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

Re: [8023-10GEPON] [FEC Superating] - kickoff preso



Mike,

Thanks for the correction.  I wrote 10B because I was thinking of the 8B/10B
encoding that XAUI uses.  

For the PMD, I realize that we don't define a serial electrical interface at
10.3125 GBd, just a serial optical interface.  Presumably, we want to re-use
as many interfaces as possible within 802.3, or at the very least, reference
an existing interface from some other group.  Other groups, such as those
involved in the XFP MSA, have taken on the burden of defining the serial
electrical interface at 10.3125 (and other speeds).  

Before the super rate is transmitted across the channel or into the optical
module, however, it needs to increase from 10.3125 to 11.09.  Is there any
additional buffering that needs to be done in order to go between these two
rates?  It seems that since the clock frequencies aren't nice multiples of
each other that this may be the case.  

- Eric

-----Original Message-----
From: Mike Dudek [mailto:mike.dudek@picolight.com] 
Sent: Wednesday, January 31, 2007 4:58 PM
To: Eric Lynskey; STDS-802-3-10GEPON@listserv.ieee.org
Subject: RE: [8023-10GEPON] [FEC Superating] - kickoff preso

 A quick correction
- XAUI is a 4-bit wide differential interface that runs at 3.125 GBd
PMD is a generic term that doesn't just apply to 10G.   
There isn't an electrical interface that runs at 10.3125 GBd
standardized in 802.3ae.   There are two electrical interfaces that
operate at this speed that are used for 802.3ae.   XFI defined in the
XFP MSA and SFI being defined in the SFF8431 (SFP+) committee.  Both
these interfaces cover more than one exact data rate.   They not only
cover 10.3125GB (for 10GBE LAN phy), also SONET,ITU, and 10GBE WAN phy
at 9.95GBd, 10G Fibre Channel at 10.51GBd, OTU-2 at 10.7GBd and the FEC
G.709version on 10GBE LAN phy at 11.1GBd.  Sorry I don't have the exact
data rates for these applications.  However in my experience most PMD's
can operate over a somewhat restricted range of speeds provided they are
given a reference clock at the correct frequency. 

Regards,
Mike Dudek

-----Original Message-----
From: Eric Lynskey [mailto:eric.lynskey@TEKNOVUS.COM] 
Sent: Wednesday, January 31, 2007 4:26 PM
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] [FEC Superating] - kickoff preso

Frank,

To start with, thanks for putting these slides together.  It's a very
good start to getting the discussion going.  

Here is a list of the defined interfaces in the existing 802.3ae
standard:

- XGMII is a 32-bit wide interface that runs at 156.25 MHz * 2
- XAUI is a 10-bit wide interface that runs at 3.125 GBd
- XSBI is a 16-bit wide interface that runs at 644.53125 MHz
- PMD is a serial interface that runs at 10.3125 GBd

For the FEC with a higher line-rate, we could have something like this:

- FEC/PMD is a serial interface that could run at 11.049 GBd

The increase in baud rate is needed to account for the added parity
bytes.
If we go with the higher line rate, will we need a buffer to add/delete
idles in order to adapt between the different line speeds?

I recognize that there is an existing mechanism mentioned in Clause 49
that can help with this (49.2.5 for Transmit, 49.2.11 for Receive).  I
think that with either the line-rate or MAC-rate modification there will
still need to be additional idle insertion/deletion within the PHY.  In
one case we are adding idles to decrease the data rate to account for
the parity insertion, and in the other case we are adding idles to adapt
between clock boundaries and/or rates.

If we choose to go with a line-rate modification, than we are probably
going to be forced to go with an existing rate.  I haven't heard anyone
propose that we use a line rate not currently used by any other
applications.  If that is the case, then would that restrict us to the
exact FEC algorithms already being used at those higher rates?  There
may be more flexibility in selecting an FEC algorithm if we stick to the
10.3125 rate.  

>From yesterday's ad-hoc call, it looks like we may be going with a 29dB
channel loss and up to 3dB in additional penalties, for a 30dB-32dB
power budget, so we need to keep that in mind, too.  

- Eric Lynskey


-----Original Message-----
From: Frank Effenberger [mailto:feffenberger@HUAWEI.COM]
Sent: Wednesday, January 31, 2007 12:32 PM
To: STDS-802-3-10GEPON@LISTSERV.IEEE.ORG
Subject: [8023-10GEPON] [FEC Superating] - kickoff preso

Dear All,

I have put together the following presentation on the issue of FEC and
line-rate vs. MAC-rate modification.  I tried to include in these slides
all the arguments I have heard favoring one method or the other.  If I
have forgotten your favorite, you can shoot an Email to me, and I'll add
it to the list.  

You may also note that the last slide, entitled "Reaching a decision" is
blank.  I don't know a truly objective way to solve this problem... It
seems to me that when you stack up the pros and cons, these two schemes
are pretty equal.  

One last thought: The one 'hard' (objective) con for the super-rating
scheme is the loss of 0.3 dB of sensitivity.  The one 'hard' con for the
sub-rating scheme is the loss of bandwidth (7% lost).  How can we put
these two items
on a common comparative base?   Usually, the common denominator in these
situations is cost, so...  
What is the relative system cost increase due to 0.3dB optical loss? 
What is the relative system cost increase due to a 7% capacity loss?  
If someone wants to hazard an answer to these questions, please do.

Regards,
Frank E.