Information Theory & Coding信息论与编码（英文版） 梁建武 第6章 限失真编码

《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