5位。上述六個(gè)級(jí)別細(xì)分網(wǎng)格由然占據(jù)了2位。GeoIPGeoIP68比特。當(dāng)代碼的長(zhǎng)度是不夠的，完整的代碼，其中x（在本文中，x是指以二進(jìn)制編碼的任何值。x1GeoIPx比特，在細(xì)分掩模的對(duì)應(yīng)位被設(shè)置為0。（在碼是從0到31，所以第二細(xì)分級(jí)別代碼標(biāo)記為0000100001成01100-01前面。在此步驟之后的當(dāng)前代碼是00001-01100-01，小于68位，其中x完成GeoIP00001-01100-01-XX.XX（01，第XXXXx比特的縮寫(xiě)。根據(jù)細(xì)分面具的設(shè)計(jì)策略，該網(wǎng)格的細(xì)分面具是11111-11111-11-0000（00000位的縮寫(xiě)ANDGeoIP因?yàn)榧?xì)分面具可以掉不固定位-XX.XX，每一個(gè)細(xì)分網(wǎng)格有幾個(gè)GeoIP以使一個(gè)細(xì)分網(wǎng)格覆蓋多個(gè)節(jié)點(diǎn)，這是非常有幫助的地理為基礎(chǔ)的模- 的地理為基礎(chǔ)的模態(tài)中的應(yīng)務(wù)器。我們結(jié)合每一個(gè)細(xì)分網(wǎng)格與服務(wù)器的地址GeoIP的。即位于細(xì)分網(wǎng)格的空間數(shù)據(jù)由節(jié)點(diǎn)管理。根據(jù)EMD編碼模型，我們可以得到不同的細(xì)分之間的特殊關(guān)系，包括水平。因此，每一個(gè)節(jié)點(diǎn)（除根節(jié)點(diǎn)）服務(wù)器可以很容易地找其對(duì)應(yīng)的細(xì)分網(wǎng)格00101-10010-01-10100-01-01-0101-XXXX。根據(jù)特GeoIP0010010010-01-10100-01-01-XXXXGeoIP00110-10010-01-10100-01-01-0101-00-XXXX，0011010010-01101000101010101XXXX，0011010010-01-10100-01-01-0101-10-XXXX，00110-10010-01-10100-01-01-010111XXXX。我們可以通過(guò)上面提到的特殊的包含關(guān)系轉(zhuǎn)移各個(gè)節(jié)點(diǎn)的地址。動(dòng) BINES使用IPV6的方GeoIP68IPv4GeoIP基本研究，IPv6地址包128位。要結(jié)GeoIPIP地址68位的IPv6GeoIPGeoIPIPv6網(wǎng)絡(luò)通過(guò)IPv6地址，我們采用的是：12AB：0：0：CD30::/60，GeoIP的是應(yīng)用IPv6地址的主機(jī)地址。子網(wǎng)地址的結(jié)構(gòu)如下：絡(luò)。因此，每個(gè)子網(wǎng)可以代表一些細(xì)分15mesh。在本文中，IPv6地址分為三部號(hào);6670ID;58ID。其他級(jí)別的子網(wǎng)地址的構(gòu)顯示在圖5。IDID標(biāo)識(shí)該細(xì)分水平;2ID標(biāo)識(shí)第一細(xì)分網(wǎng)格序列號(hào);在ID2ndsubdivision網(wǎng)格序列號(hào);4級(jí)子網(wǎng)地址的子網(wǎng)ID標(biāo)識(shí)第三細(xì)分網(wǎng)格序列號(hào);在左側(cè)水平子網(wǎng)ID是通過(guò)類比來(lái)定義。IPv6GeoIP的到來(lái)，它的目的地址進(jìn)行查找路由LAN，當(dāng)子網(wǎng)介紹，路由表發(fā)生變化，增加的形式（這個(gè)網(wǎng)絡(luò)，子網(wǎng)，0此網(wǎng)一些實(shí)驗(yàn)結(jié)本文比較了GeoIP的使用DNS在以下三個(gè)方面：DNSEMDIP地址之間的映射關(guān)GeoIPIP地址，所以，GeoIP當(dāng)我們需要增加一個(gè)新的節(jié)點(diǎn)中，DNS解決方案，需要更新DNS斷絕記錄，GeoIPGeoIP的需要不是維持在DNS服務(wù)器的數(shù)據(jù)。通過(guò)GeoIP的，我們只需要小于1ms，從EMD細(xì)分網(wǎng)格代碼獲取IP地址，量級(jí)只有0.001ms的。總結(jié)和展雖然轉(zhuǎn)化細(xì)分成網(wǎng)格的長(zhǎng)度固定的GeoIP，我們插入的GeoIP到當(dāng)前的IPv6地GeoIPIPv6IPv6和地理位置空間，其中至少包括268主機(jī)ID?！?×4×9×421）=523991×249IPv6IPv6果我們發(fā)現(xiàn)了第27個(gè)分區(qū)從根節(jié)點(diǎn)，我們需要擊潰27倍。Proceedingsofthe201014thInternationalConferenceonComputerSupportedCooperativeWorkinDesignandResearchonDongInstituteofRemoteSensingandGIS,PekingUniversity,Beijing100871InstituteofRemoteSensingandGIS,Peking-BasedonthecodingofTheEMD(Theextendedmodelbasedonmapdivision)subdivisionmesh,wedesignedannetworkaddresscodingofsubdivisionmesh(GeoIP).Inthispaper,themethodthatthecodingofTheEMDsubdivisionmeshcombineswithIPv6isproposed.ThroughthemethodwecanorganizeSpatialDatabythestoragearchitecturemodelthateveryserverstandsforageographicarea.Themainideaisfollowing:first,extendthelengthofEMDsubdivisionmeshcodeto68bits;theninserttheGeoIPintoIPv6address,afterthattheIPv6networkisdividedbyVLSM(VariableLengthSubnetMask)inaccordancewiththeDivisionstrategyoftheEMDsubdivision

