Modern characterization of electromagnetic systems and its associated metrology (wiley - ieee) 1st e

Page 1


https://ebookmass.com/product/modern-characterization-of-

Instant digital products (PDF, ePub, MOBI) ready for you

Download now and discover formats that fit your needs...

Electromagnetic Metasurfaces: Theory and Applications (Wiley - IEEE) 1st Edition Karim Achouri

https://ebookmass.com/product/electromagnetic-metasurfaces-theory-andapplications-wiley-ieee-1st-edition-karim-achouri/

ebookmass.com

Principles of Materials Characterization and Metrology

https://ebookmass.com/product/principles-of-materialscharacterization-and-metrology-kannan-m-krishnan/

ebookmass.com

Introduction To Modern Planar Transmission Lines: Physical, Analytical, and Circuit Models Approach (Wiley

IEEE)

1st Edition Anand K. Verma

https://ebookmass.com/product/introduction-to-modern-planartransmission-lines-physical-analytical-and-circuit-models-approachwiley-ieee-1st-edition-anand-k-verma/ ebookmass.com

Co-Production of Public Services and Outcomes 1st ed.

https://ebookmass.com/product/co-production-of-public-services-andoutcomes-1st-ed-edition-elke-loeffler/

ebookmass.com

https://ebookmass.com/product/for-you-no-one-else-roni-loren-2/

ebookmass.com

The Palgrave Handbook of Climate History 1st ed. Edition Sam White

https://ebookmass.com/product/the-palgrave-handbook-of-climatehistory-1st-ed-edition-sam-white/

ebookmass.com

The Outrageous Idea of Christian Teaching Perry Glanzer

https://ebookmass.com/product/the-outrageous-idea-of-christianteaching-perry-glanzer/

ebookmass.com

Queens of London Heather Webb

https://ebookmass.com/product/queens-of-london-heather-webb/

ebookmass.com

Four Ways to Wear a Dress Gillian Libby

https://ebookmass.com/product/four-ways-to-wear-a-dress-gillianlibby-2/

ebookmass.com

3 Days to Live James Patterson

https://ebookmass.com/product/3-days-to-live-james-patterson/

ebookmass.com

ModernCharacterizationof ElectromagneticSystemsandIts AssociatedMetrology

ModernCharacterizationof ElectromagneticSystemsandIts AssociatedMetrology

TapanK.Sarkar†

SyracuseUniversity

11WexfordRoad,Syracuse,NewYork13214

MagdalenaSalazar-Palma

CarlosIIIUniversityofMadrid

Avda.delaUniversidad30,28911Leganés,Madrid,Spain

MingDaZhu

XidianUniversity

No.2SouthTaibaiRoad,Xi’an,Shaanxi,China

HengChen

SyracuseUniversity

211LafayetteRd.Room425,Syracuse,NY,USA

Thiseditionfirstpublished2021

©2021JohnWiley&Sons,Inc.

Allrightsreserved.Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,or transmitted,inanyformorbyanymeans,electronic,mechanical,photocopying,recordingor otherwise,exceptaspermittedbylaw.Adviceonhowtoobtainpermissiontoreusematerialfrom thistitleisavailableathttp://www.wiley.com/go/permissions.

TherightofTapanK.Sarkar,MagdalenaSalazar-Palma,MingDaZhu,HengChentobeidentified astheauthorsoftheeditorialmaterialinthisworkhasbeenassertedinaccordancewithlaw.

RegisteredOffice

JohnWiley&Sons,Inc.,111RiverStreet,Hoboken,NJ07030,USA

EditorialOffice 111RiverStreet,Hoboken,NJ07030,USA

Fordetailsofourglobaleditorialoffices,customerservices,andmoreinformationaboutWiley productsvisitusatwww.wiley.com.

Wileyalsopublishesitsbooksinavarietyofelectronicformatsandbyprint-on-demand.Some contentthatappearsinstandardprintversionsofthisbookmaynotbeavailableinotherformats.

LimitofLiability/DisclaimerofWarranty

MATLAB® isatrademarkofTheMathWorks,Inc.andisusedwithpermission.TheMathWorksdoes notwarranttheaccuracyofthetextorexercisesinthisbook.Thiswork’suseordiscussionof MATLAB® softwareorrelatedproductsdoesnotconstituteendorsementorsponsorshipbyThe MathWorksofaparticularpedagogicalapproachorparticularuseoftheMATLAB® software.While thepublisherandauthorshaveusedtheirbesteffortsinpreparingthiswork,theymakeno representationsorwarrantieswithrespecttotheaccuracyorcompletenessofthecontentsofthiswork andspecificallydisclaimallwarranties,includingwithoutlimitationanyimpliedwarrantiesof merchantabilityorfitnessforaparticularpurpose.Nowarrantymaybecreatedorextendedbysales representatives,writtensalesmaterialsorpromotionalstatementsforthiswork.Thefactthatan organization,website,orproductisreferredtointhisworkasacitationand/orpotentialsourceof furtherinformationdoesnotmeanthatthepublisherandauthorsendorsetheinformationorservices theorganization,website,orproductmayprovideorrecommendationsitmaymake.Thisworkis soldwiththeunderstandingthatthepublisherisnotengagedinrenderingprofessionalservices. Theadviceandstrategiescontainedhereinmaynotbesuitableforyoursituation.Youshouldconsult withaspecialistwhereappropriate.Further,readersshouldbeawarethatwebsiteslistedinthiswork mayhavechangedordisappearedbetweenwhenthisworkwaswrittenandwhenitisread.Neither thepublishernorauthorsshallbeliableforanylossofprofitoranyothercommercialdamages, includingbutnotlimitedtospecial,incidental,consequential,orotherdamages.

LibraryofCongressCataloging-in-PublicationData

Names:Sarkar,Tapan(TapanK.),editor.|Salazar-Palma,Magdalena, editor.|Zhu,MingDa,editor.|Chen,Heng,editor.

Title:Moderncharacterizationofelectromagneticsystemsandits associatedmetrology/editedbyTapanK.Sarkar,Magdalena Salazar-Palma,MingDaZhu,HengChen.

Description:Hoboken,NJ:Wiley,2020.|Includesbibliographical referencesandindex.

