Dear Frank,
Your comments are
well taken. I comment on them below:
From: Frank
Chang [mailto:ychang@VITESSE.COM]
Sent: Wednesday, September 26, 2007 5:19
AM
To:
STDS-802-3-10GEPON@LISTSERV.IEEE.ORG
Subject: Re: [8023-10GEPON] Loss Budget
compilation
Very nice initiative,
and great strawman proposal on top of the group consensus so far.
Quick comments for
consideration to further tweak the numbers:
-
I
suggest to include the baseline ER, the group still need to agree on
somewhere.
The ER used in these
slides is ~9dB, and these are average power “ITU” sensitivities. In
subsequent contributions, we will put these numbers into the real
spreadsheets, and so all of these items will be spelled out exactly.
[Marek Hajduczenia] so
far, the TF has not decided on the baseline ER. Feel free to propose
a value and substantiate it. I believe there were suggestions for 6dB ER
for ONU and OLT side, though EML does not get that bad I think.
-
Suggest
channel loss reflect the wavelength dependence for US and
DS.
It is a widely held
belief that the overall channel loss should be specified independently of the
wavelength. This is how operators specify it. In our spreadsheet,
the wavelength dependent loss is calculated, and its impact can be seen in the
excess loss.
[Marek Hajduczenia] I agree with Frank. The difference
in fibre attenuation between 1580 and 1270 nm windows are relatively
small (below 0.5dB) - 0.34 dB/km and 0.36 dB/km, respectively. Do You
feel like playing around with that ?
-
I think
the group has some consensus to budget FEC coding gain separately (in
particularly for pin or apd).
The sensitivities
given here are with some kind of EFEC, with an optical gain of something like
4 dB for PINs and 5 dB for APDs. This is the target. Once we get
this basic framework laid down, then we can move to the evaluation of EFEC
algorithms, and if necessary we can adjust our assumptions. But, we have
to start somewhere, you know?
[Marek Hajduczenia] I also agree. We need to start
narrowing the options down otherwise we will hear all the time statements
like "we cannot move forward because the power budget is not defined" etc....
Let's agree on the rough numbers, cool down the budget to the point where we
know where we stand and then talk to wee hours about the FEC gain. Then we can
revisit the budgets and weigh the changes in the FEC gain versus the added
cost for optics. I think Frank proposes the right way to move from the current
stand-still.
-
Allocation of 1dB
penalty for all confg is too loose, for further consideration, suggest
D/S include the jitter budget, while U/S add extra BM penalty (with
fiber).
Jitter and BM
penalties are included in the receiver sensitivities. This is better
than budgeting them separately, because it allows the implementer to allocate
them as they wish.
[Marek Hajduczenia] Frank C. - please have a
look at the new power budget definition. We kind of moved away from the
standard IEEE formalism thus Your point does not apply to the presented
values.
-
Take the
dynamic range requirements into account for each confg before specifying
overload specs. PR30 DS overload of -10dBm sounds unnecessarily loose.
I don’t know which
direction ‘loose’ means. The dynamic range of 15 dB for the PON loss is
a worldwide standard - I’d tend to keep using
it.
-
Specified the
guaranteed corners (temp./ps range) in considering worst-case
numbers, for example I feel DS 3dB launch power variation is not enough.
You will note that
the OLT types are 3 dB range – these optics are assumed to be cooled, so 3dB
is possible. However, 4 dB would be easier, and there is some precedent
for that, especially on the uncooled optics types. On the ONT types, the
range is 5 dB, which is well supported by precedent. So, that is ok as
it stands.
[Marek
Hajduczenia] DS is relatively safe, since the CO equipment will be
cooled whether we like it or not. ONU side with 5 dB variation is
probably a very natural value for the oncooled equipment. I find the proposed
values realistic and following the current trends in terms of CO/Subscriber
equipment.
-----Original
Message-----
From: Frank
Effenberger [mailto:feffenberger@huawei.com]
Sent: Tuesday, September 25, 2007 9:31
AM
To:
STDS-802-3-10GEPON@listserv.ieee.org
Subject: [8023-10GEPON] Loss Budget
compilation
Dear All,
Per request of the chairman, I
have compiled the loss budgets as best I can assess them.
In doing so, I have attempted to
minimize the number of distinct component levels.
Please take a look, and comment
freely on the exploder.
Sincerely,
Dr. Frank J.
Effenberger 弗兰克 埃芬博格
Huawei Technologies USA
1700
Alma Drive, Plano TX
75075
Office (732) 625
3002
Cell (908) 670
3889