ChengChengqi,GuoShideInstituteofRemoteSensingandGIS,PekingUniversity,Beijing100871InstituteofRemoteSensingandGIS,Pekingdivisionwaytoachievetheobjectivesofdirectstorageandindexforexistingspatialdata. mesh.Bythisway,everySubnetoreverynetworknodecan forsomesubdivisionmesh,wecanrealizethegoalthatorganizetheglobalSpatialDatabygeographicposition,sothattheglobalSpatialDatacanbemanagedinorderandinahighly-

efficientway

Fig.1(a)Thefirstlevel Fig.1(b)ThesecondandthethirdCurrently,withtherapiddevelopmentofremote,theamountofthespatialdatais

dramatically.However,therearesomanyproblemsinthe

organizationandmanagementofthespatialdata.Themainproblemis:Thecurrentmanagementofthespatialdataisthatthedataisstoredintimeorder,forexample,therecentdaysdataisrecordedinonedevice,severaldayslater,thedatawillberecordedinanotherdevice.Whenwerequirethesameareadata,weneedtoaccessmanystoragedevices.Itisverydifficultandtime-consumingtogetasameareadata.Tosolvethisbigproblem,Goodchild[2]hasproposedanewmodeltoorganizespatialdatabygeographic-basedservers.Themodeldefinedasthis:storeallthedatainfieldAbythecorrespondingsever:severA.WhenusersrequiredatainfieldA,ThemodelcansearchthedataonlyinserverA.Bythismodelwecanimprovethemanagementefficiencyoftheglobalspatialdatagreatly.Thispaperpointsoutthewaytoindicatethemaprelationshipbetweengeographicalareasandtheservers’networkaddresses.ThewayisthroughGeoIPtoachievethemaprelationship.TheGeoIPdesignprincipleandthemethodthatEMDsubdivisionmeshcodecombineswithIPv6addressareintroducedasfollows.EMDTheEMDmodelmadethehierarchicalsubdivisionlongitudeandlatitudeinterval,basedontraditional

