Initial Design of the TFN Regional Fiber Network

1
Initial Design of the TFN Regional Fiber Network

• The Quilt Optical Workshop
• June 22, 2004
• pschopis_at_oar.net

2
Topics

• First Cut
• Defining Mission
• Buy or Build
• Getting your Ducks in a row - Which Vendor
• Tipping Points

3
First Cut

• Everyone is doing it
• Is there a solid business case?
• What are the costs of provisioned services vs.
  owning your own fiber plant
• Does it fit within scope of primary mission
• Does it serve the primary stake holders in your
  community?
• We have 91 member institutions
• 10 I2 members
• 25 SEGP

4
Defining Mission

• So what do you want this thing to do?
• Reduce long term costs
• It appears that long term costs can be contained
  by making large upfront capital investment
• Allow dedicated light paths through regional
  fabric
• Research community believes that Lambda based
  service provides resource that other mechanisms
  dont
• Allow interaction with other Gigapops and
  entities in a dedicated fashion (HOPI?)
• Problem inter vendor compatibility

5
Buy or Build?

• Rule of thumb It costs 50K in urban environment
  and 40K per mile in rural
• Based on this Ohio could have built 84 miles
  urban or 105 miles rural
• We bought (IRU) 1600 miles for 4,200,000 e.g.
  about 2625 (Average price)
• Primary Wiltel and AEP
• Others Qwest, AFS, Buckeye Telcom, and CityNet

6
Getting Ducks in A Row

• TFN uses SBC as integrator
• Get the target off our back
• Primarily looked at Nortel vs Cisco
• We did talk to other vendors
• SBCs staff seemed fairly objective
• We require production service e.g. we are a
  little reluctant to go to far out of the box
• Due to size of state and budget went with Cisco
  ONS 15454

7
Tipping Points

• Nortel
• Primary caveat too expensive and lack of
  flexibility
• Went with Cisco Transponder (G.709) based service
• Good - Less expensive in short term uses FEC for
  greater reach leverage Layer 23 for resiliency
• Bad - adding Sonet or RPR later adds cost to that
  service
• Going with Cisco has caveats
• They insist on providing complete solution

8
General Principles

• We used full 15454 implementation
• Used the AMP and that go into 15454
• Makes management uniform
• Makes certain aspects of sparing easier
• We over engineered by 4 to 10 for degradation
• Accounts for aging fiber (aerial)
• Accounts for fiber cuts
• Constrained only by fiscal considerations

9
(No Transcript)
10
(No Transcript)
11
(No Transcript)