toward new generation network - beyond the internet and...

OFC NFOEC 2009

Toward New Generation Network- Beyond the Internet and Next

Generation Network (NGN) -

March 24, 2009

Tomonori Aoyama

Keio University/NICT

Content

- The Internet, Next Generation Network and New Generation Network

- Requirement for New Generation Network

- Some important research items for New Generation Network

- Some future applications over New Generation Network

2

20202008

An image of a network evolution

Legacy Telephone NW

Cellular Phone NW

The Internet

Started NXGN(NGN)

Future NWGN

NXGN: Next Generation Network (NGN)NWGN: New Generation Network

IP-based NW

Post-IP NW

3

NXGN and NWGN(Next) (New)

NXGN (NeXt Generation Network): NGN

Replacement of legacy telephone

networks using IP-based networks to provide

triple-play/quadruple-play services

NWGN NeW Generation Network)

Clean-slate designed network architectures and main protocols different from IP-based

networks , which can be post-Internet/NGN

Industry is investing their resources to NXGN deployment.

NICT contributes to promote R&D on NWGN. 4

Can IP Protocol handle the explosion of information in 2020s ?

ApplicationsWebEmailVideo/Audio

TCP/IPAccess Networks

Ethernet (LAN)Wireless (LMDS, WLAN, Cellular)CableADSLSatellite

Applications

AccessNetworks

5

© National Institute of Information and Communications Technology

Internet Too Much Complicated

Cannot add new functions Cannot provide services for future society

Original Internet Architecture

mobility

anycast

flow-label

guaranteedservice?

universal communication?

small devices? authentication?

L2: Datalink Layer

L3: Internet Layer

L4: Transport Layer

Entrust network with your life & living ? (tele-medicine, ITS & anticrime, finance)Rich life? (connecting sensor, RFID)Safe? Secure? (spam, DDoS)Never broken? how long? (sustainable society)Flexible to future change? (nobody knows future)

Adding Functions

Individual optimum but NOT global optimumDesign from scratch has come!

Increasing Layers

dependability?

6/38

Architecture is different from that of the Internetalthough both are based on the IP protocol

Application

APINNI

Transport Stratum

Service StratumService Control Function

Transport Function

Other N

etwork

End U

ser System

UNI

-to-7

Aoyama 6

Best effort bearer function to interconnect multiple router based network

Terminal Terminal Server Terminal Terminal Server

IP router networkBest effort



Internet

NGNVersatile

Bearer functions

NGNVersatile

Bearer functions

NGNVersatile bearer functions

QoS controlled bearer function to interconnect multiple networks with clear responsibility

NGN

A comparison between The Internet and NGN

No over all network planningTCP/IP Protocol is the only common ruleBest effort based network and no clear responsibility and control rule exist among networks

User can have freedom to install applications

IP based network with network control function and with clear responsibility for the control

Qos control and security functions are installedMaintain the Internet connection function



NGNVersatile bearer function

8

R&D Marathon to realize a new communication paradigm has just started globally.

Who will win ?

Now !

9/38

Research on New Generation Networks started globally

To realize NWGN, tight collaboration among the E.C., U.S., China, korea and Japan is important.

the future Internet architecture through a clean slate approach. make a small scale, but a large number of projects converge into a fewer number of full scale architectures, and to verify using GENI A total of 1,200 million dollars to 26 projects by 2006

GENI

development. programmable

International cooperation is also included as a target.Project Bureau (GPO) is BBN. Planning to offer prototype, 1,500 million dollars in December 2007 over a two year period. A scale of construction budget is 367 million dollars,

total fields of science and technology

projects in Call 1 Call 2 is investigating

E.C.FP7FIND

main individual programs-Network of the Future

(Call 1, D bureau ) 2 billionEuros)

-FIRE(Call2, F bureau ) 4,000 million Euros

NWGN AKARI architecture development project

specification in April 2007)R&D on evaluation and establishment in

the structure of NWGN) (NICT commitment, up to 20 million Yen times 2 years plural cases)

R&D for dynamic network technology (NICT commitment, 4 years plural cases)

NWGN Research projects

JGN2plus Testbed

development2008

2010.

The E.C. Budget for GÉANT was 9,300 million Euros over 4 years since 2004. It is in the middle of preparations for upgrading and for greater capacity by transforming it into GEANT3 from 2008.

GÉANT3

The United States

NWGN Promotion Forum

November, 2007 Working in 4 WGsMinistry of International Affairs and Communications and NICT are a co-secretariat.

Japan

10/38

of communication networks !