Fig.1(c)Theforthlevel Fig.1(d)ThefifthlevelFig.1(e)Thesixthlevel Fig.1(f)ThelevelsabovesixthFigure1.ThesubdivisionmethodofInthismodel,thefirstlevelsubdivisiondividestheglobalinto22parts:everypartisafirstlevelsubdivisionmesh.Fromthesouthlatitude88°circletotheSouthPoleisthesubdivisionmeshwhosenumberis0.Fromthenorthlatitude88°circletotheNorthpoleisthesubdivisionmeshwhosenumberis21.Thelineofsouthlatitude88°,thelineofnorthlatitude88°andtheequatorsplittheglobalsurfaceintotwopart.StartfromW180°,every36°isasubdivisionmesh,wecangetsubdivisionmeshes,encodethe20subdivisionmeshesfr978-1-4244-6763-1/10/$26.00?2010southtonorth,fromwesttoeastinorder.Thefirstlevelsubdivisionequallydivideseverysecondsubdivisionmeshinto33parts.Thesixthlevelsubdivisionequallydividesthefifthsubdivisioninto9parts.Exceptthethirdlevelandthesixthlevel,alltheothersubdivisionlevelsequallydividetheformerlevelsubdivisionmeshinto4parts.ThecodingofsubdivisionmeshisshowinginFig.1[3].Thetwofirstlevelsubdivisionmeshwhosecodeis021aresubdividedintotrianglesFollowthe"regularpolyhedron"principle.BythemethodproposedbyDutton-ZOT(ZenithalOrthoTriangular)projection,Projectthe"regularpolyhedron"ontotheglobal.Connecttheintersectionswithstraightline,theintersectionsarethecrossingpointsofthenorthlatitude88°lineandthe10linesofsecondlevelsubdivisionlongitude.Andthenconnecttheintersectionsandthearctic.SodoesthefirstlevelsubdivisionmeshofAntarctica.Encodethenextlevelsubdivisionmeshesfromsouthtonorth,fromwesttoeast.TheDetailsareinReference[3].CodingofthesubdivisionTheGeoIPdesignedinthispaperisbasedonthecodingofthesubdivisionmeshes.Thesubdivisionmeshesarecodedbybinaryencoding.Thesubdivisionmeshescodeconsistsoftwoparts:TheformersixlevelscodingandtheabovesixlevelsTheformersixlevelsareencodedbyLinesequence,thisparttakesup21bits.Thefirstlevelsubdivisioncodetakesup5bits,thesecondlevelsubdivisioncodetakesup2bits,thethirdlevelsubdivisioncodetakesup6bits,thelevelsubdivisioncodetakesup2bits,thefifthlevelsubdivisioncodetakesup2bits,thesixthlevelsubdivisioncodetakesup4bits.TheabovesixlevelscodingareencodedbytheHilbertcurvetoidentitythespatialsequence.Everylevelinthisparttakesup2bits.Tillthe27thlevelsubdivision,thelargestsubdivisionmeshisabout4cm2,andthe27thlevelsubdivisionmeshcodeneed5+2+6+2+2+4+2×21=63bits.Inordertoexpressconveniently,weuseaLongintegerof64bitstoidentifythecentimeterlevelsubdivisionmeshes.Thestructureofthesubdivisioncodeisas1st3rd6thAbove6level6 Figure2.TheStructureofEMDSubdivisionMeshTherearesomeproblemsinaddressingbylocationinformationinnetwork,usingthiscodetomatchwithThelengthofdifferentlevelsubdivisioncodeisdifferent,incontradictionwiththeexistinglength-fixedIPaddress.

Thelongestsubdivisioncodeis64bit,whichislongerthanthelengthofIPv4address.Thesubdivisioncodehassomanypartsanddifferentparthasdifferentmeaning,whichisdifferentfromtheexistingIPaddress.DesignStrategyofInordertoachievethepurposethatpeoplecanaddressbylocation,wedesignGeoIPbasedontheEMDsubdivisionmeshcode.ThedesignStrategyofGeoIPisfollowing:First,insert5bitsintothefrontoftheEMDsubdivisionmeshcodetoidentifythesubdivisionlevel,wecallthispartsubdivisionlevelcode.Second,thewholeoftheEMDsubdivisionmeshcodeismovedbackward5bits.TheabovesixlevelssubdivisionmeshesareencodedbylinesequenceinsteadoftheHilbertcoding.Althoughwechangedtheencodingschemesoftheabovesixlevelssubdivisionmeshes,butthelengthofsubdivisionmeshescodeisnotchange,aftersixthSubdivisionlevel,everysubdivisionlevelstilltakesup2bits.ThestructureofGeoIPisaslevel1st2nd3rd6thAbove6level605101218202226- Figure3.ThestructureofextendedbinaryThelengthofGeoIPisfixed,allGeoIPare68bits.Whenthelengthofthecodeisnotenough[13],completethecodewithx(Inthispaper,xmeansanyvalueinbinaryencoding).Foreachsubdivisionmesh,correspondingwiththeGeoIP,thereisasubdivisionmask.Allofthenon-xbitsintheGeoIP,Thecorrespondingbitinsubdivisionmaskissetto1.AllofthexbitsintheGeoIP,Thecorrespondingbitinsubdivisionmaskissetto0.Forexample,asecondlevelsubdivisionmeshwhosesubdivisioncodeis01100-01(InordertoobserveEasily,insert“-”betweendifferentsubdivisionlevels),thesubdivisionlevelis2,sothesubdivisionlevelcodeoftheGeoIPis00001(thecodeisfrom0to31,sothesecondsubdivisionlevelcodeismarkedas00001),insert00001intothefrontof01100-01.Afterthisstepthecurrentcodeis00001-01100-01,lessthan68bits,completeitwithx,finallytheGeoIPwegetis00001-01100-01-xx……xx(xidentify0or1,xx……xxistheabbreviationsofthexbits).Accordingtothedesignstrategyofsubdivisionmask,thesubdivisionmaskofthismeshis11111-11111-11-00….00(00….00istheabbreviationsofthe0bits).ThenusethelogicalANDbetweentheGeoIPandthesubdivisionmesh,theresultis00001-01100-01-00…00.-xx….xx,everysubdivisionmeshhasseveralGeoIPandthecorrespondingsubdivisionmask.AGeoIPbelongstooneandonlyonesubdivisionmesh.(Thisfeaturecanmakeasubdivisionmeshcoverseveralstoragenodes,itisveryhelpfultothegeographic-basedstoragemodel).ThesubdivisionmeshcodeandthecorrespondingGeoIPofsomeformer3levelsubdivisionmeshasfollows:TABLE someGeoIPlevelMesh100000-11111-100000-11111-100000-11111-100000-11111-100000-11111-20000-00001-00-11111-11-20000-00001-01-11111-11-20000-00001-10-11111-11-20000-00001-11-11111-11-30000-00010-11111-01-11-30000-00010-11111-01-11-30000-00010-11111-01-11-Differentpartintheabovetableisseparatedby“-”,themeaningofeverypartdefinedinFig.3.TheGeoIPisencodedbythedesignstrategy.TheapplicationofGeoIPinthegeographic-basedstoragemodal

