Optical's Packet Magic

One of the biggest optical networking developments in 2007 will also be the key one to watch in 2008. I'm referring to the emergence of the packet-optical transport system. These new systems essentially represent the convergence of metro WDM systems, MSPPs, and Ethernet transport. Heavy Reading is now tracking this emerging market in its new Packet-Enabled Optical Networking Quarterly Market Tracker.

Packet-optical transport systems will be the replacements for the "legacy" MSPPs that are the bread and butter of major optical suppliers today. They also represent the next generation of metro/regional WDM systems (another large optical market). As such, it would be difficult to overstate the impact these products will have on the optical market.

These products are very new and features are still evolving, but Heavy Reading is looking for the following functionality for a platform to qualify as a packet-optical transport system:

  • WDM transport and wavelength-level switching, via ROADM or OEO crossconnects
  • Full support for Sonet/SDH within the chassis (or "in skin")
  • Layer 2 connection-oriented Ethernet switching and aggregation
  • "Transport class" network element, including 50ms recovery and five-nines (or greater) availability, software upgradeability, and OAM to match Sonet/DWDM

We will continue to refine our definition as the segment evolves, but we think this makes a good starting point to begin evaluating products.

To date, we are tracking products from eight vendors, including:

As of the end of 3Q07, no vendors had all of the pieces in place to qualify in this product category, so we are in the earliest days of the market right now. Typically missing were either the ROADM functionality, the Layer 2 connection-oriented Ethernet functionality, or in some cases both. It's possible that some revenue was booked at the tail end of 2007, but we expect the first meaningful shipments to occur during the first half of 2008.

We believe the packet-optical transport market will grow from essentially $0 in 2007 to $144 million worldwide by the end of 2008. By the end of 2012, we forecast that this market will reach $2.8 billion – increasing at a 110% CAGR in the four-year period from 2008 to 2012. This will be the fastest-growing area of optical transport.

Challenges & Unknowns
Our Heavy Reading and Light Reading surveys have shown that carriers are keenly aware of the "packet problem" in their networks, but that they are also sharply divided on how best to solve it. Besides the new packet-optical transport system, other options for migrating networks from TDM circuits to packets include prolonging the lifespan of MSPPs with packet advancements or investing in overlay carrier Ethernet networks.

In a Heavy Reading survey of 72 respondents from network operators around the world, published in June 2006, transport network migration strategies broke down as follows:

  • Converged TDM/WDM platforms: 24%
  • Carrier Ethernet overlays: 29%
  • New packet-optimized MSPPs: 30%
  • Linecards on existing MSPPs: 17%

We also conducted a follow-up poll during an August 2007 Light Reading Webinar entitled "The Promise of Packets," in which 56 network operators from around the world responded to the same question as above. In this more recent poll, the breakdown was this:

  • Converged TDM/WDM platforms: 37.5%
  • Carrier Ethernet overlays: 37.5%
  • New packet-optimized MSPPs: 12.5%
  • Line cards on existing MSPPs: 12.5%

Our conclusion from these two polls is that the battle to be waged is between building overlay carrier Ethernet networks or building networks with newer packet-optical transport products. MSPPs have had a great run, but we believe they are now poised to decline, as both packet-optical transport and carrier Ethernet deployments grow – a no-win scenario for MSPPs.

ROADM-enabled WDM has a role to play moving forward in either network architecture, but the key difference is whether it's the star performer or a member of the supporting cast. Packet-optical transport products must be able to switch wavelengths, so ROADMs will be key. Carrier Ethernet proponents put the Layer 2 or 3 intelligence in switches and routers, but still see a requirement for a dynamic transport layer that includes ROADMs. In this scenario, it would play more of a "dumb" (but reliable) transport role.

As 2008 gets underway, our eyes will be on the initial commercial deployments of packet-optical transport systems and the commencement of the competition between packet-optical transport and carrier Ethernet.

– Sterling Perrin, Senior Analyst, Heavy Reading

For more information about Heavy Reading's Packet-Enabled Optical Networking Quarterly Market Tracker, or for a FREE PREVIEW of this exciting new product, please contact:

Dave Williams
Sales Director, Heavy Reading
[email protected]

Be the first to post a comment regarding this story.
Sign In