============Telephone Network

Digitalization ISDN B-ISDN NGN over IP

IP Network TestbedARPANET / NSFNET

Commercial Internet1969

2008

R&D for beyond IP started worldwide

2006/2007

GENI/FIND, NWGN, FP7

Commercial beyond IPnetworks will start

It is time to start the second ARPANET/NSFNETtype academic networks globally for establishing beyond the Internet/NGN based on IP.

1990

1995

2020

11

World market shares for main ICT systems (2005)

Router

Server

Cellular Phone

Desk-top PC

Note PC

LSI

LCD

Japan US Korea China

Legacy TV setDigital Camera

12

3 Approaches are required to make a clean-slate designed architecture

Top-down ApproachSocial requirements and requirements for long tail applicationsIdeas to meet those requirements

Bottom-up ApproachTechnical breakthroughWhat can be realized by this breakthrough ?

New generation network architecture Design Principle

13


New Generation Social & Design Requirements in 2015~2020

Peta-bps backbone, 10G-bps FTTH100-billion devices, M2M, 1-Mega stationsCompetitive industry and user-oriented servicesMedical care, traffic control, emergency, four-ninePrivacy, financing, food tracking, anti-disasterRich society, handicapped, aged supportEarth & human monitoringBroadcasting & communication, web 2.0Economic Incentive (Business-cost model)Ecology, sustainable society Human possibility, universal communication

Design RequirementsSocial RequirementsCapacityQuantityOpennessRobustnessSafety & SecurityDiversity (Long-tail) Ubiquity (Pervasive)Converge & SimplifyNetwork ModelEnergy-savingEvolvability

New Generation Network Designed by the Architecture

14

K M G

Access frequency [page/day]

K

M

G

T

P

Yahoo300Mpage/day

InternetTV11Mpage/day

Web content

Cap

acity

of

cont

ent

SDTVDVD>GB

Web 10kB/page

MP3>MB/music

DigitalCinema> 100GB

e Commerce

IP TV

HDTV

Cine-grid

B2C

B2B

[bit]

Contents in the ubiquitous societyFrom tiny to huge

Sensor & RF ID

S2M

P2P

Both directionsTiled Display

15

AKARI Architecture Design Project Toward NewGeneration Network

Designing the future, diverse, new generation network beyond 2020- pick up techniques for the future under the principles- integrate & simplify them with design methods- Clean-slate design approach

http://akari-project.nict.go.jp/http://www.akari-project.jp/

AKARI a small light in the dark pointing to the future 16


AKARI Sustainable Architecture Principles

ID-Locator separationBi-directional authenticationTraceability

Self-* properties (emergent)Autonomic distributed controlScalableSocial Selection

KISS (Keep It Simple, Smart)Crystal synthesis Common layer (layer degeneracy)End-to-end (original Internet)

Sustainable EvolutionReliable Network Space

Diversity Inclusion

17

Overlay Network

Underlay Network

Cross-layer Control Mechanism

Application

Study Items for NWGN Architecture

Photonic NWMobile NW

Sensor NW

All the elem

ents should be redesigned.

18

Fundamental Principles

End-to-end argumentOpen, Transparent vs SecurityHandling data format

Datagram PacketFlowPath/CircuitMessage

Layering StructureIdentifier / Locator SeparationIn band vs Out of band

19/38

AKARI Architecture Focus (I)(1) AKARI Optical Switching - Path / Packet Integration -

QoS Guaranteed

Keep Both QoS Assuredness & Best Effort Efficiency=> OPS + OCS (Not OPS over OCS)

A,B,C,D E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T

Wavelength

Pow

er S

pect

rum

Header Payload

Sustainable in Capacity

Optical Circuit Switch

Best Effort

Optical WaveSharing Optical Packet Switch

(with Fiber Delay Line)

Parallel Optical Packet Transmission

- Scalable- Energy-Saving

Sustainable in Usage - Allow Usage Diversity

Requirements:- Handle different types of traffic- Peta-bps switching capacity- Tera-bps link speed (40G x 100)- 100 billion tiny terminals

20

AKARI Architecture Focus IISeparation of Identifier & Locator

identifier

locator

identifier

identifier

locator

locator

locator

locator

Applicationlayer

Transportlayer

Networklayer

Identification sublayer

Delivery sublayer

Mapping

Destination Node is defined by Identifier

Destination Node is defined by Identifier

Destination Node is defined

by Locator

Bypass

ConventionalID/Locator Separation

Destination Node is defined by Locator

21

AKARI Architecture Focus (III)

Self-organizing Control (Concept)

nageableFast Recovery

ultaneousSim sSimultaneousSimultaneousfailuresfailure

ChainCreaction

Softwaree

Bad provisioning

Optimum

Adapted

Robust

Efficient

Base StationsBase Stations

User a, b, c

#1 #2#3 #N

WideSharedChannel

User a, b, c

- PDMA (Packet Division Multiple Access)

