Study HighLights Long-Haul Differences

An exercise that highlights different vendors’ approaches to building long-haul optical backbones is analyzed in a Light Reading report published today: Long-Haul WDM: RFI Exercise. In the exercise, four equipment manufacturers submitted proposals addressing a simplified, fictitious RFI (request for information) for typical requirements for a portion of a pan-European DWDM (dense wavelength-division multiplexing) backbone. Two vendors proposed an ambitious leap in technology -- migrating to 40 Gbit/s in the second year and adopting all-optical switches. The other two were more conservative, sticking to 10-Gbit/s transmission speeds and electrical core switches.

The RFI was created by Stuart Barnes, entrepreneur-in-residence at Atlas Venture; the vendors’ proposals were presented at Lightspeed Europe, Light Reading’s conference in London, last December. Since then, Barnes has analyzed the submissions to cast light on some key issues that carriers are likely to face when they get back to building out their long-haul infrastructures.

The submissions came from the following vendors:

The fictitious "Euro 1" network contained five nodes in a ring configuration with link distances ranging from 300 to 600 km and traffic requirements evolving from tens of Gbit/s up to hundreds of Gbit/s in the third year of operation. Two nodes were specifically set up as key cross-connection nodes with very high capacity requirements to connect Euro 1 to the rest of a larger pan-European network. In terms of line rates, two very distinct camps arose. Ciena and Fujitsu/Tellium proposed 10-Gbit/s line rates throughout the network to satisfy even the very high capacity requirements in year three -- Ciena using more densely packed wavelengths, Fujitsu/Tellium requiring more fiber to be installed. Lucent and Nortel flew the flag for 40-Gbit/s DWDM, however, with both proposals beginning to make use of the higher line rate in year two. It was a similar story with the add/drop and crossconnect technologies being proposed. Lucent and Nortel presented designs incorporating the latest in all-optical crossconnects at the high-capacity nodes, whereas Ciena and Fujitsu/Tellium were confident that electrical switching was still the most sensible solution.

For more detail, check out the report itself: Long-Haul WDM: RFI Exercise — Craig Williamson, Associate Editor, Light Reading
Page 1 / 2   >   >>
edgecore 12/4/2012 | 10:48:46 PM
re: Study HighLights Long-Haul Differences With the "PX" going to the glue factory, how do they propse to do this?

flanker 12/4/2012 | 10:48:45 PM
re: Study HighLights Long-Haul Differences
Good point, edgecore. This "bogus RFI" is a little self-serving. In order to justify 40G line rates and OXCs, Stuart must have been projecting more data traffic than the entire continent currently gernerates. If this guy from Atlas were actually to build the network, he'd be lucky to run 5Gbps peak traffic in three years.

edgecore 12/4/2012 | 10:48:44 PM
re: Study HighLights Long-Haul Differences I agree Flanker,

Plus, when you plan an optical network, you usually talk a bit more about the services that will come into play, not a single mention of services...

What exactly is generating all this traffic? OC192 Leased Lines? 10G Catalyst 6500 Ethernet switches in every large enterprise...Storage over IP San for everyone...details please!

flanker 12/4/2012 | 10:48:43 PM
re: Study HighLights Long-Haul Differences
Ok, so I read the report. Fine, it's an *academic* exercise. so why not do a metro comparison where, as edgecore mentioned, you can get into messy service delivery issues?

Now there's a food fight.

I think the biggest thing left unmentioned in the full report (which is pretty good considering it's free to us), is cost.
zweisel 12/4/2012 | 10:48:40 PM
re: Study HighLights Long-Haul Differences flanker, I agree with you. Without any costs, the comparison is totally useless just like most of Light Reading. only comparing technology without costs is what got this industry into the tank in the first place.
flanker 12/4/2012 | 10:48:39 PM
re: Study HighLights Long-Haul Differences Only comparing technology without costs is what got this industry into the tank in the first place.

Bingo! Wonder why all those CEOs with 20 yrs telco or data experience couldn't figure that out. Plus I get nervous when I hear VCs talk about paradigm shifts but they can't read a financial statement.
edgecore 12/4/2012 | 10:48:38 PM
re: Study HighLights Long-Haul Differences I know how to replace the PX, get a MEMS based DMS100, all optical class 5!

dodo 12/4/2012 | 10:48:38 PM
re: Study HighLights Long-Haul Differences As the mantra used to be, It would be interesting to know " what the value proposition" that Marketing will be promoting after the death of PX
jim_smith 12/4/2012 | 10:48:37 PM
re: Study HighLights Long-Haul Differences In a way the results of the study were obvious...

Companies will either peddle what they have, or
they will peddle what their competitors don't

So, Fujitsu and Ciena peddle 10G with OEO switch
and Lucent and Nortel peddle 40G with OOO switch.
PhotonGolf 12/4/2012 | 10:48:35 PM
re: Study HighLights Long-Haul Differences
I'll bet someone on these boards can take a swag at some of these costs.

Anyone wanna take a shot? +/- 25%, right?

Page 1 / 2   >   >>
Sign In