Identifiers:LCCN2020008264(print)|LCCN2020008265(ebook)|ISBN 9781119076469(hardback)|ISBN9781119076544(adobepdf)|ISBN 9781119076537(epub)

Subjects:LCSH:Electromagnetism–Mathematics.|Electromagnetic waves–Measurement.

Classification:LCCQC760.M532020(print)|LCCQC760(ebook)|DDC 537/.12–dc23

LCrecordavailableathttps://lccn.loc.gov/2020008264

LCebookrecordavailableathttps://lccn.loc.gov/2020008265

CoverDesign:Wiley

CoverImage:©zfL/GettyImages

Setin10/12ptWarnockbyStraive,Pondicherry,India 10987654321

Contents

Preface xiii

Acknowledgments xxi

TributetoTapanK.Sarkar – MagdalenaSalazarPalma,MingDaZhu, andHengChen xxiii

1MathematicalPrinciplesRelatedtoModernSystemAnalysis 1

Summary 1

1.1Introduction 1

1.2Reduced-RankModelling:BiasVersusVarianceTradeoff 3

1.3AnIntroductiontoSingularValueDecomposition(SVD)andthe TheoryofTotalLeastSquares(TLS) 6

1.3.1SingularValueDecomposition 6

1.3.2TheTheoryofTotalLeastSquares 15

1.4Conclusion 19 References 20

2MatrixPencilMethod(MPM) 21

Summary 21

2.1Introduction 21

2.2DevelopmentoftheMatrixPencilMethodforNoiseContaminated Data 24

2.2.1ProcedureforInterpolatingorExtrapolatingtheSystemResponse UsingtheMatrixPencilMethod 26

2.2.2IllustrationsUsingNumericalData 26

2.2.2.1Example1 26

2.2.2.2Example2 29

2.3ApplicationsoftheMPMforEvaluationoftheCharacteristic ImpedanceofaTransmissionLine 32

2.4ApplicationofMPMfortheComputationoftheS-Parameters WithoutanyAPrioriKnowledgeoftheCharacteristicImpedance 37

2.5ImprovingtheResolutionofNetworkAnalyzerMeasurements UsingMPM 44

2.6MinimizationofMultipathEffectsUsingMPMinAntenna MeasurementsPerformedinNon-AnechoicEnvironments 57

2.6.1ApplicationofaFFT-BasedMethodtoProcesstheData 61

2.6.2ApplicationofMPMtoProcesstheData 64

2.6.3PerformanceofFFTandMPMAppliedtoMeasuredData 67

2.7ApplicationoftheMPMforaSingleEstimateoftheSEM-Poles WhenUtilizingWaveformsfromMultipleLookDirections 74

2.8DirectionofArrival(DOA)EstimationAlongwithTheir FrequencyofOperationUsingMPM 81

2.9EfficientComputationoftheOscillatoryFunctionalVariationinthe TailsoftheSommerfeldIntegralsUsingMPM 85

2.10IdentificationofMultipleObjectsOperatinginFreeSpaceThrough TheirSEMPoleLocationsUsingMPM 91

2.11OtherMiscellaneousApplicationsofMPM 95

2.12Conclusion 95 Appendix2AComputerCodesforImplementingMPM 96 References 99

3TheCauchyMethod 107

Summary 107

3.1Introduction 107

3.2ProcedureforInterpolatingorExtrapolatingtheSystemResponse UsingtheCauchyMethod 112

3.3ExamplestoEstimatetheSystemResponseUsingtheCauchy Method 112

3.3.1Example1 112

3.3.2Example2 116

3.3.3Example3 118

3.4IllustrationofExtrapolationbytheCauchyMethod 120

3.4.1ExtendingtheEfficiencyoftheMomentMethodThrough ExtrapolationbytheCauchyMethod 120

3.4.2InterpolatingResultsforOpticalComputations 123

3.4.3ApplicationtoFilterAnalysis 125

3.4.4BroadbandDeviceCharacterizationUsingFewParameters 127

3.5EffectofNoiseContaminatingtheDataandItsImpactonthe PerformanceoftheCauchyMethod 130

3.5.1PerturbationofInvariantSubspaces 130

3.5.2PerturbationoftheSolutionoftheCauchyMethodDue toAdditiveNoise 131

3.5.3NumericalExample 136

3.6GeneratingHighResolutionWidebandResponsefromSparseand IncompleteAmplitude-OnlyData 138

3.6.1DevelopmentoftheInterpolatoryCauchyMethodfor Amplitude-OnlyData 139

3.6.2InterpolatingHighResolutionAmplitudeResponse 142

3.7GenerationoftheNon-minimumPhaseResponsefrom Amplitude-OnlyDataUsingtheCauchyMethod 148

3.7.1GenerationoftheNon-minimumPhase 149

3.7.2IllustrationThroughNumericalExamples 151

3.8DevelopmentofanAdaptiveCauchyMethod 158

3.8.1Introduction 158

3.8.2AdaptiveInterpolationAlgorithm 159

3.8.3IllustrationUsingNumericalExamples 160

3.8.4Summary 171

3.9EfficientCharacterizationofaFilter 172

3.10ExtractionofResonantFrequenciesofanObjectfromFrequency DomainData 176

3.11Conclusion 180 Appendix3AMATLABCodesfortheCauchyMethod 181 References 187

4ApplicationsoftheHilbertTransform – ANonparametricMethodfor Interpolation/ExtrapolationofData 191

Summary 191

4.1Introduction 192

4.2ConsequenceofCausalityandItsRelationshiptotheHilbert Transform 194

4.3PropertiesoftheHilbertTransform 195

4.4RelationshipBetweentheHilbertandtheFourierTransforms fortheAnalogandtheDiscreteCases 199

4.5MethodologytoExtrapolate/InterpolateDataintheFrequency DomainUsingaNonparametricMethodology 200

4.6InterpolatingMissingData 203

4.7ApplicationoftheHilbertTransformforEfficientComputation oftheSpectrumforNonuniformlySpacedData 213

4.7.1FormulationoftheLeastSquareMethod 217

4.7.2HilbertTransformRelationship 221

4.7.3MagnitudeEstimation 223

4.8Conclusion 229 References 229

5TheSourceReconstructionMethod 235

Summary 235

5.1Introduction 236

5.2AnOverviewoftheSourceReconstructionMethod(SRM) 238

5.3MathematicalFormulationfortheIntegralEquations 239

5.4Near-FieldtoFar-FieldTransformationUsinganEquivalent MagneticCurrentApproach 240

