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

Re: [802.3_EXTND_EPON] November 2011 plenary meeting



Hi Ed,

 

Thank you for the feedback. See some more discussion inline

 

Marek

 

From: Edwin Mallette [mailto:edwin.mallette@xxxxxxxxx]
Sent: 19 October 2011 15:46
To: Marek Hajduczenia
Cc: STDS-802-3-EXTND-EPON@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_EXTND_EPON] November 2011 plenary meeting

 

Marek,

 

I think I'm reasonably aligned with the target objectives.  My comments are in-line.

 

Ed

On Mon, Oct 10, 2011 at 10:09 AM, Marek Hajduczenia <marek.hajduczenia@xxxxxx> wrote:

Dear colleagues,

 

I would like to start discussion on the target objectives for the project. While I understand that we still need some more contributions on individual topics, including operator requirements, here is something that I think might be the feeling in the room we had at the last meeting in Chicago.

 

So here is the first pass at the proposal objectives:

1)      Define symmetric 29dB budget for 1G-EPON by adding downstream power budget and coupling it with PRX30 class specs already in place in 802.3av. We can decide then whether we keep PRX30 specs in Clause 77 or move then to Clause 60. My personal preference would be to do just that.

 

[EJM] This one is definitely low-hanging fruit IMHO and is valuable given the varying optical characteristics of current shipping super-compliant optics (PX20+, PX20++, PX20-E).

 

I'm not clear whether you're proposing to keep the PRX30 specs in Clause 77 or move them to Clause 60.  It seems to make the most sense to muck with as little of the existing spec as possible and if that means to keep the PRX30 specs in Clause 77 (as makes sense to me) then I think we agree.

 

[mh1020] My personal preference would be to keep 1G-EPON specs confined to Clause 60 and keep Clause 75 for 10G-EPON. It does seem like a lot of work, but all we do is move content of one table column to Clause 60, create a pointer back in Clause 75 and add new column for downstream for 1G-EPON. In either case, at this point of time what is important is to agree that this something that we all want to work on, which is what your opinion seems to be. I would certainly like to hear more opinions, if possible.

 

2)      Define 33dB budget for EPON at least 1G-EPON, 10/1G-EPON, and hopefully 10/10G-EPON. Here, contributions on technical feasibility would be much appreciated, especially from optics vendors.

The reason to go with 33dB and not 32dB is that an extra 1:2 split adds about 3.5dB to channel insertion loss over 29dB channel insertion loss supported by PR(X)30 PMDs.

 

[EJM] I agree here as well; I can't imagine that this is not feasible, though I as you would also like additional feasibility contributions from optics vendors.

 

[mh1020] I believe we heard one contribution with concerns on 33dB power budget feasibility for 10G symmetric and another one demonstrating that there was enough margin to do so. Probably the whole discussion will have to get down to better characterization of ODN and its components and trying to tighten margins, though we will have to be very careful to avoid producing a spec which would be too complex, too complicated or too expensive to implement. The strength of EPON was always in balancing cost versus technical complexity and I think we ought to follow that path closely. What do you think?

 

3)      Define power budget extender (PBEx) for power budgets beyond that point.

 

[EJM] As much I would like to keep everything completely passive and reach ~128 splits on a single fiber, I'm a pragmatist and accept that this is a reasonable approach which could enable operators at some point to go even higher than 128 splits.

 

[mh1020] Do you think there are any additional limitations for PBEx devices which we need to take into account like compatibility with other wavelength band plans, target power budget boost etc.?

 

4)      Maintain PR(X)30 ONU definitions for items 2) and 3) above, targeting OLT-side changes only

Please comment.

 

Regards

 

Marek Hajduczenia

 

From: Marek Hajduczenia [mailto:marek.hajduczenia@xxxxxx]
Sent: 03 October 2011 07:48
To: stds-802-3-EXTND-EPON@xxxxxxxxxxxxxxxxx
Subject: November 2011 plenary meeting

 

Dear colleagues

 

This email serves as a starting point for preparation of contributions for the November plenary meeting in Atlanta. I would very much appreciate your feedback and active discussion on the email reflector on individual topics, if possible.

 

We received many good comments on existing contributions, which I hope will facilitate development of updated versions and improve the clarity of messages. In the following days, I will be working with individual contributors from the last meeting to focus on technical and organizational contributions for the following meeting, trying to address requests for clarification and changes as requested in Chicago.

 

Here is the list of target topics under the Call for Contributions for the November plenary meeting:

-          Comparison of cost between existing 1G-EPON and 10G-EPON PMDs and extended EPON PMDs, with the definition of what power budget is considered as “extended”. It would be especially interesting to see what cost increase (in %) would be related with adding X dB to the available power budget.

-          Cost comparison for various implementation scenarios for Extended EPON: compare scenarios of serving a given group of subscribers with multiple lines of existing EPON PMDs versus using fewer lines but using Extended EPON PMDs or using power budget extender (outside device). It would be nice to see when it becomes economically viable to use Extended EPON PMDs or power budget extenders.

-          Technical feasibility of various architecture for power budget extenders, their strengths and weakness, current status of their components, their availability, reliability etc. Any known use cases, lab test results for such devices and their performance data would be most welcome.

-          Market drivers for Extended EPON, with the special focus on:

o   Specific market applications to be covered by Extended EPON that are not covered by EPON today, indicating whether it is targeting extended reach, extended split or both that are needed in the given scenarios. Provide specific data i.e. distance between the OLT and the farthest ONU, number of splits, ability to use active devices in the outside plant (power budget extenders)

o   Size of the market for Extended EPON i.e. for an increased reach application (e.g. with 40 km reach), what % of deployed lines would use the Extended EPON?

o   How many dBs need to be added to existing EPON PMDs to guarantee coverage of 90%, 95% and 99% of the target market for access applications.

 

Please let me know whether I captured everything that was discussed.

Regards
 
Marek Hajduczenia, PhD

Chair, Extended EPON Study Group, http://www.ieee802.org/3/EXTND_EPON/
 
ZTE Portugal
Edifício Amoreiras Plaza,
Rua Carlos Alberto da Mota Pinto, nr. 9 - 6 A,
1070-374 Lisbon, Portugal
 
Office: +351 213 700 090
Fax: + 351 213 813 349
Mobile: +351 961 121 851 (Portugal)

Mobile: +1 571 225 2781 (USA)