Presented by: 张成伟 Email: zhangcw@hust Phone: 1398 6214 512

1. 2011 Computer CommunicationsArchitectures for the future networks and the next generation Internet: A survey Author:Subharthi Paul, Jianli Pan, Raj Jain Department of Computer Science and Engineering, Washington University in Saint Louis, United States Presented by: 张成伟 Email: zhangcw@hust.edu.cn Phone: 1398 6214 512

2. Outline Introduction Internet 3.0 (GINA) Content Delivery Mechanisms Content-Centric Networks(CCN) Data-Oriented Network Architecture (DONA) Challenged Network Environments Future Internet Projects Top 10 Features of Next Generation Internet Research Study Report

3. Introduction • Internet 1.0 (1969 – 1989) – Research project • RFC1 is dated April 1969. • ARPA project started a few years earlier. • IP, TCP, UDP • Mostly researchers • Industry was busy with proprietary protocols: SNA, DECnet,AppleTalk, XNS • Internet 2.0 (1989 – Present) – Commerce • Security RFC1108 in 1989 • NSFnet became commercial • Inter-domain routing: • BGP (Policy-based) • Address Shortage IPv6 • Congestion Control, Quality of Service,… • What the like of the Next Generation Internet? Research Study Report

4. Introduction R. Jain, ``Internet 3.0: Ten Problems with Current Internet Architecture and Solutions for the Next Generation," Proceedings of Military Communications Conference (MILCOM 2006), Washington, DC, October 23-25, 2006 • Why we need the next Generation Internet? • Security • No concept of ownership • Identity and location in one (IP Address) • Assumes live and awake end-system • No representation for real end system: the human • Clean-State Design • Not take the current architecture into consideration • all clean, all fresh Research Study Report

5. Names, IDs, Locators Name: John Smith ID: 012-34-5678 Locator: 1234 Main Street Big City, MO 12345 USA • Locator changes as you move, ID and Names remain the same. • Examples: • Names: Company names, DNS names (Microsoft.com) • IDs: Cell phone numbers, 800-numbers, Ethernet addresses, • Skype ID, VOIP Phone number • Locators: Wired phone numbers, IP addresses Research Study Report

6. Internet 3.0(GINA) Internet 3.0 is the name of the Washington University project on the next generation Internet Goal 1: Represent the commercial reality of distributed Internet ownership and organization Goal 2: Develop a clean slate architecture to overcome limitations of the current internet Goal 3: Develop an incremental approach to implement the architecture Research Study Report

7. Objects in GINA You can connect to a human, organization, or a department • Object = Addressable Entity • Current: End-Systems and Intermediate Systems • GINA: • Computers, Routers/Firewalls…. • Networks • Humans • Companies, Departments, Cities, States, Countries, Power grids • Process in a computer • Recursive ⇒ Set of Objects is also one object, • e.g., Networks of Networks Research Study Report

8. Realms in GINA Realm = Administrative Group • Object names and Ids are defined within a realm • A realm is a logical grouping of objects under an administrative domain • The Administrative domain may be based on Trust Relationships • A realm represents an organization • Realm managers set policies for communications • Realm members can share services • Objects are generally members of multiple realms • Realm Boundaries: • Organizational, Governmental, ISP, P2P,… Research Study Report

9. GINA: Multi-Tier Object-Oriented View • Objects provide services. Higher tiers specify the requirements • Tier service broker (shown by dotted line) composes a service • can negotiate with multiple realms in that tier • Higher tier may not/need not find details of lower tiers Research Study Report

10. Content distribution mechanisms Next generation CDN Next generation P2P Swarming architecture Content Centric Networking Research Study Report

11. Content distribution mechanisms • Next generation CDN • Neither the server nor the clients have any control over the “middle mile” • Fat File Paradox: • it is the length of the pipe rather than its width that determines how fast a large file can travel through it • Highly distributed CDNs place servers at the edge networks • Security, management, scalability and synchronization problems. Research Study Report

12. Next generation P2P • The self-organizing and self-healing properties of P2P networks have the potential to become the predominant content distribution mechanism of the future Internet. • Bandwidth provision (Asymmetric) • Dynamic sharing • Tussle of interests between P2P networks and ISPs • Aware of the underlying topology and location of peers. Research Study Report

13. Swarming architecture • Uswarm • A “swarm” is 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 model Research Study Report

14. Content-Centric Networks • IP cares about “Where”: forward packets from A to B • Users care about “What”: Movie X • Replace “packets” with “Data Objects” or “Interests” (requests) • Replace “Addresses” with “Names of Objects” V. Jacobson, et al, “Networking Named Content,” to appear in CoNEXT 2009, December 2009 Research Study Report

