Future Optical Internet ATM will play a key role

1
Future Optical InternetATM will play a key role
Different Protocol Stacks Integrated to provide
different size bandwidth pipes and CoS
IP/ATM Network
IP SONET Network
IP/ATM Network
IP SONET Network
HDWDM OC-3084
IP Optical Network
IP Optical
IP Sonet
QoS VPNs up to OC3
Greater than OC-48
OC3, OC12, OC48
2
Future Optical InternetMPLS ATM
ATM VCs
ATM VCs
SONET LSP
SONET LSP
DWDM LSP
IP Optical
IP Sonet
QoS VPNs up to OC3
OC-48, OC-192
OC3, OC12
3
Gigabit Ethernet, 10xGbE or SDL

• Gigabit Ethernet, 10xGbE or SDL Framing
  advantages
• frame size packet size therefore packet
  switching and SAR more efficient and easier to
  implement
• data format consistent with LAN format with no
  translation
• low cost tributary service - do not need to
  terminate link on a router or SONET DCS equipment
• new 10xGigabit Ethernet will approximately equal
  OC-192
• standard SNMP MIBs, but not accessible by out of
  band
• interoperable standard from many vendors
• No scrambling sync or packet loss with state
  based scrambling
• Proposed rate adaptive protocol for gopher bait
  fiber
• Gigabit Ethernet Framing disadvantages
• not very efficient with 8B/10B block coding
• new 10xGigabit Ethernet or SDL may use more
  efficient coding
• No standard out of band management or monitoring
• But some WDM suppliers provide this

4
Packet over SONET Framing

• SONET framing advantages
• well established jitter specifications
• out of band management systems
• can be used in SONET networks for fast restoral
  and protection
• very high efficiency - over 98
• SONET framing disadvantages
• Susceptible to sync attacks even with X43
• SAR processing more complex as there can be
  multiple packets per frame, or packets can cross
  frame boundaries
• POS is packet over PPP over HDLC over SONET
• tributary services require SONET mux services
• NOT and interoperable standard for SONET/SDH
  transport networks
• Interoperable on the other side of ADMs
• Requires DCC bypass to work around single vendor
  solutions
• Too much overhead for transparent networks

5
Optical Internet ArchitectureRings are Dead
Both sides of fiber ring ring used for IP
traffic
Traditional SONET Transport Node
Traditional SONET Transport Node
WDM
WDM
3 0C-48 Tx 2 OC-48 Rx
High Priority Traffic Cannot exceed 50 of
bandwidth in case of fiber cut
Asymmetric Tx/Rx lambdas that can
be dynamically altered
Traditional SONET Restoral
Low priority traffic that can be buffered or have
packet loss in case of fiber cut
6
10xGbE CWDM

• Several companies have announced long haul GbE
  and CWDM with transceivers at 50km spacing
• 10GbE coming shortly
• Costs are as little as 12K US per node (or
  transceiver)
• Future versions will allow rate adaptive clocking
  for use with gopher bait fiber, auto discovery,
  CPE self manage
• Excellent jitter specification
• Most network management and signaling done at IP
  layer
• Anybody with LAN experience can build a long haul
  WAN all you need is dark fiber
• Still some issues remain with OAMP and frame
  jitter

7
Market drivers for GbE in the WAN

• Many ISPs, regional networks, municipalities,
  school districts are purchasing dark fiber or
  building dark fiber networks up to 1000 km rather
  than managed bandwidth
• With dark fiber increased bandwidth only entails
  upgraded equipment costs and no additional
  monthly charges
• Significant savings in relocating servers to
  central site and using VoIP
• Also many carriers willing to sell gopher bait
  fiber (fiber that does not meet stringent
  SONET/DWDM requirements) at a discount
• As such, cost of transmission equipment is
  becoming a significant factor versus cost of
  fiber
• SONET and ATM networks require specialized
  engineering knowledge and skills
• Customers want a technology in the WAN they are
  familiar with and that is easily extensible from
  the LAN e.g. Ethernet
• Dont require the same reliability as telco
  systems

8
Optical Networks for the Rest of Us

• With customer owned dark fiber, 10GbE and 4
  channel CWDM anybody can build a 40Gbps network
  up to 1000km or greater at a fraction of cost of
  traditional telco network
• May not be suitable for mission critical traffic
  (at least not yet)
• But ideal for high bandwidth Internet to the
  school, small business and home
• Ring structures are a customer option not a
  mandatory requirement
• The driver is NOT new applications, but cost
  savings 1 year payback
• Typical cost is one time 20K US per school for a
  20 year IRU
• In Ottawa we are deploying a 60km- 96 strand
  network connecting 22 institutions cost 500K
  US

9
Historical Context

• In the 80s the Information Highway was conceived
  as being a gateway service that would be
  operated by telcos and cablecos
• They would define and deliver the services to the
  end used
• But then came along the Internet
• Internet was built by research and education
  community as set of independent peering networks
  that exchanged information by a mutually agreed
  upon set of protocols TCP/IP
• There was no hierarchy or gateways as in the
  traditional carrier centric view of the world
• The Internet empowered the end user not only to
  be a consumer of services but also an originator
  of services

10
Where are we going?

• Today the Internet is virtual network riding on
  top of a traditional connection oriented
  network of cooper and fiber
• With optical technology such as customer owned
  dark fiber, customer owned wavelengths, 10GbE etc
  we can extend the model of the Internet as tool
  to empower the user to build networks autonomous
  peering optical
• The future telecommunications world may be
  dominated by thousands of customer owned networks
  that peer at the physical as well as at the
  virtual level, Optical Networks for the Rest of
  Us
• A national or provincial K-12 network with its
  own wavelengths and dark fiber
• A national bank network with its own wavelengths
  and dark fiber
• A national health network with its own
  wavelengths and dark fiber
• A radical departure from the carrier centric
  view of the universe

11
3 Different Views
SONET access ring
Telco Network 99.999 reliability only in the
SONET Ring for the telco, no guarantees for
the customer
CO
Cableco Network 99.999 reliability only in the
SONET Ring for the cableco, no guarantees for
the customer
ISP B
Customer Empowered Network 99.999 reliability to
the customer but no guarantees for the ISP
ISP A
2 separate dark fiber builds
ISP C
12
Customer Empowered Networks
Dark fiber Network City C
Dark fiber or CWDM Network City A
ISP B
ISP D
ISP C
ISP A
Dim Wavelength
Customer achieves 99.999 reliability by
multi-home to different ISPs
Optical Label Switched Router
Long Haul DWDM
ISP C
ISP A
First Dark fiber Network City B
ISP B
Second Dark Fiber Network
13
New Challenges and Opportunities

• Customer empowered networks present a whole new
  set of research challenges
• Peering and topology protocols in the optical
  domain what will be the equivalent to BGP and
  OSPF in the optical domain
• Multi Protocol Lambda Switching?
• Defining LSP attributes such as power level,
  wavelength, encoding, etc?
• Interdomain optical MPLS?
• Customer controlled establishment of wavelengths,
  routing and service delivery
• Auto discovery of wavelengths?
• Management and interface systems, etc, etc