简介：Thetrafficbufferingproblemsintheethernetoversynchronousdigitalhierarchy(EOS)areintroducedandanalyzed.Differentsolutionsarealsopresentedindetail.SynchronousDRAM(SDRAM)isusedasoff-chipbuffertostore-and-retransmissionethernetframes.Anewandeasycontroldesignisintroducedhere.Thebufferareasizeonchipisgreatlyreducedandthepowerdissipationislowedatthesametime.

简介：Inthispaper,theauthorspresentamethodtohandletheEchoCancellerasanon-didejobofLD-CELPcodecandacircuitrytoembedechocanecellerintoaLD-CELPcodec.ThepossibilitytoimplementasystemwiththeintegrationofLD-CELPcodecandechocancellerinrealtimebytwochipsofTMS320C30isdiscussed.

简介：Inthispaper,anewrecursivealgorithmofadaptivearrayweightvectorsanditssystolicarrayimplementingstructureisproposed,whichisbasedontheQRdecompositionalgorithmsandtheideaofalgorithmicengineering.Thealgorithmcanrealizefullyparallelandpipelineprocessing.

简介：Modularinversearithmeticplaysanimportantroleinellipticcurvecryptography.BasedontheanalysisofMontgomerymodularinversionalgorithm,thispaperpresentsanewdual-fieldmodularinversionalgorithm,andanovelscalableandunifiedarchitectureforMontgomeryinversehardwareinfinitefieldsGF(p)andGF(2n)isproposed.Furthermore,thisarchitecturebasedonthenewmodularinversionalgorithmhasbeenverifiedbymodelingitinVerilog-HDL,andaccomplisheditunder0.18μmCMOStechnology.Theresultindicatesthatourworkhasbetterperformanceandflexibilitythanotherworks.

简介：Thewide-banddigitalreceivingsystemsrequiredigitaldownconversion(DDC)withhighdatarateandshorttuningtimeinordertointerceptthenarrow-bandsignalswithinbroadtuningbandwidth.ButtheserequirementscannotbemetbythecommercialDDC.Inthispaperanefficientimplementationarchitectureispresented.ItcombinestheflexibilityofDFTtuningwiththeefficiencyofthepolyphasefilterbankdecomposition.Byfirstdecimatingthedatapriortofilteringandmixing,thisarchitecturegivesabettersolutiontothemismatchbetweenthelowerhardwarespeedandhighdatarate.Thecomputersimulationsshowthefeasibilityofthisprocessingarchitecture.

简介：Thecostfunctionforeigenstructuresextractionisdiscussedindetailinthispaper,onecanobtainthelargesteigenvectorbyminimizingthecostfunction.Inordertoobtainothereigenvectors,acovariancematrixseriesisconstructed.Ifonecomparesthecostfunctionwiththeenergyfunctionofaneuralnetworks,theneuralnetworkscanbeeasilyintroducedtoextracttheeigenvectors.Theoreticalanalysisandcomputersimulationsshowthattheproposedmethodisreasonableandfeasible.

简介：BasedonthestructureandbootloaderprincipleofDSPsystem,amethodfortheremoteupdation,debuggingandself-loadingofDSPsystemisdevelopedbyEthernet.HardwarecircuitandDM9000drivenprogramarepresentedandTCP/IPprotocolisembeddedintoDSP.Throughthere-mappingofexternalmemoryaddress,itiseasytoimplementtheloadofprogramsectionselectivelyandDSPself-boot.Theexperimentalresultsshowthattheproblemofhighcostinsystemmaintenancebyconventionalfielddebuggingbyemulatorandthelimitationinchip-levelbootareresolved.

简介：InthispaperthedesignandimplementationofMulti-Dimensional(MD)filter,particularly3-Dimensional(3D)filter,arepresented.Digital(discretedomain)filtersappliedtoimageandvideosignalprocessingusingthenovel3Dmultiratealgorithmsforefficientimplementationofmovingobjectextractionareengineeredwithanexample.Themultirate(decimationand/orinterpolation)signalprocessingalgorithmscanachievesignificantsavingsincomputationandmemoryusage.Theproposedalgorithmusesthemappingrelationsofz-transferfunctionsbetweennon-multirateandmultiratemathematicalexpressionsintermsoftime-varyingcoefficientinsteadoftraditionalpolyphasede-compositioncounterparts.ThemappingpropertiescanbereadilyusedtoefficientlyanalyzeandsynthesizeMDmultiratefilters.