15. Content-Centric Networks: Routing Policy • Content Store(CS): Local cache of data • Pending Interest Table (PIT): Recent requests forwarded • Forwarding Information Base (FIB): Known data locations • Faces: Requesting processes and hardware interfaces Research Study Report

16. Content-Centric Networks: Routers Operation • Applications send “Interest” in data X • Router looks up in local store and sends if found • Router looks up in PIT, if entry already exists • someone requested it recently • adds the interest • face to the same entry • Router looks up in FIB, if entry exists • data location is known • a PIT entry is made • the interest is multicasted to all faces in the FIB entry • the FIB entry is removed • If there is no FIB entry, interest is discarded • router does not know how to get the data • When data arrives, • Content Store match ⇒ duplicate, discard • PIT match ⇒ Forward to all faces • FIB match ⇒ No PIT ⇒ Unsolicited ⇒ Discard • Data providers register their data ⇒ Creates FIB entries Research Study Report

17. DONA :Data-Oriented Network Architecture DONA proposes a novel mechanism of explicitly naming the data or service and routing on these names for data or service access. DONA proposes replacing DNS names with flat, self-certifying names, and replacing DNS name resolution with a name-based anycast primitive that lives about the IP layer. Research Study Report

18. DONA: Basic Design • Persistence • The name of a service or data object remains valid as long as the service or data are available • Availability • Data or service should have a high degree of reliability and acceptable latency • Authenticity • Data can be verified to have come from a particular source T. 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. Research Study Report

19. DONA: Naming Mechanism • DONA names are organized around principals. • Each principal is associated with a public-private key pair • Each datum or service or any other named entity (host, domain, etc.) is associated with a principal • Names are of the form P:L • P is the cryptographic hash of the principal’s public key • L is a label chosen by the principal, which ensures that these names are unique • Granularity of naming is left up to principals • A principal might choose to just name her web site, or name her web site and each page within it, or name at a finer granularity Research Study Report

20. DONA: Name Resolution • Name Resolution’s goal is to achieve high availability, by finding close-by copies and avoiding failures. • Route-by-name paradigm • Infrastructure :Resolution handlers(RHs) • Two basic primitives operations in this design • FIND(P:L) and REGISTER(P:L) • FIND(P:L) locate the object named P:L • REGISTER messages set up the state necessary for the RHs to route FINDs effectively Research Study Report

21. Establishing REGISTER state • Any machine authorized to serve a datum or service with name P:L sends a REGISTER(P:L) command to its first-hop RH • RH maintains a registration table that maps a name to both a new-hop RH and the distance to the copy (in some metric) • REGISTERS are forwarded according to interdomain policies: • REGISTERs from customers (child) to both peers and providers • REGISTERs from peers optionally to providers/peers Research Study Report

22. Forwarding FIND(P:L) • When FIND(P:L) arrives to a RH: • If there is an entry in the registration table, the FIND is sent to the next-hop RH • If there is no entry, the RH forwards the FIND towards to its provider (parent) • In case of multiple equal choices, the RH uses its local policy to choose among them Research Study Report

23. DONA: Challenge Research Study Report

24. DONA: Using Functionalities • Using Functionalities • Using name-based anycast • Server Selection • Mobility and Multihoming • Multicast State Establishment • Improving content delivery • Caching • Subscriptions • Avoiding Misbehaving and Overloaded Servers Research Study Report

25. CCN VS. DONA Both DONA and NNC, advocate a paradigm shift from the present host centric architecture of the Internet to a data centric architecture. NNC 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. DONA on the other hand emphasizes a novel mechanism for the naming of content and name resolution to build an architecture around service and data access. Research Study Report

26. Challenged Network Environments • Delay Tolerant Network • Bundle protocol • Delay/fault tolerant mobile sensor networks (DFT-MSN) • Postcards from the edge • Disaster day after networks (DAN) • Selectively Connected Networking (SCN) Research Study Report

27. Future Internet Projects • In 2005 US National Science Foundation started a large research and infrastructure program on next generation Internet • Q: How would you design Internet today? Clean slate design. • “Future Internet Design” (FIND): 48+ projects • www.nets-find.net • Stanford, MIT, Berkeley, CMU, … • “An Architecture for Diversified Internet” at WUSTL • “Global Environment for Networking Innovations” (GENI):29+ projects • http://www.geni.net/ • European Union: 7th Framework program • http://cordis.europa.eu/fp7/ • Japan: AKARI (A small light in the dark pointing to the future) • http://akari-project.nict.go.jp/eng/index2.htm • China, Korea, Australia, …20+ countries Research Study Report

