I would like
to propose adding additional
data points to your diagram, http://www.dovenetworking.com/images/ratevschannel.jpg,
one representing DQPSK, with number of channels =1, baud rate = 50 and
representing Coherent WDM, with number of channels = 1, baud rate =10.
points for 80 and 120 Gbit/s)
Fellow HSSG Members,
The Fiber Optic Adhoc met
on 10/18/06 at 9am PDT to discuss
Objectives of the AdHoc and plan for our upcoming plenary meeting.
Thomas Fischer - Siemens Germany
Rick Pimpenella - Panduit
Pete Anslow - Nortel
Peter Dartnell - Brookham
Allesandro Barbieri –
We started off with a
monologue by the chair (me) about how
I envision the adhoc moving forward. First off, I re-iterated that the
objective of the adhoc is to evaluate the technical and economic
various fiber-optic approaches to higher speed fiber, and that we are
to provide a recommendation to the HSSG. Then I indicated that due to
time, we are unlikely to have an opportunity to take in presentations
the plenary, so presentations should be intended for the November
we meet as an adhoc in November, it will be to organize future meetings
but presentations should be intended for the larger group.
Regarding our process for
moving forward, I offered was that
we could create a graphic representation for a given data rate that
the relationship between baud rate, number of channels, and data rate
areas where technical and economic feasibility were more likely, and
where T & E were less likely, and from that basis provide direction
HSSG. There appeared (through lack of an alternative proposal) to be
on this approach.
Then we chose to start
off the process by excluding areas
where we determined reasonable bounds would exist. For example, 100
1G/s we determined as a reasonable exclusion. So we determined that a
channel assumptions could be made (note, these channels may be parallel
or WDM or a combination thereof) and we decided to exclude all data
than 80G, 100G, and 120G.
I should note that the
adhoc made these exclusions and
assumptions with an understanding that the HSSG may disagree and
request us to
re-address any particular baud or channel number. The primary objective
exercise was to give the adhoc a direction to proceed until we have an
opportunity to narrow down our activity based on HSSG direction.
- Thomas Fischer - Siemens
One cube per data rate per reach and timeframes
- Hugh Barass - Cisco -
number of fibers and numbers of wavelengths not necessarily independent
- Jugnu - Avago - number of
fibers and number of wavelengths are related.
- Piers 80G, 120G
- Chris Cole - Finisar 100G
- Thomas F - Long Haul
(40Km) should be considered.
- Dan D - Short Haul (100m)
and a MAN range of (10km) would be basis for study.
- Thomas F - FEC as an
example… needed for long haul and thus excess bandwidth important
- Dan D - Timeframe
assumption (3-4 years)
- Chris - Should further
narrowing be considered? 10x10 and 5x20?
- Piers 8 and 12 channels
for 80 and 120
- Hugh 1,2,4 channels
- Peter Anslow 6, 12 for 120
- Cumulative list of
“channels to consider” (1,2,4,5,6,8,10,12)
- Piers - WDM finds
efficiency with number of wavelengths
- Peter - WDM will be
bounded to a given a given number of wavelengths
- Chris - PO should be
kept on the table for 100m study.
- Rick Pimpenella - Can we
eliminate 1 and 2 channels?
- Should we leave more
complex modulation methods on the table or exclude the analysis to NRZ.
- Hugh - we should not
exclude 1,2 at this time.
I made any errors or omissions on this report, please contact me
directly and I
will update it for publication on the IEEE802.3 HSSG website.
Chairman IEEE802.3 HSSG FO adhoc
Principal Engineer - LAN PHY Technology
Dove Networking Solutions - Serving ProCurve
(530)389-9315 (home office)