Mike and Marek,
The latest IEEE link budget methodology
that Mike stated is optimal for transceiver makers because it distills
a collection of separate transmitter tests into a single test parameter
called TDP. It also permits maximal trade-off between all the parameters
that contribute to TDP.
While these are positive attributes
to this approach, it is not without its down sides. By lumping parameters
together, the standard no longer needs to specify limits on the underlying
constituents. This removes the boundary conditions on the constituents
that if specified would allow optimizing solutions to address various cases
not covered within the scope of the standard. These cases include
variations in topologies and media types not included in the standard.
It is my opinion that allowing this kind of flexibility is advantageous
because it permits optimal deployment in a greater variety of situations.
Please consider these trade-offs as
the specifications are written into the standard. One way to address
both sides is to use TDP as the normative metric, and provide informative
values on the limits of the constituent parameters consistent with the
TDP spec.
Regards,
Paul Kolesar
CommScope Enterprise® Solutions
1300 East Lookout Drive
Richardson, TX 75082
Phone: 972.792.3155
Fax: 972.792.3111
eMail: pkolesar@commscope.com
Please respond to
"Hajduczenia, Marek" <marek.hajduczenia@SIEMENS.COM>
To
STDS-802-3-10GEPON@listserv.ieee.org
cc
Subject
Re: [8023-10GEPON] Ad hoc on 10GEPON
channel model - definitions
Dear Mike,
The FC-PI-2 and FC-PI-4 are inherently the same in terms of the definitions
I quoted but I will change the reference for the sake of using only stabilized
and released specifications. Thank You for that comment.
Regarding the receiver sensitivity, there are two definitions provided.
Number one is quoted after FC-PI-2 and says "The minimum acceptable
value of re-ceived signal at point γR to achieve a BER < 10-12.".
The alternative is quoted after T1.523-2001 and says "The minimum
value of average received power to achieve a 1 x 10–10 bit error ratio.
It includes power penalties caused by use of a transmitter under standard
operating conditions with worst-case values of extinction ratio, pulse
rise and fall times, optical return loss, receiver connector degradations,
and measure-ment tolerances." Is that clear right now ? Sorry for
the wording in the original document. I will try to clarify that ...
Thank You also for the suggestion on the power budget estimation based
on OMA - it was also suggested by Piers to have a look in that direction
and that is what I am doing right now. I am afraid that I am still a little
bit confused with the spreadsheet but I will figure it out with time. I
will have to look into which terms require proper definitions in order
to have the power budget defined in the way You mentioned before. I will
try to have a look at it today and come back to You tomorrow.
Thank You for Your time
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
-----Original Message-----
From: Mike Dudek [mailto:mike.dudek@PICOLIGHT.COM]
Sent: quarta-feira, 13 de Dezembro de 2006 16:43
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] Ad hoc on 10GEPON channel model - definitions
A few comments on these definitions.
I would caution the use of FC-PI-4. This is a draft standard that
is in development, and therefore is likely to contain errors. That
said the definitions are unlikely to change and are mostly copied (maybe
all copied) from the released document FC-PI-2. At this point however
I would recommend that FC-PI-2 is used.
I know that this is not easy to incorporate into a matrix, but it is fairly
common practice in the datacom world to create the exact definitions of
what is meant by a particular term through the use of a test procedure.
Eg. Transmitter and Dispersion penalty (TDP) for the 1G EPON
clause 60 (which is in 802.3 2005) is defined by reference to the test
procedure of clause 58.8.9.
As you noted there are significant differences between link penalties (IEEE)
and Optical Path penalty (ITU) and lumping them into the same row in the
matrix may add to confusion. However as these use different names
it is not that confusing. More of a problem is "receiver sensitivity"
AKA "intrinsic sensitivity". As normally used in the IEEE
and Fibre Channel context (including it's use in the various link budget
spreadsheets) this is the performance of the receiver when tested with
a perfect transmitter. I don't see the definition you have provided
in the matrix stated as coming from FC-PI-4 in FC-PI-4. I would be
grateful if you would provide the exact page reference for it, because
in the FC-PI-4 document this would be an incorrect definition and needs
to be corrected if it exists. The definition as stated I think is
more like how ITU defines sensitivity including the effect of transmitter
impairments, (the definition may have come from FDDI which is a much older
A!
NSI standard).
Note that in my opinion the budgeting method used in 802.3 Clause 52 (10G
serial) is probably the best method to achieve plug and play operation
for a system which has a challenging budget to meet.
In summary this method is as follows.
The budget is based on Optical Modulation Amplitude at both the Tx and
Rx, (So variations in extinction ratio don't affect the budget).
The difference between the sensitivity measured with a reference receiver
between a near perfect Tx with a short fiber, and the device under test
Tx with a long fiber with a worst case dispersion characteristic is measured
and called Transmitter and Dispersion penalty (TDP). (hence
TDP includes all Transmitter and Dispersion penalties relative to a near
perfect Tx.).
The Receiver is tested with a signal that emulates the worst case signal
at the end of a worst case dispersive fiber. (ie emulating a Tx
with the worst case Transmitter and Dispersion penalty.). This is
the Receiver's Stressed receiver sensitivity.
The link budget closes (ie system is OK) provided the loss in the link
is less than "minimum Tx OMA -TDP" - Rx stressed receiver
sensitivity. Note that the key Tx parameteter specified is
the (Tx OMA - TDP).
-----Original Message-----
From: Hajduczenia, Marek [mailto:marek.hajduczenia@SIEMENS.COM]
Sent: Wednesday, December 13, 2006 5:39 AM
To: STDS-802-3-10GEPON@LISTSERV.IEEE.ORG
Subject: Re: [8023-10GEPON] Ad hoc on 10GEPON channel model - definitions
Dear Frank,
Attached please find an updated matrix list. It contains several new definitions,
which - being unable to find them in the 802.3, I went to ANSI standards
to find the respective terms defined. Some of them are still missing and
I am trying to find some comprehensive and yet recognized defnitions for
such parameters.
I also added a wish list, which would probably need to be added if we want
to go into details with the channel impairments and penalties.
Feel free to modify/add/change ...
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
-----Original Message-----
From: EffenbergerFrank 73695 [mailto:feffenberger@HUAWEI.COM]
Sent: quarta-feira, 13 de Dezembro de 2006 8:09
To: STDS-802-3-10GEPON@listserv.ieee.org
Subject: Re: [8023-10GEPON] Ad hoc on 10GEPON channel model - definitions
> All,
>
> Marek offers a good starting point of a definition cross-reference
> matrix.
> I have taken this, and filled in the ITU side.
>
> We should probably make more items in this matrix, to include at
> least the Tx min and max, and Rx sensitivity and overload. Those
> are somewhat different between the two systems, too.
>
> Regards,
> Frank E
>
>