简介：DigitalnetworkedcommunicationsarethekeytoallInternet-of-Thingsapplications,especiallytosmartmeteringsystemsandthesmartgrid.Inordertoensureasafeoperationofsystemsandtheprivacyofusers,thetransportlayersecurity(TLS)protocol,amatureandwellstandardizedsolutionforsecurecommunications,maybeused.WeimplementedtheTLSprotocolinitslatestversioninawaysuitableforembeddedandresource-constrainedsystems.ThispaperoutlinesthechallengesandopportunitiesofdeployingTLSinsmartmeteringandsmartgridapplicationsandpresentsperformanceresultsofourTLSimplementation.Ouranalysisshowsthatgivenanappropriateimplementationandconfiguration,deployingTLSinconstrainedsmartmeteringsystemsispossiblewithacceptableoverhead.

简介：Thequestforenergyefficiencyhasgrowingimportanceinhighperformancemany-coresystems.However,incurrentpractices,thepowerslacks,whicharethedifferencesobservedbetweentheinputpowerbudgetandtheactualpowerconsumedinthemany-coresystems,aretypicallyignored,thusleadingtopoorenergyefficiency.Inthispaper,weproposeaschemetoeffectivelypowertheon-chipcommunicationsbyexploitingtheavailablepowerslackthatistotallywastedincurrentmany-coresystems.Asso,thedemandforextraenergyfromexternalpowersources(e.g.,batteries)isminimized,whichhelpsimprovetheoverallenergyefficiency.Inessence,thepowerslackisstoredateachnodeandtheproposedroutingalgorithmusesadynamicprogrammingnetworktofindthegloballyoptimalpath,alongwhichthetotalenergystoredonthenodesisthemaximum.Experimentalresultshaveconfirmedthattheproposedscheme,withlowhardwareoverhead,canreducelatencyandextraenergyconsumptionby44%and48%,respectively,comparedwiththetwocompetingroutingmethods.IndexTerms?Adaptiverouting,dynamic

简介：一个高速度的X光检查非破坏性的察觉者在这份报纸被设计。X光检查非破坏性的察觉的原则被分析，并且高速度的X光检查非破坏性的察觉者的一个一般系统计划被建议。Virtex-4系列Fx12FPGA芯片被用来设计它的硬件电路，PowerPC405嵌入系统被开发，处理算法的高速度的图象被使用编译它的处理软件，并且TCP/IP协议被采用编译通讯软件，认识到高速度的X光检查信号聚会，处理并且传播。试验性的结果显示察觉者能在关联词工业地里与钢电线绳被用于生产线的长途、联机的非破坏性的察觉，例如矿，港口和码头。当察觉者的调查宽度是1.6m时，conveyer带的跑的率能完成6m/s。

简介：AT-InSAR(AlongTrackInterferometricSAR)isatechniquetodetectslow-movingtargets.However,thedetectionperformanceisgreatlyinfluencedbynoiseandclutter.Inthispaper,theinfluenceofnoiseandclutteronthedetectingperformanceisanalyzed.Bysimulatingdifferentbackgroundclutterandnoise,theperformancesofthephasethresholdanddual-thresholdmethodsarediscussedindetail,andthentheadaptive-thresholdmethodisproposedwhichcangreatlyimprovethedetectionperformance.

简介：Inthispaper,anewmodelbasedonanimprovedBrentKung(BK)parallelprefixnetwork(PPN)algorithmisproposedandrealizedinthefieldprogrammablegatearray(FPGA).Thismodelisemployedintheimplementationof20Gb/sdifferentialquadraturephase-shiftkeying(DQPSK)precoderin40Gb/spolarizationdivisionmultiplex(PolDM)DQPSKsystem.Inthecomputationprocess,thecomputationcomplexity(area)optimizationwithfan-outlimitedisachieved.Intheimplementation,770FPGAsliceregistersareutilized,whichsaveabout60%logicresourcescomparedwiththepreviousKoggeStone(KS)algorithm.

