| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Resending, somehow this did not seem to get on the first try.
DP
I agree with Frank that un-audited calculations are
problem-some in that the the CNR depends on a number of parameters, not all
relevant to all applications. I hereby offer my unaudited
calculation.
I have attached a document in which I do not calculate
CNR, but I compare the Raman noise contribution from a 10G signal, 20 nm above
the video signal, to the contribution from a class B, B-PON signal at 1490 nm on
a video signal at 1555 nm.
The B-PON is widely deployed with analog video in North
America. As long as the "Raman noise" from the 10G signal is less than the
B-PON "Raman noise" the 10G-EPON should be considered deployable with a
video overlay.
The B-PON "Raman noise" limit is however
arbitrary in that one can certainly deploy RF video with 10G-EPON systems
with noise that exceed the B-PON limit. In some cases one may need to mitigate
the Raman effect by one or more of the following:
Use digital modulation (QAM) or more of it with a
greater fraction of the RF carriers.
Use greater RF pre-emphasis on low frequency RF
channels.
Use separate feeder fibers for 10G-EPON and RF
video
etc.
I am comfortable with a +14 dBm 10G launch power as a
maximum power limit.
I also agree with Frank, that as a practical point,
analog video is becoming less important.
Effective 2009, in the USA all over the air channels
will go all digital, In addition all wireline video providers who do not provide
a "CableCard" device to enable the use conditional access on third party set top
boxes, and TV sets, will also have to go all-digital. As far as I know
aside from legacy cable TV providers, the conclusion is that going all digital
is easier than building a cable-card infrastructure. It seems as if
Verizon, for example would fall into this category.
In addition, I also do not believe that the
incompatiblablity between GE-PON idle frame interference at 62.5 MHz and its
harmonics and analog video deterred the deployment of
GE-PON;
I welcome comments.
David Piehler
Alphion
mobile: +1 732 692 4581
From: Frank Effenberger [mailto:feffenberger@HUAWEI.COM] Sent: Friday, September 07, 2007 12:54 AM To: STDS-802-3-10GEPON@LISTSERV.IEEE.ORG Subject: [8023-10GEPON] Raman cross-talk calculations Dear all,
I note that there was a contribution
regarding Raman interactions. I have concerns that the calculations used
there do not reflect the actual operational cases and trade-offs of fiber
distance loss and Raman interaction strength, the significant effect of
intra-office losses, the use of low-channel pre-emphasis, and so on.
The bottom line is that these
calculations have many parameters, and drawing any conclusions from un-audited
calculations is dangerous. To help our progress, I have
attached my calculation model for Raman cross talk with a video channel and 2
digital waves. (This is a slight update of the file I used to prepare my
January 07 presentation on this same topic.) There are two independent sheets.
The first computes the delivered CNR. My calculations show that a 1G
launch power of 2 dBm and a 10G launch power of 8 dBm seems to deliver more than
46 dBm (which is the current performance spec of important operators). One
must remember that typical in-office losses are 2 dB, so the maximum launch
powers of the respective systems would be +4 and +10 dBm at the OLT face-plate.
That is a significant 3dB more than a number suggested recently.
I invite others to work with the
model, and provide their own comments. As a parting comment, I note that
analog overlay use is rapidly declining. I know of only one large
application of such systems, and that user is now phasing out the support of
analog. So, the entire point may become moot.
Sincerely, Dr. Frank J.
Effenberger ¸¥À¼¿Ë °£·Ò²©¸ñ Huawei Technologies Office (732) 625
3002 Cell (908) 670
3889 |
Alphion - Raman crosstalk 10G-EPON vs. B-PON.pdf