5.4.1DescriptionoftheProposedMethodology 241

5.4.2SolutionoftheIntegralEquationfortheMagneticCurrent 245

5.4.3NumericalResultsUtilizingtheMagneticCurrent 249

5.4.4Summary 268

5.5Near-FieldtoNear/Far-FieldTransformationforArbitrary Near-FieldGeometryUtilizinganEquivalentElectricCurrent 276

5.5.1DescriptionoftheProposedMethodology 278

5.5.2NumericalResultsUsinganEquivalentElectricCurrent 281

5.5.3Summary 286

5.6EvaluatingNear-FieldRadiationPatternsofCommercial Antennas 297

5.6.1Background 297

5.6.2FormulationoftheProblem 301

5.6.3ResultsfortheNear-fieldToFar-fieldTransformation 304

5.6.3.1ABaseStationAntenna 304

5.6.3.2NFtoFFTransformationofaPyramidalHornAntenna 307

5.6.3.3ReferenceVolumeofaBaseStationAntennaforHuman ExposuretoEMFields 310

5.6.4Summary 311

5.7Conclusions 313 References 314

6PlanarNear-FieldtoFar-FieldTransformationUsingaSingleMoving ProbeandaFixedProbeArrays 319 Summary 319

6.1Introduction 320

6.2Theory 322

6.3IntegralEquationFormulation 323

6.4FormulationoftheMatrixEquation 325

6.5UseofanMagneticDipoleArrayasEquivalentSources 328

6.6SampleNumericalResults 329

6.7Summary 337

6.8DifferencesbetweenConventionalModalExpansionandthe EquivalentSourceMethodforPlanarNear-FieldtoFar-Field Transformation 337

6.8.1Introduction 337

6.8.2ModalExpansionMethod 339

6.8.3IntegralEquationApproach 341

6.8.4NumericalExamples 344

6.8.5Summary 351

6.9ADirectOptimizationApproachforSourceReconstruction andNF-FFTransformationUsingAmplitude-OnlyData 352

6.9.1Background 352

6.9.2EquivalentCurrentRepresentation 354

6.9.3OptimizationofaCostFunction 356

6.9.4NumericalSimulation 357

6.9.5ResultsObtainedUtilizingExperimentalData 358

6.9.6Summary 359

6.10UseofComputationalElectromagneticstoEnhancetheAccuracy andEfficiencyofAntennaPatternMeasurementsUsingan ArrayofDipoleProbes 361

6.10.1Introduction 362

6.10.2DevelopmentoftheProposedMethodology 363

6.10.3PhilosophyoftheComputationalMethodology 363

6.10.4FormulationoftheIntegralEquations 365

6.10.5SolutionoftheIntegro-DifferentialEquations 367

6.10.6SampleNumericalResults 369

6.10.6.1Example1 369

6.10.6.2Example2 373

6.10.6.3Example3 377

6.10.6.4Example4 379

6.10.7Summary 384

6.11AFastandEfficientMethodforDeterminingtheFarField PatternsAlongthePrincipalPlanesUsingaRectangular ProbeArray 384

6.11.1Introduction 385

6.11.2DescriptionoftheProposedMethodology 385

6.11.3SampleNumericalResults 387

6.11.3.1Example1 387

6.11.3.2Example2 393

6.11.3.3Example3 397

6.11.3.4Example4 401

6.11.4Summary 406

6.12TheInfluenceoftheSizeofSquareDipoleProbeArray MeasurementontheAccuracyofNF-FFPattern 406

6.12.1IllustrationoftheProposedMethodologyUtilizingSample NumericalResults 407

6.12.1.1Example1 407

6.12.1.2Example2 411

6.12.1.3Example3 416

6.12.1.4Example4 419

6.12.2Summary 428

6.13UseofaFixedProbeArrayMeasuringAmplitude-Only Near-FieldDataforCalculatingtheFar-Field 428

8.2TheConjugateGradientMethodwithFastFourierTransformfor ComputationalEfficiency 495 Contents x

6.13.1ProposedMethodology 429

6.13.2SampleNumericalResults 430

6.13.2.1Example1 430

6.13.2.2Example2 434

6.13.2.3Example3 437

6.13.2.4Example4 437

6.13.3Summary 441

6.14ProbeCorrectionforUsewithElectricallyLargeProbes 442

6.14.1DevelopmentoftheProposedMethodology 443

6.14.2FormulationoftheSolutionMethodology 446

6.14.3SampleNumericalResults 447

6.15Conclusions 449 References 449

7SphericalNear-FieldtoFar-FieldTransformation 453 Summary 453

7.1AnAnalyticalSphericalNear-FieldtoFar-Field Transformation 453

7.1.1Introduction 453

7.1.2AnAnalyticalSphericalNear-FieldtoFar-Field Transformation 454

7.1.3NumericalSimulations 464

7.1.3.1SyntheticData 464

7.1.3.2ExperimentalData 465

7.1.4Summary 468

7.2RadialFieldRetrievalinSphericalScanningforCurrent ReconstructionandNF–FFTransformation 468

7.2.1Background 468

7.2.2AnEquivalentCurrentReconstructionfromSpherical MeasurementPlane 470

7.2.3TheRadialElectricFieldRetrievalAlgorithm 472

7.2.4ResultsObtainedUsingThisFormulation 473

7.2.4.1SimulatedData 473

7.2.4.2UsingMeasuredData 475

7.3Conclusion 482 Appendix7AAFortranBasedComputerProgramforTransforming SphericalNear-FieldtoFar-Field 483 References 489

8DeconvolvingMeasuredElectromagneticResponses 491 Summary 491

8.1Introduction 491

8.2.1Theory 495

8.2.2NumericalResults 498

8.3TotalLeastSquaresApproachUtilizingSingularValue Decomposition 501

8.3.1Theory 501

8.3.2TotalLeastSquares(TLS) 502

8.3.3NumericalResults 506

8.4Conclusion 516 References 516

9PerformanceofDifferentFunctionalsforInterpolation/ ExtrapolationofNear/Far-FieldData 519 Summary 519

