欢迎来到天天文库
浏览记录
ID:40238717
大小:895.00 KB
页数:67页
时间:2019-07-28
《Information Theory & Coding信息论与编码(英文版) 梁建武 第6章 限失真编码》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、Chapter6LimiteddistortionLossSourceCodingBriefintroduction:comparedwiththelosslessencoding,(comparisonondistortionandentropyrate).Iftheentropycompressedencodingpermitsdistortion,thenentropyratewillbecompressedtotheleastafterencoded.(withoutdecoder)(explai
2、nthenecessityoftwokindsofencodemethod)Info.ratedistortionfunctionR(D):isthebasisofentropycompressedencoding.Itcombinesthetwomeasurements:informationanddistortiontoprovidesolutiontoconsidertwofactorsatthesametimeinsignalprocessing.Thenecessityofintroductio
3、nofdistortion1)Distortionisinevitable;2)Whetherthereceiver(info.host)ishumanbeingormachinedevice,hehascertainanalyzeabilityandsensitivity.Thetransmissionprocesssurpasstheseabilitywillbenosignificance.3)Eventhereceivercanbedistinguishedandbedifferentiated,
4、ithaslittleinfluenceoncommunicationquality.Sowealsocallitpermitteddistortion;4)Ourgoalistoresearchvariousobjectivesourceandsourcereceiverandtogetthebiggest(permitted)distortionDanditsleastcorrespondingsourceinfo.rateR(D)whentheQosisassigned.5)Tolimiteddis
5、tortionsource,theleastinfo.ratemustbetransmittedisR(D),notthesourceentropyH(U)underthelosslesssituation.Obviously,H(U)≥R(D).onlywhenD=0,theequalityistenable;6)WemustestablishobjectivedistortionmeasureofthesourceinordertomeasureD,shouldwealsoestablishstint
6、relationwithD;7)FunctionR(D)isthetheorybasisoflimiteddistortionsourceprocessing;GoalstoachieveinthischapterUnderstandthesignificanceoflimiteddistortionsourceencodingincommunicationsystemDiscernthemeaningandbasicmethodofdistortionmeasureUnderstandtheconcep
7、t,physicalsignificanceandnatureoffunctionR(D)DiscernthecalculationmethodoffunctionR(D)inspecificsituationDiscerntheusageoffunctionR(D)inchannelencodingtheoremUnderstandthephysicalandpracticalsignificanceoftheThirdShannontheoremDiscerntheinterrelationandco
8、mparisonofthethreeShannontheoremContentsofthischapteroutlineMeasureoflossInfo.ratelossfunctionLimitedlosssourcecodingtheoremRelationsandcomparisonofthethreeShannontheorems§6.1:Outline-1ProblemintroducedReviewofnoise
此文档下载收益归作者所有