ebook img

Machine Learning for Future Fiber-Optic Communication Systems PDF

404 Pages·2022·21.598 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Machine Learning for Future Fiber-Optic Communication Systems

MACHINE LEARNING FOR FUTURE FIBER-OPTIC COMMUNICATION SYSTEMS This page intentionally left blank MACHINE LEARNING FOR FUTURE FIBER-OPTIC COMMUNICATION SYSTEMS Editedby ALANPAKTAOLAU FAISALNADEEMKHAN AcademicPressisanimprintofElsevier 125LondonWall,LondonEC2Y5AS,UnitedKingdom 525BStreet,Suite1650,SanDiego,CA92101,UnitedStates 50HampshireStreet,5thFloor,Cambridge,MA02139,UnitedStates TheBoulevard,LangfordLane,Kidlington,OxfordOX51GB,UnitedKingdom Copyright©2022ElsevierInc.Allrightsreserved. MATLAB®isatrademarkofTheMathWorks,Inc.andisusedwithpermission. TheMathWorksdoesnotwarranttheaccuracyofthetextorexercisesinthisbook. Thisbook’suseordiscussionofMATLAB®softwareorrelatedproductsdoesnotconstituteendorsementor sponsorshipbyTheMathWorksofaparticularpedagogicalapproachorparticularuseoftheMATLAB®software. Nopartofthispublicationmaybereproducedortransmittedinanyformorbyanymeans,electronicor mechanical,includingphotocopying,recording,oranyinformationstorageandretrievalsystem,without permissioninwritingfromthepublisher.Detailsonhowtoseekpermission,furtherinformationaboutthe Publisher’spermissionspoliciesandourarrangementswithorganizationssuchastheCopyrightClearanceCenter andtheCopyrightLicensingAgency,canbefoundatourwebsite:www.elsevier.com/permissions. ThisbookandtheindividualcontributionscontainedinitareprotectedundercopyrightbythePublisher(other thanasmaybenotedherein). Notices Knowledgeandbestpracticeinthisfieldareconstantlychanging.Asnewresearchandexperiencebroadenour understanding,changesinresearchmethods,professionalpractices,ormedicaltreatmentmaybecomenecessary. Practitionersandresearchersmustalwaysrelyontheirownexperienceandknowledgeinevaluatingandusingany information,methods,compounds,orexperimentsdescribedherein.Inusingsuchinformationormethodsthey shouldbemindfuloftheirownsafetyandthesafetyofothers,includingpartiesforwhomtheyhaveaprofessional responsibility. Tothefullestextentofthelaw,neitherthePublishernortheauthors,contributors,oreditors,assumeanyliability foranyinjuryand/ordamagetopersonsorpropertyasamatterofproductsliability,negligenceorotherwise,or fromanyuseoroperationofanymethods,products,instructions,orideascontainedinthematerialherein. LibraryofCongressCataloging-in-PublicationData AcatalogrecordforthisbookisavailablefromtheLibraryofCongress BritishLibraryCataloguing-in-PublicationData AcataloguerecordforthisbookisavailablefromtheBritishLibrary ISBN:978-0-323-85227-2 ForinformationonallAcademicPresspublications visitourwebsiteathttps://www.elsevier.com/books-and-journals Publisher:MaraConner AcquisitionsEditor:TimPitts EditorialProjectManager:MicaEllaOrtega ProductionProjectManager:SuryaNarayananJayachandran Designer:ChristianJ.Bilbow TypesetbyVTeX Dedicated to my family for their unwavering love, support, and encouragement. Without their monumental help, this book would never have come to fruition. − Faisal Nadeem Khan To my parents for their unconditional love and support throughout my life. To my wife Sandy and my son Sean, who give me meaning and purpose in life. − Alan Pak Tao Lau This page intentionally left blank Contents Contributors xi Preface xv Acknowledgments xvii 1. Introductiontomachinelearningtechniques:Anopticalcommunication’s perspective 1 FaisalNadeemKhan,QiruiFan,ChaoLu,andAlanPakTaoLau 1.1. Introduction 1 1.2. Supervisedlearning 5 1.3. Unsupervisedlearning 18 1.4. Reinforcementlearning(RL) 24 1.5. Deeplearningtechniques 27 1.6. FutureroleofMLinopticalcommunications 37 1.7. OnlineresourcesforMLalgorithms 37 1.8. Conclusions 39 Appendix1.A 39 References 40 2. Machinelearningforlong-haulopticalsystems 43 