9.1Background 520

9.2ApproximatingaFrequencyDomainResponsebyChebyshev Polynomials 521

9.3TheCauchyMethodBasedonGegenbauerPolynomials 531

9.3.1NumericalResultsandDiscussion 537

9.3.1.1ExampleofaHornAntenna 537

9.3.1.2Exampleofa2-elementMicrostripPatchArray 539

9.3.1.3ExampleofaParabolicAntenna 541

9.4Near-FieldtoFar-FieldTransformationofaZenith-Directed ParabolicReflectorUsingtheOrdinaryCauchyMethod 543

9.5Near-FieldtoFar-FieldTransformationofaRotatedParabolic ReflectorUsingtheOrdinaryCauchyMethod 552

9.6Near-FieldtoFar-FieldTransformationofaZenith-Directed ParabolicReflectorUsingtheMatrixPencilMethod 558

9.7Near-FieldtoFar-FieldTransformationofaRotatedParabolic ReflectorUsingtheMatrixPencilMethod 564

9.8Conclusion 569 References 569

10RetrievalofFreeSpaceRadiationPatternsfromMeasuredData inaNon-AnechoicEnvironment 573 Summary 573

10.1ProblemBackground 573

10.2ReviewofPatternReconstructionMethodologies 575

10.3DeconvolutionMethodforRadiationPatternReconstruction 578

10.3.1EquationsandDerivation 578

10.3.2StepsRequiredtoImplementtheProposedMethodology 584

10.3.3ProcessingoftheData 585

10.3.4SimulationExamples 587

10.3.4.1ExampleI:OnePECPlateServesasaReflector 587

10.3.4.2ExampleII:TwoPECPlatesNowServeasReflectors 594

10.3.4.3ExampleIII:FourConnectedPECPlatesServeasReflectors 598

10.3.4.4ExampleIV:UseofaParabolicReflectorAntennaastheAUT 604

10.3.5DiscussionsontheDeconvolutionMethodforRadiationPattern Reconstruction 608

10.4EffectofDifferentTypesofProbeAntennas 608

10.4.1NumericalExamples 608

10.4.1.1ExampleI:UseofaYagiAntennaastheProbe 608

10.4.1.2ExampleII:UseofaParabolicReflectorAntennaastheProbe 612

10.4.1.3ExampleIII:UseofaDipoleAntennaastheProbe 613

10.5EffectofDifferentAntennaSize 619

10.6EffectofUsingDifferentSizesofPECPlates 626

10.7ExtensionoftheDeconvolutionMethodtoThree-Dimensional PatternReconstruction 632

10.7.1MathematicalCharacterizationoftheMethodology 632

10.7.2StepsSummarizingfortheMethodology 635

10.7.3ProcessingtheData 636

10.7.4ResultsforSimulationExamples 638

10.7.4.1ExampleI:FourWidePECPlatesServeasReflectors 640

10.7.4.2ExampleII:FourPECPlatesandtheGroundServeas Reflectors 643

10.7.4.3ExampleIII:SixPlatesForminganUnclosedContourServeas Reflectors 651

10.7.4.4ExampleIV:AntennaMeasurementinaClosedPECBox 659

10.7.4.5ExampleV:SixDielectricPlatesFormingaClosedContour SimulatingaRoom 662

10.8Conclusion 673

AppendixA:DataMappingUsingtheConversionbetweenthe SphericalCoordinateSystemandtheCartesianCoordinate System 675

AppendixB:Descriptionofthe2D-FFTduringtheData Processing 677

References 680

Index 683

Preface

Theareaofelectromagneticsisanevolutionaryone.Intheearlierdaysthe analysisinthisareawaslimitedto11separablecoordinatesystemsforthesolutionofHelmholtzequations.Theelevencoordinatesystemsarerectangular, circularcylinder,ellipticcylinder,paraboliccylinder,spherical,conical, parabolic,prolatespheroidal,oblatespheroidal,ellipsoidalandparaboloidal coordinates.However,Laplace ’sequationisseparablein13coordinatesystems, theadditionaltwobeingthebisphericalandthetoroidalcoordinatesystems. Outsidethesecoordinatesystemsitwasnotpossibletodevelopasolution forelectromagneticproblemsintheearlierdays.However,withtheadvent ofnumericalmethodsthissituationchangedanditwaspossibletosolvereal practicalproblemsinanysystem.Thisdevelopmenttookplaceintwodistinct stagesandwasprimarilyaddressedbyProf.RogerF.Harrington.Inthefirst phaseheproposedtodevelopthesolutionofanelectromagneticfieldproblem intermsofunknowncurrents,bothelectricandmagneticandnotfieldsby placingsomeequivalentcurrentstorepresenttheactualsourcessothatthese currentsproduceexactlythesamedesiredfieldsineachregion.Fromthese currentshecomputedtheelectricandthemagneticvectorpotentialsinany coordinatesystem.Intheintegralrepresentationofthepotentialsintermsof theunknowncurrents,thefreespaceGreen’sfunctionwasusedwhichsimplifiedtheformulationconsiderablyasnocomplicatedformoftheGreen’sfunctionforanycomplicatedenvironmentwasnecessary.Fromthepotentials,the fields,bothelectricandmagnetic,weredevelopedbyinvokingtheMaxwellHertz-Heavisideequations.Thismadethemathematicalanalysisquiteanalytic andsimplifiedmanyofthecomplexitiesrelatedtothecomplicatedGreen’stheorem.Thiswasthemainthemeinhisbook “TimeHarmonicElectromagnetic Fields”,McGrawHill,1961.Attheendofthisbookhetriedtodevelopavariationalformforalltheseconceptssothatanumericaltechniquecanbeapplied andonecansolveanyelectromagneticboundaryvalueproblemofinterest.This themewasfurtherdevelopedinthesecondstagethroughhissecondclassic book “FieldComputationsbyMomentMethods”,MacmillanCompany,1968. Inthesecondbookheillustratedhowtosolveageneralelectromagneticfield

problem.Thisgradualdevelopmenttookalmosthalfacenturytomature.Inthe experimentalrealm,unfortunately,nosuchprogresshasbeenmade.Thismay bepartiallyduetodecisionstakenbythepastleadershipoftheIEEEAntennas andPropagationSociety(AP-S)whofirstessentiallydisassociatedmeasurementsfromtheirprimaryfocusleadingantennameasurementpractitioners toformtheAntennaMeasurementsTechniquesAssociation(AMTA)asan organizationdifferentfromIEEEAP-S.AndlateroneventhenumericaltechniquespartwasnotconsideredinthemainthemeoftheIEEEAntennasand PropagationSocietyleadingtotheformationoftheAppliedComputational ElectromagneticSociety(ACES).However,inrecenttimestheseshortcomings ofthepastdecisionsoftheAP-Sleadershiphavebeenaddressed.

