Briefing.com Photonics Tutorial, Part the Third
tekboy/Ares@forpartsone&two,see25229and25676above.com
Photonic Revolution III: Optical Switches
[BRIEFING.COM - Gregory A. Jones] In part II of our series on the optical networking sector, we focussed on DWDM -- the technology that has made possible a bandwidth explosion at the core of the telecommunications network. But revisiting our transportation network analogy in part I, an explosion in core bandwidth without any improvement in routing that traffic is like increasing the traffic on Interstate 80 by a factor of 40 without adding any capacity at the on and off-ramps -- you would end up with most of the cars in the ocean off of New York and California.
Being able to deliver huge amounts of bandwidth on a strand of fiber is great, but it's not worth much if you can't direct the traffic. That's where optical switches -- sometimes called optical cross-connects -- come into play. The promise of the optical switch is that it will be capable of directing a tremendous amount of bandwidth over fiber networks. We say promise, because at this point there is a lot more promise than reality.
Note: Briefing.com subscribers can access parts I and II of the series via the links below. If you are not a subscriber, please take a free trial at www.briefing.com so you can access the Stock Briefs' Archives.
Photonic Revolution I: The Fiber Landscape
Photonic Revolution II: DWDM
The OEO Dilemma
Some hybrid optical-electrical-optical (OEO) switches are being shipped today, most notably by Ciena (CIEN), Sycamore (SCMR), and privately held Tellium. While these hybrid switches do offer cost savings for telecom carriers today, they are viewed as merely a stepping stone on the path to the all-optical network.
Because they convert signals from optical to electrical to be switched and then back to optical, they are still subject to the speed limitations of electrical switching. They are not expected to be capable of handling the explosion in bandwidth still to come with OC-768 (40 gigabit, or four times the current maximum) fiber and additional channels of DWDM. For that reason, most OEO switches are built with the capability to be transitioned to all-optical switches when the technology catches up. Ultimately, though, there is little doubt that all-optical switches will win in this market.
The MEMS Makers
All-optical switches are still more prevalent in white papers than in telco POPs (points of presence). But despite being in its early stages, optical switch technology is tending to focus on a preferred approach: MEMS. Micro-electro-mechanical systems is just a fancy name for tiny mirrors. These mirrors can be repositioned in fractions of a second to redirect light signals from one wavelength onto another.
Obviously, these mirrors are not gleaning any information from the photons, they are simply redirecting their path based on software instructions. This is what makes an optical switch far less intelligent than an electrical router, which can actually look inside a packet of data to read information that helps deliver that packet to its final destination.
But even without that intelligence, an all-optical switch will be of tremendous value to telco carriers who are struggling to keep up with bandwidth demand. Relative to current technology such as digital cross-connects and optical add/drop multiplexers, optical switches hold the promise of not only massive bandwidth capabilities, but also fast provisioning times (seconds as opposed to weeks).
The MEMS companies detailed in the table below offer many different solutions in terms of specifics, but they share the vision of switching optical traffic with mirrors. (Companies without tickers are still private; the "Unit" column indicates the recent acquisitions that have facilitated public companies' entry into the optical switch market.)
Company Ticker Unit
Optical-Electrical-Optical (OEO)
Brightlink Networks ---
Ciena CIEN Lightera
Cisco Systems CSCO Monterey
Sycamore Networks SCMR
Tellium ---
All Optical Micro-Electro-Mechanical Systems (MEMS)
Astarte Fiber Networks ---
Calient Networks ---
Corvis CORV*
C Speed ---
Lucent Tech LU
Nortel Networks NT Xros
Optical Micro Machines ---
Siemens AG ---
All Optical InkJet
Agilent A
All Optical Router
ilotron ---
* Corvis has filed for its IPO.
Blowing Bubbles
One seemingly odd entry on the table above is the InkJet section occupied by Agilent. The HWP spin-off is indeed in the optical switch business by virtue of its old InkJet technology. As light travels through a narrow opening, the bubbles in the InkJet technology are used to redirect the signal onto a new wavelength. This is an entirely different and in some way more elegant solution than MEMS, but there is one big question mark: scalability. Carriers ultimately want thousands of port on each switch, and while MEMS offer that promise, it's still too early to say if bubbles do.
Holy Grail: Optical Router
The optical router may or may not be the future. Carriers want it to be the future. But unlike all-optical switches, which are already in the beta-test phase, all-optical routers are a figment of the imagination at this point. We could only find one private company -- ilotron -- that even claims to be working on an optical router. Most feel that the technology that will peer into a light signal and read information in that signal is still many years off. But when and if it happens, it will render OEO, MEMS, bubbles, and any other switches obsolete. But from an investment standpoint, it's a bit early to worry about this technology.
End-to-End
As we have already noted in this series, a key to success in the optical networking sector is providing an end-to-end solution. From an investment perspective, it is therefore critical to either buy a company with such a solution, or a company whose technology will need to be acquired by a company seeking that solution. Nortel is an example of the former, while its recent optical switch acquisition Xros is an example of the latter.
Having looked at DWDM and now optical switches, the leading horizontal players are becoming clear -- Cisco, Nortel, and Lucent. Trying to gain entry into that group are Alcatel (ALA), Ciena, and Sycamore, though as we will see, these companies do not yet have the product portfolio of the leaders.
Then there are the privately-held companies that seek to be the next Xros, Cerent, or Chromatis. Some will be coming public soon, such as Corvis in the switching and long haul transmission space. To make solid investment candidates, these single-solution companies had better have a technology advantage that a larger player will want to acquire. Most of them will have difficulty surviving on their own against the horizontal giants. We'll continue to investigate these issues as the series progresses.
Parting Shot
Hats off to Light Reading, the optical networking information site noted in part II for its prediction that Sycamore was on the prowl for an optical access company. Shortly after that mention, Sycamore acquired one of the candidates, Sirocco Systems. With that success in mind, we'll note that Light Reading also wrote that Sycamore might acquire Tenor Networks, which makes so-called provisioning equipment, which can translate optical bandwidth into actual revenue-producing services. Tenor is a competitor of ONI Systems (ONIS), the latest hot optical IPO.
Greg Jones - gjones@briefing.com |
