Transcript Electrons to Photons

Advanced Networking Infrastructure and
Research (ANIR)
“Future Directions in Networking”
Aubrey Bush
Division Director, ANIR
National Science Foundation
Internet2 Members Meeting Washington, DC
1
Outline









Electrons to Photons: Network Speeds
Electrons to Photons: Network Components
Three Network Model
Cyberinfrastructure
Abilene Research Network
Experimental Networks
Experimental Network Characteristics
Research Network
International Participation
Internet2 Members Meeting Washington, DC
2
Electrons to Photons: Network Speeds
Photons
Production and or in development
Electrons
Research
Research
Production and or in development
Desktop connections
Future Desktop
???
Network capacity
Phone Modems
DSL
Cable Modems
1 Kbs
10 Kbs
100 Kbs
Future Network
T1
T3
OC3
OC48
OC768
Ethernet Fast Ethernet GigE
10Gig
DWDM
1 Mbs
10 Mbs
100 Mbs
1 Gbs
10 Gbs
Bandwidth
Internet2 Members Meeting Washington, DC
100 Gbs
???
1 Tbs
???
10 Tbs
100 Tbs
3
1 Pbs
Electrons to Photons: Network
Components
Electrical
Circuit switches
Packet switches
IP Routers
ATM, Ethernet,
etc
Electrical to
Optical conversion
SONET
Packet over SONET
DTM
Today’s Technologies
Optical
Lambda switching
Optical packet switching
Optical burst switching
MEMS, Holography,
etc
Tomorrow’s Technologies
Internet2 Members Meeting Washington, DC
4
Three Network Model
Abilene
Circuit switches
Packet switches
IP Routers
ATM, Ethernet,
etc
Electrical
NSF Experimental
Network
SONET
Packet over SONET
DTM
NSF Research
Network
Lambda switching
Optical packet switching
Optical burst switching
MEMS, Holography,
etc
Electrical to
Optical Conversion
Internet2 Members Meeting Washington, DC
Optical
5
Cyberinfrastructure
 Recognition of the new requirement of infrastructure to
support leading edge science and research
 Dependence on technology for future scientific advances
 Cyberinfastructure capabilities include






computational power and high speed networks
distributed, ad hoc and embedded sensor networks and arrays
large data repositories
systemic security
large-scale interoperability and collaborative tools
middleware
Internet2 Members Meeting Washington, DC
6
Abilene Research Network *
 Production Research Network




24x7 support
High speed capacity
53 direct connections
207 participants
 Extended service until October 2006
 Expanded capacity upgrades in process
 moving to OC 192
 International connections through STARTAP
and MOUs
* Steve Corbato, Director - Backbone Network Infrastructure
Joint Techs, Tempe Arizona, January 2002
Internet2 Members Meeting Washington, DC
7
Experimental Networks
 Experimental networks are required to solve many
e-Science problems
 Experiments will be driven by Grand Challenge
Applications
 vantage point is the end-user
 application performance is the success indicator
 International connections required for multi-national
e-science projects and programs
 Collaboration across disciplines a hallmark
 Systemic approaches to solve networking issues
Internet2 Members Meeting Washington, DC
8
Experimental Network Characteristics
 Quality of Service
 Dedicated Provisioning for guaranteed data rates
 Support of other network layers such as GMPLS
and OBGP
 Repeatable network experiments
 Reconfigurability
 Experimental protocols and approaches for high
throughput, low latency, large bursts
 End-to-End performance and support
NSF CISE Grand Challenges in e-Science Workshop
http://www.evl.uic.edu/activity/NSF/index.html
Internet2 Members Meeting Washington, DC
9
Research Network Characteristics







Disruptive technologies and approaches
Hybrid and experimental designs
End-device research
Core technology development
New protocol research
Alternative network architectures
Testbed implementations
Internet2 Members Meeting Washington, DC
10
International Participation
 Coordination of advanced networking research activities
and programs, including research production networks
 Support of Grand Challenge applications which transcend
geographic and national boundaries
 Direct connections to the NSF Experimental Networks
 multiple projects and connections
 some problems can only be resolved by international
partnerships
 Joint strategic planning efforts for longer term support
of e-Science and research
Internet2 Members Meeting Washington, DC
11