Wavelength Services Catch On

Announcements this week from leading carriers highlight the growing popularity of wavelength services, in which optical wavelengths are packaged as bandwidth products. But the jury's still out on their potential impact, and service providers are keeping their hopes to themselves.

Global Crossing Ltd. (NYSE: GX) announced this morning that it's launched a worldwide wave service, one that allows carrier customers to tap 2.5- or 10-Gbit/s connectivity among "more than 90 cities in Asia, North America, and Europe" (see Global Crossing Crosses Globe).

Separately, Level 3 Communications Inc. (Nasdaq: LVLT), which also claims international wave services, says it's sealed a pact with America Online Inc., the Internet division of AOL Time Warner Inc. (NYSE: AOL), by which AOL will buy wavelengths in North America, private lines between North America and Europe, and colocation services in six U.S. cities (see Level 3 Wins AOL Contract).

Neither Global Crossing nor Level 3 would provide many details of the announcements at press time. Information was particularly skimpy regarding the Level 3/AOL deal, about which AOL hasn't issued a release at all.

"We prefer to let the services speak for themselves," says America Online spokesman David Theis.

So far, they're mute. Carriers apparently aren't willing to tip their hands in this most competitive -- and distressed -- of markets. "They hate to provide revenue figures of any kind, but especially for new services where adoption is only beginning," says Nick Maynard, senior analyst at The Yankee Group. If carriers can't show more than a handful of customers for wave services, Maynard says, they'd rather say as little as possible.

Wavelength services, in which service providers lease wavelengths of a fiber strand owned by another carrier, offer a money-saving alternative to Sonet and private-line services. They represent a market forecast to grow an order of magnitude over the next five years, reaching a high of nearly $8 billion by 2005, according to the Yankee Group.

Scott Clavenna, president of PointEast Research LLC and director of research at Light Reading, says the savings from wave services are compelling enough to win converts, despite the fact that they don't yet offer the protection schemes or net management of Sonet (see On the Crest of a Wave).

"Wave services help service providers get a bigger bang from their shrinking capex budgets – not only because they're less expensive than Sonet connections, but also because they can be provisioned much more rapidly," Clavenna maintains.

Analysts are careful to point out that wave services won't be a panacea for the current capex downturn. "We've had confirmation of our forecasts, and some even expect growth to be brisker than estimate," Maynard says. "But wavelength services [alone] aren't going to reinvigorate the market."

Still, investors are clearly thinking waves will help carriers whose outlooks have been rickety. In trading today, shares of Level 3 shot up 0.84 (14.74%) to close at 6.54; and shares of Global Crossing rose 0.04 (2.37%) to 1.73.

Maynard says long-haul wave services, such as those offered by Level 3 and Global Crossing, will account for the lion's share of revenues over the next few years. Other long-haul players include America's Fiber Network (AFN), Broadwing Communications Inc. (NYSE: BRW), Genuity Inc. (Nasdaq: GENU), and Williams Communications Group (NYSE: WCG). Qwest Communications International Corp. (NYSE: Q) has pledged offerings too.

Metro carriers with wavelength services are set for growth rates comparable to those for long-haul, albeit on a smaller scale. Providers include Sigma Networks Inc., Sphera Optical Networks Inc., Telseon Inc., and XO Communications Inc. (Nasdaq: XOXO).

— Mary Jander, Senior Editor, Light Reading
<<   <   Page 3 / 4   >   >>
lightmaster 12/4/2012 | 7:30:54 PM
re: Wavelength Services Catch On BOD Analyst,

I have always found CIR, specifically Gassman and Gross, to be much more pragmatic in their forcasting and much less subject to marketing hype.

