Frank: A further clarification
for 10/1GBASE-PRX might be: 10D1UGBASE-RPX , unless that violates some
other naming convention.
I would amplify the comment that "...27
port types is alot" to: 27 port types is wayyyy too much,
and will ensure a lack of market
traction. You may recall that
Jonathan Thatcher tried to (with limited success) reduce the possible number
of port types in the
initial 10G (802.3ae) standards
meetings. Had there been more, market traction in 10G would even
be less than is at the
present time, although these types of
things end up in endless debates, best had over a beer.
Reducing port types to a few good ones
is a challenge for the whole group, and will require setting altruism before
personal bias.
Best Regards
Maurice Reintjes
MindspeedTM
Hillsboro, Oregon,USA
Office Phone (503)-914-5370
Mobile (503)-701-0797
Frank Effenberger <feffenberger@HUAWEI.COM>
03/06/2007 11:41 AM
Please respond to
Frank Effenberger <feffenberger@HUAWEI.COM>
To
STDS-802-3-10GEPON@listserv.ieee.org
cc
Subject
Re: [8023-10GEPON] Reminder to presenters
and Time slot request
Duane and Marek,
I am very happy that you have looked into these editorial matters. Your
plans seem very logical.
Some musing on the issue of naming: In my presentation on power budgets,
I
give some consideration for names. The "X" in the previous
PON optics was
associated with the 8b10b coding. Since all the 10G proposals we
are
talking about so far (I think) use the 64b66b coding. Therefore,
the
correct letter to use is "R".
The asymmetric case raises interesting questions regarding what to call
it,
since this case will use (I think) different data rates AND different coding
in either direction. I think the simplest way to extend this would
be to
just put both designations into the name.
So, the symmetric PONs would have the letter designation: PR, and the
asymmetric PONs would be called: PRX (which has the added benefit of
sounding vaguely sexy.)
In the presentation, I tally up the port types, and end up with 12 (3 loss
grades * 2 speed combinations * 2 ends). There are actually only
7 new
power budgets (6 10G budgets, plus the 29dB upstream 1G budget).
It is interesting to note that the 12 port types listed above imply the
support of a single data rate in either direction. Actually, if we
are
truly rigorous, there may be additional OLT port types, because I think
it
is good if the OLT phy can support both 10G and 1G at the same time. I
have
revised my presentation (attached) to include the 'truth table' that
considers all of these combinations.
On the leading speed grade designation: we start with "1000BASE"
and
"10GBASE". If we just hybridize, we get the unwieldy "10G1000BASE"
- that's
no good. We could shorten to "10/1GBASE". Or, we
could follow Glen's
advice and say "11GBASE". Actually, both of these could
be useful, because
the 10/1 could be used for the asymmetric case, while 11 could be used
for
dual-support. I add these thoughts into the revision, as well.
Of course, we can try to pare down all the combinations to a smaller set.
Certainly, 27 port types (including the existing 1G PONs) is a lot.
Sincerely,
Frank Effenberger.
-----Original Message-----
From: Duane Remein [mailto:duane.remein@ALCATEL-LUCENT.COM]
Sent: Monday, March 05, 2007 6:00 PM
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: [8023-10GEPON] Reminder to presenters and Time slot request
All,
Please remember that Glen is traveling and has asked me to upload
presentations to the IEEE WEB site. So far I have 1 presentation
from
Peter Anslow, 5 from Frank E., 1 from Marek H. and 1 from Glen (along
with the one from myself attached).
Glen,
Can I please get about 15 minutes to present the attached file on behalf
of Marek and myself at the Orlando meeting? The presentation covers
some preliminary logistics to help provide us with direction in
preparing for Draft 1 editing tasks.
Thanks
Duane