简介:Real-timevideotransportoverwirelessInternetfacesmanychallengesduetotheheterogeneousenvironmentincludingwirelineandwirelessnetworks.Arobustnetworkconditionclassificationalgorithmusingmultipleend-to-endmetricsandSupportVectorMachine(SVM)isproposedtoclassifydifferentnetworkeventsandmodelthetransitionpatternofnetworkconditions.End-to-endQuality-of-Service(QoS)mechanismslikecongestioncontrol,errorcontrol,andpowercontrolcanbenefitfromthenetworkconditioninformationandreacttodifferentnetworksituationsappropriately.TheproposednetworkconditionclassificationalgorithmusesSVMasaclassifiertoclusterdifferentend-to-endmetricssuchasend-to-enddelay,delayjitter,throughputandpacketloss-ratefortheUDPtrafficwithTCP-friendlyRateControl(TFRC),whichisusedforvideotransport.Thealgorithmisalsoflexibleforclassifyingdifferentnumbersofstatesrepresentingdifferentlevelsofnetworkeventssuchaswirelinecongestionandwirelesschannelloss.Simulationresultsusingnetworksimulator2(ns2)showedtheeffectivenessoftheproposedscheme.
简介:为了分析一个浅广场的稳定性,掘,新弄弯的失败机制,代表机械特征和这类隧道的崩溃形式,被构造。基于限制分析和Hoek布郎非线性的失败标准的上面的界限定理,从虚拟工作率方程导出的支持的压力被认为是客观功能完成最佳的计算。由采用变化计算优化客观函数,支持的压力和一条浅方形的隧道的崩溃的块形状的一个上面的界限解决方案被获得。为了评估失败机制的有效性,在这份报纸求婚了,弄弯的失败机制计算的答案与当Hoek布郎非线性的失败标准被变换成Mohr库仑线性标准时,线性多重块失败机制计算的结果相比。支持的压力和崩溃的块形状上的岩石团参数的影响被讨论。
简介:Therealizationofanefficientopticalsensorbasedonaphotoniccrystalmetasurfacesupportingboundstatesinthecontinuumisreported.Liquidswithdifferentrefractiveindices,rangingfrom1.4000to1.4480,areinfiltratedinamicrofluidicchamberbondedtothesensingdielectricmetasurface.Abulkliquidsensitivityof178nm/RIUisachieved,whileaQ-factorofabout2000givesasensorfigureofmeritupto445inairatbothvisibleandinfraredexcitations.Furthermore,thedetectionofultralow-molecular-weight(186Da)moleculesisdemon-stratedwitharecordresonanceshiftof6nmperlessthanaInmthicksinglemolecularlayer.Thesystemexploitsanormal-to-the-surfaceopticallaunchingscheme,withexcellentinterrogationstabilityanddemonstratesalignment-freeperformances,overcomingthelimitsofstandardphotoniccrystalsandplasmonicresonantconfigurations.