I personally think its absurd, in this environment, to make predictions that go out 5 years. Most of the component technologies that will be used in 5 years haven't even been invented yet.
optigirl 12/4/2012 | 7:30:52 PM
re: Wavelength Services Catch On Metro carriers with wavelength services are set for growth rates comparable to those for long-haul, albeit on a smaller scale. Providers include Sigma Networks Inc., Sphera Optical Networks Inc., Telseon Inc., and XO Communications Inc. (Nasdaq: XOXO - message board).

If we are expecting the growth of this market to happen based on the SPs mentioned above then these assumptions are full of bunk. I have a suspicion that Yankee has these companies as clients and so they are basing their growth rates on what they are being told. Cynical? Perhaps but if your look at the financial results and records for these carriers I just don't get it. And, since the RBOCs are not going to play in the wholesale services business I just can't see these numbers being accurate.

(BTW Dodo, that was a good paper by CIR about hype based forecasting..... had a great laugh cuz it was dead on!!!)
cable_guy 12/4/2012 | 7:30:52 PM
re: Wavelength Services Catch On "In this case, you have to patch panel your way over to the SP's LH DWDM system from your cage!"

"In the metro, you would still need to have the SP drop a box at the enterprises site."

The latter is a much bigger issue than people realize. Customers do not want to hear the possibility that another CUSTOMER could take down their traffic (or just that their traffic passes through another customer's site, even though this is the case in SONET). If the customer is already in the CO, then you're all set, patch it over the wall to the carriers LH DWDM. Possible in a wholesale model where carriers are often colocated, but fails in the enterprise model. So to efficiently offer a wavelength service to customers in a metro, you need to first aggregate their wavelengths back to the CO - this means a ring topology and OADM's, and this means explaining that the guy down the road might back into his OADM and take down your wavelength. Granted, protection can nearly eliminate this scenario, but a customer who has enough bandwidth demand to lease a wavelength is going to be deadly serious about accountability for the uptime of that wavelength. This is why certain managed service companies pitch private networks - they play on and spread the FUD and paranoia of dependind on a shared DWDM infrastructure.

Nonetheless, I'm convinced that if the price right, wavelength services will with the day for medium businesses. It goes like this:

Fortune 500: private network
Medium: wavelength services
small: SONET, etherent, DSN, etc...

just my 6 cents.

dodo 12/4/2012 | 7:30:50 PM
re: Wavelength Services Catch On Optigirl

Thanks for the pun. Was doing some research on OADM and landed on this paper and thought that it may be a good backup for your comments
edgecore 12/4/2012 | 7:30:50 PM
re: Wavelength Services Catch On OUCH!!!

Never said the business model made sense, but the marketing BS around managed wavelength services are:

-capacity on demand
-short term bandwidth contarcts

Never said I beleived it to be a gold mine!

But, just to stretch my point to the extreme.

If a retail customer has fibre access to their building and the local PoP that serves them has a metro ring that suports OC48 lambdas, could you not drop in small box (i.e. Optera Metro 5100) into their building for a short period of time and have a pt. to pt connection from their building to another box i.e. Optera metro 5200 at the PoP? With even up front notice, could this not be done! The SP would take his box back afterwards, and he charges for the set up, the bw costs and the tear down. No laterals are required and I need another beer....

lightmaster 12/4/2012 | 7:30:48 PM
re: Wavelength Services Catch On edgecore,

Some day the technology will reach the point where retail switched wavelengths make sense, but I'll be retired by then. Technology can always accomplish more that the carriers are able to swallow. Hey, I'm still waiting for frame relay SVCs.
edgecore 12/4/2012 | 7:30:47 PM
re: Wavelength Services Catch On lightmaster,

Sounds like we both had it with conversations about managed wavelength...cheers!

Had JDU, NT and SCMR not entered my life, you and I could both retire tommorow...


Shareholdervalue 12/4/2012 | 7:30:45 PM
re: Wavelength Services Catch On Let's see, Victoria's Secret got famous for crashing the net after they ran a superbowl ad. How much money do you think they will pay for 8 hour of OC-48 just after they run a multi-million dollar campaign? Think about it.

