Dekun,
Happy New Year, and sorry for my late responses, which I have added in-line below.
Ed
From: Liudekun [mailto:liudekun@xxxxxxxxxx]
Sent: Friday, December 23, 2016 10:52 PM
To: Harstead, Ed (Nokia - US) <ed.harstead@xxxxxxxxx>; STDS-802-3-NGEPON@xxxxxxxxxxxxxxxxx
Subject: RE: 802.3ca (100G-EPON) consensus building agenda for 12/22/16
Ed:
I am sorry that I misunderstood your contribution yesterday. Yes,
proposed sensitivity is indeed derived for the 25G InP APD using conventional NRZ detection. I am OK with this.
I think we should have the main information to proceed if we assume “No improved FEC will be need” and “diplexer loss is same with 10G EPON”, one thing may be missing is the coexistence element with
10G EPON (in your following contribution sent out by today). Of course, again, I hope I didn’t miss something in your contribution(without your presentation).
[Ed Harstead] You are correct that in the downstream direction a “brownfield” OLT module will require an additional filter in the transmit path, as shown below. This does not affect the spec
since it is internal to the OLT, but it will require another ~0.5 dB from the 25G transmitter.
Just for some more clarification,
In your motion, about the following two items (your default assumption in your analysis in the contribution):
1. No improved FEC or 0.5dB improved FEC gain @1E-3
2. diplexer loss is same with 10G EPON
do you mean we make a decision that the above items are adopted as default values or they are only an assumed precondition of “-24.2dBm sensitivity” ?
[Ed Harstead] Probably I could use some help with the wording of the Motion. I do not mean to say that -24.2 dBm should be the final specification, but that we set it as a “baseline” for determining
the specification, pending decisions on FEC and DS/US gap, which we can now focus on.
For the improved FEC, we will try to have a further contribution to Huntington (but may be to whistler meeting due to the time is so urgent this time, there are a lot of holidays before next meeting)
The power budget of 100G PON is so challenging, from me, we should try to gain every potential to alleviate the pressure on optics. Such as the potential improved FEC capability.
[Ed Harstead] Actually I am not sure that the 100G EPON budget is more challenging: we just add SOA amplifiers. I think getting 25G EPON without amplifiers is the bigger challenge, but I think
we can do it, we just need another ~0.5-1 dB.
For the diplexer loss, I will bring one contribution to Huntington. Based on the feedback from the filter vendors , just like the analysis in John’s
contribution, due to the polarization dependency for 45 degree filters(which will run out ~10nm transition band of the filter), it indeed will bring some more extra insertion loss for ~40nm DS/US gap. There can be some way to overcome this, such as small
incident angle (13 degree filter) , new thin film design for depolarization, but such way will inevitable bring some extra cost , non-common technology , and bring some risk for volume availability.
So before we make a decision that we use C/S-band for downstream and O band for upstream, we most probably need to take care on this.
[Ed Harstead] Agree.
Best regards
Dekun Liu
____________________________________________________
Advanced Access Technologies Dept.
网络研究接入技术部
Huawei Technologies Co., Ltd.
华为技术有限公司
Phone: +86 027-59267217 Email: liudekun@xxxxxxxxxx
湖北武汉市关山一路光谷软件园A7-9
邮编：430074
Huawei Technologies Co., Ltd.
A7-9 Wuhan Optical Valley Software Park,Guan Shan Road,Wuhan,Hubei, P.R.China
From: Harstead, Ed (Nokia - US) [mailto:ed.harstead@xxxxxxxxx]
Sent: Saturday, December 24, 2016 2:47 AM
To: Liudekun; STDS-802-3-NGEPON@xxxxxxxxxxxxxxxxx
Subject: RE: 802.3ca (100G-EPON) consensus building agenda for 12/22/16
Dekun,
Please read the contribution once more! The proposed sensitivity is derived for the 25G InP APD using conventional NRZ detection. Do you disagree? We included the analysis for EDB as a check to
see that EDB would not be ruled out as an option. As you well know, we have never proposed EDB as a required implementation, only as an option!
