1985_Robust Techniques for Signal Processing A Survey_PIEEE

《1985_Robust Techniques for Signal Processing A Survey_PIEEE》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、RobustTechniquesforSignalProcessing:ASurveyInrecentyearstherehasbeenmuchinterestinrobustnessissueswasonthederivationofoptimumschemesforuseiningeneralandinrobustsignalprocessingschemesinparticular.specifiedsignalandnoiseenvironments.AclassicexampleRobustschemesareusefulins

2、ituationswhereimpreciseaprioriofthisisthematchedfilterwhichisoptimumforaparticu-knowledgeofinputcharacteristicsmakesthesensitivityofperfor-mancetodeviationsfromassumedconditionsanimportantfactorlarsignalandnoisemodel.Becausethesignalsandnoiseininthedesignofgoodsignalproce

3、ssingschemes.Inthissurveywesignalprocessingapplicationsareusuallymodeledasran-discusstheminimaxapproachforthedesignofrobustmethodsfordomprocessesandperformancemeasuresthereforeusuallysignalprocessing.Thishasproventobeaveryusefulapproachinvolveprobabilisticquantities(sucha

4、smeansquarederrorbecauseitleadstoconstructiveproceduresfordesigningrobustschemes.Ouremphasisisonthecontributionswhichhavebeenorprobabilityoferror),thetheoryofstatisticshasplayedamadeinrobustsignalprocessing,althoughkeyresultsofotherfundamentalroleinthedevelopmentofoptimum

5、signalrobuststatisticalproceduresarealsoconsidered.Mostoftheresultsprocessingtechniques.wesurveyhavebeenobtainedinthepastfifteenyears,althoughSupposeasignalprocessingscheme,sayadetectorforasomeinterestingearlierideasforminimaxsignalprocessingarealsosignalwithknownwaveform

6、inadditivenoise,isdesignedmentioned.Thissurveyisorganizedintofivemainparts,whichdealsep-togiveoptimumperformancefornoisepossessingaspecificaratelywithrobustlinearfiltersforsignalestimation,robustlinearstatisticaldescription.Forexample,onewidespreadmodelfiltersforsignaldet

7、ectionandrelatedapplications,nonlinearmeth-fornoiseisthatitisaGaussianprocess.Animportantodsforrobustsignaldetection,nonlinearmethodsforrobustesti-questionthatarisesis,howsensitiveistheperformanceofmation,androbustdataquantization.Theinterrelationshipsamongmanyoftheseresu

8、ltsarealsodiscussedinthesurvey.suchanoptimumschemetodeviationsinthesignalandnoisecharacteristics