[802.3_NGEPON] Gilbert burst noise model

[Mark Laubach <000006d52ee8f1bc-dmarc-request@xxxxxxxx>]

Hello  NGEPON Task Force,

 

Dora Van Veen mentioned the Gilbert-Elliot burst error model in  vanveen_3ca_1a_0317.pdf that was presented at the most recent P802.3ca Vancouver meeting.   During the 802.3 Working Group closing plenary on Thursday, Peter Anslow requested the addition of â??(or the frame loss ratio equivalent)â?? be added to the BER objective for the Multi-Gig Automotive Ethernet PHY Study Group citing â??to account for burst errorsâ??.  Taking notice, I went and asked both him and Adam Healey what is being used beyond AWGN for burst noise modeling in 802.3.  Both of them referenced the Gilbert Burst Model in the following presentation to account for DFE error propagation (see page 5):

 

http://www.ieee802.org/3/bj/public/nov11/cideciyan_02a_1111.pdf

 

We have been discussing FEC codes for the past several meetings and sooner or later are going to get into performance simulation details of proposed candidates.  Based both on Doraâ??s mention above and that it being used and recommended in other past and current 802.3 efforts, Iâ??d like to raise awareness that the Gilbert burst noise model may also be of interest in our Task Forceâ??s future FEC selection considerations.

 

Comments?

 

Yours truly,

Mark Laubach