re: Tellium Gets Some Competition > 64 OC-192's in 1/2 a rack is physically impossible
I wouldn't say that so certianly. Just a couple of years ago I worked on OC-48 equipment that required a whole rack. Now most of that functionality is in a chip, and packaged in a single slot easily.
I have no direct connection with them, but White Rock Networks has announced an OC-192 ADM in 2U's. It looks like 4 interfaces. That means that 64 of their interfaces would fit in 32U's. Not that far off from a "half-rack"
re: Tellium Gets Some CompetitionThat's right. This isn't a grooming switch. It can't do sub-rate multiplexing.
That's why I pitched it as similar to Tellium in that it handles whole wavelengths and has an electrical core - and dissimilar to Tellium in switching analog rather than digital signals.
The business of converting digital signals into analog ones is more complicated than I made it sound in the article. It isn't just a case of using a codec lke the ones used to translate analog voice signals into strings of bits and vice versa. It has something to do with using Fourier transforms, which is beyond my understanding.
re: Tellium Gets Some CompetitionIn response to the question below aboput O/E and E/O converters, it seems they have 3R wavelength converting transponders on the input and output of every channel. This gives them the ability to do PM and 3R regeneration on incoming and outgoing wavelengths. Unfortunately, inlike an all-optical matrix, they don't have the option of allowing some channels to pass through via OOO (ie without transponders) if power levels allow for it. I guess their pitch is that this is more scalable and/or cheaper than digital matrices such as Tellium's.
Detect an incoming photon stream convert to some sort of modulated RF signal (error coding?), send it to a GAS based RF switching matrix, send it back out to a Lambda modulator. Would have develop some power dividers, multi octave will be tricky bit but can be done. Switching speed will be fine, not low ns rate, lots of elements to switch but should in the 100GÇÖs of ns. Gain budget would be somewhat tricky but is mostly book keeping exercise with some SNR trades thrown in.
Yep that might work.
40 Gb, well not sure what state of the art is for GAS switching devices, probably have a tough time getting something with that much bandwidth right now, but lots going on in the cell phone 3G area so chips are becoming available.
re: Tellium Gets Some CompetitionIf the alarm bells aren't ringing yet, they soon will be:
(From the Teraburst web site)
Christopher Weller Vice President, Network Design
Chris brings extensive technological and corporate development experience in the fiber-optic telecommunications industry to TeraBurst. He has patents-pending in optical switching and add/drop products and has design expertise in optical network solutions for inter-exchange carriers and global oil companies. Most recently with SilkRoad, a San Diego telecommunications start-up...
re: Tellium Gets Some CompetitionI don't assume it can do grooming.
Probably notGǪwould have to DeMux data stream, sort, then redistribute based on what field?
How could it, if it's millimeter-wave analog? Also, why does it convert from optical->digital->analog->digital->optical? Why not optical->analog->optical?
> 64 OC-192's in 1/2 a rack is physically impossible
I wouldn't say that so certianly. Just a couple of years ago I worked on OC-48 equipment that required a whole rack. Now most of that functionality is in a chip, and packaged in a single slot easily.
I have no direct connection with them, but White Rock Networks has announced an OC-192 ADM in 2U's. It looks like 4 interfaces. That means that 64 of their interfaces would fit in 32U's. Not that far off from a "half-rack"
http://www.whiterocknetworks.c...