Re: [BP] Channel Data Posted - Moving Forward!
IEEE P802.3ap Task Force Members:
Enough is enough! November is right around the corner and we are not
proceeding forward with the evaluation of the signaling techniques. Why? for a
very simple reason. Politics. Obviously, different companies have different
agendas regarding where they would like to see the channel model. This goes
without saying. Anyone who denies this is not being honest. The point is, we
are supposed to be picking the best solution to a problem by experimentation.
By this I mean we should look at the performance of the proposed signaling
techniques through a variety of channels that are IN THE NEIGHBORHOOD of the
channel model. Any good scientist always works this way when trying to prove
out a solution. Denying people the right to look at channels slightly below the
channel model line is not logical. It is purely political!
Let us put the politics behind us and move forward with the analysis. Clearly
any result obtained with a channel that falls below the model is not going to
result in the immediate dismissal of an approach. But, in reality, margins DO
matter. What if someone measures a backplane channel and due to unexpected
thermal conditions in the field, the response drops marginally below the model
line? What if this is combined with some measurement error? What if you add in
some mismatch at the source and load? A good engineering solution REQUIRES
looking at such cases. We have a nice collection of channels to look at now.
Some with ripple, some above the line, some below the line, etc.
Enough said. Just for a little while, lets try to be engineers, not
politicians, and lets get some work done!
Jeff
On 10/6/2004 12:44 AM, Healey, Adam B (Adam) wrote:
> IEEE P802.3ap Task Force Members:
>
> The channel data contributed by John D'Ambrosia has been posted in the
> Channel Model Library section of the web site:
>
> http://ieee802.org/3/ap/public/channel_model/index.html
>
> Thanks to John and Tyco Electronics for making this data available to
> the Task Force.
>
> As a sidenote, I wanted to strongly encourage open discussion of
> backplane channel issues on the reflector. There is limited time
> between now and the November plenary meeting, but a great deal to
> accomplish. Please remember that our objective was to leave the
> November plenary meeting with complete specification baseline. We can
> cannot have a complete baseline without a channel specification or a
> 10GBASE-KR PMD, and we cannot close on the 10GBASE-KR PMD until we have
> an usable channel model.
>
> Based on the spirited discussions that took place on during our interim
> meeting last week, I know many of you have strong opinions on these
> topics. Now is the time to clearly articulate your respective positions
> to the reflector, so that we can work towards finding the middle
> ground. Please carefully consider the discussion that has been posted
> to the reflector to date, and propose your ideas on what we might do to
> move this effort forward.
>
> Thank you,
> -Adam
>
--
_____________________________________________
Jeffrey Sinsky
Photonic Networks Research Department
Bell Laboratories
Room L-245
791 Holmdel-Keyport Road
Holmdel, NJ 07733
Phone: (732) 888-7144
Fax: (732) 888-7074
Email: sinsky@bell-labs.com
_____________________________________________