sep 6th, rio de janeirointernet: as of tomorrow phonephone econo financ econo financ educeduc...
TRANSCRIPT
Sep 6th, Rio De JaneiroSep 6th, Rio De Janeiro
1750 1800 1850 1900 1950 2000 2050
1876: Alexander G. BELL invents the PHONE1838: Samuel MORSE invents the TELEGRAPH
• ”The telephone would be used only to inform people of arrival of telegrams."
-- 15 % Penetration-- 2 Days Walk for 1 Billion people
New York 1891,New York 1891,Broadway/John Street
Analog Communication
PhonePhoneEconoFinancEconoFinanc
EducEducTrans-Port-ation
Trans-Port-ation
Agri-cultureAgri-
cultureTV
RadioTV
Radio
PhoneInfra
PhoneInfra
BCASTInfra
BCASTInfra
Internet: as of Yesterday
PhonePhone EconoFinancEconoFinanc
EducEducTrans-
portationTrans-
portationAgri-
cultureAgri-
cultureTV
RadioTV
Radio
Phone Infra
InternetBCAST
InfraBCAST
Infra
Internet: as of Today
PhonePhoneEconoFinancEconoFinanc
EducEducTrans-
portationTrans-
portationAgri-
cultureAgri-
cultureTV
RadioTV
Radio
Phone Infra
InternetBCAST
InfraBCAST
InfraInternet Infra
(fiber, cable &wireless)
Internet: as of Tomorrow
PhonePhoneEconoFinancEconoFinanc
EducEducTrans-
PortationTrans-
PortationAgri-
cultureAgri-
cultureTV
RadioTV
Radio
Internet
Internet Infrastructure(fiber, cable, and wireless)Internet Infrastructure
(fiber, cable, and wireless)
Phone Infra
BCA
ST Infra
Internet Infrastructurefor digital communication society
PhonePhone EconoFinancEconoFinanc
EducEducTrans-
portationTrans-
portationAgri-
cultureAgri-
cultureTV
RadioTV
Radio
InternetInternet(Integrated) Internet Infrastructure(Integrated) Internet Infrastructure
A testbed is VERY important!
IPv4: Not enough IP addresses and growing requirements
CableCable
IPv4IPv4
Tans-portationTans-
portationAgri-
cultureAgri-
cultureEcono
FinanceEcono
FinanceEducEducPhonePhone
TVRadioTV
Radio
OpticalFiber
OpticalFiber
WirelessWireless SatelliteSatellite
IPv6: The Internet Infrastructure
CableCableOpticalFiber
OpticalFiber
WirelessWireless SatelliteSatellite
IPv6
Trans-portationTrans-
portationEcono
FinanceEcono
Finance
TVRadioTV
RadioPhonePhone EducEduc Agri-
cultureAgri-
culture
Perc
enta
ge o
f Ow
ners
hip
Perc
enta
ge o
f Ow
ners
hip
100100
8080
6060
4040
2020
00
ElectricitElectricityy
(1873)(1873)TelephonTelephon
ee(1876)(1876)
AutomobileAutomobile(1886)(1886)
TelevisioTelevisionn
(1926)(1926)RadioRadio(1905)(1905)
VCRVCR(1952)(1952)
MicrowaveMicrowave(1953)(1953)
PCPC(1975)(1975)
InternetInternet(1975)(1975)
1001008080606040402020 120120
Years Since IntroductionYears Since Introduction
ARPANET DEMONSTRATIONJohannesburg, South Africa, 1974
X
under
1 G 2 G 3 G 10 * 4 10 * 8 10 * 10 Professionals Innovators Everyone Email, Ftp WWW Wireless, Streaming Media 9.6 k 56 k 1 + MGovernment Internet Public Internet Global Internet
1 G 2 G 3 G 10 * 4 10 * 8 10 * 10 Professionals Innovators Everyone Email, Ftp WWW Wireless, Streaming Media 9.6 k 56 k 1 + MGovernment Internet Public Internet Global Internet
1 G 2 G 3 G 10 * 4 10 * 9 10 * 38 Professionals Innovators Every-One & -Thing Email, Ftp WWW Wireless, Streaming Media 9.6 k 56 k 1 + MGovernment Internet Public Internet Global InternetDIALDIAL--UP INTERNETUP INTERNET ALWAYSALWAYS--ONON
