Slashdot Mirror
Pure Optical Network Switches
richi wrote to us about the all optical switch that was announced from Agilent. The primary reason for the coolness factor is that an optical switch means that an optical signal doesn't need to be converted into electric at the switch, then back to optical.
See The Coming of the Fibersphere, a great essay about something that this switch makes possible.
Lucent did this about, what, a year ago? Read here. Nortel has already done this as well. Here's the Nortel link.
Optical switching is really cool, don't get me wrong! The slowest link in any big telco network is the switching equipment. How are you going to resolve Nortel's new 2.4Terabit pipes with OC-3 or OC-12 switching technology? Guess what, not happening.
Have a read about what Lucent and Nortel are doing...very cool stuff.
Working for the (other) man
My reading of the article is that it is a less responsive level of switching. It is the equivalent of a phone exchange or patch panel that allows you to change the connectivity between a set of fibres. Obviously there is some other method of sending information on which pairs to connect together. Also I would guess the mechanism that generates the bubble is electricity based (though I may be wrong).
I don't see that this particular technology could be used to switch at the packet level simply because bubbles cannot be created or removed fast enough (we're talking moving physical matter so information would be lost in transition). Or maybe the system can only switch a thousand packets a second. Big gaps but boy do they move fast.
Optical computers require the switching to be instantaneous and driven by light itself (photoreactive or some such). Using this technology in optical computers would be like using relay switches in electrical computers. However relays at least used to be used in MAUs to connect computers together in a network.
This technology is useful for the infrastructure of optical computing systems but we need the optical equivalent of the valve or transistor (light amplified or switched using only light) to build the actual optical processor.
Gamma Testing - Where testing is extended to the full user community (AKA Shipping the Program)
This is a rather important development for optical networks. Other than the initial energy source and the receptor at the end of the line, this technology may be the last piece required for a fully optical system.
The previous important development in optical transmission networks was the optical amplifier. By doped sections of fiber with a high voltage placed across each section, they were able to create an amplifier. The light waves coming into the doped section of fiber would be amplified.
To briefly describe it, imagine blinking a small flashlight into a window. On the other side of the glass it looks like a car's headlight rather than a small flashlight.
Ok, this amplifier is cool but why is it important. The amplifier allowed companies to move away from light to electric to light. In fiber optic system, the signal decays as it travels through the fiber. Once it reaches a certain level, it will become unusable. Thus, you have to boost the signal periodically. Traditionally, communication companies had to take the decayed signal, convert it to electric, then you drive another laser with the weak electric signal. This then completes the light to electric to light again, and you now have a usable signal again. (Until you have to boost the signal again) Anyway, this conversion process created a whole new set of problems. By converting it to electric, you introduced noise. After a "amplifying" the signal a few times, your signal was no longer clean buy now included some noise.
This optical amplifier was important and lead to the FLAG system. If you read Wired a few years ago, they did a really good job of discussing FLAG. FLAG is the "Fiber Optic Link 'round the Globe" and is a _large_ undersea fiber optic cable reaching from UK, to Egypt, to Malaysia, to Japan. This project and other like it, became feasible because of optical amplifiers.
Optical switches are important for the same reason. It will reduce noise and increase reliability of the system.
I would not look to see these switches replace the current technology (as mentioned in a previous post). However, I would expect to see these switches start to become incorporated into new systems such as subsea communication, pipeline right-of-way bandwidth, or other cross country communication networks. There already is a big market for piping ESPN to parts of the country. Imagine that you could route the information more like a phone system than a gas pipeline. This could create a whole new market.
Well, as a Network Engineering Consultant in a Heavy Nortel Shop, I have had a chance to see the OPTera and Verselar 25000 in action. These devices are Electro/Optical not pure optical and I believe the break through with the Aligent device is that it is purely optical.
The Economics of Website Security
That's not what this is designed for. You'd actually use it for things like this:
If a backbone node goes down, it's not going to go back up a microsecond later. You want to switch _all_ traffic to an alternate route, and then switch it back a few minutes or hours later when the node goes up again.
Think of this switch as acting something like a crossbar bus, connecting pipes point-to-point. Need more bandwidth between point a and point b? Allocate an additional pipe connecting them. Not using all bandwidth? Remove a pipe and allocate it to another pair of servers. Load patterns vary over minutes or hours, not microseconds, so this works fine.
Now, a purely-optical switch that _could_ work on the microsecond or nanosecond level would be very nice; however, a slower switch is still very useful.