10010-01-10100-01-01-0101-xx….xx.Accordingtothespecialinclusionrelationship,thecorrespondingGeoIPofitsfathernodeis00100-10010-01-10100-01-01-xx….xx,itscorrespondingGeoIPofchildnodesare00110-10010-01-10100-01-01-0101-00-xx….xx,00110-10010-01-10100-01-01-0101-01-xx….xx,00110-10010-01-10100-01-01-0101-10-cantransfertheaddressbetweenvariousnodesbythespecialinclusionrelationshipmentionedabove.BythespecialinclusionrelationshipofdifferentGeoIP,WhentheGeoIPthattheaccessdoesnotmatchwiththeGeoIPofthecurrentsever,wecangettheshortestpaththatfromthecurrentsevertotheseverwhichincludestherequireddata,soanynodeservercanbefoundinthisway.Hencethisgeographic-basedstoragestructureorganizesthespatialdatainthespace-timeorder.Thisgeographic-basedstoragestructurealsocanprovidethesupportinfastsearchingthespatialdataanddistributingthedatathe bineswithThelengthofGeoIPis68bits,[9w],hichislongerthanthelengthofthecurrentIPv4addressanddifferentpartwechosetheIPv6asthebasictoresearch,theIPv6addressincludes128bits[4][6].TocombinetheGeoIPwiththeIPaddress,thelast68bitsinIPv6addressareinsteadofGeoIP.ThenwedividetheIPv6networkbyVLSM(VariableLengthSubnetmask))inaccordancewiththeGeoIPdivisionstrategy.So,everysubnetcanstandforsomesubdivisionmesh.Forexample,supposetheIPv6addressweappliedis:12A:0:0:CD30::/60,GeoIPisthehostaddressoftheappliedIPv6address.Thesubnetaddressstructureisfollowing: 60 1stlevelsubnetaddressHostHostID(58Figure4.TheEMDSubdivisionMeshandtheCorrespondingnetworknetworkHostID(56networkIntheorganizationandmanagementframeworkoftheglobalspatialdata,everysubdivisionmeshhasacorrespondingstorageserver[5].WecombinetheGeoIPworknetworkHostID(56network

3rdlevelsubnetaddress4thlevelsubnetaddress27stlevelsubnetaddress

HostID(50inclusionrelationshipbetweendifferentsubdivisionlevels.Soeverynode(excepttherootnode)servercaneasilyfinditsfathernodeaddressandchildnodesaddress.Forexample,supposethereissomesixthlevelstoragenode,theGeoIPofwhosecorrespondingsubdivisionmeshis00101-

