〔原文〕無線傳感器網(wǎng)絡(luò)路由協(xié)議爭論普遍的網(wǎng)絡(luò)化、微機電系統(tǒng)(MEMS)和片上系統(tǒng)(SOC)的快速進展，以及這三Internet技術(shù)加上Ad—hoc路由機制對傳感器網(wǎng)絡(luò)的設(shè)計是足夠充分的，但深人心，后者以傳輸數(shù)據(jù)為目的。為了適應(yīng)廣泛的應(yīng)用程序，傳統(tǒng)網(wǎng)絡(luò)的設(shè)計遵循著網(wǎng)絡(luò)有幾個不同于傳統(tǒng)無線Ad—hoc網(wǎng)絡(luò)的特點，因此它的路由格外有挑戰(zhàn)性。有顯著的冗余性，因此可以利用數(shù)據(jù)聚合來提高能量和帶寬的利用率；第三，節(jié)點能量和處理存儲力量有限，需要精細(xì)的資源治理；最終，由于網(wǎng)絡(luò)拓?fù)渥兓l繁，需要路由協(xié)議有很好的魯棒性和可擴展性?，F(xiàn)有方案分等級的、基于位置的、基于網(wǎng)絡(luò)流的。以數(shù)據(jù)為中心的路由協(xié)議SPIN是第一種以數(shù)據(jù)為中心D．Estrin等人開發(fā)出一種突破性的路由協(xié)議肯定向集中(DirectedDifusion)sink組合定義一個興趣(interest)消息，并數(shù)據(jù)為中心路由協(xié)議合．SPIN是第一個數(shù)據(jù)為中心的路由議，定向集中DD(directeddifusion)消滅后，很多路由協(xié)議的提出都是基于DD或遵循相像思想的．定向集中協(xié)議定向集中DD【431化算法，能夠很好地滿足WSN對節(jié)能、魯棒性以及可擴展性的要求．但由于承受了按需查詢驅(qū)動的數(shù)據(jù)模型，所以不適用于環(huán)消耗更多的能量．基于規(guī)律模糊掌握器的多路徑數(shù)據(jù)為中心路由協(xié)議文獻[5]提出了一個的以數(shù)據(jù)為中心的多路徑路由方案，該方案人可以保證網(wǎng)絡(luò)從路徑失效中快速恢復(fù)以節(jié)約能量．層次路由協(xié)議簇的思想，馬上整個WSN區(qū)域分為假設(shè)干個簇，每個簇中選舉一個簇頭，通過多跳LEACH(1ow—energyadaptiveclusteringhierarchy)協(xié)議是最早的分簇協(xié)議，很多其他層次路由協(xié)議都是基于LEACH的．LEACHLEACH的根本思想是以循環(huán)的方式隨機選擇簇頭，將整個網(wǎng)絡(luò)的能量負(fù)載平均安排到每個傳感器節(jié)點，從而降低網(wǎng)絡(luò)能源消耗，提高網(wǎng)絡(luò)生存時間．LEACH協(xié)議每一輪循環(huán)可分為簇的建立和數(shù)據(jù)通信2個階段．在簇的建立階段，相鄰節(jié)簇頭進展數(shù)據(jù)融合并把結(jié)果發(fā)給會聚節(jié)點．與節(jié)點直接與會聚節(jié)點通信相比，LEACH協(xié)議降低了74—8利用LEACH算法選舉出的簇頭分布不均勻，不適合大規(guī)模網(wǎng)絡(luò)及能量不均衡的網(wǎng)絡(luò)．PEGASIS和分層PEGASISPEGASISLEACH根底上改進設(shè)計的，其目的是提高網(wǎng)絡(luò)的生存周與它們最近的鄰居之間進展通信．仿真結(jié)果說明，與LEACH協(xié)議相比，其生存周期提高近2倍．但離鏈?zhǔn)纵^遠(yuǎn)的節(jié)點會引起過多的數(shù)據(jù)延遲，而且鏈?zhǔn)坠?jié)點的唯Lindsey等提出的分層PEGASIS協(xié)議是對PEGASIS的擴展，承受數(shù)據(jù)并行傳輸機制，并利用CDMA方式避開信號干擾．仿真結(jié)果說明，分層PE—GASIS較PEGASIS使網(wǎng)絡(luò)的生存期提高了約60倍．HMRP文獻[8]提出了一個分層的多路徑路由協(xié)議HMRP(hierarchicalmultiple—choiceroutingpath)，該協(xié)議首先把WSN構(gòu)造成一個分層的網(wǎng)絡(luò)，基于此層次構(gòu)造，傳感器節(jié)點通過候選的雙親樹型構(gòu)造相比，延長了網(wǎng)絡(luò)的生存周期．地理位置路由協(xié)議在WSN的很多應(yīng)用中(如森林防火系統(tǒng))，節(jié)點通常需要獵取其位置信息．而在很多路由協(xié)議中，通常也需要位置信息來計算2個特定節(jié)點的距離以便估算所了能耗．MECNSMECN最小能量通信網(wǎng)絡(luò)MECN(minimumenergycommunicationnetwork)協(xié)議最初是為MANET設(shè)計的，但同樣適用于WSN．SMECN(smallminimumenergycommunicationnet—work)協(xié)議是MECN協(xié)議的擴展，它充分考慮到任意2個節(jié)點間可能存在障礙而導(dǎo)致節(jié)點間不能直接通信的狀況．MECN和SMECN協(xié)議都利用低功耗的GPS算和構(gòu)建能源有效的子網(wǎng)實現(xiàn)節(jié)能目標(biāo)．MECN協(xié)議為每個節(jié)點設(shè)置了轉(zhuǎn)發(fā)區(qū)域，發(fā)送節(jié)點通過轉(zhuǎn)發(fā)區(qū)域內(nèi)的節(jié)點發(fā)送比直接將數(shù)據(jù)傳遞到目的節(jié)點需要更少的連通，這在實際環(huán)境中不易做到．GAFGAF(geographicaladaptivefidelity)協(xié)議是以節(jié)點地理位置為依據(jù)的分簇協(xié)議，也是MANET的一種路由協(xié)議，但同樣適用于WSN．該協(xié)議把監(jiān)測區(qū)域劃分成假設(shè)干虛擬單元格，將節(jié)點依據(jù)其位置信息劃入相應(yīng)的單元格，在每個單元格中定期選舉產(chǎn)生1個簇頭，只有簇頭節(jié)點保持活動，其他節(jié)點進入休眠狀態(tài)，同一單元格的節(jié)點之間通過GPS標(biāo)識的位置信息進展通信．仿真結(jié)果說明，GAF在時延、丟包率和生命期等方面均優(yōu)于一般的MA—NET協(xié)議．但這種基于地理位置進展分簇的算法對傳感器網(wǎng)絡(luò)提出了更高的要求，而且GAF算法基于平面模型，沒有考慮到在實際網(wǎng)絡(luò)中節(jié)點之間的接近問題，P．Santi等人提出了一種改進GAF算法，設(shè)計了2種不同的簇頭選擇機制，并具體分析了簇頭節(jié)點產(chǎn)生后的網(wǎng)絡(luò)運行方式，與GAF相比，延長了傳感器網(wǎng)絡(luò)的生存周期．