Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : JDS Uniphase (JDSU) -- Ignore unavailable to you. Want to Upgrade?

To: quidditch who wrote (722)	8/3/1999 9:53:00 PM
From: Exciton	Read Replies (4) \| Respond to of 24042

Steven: RE: <I had thought that the fiber bragg gratings functioned in a manner that could be compared to "fixed wavelength routers", but you put these two components in distinct categories. If you don't mind, what is the difference?> Sorry, I was a little unclear with my terminology. When I referred to fixed wavelength routers, I should have used the term optical crossconnects. There are several concepts that are getting confused here. I will try to give a better explanation. Fiber bragg gratings and arrayed waveguide gratings are passive devices that are used for multiplexing and demultiplexing wavelengths as you probably already know. One could refer to them as fixed wavelength routers in that the various waveguides have the effect of coupling many wavelengths into one fiber and then separating them out at the other end of a WDM point-to-point link. WDM point-to-point links are the simplest WDM architecture and naturally the first to be commercially implemented. The next stage in WDM evolution--optical networking in high capacity optical backbones--requires two more elements according to the vision that Lucent and others are promoting. These elements are optical add/drop nodes and optical crossconnects. The optical add/drop nodes allow a customer that lies between end points of a WDM point-to-point link to drop one of the wavelengths from the fiber and add that wavelength back to the fiber all in the optical domain. Without optical add/drop nodes, the only way to do this is to demultiplex all of the wavelengths, convert them back to electronic form, demultiplex these time division multiplexed (ie SONET) bit streams, decode the one signal he wants and reverse the whole process to get all of the data back into the optical domain and back on the fiber to its next delivery point. The above process requires large, very expensive digital crossconnects. The optical add/drop nodes eliminate the digital crossconnects. The next stage in the WDM evolution involves optical crossconnects connecting one or several WDM rings and permitting any wavelength to be routed to any of these rings. These systems are beginning to emerge and I believe several carriers are currently testing them. An example would be the Monterey Networks 20000 Optical Crossconnect. The next stage in the evolution and one that is probably still a few years away, is wavelength switching optical crossconnects. The added feature here is that these crossconnects can not only connect any wavelength to any port, they can also change the wavlength that a particular signal is on. Several component advances are required for this and one candidate device is the semiconductor optical amplifier. Prototypes of all optical wavelength switching crossconnects have been demonstrated by Alcatel and others, but these are not yet ready for commercial production. Note that in all of the above devices, optical switching is occuring, but this is not the type of optical switching that most people think of when they here this term. Switching speeds for the above systems can be in the milliseconds--not exactly high speed. They don't have to be because at this level of signal aggregation--in high capacity backbones or metro networks--the these optical crossconnects are used primarily for network provisioning and restoration after a fiber cut--in other words high level network management stuff. The final piece of the puzzle, actual optical routing and switching of individual optical bit streams where each packet address is optically read, processed, and sent on its way is very far in the future. This obviously will require very very fast optical switching. Even in this case, the intelligence in the network will have to be electronic more than likely because optics although great at transport are poor for computation. Hope this helps clarify the issues to some degree.