Theobjectiveofthisbookistoadvancethestateoftheartofantennameasurementsandnotbeinglimitedtothesituationthatmeasurementscanbe madeinoneoftheseparablecoordinatesystemsjustlikethestateofelectromagneticsoverhalfacenturyago.Weproposetocarryoutthistransformation intherealmofmeasurementfirstbytryingtofindasetofequivalentcurrents justlikewedointheoryandthensolvefortheseunknowncurrentsusingthe Maxwell-Hertz-HeavisideequationsviatheMethodofMomentspopularized byProf.Harrington.Sincetheexpressionsbetweenthemeasuredfieldsand theunknowncurrentsareanalyticandrelatedbyMaxwell-Hertz-Heaviside equations,themeasurementscanbecarriedoutinanyarbitrarygeometry andnotjustlimitedtotheplanar,cylindricalorsphericalgeometries.The advantageofthisnewmethodologyaspresentedinthisbookthroughthetopic “SourceReconstructionMethod” isthatthemeasurementofthefieldsneed notbedoneusingaNyquistsamplingcriteriawhichopensupnewavenues particularlyintheveryhighfrequencyregimeoftheelectromagneticspectrum whereitmightbedifficulttotakemeasurementsampleshalfawavelength apart.Secondlyaswillbeillustratedthesemeasurementsamplesneednot evenbeperformedinanyspecifiedplane.Alsobecauseoftheanalyticalrelationshipbetweenthesourcesthatgeneratethefieldsandthefieldsthemselves itispossibletogobeyondtheRaleighresolutionlimitandachievesuperresolutioninthediagnosisofradiatingstructures.IntheRaleighlimittheresolutionislimitedbytheuncertaintyprincipleandthatisdeterminedbythe lengthoftheaperturewhoseFouriertransformwearelookingatwhereas inthesuperresolutionsystemthereisnosuchrestriction.Anotherobjective ofthisbookistooutlineaverysimpleproceduretorecoverthenon-minimum phaseofanyelectromagneticsystemusingamplitude-onlydata.Thissimple procedureisbasedontheprincipleofcausalitywhichresultsintheHilbert transformrelationshipbetweentherealandtheimaginarypartsofatransfer functionofanylineartimeinvariantsystem.Thephilosophyofmodelorder reductioncanalsobeimplementedusingtheconceptsoftotalleastsquares alongwiththesingularvaluedecomposition.Thismakestheill-poseddeconvolutionproblemquitestablenumerically.Finally,itisshownhowto

interpolateandextrapolatemeasureddataincludingfillingupthegapofmissingmeasurednear/far-fielddata.

Thebookcontainstenchapters.InChapter1,themathematicalpreliminaries aredescribed.Inthemathematicalfieldofnumericalanalysis,modelorder reductionisthekeytoprocessingmeasureddata.Thisalsoenablesustointerpolateandextrapolatemeasureddata.Thephilosophyofmodelorderreduction isoutlinedinthischapteralongwiththeconceptsoftotalleastsquaresandsingularvaluedecomposition.

InChapter2,wepresentthematrixpencilmethod(MPM)whichisamethodologytoapproximateagivendatasetbyasumofcomplexexponentials.The objectiveistointerpolateandextrapolatedataandalsotoextractcertainparameterssoastocompressthedataset.Firstthemethodologyispresented followedbysomeapplicationinelectromagneticsystemcharacterization. Theapplicationsinvolveusingthismethodologytodeembeddevicecharacteristicsandobtainaccurateandhighresolutioncharacterization,enhance networkanalyzermeasurementswhennotenoughphysicalbandwidthis availableformeasurements,minimizeunwantedreflectionsinantennameasurementsand,whenperformingsystemcharacterizationinanon-anechoic environment,toextractasinglesetofexponentsrepresentingtheresonant frequencyofanobjectwhendatafrommultiplelookanglesaregivenand computedirectionsofarrivalestimationofsignalsalongwiththeirfrequencies ofoperation.Thismethodcanalsobeusedtospeedupthecalculationofthe tailsencounteredintheevaluationoftheSommerfeldintegralsandinmultiple targetcharacterizationinfreespacefromthescattereddatausingtheir characteristicexternalresonancewhicharepopularlyknownasthesingularity expansionmethod(SEM)poles.Referencestootherapplications,including multipathcharacterizationofapropagatingwave,characterizationofthe qualityofpowersystems,inwaveformanalysisandimagingandspeedingup computationsinatimedomainelectromagneticsimulation.Acomputer programimplementingthematrixpencilmethodisgivenintheappendixso thatitcaneasilybeimplementedinpractice.

Innumericalanalysis,interpolationisamethodofestimatingunknowndata withintherangeofknowndatafromtheavailableinformation.Extrapolationis alsotheprocessofapproximatingunknowndataoutsidetherangeoftheknown availabledata.InChapter3,wearegoingtolookattheconceptoftheCauchy methodfortheinterpolationandextrapolationofbothmeasuredandnumericallysimulateddata.TheCauchymethodcandealwithextendingtheefficiency ofthemomentmethodthroughfrequencyextrapolation.Interpolatingresults foropticalcomputations,generationofpassbandusingstopbanddataandvice versa,efficientbroadbanddevicecharacterization,effectofnoiseontheperformanceoftheCauchymethodandforapplicationstoextrapolatingamplitudeonlydataforthefar-fieldorRCSinterpolation/extrapolation.Usingthis methodtogeneratethenon-minimumphaseresponsefromamplitude-only

data,andadaptiveinterpolationforsparselysampleddataisalsoillustrated.In addition,ithasbeenappliedtocharacterizationoffiltersandextractingresonantfrequenciesofobjectsusingfrequencydomaindata.Otherapplications includenon-destructiveevaluationoffruitstatusofmaturityandqualityoffruit juices,RCSapplicationsandtomultidimensionalextrapolation.Acomputer programimplementingtheCauchymethodhasbeenprovidedintheAppendix againforeaseofunderstanding.