The proposed receiver sensitivity is supported by 6 component vendors. In the draft loss budget contribution I sent today, the launch power is based on a value that also comes from input from 6
component vendors. Do you agree that we now have enough information to proceed with decisions? If not, what is missing, and will you bring it to the January meeting?
Ed
From: Liudekun [mailto:liudekun@xxxxxxxxxx]
Sent: Thursday, December 22, 2016 9:48 PM
To: Harstead, Ed (Nokia - US) <ed.harstead@xxxxxxxxx>;
STDS-802-3-NGEPON@xxxxxxxxxxxxxxxxx
Subject: RE: 802.3ca (100G-EPON) consensus building agenda for 12/22/16
Dear Ed:
Merry charismas and Happy new year! Have a good holiday for all of you US friends. ( It’s only for you, I need continue to work on your Christmas
dayL)
Thanks for your good analysis on the 25G receiver sensitivity , it’s a good reference when we make a decision on this.
Due to I didn’t attend the telephone meeting yesterday (the time was too bad for my time zone), here are some of my comments:
If I didn’t understand wrong, you get the -24.21dBm sensitivity from that of 10G PON as a base line(-28.5dBm), and subtract 1dB for APD improvement
, then add 5dB for the gap between 25G EDB detection( based on 10G APD) and 10G NRZ detection (based on 10G APD). While, we haven’t make a decision that we will based on 25G EDB detection(re-using 10G APD) for 25Gb/s receiver. And 5dB gap between 25G EDB
detection and 10G NRZ detection is also only a one-time experimental data rather than a theory principle . This gap still may be improved such as by some equalization algorithm.
From me, it would be better for us to list the capability of transmitter (the power levels of 25G EML/DML ), capability of the 25G APD(including both
InP and GeSi), OPP on different wavelength band , the “real” power budgets on the link(including the insertion loss of the mux/demux, CEX with 10G PON, BOSA filters…) on the table first, and then make a decision on the sensitivity and the power levels.
Anyway, we can continue to discuss in Huntington next month.
Best regards
Dekun Liu
____________________________________________________
Advanced Access Technologies Dept.
网络研究接入技术部
Huawei Technologies Co., Ltd.
华为技术有限公司
Phone: +86 027-59267217 Email: liudekun@xxxxxxxxxx
湖北武汉市关山一路光谷软件园A7-9
邮编：430074
Huawei Technologies Co., Ltd.
A7-9 Wuhan Optical Valley Software Park,Guan Shan Road,Wuhan,Hubei, P.R.China
From: Harstead, Ed (Nokia - US) [mailto:ed.harstead@xxxxxxxxx]
Sent: Thursday, December 22, 2016 10:45 AM
To: STDS-802-3-NGEPON@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_NGEPON] 802.3ca (100G-EPON) consensus building agenda for 12/22/16
All,
Please see the two attachments pertaining to these topics. For those not able to attend the call (for reason of time zone or other), please feel free to provide comments by email.
Ed
From: Curtis Knittle [mailto:C.Knittle@xxxxxxxxxxxxx]
Sent: Wednesday, December 21, 2016 7:54 PM
To: STDS-802-3-NGEPON@xxxxxxxxxxxxxxxxx
Subject: [802.3_NGEPON] 802.3ca (100G-EPON) consensus building agenda for 12/22/16
Dear Colleagues,
We have two agenda items for tomorrow, and possibly a third:
1.
Some theoretical results on APD receiver sensitivity vs. SOA+PIN sensitivity as a function of SOA ASE filter width—this pertains to the performance of 100G OLT receivers in TDM and WDM co-existence cases. (Harstead)
2.
Proposal for 25G ONU and OLT receiver sensitivity specifications. (Harstead)
Please join the meeting at 11:30 am MST.
1. Please join my meeting.
https://global.gotomeeting.com/join/838401493
2. Use your microphone and speakers (VoIP) - a headset is recommended. Or, call in using your telephone.
Dial +1 (571) 317-3122
Access Code: 838-401-493
Audio PIN: Shown after joining the meeting
Meeting ID: 838-401-493
Curtis Knittle
VP Wired Technologies – R&D
CableLabs
desk: +1-303-661-3851
mobile: +1-303-589-6869
Stay up to date with CableLabs: Read the blog and
follow us on Twitter