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

Re: [802.3_NGBIDI] Wavelength plans



All,

The following Email was sent by Mr. Bottoni, but the reflector rejected it.

I resend it here, so you may see it.

Thank you,

Frank E.

 

 

From: Fabio Bottoni (fbottoni) [mailto:fbottoni@xxxxxxxxx]
Sent: Tuesday, February 26, 2019 10:21 AM
To: frank effenberger <frank.effenberger@xxxxxxxxxx>; STDS-802-3-NGBIDI@xxxxxxxxxxxxxxxxx
Cc: Alessandro Cavaciuti (acavaciu) <acavaciu@xxxxxxxxx>
Subject: RE: Wavelength plans

 

Dear All,

 

This is Fabio Bottoni, HW engineer in the Transceiver Module Group of Cisco. I would like to provide Cisco opinion on this point.

 

·       10G BiDi: we strongly support the 1270/1330nm wavelength plan. Rationale is ensuring compatibility with most of the 10G BiDi pluggables, already available at 1270/1330nm.

·       25G BiDi:

o   We do support the statement that dual-rate 10/25G optic would be an interesting feature to be implemented.

o   At the same time, we are aware that DML-based solutions @1330nm, above 20km, can be affected by significant penalties.

§  We made some naïve experiments, where the transmission penalty was about 3dB, at 40km

§  and we didn’t even explore worst case conditions: such as temperature corners, lambda > 1330nm,…

o   Given that, we would be particularly interesting in getting a feedback, from main suppliers, about this topic:

§  EML-based Tx @1330nm (or DML-based with chirp properly optimized), can be a viable solution for 25G links up to 40km?

§  If answer is yes, we would strongly support 1270/1330nm plan, in this case as well.

 

Regards,

Fabio

 

 

From: frank effenberger <frank.effenberger@xxxxxxxxxx>
Sent: martedì 26 febbraio 2019 14:53
To: STDS-802-3-NGBIDI@xxxxxxxxxxxxxxxxx
Subject: [802.3_NGBIDI] Wavelength plans

 

All,

I wanted to make some observations regarding the possible wavelength plans for the bidirectional optical PHYs.

My intention is to strike up a conversation on these issues, so that we can gain a better understanding.  

(Everything I say below is just my personal observation, and has no special status.  If you have another idea, please tell us!) 

 

For 10G, it seems that there is wide support for using 1331 nm downstream, 1271 nm upstream, using the typical CWDM bandwidths, and assuming uncooled directly modulated lasers (DMLs).  Additionally, for 10G there is no dispersion reason why we could not use these wavelengths for all distances. 

Since keeping the same plan has its advantages, it would seem that is a good solution.  Question for the group: Do you disagree?  Talk about it. 

 

For the higher speeds, the dispersion penalty associated with a directly modulated laser starts to be an issue.

For 25G, we can note that the 25GBASE-ER uses the band 1295-1310 nm, while the LR goes 1295-1325 nm.  

These are fairly wide bands, and so again uncooled or semi-cooled DMLs might be a possibility here. 

We could look to reusing the 1295-1310 band for the downstream, while using the 1271nm CWDM band for the upstream. 

There are two complications:

1)       For the 10km distance, we are not so limited.  In fact, we can even use the 1331nm assignment.  This is nice in that it allows a dual-rate 10/25G optic to be implemented. 

2)       The guard band for the diplex filter is getting pretty narrow (~20nm).  This makes 45 degree diplex filters impractical, and would push us into a more costly packing scheme.

An alternative solution to the dispersion problem is to use externally modulated lasers (EMLs).  An EML can easily operate all the way up to 1360 nm, and so we could continue to use the 1331 / 1271 nm channel assignments, and we could have dual-rate possibilities. 

Question for the group: What do you think?  Talk about it. 

 

For 50G, at our last meeting we saw the proposal to use the 1331 / 1271 nm scheme for the 10km link, and to use 1306~1310 / 1292~1296 nm for 40 km.  I would observe that since we expect to use PAM4, from a dispersion perspective and from a bandwidth perspective, we could reuse whatever wavelength plan we choose for the 25G optics.  This could again have some commonality advantages; however, it should also be said that the 50G optics are going to need higher output power given the same receiver, so it might not be 100% common.  

Question for the group: What do you think?  Talk about it. 

 

Thanks for your attention.

 

Sincerely, 

Frank Effenberger, VP fixed access networks lab, Futurewei Technologies

Chair IEEE P802.3cp Bidirectional 10, 25, and 50 Gb/s Optical Access PHYs Task Force

Rapporteur ITU-T Q2/15 Optical Access Networks Question

For urgent matters, contact sophie.johnson1@xxxxxxxxxx

 


To unsubscribe from the STDS-802-3-NGBIDI list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-NGBIDI&A=1


To unsubscribe from the STDS-802-3-NGBIDI list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-NGBIDI&A=1