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

[EFM] FEC proposal - additional conference call



Hi all,
We had last Monday a conference call regarding FEC baseline proposal. We
discussed there the proposal based on:
http://www.ieee802.org/3/efm/public/may02/maislos_3_0502.pdf
This proposal describes a method to add FEC with a compatible frame format
to current 1000BaseX PCS, meaning that an FEC coded packet can be read
without errors in non-FEC legacy PCS and vise-versa. The FEC in the proposal
is located in the PCS layer or in a shim layer above PCS with some
modifications to the layer.
In the call, and afterwards in off-line conversations, people have expressed
a concern regarding the layering of the FEC and regarding the changes in the
current PCS in order to add FEC. They would rather have a separate FEC layer
connected to legacy PCS even in the layer becomes more complicated.

Therefore we have made some modifications in the proposal. Attached is an
updated version of it. The framing of the FEC in the new proposal remains
the same, thus keeping compatibility with legacy 1000BASE-X.
The layering was changed. FEC is added now in a shim layer below PCS. FEC
layer has an RX and TX part interfaced with a Ten Bit Interface from both
sides.
In that case PCS layer remains unchanged.

The new layering does not changes the FEC framing and functionality. Nor
will it probably change any implementation of FEC. The changes are "only"
for a neater partitioning of the layers and for preserving current PCS.

I hope that people would feel more comfortable with this suggestion.
Please send me your remark, any feedback would be appreciated.

I want to schedule another call next Monday to discuss further.
I suggest this time and date for the call:
Date: Monday June 24
Time: 1:00 PM EDT
I will provide the bridge number and access code later.


Lior Khermosh
Passave

fec_tech20_6-f.pdf