Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [8023-10GEPON] [Channel link model ad-hoc]




Dear David,
Thank You very much for the reference.
In the light of the presented and well justified reasons, I decided to change the entry in the XPM field to the following:
 
 
I believe the impact of the XPM especially in the case of digital channel being located in the 1550nm+ window deserves some attention since it was not examined before - all system analysis assumed the downstream digital signal to be transmitted in the 1490 nm window.
 
The presentation will be updated along with the comments from Tatsuta-san.
 
Thank You for Your input
 
Best wishes
 

Marek Hajduczenia (141238)
SIEMENS Networks S.A. - IC COM D1 R
Rua Irmãos Siemens, 1
Ed. 1, Piso 1
Alfragide
2720-093 Amadora
Portugal
* Marek.Hajduczenia@siemens.com
http://marekhaj.easyisp.pl/index.php
(+351.21.416.7472  4+351.21.424.2082

 

 


From: David Piehler [mailto:dpiehler@ALPHION.COM]
Sent: quarta-feira, 10 de Janeiro de 2007 21:37
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] [Channel link model ad-hoc]

Dear Marek and Tsutomu:

 

Here is the reference:

 

ECOC 2006, paper We4.5.7:

Analysis on SRS-Induced Performance Impairment of Downstream Data Signal in Video Overlay EPON Systems

A. Agata | A. Murakami | Y. Horiuchi

KDDI R&D Laboratories Inc., Saitama , Japan

 

To quote (and add to) the conclusion from the paper:

 

…we have found that the impact of AC crosstalk (from SRS)  on downstream data signal (at 1490 nm)  exceeds 0.1dB on average only when these conditions as follows are satisfied simultaneously: (1) PRF  (optical launch power at 1550 nm) becomes ~18dBm or more, (2) RF video signal contains VHF component  (that is operates in the ~<100 MHz region), and (3) OMI (optical modulation index) of the VHF component in the optical RF signal becomes more than ~10%.

 

 

The authors show that cross talk from a discrete RF tone in the video domain onto a baseband GbE signal, can result in a power penalty.  This is in addition to any loss in optical power suffered by GbE signal due to SRS.

 

In the case of 10GEPON:

 

(1) The crosstalk would be 1545(?)nm <-> 1550nm XPM instead of 1490(?)nm <-> 1550nm SRS.  The relative magnitude of these effects can be seen in M. R. Phillips, et al., J. Lightwave Technol., vol.17,

(1999), pp1782-1792.

 

(2) Instead of a ~ 10% RF subcarrier at ~ 50 – 100 MHz, the RF “tones” (typically two)  are in the 2 – 10 GHz range.  The tones are chosen such that f1-f2, f1-2*f2, 2*f1-f2, f1-2*f2, etc lie outside of the CATV signal band (50-870 MHZ).  Typically the frequencies of the tones may be in the 2 and 6 GHz ranges.  The tones can result from phase and or intensity modulation of the source laser, and the resulting RF power is usually much higher than any one CATV RF sub carrier.  (Laser chirp + fiber dispersion rapidly convert pure phase modulation into intensity modulation)

 

I have not carried out experiment or attempted to calculate this effect.  I will be in Monterey next week and will be happy to discuss this topic over a beer. 

 

 

 

David Piehler

Alphion

mobile: +1 732 692 4581


From: Hajduczenia, Marek [mailto:marek.hajduczenia@SIEMENS.COM]
Sent: Wednesday, January 10, 2007 12:18 AM
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] [Channel link model ad-hoc]

 

Dear David,

 

Would You be so kind and provide the complete reference for the said paper ? I am also curious but a broad search is lengthy - I am sure we could benefit from this publication if we could identify which one it is ..

 

Regarding the issue number 2: what we are interested in is the impact of the video overlay signal on the digital transmission rather than trying to model the RF channel transmission in the 1550 nm band. The video overlay is not an obligatory part of the standard and as far as I know, current IEEE 802.3 does not specify the parameters of the video overlay apart from the statement that the 1550 nm band is reserved for other services. As such, I believe that the 10G standard should go the same way and focus on the EPON PHY rather than on the RF overlay specifications, which are outside of the scope of our PAR.

 

Thank You for Your comments. I would appreciate if You had a look at the modelling of the SBS and SRS once the updates spreadsheet is released and shared Your thoughts with us ...