無線傳感器網(wǎng)絡(luò)路由協(xié)議比較分析WSNWSN路由協(xié)議在設(shè)計上應(yīng)充分考慮節(jié)點能量的有對各類WSN的路由協(xié)議進展比較分析，如表1所示．結(jié)論WSN因其自身的特點使其路由協(xié)議的爭論較之以往的網(wǎng)絡(luò)更為簡單．通過以上分析，可以看出WSN路由協(xié)議將來的爭論熱點和進展趨勢：立和特地選取節(jié)點作為簇頭的額外開銷，但在簡單查詢中該方法就顯得力不從一個有待解決的問題．頭負(fù)責(zé)數(shù)據(jù)融合并向會聚節(jié)點傳遞數(shù)據(jù)，使WSN中節(jié)點均衡了全網(wǎng)的能耗，且提是將來的爭論方向．在地理位置路由協(xié)議中，通過節(jié)點的地理位置信息，可以到達節(jié)約能量的目的．如何智能地利用地理位置信息來關(guān)心能量有效路由是一個值得探討的問還要考慮全網(wǎng)的通信負(fù)載平衡，路由協(xié)議的容錯性以及節(jié)點的移動性等．(譯文)forWirelessSensorNetworksResearchandSystem-on-Chip(SOC)oftherapiddevelopment,aswellmode,theisgenerallybelievedthatshort-rangelow-powerwirelesscommunicationtechnologybestsuitedtotheuseofsensorwirelesssensornetworks.Atfirst,peoplethoughtthatInternetstudyshowsthat:sensornetworksandtraditionalnetworkshavesignificantlydifferenttechnicalrequirements.Theformerdata.Inordertoadapttoawiderangeofapplications,designedtheedgerelatedtotheclientsystemonthenetwork,theintermediatenodeisonlyresponsiblefordatapacketforwarding,Forsensornetworks,thismaynotbeareasonablechoice.Inthewirelesssensornetworkarchitecture,networklayerandnetworkinfrastructurevary.Networklayeristhemainandreliabletoestablishtheroutingofthedatatransmissionmethod,sothatthelongestlifeexpectancynetwork.AsaresultofanumberofsensornetworksisdifferentfromtraditionalwirelessAd-hocnetworkcharacteristics,soitofmanynodes,itisimpossibletoestablishamechanismtoaddresssignificantredundancy,dataaggregationcanbeusedcapacity,asaresultoffrequentchangesinnetworktopology,routingprotocolsneedtohavegoodrobustnessandscalability.existingprogramsConsiderthecharacteristicsofsensornetworksandapplicationsandinfrastructureneeds,theresearchersputforwardmanyproposalstosolvethesensornetworkroutingproblems.Wecandividethemintofourcategories:data-centric,hierarchical,location-based,network-basedstream.data-centricroutingprotocolofthetargetaddresssignificantdifferences.SPINisthefirstdata-centricAdaptiveRoutingProtocol,itisaskedthroughthenodestoD.ofcertainroutingprotocols(DirectedDifusion)togenerateusedtodefineacombinationofattributesofinterest(interest)information,andData-centricRoutingProtocolnaminginquiriestoobtaininformationanddatatransmissionindatafusionprocess.SPINisthefirstdata-centricroutingyou,manyoftheproposedroutingprotocolsarebasedonDDorfollowsimilarideas.ProliferationofdirectionalorientationDD-proliferationagreementsdescribeDDintherunning,includingthepathtoestablish,asfarasshowninFigure1.DDintroducestheconceptofgradientnetwork,canbefoundfromthesourcenodetothetargetnodeofthetomeetrequirements.However,duetotheuseoftheon-demandquery-drivendatamodel,itdoesnotapplytoCentralnodeswillconsumemoreenergy.Fuzzylogiccontrollerbasedonthemulti-pathdata-centricofsensornodesbasedonbatteryenergylevelandthetypeofdatatonodestoavoidalargeamountofdatatransmission,effectivelyextendingthenetworklifecycle.Multi-pathmechanismandthefoodcanguaranteefailureinthenetworkenergy.2.2-levelroutingprotocolInordertoensurenetworkscalabilitywithoutreducingusetheofregionselectionofacluster,throughthemulti-hopcommunicationsandtransmissionofdatafusiontoreducetheamountofinformationtoclusternodes,thussavingenergy.LEACHoneofthefirstsub-clusteragreement,anumberofotherroutingprotocolsarebasedonthelevelofLEACH.LEACHcycleofthefirstcluster,theenergyoftheentirenetworkloadtheofthenetwork.LEACHprotocolforeachcyclecanbedividedintoclustersandtheestablishmentoftwostagesofdatatheadjacentsectioninthedatasenttotheclusterhead,clusterheadfordatafusionandtheresultsdistributednodeclustering.Withnode-nodeLEACHprotocoltoreduceenergyconsumptionby7times,withtheminimumtransmissionenergyroutingcomparedto4-8timeslowerpowerconsumption.ButtheuseofLEACH”sclusteralgorithmfortheelectionofthefirstnon-uniformdistribution,notsuitableforlarge-scalenetworksandenergyimbalance網(wǎng)絡(luò).PEGASISandstratifiedPEGASISLEACH,anditspurposeistoimprovethenetworklifecycle.LEACHalgorithmwithmulti-clusterstructure,PEGASISchaintheirmostrecentcommunicationbetweenneighbors.ThesimulationresultsshowthatcomparedwiththeLEACHprotocol,anditsfromthedelays,bottleneckofcommunicationnetworks.Lindseyputforwardbytheusingavoidingresultsaviewtoimprovingthesurvivalofabout60times.RoutingProtocolHMRP(hierarchicalmultiple-choicetoatieredsensornoderesultstocycle.LocationRoutingProtocolInmanyWSNapplications(suchasforestfirepreventionusuallyalsoneedtocalculatethetwospecificnodesinordertoknowtheirpurposeorobjectivesoftheregionalnodeofthethebasisforrouting,nodesforwardthedataaccordingtocertainreleaseinformationtothedesignatedarea,iteffectivelyreducesMECNandSMECNminimumenergycommunicationnetworkMECN(minimumenergycommunicationnetwork)protocolwastoWSN.SMECN(smallminimumenergycommunicationnet-work)fullytakeintoaccountanytwonodestheremaybearesultofobstaclestodirectcommunicationbetweennodescannotbethecase.SMECNagreementsMECNandlowpoweruseofGPSpositioningsystem,andbuildingenergybycalculatingthesetupforeachnodeforwardingtheregion,sentthroughtheforwardingnodesintheregionthanthenodetosenddataenergy.asendingofauto-configurationcanbedynamicallyadaptivenodefailureandnetworkdistribution.However,theagreementassumesthatthefullnetworkconnectivity,whichintheactualenvironmentisnoteasytodo.GAFGAF(geographicaladaptivefidelity)Agreementisbasedonlocation-basednodeclusteringagreementisaMANETroutingprotocol,butitisequallyapplicabletoWSN.Theagreementtomonitortheregionintoanumberofvirtualcells,nodesinaccordancewithitseachcellheadnodestate,thesamecellbetweenthenodelabeledGPSlocationthat,arebetterthanordinaryMA-NETprotocol.However,location-baseddemand,andtheGAFalgorithmisbasedontheplanarmodeldoesnottakeintoaccounttheactualnetworknodesinbetweendoesnotmeanthatitcanbeclosetodirectcommunication,andenergyofnodes.Inresponsetothisproblem,P.SantietalGAF,differentoftheclusterheadnodeofthenetworkaftertheoperation,comparedwiththeGAF,theextensionofthesensornetworklifecycle.Analysisdesignenergy.Furthermore,itshouldbeinaccordancewiththespecificcharacteristicsoftheapplicationtomeetthefollowingpaths,QoSWSNroutingprotocolbasedonanalysisofroutingprotocolsanalysisofroutingprotocol