电子信息类专业英语（西电第二版）Unit9DigitalSignalProcessing.ppt

《电子信息类专业英语（西电第二版）Unit9DigitalSignalProcessing.ppt》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、Unit9DigitalSignalProcessingPassageABasicConceptsofDSPPassageBDigitalSignalProcessorsPassageCComparisonofDSPandASPPassageABasicConceptsofDSPWedon’tspeakinadigitalsignal.Adigitalsignalisalanguageof1sand0sthatcanbeprocessedbymathematics.Wespeakinreal-world,analogsignals.Analogsignalsarerealw

2、orldsignalsthatweexperienceeveryday-sound,light,temperature,andpressure.Adigitalsignalisanumericalrepresentationoftheanalogsignal.Itmaybeeasierandmorecosteffectivetoprocessthesesignalsinthedigitalworld.Intherealworld,wecanconvertthesesignalsintodigitalsignalsthroughtheanalog-to-digitalconve

3、rter,processthesignals,andifneeded,bringthesignalsbackouttotheanalogworldthroughthedigital-to-analogconverter.1.Theessentialsofanalog-to-digitalanddigital-to-analogconversionThefirstessentialstepinanalog-to-digital(A/D)conversion(asshowninFigure9.1)istosampleananalogsignal.Thisstepisperf

4、ormedbyasampleandholdcircuit,whichsamplesatregularintervalscalledsamplingintervals.Thelengthofthesamplingintervalisthesameasthesamplingperiod,andthereciprocalofthesamplingperiodisthesamplingfrequencyfs.AccordingtotheNyquisttheorem,asignalwithamaximumfrequencyofWHz(calledaband-limitedsignal)

5、mustbesampledatleast2Wsamplespersecondtoensureaccuraterecording.Whenthisminimumisnotrespected,distortioncalledaliasingoccurs.Aliasingcauseshighfrequencysignalstoappearaslowerfrequencysignals.Tobesurealiasingwillnotoccur,samplingisalwaysprecededbylowpassfiltering.Thelowpassfilter,calledthean

6、ti-aliasingfilter,removesallfrequenciesabovehalftheselectedsamplingrate.Figure9.1Analog-to-DigitalConversionsAfterabriefacquisitiontime,duringwhichasampleisacquired,thesampleandholdcircuitholdsthesamplesteadyfortheremainderofthesamplinginterval.ThisholdtimeisneededtoallowtimeforanA/Dconver

7、tertogenerateadigitalcodethatbestcorrespondstotheanalogsample.TheA/Dconverterchoosesaquantizationlevelforeachanalogsample.AnN-bitconverterchoosesamong2Npossiblequantizationlevels.Thelargerthenumberoflevels,thesmallerthequantizationerrors,calculatedasth