《华中科技大学博士论文答辩会》由会员分享，可在线阅读，更多相关《华中科技大学博士论文答辩会（43页珍藏版）》请在金锄头文库上搜索。
1、2011 Computer Communications Architectures for the future networks and the next generation Internet: A surveyPresented by: 张成伟Email: zhangcwhust.edu Phone: 1398 6214 512Author:Subharthi Paul, Jianli Pan, Raj JainDepartment of Computer Science and Engineering, Washington University in Saint Louis, Un
2、ited StatesOutline1.Introduction2.Internet 3.0 (GINA)3.Content Delivery Mechanisms4.Content-Centric Networks(CCN)5.Data-Oriented Network Architecture (DONA)6.Challenged Network Environments7.Future Internet Projects8.Top 10 Features of Next Generation InternetResearch Study Report29/18/2024Introduct
3、ionnInternet 1.0 (1969 1989) Research projectRFC1 is dated April 1969.ARPA project started a few years earlier.IP, TCP, UDPMostly researchersIndustry was busy with proprietary protocols: SNA, DECnet,AppleTalk, XNSnInternet 2.0 (1989 Present) CommerceSecurity RFC1108 in 1989NSFnet became commercialIn
4、ter-domain routing:BGP (Policy-based)Address Shortage IPv6Congestion Control, Quality of Service,nWhat the like of the Next Generation Internet?Research Study Report39/18/2024IntroductionnWhy we need the next Generation Internet?SecurityNo concept of ownershipIdentity and location in one (IP Address
5、)Assumes live and awake end-systemNo representation for real end system: the humannClean-State Design Not take the current architecture into considerationall clean, all freshResearch Study Report49/18/2024R. Jain, Internet 3.0: Ten Problems with Current Internet Architecture and Solutions for the Ne
6、xt Generation, Proceedings of Military Communications Conference (MILCOM 2006), Washington, DC, October 23-25, 2006Names, IDs, LocatorsName: John SmithID: 012-34-5678Locator:1234 Main StreetBig City, MO 12345USAnLocator changes as you move, ID and Names remain the same.nExamples:nNames: Company name
7、s, DNS names (Microsoft)nIDs: Cell phone numbers, 800-numbers, Ethernet addresses,nSkype ID, VOIP Phone numbernLocators: Wired phone numbers, IP addressesResearch Study Report59/18/2024Internet 3.0(GINA)nInternet 3.0 is the name of the Washington University project on the next generation InternetnGo
8、al 1: Represent the commercial reality of distributed Internet ownership and organizationnGoal 2: Develop a clean slate architecture to overcome limitations of the current internetnGoal 3: Develop an incremental approach to implement the architectureResearch Study Report69/18/2024nObject = Addressab
9、le EntitynCurrent: End-Systems and Intermediate SystemsnGINA:Computers, Routers/Firewalls.NetworksHumansCompanies, Departments, Cities, States, Countries, Power gridsProcess in a computerRecursive Set of Objects is also one object, ne.g., Networks of NetworksObjects in GINAResearch Study Report79/18
10、/2024You can connect to a human, organization, or a departmentRealms in GINAnObject names and Ids are defined within a realmnA realm is a logical grouping of objects under an administrative domainnThe Administrative domain may be based on Trust RelationshipsnA realm represents an organizationRealm m
11、anagers set policies for communicationsRealm members can share servicesObjects are generally members of multiple realmsRealm Boundaries: nOrganizational, Governmental, ISP, P2P,Research Study Report89/18/2024Realm = Administrative GroupGINA: Multi-Tier Object-Oriented ViewnObjects provide services.