ShaoliangZhangandChristianHäger 2.1. Introduction 43 2.2. Applicationofmachinelearninginperturbation-basednonlinearitycompensation 44 2.3. Applicationofmachinelearningindigitalbackpropagation 52 2.4. Outlookofmachinelearninginlong-haulsystems 61 References 62 3. Machinelearningforshortreachopticalfibersystems 65 BorisKaranov,PolinaBayvel,andLaurentSchmalen 3.1. Introductiontoopticalsystemsforshortreach 65 3.2. Deeplearningapproachesfordigitalsignalprocessing 67 3.3. OpticalIM/DDsystemsbasedondeeplearning 69 3.4. Implementationonatransmissionlink 84 3.5. Outlook 86 References 87 4. Machinelearningtechniquesforpassiveopticalnetworks 91 LilinYi,LuyaoHuang,ZhengxuanLi,YongxinXu,andWantingXu 4.1. Background 91 4.2. ThevalidationofNNeffectiveness 93 4.3. NNfornonlinearequalization 98 vii viii Contents 4.4. Endtoenddeeplearningforoptimalequalization 106 4.5. FPGAimplementationofNNequalizer 109 4.6. Conclusionsandperspectives 111 References 112 5. End-to-endlearningforfiber-opticcommunicationsystems 115 OgnjenJovanovic,FrancescoDaRos,MetodiYankov,andDarkoZibar 5.1. Introduction 115 5.2. End-to-endlearning 117 5.3. End-to-endlearningforfiber-opticcommunicationsystems 120 5.4. Gradient-freeend-to-endlearning 133 5.5. Conclusion 134 Acknowledgments 136 References 136 6. Deeplearningtechniquesforopticalmonitoring 141 TakahitoTanimuraandTakeshiHoshida 6.1. Introduction 141 6.2. Buildingblocksofdeeplearning-basedopticalmonitors 144 6.3. Deeplearning-basedopticalmonitors 150 6.4. TipsfordesigningDNNsforDL-basedopticalmonitoring 163 6.5. Experimentalverifications 167 6.6. Futuredirectionofdata-analytic-basedopticalmonitoring 182 6.7. Summary 184 Acknowledgment 185 References 185 7. MachineLearningmethodsforQuality-of-Transmissionestimation 189 MemedheIbrahimi,CristinaRottondi,andMassimoTornatore 7.1. Introduction 189 7.2. ClassificationandregressionmodelsforQoTestimation 191 7.3. ActiveandtransferlearningapproachesforQoTestimation 199 7.4. OntheintegrationofMLinoptimizationtools 205 7.5. Illustrativenumericalresults 208 7.6. Futureresearchdirectionsandchallenges 219 7.7. Conclusion 221 References 221 8. MachineLearningforopticalspectrumanalysis 225 LuisVelasco,MarcRuiz,BehnamShariati,andAlbaP.Vela 8.1. Introduction 225 8.2. Feature-basedspectrummonitoring 230 8.3. Residual-basedspectrummonitoring 247 Contents ix 8.4. Monitoringoffilterlessopticalnetworks 262 8.5. Concludingremarksandfuturework 276 Listofacronyms 278 References 279 9. Machinelearninganddatascienceforlow-marginopticalnetworks 281 CamilleDelezoide,PetrosRamantanis,andPatriciaLayec 9.1. Theshapeofnetworkstocome 281 9.2. CurrentQoTmargintaxonomyanddesign 283 9.3. Generalizationofopticalnetworkmargins 284 9.4. Largescaleassessmentofmarginsandtheirtimevariationsinadeployednetwork 290 9.5. Trade-offbetweencapacityandavailability 296 9.6. Data-drivenrateadaptationforautomatednetworkupgrades 302 9.7. Machinelearningforlow-marginopticalnetworks 306 9.8. Conclusion 313 References 314 10.Machinelearningfornetworksecuritymanagement,attacks,andintrusions detection 317 MarijaFurdekandCarlosNatalino 10.1. Physicallayersecuritymanagement 317 10.2. Machinelearningtechniquesforsecuritydiagnostics 320 10.3. AccuracyofMLmodelsinthreatdetection 325 10.4. RuntimecomplexityofMLmodels 328 10.5. InterpretabilityofMLmodels 332 10.6. Openchallenges 333 10.7. Conclusion 335 Acknowledgments 335 References 335 11.Machinelearningfordesignandoptimizationofphotonicdevices 337 KeisukeKojima,ToshiakiKoike-Akino,YinghengTang,YeWang,andMatthewBrand 11.1. Introduction 337 11.2. Deepneuralnetwork(DNN)models 339 11.3. Nanophotonicpowersplitter 341 11.4. Metasurfacesandplasmonics 357 11.5. Othertypesofopticaldevices 365 11.6. Discussion 367 11.7. Conclusion 369 References 370 Index 375

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.