Free from:-Frequency Channel Allocations-Cell Design

Sustainable in Management and Capacity

StaticChannelAllocation

CSMA/CA(packet)

- Self-Organizing Mesh Network

New Network Science RequiredScale-free NetworkBio-inspired Network 22

AKARI Architecture Focus (IV)

(4) Network Virtualization (Concept)

Physical Network

Real TestBed Network

VN1 VN2

(a) Isolated Virtual Networks

Real Operational Network

VN1 VN2

(b) Transitive Virtual NetworksSelf-evolvable

Real Operational Network

VN1

VN2

(C) Overlaid Virtual Networks

Optical Path NetworkWireless Network

23


Various Virtualization Techniques

Computational VirtualizationHypervisor-Based, Host-Based, OS-Level VirtualizationHardware-support (Intel VTx, AMD-V)

Network VirtualizationI/O Virtualization (Intel VTd , AMD IOMMU) Resource ContainerNetwork Resource Allocation Management

Hardware

HostOS

Apps VM VM

GuestOSGuestOS

Apps Apps

Hardware

VMM

VM VM

GuestOSGuestOS

Apps apps

VM

MngOS

Manager

Hardware

HostOS

Apps VE VE

Apps Apps

Hypervisor-Based Host-Based OS-Level

24/38


Overlay to Network Virtualization

VMM/Hypervisor

Mgm

t S/

WVM

(Slic

e)VM

(Slic

e)

VM(S

lice)

A Slice of Computation Resources

PC Server

switch

25/38

1999

(year)

2003 2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

JGN Project Mile Stone

(19999.49 -2004.3)0 ((2004.4( -2008.33)3

JGNN2pllusN lTo New Testbed

NICT2nd Mid-term Plan

(2006.4-2011.3)

CRL (NICT)1st Mid-term Plan(2001.4-2006.3)

2008.4-2011.3 2011.4-

NICT3rd Mid-term Plan

2011.4-

26

Network Research and Operation [Predicted Research Projects]

Network Operation TechnologyTraffic Measurement TechnologyNew Generation Network ArchitectureGrid NetworkOverlay NetworkAdvanced InternetVerification of Interconnection

CSTNET(China)

SingAREN(SingAREN)

JGN2 plus

Tokyo

CA*net4(Canada) SURFnet

(Netherlands)

Internet2(USA)

NLR(USA)

MREN(USA)

AARNet(Australia)

IEEAF(USA)

UKLight(UK)

GLORIAD(USA, Russia, China)

ThaiSarn(Thailand)

TWAREN(Taiwan)

HARNET(Hong Kong)

StarLight (USA)

UniNet(Thailand)

KOREN(Korea)

TransPAC2(USA)

LA

Chicago

HK

BKK

SG

CERNET(China)

GEANT2(Europe)

TEIN2(Asia, Europe)

APAN(Asia)

KR

Pacific Wave(USA)

27

Current Photonic Network R&D supported by NICT

All-optical transport1996 2005

Photonic node enabling broadband access2000 2005

Optical burst switching network2001 2005

Control plane for terabit-class network2001 2005

PHASE I

Photonic node with multi-granularity

switching capability 2005 2009

Access 2006 2010

Utility 2005 2009

Photonic RAM 2005 2009

PHASE II

NICT R & DNICT R & D29

NICT funding projects for photonic NW R&D

Photonic node with multiple granularity switching capability2005 2009

l Utility2006 2010

ll Access2006 2010

Optical RAM for all-optical packet switching2006 2010

Note: ll refers to a large-capacity and transparent information channel, identified by the wavelength value

Next phase program is discussed in PIF.

PIF: PhotonicInternet Forum

All-optical terabit-class LANAccess-to-backbone seamless networking

Establishing customer-initiated end-to-end 100Gbps optical path and QoS service provisioning between terabit-class LANs

Photonic crystal bit memory

100Tbps-Class photonic node with multiple granularity switching

30/38

Proposal for Phase III photonic Network R&D

PIF (Photonic Internet Forum ) is making the proposal for MIC

Green ICT

- Self-controllable power-minimum photonic network

100 times higher speed with the same power consumption as today

Application-oriented photonic network Network Virtualization

- Optimal resource allocation of photonic networkswith multiple QoS adapting application requirements

31/38


NWGN related R&D $ 21M

Photonic Network Projects ( 7 Groups ) $ 22 M

Security Technology Projects $ 11 M

Wireless Network Project ( 1 Group ) $ 3 M

Total about $ 60 M

NICT Funding Projects in 2008 Fiscal Year

$ 1 = ¥100

32/38


NICT Research Projects and NICT Joint Research Projects for NWGN

NWGN related R&D $ 30M

Network Testbeds $ 40M

MIC Funding Projects related to New Generation NetworkUbiquitous Network Platform ( 1 Group ) $ 15M( Ubiquitous Networked Robot Projects Proposing for 2009 Budget )

