Advances in Network Analysis and its Applications PDF
Preview Advances in Network Analysis and its Applications
MATHEMATICS IN INDUSTRY 18 Editors Hans-GeorgBock FrankdeHoog AvnerFriedman ArvindGupta Andre´ Nachbin HelmutNeunzert WilliamR.Pulleyblank TorgeirRusten FadilSantosa Anna-KarinTornberg Forfurthervolumes: http://www.springer.com/series/4650 Evangelos Kranakis Editor Advances in Network Analysis and its Applications With 95 Figures and 51 Tables 123 Editor EvangelosKranakis SchoolofComputerScience CarletonUniversity Ottawa,Ontario Canada ISSN1612-3956 ISBN978-3-642-30903-8 ISBN978-3-642-30904-5(eBook) DOI10.1007/978-3-642-30904-5 SpringerHeidelbergNewYorkDordrechtLondon LibraryofCongressControlNumber:2012950143 MathematicsSubjectClassification(2010):68M10,68M11,94A60,94A62,91D30 ©Springer-VerlagBerlinHeidelberg2013 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof thematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation, broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionorinformation storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodology nowknownorhereafterdeveloped.Exemptedfromthislegalreservationarebriefexcerptsinconnection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’slocation,initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer. PermissionsforusemaybeobtainedthroughRightsLinkattheCopyrightClearanceCenter.Violations areliabletoprosecutionundertherespectiveCopyrightLaw. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse. While the advice and information in this book are believed to be true and accurate at the date of publication,neithertheauthorsnortheeditorsnorthepublishercanacceptanylegalresponsibilityfor anyerrorsoromissionsthatmaybemade.Thepublishermakesnowarranty,expressorimplied,with respecttothematerialcontainedherein. Printedonacid-freepaper SpringerispartofSpringerScience+BusinessMedia(www.springer.com) Preface Networks composed of interacting, communicating, and co-operating processes providepowerfulmodelsforunderstandingthebehaviorofcomplexsystems.The explosivedevelopmentoftheInternetinthelastdecadehasmadethempervasivein all aspects of our lives. It has also made possible their emergenceas new models for applying computational techniques for solving and providing understanding for old problems with new insights. As such they often provide the framework for new methodologiesthat lead to better decision making in many fields such as transportation,communication,health,finance,andsocialengineering.Solvingthe manyemergingproblemsinthisareahasrequiredthecollaborationofresearchers from fields as diverse as mathematics, computer science, biology, economics, sociology,managementscience,andengineering. Papers included in this volume originate from participants in a sequence of sevenworkshopsonmathematicsofnetworking(FP-NETS)Iorganizedonbehalf of Mitacs in the period 2010–2012. One important area of Mitacs research was the mathematical study and analysis of complex systems as this relates to and is inspired by the study and research and development of information technologies inthescientificandengineeringcommunities.Amajorideaunderpinningmanyof Mitacs’pastresearchprojectswasalsorelatedtothatofdynamicnetworkanalysis whereby interacting communicating entities process, exchange, and compute in order to attain optimal design goals. Applications can be found in all scientific and engineering areas: from wireless communication to network security, from cooperative and large-scale computing to social networking, and from financial analysisandriskassessmenttocyber-warfareandunderstandingofwar. Overall, it appears that the networking field is somewhat siloed, with research approaches to problems in one type of networks, say biological, having common elements with those in linguistic networks. For example, many questions can be rephrased as shortest path problems, routing problems, max flow problems, min costflowproblems,etc.,andheuristictechniquesdevelopedinonesub-areamaybe applicable in others. At the same time, networking often transcends the scientific boundaries of traditional fields like biology, economics, physics, and computer science.Forexample,commonalitiesareeasilyfoundinthestudyoftheformation v vi Preface ofcellularsystemsbybiologists,theoriginsofcompanynetworksbyeconomists, alignmentofatomsbyphysicists,anddesignofnetworksofcomputersbycomputer scientists. In general, we are also interested in exploring how to organize and facilitate informationacquisition,processingandactinginalarge-scalesystemusingadaptive techniques and local, sometimes restricted, channels of communication. One of the most powerfultools to emergefromthe study of networkingis computational methodologies that allow the testing and exploration of a wider range of more realistic models that may include dynamic parameters such as noise, motion, locality, etc. At the same time, models developed have been enriched by a vast wealth ofapplicablemathematicalmethodologiesrangingfromprobabilitytheory and statistics to graph theory, from combinatorial optimization to mathematical analysisandPDEs,fromnumbertheorytoalgebra,andfromdistributedcomputing to mechanism design that find applications in networking. The goal of this focus period on networking was to highlight application areas relevant to network analysis, identify new mathematicalresearch areas that may provideinsights, and enablecross-fertilizationofideas. ThefocusperiodFP-NETS:FocusPeriodonRecentAdvancesinNetworkingor- ganizedandcoordinatedconferences,problemsolvingworkshops,summerschools, plenary talks, and industrial academic panel discussions in selected, key areas of networking. The aim was to organize and run events pertinent to networking, promotethecross-fertilizationofnewideas,aswellastosupporttheparticipationof leadingexperts,facultymembers,postdocs,andstudentsfromCanadianuniversities andinternationalpartnerorganizations.Activitiesincluded(1)tutorialsthatbrought students and interested researchers up to speed, (2) invited talks by leaders in the fieldthatilluminatedstate-of-the-artproblems,(3)contributedtalksbyresearchers, (4) panel discussions that elaborated and discussed important issues transcending current research problems, and (5) industrial and interdisciplinary exchanges. It also providedseveralopportunitiesforacademicsto brainstormwith researchend usersandidentifyrelevantopenproblems.Theoutcomefromeachconferenceand workshopincludedtheidentificationofopenproblemsandideassuitableforfurther exploration and collaboration. The focus period ended with