| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Here is the problem with using a single downstream wavelength for the first channel.
If you do it my way, then we will need the following interfaces:
Option 1 25G.
Option 2 25G.
50G (using channels 1 and 2)
100G. (the channels are associated in the same way as the options 1 and 2).
If you do it your way, we will then need:
Option 1 25G.
Option 2 25G.
50G (using channels 1 and 2)
Option 1 100G
Option 2 100G These last two are different because the association of the channels are different.
If we consider using the two option channels for 50G, then again, if we don’t use separate downstream channels, we end up with two types of 50G.
To the issue of cost: we are doing the options because we believe there is significant volume (>10M, eventually) in both of these applications.
The cost of bringing up a new laser wavelength is not so much, and it will be divided over a large volume. It becomes negligible.
Sincerely,
Frank E.
From: John Johnson [mailto:john.johnson@xxxxxxxxxxxx]
Frank, I would agree with Ed that the operational justification for and cost implications of two different downstream versions hasn't been sufficiently covered yet. A supposed coexistence advantage
with a single-channel 50G PON also needs a contribution to justify it. Regards, John On Wed, Jul 12, 2017 at 11:25 AM, frank effenberger <frank.effenberger@xxxxxxxxxx> wrote: I’d rather do it as a package.
Frank E. From: Harstead, Ed (Nokia - US) [mailto:ed.harstead@xxxxxxxxx]
Frank, We are thinking along the same lines. However I propose that we take the motions one step at a time. Dekun and I drafted this motion, which only addresses two upstream options:
802.3ca shall adopt a wavelength plan with two 20 nm wide wavelength options for the first 25G upstream channel:
1. “Plan A” option: center wavelength between 1300 and 1320 nm, exact value to be determined
2. “Plan B” option: 1270 nm
Only the “Plan A” option will be required to co-exist with 10G EPON. Your motion bundles in two downstream options, and I don’t think we have the same degree of consensus on that. What do you think of starting with the above upstream option motion,
and then follow that with a downstream option motion, extracted from your text below? Ed From: frank effenberger [mailto:frank.effenberger@xxxxxxxxxx]
I have some motions I would like to make.
Given that I think there is support for defining at least the first 25G channel, and that the ‘either or’ plan has a lot of support,
I suggest: 802.3ca shall adopt a wavelength plan for the first 25G channel with two options, Option 1: downstream at 1358.4nm width 3nm, upstream at 1310nm width 20nm; coexistent with 10GE-PON Option 2: downstream at 1336.6nm width 3nm, upstream at 1270nm width 20nm; coexistent with G-PON reduced wavelength set.
Consequent to this motion, I propose another motion:
802.3ca shall add the objective:
Wavelength allocation allowing concurrent operation with G-PON reduced wavelength set (1290-1330nm) PHYs Sincerely, Frank Effenberger |