简介:在无线传感器节点的动态电源管理(DPM)是为减少闲散精力消费的一种著名技术。DPM由动态地基于事件出现的预言翻转它的单位的开/关地位控制一个节点操作模式。然而,后来,每个模式变化在它的自己的权利导致一些开销,保证DPM效率不是在展出有未知统计的非宿命论和无常的环境的吝啬的功绩。我们在这份报纸的解决方案套房,一起指了同样认知的电源管理(CPM),是向在统计上未知的设置的创新DPM的一次原则性的尝试并且给二不同分析保证。我们的第一个图案在面临非静止的事件过程时基于学习自动机和保证better-than-pure-chanceDPM。我们的第二个解决方案迎合事件出现可以在雇用一个对手的人物的甚至更一般的设置。在这种情况中,我们以节点在依靠一个no-external-regret过程以一种联机方式学习它的得最高分的战略策略的一场重复零和的比赛与它的环境提出单个尘埃的相互作用。我们进行数字实验以网络一生和事件损失百分比测量我们的计划的表演。
简介:目标范围是在无线传感器网络的一个重要还挑战性的问题,特别当范围和精力限制应该被考虑时。由于它的非线性的性质,这个问题的以前的研究主要集中了于启发式的算法;理论界限仍然保持未知。而且,最流行的方法在以前的文学使用了,即,连续时间的discretization,还得被认为正当。这份报纸用二理论结果填这些差距。第一个是为方法的一个正式理由。我们使用一个简单例子说明及时转变一个解决方案的过程领域进在有一样的网络一生的模式域的一个相应解决方案并且获得二关键观察。在那以后,我们正式证明这是二观察并且把他们用作基础认为方法正当。第二结果是能保证网络一生是的一个算法至少(1-)最佳的网络一生,在此能被使任意地小取决于要求的精确。算法基于列产生(CG)理论,它把原来的问题分解成二亚问题并且反复地以接近最佳的答案的一个方法解决他们。而且,我们开发了几条建设性的途径进一步优化算法。数字结果验证我们的基于CG的算法的效率。
简介:ForTDMAMACprotocolsinwirelesssensornetworks(WSNs),redundancyandretransmissionaretwoimportantmethodstoprovidehighend-to-endtransmissionreliability.Sincereliabletransmissionswillleadtomoreenergyconsumption,thereexistsanintrinsictradeoffbetweentransmissionreliabilityandenergyefficiency.Foreachlink,wenamethenumberofitsreservedtimeslotsineachMACsuperframeasareplicatorfactor.Inthefollowingpaper,weproposeareliability-lifetimetradeoffframework(...
简介:Basedonthesensingmechanismofmicrosensor,asimulationmodelofapracticalsiliconbeamresonatorattachedtoanE-typerounddiaphragmandusedformeasuringconcentratedforceisestablished.Therelationshipbetweenthebasicnaturalfrequencyofthebeamresonatorandtheconcentratedforceiscalculated,analyzedandinvestigated.AsamicrosensorFEMisusedtostudysomeimportantsimulationresultsonthevibrationfeaturesofthebeamresonators.Basedonthedifferentialoutputsignals,asetofoptimumparametersoftheproposedsensingunitisdetermined.
简介:Thispaperpresentsasymmetriccooperationstrategyforwirelesssensornetworks,aimingtoimprovethetransmissionefficiencyofthenetwork.Thecooperationstrategyisimplementedbypartitioningthenodesintoseveralcooperativegroups.Then,ineachgroup,theoptimalcooperativebandwidthallocationisobtainedbasedonRaiffa-KalaiSmorodinskybargainingsolution(RBS).Numericalresultsshowthatthesymmetriccooperationstrategycanimprovethesensornode'stransmissionefficiencydramatically.
简介:Thepaperproposesacooperativedistributedtargettrackingalgorithminmobilewirelesssensornetworks.Therearetwomaincomponentsinthealgorithm:distributedsensor-targetassignmentandsensormotioncontrol.Inthekeyideaofthesensor-targetassignment,sensorsareconsideredasautonomousagentsandthedefinedobjectivefunctionofeachsensorconcentratesontwofundamentalfactors:thetrackingaccuracyandthetrackingcost.Comparedwiththecentralizedalgorithmandthenoncooperativedistrib...
简介:Efficientestimationofpopulationsizeisacommonrequirementformanywirelesssensornetworkapplications.Examplesincludecountingthenumberofnodesaliveinthenetworkandmeasuringthescaleandshapeofphysicallycorrelatedevents.Thesetasksmustbeaccomplishedatextremelylowoverheadduetothesevereresourcelimitationofsensornodes,whichposesachallengeforlarge-scalesensornetworks.Inthisarticlewedesignanovelmeasurementtechnique,FLAKEbasedonsparsesamplingthatisgeneric,inthatitisapplicabletoarbitrarywirelesssensornetworks(WSN).Itcanbeusedtoefficientlyevaluatesystemsize,scaleofevent,andotherglobalaggregatingorsummationinformationofindividualnodesoverthewholenetworkinlowcommunicationcost.Thisfunctionalityisusefulinmanyapplications,buthardtoachievewheneachnodehasonlyalimited,localknowledgeofthenetwork.Therefore,FLAKEiscomposedoftwomaincomponentstosolvethisproblem.OneistheInjectedRandomDataDissemination(Sampling)method,theotherissparsesamplingalgorithmbasedonInverseSampling,uponwhichitimprovesbyachievingatargetvariancewithsmallerrorandlowcommunicationcost.FLAKEusesapproximatelyuniformrandomdatadisseminationandsparsesamplinginsensornetworks,whichisanunstructuredandlocalizedmethod.Atlastweprovideexperimentalresultsdemonstratingtheeffectivenessofouralgorithmonbothsmall-scaleandlarge-scaleWSNs.Ourmeasurementtechniqueappearstobethepracticalandappropriatechoice.
简介:Ahybridmethodologyisproposedforuseinlowpower,safetycriticalwirelesssensornetworkapplications,wherequality-of-serviceorientatedtransceiveroutputpowercontrolisrequiredtooperateinparallelwithradiofrequency-basedlocalization.Thepracticalimplementationisframedinanexperimentalproceduredesignedtotrackamovingagentinarealisticindoorenvironment.Anadaptivetimesynchronizedapproachisemployedtoensurethepositioningtechniquecanoperateeffectivelyinthepres...