If fiber-optic transport networks are the highways of the Information Age, then optical switching provides the intelligence at the intersections. And because the traffic keeps getting heavier and less predictable, the traffic lights have to keep getting smarter.
Optical switching has gone through many transitions over the years, but it remains a key element in the success and continued growth of the Internet. The earliest optical add/drop multiplexers (OADMs) delivered simple, colored add/drop at intermediate network nodes. With the growth of more dynamic traffic and networks, there has been the development of increasingly sophisticated reconfigurable OADMs (ROADMs) – available in a gradually larger array of configurations based on different technologies. ROADMs can boost the flexibility, scalability and remote configurability of a network while significantly reducing a bandwidth provider’s operational expenditures.
But the ROADM landscape is sprawling, and bandwidth providers must be careful to select the right technology for a particular situation in order to maximize network potential and cost efficiencies. So, what is the optimal type of ROADM – dynamic, colorless, directionless, etc. – for a given network and node configuration?
The Road to ROADMs
Carrier networking was once based solely on electrical switching, and then came Wavelength Division Multiplexing (WDM). WDM transport technology successfully multiplied the bandwidth capacity of optical networks without requiring installation of additional fiber, but it also demanded “OEO” conversion as data streams had to be converted – optical to electrical to optical – before electrical switching could occur.
On came OADMs. Based on dielectric filters, OADMs can drop an appropriate color from the rainbow carried via WDM along a given strand of fiber and then re-add incoming traffic of the same color back to the optical aggregate. Relatively inexpensive, these OADMs remain popular for fixed network installations.
Because OEO conversion proved to be power hungry and complex to scale as data rates and channel counts increased, optical cross-connect (OXC) photonic switches were developed. OXCs are transparent to data rates, meaning bandwidth providers did not have to upgrade to new generations of products as services became faster.
Deployment of these early OXCs uncovered the first instances of “blocking.” When these first OXCs attempted to switch colored WDM traffic among paths, it was statistically possible that a wavelength could be blocked by a subsequent WDM filter or combiner.
