Sorrento Gets ROADM-Like

The reborn Sorrento Networks Corp. trotted out its first Ethernet transport card today, pitching it as a Layer 2-based analogue to the reconfigurable optical add-drop multiplexer (ROADM).

Sorrento, which counts itself a ROADM supplier, isn't saying the OET is going to compete with full-sized, 40-channel ROADMs. But it claims that for smaller cases, where only a handful of wavelengths need switching, its 10-Gigabit Optical Ethernet Transport (OET) card is a viable alternative. (See Sorrento Does Optical Ethernet.)

The goal, Sorrento says, is to let carriers set up point-to-point Gigabit Ethernet services more quickly. Most providers are keen to save money by not setting up Ethernet connections unless they're needed, but increasingly tough competition means that when those connections come into demand, they have to be created fast.

So, Sorrento reasons, why not apply Ethernet aggregation to the task? OET cards, situated on Sorrentos GigaMux transport systems, can send one 10-Gbit/s Ethernet wavelength around a network ring. Then, Gigabit Ethernet connections can be switched on and off to whatever nodes are appropriate, and they can be aggregated into that 10-Gbit/s wavelength.

Yes, a ROADM could do the same kind of thing, albeit at the optical layer rather than Layer 2 -- but the OET is targeting situations where just a handful of these connections are needed. Most ROADMs aren't built for that small of a scale that some providers need, says Jim Nevelle, Sorrento's CEO.

One exception would be Nistica , whose Fledge 3 is a ROADM carrying only eight wavelengths. Nistica has gotten some affirmation in the form of a Finisar partnership and a recent round of funding. (See Finisar Moves Into ROADMs and Nistica's Early Xmas.)

Sorrento isn't talking out loud about plans to make the OET compete with bigger ROADMs. What might be more interesting to the company, Nevelle says, is the integration of other protocols such as the Optical Transport Network (OTN).

Sorrento spun itself back out of Zhone Technologies Inc. (Nasdaq: ZHNE), its one-time acquirer, about a year ago. (See Zhone Sheds Old Nortel, Sorrento Gear.)

— Craig Matsumoto, West Coast Editor, Light Reading

DCITDave 12/5/2012 | 4:14:26 PM
re: Sorrento Gets ROADM-Like Are there a lot of scenarios where providers would need to carry some Ethernet traffic quickly, but where a more thorough upgrade would be impractical?

I don't yet understand what area/scenario this device is aimed at.

DCITDave 12/5/2012 | 4:14:25 PM
re: Sorrento Gets ROADM-Like Ah, yes, the tough economy pitch. Very good.

I guess I'm just skeptical that such a band-aid would lend itself to future equipment sales.

Seems like they'd do well to come up with a more complete solution that includes stuff like the ROADM-like announcement.

Is it just me, or is part of the deal with the new Sorrento is that it's tough to figure out where they fit in. What do they do that ADVA doesn't?
Pete Baldwin 12/5/2012 | 4:14:25 PM
re: Sorrento Gets ROADM-Like Part of the pitch is the assumption that the carrier doesn't want to fork out for a more thorough upgrade. Better (so the theory goes) to slap down one 10-Gbit/s wavelength now, and use it to carry point-to-point GigE links later between arbitrary nodes on the ring whenever the need arises.

The GigEs could be turned up very quickly, in theory, so that would be part of the attraction. But yes, you'd have to set up that 10-Gig wavelength first.
nodak 12/5/2012 | 4:14:24 PM
re: Sorrento Gets ROADM-Like This reminds me of the 4:1 muxes a lot of people built to bring OC-48's onto 192 signals. At each regeneration point you had the option of regenerating the entire 192 or demuxing it and then adding/dropping a 48. Not really sure how this is ROADM like since this is all electrical and now you have to have a 10GigE card at every node you may want to drop at to break the signal apart versus a GigE card. I wonder what the cost differential will be and if the efficiency is worth it?
paolo.franzoi 12/5/2012 | 4:14:23 PM
re: Sorrento Gets ROADM-Like
The only thing I don't understand about this is why not deal with this a different way around. Let's say that the "upgrade" is required not because there is WDM equipment present, but because its NOT present. If there is WDM equipment there, it would seem to be simpler to enable a wavelength and put out a switch.

So, IF we take my base assumption as true - then basically we are talking (very likely) about a 1310nM SONET ring (or SDH ring). Couldn't one just deploy an overlay of an Ethernet Switch using 1550 nM optics and a passive?

Sign In