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

[8023-10GEPON] Ad-Hoc to 10G EPON TF [High split ratio study] - summary week 1



Title:
Dear all,
Here is a small summary of the current activities in the "High split ratio study" ad hoc.
 
I received participation requests from the following people:
1. Marek Hajduczenia, Siemens Networks S.A., Portugal
2. Wael William Diab, Broadcom
3. Piers Dawe, AvagoTech
4. David Li, Ligent Photonics, Inc.
5. Bin Yeong Yoon, Electronics and Telecommunications Research Institute (ETRI)
6. Tsutomu Tatasuta, NTT
7. Petre Popescu, Astar Original Design Silicon Manufacturer Inc.
8. Keiji Tanaka, KDDI labs
9. Glen Kramer, Teknovus Inc.
10. Mitsunobu Kimura, Hitachi Communication Technologies, Ltd.
11. Dongsoo Lee, Electronics and Telecommunications Research Institute (ETRI)
12. Haim Ben-Amram, PM Sierra
13. Russell Davey, Optical Systems, BT
Should anybody happen to be excluded (and sent me an email) or there is any mistake in the name/company, please let me know.
 
Regarding the tasks that the ad-hoc strives to fulfil, here is the list with the current association valid as for today:
 
- estimation of 1x64 and 1x128 port splitter power loss value, with respect to various manufacturing technologies - almost complete - needs verification only
    Participants: Marek Hajduczenia, Bin Yeong Yoon, Dongsoo Lee
- estimate the maximum launch power into fiber which can be achieved without introducing non-linear effects
- estimate the currently available transceiver technology (maximum launched power at Tx and sensitivity at Rx)
    Participants: Mitsunobu Kimura
- estimate the overall required system power budget for 1x64 and 1x128 port systems (various stage counts, based on results from the first  task)
- estimate the required power margin, based on a little survey margin service providers
- evaluate the 1x64 and 1x128 systems from the carrier's point of view:
    Participants: Russell Davey 
- any other topics of interest that You may identify, subject to confirmation by the ad hoc ...
I do hope that I got the associations right - some tasks were not yet occupied and I would like to ask kindly the people who expressed their will to participate in the as-hoc, to let me know soon as to what they would like to contribute with. There is only one more month to go and we covered only part of the chartered tasks ...
 
As far as the current results are concerned:
 
We have initial results of the evaluation of the power loss for 1x64 and 1x128 port splitters in single and multistage configurations for FBT (standard fibre fused) as well as PLC devices. The spreadsheet with the calculations and all the input data can be found at: http://www.ieee802.org/3/av/working_docs/splitter_loss.xls
All the data included in the spreadsheet is publicly available - the links to websites of individual manufacturers are included in the tab entitled "Companies".
The tab "Splitter comparison" presents a comparison between two available technologies for manufacturing PSC modules, namely FBT and PLC. Since 1x64 and 1x128 FBT as well as 1x128 PLC modules are not available commercially, the values were extrapolated the loss based on lower split devices. The excess loss for individual port counts was estimated by subtracting the average/maximum total insertion loss value (from data sheets) and the theoretical insertion loss for the given port count. Next, the excess loss for each port count was plotted against the port number, approximated with the logarithmic curve and the obtained curve formula was used to estimate the excess loss for 1x64 and 1x128 devices. I would expect these values to be correct down to 0.5 dB, though I have no way of confirming that for now - no manufacturers have/are planning to have such high port count devices.
Based on the values estimated in tab "Splitter comparison", the 1x64 and 1x128 power loss was estimated with various stage numbers and combinations of splitters - see tabs "1x64 splitter" and "1x128 splitter".
Finally, using the respective values for 1x64 and 1x128 splitters, the loss budget was estimated - see tab "Power budget". There are two cases considered, namely 10 and 20 km reach with the fiber loss (including splices) estimated in accordance with the IEEE 802.3 - 2005 values (table 60.1). Connector loss was accounted for - SC type connectors with typical loss of 0.2 dB. The examined PSC modules included various combinations leading to 1x64 and 1x128 port system. Additionally, I includes also the cases of PSC modules spliced into the fiber plant (1st and 3rd column) as well as connectorized splitters (2nd and 4th columns).
I hope that this can be a starting point for other tasks of the study group, namely estimation of the required launch power as well as Rx module sensitivity in the case of 1x64 and 1x128 port systems.
I am looking forward to any feedback from the group
 
PS. Should any materials require publishing, please email me the document and I will publish it on the IEEE website - this way we can skip the email storms :-)

Marek Hajduczenia (141238)
(PhD Student - COM RD1)
SIEMENS Networks S.A. - IC
Rua Irmãos Siemens, 1
Ed. 1, Piso 1
Alfragide
2720-093 Amadora
Portugal
* Marek.Hajduczenia@siemens.com
(+351.21.416.7472  4+351.21.424.2082