a problem-solving interdisciplinary workshop with selected participants from all the networking workshops to interact and share expertise and ideas on important problems in networking. Overall, the focus period FP-NETS attempted to provide a diverse and com- prehensive forum to all interested researchers for understanding the most recent advancesanddevelopmentsinthisimportantarea.Inparticular,therewereactivities inthefollowingnetworkingthemes:(1)WirelessNetworkingandMobileComput- ing,(2)NetworkSecurityandCryptography,(3)SocialNetworks,(4)Internetand Network Economics,(5) BiologicalNetworksand Systems Biology,(6) Financial Networks and Risk Assessment, as well as a Problem solving workshop which concentratedon the solution of openproblemsresulting fromthe workshops.The currentproceedingsrepresentonlysamplesof extensivediscussionsand scientific Preface vii presentationsfromthreeoftheseworkshops,namely,FinancialNetworks,Network SecurityandCryptography,andSocialNetworks. Needlesstosay,organizingalltheseeventswouldhavenotbeenpossiblewithout the support and encouragementI received from several people and organizations. Firstofall, inthelast14years,myinvolvementwith Mitacs(whenitmanagedan NCE grant for research in the mathematicalsciences) has been pivotalin shaping andenhancingmyevolvingunderstandingofthenatureandbeautyofmathematics. Mitacs has been a truly transformative organization in its efforts to change the mathematical culture not only in Canada but globally. The scientific discussions with the other members of the Research Management Committee shaped my mathematical focus. My interactions with Arvind Gupta have been inspirational. ThelogisticalsupportandefficiencyofOlgaStachovewerealwaystrulyamazing. NilimaNigamandRebeccahMarshwereveryhelpfulintheinitialstagesofthe planningwhileMichaelLynchwasalwayspresentinsupportingandguidingallthe organizationalaspectsoftheevents.AlsomanythankstoOscarMoralesPoncefor helpingtointegratetheelectronicfilesintoasinglevolume. Ottawa,ON,Canada EvangelosKranakis Contents PartI FinancialNetworks 1 MathematicalModelingofSystemicRisk ............................... 3 HamedAminiandAndreeaMinca 1.1 Introduction........................................................... 3 1.2 FinancialLinkagesandContagion .................................. 5 1.2.1 FinancialNetworks......................................... 7 1.2.2 InsolvencyCascades........................................ 8 1.2.3 IlliquidityCascades......................................... 11 1.2.4 LiquidationandPriceFeedbackEffects................... 15 1.3 RandomFinancialNetworkModels................................. 16 1.4 AsymptoticAnalysisofDefaultCascades.......................... 20 References.................................................................... 24 2 SystemicRiskinBankingNetworksWithoutMonteCarlo Simulation................................................................... 27 JamesP.Gleeson,T.R.Hurd,SergeyMelnik, andAdamHackett 2.1 Introduction........................................................... 27 2.2 ModelsofContagioninBankingNetworks ........................ 29 2.2.1 GeneratingModelNetworks............................... 30 2.2.2 ContagionMechanisms .................................... 32 2.2.3 LiquidityRisk............................................... 34 2.2.4 MonteCarloSimulations................................... 35 2.3 Theory ................................................................ 35 2.3.1 ThresholdsforDefault ..................................... 35 2.3.2 GeneralTheory............................................. 37 2.4 SimplifiedTheory.................................................... 39 2.4.1 SimplifiedTheoryforGK.................................. 39 2.4.2 FrequencyofContagionEvents............................ 42 ix x Contents 2.5 Results ................................................................ 42 2.5.1 GKModel................................................... 42 2.5.2 NYYABenchmarkCase ................................... 43 2.5.3 NetworkswithFat-TailedDegreeDistributions........... 45 2.6 Discussion ............................................................ 48 References.................................................................... 55 3 SystemicValuationofBanks:InterbankEquilibriumandContagion 57 GrzegorzHałaj 3.1 Introduction........................................................... 58 3.2 TheModel............................................................ 61 3.2.1 ValuationFundamentals.................................... 61 3.2.2 InterbankLiquidityandFunding .......................... 62 3.2.3 TransferofCreditRisk..................................... 63 3.2.4 TheEquilibrium ............................................ 64 3.2.5 SecondaryDefaults:DominoEffect....................... 66 3.3 ExistenceofEquilibriumandNumericalProcedure................ 67 3.3.1 HowDoestheEquilibriumWork?AnExample.......... 67 3.4 BacktoBankValuationFormula.................................... 69 3.5 ValuationofUSBanks............................................... 70 3.5.1 Data ......................................................... 71 3.5.2 Simulation .................................................. 73 3.5.3 Discussion .................................................. 74 3.6 Conclusions........................................................... 79 3.7 IsotoneΨ:TheProof................................................. 80 References.................................................................... 82 4 AnOpenProblem .......................................................... 85 JohnB.Walsh PartII NetworkSecurity 5 DynamicTrustManagement:NetworkProfilingforHigh AssuranceResilience....................................................... 91 MikeBurmesterandW.OwenRedwood 5.1 Introduction........................................................... 91 5.2 OverviewofRelatedWork........................................... 92 5.2.1 AccessControlandTrustManagement ................... 92 5.2.2 IntrusionDetection/PreventionSystems................... 93 5.2.3 SignatureDetectionSystems............................... 94 5.2.4 AnomalyDetectionSystems............................... 95 5.2.5 BinaryVersusGraduatedResponseMechanisms......... 96 5.3 ThreatManagement.................................................. 96 5.3.1 ADynamicTrustManagementInfrastructure ............ 96 5.3.2 HowtheThreatLevelChanges ............................ 98 5.3.3 FeatureSelection ........................................... 99
The list of books you might like