28. Next Generation Testbeds Past: PlanetLab, Emulab, VINI, OneLab Federation GENI, Requirements, Subsystems GENI Prototype Clusters Wireless Network Virtualization Supercharged PlanetLab Platform (SPP) FIRE, FEDERICA AKARI Research Study Report

29. PlanetLab • Global networking research testbed • 1055 nodes at 490 sites [Nov 2009] • Researchers use it experiment with new ideas on distributed storage, network mapping, peer-to-peer systems, distributed hash tables, and query processing http://www.planet-lab.org/ Research Study Report

30. PlanetLab (Cont.) • Linux virtual server software on Interned nodes • Slivers = Piece of a resource • Node manager (NM) manages the node's virtual servers • Planet Lab Control (PLC) interacts with NM • Experimenters request a "Slice" = slivers in various sites Research Study Report

31. Emulab • Networking research testbed at University of Utah • Available for public use for research and education • Software implemented at two dozen sites around the world • Allows simulated links and nodes in slices • ⇒Allows fault studies • Provides repeatability http://www.emulab.net/ Research Study Report

32. Internet for Masses Make it easy for naive users Function in infrastructure poor environments Intermittent power Device sharing Allow sneaker nets Long distance wireless CDMA450 Research Study Report

33. OneLab • Federation of European PlanetLabs • Enhanced monitoring infrastructure • Wireless Testbeds • IPv6 multihoming • Emulation tools • Deal with unstable connectivity http://www.onelab.eu/ / Research Study Report

34. Federation • Larger testbeds • Testbeds for specialized resources such as access technologies • Specialized research communities and cross-discipline • Challenges: • Homogenization of diverse context • Interoperability of security protocols • Political or social-economic issues • Intellectual Property rights • Commercial and non-commercial interests Research Study Report

35. GENI http://www.geni.net/ Global Environment for Network Innovations Dedicated shared substrate facility for large-scale experiments US National Science Foundation project Dedicated backbone links through LambdaRail and Internet2 Diverse and extensible set of technologies Research Study Report

36. GENI Requirments Sliceability: Sharing with isolation. Programmability: All components should be programmable Virtualization: Slicing via virtualization or space/time sharing. Federation: Combination of independently owned testbeds Observability: Allow specifiable measurement framework Security: Should not harm production Internet Research Study Report

37. GENI Subsystems Research Study Report

38. GENI Prototype Clusters GENI Spiral 1, http://groups.geni.net/geni/wiki/SpiralOne • Five Clusters in Spiral 1: • 1.Trial Integration Environment with DETER (TIAD): • Emulab based security experiments testbed • 2.PlanetLab: • Federate all slice-based substrates PlanetLab, Emulab, VINI, and GENI • 3.ProtoGENI: • Federation of Emulab testbeds, • Enhanced Emulab Control • 4. Open Resource Control Architecture (ORCA): • Resource manager runs under the host operating system • Uses virtualization to allocate containers • 5. Open Access Research Testbed (ORBIT): • Wireless testbed with emulated and real nodes Research Study Report

39. Top 10 Features of Next Generation Internet Security Mobility User/Data-Centric: Network support of data objects Easy to use: Self-organizing, better user control Disruption Tolerant Green: Proxy, Sleep Modes,(Green Network) Services: Storage, Translation, Monitoring Organizational Representation Virtualizable to create Application Specific Context Policy Enforcement Research Study Report

40. Summary NSF FIND program has funded a number of architectural component research programs. GENI testbed consists of 5 clusters in Spiral 1 FIRE in Europe and AKARI in Japan are similar to GENI. Internet 3.0 is an industry sponsored full architecture program Research Study Report

41. Other References • Internet 3.0 Talk by Raj Jain • http://research.microsoft.com/apps/video/default.aspx?id=104012 • R. Jain, ``Internet 3.0: Ten Problems with Current Internet Architecture and Solutions for the Next Generation," Proceedings of Military Communications Conference (MILCOM 2006), Washington, DC, October 23-25, 2006 • V. Jacobson, et al, “Networking Named Content,” to appear in CoNEXT 2009, December 2009 • T. 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. Research Study Report

42. Future Testbeds • OneLab: http://www.onelab.eu/ • User cases :http://www.onelab.eu/index.php/services/testbed-access/use-cases.html • GENI • Fresh user: http://groups.geni.net/geni/wiki/NewExpermenterQandA • GENI Experiments : http://groups.geni.net/geni/wiki/GeniExperiments • Experiments Archive :http://groups.geni.net/geni/wiki/ExperimentationEventArchive Research Study Report

43. ThanksQ&A 2011-8-4