Re: [STDS-802-3-EPOC] Additional Eval Criteria item to support ToD (time of day) delivery for Mobile Backhaul
Steve,
I'd like to move that the 3rd bullet from Bill is already addressed by the
set of registers added under 802.3bf project. All we need to do is use them,
i.e., require their use in the respective EPoC register package (mark them
as mandatory, and not optional). Once done, vendors would have the way to
populate min/max delay for Tx/Rx path across the given layer implemented in
the device. That should take care of this requirement.
On the whole concept of equalizing delay - given the level of moving parts
the PCS will likely have, I find it technically challenging to figure out
even at the conceptual level how that could be achieved with any level of
precision, especially if the effective data rate fluctuates over time and we
need to constantly play catch with the depth of delay adaptation queues.
Combine that with burst mode transmission and only statistically predictable
scheduling delay that is network load dependent, I'd like to see a detailed
proposal before we agree to do any delay equalization.
Marek
From: Shellhammer, Steve [mailto:sshellha@xxxxxxxxxxxxxxxx]
Sent: Thursday, 24 January, 2013 2:54 PM
To: STDS-802-3-EPOC@xxxxxxxxxxxxxxxxx
Subject: Re: [STDS-802-3-EPOC] Additional Eval Criteria item to support ToD
(time of day) delivery for Mobile Backhaul
Bill,
I added that as a possible evaluation criteria. I also
include the changes we made this morning.
Steve
From: Bill Powell [mailto:bill.powell@xxxxxxxxxxxxxxxxxx]
Sent: Thursday, January 24, 2013 8:18 AM
To: Shellhammer, Steve; STDS-802-3-EPOC@xxxxxxxxxxxxxxxxx
Subject: Additional Eval Criteria item to support ToD (time of day) delivery
for Mobile Backhaul
Hi Steve,
Related to the Eval criteria delay items that we currently have in the Eval
Criteria table, I don't think this will adequately capture an Eval criteria
that will meet the needs of Moble BackHaul synchronization for EPoC. I'd
like to suggest adding the following additional Eval Criteria item, based on
my EPoC Synchronization transport presentation from the Phoenix meeting:
Design to support a time transfer error performance of +/-120 ns or less
from the CLT MAC/PLS to the CNU MAC/PLS
There are a number of ways this can be met:
- Fixed & equal delay from MAC/PLS to the Medium and the Medium to MAC/PLS
(but a not likely case)
- Equal TX / RX delay to/from the MAC/PLS interface to/from the medium
- Reportable time delay across the PHY layers to a known level of precision
in each direction
(may depend on configured OFDM, FEC, and interleaving configured
parameters)
- Others?
I'm open to adjusting this number - up or down (preferable) as we move
forward with choices of PHY parameters and reference models. Our
competition (GPON, DOCSIS, P2P Ethernet) is currently doing much better than
this performance level.
Comments welcome...
Regards,
Bill
--
Bill Powell
Alcatel-Lucent
Fixed Access Systems Engineering
2301 Sugar Bush Road
Raleigh, NC 27612
bill.powell@xxxxxxxxxxxxxxxxxx
(O) 919-850-6462
(Cell) 919-614-3225
_____
________________________________________________________________________
To unsubscribe from the STDS-802-3-EPOC list, click the following link:
https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-EPOC&A=1