12、Higher tiers specify the requirementsnTier service broker (shown by dotted line) composes a servicecan negotiate with multiple realms in that tiernHigher tier may not/need not find details of lower tiersResearch Study Report99/18/2024Content distribution mechanismsnNext generation CDNnNext generatio
13、n P2PnSwarming architecturenContent Centric NetworkingResearch Study Report109/18/2024Content distribution mechanismsnNext generation CDNNeither the server nor the clients have any control over the “middle mile”Fat File Paradox:nit is the length of the pipe rather than its width that determines how
14、fast a large file can travel through itHighly distributed CDNs place servers at the edge networksSecurity, management, scalability and synchronization problems.Research Study Report119/18/2024Next generation P2PnThe self-organizing and self-healing properties of P2P networks have the potential to be
15、come the predominant content distribution mechanism of the future Internet.nBandwidth provision (Asymmetric)nDynamic sharingnTussle of interests between P2P networks and ISPsAware of the underlying topology and location of peers.Research Study Report129/18/2024Swarming architecturenUswarmA “swarm” i
16、s a set of loosely connected hosts that act in a selfish and highly decentralized manner to provide local and system level robustness through active adaptation. (BitTorrent)A unified swarming modelResearch Study Report139/18/2024Content-Centric NetworksnIP cares about “Where”: forward packets from A
17、 to BnUsers care about “What”: Movie XnReplace “packets” with “Data Objects” or “Interests” (requests)nReplace “Addresses” with “Names of Objects”Research Study Report149/18/2024V. Jacobson, et al, “Networking Named Content,” to appear in CoNEXT 2009, December 2009Content-Centric Networks: Routing P
18、olicynContent Store(CS): Local cache of datanPending Interest Table (PIT): Recent requests forwardednForwarding Information Base (FIB): Known data locationsnFaces: Requesting processes and hardware interfacesResearch Study Report159/18/2024Content-Centric Networks: Routers OperationnApplications sen
19、d “Interest” in data XnRouter looks up in local store and sends if foundnRouter looks up in PIT, if entry already existssomeone requested it recently adds the interestface to the same entrynRouter looks up in FIB, if entry exists data location is knowna PIT entry is made the interest is multicasted
20、to all faces in the FIB entry the FIB entry is removednIf there is no FIB entry, interest is discarded router does not know how to get the datanWhen data arrives, Content Store match duplicate, discardPIT match Forward to all facesFIB match No PIT Unsolicited DiscardnData providers register their da
21、ta Creates FIB entriesResearch Study Report169/18/2024DONA :Data-Oriented Network ArchitecturenDONA proposes a novel mechanism of explicitly naming the data or service and routing on these names for data or service access.nDONA proposes replacing DNS names with flat, self-certifying names, and repla
22、cing DNS name resolution with a name-based anycast primitive that lives about the IP layer.Research Study Report179/18/2024DONA: Basic DesignGoalCurrent InternetDONA DesignProvided byPersistenceDNS, HTTP redirectNamesFlat names, remain invariantAuthenticityIPsec, PKI,TLSNamesSelf-certifying names, e
23、nable easy authenticationAvailabilityCDNs, P2PName resolutionRoute-by-name paradigmnPersistence The name of a service or data object remains valid as long as the service or data are availablenAvailability Data or service should have a high degree of reliability and acceptable latencynAuthenticityDat
24、a can be verified to have come from a particular sourceResearch Study Report189/18/2024T. Koponen, M. Chawla, B. chun, et al., “A data oriented (and beyond) network architecture,” ACM SIGCOMM Computer Communication Review, Volume 37, Issue 4, pp 181-192, October 2007.DONA: Naming MechanismnDONA name
25、s are organized around principals. Each principal is associated with a public-private key pairEach datum or service or any other named entity (host, domain, etc.) is associated with a principalnNames are of the form P:LP is the cryptographic hash of the principals public keyL is a label chosen by th