• NCP Conception
• NCP Roll-Out
1961 19691961 19691969 19821969 19821972 19821972 1982
1983 Now!1983 Now!
1991 1995-Now1991 1995-Now
IP Conception
IPv4 Roll-Out
IPv6 ConceptionIPv6 Roll-Out2000 2000
IPv6 Technical DirectorateIPv6 Technical DirectorateChair: Jim Bound
Steve DeeringBrian CarpenterScott Bradner Christian Huitema Allison Mankin
Charlie PerkinsMatt Crawford Thomas Narten Erik Nordmark Thomas Ecklund
IPv6 Technical DirectorateIPv6 Technical Directorate
Brian Zill
Carl WilliamsStig Venas
Francis Dupont Alain Durand Jack McCann Henk Steenman
Cyndi Jung
Peter Tattam
Brian Haberman
Ivano GuardiniDale Finkelson
Yanick Pouffary
Stephen Hayes
Internet User Trends
0500
10001500200025003000350040004500
1995
1997
1999
2001
2003
2005
2007
2009
Year
User
s (M
illion
s)
Source: Nua Internet Surveys + vgc projections
PersonPerson
PersonPerson
ApplianceAppliance
ApplianceAppliance
VoiceEmail
VoiceMail
Telemetry
IPsensors
Alarm
NotificationsEmerging
Growing
Growing
MaturingVoice Control
Fax Computer TeIephony Integration
Home Banking
InfoCom Application AreasInfoCom Application Areas
UMTS/IMT-2000MBS BB-WLANs
1972 1992 2002 2012
MobileMultimedia
Voice + Data
Voice
1G
2G
3G
Info
rmat
ion
bit r
ate
GSM 900/1800DECT
G - Evolution in continuity
IntegratedWireless World
©JPER
?
4G
2022
Evolution from 2G to 3G and BeyondEvolution from 2G to 3G and BeyondEvolution from 2G to 3G and BeyondCircuit
switchedCircuit
switched CS voice / packet dataCS voice / packet data IP coreIP core
GSM
UMTS
UWC-136
CDMA
2GHSCSD
2GHSCSD
end to end IPend to end IP
GPRSEDGEGPRSEDGE
WLAN
HiperLAN/802.11a
IMT-2000 CDMAMulti-Carrier
Optimally Connected Anywhere, Anytime
MBS at 60 GHz?MBS at 60 GHz?
Ad hoc NetworksAd hoc Networks
New RadioInterface?
New RadioInterface?Bluetooth
2G 3G and beyond9.6-14.4 kbps
evolved 2G
64-144 kbps 384 kbps-2 Mbps 100 Mbps?384 kbps-20 Mbps
3GPP Rel 0xIMT-2000 CDMA
Direct Spread + TDD
3GPP Rel 0xIMT-2000 CDMA
Direct Spread + TDD3GPP Rel 993GPP Rel 99
D-AMPSD-AMPS
IS 95IS 95
GPRS/EDGEIMT-2000 TDMA
Single-Carrier
GPRSEDGEGPRSEDGE
CDMA2000
CDMA2000
Includes VerticalHandover between
Technologies
WirelessWirelessIPIP
SocietySocietyS-UMTS
SatelliteBroadband
DVB-SDVB-T
DAB
GSMGPRS/EDGE
DECTIR
BroadbandW-LAN
UMTS
Satellite/HAPS
Broadcasting
Cellular
Indoor
MBS 40 xMDS
Broadband WFA
Wireless Local Loop
Body LANs
PersonalArea Networks
UMTS ++
4th Generation
Local Area NetworksMBS 60 MWS
Bluetooth
Quasi-Cellular
A Galaxy of heterogeneous networksA Galaxy of heterogeneous networks
Telecom IndustryTelecom Industry
main frames
desk top computing
PC-LANPC-WAN
InternetInternetInternetInternet
electronicpublishing and entertainment
Computer IndustryComputer Industry
Media IndustryMedia Industry
The ConvergedIndustry
The ConvergedIndustry
Wireline
mobilitymobile/mobile
PSTN
ISDN
PC/Servers
NewTelecomsIndustry
NewTelecomsIndustry
3G/mobile Internet
Carrier class
MMM
WWWWWWWWWWWW
WAP
Next
battle ground
Vision: World at your fingertipsVision: World at your fingertips
ADSL, Cable TV,... Service
Gateway
What is the Key Enabler for Mass Deployment
of Mobile IP-Services ??
!! Cheap Cheap Chea
Cheap Cheap Cheap Cheap Cheap Cheap Cheap
It must be
2000 2001 2002 2003 2004 2005 (1 Billion Users Forecast)
NCP Pv4 NAT IPv6NCP Pv4 NAT NAT IPv6IPv61970 1980 2001
IPv4 is in the same state as DOS/Windows 3.1!1990
- ?