Theprevioustwochaptersdiscussedtheparametricmethodsinthecontext oftheprincipleofanalyticcontinuationandprovideditsrelationshiptoreduced rankmodellingusingthetotalleastsquaresbasedsingularvaluedecomposition methodology.Theproblemwithaparametricmethodisthatthequalityofthe solutionisdeterminedbythechoiceofthebasisfunctionsanduseofunsuitable basisfunctionsgeneratebadsolutions.Aprioriitisquitedifficulttorecognize whataregoodbasisfunctionsandwhatarebadbasisfunctionseventhough methodologiesexistintheoryonhowtochoosegoodones.Theadvantageof thenonparametricmethodspresentedinChapter4isthatnosuchchoices ofthebasisfunctionsneedtobemadeasthesolutionprocedurebyitselfdevelopsthenatureofthesolutionandnoaprioriinformationisnecessary.Thisis accomplishedthroughtheuseoftheHilberttransformwhichexploitsoneofthe fundamentalpropertiesofnatureandthatiscausality.TheHilberttransform illustratesthattherealandimaginarypartsofanynonminimumphasetransfer functionforacausalsystemsatisfythisrelationship.Inaddition,someparametrizationcanalsobemadeofthisprocedurewhichcanenableonetogeneratea nonminimumphasefunctionfromitsamplituderesponseandfromthatgeneratethephaseresponse.Thisenablesonetocomputethetimedomain responseofthesystemusingamplitudeonlydatabarringatimedelayinthe response.Thisdelayuncertaintyisremovedinholographyasinsuchaprocedureanamplitudeandphaseinformationismeasuredforaspecificlookangle thuseliminatingthephaseambiguity.Anoverviewofthetechniquealongwith examplesarepresentedtoillustratethismethodology.TheHilberttransform canalsobeusedtospeedupthespectralanalysisofnonuniformlyspaceddata samples.Therefore,inthissectionanovelleastsquaresmethodologyisapplied toafinitedatasetusingtheprincipleofspectralestimation.Thiscanbeapplied fortheanalysisofthefar-fieldpatterncollectedfromunevenlyspacedantennas. Theadvantageofusinganon-uniformlysampleddataisthatitisnotnecessary tosatisfytheNyquistsamplingcriterionaslongastheaveragevalueofthesamplingrateislessthantheNyquistrate.Accurateandefficientcomputationof thespectrumusingaleastsquaresmethodappliedtoafiniteunevenlyspaced dataisalsostudied.

InChapter5,thesourcereconstructionmethod(SRM)ispresented.Itisa recenttechniquedevelopedforantennadiagnosticsandforcarryingout near-field(NF)tofar-field(FF)transformation.TheSRMisbasedontheapplicationoftheelectromagneticEquivalencePrinciple,inwhichoneestablishesan

equivalentcurrentdistributionthatradiatesthesamefieldsastheactualcurrentsinducedintheantennaundertest(AUT).Theknowledgeoftheequivalentcurrentsallowsthedeterminationoftheantennaradiatingelements,aswell asthepredictionoftheAUT-radiatedfieldsoutsidetheequivalentcurrents domain.Theuniquefeatureofthenovelmethodologypresentedisthatit canresolveequivalentcurrentsthataresmallerthanhalfawavelengthinsize, thusprovidingsuper-resolution.Furthermore,themeasurementfieldsamples canbetakenatspacinggreaterthanhalfawavelength,thusgoingbeyondthe classicalsamplingcriteria.Thesetwodistinctivefeaturesarepossibleduetothe choiceofamodel-basedparameterestimationmethodologywherethe unknownsourcesareapproximatedbyabasisinthecomputationalMethod ofMoment(MoM)contextand,secondly,throughtheuseoftheanalyticfree spaceGreen’sfunction.Thelatterconditionalsoguaranteestheinvertibilityof theelectricfieldoperatorandprovidesastablesolutionforthecurrentseven whenevanescentwavesarepresentinthemeasurements.Inaddition,the useofthesingularvaluedecompositioninthesolutionofthematrixequations providestheuserwithaquantitativetooltoassessthequalityandthequantity ofthemeasureddata.Alternatively,theuseoftheiterativeconjugategradient (CG)methodinsolvingtheill-conditionedmatrixequationsfortheequivalent currentscanalsobeimplemented.Twodifferentmethodsarepresentedinthis section.Onethatdealswiththeequivalentmagneticcurrentandthesecond thatdealswiththeequivalentelectriccurrent.Iftheformulationissound,then eitherofthemethodologieswillprovidethesamefar-fieldwhenusingthesame near-fielddata.Examplesarepresentedtoillustratetheapplicabilityandaccuracyoftheproposedmethodologyusingeitheroftheequivalentcurrentsand appliedtoexperimentaldata.Thismethodologyisthenusedfornear-fieldto near/far-fieldtransformationsforarbitrarynear-fieldgeometrytoevaluate thesafedistanceforcommercialantennas.

InChapter6,afastandaccuratemethodispresentedforcomputingfar-field antennapatternsfromplanarnear-fieldmeasurements.Themethodutilizes near-fielddatatodetermineequivalentmagneticcurrentsourcesoverafictitiousplanarsurfacethatencompassestheantenna,andthesecurrentsareused toascertainthefarfields.Undercertainapproximations,thecurrentsshould producethecorrectfarfieldsinallregionsinfrontoftheantennaregardless ofthegeometryoverwhichthenear-fieldmeasurementsaremade.Anelectric fieldintegralequation(EFIE)isdevelopedtorelatethenearfieldstotheequivalentmagneticcurrents.Methodofmoments(MOM)procedureisusedto transformtheintegralequationintoamatrixone.Thematrixequationissolved usingtheiterativeconjugategradientmethod(CGM),andinthecaseofarectangularmatrix,aleast-squaressolutioncanstillbefoundusingthisapproach forthemagneticcurrentswithoutexplicitlycomputingthenormalformofthe equations.Near-fieldtofar-fieldtransformationforplanarscanningmaybe efficientlyperformedundercertainconditionsbyexploitingtheblockToeplitz