Today, the carriers want them to first give 90 days notice, then keep the darn thing for a year, at a price of 90K per month. (totals over $1M) They'd probably be more willing to pay $200K for 8 hours, which will cover L3s costs just fine. Everything else is gravy. L3 can take it out and use it elsewhere, or leave it there for next time.
trepanne 12/4/2012 | 7:30:44 PM
re: Wavelength Services Catch On

+++Let me get this straight: Level 3 runs a fiber to the victoria secret's location. They install a piece of equipment that supports OC-48 transmission. They only generate revenue on the assets 8 hours out of the two months. Then the assets sit idle since they only need a T3 the rest of the time.+++

no no no. level 3 runs a fiber to colocation space (either their own or neutral) where it is permanently attached & carrying all sorts of traffic from various sources. victoria's secret either parks a server in the colo space, or pays a fee to someone who is parked there.

it's going to be ethernet up to a router with high-speed optical interfaces; the router is the client of the optical WAN.

you could lose the ethernet/router part & swap it for SONET muxes. same deal for the optical Tx gear, though.

no assets are sitting idle, although capacity is sitting idle.

how much network infrastructure do you imagine a company like victoria's secret is really willing to park on their books for stuff like this? how much WAN staff are they willing to feed for the occasional moment they're needed? gimme a break.

all of this is totally separate from their T3 uplink to their ISP, which could be somebody totally different than whoever they've contracted to multicast their temporary show.

welcome to the wonderful world of outsourcing.

Frank 12/4/2012 | 7:30:43 PM
re: Wavelength Services Catch On After reading this article I posted the following message elsewhere. Comments welcome.


It's appealing, granted, to classify these as "wavelength" services, and indeed they do ride over wavelengths. But when the product is wrapped as a 2.5 or 10 Gb/s offering, it's clear that these are not purely "wavelength" deliveries. They are, instead, offerings that are framed electronically. The question, then, is how is a purely optical product defined, and how is it packaged? Hint: a purely optical delivery mechanism is also a purely analog one. Take, for example, that one should be able to pass analog content as well as binary, digitized content over a purely optical medium. Not that one would want to. Instead, one might want to be independent of what the service provider has in mind.

Say, for example, a customer didn't want SONET framing or any of the other popular formats. Instead, the user might want to have the flexibility to send any of a variety of different digital (maybe even analog at times) formats at various data rates at different points in time.

Such a product would require the provider to redefine [optical] channel parameters in such a way as to allow the user to clearly know and understand (in SLA terms) such things as line width, signal attenuation (engineered loss budget), nonlinear effects (including 4-wave mixing), bounds on chromatic and polarization mode dispersions, etc., parameters and performance limits. In effect, this would result in the definition of a conditioned private line (similar to how ancient analog voice grade data lines were categoried, having to meet 14 tariffed parameters), suitable for both analog (sinusoidal) and digital (RZ/NRZ), passage.

Lest we forget, subcarrier modulation (SCM) techniques that are designed to operate over optical line facilities employ digitally encoded analog transmission through the use of phase shift keying (PSK), qaudrature amplitude modulation (QAM), and so on.

The Blocker:

A major problem that would have to be overcome in assuring a purely optical product (service), especially in long haul systems, exists at regeneration stations. Here, o-e-o takes place and the reshaping and signal regeneration occurs in the digital domain (1s and 0s), thus blocking continuously variable (analog) signals, entirely. Not good for SCM schemes, not at all. And I suspect that this is why SCMs and other nontraditional optical systems are limited to short haul deployments in the metro, where regens seldom come into play.

Instead, analog amplification would be required for those lambdas that were being passed as being purely analog wavelengths. But as we know, once amplification occurs more than a half dozen or so times in serial fashion along a span, noise and dispersion levels of various types accrue to the point where the signal is no longer commercially acceptable.


<<   <   Page 3 / 4   >   >>
Sign In