Always-On IP Addresses
Global Address Allocation
RIPE NCC
ARIN
APNIC
5%2%
4%
53%
36%
Unallocated
PRE-RIR Historical Allocations
ARIN APNIC
RIPE NCC
Global IPv6 Distribution
JPUS
KR
DE
SEUK EU FR PL RUTW
NL
Other:
AU, AT, BE, CA, CN, DK, FR, GR, HK, IT, MX, PT, SG, CH
Million Year 2005 Year 2010Mobile Phones 1500 3000Mobile IP Phones 500 10001% Roaming 5 10400 Wireless Networks 2000 4000
0
10000
20000
30000
40000
50000
60000
70000
80000
90000
Jun
-88
Dec
-88
Jun
-89
Dec
-89
Jun
-90
Dec
-90
Jun
-91
Dec
-91
Jun
-92
Dec
-92
Jun
-93
Dec
-93
Jun
-94
Dec
-94
Jun
-95
Dec
-95
Jun
-96
Dec
-96
Jun
-97
Dec
-97
Jun
-98
Dec
-98
Jun
-99
Dec
-99
Growth in BGP Route Table
Projected routing table growth without CIDR/NAT
Deployment Period of CIDR
Moore’s Law and NATsmake routing work today
Source: http//www.telstra.net/ops/bgptable.html
But they cannot berelied on forever
120,000
Scenario 1 Scenario 2 Scenario 3
IPv6 Deployment
Successful Complete Failure Partial
Sub-secnario 1 Sub-scenario 2
ExhaustionNAT-o-NAT
IPsec Works e-2-e Limited Broken Partial
Fog Noticeble Noticeble Clears! Fog Fog
NATs between even ISPs
Address Transparency
Restored e-2e Recycling addresses
Partial
Similar to v4-2-v6 Transition
FOG Permanet Thick Fog
Issues Intranet, Proxies & Firewalls may remain
Generalised use of NAPT, RSIP?
Scenario 1 Scenario 2 Scenario 3
IPv6 Deployment
Successful Complete Failure Partial
Sub-secnario 1 Sub-scenario 2
ExhaustionNAT-o-NAT
IPsec Works e-2-e Limited Broken Partial
Fog Noticeble Noticeble Clears! Fog Fog
NATs between even ISPs
Address Transparency
Restored e-2e Recycling addresses
Partial
Similar to v4-2-v6 Transition
FOG Permanet Thick Fog
Issues Intranet, Proxies & Firewalls may remain
Generalised use of NAPT, RSIP?
Scenario 1 Scenario 2 Scenario 3
IPv6 Deployment
Successful Complete Failure Partial
Sub-scenario 1 Sub-scenario 2
Exhaustion NAT-o-NAT
IPsec Works e-2-e Limited Broken Partial
Fog Noticeble Noticeble Clears! Fog Fog
NATs between even ISPs
Address Transparency
Restored e-2e Recycling IP Addresses
Partial
Similar to v4-2-v6 Transition
FOG Permanet Thick Fog
Issues Intranet, Proxies & Firewalls may remain
Generalised use of NAPT, RSIP?
Scenario 1 Scenario 2 Scenario 3
IPv6 Deployment
Successful Complete Failure Partial
Sub-secnario 1 Sub-scenario 2
ExhaustionNAT-o-NAT
IPsec Works e-2-e Limited Broken Partial
Fog Noticeble Noticeble Clears! Fog Fog
NATs between even ISPs
Address Transparency
Restored e-2e Recycling addresses
Partial
Similar to v4-2-v6 Transition
FOG Permanet Thick Fog
Issues Intranet, Proxies & Firewalls may remain
Generalised use of NAPT, RSIP?
Scenario 1 Scenario 2 Scenario 3
IPv6 Deployment
Successful Complete Failure Partial
Sub-secnario 2.1 Sub-scenario 2.2
ExhaustionNAT-o-NAT
IPsec Works e-2-e Limited Broken Partial
Fog Noticeble Noticeble Clears! Fog Fog
NATs between even ISPs
Similar to v4-2-v6 Transition
FOG Permanet Thick Fog
Issues Intranet, Proxies & Firewalls may remain
Generalised use of NAPT, RSIP?
Address Transparency
Restored e-2e Recycling addresses
Partial
Net Net HeadsHeadsThe people who get it will run it
The Culture Shock !
Bell Bell HeadsHeadsThe people who know how to operate massive
mission - critical networks will control the Net
MeltingMelting PotPotEstablished carriers will hire Internet geeks and buy ISPs, andNG service providers will bring in experienced pros
Bell Bell HeadsHeads! Central control is a must.
