Análise de performance de algoritmos de reconstrução para um conversor analógico para informação

• Ano de defesa: 2022
• Autor(a) principal: Araujo, Hugo Bruno Santos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Engenharia Elétrica; Programa de Pós-Graduação em Engenharia Elétrica; UFPB
• Palavras-chave em Português: Engenharia elétrica; Amostragem compressiva; Conversor analógico para informação; Sinais esparsos; Algoritmos de reconstrução; Electrical engineering; Compressed Sensing; Analog-to-information Converter; Sparse signals; Re-construction algorithms; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/28171

The conventionalapproachtosamplinganalogsignalstodigitalfollowstheNyquistThe- orem,inwhichthesamplingrate,calledNyquistrate,mustbeatleasttwotimesbigger than themaximumfrequencyoftheanalogsignal.Inpracticalterms,generally,analogto digital converters’(ADC)samplingratesaremuchhighertooptimizesignalreconstruc- tion, however,theyincreasetheneedformemoryandprocessingpower,systemscosts, and alsoincreaseenergyconsumption.CompressiveSensing(CS)isatechniquethattakes advantage ofsignalsparsityinagivendomain,thatis,thenumberofnon-zerovaluesof a signal,andcapturesonlytheregionsthatconcentrateinformation,withsamplingrates lowerthanwhattheNyquistTheoremsays.AdevicethatimplementstheCSmodelis the Analog-to-InformationConverter(AIC),whichobtainsanoutputthatisadigitized and compressedversionoftheinput.Themainobjectiveistoanalyzetheperformanceof reconstruction algorithms-BasisPursuit(BP),OrthogonalMathingPursuit(OMP)and CompressiveSamplingMatchingPursuit(CoSaMP)-foranadaptedAICconfiguration based onRandomModulatorPre-Integration(RMPI).SimulationsweredoneonProteus and SimulinktovalidatetheAICconfigurationused,andsingle-toneandduo-tonessig- nals werereconstructedwithfrequencycomponentssimlartotheoriginalsignals.Mean squarederror(MSE)wascalculatedforthethreereconstructionmethodstodetermine which onehadbetterperformance.