26、e principal, which ensures that these names are uniquenGranularity of naming is left up to principalsA principal might choose to just name her web site, or name her web site and each page within it, or name at a finer granularityResearch Study Report199/18/2024DONA: Name ResolutionnName Resolutions
27、goal is to achieve high availability, by finding close-by copies and avoiding failures.nRoute-by-name paradigmnInfrastructure :Resolution handlers(RHs)nTwo basic primitives operations in this designFIND(P:L) and REGISTER(P:L)nFIND(P:L) locate the object named P:LnREGISTER messages set up the state n
28、ecessary for the RHs to route FINDs effectivelyResearch Study Report209/18/2024nAny machine authorized to serve a datum or service with name P:L sends a REGISTER(P:L) command to its first-hop RHnRH maintains a registration table that maps a name to both a new-hop RH and the distance to the copy (in
29、some metric)nREGISTERS are forwarded according to interdomain policies:REGISTERs from customers (child) to both peers and providersREGISTERs from peers optionally to providers/peersEstablishing REGISTER state9/18/202421Research Study ReportForwarding FIND(P:L)nWhen FIND(P:L) arrives to a RH:If there
30、 is an entry in the registration table, the FIND is sent to the next-hop RHIf there is no entry, the RH forwards the FIND towards to its provider (parent)nIn case of multiple equal choices, the RH uses its local policy to choose among themResearch Study Report229/18/2024DONA: ChallengeGoalConcernApp
31、roachSecurityPrivate key will get compromisedKey revocation mechanism neededUsabilityFlat names are hard to rememberExternal mechanisms neededScalabilityHuge flat namespaceMore shortly(local address)Research Study Report239/18/2024DONA: Using FunctionalitiesnUsing FunctionalitiesUsing name-based any
32、castnServer SelectionnMobility and MultihomingnMulticast State EstablishmentnImproving content deliveryCachingSubscriptionsAvoiding Misbehaving and Overloaded ServersResearch Study Report249/18/2024CCN VS. DONAnBoth DONA and NNC, advocate a paradigm shift from the present host centric architecture o
33、f the Internet to a data centric architecture. nNNC proposes a network-wide caching mechanism at various network nodes, leveraging the dipping cost of persistent storage and defining an efficient content dissemination system as an overlay over the present IP networks. nDONA on the other hand emphasi
34、zes a novel mechanism for the naming of content and name resolution to build an architecture around service and data access.Research Study Report259/18/2024Challenged Network EnvironmentsnDelay Tolerant NetworkBundle protocolnDelay/fault tolerant mobile sensor networks (DFT-MSN)nPostcards from the e
35、dgenDisaster day after networks (DAN)nSelectively Connected Networking (SCN)Research Study Report269/18/2024Future Internet ProjectsnIn 2005 US National Science Foundation started a large research and infrastructure program on next generation Internet nQ: How would you design Internet today? Clean s
36、late design.n“Future Internet Design” (FIND): 48+ projectsnnets-nStanford, MIT, Berkeley, CMU, n“An Architecture for Diversified Internet” at WUSTLn“Global Environment for Networking Innovations” (GENI):29+ projectsn :/ Union: 7th Framework programn :/cordis.europa.eu/fp7/nJapan: AKARI (A small ligh
37、t in the dark pointing to the future)n :/akari-project.nict.go.jp/eng/index2.htmnChina, Korea, Australia, 20+ countriesResearch Study Report279/18/2024Next Generation TestbedsnPast: PlanetLab, Emulab, VINI, OneLabnFederationnGENI, Requirements, SubsystemsnGENI Prototype ClustersnWireless Network Vir
38、tualizationnSupercharged PlanetLab Platform (SPP)nFIRE, FEDERICAnAKARIResearch Study Report289/18/2024PlanetLabnGlobal networking research testbedn1055 nodes at 490 sites Nov 2009nResearchers use it experiment with new ideas on distributed storage, network mapping, peer-to-peer systems, distributed