Figure5.ThestructureoftheIPv6addresswithGeoIPDividetheIPv6networkbyVLSM(VariableLengthSubnetmask))inaccordancewiththeGeoIPsubnetdivisionstrategy[10].So,everySubnetcanstandforsomesubdivisionmesh.Inthispaper,theIPv6addressisdividedintothreeparts:networkID,subnetIDandthehostID.Wedefined27kindsofsubnetsbythedifferentlengthofthethreeparts.The27kindsofsubnetsisshowedinFig.5.Inthe1stlevelsubnetaddress,theformer60bitsareNetworkID;thebitsfrom61thbitto65thbitaresubnetID;theleft63bitsareHostID.Inthe2ndlevelsubnetaddress,theformer65bitsareNetworkID;thebitsfrom66thbitto70thbitaresubnetID;theleft58bitsareHostID.TheThesubnetIDindifferentlevelsubnetaddresshasthedifferentmeaning:thesubnetIDin1stlevelsubnetaddressidentifiesthesubdivisionlevels;thesubnetIDin2ndlevelsubnetaddressidentifiesthe1stsubdivisionmeshserialnumber;thesubnetIDin3rdlevelsubnetaddressidentifiesthe2ndsubdivisionmeshserialnumber;thesubnetIDin4thlevelsubnetaddressidentifiesthe3rdsubdivisionmeshserialnumber;thesubnetIDintheleftlevelsaredefinedbyogy.TheappliedIpv6

WhenanpacketofIPv6whichisincludedtheGeoIParrives,itsdestinationaddressislookedupintheroutingtable.Ifthepacketisforadistantnetwork,itisforwardedtothenextrouterontheinterfacewhichisgivenintheroutingtable.Ifitisalocalhost(e.g.,ontherouter'sLAN),itissentdirectlytothedestination.Ifthenetworkisnotpresent,thepacketisforwardedtoadefaultrouterwithmoreextensiveroutingtables[12].Thisalgorithmmeansthateachrouteronlyhastokeepthetrackofothernetworksandlocalhosts,not(network,host)pairs,greatlyreducingthesizeoftheroutingtable[1].Whensubnettingisintroduced,theroutingtablesarechanged,addingentriesoftheform(this-network,subnet,0)and(this-network,this-subnet,host).Thus,arouteronsubnetkknowshowtogettoalltheothersubnetsandalsohowtogettoallthehostsonsubnetk.Itdoesnothavetoknowthedetailsabouthostsonothersubnets.Infact,allthatneedstobechangedistohaveeachrouterdoaBooleanANDwiththenetwork'ssubnetmasktogetridofthehostnumberandlookuptheresultingaddressinitstables(afterdeterminingwhichnetworkclassitis)[1].

SOMEEXPERIMENTALFromtheEMDsubdivisionmeshcodetotheseverIP0 2ndlevel2nd

0123012301233rdIdentifythesubnetdividedby

3rdlevel3rdlevel

IfweuseDNStosolvethe subdivisionmeshcodeandtheseverIPaddress,therewill0GeoIPsubnetDivisionoftheSubdivisionmesh

32

someunreliableelements[11][14].WhenthereissomethingwrongabouttheDNSserver,thewholesystemwillbedestroyed[7].TheGeoIPcangettheserverIPaddressbytheFigure6.ThestructureofthenetworkThecorrespondingstructureofthesubdivisionmeshorganizationisasfollowing:001Allthe1stTheNO0meshofthe2rdsubdivisionTheNO1meshoftheTheNOmeshoftheTheNO3meshofthe2rdsubdivision001Figure7.ThecorrespondingstructureoftheSubdivisionmesh001Inthenetworkconstruction,thesubnetnodesinFig.6arecorrespondingtotherouters,andthestoragenodesinfig.6arecorrespondingtothestorageservers.

abovecalculatingmethod,so,GeoIPhasHighApplicationandWhenweneedtoaddanewstoragenode,theDNSsolutionneedtoupdatetheDNSseverrecord,theGeoIPsolutiononlyneedstosettheaddressofthenewnode.AndGeoIPneedsnottomaintainthe namedataintheDNSserver.ThroughGeoIPweonlyneedlessthan1mstogettheIPaddressfromtheEMDsubdivisionmeshcode,theorderofmagnitudeisonly0.001ms.theSummaryandThoughtransformingthesubdivisionmeshintothelength-fixedGeoIP,weinserttheGeoIPintothecurrentIPv6address.AndweproposedthemethodbywhichwecancombinetheGeoIPwithIPv6address.BythecurrentnetworkaddressingtechnologyofIPv6andthegeographicallocationsmeaningoftheGeoIP,wecanrealizetheassumptionthatconstructingamodelorsysteminwhichthespatialdataisorganizedbygeographic-basedserverstructure.Therealsoexistssomeproblemsinthemethodthatwestudyinthefollowingaspects:First,thespaceoftheIPv6addressweneedtoapplyistoobig,weneedtoapplyaIPv6addressspacewhichatleastincludes268hostID.Second,usingthemethod,wecancombinetheGeoIPwithIPv6,therewillbe(268-22×4×3×4×4×9×421)=523991×249IPv6addressmayberedundant,toomanyIPv6addresswillbeidle.Third,thehigherthesubdivisionlevelis,thehigherthecorrespondingsubnetlevelis,andthedee