Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: Supporting Installed Fibre




Alan,

I'm sorry that you got the impression in Montreal that support for 300m of
installed MMF will be complex and expensive.  The WWDM approach that we have
been proposing should be a very low-cost solution for existing MMF links out
to about 300m (and also single-mode links out to about 10km).  Whether we
can achieve 300m on the worst-case fiber, or 275m, or 250m, will depend upon
the precise link budget and receiver sensitivity that the committee ends up
agreeing on.  

Best Regards,

Brian Lemoff
HP Labs

-----Original Message-----
From: Alan Flatman [mailto:a_flatman@compuserve.com]
Sent: Thursday, September 16, 1999 7:30 AM
To: IEEE 802.3 HSSG
Subject: Supporting Installed Fibre



I would like to offer some perspectives on supporting installed multimode
fibre links.

Firstly, let's remind ourselves of the statistics. I presented a survey of
installed fibre links in Montreal and compared this with existing data:

I found 25% of installed building backbone links within 100m
The maximum percentage found by other surveys for this distance was 35%
I would add approximately 3% to this figure to include in-range campus
links

I found 58% of installed building backbone links within 200m
The maximum percentage found by other surveys for this distance was 70%
I would add approximately 4% to this figure to include in-range campus
links

I found 88% of installed building backbone links within 300m
Other surveys found slightly less, so there was good correlation for 300m
I would add approximately 5% to this figure to include in-range campus
links

It's worth noting that cabling standards and installers treat horizontal
(generally copper) and backbone (generally fibre) cabling differently. The
horizontal medium is chosen and installed for generic use and for long life
(typically 10 years). The backbone medium is more application-dependent.
Standards for structured cabling have tried to specify optical fibre for
generic application but we do not believe that this is possible.
Consequently, we expect installed fibre backbone links to turnover more
frequently. As a consultant, I have been recommending a combination of MMF
and SMF in the backbone for some time.

Having said this, some organisations will not or cannot tolerate the
disruption associated with installing new fibre for 10GE. It will therefore
be useful to offer them a solution. Refer to the above statistics in making
this decision.

I tend to agree with Bruce Tolley that a 200m target for installed MMF is
much  better than 100m. I also got the impression from the Montreal meeting
that 300m was possible but complex and expensive. I would like to hear more
from the optoelectronics industry before we move the goalposts.

Regards,

Alan Flatman