39、hash tables, and query processingResearch Study Report299/18/2024http:/www.planet-lab.org/PlanetLab (Cont.)nLinux virtual server software on Interned nodesnSlivers = Piece of a resourcenNode manager (NM) manages the nodes virtual serversnPlanet Lab Control (PLC) interacts with NMnExperimenters reque
40、st a Slice = slivers in various sitesResearch Study Report309/18/2024EmulabnNetworking research testbed at University of UtahnAvailable for public use for research and educationnSoftware implemented at two dozen sites around the worldnAllows simulated links and nodes in slicesnAllows fault studiesnP
41、rovides repeatabilityResearch Study Report319/18/2024http:/ for MassesnMake it easy for naive usersnFunction in infrastructure poor environmentsnIntermittent powernDevice sharingnAllow sneaker netsnLong distance wireless CDMA450Research Study Report329/18/2024OneLabnFederation of European PlanetLabs
42、nEnhanced monitoring infrastructurenWireless TestbedsnIPv6 multihomingnEmulation toolsnDeal with unstable connectivityResearch Study Report339/18/2024http:/www.onelab.eu/ /FederationnLarger testbedsnTestbeds for specialized resources such as access technologiesnSpecialized research communities and c
43、ross-disciplinenChallenges:Homogenization of diverse contextInteroperability of security protocolsPolitical or social-economic issuesIntellectual Property rightsCommercial and non-commercial interestsResearch Study Report349/18/2024GENInGlobal Environment for Network InnovationsnDedicated shared sub
44、strate facility for large-scale experimentsnUS National Science Foundation projectnDedicated backbone links through LambdaRail and Internet2nDiverse and extensible set of technologiesResearch Study Report359/18/2024http:/ RequirmentsnSliceability: Sharing with isolation.nProgrammability: All compone
45、nts should be programmablenVirtualization: Slicing via virtualization or space/time sharing.nFederation: Combination of independently owned testbedsnObservability: Allow specifiable measurement frameworknSecurity: Should not harm production InternetResearch Study Report369/18/2024GENI SubsystemsRese
46、arch Study Report379/18/2024GENI Prototype ClustersnFive Clusters in Spiral 1:n1.Trial Integration Environment with DETER (TIAD):Emulab based security experiments testbedn2.PlanetLab: Federate all slice-based substrates PlanetLab, Emulab, VINI, and GENIn3.ProtoGENI: Federation of Emulab testbeds,Enh
47、anced Emulab Controln4. Open Resource Control Architecture (ORCA):Resource manager runs under the host operating systemUses virtualization to allocate containersn5. Open Access Research Testbed (ORBIT): Wireless testbed with emulated and real nodesResearch Study Report389/18/2024GENI Spiral 1, http:
48、/ 10 Features of Next Generation InternetnSecuritynMobilitynUser/Data-Centric: Network support of data objectsnEasy to use: Self-organizing, better user controlnDisruption TolerantnGreen: Proxy, Sleep Modes,(Green Network)nServices: Storage, Translation, MonitoringnOrganizational RepresentationnVirt
49、ualizable to create Application Specific ContextnPolicy EnforcementResearch Study Report399/18/2024SummarynNSF FIND program has funded a number of architectural component research programs.nGENI testbed consists of 5 clusters in Spiral 1nFIRE in Europe and AKARI in Japan are similar to GENI.nInterne
50、t 3.0 is an industry sponsored full architecture programResearch Study Report409/18/2024Other ReferencesnInternet 3.0 Talk by Raj Jainn :/research.microsoft/apps/video/default.aspx?id=104012nR. Jain, Internet 3.0: Ten Problems with Current Internet Architecture and Solutions for the Next Generation,
51、 Proceedings of Military Communications Conference (MILCOM 2006), Washington, DC, October 23-25, 2006nV. Jacobson, et al, “Networking Named Content,” to appear in CoNEXT 2009, December 2009nT. Koponen, M. Chawla, B. chun, et al., “A data oriented (and beyond) network architecture,” ACM SIGCOMM Compu
52、ter Communication Review, Volume 37, Issue 4, pp 181-192, October 2007.Research Study Report419/18/2024Future TestbedsnOneLab: :/onelab.eu/ nUser cases : :/onelab.eu/index.php/services/testbed-access/use-cases.htmlnGENInFresh user: :/ Experiments : :/ Archive : :/ Study Report429/18/2024ThanksQ&A2011-8-4