 

Best wishes

Marek Hajduczenia (141238)
SIEMENS Networks S.A. - IC COM D1 R
Rua Irmãos Siemens, 1
Ed. 1, Piso 1
Alfragide
2720-093 Amadora
Portugal
* Marek.Hajduczenia@siemens.com
http://marekhaj.easyisp.pl/index.php
(+351.21.416.7472  4+351.21.424.2082

 

 

 


From: TATSUTA [mailto:tatsuta@ANSL.NTT.CO.JP]
Sent: quarta-feira, 10 de Janeiro de 2007 2:42
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] [Channel link model ad-hoc]

Dear David,

(1)     I would not be so quick to dismiss XPM.
PM gets stronger as delta lambda gets small, and at higher RF frequencies. t’s not well known but most video signals have or less discrete, high modulation index, phase and/or intensity tones in the 2 10 GHz range. hese are used for SBS suppression, and they vary significantly among vendors.I have not done any calculations, but there was a paper at ECOC this year from KDDI that showed that cross-talk from RF carriers could add a penalty aside from power loss.

I looked for a paper you suggested, but I did not find it.
Anyway, if it is true, I agree with your comment.

(2)     Is there any need to define the acceptable degradation on the video signal?This is perhaps the most significant issue with the B/GPON RF video overlay.I am not concerned with power loss (or gain) but rather the effective increase in RIN (mediated by either XPM or SRS), or by discrete RF interference from idle fames again mediated by XPM and/or SRS.(Incidentally would these idle frames be in phase if you had several 10G DWDM signals originating from the same switch/router?

I do not think it is an issue of the channel link model, because we can not put it on the spread sheet and also it is an implementation issue of RF-ONU.
However, if we can consider it when we decide the output power level of 10GE-OLT and wavelength allocation of a downstream signal, it is good for us.

Sincerely yours,
Tsutomu TATSUTA


At 02:02 07/01/10, David Piehler wrote:

Just a few comments
 
(1)
     I would not be so quick to dismiss XPM. XPM gets stronger as delta lambda gets small, and at higher RF frequencies. It’s not well known but most video signals have or less discrete, high modulation index, phase and/or intensity tones in the 2 10 GHz range. These are used for SBS suppression, and they vary significantly among vendors. I have not done any calculations, but there was a paper at ECOC this year from KDDI that showed that cross-talk from RF carriers could add a penalty aside from power loss.
 
(2)
     Is there any need to define the acceptable degradation on the video signal? This is perhaps the most significant issue with the B/GPON RF video overlay. I am not concerned with power loss (or gain) but rather the effective increase in RIN (mediated by either XPM or SRS), or by discrete RF interference from idle fames again mediated by XPM and/or SRS. (Incidentally would these idle frames be in phase if you had several 10G DWDM signals originating from the same switch/router?
 
I am not trying to open a big can of worms here, and my feeling is that these effects are probably negligible, although measurable.
 
 
 
David Piehler
Alphion
mobile: +1 732 692 4581


From: Hajduczenia, Marek [mailto:marek.hajduczenia@SIEMENS.COM]
Sent: Tuesday, January 09, 2007 3:56 AM
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] [Channel link model ad-hoc]
 
Dear Tatsuta-san,
Thank You very much for Your contribution. I found it useful and if there is no problem with it, I would like to reuse part of Your conclusions in the introductory section of the presentation on the ad-hoc activities.
I agree with Your conclusions and I believe that the updated Excel spreadsheet will contain SBS and SRS power penalties - SRS is completed while SBS is still under examination to achieve 0 - order approximation with reasonable quality.
 
The presentation will be distributed today and I will ask for the potential supporters of the conclusions and implementation.
The updated Excel spreadsheet will also be distributed in due time on the reflector ...
 
Best wishes
 

Marek Hajduczenia (141238)
SIEMENS Networks S.A. - IC COM D1 R
Rua Irmos Siemens, 1
Ed. 1, Piso 1
Alfragide
2720-093 Amadora
Portugal
* Marek.Hajduczenia@siemens.com
http://marekhaj.easyisp.pl/index.php
(+351.21.416.7472  4+351.21.424.2082
 
 
 


From: TATSUTA [mailto:tatsuta@ansl.ntt.co.jp]
Sent: ter
a-feira, 9 de Janeiro de 2007 6:32
To: Hajduczenia, Marek; STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] [Channel link model ad-hoc]
Dear Marek and all,

I summarized the influence of RF signal as attached.
My conclusion is following;

1. SRS and SBS should be studied.
2. XPM does not need to be considered, if a wavelength of 10G-EPON downstream is separated in 5nm or more from RF signal wavelength.
3. The other items do not need to be considered, unless S/X value of ONU input point is specified in the standard body (I do not think we specify it.). I think S/X is implementation issue.

