application to signal processing

application to signal processing

ID:30372883

大小:5.59 MB

页数:70页

时间:2018-12-29

application to signal processing_第1页
application to signal processing_第2页
application to signal processing_第3页
application to signal processing_第4页
application to signal processing_第5页
application to signal processing_第6页
application to signal processing_第7页
application to signal processing_第8页
application to signal processing_第9页
application to signal processing_第10页
资源描述:

《application to signal processing》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、Chapter5APPLICATIONTOSIGNALPROCESSINGThegeneralconvergencetheoremsdevelopedinChapter2candealwithnoisescontainingnotonlyrandomcomponentsbutalsostructuralerrors.ThispropertyallowsustoapplySAalgorithmstoparameterestimationproblemsarisingfromvariousfields.Thegeneralapproach,roughlyspeakin

2、g,isasfollows.First,theparameterestimationproblemcomingfrompracticeistransformedtoaroot-seekingproblemforarea-sonablebutunknownfunctionwhichmaynotbedirectlyobserved.Then,therealobservationisartificiallywritteninthestandardformwithNormally,itisquitestraightforwardtoarriveatthispoint.Th

3、emaindifficultyistoverifythatthecomplicatednoisesatisfiesoneofthenoiseconditionsrequiredintheconvergencetheorems.Itiscommonthatthereisnostandardmethodtocompletetheverificationprocedure,becausefordifferentproblemsarecompletelydifferentfromeachother.InSection5.1,SAalgorithmsareappliedto

4、solvetheblindchannelidentificationproblem,anactivetopicincommunication.InSection5.2,theprinciplecomponentanalysisusedinpatternclassificationisdealtwithbySAmethods.Section5.3continuestheproblemdiscussedinSection5.1,butinmoregeneralsetting.Namely,unlikeSection5.1,thecovariancematrixofth

5、eobservationnoiseisnolongerassumedtobeknown.InSection5.4,adaptivefilteringisconsidered:Verysimpleconditionsforconvergenceofsign-algorithmsaregiven.Section5.5dis-cussestheasymptoticbehaviorofasynchronousSAalgorithms,whichtakethepossiblecommunicationdelaysbetweenparallelprocessorsintoco

6、nsideration.219220STOCHASTICAPPROXIMATIONANDITSAPPLICATIONS5.1.RecursiveBlindIdentificationInsystemandcontrolarea,theunknownparametersareestimatedonthebasisofobservedinputandoutputdataofthesystem.Thisisthesubjectofsystemidentification.Incontrasttothis,forcommunicationchannelsonlythech

7、anneloutputisobservedandthechannelinputisun-available.Thetopicofblindchannelidentificationistoestimatechannelparametersbyusingtheoutputdataonly.Blindchannelidentificationhasdrawnmuchattentionfromresearchersbecauseofitspotentialap-plicationsinwirelesscommunication.However,mostexistinge

8、stima

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。