4/10/2006 -- With carriers looking to scale their metro networks by implementing new mesh topologies, support for multi-degree rings is increasingly becoming an optical networking sine qua non. But Jason Marcheck, a principal analyst for carrier infrastructure with Current Analysis, sees a couple of unanswered questions in the burgeoning reconfigurable optical add/drop multiplexer (ROADM) rat race: How important of a requirement is multi-degree ring support, and (equally vexing) what is the best technology to enable it?

One way of enabling ROADM capabilities is by means of wavelength blocking techniques, Marcheck notes. But another, newer concept, called Wavelength Selective Switching (WSS), is also gaining support -- especially as a more efficient alternative to wavelength blockers.

Networking vendors, at least, are falling all over themselves to incorporate WSS features into their ROADM offerings, Marchek says -- and Cisco Systems Inc., in particular, has articulated what he describes as an "elegant" migration path to WSS, pushing an ONS 15454 MSTP-based stack that exploits conventional (wavelength blocking) techniques in the low end and WSS in the high end, where more robust capabilities are typically required.

On paper, Marchek says, WSS is clearly superior to wavelength blocking: It uses all-optical switching technology in order to switch wavelengths in the optical domain. "As a result, when deployed in tandem with the right network management software and colorless optics, switching patterns can be dynamically reconfigured in an 'any-to-any' fashion to account for changes in transport demands on an as-needed basis," he explains.

That's the good news. The bad news is that economies of scale haven't yet made WSS a truly affordable proposition. "WSS is hampered by cost issues. It is almost universally acknowledged that, as a new technology, cost improvements are necessary before widespread adoption of WSS-enabled switches will occur," Marchek points out. "[There's also some question about] when the presence of multi-degree rings will increase to the level that WSS-based ROADMs will become a volume requirement. Some vendors claim that their customers see little need to deploy much more than single-degree ROADMs, while other vendors will say several large carriers are seeing the need to introduce multi-degree hubbing in a majority of the rings in their networks."

In this respect, Marchek argues, even though Cisco is off to a good start with its ONS 15454-centered strategy, it can't afford to rest on its laurels. "Cisco needs to articulate more clearly how it supports topologies beyond two degree nodes," he argues, conceding that Cisco "makes a strong case for its IPLC-based approach as a cost-effective way to meet the ROADM needs of carriers when deployed in a two-degree ring."

What works for Cisco's enterprise customers might not work for large telcos, however. "In Cisco's case, with many enterprise customers, it is understandable why the IPLC approach holds value. However, as large telecom carriers look to migrate to ROADMs with WSS capabilities, Cisco could be in a position of weakness with that customer segment unless it better articulates how its wavelength blocking approach meets ROADM requirements at 'super-hub' locations," he concludes. -Stephen Swoyer