论文网
参考文献类有关论文范文资料,与网络信息文化对中学生作文的影响相关论文提纲
本论文是一篇参考文献类有关论文提纲,关于网络信息文化对中学生作文的影响相关在职研究生毕业论文范文。免费优秀的关于参考文献及计算机应用及职工教育方面论文范文资料,适合参考文献论文写作的大学硕士及本科毕业论文开题报告范文和学术职称论文参考文献下载。
erusersandnotmuchdifferenceinedgeusers.ThetheoreticalanalysisandsimulationresultsshowthatICIR-UMBCEEAcaneffectivelyimprovetheenergyefficiencyofmunicationsysteminengineering.本文网址:http://www.sxsky.net/xie/070686677.html
针对接收信号强度值(RSSI)的时变特性降低定位精度的问题,提出了一种基于二维网格特征参数融合的室内匹配定位算法.该算法融合RSSI和信号到达时间差(TDOA)构建网格特征参数模型,基于二维网格快速搜索策略降低匹配定位的计算量,采用网格特征向量的归一化欧氏距离进行最优网格匹配定位,最终由匹配网格的参考节点计算终端的精确位置.定位仿真实验中,该算法在3m网格粒度下的定位均方根误差为1.079m,平均定位误差小于1.865m,3m定位精度下的概率达到94.7%,相对于传统单一RSSI模型法提高了19.6%.所提算法能够有效提高室内定位精度,同时减少搜索数据量,降低匹配定位的计算复杂度.
Focusedontheissuethatthetime-varyingcharacteristicofindoorReceivedSignalStrengthIndicator(RSSI)drasticallydegradesthelocalizationaccuracy,anindoormatchinglocalizationalgorithmbasedontwo-dimensionalgridcharacteristicparameterfusionwasproposed.ThealgorithmfusedreceivedsignalstrengthandTimeDifferenceofArrival(TDOA)parameterstobuildgridfeaturemodel,inwhichtwo-dimensionalgridquicksearchstrategywasadoptedtoreduceputationamount.NormalizedEuclideandistanceofgridfeaturevectorwasusedtorealizetheoptimalgridmatchlocalization.Finally,thepreciseterminallocationwasputedbyreferencenodesofthematchedgrid.Inthelocalizationsimulationexperiments,theproposedalgorithmachievedthelocalizationRootMeanSquareError(RMSE)at1.079m,andtheaveragelocalizationaccuracywaswithin1.865mintheconditionof3mgridgranularity,Theprobabilityof3mlocalizationaccuracyreached94.7%,whichwas19.6%higherthanthatoftraditionalmethodonlybawsedonRSSI.Theproposedalgorithmcaneffectivelyimprovetheindoorpositioningaccuracy,meanwhilereducesthesearchdataquantityandtheputationalplexityofmatchinglocalization.
针对容延/容断网络(DTN)网络的时延高,割裂频繁,以及节点缓存和能量受限等网络特性,为提高容延网络的传输率,同时降低网络开销和网络时延,提出了一种基于节点相似性的容延网络路由算法(RABNS).该算法利用历史相遇信息预测节点未来相遇概率,并且把历史相遇的节点录入为集合,利用集合的交集运算来评估一对相遇节点的相似性,并以此为判定条件控制网络中的副本数量.在模拟器TheONE上采用RandomWaypoint运动模型进行仿真,其中RABNS在消息投递率方面优于PROPHET,网络负载约为PROPHET的50%,较大程度上提高了网络资源利用率,平均时延稍高于Epidemic但低于PROPHET,节点缓存空间大小对算法的平均跳数影响不大,且RABNS的平均跳数约为PROPHET的一半.仿真结果表明,RABNS能有效地限制消息洪泛,获取更高的消息投递率,更低的网络开销和数据时延,因此尤其适用于节点存储空间有限的DTN环境和具有群居特性的社交容延网络中.
Delay/DisruptionTolerantNetwork(DTN)hascharacteristicsoflongdelay,intermittentdisruption,andlimitationofbufferspaceandenergy.Toimprovethedeliveryrateofmessages,whilereducingworkoverheadandtheaveragelatency,anewRoutingAlgorithmBasedonNodeSimilarity(RABNS)inDTNwasproposed.Thealgorithmusedhistoricalinformationtopredictnodeencounterprobabilityinfuture.Thenodeswhichencounteredhistoricallywererecordedasacollection,thenthesetintersectionoperationwasappliedtoevaluatethesimilarityofapairofnodes.Andthesimilaritywasusedtocontrolthenumberofcopiesinthework.SimulationswereconductedonTheONEplatformusingRandomWaypointmotionmodel.Inthesimulation,RABNSperformedbetterthanPROPHET(ProbabilisticROutingProtocolusingHistoryofEncountersandTransitivity)inthemessagedeliveryrate.AndtheworkoverheadofRABNSwasabouthalfofPROPHET,whichgreatlyimprovedtheutilizationofworkresources.TheaveragelatencyofRABNSwasalittlehigherthanEpidemicbutlowerthanPPROPHET,thenodecachesizedidnothaveasignificantimpactonaverage-hops,anditsaverage-hopswasabouthalfofPROPHET.ThesimulationresultsshowthatRABNScaneffectivelylimitthemessagefloodingwithhighermessagedeliveryrate,lowerworkoverheadandaveragelatency,thereforeitissuitablefortheDTNsceneswithlimitednodes'storageandalsoapplicableinsocialDTNwithgregariouscharacteristics.
为解决基于帕累托(Pareto)支配解排序的多目标进化算法高时间复杂度问题,依据非支配解排序潜在特性,介绍了一种快速的非支配解排序方法,每次只处理当前种群中最高等级个体,且在分配等级的同时,能选择个体进入下一代,下一代被选足时即结束程序,减少了排序处理个体的数量,大幅度降低时间复杂度,另外,给出一种均匀的拥挤距离计算方法,最后,将快速非支配解排序和均匀拥挤距离计算与微分进化算法结合,提出基于非支配解排序的快速多目标微分进化算法(FMODE).采用标准多目标优化问题ZDTl～ZDT4和ZDT6进行仿真实验:当种群个体较多(大于500)时,FMODE所用时间远小于NSGAⅡ,FMODE的总体性能上均优于经典的NSGAⅡ,SPEAⅡ和DEMO,在FMODE框架内,采用均匀拥挤距离在性能上也明显优于经典拥挤计算方法,并通过实验确定了FMODE算法的参数.实验结果表明FMODE能够减少计算等级时的处理时间,并在收敛性和多样性指标上明显优于对比算法.
Concerningthehightime-plexityofmulti-objectiveevolutionaryalgorithmbasedonParetodominatedsolutionsorting,consideringthepotentialfeaturesofnon-dominatedsolutionsorting,afastsortingmethodwhichonlyhandlesindividualswiththehighestrankincurrentpopulationwasintroduced.Theindividualscouldbechosenintothenextgenerationduringthesorting.Thealgorithmwasterminatedwhenthepopulationofnextgenerationwasselectedenough,whichreducedthenumberofindividualsforsortingprocessandthetimeplexity.Inaddition,amethodofuniformcrowdingdistancecalculationwasgiven.Finally,aFastMulti-ObjectiveDifferentialEvolution(FMODE)algo
参考文献类有关论文范文资料,与网络信息文化对中学生作文的影响相关论文提纲参考文献资料：