Sincerely yours,
Tsutomu TATSUTA


At 21:41 07/01/04, Hajduczenia, Marek wrote:

Dear Sergey,
I would appreciate any help You can provide me with. I have some formulas to estimate the impact of the video channel overlay on the downstream and upstream channels but I have a very hard time trying to figure out how to calculate / estimate the values of individual parameters.
Do You have any idea how to apply the formulas with the set of parameters that we have in the Excel spreadsheet?
Best wishes
 
Marek Hajduczenia (141238)
SIEMENS Networks S.A. - IC
COM D1 R
Rua Irms Siemens, 1
Ed. 1, Piso 1
Alfragide
2720-093 Amadora
Portugal
* Marek.Hajduczenia@siemens.com
http://marekhaj.easyisp.pl/index.php

(+351.21.416.7472  4+351.21.424.2082<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

 
 


From: Ten, Sergey Y [mailto:TenS@CORNING.COM]
Sent: quarta-feira, 3 de Janeiro de 2007 19:32
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] [Channel link model ad-hoc]

Marek,

The impact of the RF overlay may be modeled (in zero approxiamtion) as increased loss at data signal wavelength since Raman scattering will transfer energy from digital signal to video signal.

In the next order approximation one has to take into account transfer of the modulation.

I can help you with taking into account zero approximation.

SergeyFax  +1 607 974 4354

-----Original Message-----
From: Hajduczenia, Marek [mailto:marek.hajduczenia@SIEMENS.COM]
Sent: Wednesday, January 03, 2007 11:14 AM
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: [8023-10GEPON] [Channel link model ad-hoc]

 

Dear all,

since there was no activity in the said ad-hoc for some time, I decided to run the first update of the channel link model spreadsheet (please find it attached in a zipped form - 10GEPON-D.0.0.zip). The main changes included in this release are as follows:

  • the examined target range is always limited from L_start to Target reach, with the granularity depending on the Target reach value. The formula used to calculate the individual steps in column [A18:A34] is defined as follows: =(($L$3-$L$4)/16*(ROW($A18)-ROW($A$18))+$L$4)
  • the resulting chart for power penalties versus distance has now the auto ranges for X and Y /top value/ enabled - this way the chart auto adjusts to the selected range
  • L_step was removed since the granularity is auto adjusted ...
  • the "Add Ins Loss" parameter is now calculated based on the split count for the PSC module used in the EPON system. Cell L5 contains the split count (Split_count parameter). The cell L8 value is calculated using the following formula: =10*LOG(L5)+$AM$121*LN($L$5)+$AN$121, where the theoretical loss for N way splitter are summed with the approximated excess loss, based on the approximation curves estimated for typical, commercially available splitters (Cell range [Y69:AL130]). The curve coefficients for average expected splitter parameters are as follows: $AM$121 (A=0.5636) and $AN$121 (B=0.3979) with the approximation curve of logarithmatic type: A*ln(N)+B


To be added, if required: changing the average splitter type into worst/best case scenario values - preferably with the switch variables defined in the spreadsheet (may be tough to find some space for that though :-9)

 

As for the other parameters in the spreadsheet: as Dawe suggested previously, we should probably examine the "TP4 offset cell" (T8) which currently follows "Tx mask top" (G14) in similar style to 10GEPBud3_1_16a.xls. Dawe thinks that this cell needs a thorough review and we should adjust its value in such a way that it compensates for the upstream channel burst mode transmission.

 

Additionally, we should try to numerate the impact of the optional video overlay channel on the overall power budget. Is anyone willing to help me with that ?

 

Thank You for Your time and attention

 

Best wishes

Marek Hajduczenia (141238)
SIEMENS Networks S.A. - IC COM D1 R
Rua Irms Siemens, 1
Ed. 1, Piso 1
Alfragide
2720-093 Amadora
Portugal
* Marek.Hajduczenia@siemens.com
http://marekhaj.easyisp.pl/index.php
(+351.21.416.7472  4+351.21.424.2082