Eve Griliches, managing partner, ACG Research on Packet Optical Transport System trends

Eve Griliches, ACG ResearchWhile there are many definitions to just what a Packet Optical Transport System (P-OTS) is since the concept began to bubble up on the scene just a few short years ago, carriers are clearly interested in leveraging these platforms to converge their legacy TDM services and next-gen IP service sets. While definitions vary, a P-OTS platform combines three essential elements: a WDM piece with a reconfigurable add-drop module (ROADM); a TDM interface to multiplex and groom traditional TDM traffic; and an Ethernet switching interface. FierceTelecom Editor Sean Buckley caught up with Eve Griliches, managing partner at ACG Research, a speaker on Packet Optical Network Platforms: Making the 100 Gbps Connection webinar, to talk about how this market segment is shaping up.

FierceTelecom: There are a few different approaches to P-OTS. In researching the space, do you think there any pros/cons to these approaches or does it really depend on the carrier's own preferences?

Griliches: Well, there are quite a few different approaches. I think the Packet Optical Transport market as it is loosely defined basically says that you are utilizing optical transport, OTN, SONET and Ethernet technologies. To date, I can say there's not one product out there that's actually doing that. Everyone is claiming that their system is packet optical transport capable and things are going to be cobbled together.

It's an interesting market and everyone is offering pieces of it and different service providers are going to use different portions of it. When we talk to AT&T (NYSE: T), they say they want to keep DWDM and OTN together, but keep their packet infrastructure separate. They don't want Layer-2 features; they don't want core routers to be incorporated; they want those functions to be standalone and separated. AT&T will never go to a full optical transport optic box.

Obviously, Verizon is leading the edge on this and have already deployed what should packet optical transport in the metro, but the packet support to date has been a bit lacking. I think they are still waiting for vendors to deliver on the promise of that and are looking for the same thing in the long-haul. This is not something that I would consider a God Box.

Both AT&T and Verizon (NYSE: VZ) asked for things that were God Box-like and were too expensive for anyone to build. I think this is definitely it's something capable of being built. The real question is, are there P-OTS boxes that can be made to be smaller and are there applications for that? The honest answer is today is yes. Today, Cyan Optics, which has had good revenues, has proven that a smaller packet optical transport box has worked. ECI has been doing their transport box, bringing packet capabilities into it and adding transport capabilities to their packet products. I think everyone is working to make any combination thereof.

What‘s clear to me is that there has been a shift in the market. We have been talking about packet optical for a little over two years now, and when we talk about the packet side, it's like we're talking about Layer 2 and Layer 2 1/2 switching.

This brings us an interesting crossover point: it's not Layer 3, it's not routing, and people have said, how much is Layer 2 1/2? Is it just MPLS TP and does it really stop there? Those were a lot of the questions that the router vendors had early on for Verizon. Are you sure that's all you want because we can't build a product like that and then go back in and add routing features? To date, the answer is yes we want a simple MPLS TP switched box that operates at Layer 2 and Layer 2 1/2.

That brings a more interesting question to the market: have the optical players delivered on their optical technologies to date over the past two years? Many of them have tried and many have delivered on some things, but there have not been a reduced number of vendors in the packet space for a reason because it's hard to do--it's really hard to do. We're not in this case looking at supporting BGP and necessarily multicast capabilities, but doing really good packet switching and doing it with flow control and making sure you can monitor it requires some additional features. Those that have been good at it for a while are going to start delivering on this stuff.

What I am saying is because some of the market leaders haven't delivered on the packet side of these products, the potential is shifting towards these packet-centric guys like Alcatel-Lucent (NYSE: ALU), Cisco (Nasdaq: CSCO) and Juniper (NYSE: JNPR). They may actually end up with a stronger foothold in the P-OTS market than anyone ever thought.