structureofthematrixandusingtheconjugategradientmethod(CGM)andthe fastFouriertransform(FFT),therebydrasticallyreducingcomputationtime andstoragerequirements.Numericalresultsarepresentedforseveralantenna configurationsbyextrapolatingthefarfieldsusingsyntheticandexperimental near-fielddata.Itisalsoillustratedthatasinglemovingprobecanbereplaced byanarrayofprobestocomputetheequivalentmagneticcurrentsonthe surfaceenclosingtheAUTinasinglesnapshotratherthantediouslymoving asingleprobeovertheantennaundertesttomeasureitsnear-fields.Itis demonstratedthatinthismethodologyaprobecorrectionevenwhenusing anarrayofdipoleprobesisnotnecessary.Theaccuracyofthismethodology isstudiedasafunctionofthesizeoftheequivalentsurfaceplacedinfrontof theantennaundertestandtheerrorintheestimationofthefar-fieldalongwith thepossibilityofusingarectangularprobearraywhichcanefficientlyandaccuratelyprovidethepatternsintheprincipalplanes.Thiscanalsobeusedwhen amplitude-onlydataarecollectedusinganarrayofprobes.Finally,itisshown thattheprobecorrectioncanbeusefulwhenthesizeoftheprobesisthatofa resonantantennaanditisshownthenhowtocarryitout.

InChapter7,twomethodsforsphericalnear-fieldtofar-fieldtransformation arepresented.Thefirstmethodologyisanexactexplicitanalyticalformulation fortransformingnear-fielddatageneratedoverasphericalsurfacetothefarfieldradiationpattern.Theresultsarevalidatedwithexperimentaldata. Acomputerprograminvolvingthismethodisprovidedattheendofthechapter.Thesecondmethodpresentstheequivalentsourceformulationthroughthe SRMdescribedearliersothatitcanbedeployedtothesphericalscanningcase whereonecomponentofthefieldismissingfromthemeasurements.Againthe methodologyisvalidatedusingothertechniquesandalsowithexperimentaldata.

TwodeconvolutiontechniquesarepresentedinChapter8toillustratehow theill-poseddeconvolutionproblemhasbeenregularized.Dependingonthe natureoftheregularizationutilizedwhichisbasedonthegivendataonecan obtainareasonablygoodapproximatesolution.Thetwotechniquespresented herehavebuiltinself-regularizingschemes.Thisimpliesthattheregularization process,whichdependshighlyonthedata,canbeautomatedasthesolution procedurecontinues.Thefirstmethodisbasedonsolvingtheill-poseddeconvolutionproblembytheiterativeconjugategradientmethod.Thesecond methodusesthemethodoftotalleastsquaresimplementedthroughthesingularvaluedecomposition(SVD)technique.Themethodshavebeenappliedto measureddatatoillustratethenatureoftheirperformance.

Chapter9discussestheuseoftheChebyshevpolynomialsforapproximating functionalvariationsarisinginelectromagneticsasithassomeband-limited propertiesnotavailableinotherpolynomials.Next,theCauchymethodbased onGegenbauerpolynomialsforantennanear-fieldextrapolationandthefarfieldestimationisillustrated.Duetovariousphysicallimitations,thereareoften

missinggapsintheantennanear-fieldmeasurements.However,themissing dataisindispensableifwewanttoaccuratelyevaluatethecompletefar-fieldpatternbyusingthenear-fieldtofar-fieldtransformations.Toaddressthisproblem,anextrapolationmethodbasedontheCauchymethodisproposedto reconstructthemissingpartoftheantennanear-fieldmeasurements.Asthe near-fielddatainthissectionareobtainedonasphericalmeasurementsurface, thefarfieldoftheantennaiscalculatedbythesphericalnear-fieldtofar-field transformationwiththeextrapolateddata.Somenumericalresultsaregivento demonstratetheapplicabilityoftheproposedschemeinantennanear-field extrapolationandfar-fieldestimation.Inaddition,theperformanceofthe GegenbauerpolynomialsarecomparedwiththatofthenormalCauchymethod usingPolynomialexpansionandtheMatrixPencilMethodforusingsimulated missingnear-fielddatafromaparabolicreflectorantenna.

Typically,antennapatternmeasurementsarecarriedoutinananechoic chamber.However,agoodanechoicchamberisveryexpensivetoconstruct. Previousresearcheshaveattemptedtocompensatefortheeffectsofextraneous fieldsmeasuredinanon-anechoicenvironmenttoobtainafreespaceradiation patternthatwouldbemeasuredinananechoicchamber.Chapter10illustrates adeconvolutionmethodologywhichallowstheantennameasurementundera non-anechoictestenvironmentandretrievesthefreespaceradiationpatternof anantennathroughthismeasureddata;thusallowingforeasierandmore affordableantennameasurements.Thisisobtainedbymodellingtheextraneousfieldsasthesystemimpulseresponseofthetestenvironmentandutilizinga referenceantennatoextracttheimpulseresponseoftheenvironmentwhichis usedtoremovetheextraneousfieldsforadesiredantennameasuredunderthe sameenvironmentandretrievetheidealpattern.Theadvantageofthisprocess isthatitdoesnotrequirecalculatingthetimedelaytogateoutthereflections; therefore,itisindependentofthebandwidthoftheantennaandthemeasurementsystem,andthereisnorequirementforpriorknowledgeofthetest environment.

Thisbookisintendedforengineers,researchersandeducatorswhoareplanningtoworkinthefieldofelectromagneticsystemcharacterizationandalso dealwiththeirmeasurementtechniquesandphilosophy.Theprerequisiteto followthematerialsofthebookisabasicundergraduatecourseinthearea ofdynamicelectromagnetictheoryincludingantennatheoryandlinearalgebra. Everyattempthasbeenmadetoguaranteetheaccuracyofthecontentofthe book.Wewouldhoweverappreciatereadersbringingtoourattentionany errorsthatmayhaveappearedinthefinalversion.Errorsand/oranycomments maybeemailedtooneoftheauthors,atsalazar@tsc.uc3m.es,mingda.zhu@live. com,hchen43@syr.edu.