33/38

Long Tail Application

Transfer Rate100 Mb/s

100 Gb/s

Enterprise Users Researchers & Scientists

Long Tail Application

No of Users

Con

sum

er Users

Innovations have been derived from Long Tail Applications.

e.g. Internet , WWW,

34/38

Examples of long tail applications over NWGN

- Cloud Computing over photonic networks

- Connection of big tailed display for visualization of

e-science data

- Distribution of digital entertainment contents with

ultra high quality such as digital cinema & ODS*

ODS* : Other Digital Stuff

musical, opera, drama(Kabuki), sports, etc.

35/38

Large Scale Tiled Display System developed by UC San Diego

36/38

2K vs 4K: Beyond HDTV for Digital Cinema K 1K=1024 Pixel

2K=20484K=4096

=1920 1080

TV

DDiiggiittaall TTVV

Digital HDTV

DDiiggiittaall CCiinneemmaa

=2048 1080

4096 2160 4K

30 frames/s Interlace 24 frames/s Progressive

Digital Cinema standard was made by DCI and SMPTE is proceeding with the detailed specifications

37/38

Japan USA

LA

SeattleTokyo

YokosukaOsaka

Burbank

Production StudioCinema Theater

Local Distribution Center

Global Distribution

Center

CompressionEncryptionKey Center

1Gb/s IP

Link

Source: NTT

4K Pure Cinema Trial by Warner Brothers, Toho & NTT

38/38


-Theater System

In the projection room of VIRGIN TOHO CINEMAS

4K Projector

SecureMedia Box

(SMB)

TheaterControl Box

(TCB)

Corpse BrideHarry Potter 4V for VendettaDaVinci Code

PoseidenMission Impossible 3

+Tokyo Film Festival

Batman BeginsStealth

Source: NTT39/38

Requirement for Networking

2-hour 4K digital cinema contents Non-compressed : 5TBJPEG2000 Compressed : 250GB (1/20 compressed ratio)

File Transfer10 hours over 1Gb/s link Non-compressed30 minutes over 1Gb/s link 1/20 compressed

Real time streaming6Gb/s Non-compressed300Mb/s 1/20 compressed

Multicast functionDistribution of contents to theaters, halls and home theaters

QoS RequirementsPacket LossLatencySynchronization

40/38


Olympus 4K Digital Camera

42/38

Image of digital cinema and ODS distribution over a network

Projector

City hall, hospital,

Cinema Theater

Cinema Data Center School

Education

Live of Musical

Live of sports

Broadband Network

Home

43/38

Time

3 Axis for Digital Content Evolution

Content SearchContent Distribution

Content Creation Archive

Applications e-Commerce, e-Science, e-Learning, e-Medicine, Digital Library/ Musium, Digital Entertainment, Web2.0, Web3.0,

Volume

QualityFederated Structure

44/38

AMPAS ( Academy of Motion Pictures Arts and Sciences) published a report

strategic issues in archiving and accessing digital motion picture materials.

45/38

Worldwide Collaboration is indispensable

Global Collaboration from Japan point of viewsUS-Japan Joint Workshop on NWGN/FI

October31-Novbember 1,2008 at Palo Alto in 2009EU-Japan Symposium on Future Network

June 9-10, 2008 at BrusselsJuly 6-7, 2009 in Japan (Plan)

China-Japan ICT Forum (China Academy of Science NICT)July 8-9, 2008 in Shanghai

StandardizationITU-T established a focus group (FG) in Study Group 13 to discuss about future networks

ITU-T hosted the Kaleidoscope Academic Conference jointly with IEEE ComSoc.

46/38


US-Japan Joint Workshop on New Generation Network/ Future Internet

October 31-November 1, 2008 at Palo Alto

47/38


ConclusionThe Internet is expanding rapidly and creating new applicationssuch as Web2.0.NGN started in 2008 in Japan, and industry is concentrated their resources on the deployment of NGN.R&D on NWGN has started globally toward a new paradigm of communications. - Network architecture should be studied based on

requirements for ubiquitous networking and new networking technologies such as advanced Photonic networking technologies.

- Prototype of NWGN and the standard will be realized in 2015-2020 time frame.

- Network Testbeds are essential for R&D on NWGN. - Global collaboration among universities, industries and

governments is important to accelerate R&D. 48/38

Thank you for your attention !

49/38

toward new generation network - beyond the internet and...

Documents