简介：Anovelarchitectureforthecolorlessopticalnetworkunit(ONU)isproposedandexperimentallydemonstratedwithdirect-detectionopticalorthogonalfrequencydivisionmultiplexing(DDO-OFDM).Inthisarchitecture,polarization-divisionmultiplexingisusedtoreducethecostatONU.Inopticallineterminal(OLT),quadratureamplitudemodulation(QAM)intensity-modulatedOFDMsignalwithx-polarizationat10Gbit/sistransmittedasdownstream.AteachONU,theopticalOFDMsignalisdemodulatedwithdirectdetection,andy-polarizationsignalismodulatedforupstreamon-offkeying(OOK)dataat5Gbit/s.SimulationresultsshowthatthepowerpenaltyisnegligibleforbothopticalOFDMdownstreamandtheon-offkeyingupstreamsignalsafterover50kmsingle-modefiber(SMF)transmission.

简介：Inthispaper,buildingtogrid(B2G)andvehicletogrid(V2G)havebeendefinedwithclearandpracticalunderstanding.Bothofthemarenewgenerationtechnologieswhicharetheessentialpartofsmartcitylivingandcrowdenergyclustering.Firstly,anin-detailedoverviewhasbeenprovidedwithanintroductiontoB2GandV2Gfollowedbyahistoricaloverviewandtheoreticalanalysisinrespecttosmartcityplanning.Next,areviewisconductedoncurrentandprevioussmartlivingresearch,whichdealswithB2GandV2G.EfficientB2GandV2Gimplementationsinpracticalcasesthenhavebeendiscussed.Lastly,bothofthesetechnicalprospectshavebeenanalyzedincrowdenergydiagram.

简介：AnewmethodisproposedtoimplementanopticalS-Rflip-flopbypolarizationencodedlightsignal,necessaryopticalnonlinearmaterialandhalf-waveplate.Inthissystemtherealtimespeedofoperationcanbeachieved,andatthetimeoftransmissiontheaveragepowerofabyteremainsconstant.Thispolarizationencodedflip-flopcanactasamemorycell.

简介：Withthedevelopmentofwirelesscommunicationsandpositioningtechnologies,trackingthepositionsofmovingobjectshasincreasednecessary.ThispaperproposesaCache-ConsciousTPR*-LinktreecalledCTPR*-Linktreewhichstoreinmainmemory.Tosatisfycontinuousmovement,theQRMBRdefinitionismodified.Thecompressionleadstothereductionofthetreeheight,whichimprovesthecachebehavioroftheindexandreducesthememoryaccesstime.Inordertoachievehighconcurrencycontrol,optimisticdynamicversioningandsibling-linkschemeispresented,whichnotonlyenableread-onlytransactionsnottofailwithlatch-freebutalsoreducecachemissesduringindexupdates.

简介：IEEE802.16d通讯标准使用直角的频率部门multiplexing(OFDM)。在广泛地使用的OFDM系统，快Fourier变换(快速傅里叶变换)和反的快Fourier变换对被用来调制并且使解调副载波上的数据星座。在这份报纸，为OFDM调节的人和解调器的高效快速傅里叶变换的高级实现被介绍。设计在Verilog被编码了并且指向了可编程的门数组进XilinxSpartan3地。Radix-22算法被建议并且为OFDM通讯系统使用了。FFT的设计被实现并且适用于固定WiMAXIEEE802.16dcommuni-阳离子标准。结果被公布，硬件参数被比较。建议体系结构在使用的multipliers的数字是最少的，到在蝮蛇的数字的最少的存储器尺寸，和秒使用。

简介：Inthelaserwarningsystem,anon-mechanicalscanningFabry-Perot(F-P)typelaserwarningmethodanditscrucialcomponent-laserincidentsensordesignmethodarebroughtout,andtheF-Petalonwiththeinterferometricfiltercombinationisadopted.Fortheincidentlaserpulsewithacertainwavelength,thetransmittancedifferenceoftheinterferometricfiltersisindependentwithpolarizationofincidentlight,anddescendsalmostlinearlyastheincidentanglegetswider.Alaserincidentanglesensorexperimentalplatformisbuiltup,andtheexperimentiscarriedoutwithagivenlaserrangefinder.Astheresultsindicate,whentheincidentangleislessthan5o,thetransmittancedifferencemeasuredbythelaseranglesensorissignificantlylower,andthedeviationismorethan10%.Whiletheincidentangleisgreaterthan10o,thedeviationisreducedsignificantly.

简介：