! The only way to scale the infrastructure is with a
hierarchical, unified system architecture.
! ! ! ! Web-centric computing is an extension of the mainframe era.
! ‘Bet you `re glad you don`t have to reboot your phone when
you need to dial 911!
! Contradiction: Deregulation and technical innovation have
made it impossible to maintain control of ifThe System.lT
NetNet HeadsHeads
! ! ! ! #### Chaos is good for creativity.
! The only way to scale the Net is with decentralized architecture
and control.
! ! ! ! Peer-to-peer networking is an extension of the PC revolution.
! ! ! ! #### Information (software, music, etc.) wants to be free. »
! Contradiction: Now that the 25-year adolescence of the Net is
over, and businesses want to use it for e-commerce, the Net
must grow up or get adult supervision.
IPv6 : An e-Buisness Enabler
Critical Success Factor Today with IPv4 IPv6
Cost Effectiveness Costly workarounds >1 billion addresses / person
Flexibility Frequent renumbering Simplified network planningas site grows and management
Reliability Operational complexity Return to simple andscalable architecture
Availability Single points of failure 24x7 operation
Scalability Client/server Peer-to-peer
Accessibility Obstacles to deploying Pervasive enablernext generation Simplified applicationapplications (e.g., VoIP ) development
Security Interferes with some Enabler for end-to-endapplications security
IPv4
IPv6Total
Time
Volu
me
The BIG Questions!• Why √√√√• When ? • Where ?• What is the cost ?
Strategic IPv6 IcebreakersStrategic IPv6 IcebreakersEuro6IXEuro6IX6NET6NET
100% IPv6 readiness by 2005• Prime Mister of
Japan YoshiroMori – 80 M$
• Korean MIC followed Japan Feb 23, 2001– 68 M$
Japanese Deployment Efforts• Governmental
– IPv6 Council (Kyogikai)– JGN (Japan Gigabit Network) IPv6
• R&D– WIDE IPv6 / NSPIXP6 / Kame / Usagi / TAHI
• Industrial– IAJapan IPv6 Deployment Committee– JPNIC IPv6 project– IPv6 Operation Study Group
• Publication– IPv6 Journal (RIIS)– v6start (Nikkei BP)
IPv6 Council
• Initiated by Ministry of Public Management, Home Affairs, Posts and Telecommunications
• Chair: Jun Murai• Not only router vendors and service providers, but
home appliance developers etc. are involved• TAO (Telecommunications Advancement Organization of Japan)
conducts a nation-wide IPv6 experiment including home appliance application development, using budget of 8 billion Yen (= $ 800k).
Telecommunications• connection-oriented• services are tightly-coupled to
the network• centralised control• hardware is fault tolerant• features selected at call setup
• obsession with QoS• low latency
Data Communications• connectionless• services are loosely-coupled
to the network• distributed control• software is fault tolerant• features selected during
sessions• little attention to QoS• high latency
The telecoms and data worldshave different roots:
!Common applications !
Common infrastructure
The telecoms and data worldshave different roots:
!Common applications !
Common infrastructure
Gracefull Transitions
IPv4IPv4
TB6to4
6over4NAT-PT
BISSOCKS
TB6to4
6over4NAT-PT
BISSOCKS
NAT-PTBIS
SOCKS
NAT-PTBIS
SOCKSIPv6IPv6
IPv4 OnlyIPv4 Ocean
IPv6 Island
IPv4 Island
IPv6 Ocean
IPv6 Only
IPv4/IPv6 Translation Required
ExperimentalIPv6 Network
DSTMNAT-PT
BISSOCKS"4to6 ?"
DSTMNAT-PT
BISSOCKS"4to6 ?"
IPv6 - a small step for IP but a giant leap for Telcos
Early conclusion
+ Wireless
ΣΣΣΣΣΣΣΣ Internet 2001 is a baby!Internet 2001 is a baby!
Internet Growth
100 IP Adds/person!
+ Always-On+ Info Soft
Appliances
FinallyFinally
The longer the upgrade is postponed, the costlier it will be
and the more complex the transition will be !
(compared to
IPv4
• Y2K is Over; Media Needs a New High-Tech Story
Why IPv6 is a Hot Story
• Any Doomsday Scenario for the Internet is Big News
–Denial of Service Attacks• Internet Luminaries Are Talking
about IPv6 More
- Cerf, Dyson, etc.
IPv4
The New Internet
1969 2000 2001 2002 2003 2004
IP E
volu
tion
1 billion +Connected Devices
100m
NCPNCP
1983
IPv6 FORUMIPv6 FORUM
Vint CERFHonorary Chairman