Acknowledgments

GratefulacknowledgementismadetoProf.PramodVarshney,Mr.PeterZaehringerandMs.MarilynPoloskyoftheCASECenterofSyracuseUniversityfor providingfacilitiestomakethisbookpossible.ThanksarealsoduetoProf.Jae Oh,Ms.LauraLawsonandMs.RebeccaNobleoftheDepartmentofElectrical EngineeringandComputerScienceofSyracuseUniversityforprovidingadditionalsupport.ThankstoMichaelJamesRice,Systemsadministratorforthe CollegeofEngineeringandComputerScienceforprovidinginformationtechnologysupportinpreparingthemanuscript.AlsothanksareduetoMr.Brett Kurzmanforpatientlywaitingforustofinishthebook.

TapanK.Sarkar

MagdalenaSalazarPalma(salazar@tsc.uc3m.es) Ming-daZhu(mingda.zhu@live.com) HengChen(hchen43@syr.edu) Syracuse,NewYork

TributetoTapanK.Sarkar

ProfessorTapanK.Sarkar,PhD,passed awayon12March2021.Thereviewofthe proofsofthisbookisprobablythelasttask hewasabletoaccomplish.Thus,forus,his coauthors,thisbookwillbealwayscherished andvaluedashislastgifttothescientific community.

Dr.SarkarwasborninKolkata,India,in August1948.HeobtainedhisBachelorof Technology(BT)degreefromtheIndian InstituteofTechnology(IIT),Kharagpur, India,in1969,theMasterofScienceinEngineering(MSCE)degreefromtheUniversity ofNewBrunswick,Fredericton,NB,Canada, in1971,andtheMasterofScience(MS)and Doctoral(PhD)degreesfromSyracuseUniversity,Syracuse,NY,USA,in1975.Hejoined thefacultyoftheElectricalEngineeringandComputerScienceDepartmentat SyracuseUniversityin1979andbecameFullProfessorin1985.Priortothat,he waswiththeTechnicalApplianceCorporation(TACO)DivisionoftheGeneral InstrumentsCorporation(1975–1976).HewasalsoaResearchFellowatthe GordonMcKayLaboratoryforAppliedSciences,HarvardUniversity,Cambridge,MA,USA(1977–1978),andwasfacultymemberattheRochesterInstituteofTechnology,Rochester,NY,USA(1976–1985).ProfessorSarkar receivedtheDoctorHonorisCausadegreefromUniversitéBlaisePascal,ClermontFerrand,France(1998),fromPolytechnicUniversityofMadrid,Madrid, Spain(2004),andfromAaltoUniversity,Helsinki,Finland(2012).Hewasnow emeritusprofessoratSyracuseUniversity.ProfessorSarkarwasaprofessional engineerregisteredinNewYork,USA,andthepresidentofOHRNEnterprises, Inc.,asmallbusinessfoundedin1986andincorporatedintheStateofNew York,USA,performingresearchforgovernment,private,andforeignorganizationsinsystemanalysis.

TributetoTapanK.Sarkar – MagdalenaSalazarPalma,MingDaZhu,andHengChen

Dr.Sarkarwasagiantinthefieldofelectromagnetics,aphenomenal researcherandteacherwhoalsoprovidedaninvaluableservicetothescientific communityinsomanyaspects.

Dr.Sarkarresearchinterestsfocusedonnumericalsolutionstooperator equationsarisinginelectromagneticsandsignalprocessingwithapplication toelectromagneticsystemsanalysisanddesignandwithparticularattention tobuildingsolutionsthatwouldbeappropriateandscalableforpracticaladoptionbyindustry.Amonghismanycontributionstogetherwithhisstudentsand coworkers,itmaybementionedthedevelopmentofthegeneralizedpencil-offunction(GPOF)method,alsoknownasmatrixpencilmethod,forsignalestimationwithcomplexexponentials.BasedonDr.Sarkargroup’sworkonthe originalpencil-of-functionmethod,thetechniqueisusedinelectromagnetic analysesoflayeredstructures,antennaanalysis,andradarsignalprocessing. HeisalsocoauthorofthegeneralpurposeelectromagneticsolverHOBBIES (HigherOrderBasisBasedIntegralEquationSolver).ThelistofProfessorSarkar’soriginalandsubstantivecontributionstothefieldofcomputationalelectromagneticsandantennatheoryisquitelong.Justtonameafew,theseinclude methodsofevaluatingtheSommerfeldintegrals,thealreadymentionedmatrix pencilmethodforapproximatingafunctionbyasumofcomplexexponentials, theconjugategradientmethodandfastFouriertransformmethodfortheefficientnumericalsolutionofintegralequationshavingconvolutionalkernels,the introductionofhigherorderbasisfunctionsinthenumericalsolutionofintegralequationsusingthemethodofmoments,thesolutionoftimedomainproblemsusingtheassociatedLaguerrefunctionsasbasisfunctions,theapplication oftheCauchymethodtothegenerationofaccuratebroadbandinformation fromnarrowbanddata,broadbandantennadesignandanalysis,andnear-field tofar-fieldtransformation,andmanymore.Dr.Sarkar’sworkhasmodernized manysystemsthatincludewirelesssignalpropagation,hasmadepossiblethe designofantennasconsideringtheeffectsoftheplatformswheretheyare deployedforthecurrentandnextgenerationsofairbornesurveillancesystem, andhasdevelopedadaptivemethodologiesthatmadeperformanceofadaptive systemspossibleinrealtime.Hisadvancedcomputationaltechniqueshave beenimplementedforparallelprocessingonsupercomputersforfastandefficientsolutionofextremelylargeelectromagneticfieldproblems.Hehasalso developedantennasystemsandprocessingforultrawidebandapplications. Heappliedphotoconductiveswitchingtechniquesforgenerationofkilovolts amplitudeelectricalpulsesofsubnanoseconddurationwithapplicationsin manyfieldsincludinglowprobabilityinterceptradarsystems.Itisremarkable thatDr.Sarkarhasbeenabletokeepinnovatingforsuchasustainedperiodof timethroughouthiscareer.ProfessorSarkarhasauthoredorcoauthoredmore than380journalarticles,innumerablecontributionsforconferencesandsymposia,16booksand32bookchapters,with24549citationsandh-indexof74 (GoogleScholar).Inthepast,hewaslistedamongtheISIHighlyCited

Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.
Modern characterization of electromagnetic systems and its associated metrology (wiley - ieee) 1st e by Education Libraries - Issuu