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

Re: [10GBT] phy parameter list


I have one comment:

In your xls proposal, first item, R11: "assumed additive Gaussian noise
of receiver". Where is this noise measured ? At the input of the slicer ?
I do not think that the noise is Gaussian after the filtering in the
receiver. Sailesh gave a presentation on this issue at the 802.3ab.

Jaime Kardontchik
Mountain View, CA 94041

-----Original Message-----
From: Sanjay Kasturia [mailto:SKasturia@TERANETICS.COM]
Sent: Friday, April 02, 2004 2:45 PM
Subject: Re: [10GBT] phy parameter list

I had earlier sent a list of phy parameters that we must specify to get
to a relatively complete Draft 1.0.

Brad will put this list up on the website. If there are other items to
add to the parameter list, please send me email or send email to the
reflector.  I will collect all the input and we can discuss
adding/modifying the list at the May meeting though it would be great to
get consensus before via the reflector.

As for the process of coming up with consensus values for the
parameters, we have two choices:
A) Get consensus on specific parameter values
B) Look at relatively complete PHY proposals, evaluate these, and then
the parameter values will fall out of the PHY proposal that is selected.

I suspect we will probably end up with some mix of both approaches. We
may be able to get consensus on some parameters with a point discussion
on the specific parameter and some will fall out of a comparison of
relatively complete phy proposals. If you have suggestions on this
front, get them out on the reflector or to Brad Booth or me.

To give the group some sense of what is a relatively complete PHY
proposal, I have put together another spreadsheet - which is attached.
This is based on an earlier spreadsheet from 1000Base-T days that
Sailesh Rao provided and on input from a number of individuals active in
the study group/task force. This isn't final, and is open to discussion
so please don't hesitate to provide feedback. I hope we can get broad
agreement on this spreadsheet format at the May meeting if not before.
In any case, this should give an idea of the level of detail that could
allow a reasonably informed comparison of the different proposals. If
you are planning to make a phy proposal at an upcoming meeting of the
taskforce, please fill in as much as you can.

This spreadsheet - has entries for three categories of information -
i) Information to determine if the proposal meets the reach objectives
and hence is required. Examples include - does the proposal meet the BER
target over the specified channel models? If it does, with how much
margin? Information needed to determine if the proposal will be
compliant with relevant EMI or other objectives etc.

ii)Information that will be in the final standard and hence is required-
examples include symbol rate, FEC encoding, transmit filtering, transmit
voltage levels etc.

iii) Information that provides some details of a possible implementation
to achieve the performance claimed in the PHY proposal. This does not
have to reflect your final implementation and will not be in the
standard but will help other voters on the task force to better
understand your performance claims, the complexity of the associated
implementation and also possibly contribute to improving the proposal.
Other information/parameters specific to your proposal that will help
others better understand it should be added in.

I would like to thank the following people for providing inputs that
helped me put this spread sheet together: Oscar Agazzi, Brad Booth,
Scott Powell, Sailesh Rao, Katsutoshi Seki, Jose Tellado, Geoff
Thompson, Ariel Yagil, George Zimmerman. My apologies to others I forgot
to list. Errors are my responsibility.


Sanjay Kasturia
cell (650) 704-7686
office (408) 653-2235

-----Original Message-----
From: stds-802-3-10gbt@IEEE.ORG [mailto:stds-802-3-10gbt@IEEE.ORG] On
Behalf Of Booth, Bradley
Sent: Thursday, April 01, 2004 5:59 PM
Subject: Re: [10GBT] phy parameter list


Thanks for putting the list together.  I will try to get the list up on
the website by early next week (I'm on the road this week).

Just a note to the Task Force, I am a bit concerned about the lack of
discussion around this material and lack of technical discussions in
general.  For the Task Force to make progress, the reflectors should be
used as much as possible to discuss ideas and concepts.  Please do not
wait until our meetings to bring up issues or concerns, as we should be
using our meeting time to present information that is establishing or
has established consensus (hopefully via the reflector) to build out the
baseline and eventually a completed draft specification for Working
Group ballot.


Bradley Booth
Chair, IEEE P802.3an Task Force
512-732-3924 (W)
512-422-6708 (C)

-----Original Message-----
From: stds-802-3-10gbt@IEEE.ORG [mailto:stds-802-3-10gbt@IEEE.ORG] On
Behalf Of Geoff Thompson
Sent: Saturday, March 27, 2004 4:15 PM
Subject: Re: [10GBT] phy parameter list


I think it would be helpful to have the values from 1000BASE-T preserved
on the spreadsheet as a basis for comparison.



At 05:05 PM 3/26/2004 -0800, Sanjay Kasturia wrote:
>I hope all of you have recovered from the visit to Disney World, from
>the ethernet guitar music and from the wild nights at Pleasure Island.
>While we made good progress in arriving at an agreement on the channel
>models and on the link specification front, I am obliged to point out
>that we will need to reach agreement on a number other critical
>parameters to have a reasonably complete draft 1.0
>I looked through the draft that George Eisler generated by modifying
>1000BASE-T spec and the large number of TBDs and listed some of the
>important ones in an excel spreadsheet. The spreadsheet is attached and

>I am sure Brad can help us put it on the website so it becomes a living

>document. If there are other critical parameters that should be listed
>in it, please send me email.
>I think participants should start thinking of presentations for the
>802.3an meeting to make a case for specific values for the parameters
>listed in the spreadsheet.
>I am collecting information on how the selection was made for
>and will come up with suggestions on how we can start the process of
>comparing different phy proposals in case there isn't unanimity on all
>the parameter values right from the start!
>Sanjay Kasturia
>Editor, IEEE 802.3an
>cell (650) 704-7686
>office (408) 653-2235
>Item #  Description                             Current proposals
>         PCS
>1       Symbol rate
>2       Modulation
>3       Frame structure
>4       Transmit encoding for FEC
>5       Transmitter bit to symbol mapping
>6       Transmit processing
>7       Transmit latency through PCS
>         PMA
>8       Transmit voltage specification
>9       Transmit pulse shaping
>10      Transmit master and slave jitter specifications
>11      Transmit linearity specifications
>12      Maximum allowable transmit distortion
>13      Transmit noise floor
>14      Transmit latency through PMA
>         Startup protocol
>15      state diagram for training
>16      Coefficient exchange if required
>17      Coefficient initialization if required
>18      Mode selection method if phy operates in multiple modes
>         Receiver performance requirement
>19      BER or FER over specified channel models             10^-12 for
>20      Receiver latency requirement