Verizon's Ihab Tarazi on driving 100G, control plane technology into the metro network
Ihab Tarazi, Vice President of Global IP & Transport Planning & Technology, Verizon
Verizon Communications (NYSE: VZ) has been an early mover on 100G optical, having integrated the technology into a number of its European backbone routes, with a similar movement taking place in the U.S. this year. But as its capacity needs grow to support business and consumer traffic, the service provider is finding that 100G will have a place in its global metro networks. In addition, it plans to deploy Ciena's 5430 Reconfigurable Switching System, a packet optical transport system (P-OTS)--a device that includes OTN (Optical Transport Network) aggregation and switching to extend control plane technology--in the United States. The control plane technology will be able to provide U.S. customers with the same mesh network capabilities for restoration on other network routes. Sean Buckley, Senior Editor of FierceTelecom, recently caught up with Ihab Tarazi, vice president of global IP and transport planning and technology for Verizon, about its latest 100G and control plane technology moves.
FierceTelecom: Verizon recently announced plans to incorporate control plane and 100G into its metro networks. Can you give an overview of the drivers behind those plans?
Ihab Tarazi: We've had a long history with 100G. We enabled the European backbone with 100G more than a year ago. Then, we also said we're deploying 100G in the key U.S. markets this year in terms of long haul. Now, we're going to continue 100G deployment in the metro sometime next year. That means the strategy is not for the backbone only, but also to extend it to the local distribution for all of the business and consumer traffic. The second piece is that we're also expanding on our global mesh. We've used Ciena (Nasdaq: CIEN) Core Directors for optical switching, restoration and aggregation. Now we're going to expand that into the U.S. with their 5340 product so we can build a logical optical network that has all of the benefits of control plane technology on a global basis leveraging all of these elements. We provided an example of that with our announcement with the Verizon Financial Network (VFN), which is about providing low latency services for financial customers. That's an example of how we're using that network for industry-specific solutions.
FT: Okay, so what's required to bring 100G to the metro? Are there different things you need to do versus the long haul network?
IT: It's not a transition. It's just that for the long haul network, you have different drivers: you worry about latency and you have to use equipment that can take the signal long distances without a lot of repeaters. In the metro, you can still take advantage of the same technology, but you don't have to worry as much about the power of the lasers and the electronics to get long distances, but you'll still take advantage of it. Long story short, it's probably easier to do 100G in the metro, but the scale in long haul is much larger in some places because you have these big pipes. In the metro, we have gotten to the point that the scale requires 100G and we're going to take advantage of it.
FT: In addition to carrying all of your FiOS, DSL and business services traffic, are you seeing interest from customers in 100G-based services?
IT: It will carry all of that. It will carry FiOS where we have FiOS. It will carry business customers and provide access to Terremark and other cloud data centers we have and some customers like government and major corporations that require very high speeds. We'll also use it for connectivity for wireless and other applications. Long story short, we see the demand from all of these things and depending what metro, it could be used for business and consumer and be used just for business. This is a global strategy so where we see demand we'll deploy it in any city. That allows us to scale the network, get to the next generation technology, and take advantage of a lot of the capabilities that come with this architecture including control plane.
FT: Is this going to be rolled out on an incremental basis, or basically be deployed where there are specific metro optical pain points?
IT: Yeah, I think that's a good way to put it. We do have already next-gen ROADM equipment in many places. It's a matter of expanding on that and also instead of having 10G circuits on them going to 100G. It's a combination of additional nodes and scaling the ones we already have because we have been deploying next-gen optical technology for a number of years. It's an evolution. The most significant part of the evolution is that we're going to start to take the metro to 100G everywhere we have capacity needs. That's a big decision because that means you have to qualify technology, install it, test it and start to make it your path to expansion. Then, the second thing is we're going to extend control plane, which out of the two announcements is the bigger deal because we are adopting optical switching and restoration that will allow us to enhance the customer experience.
FT: Besides the 100G in the metro, a big piece of your 100G evolution as you mentioned is control plane. How significant is that development and what are the benefits for Verizon and the customers?
IT: One of the first benefits is resiliency. Where we have the global mesh installed today we have been able to withstand a lot of natural disasters and outages, including things like the Japan tsunami, with no customer impact. The control plane technology allows you to go more than 1+1 for restoration. It will be able to see the entire network and maximize restoration and performance for that customer. In some places across the Atlantic Ocean we have 10 different paths and the network understands all of them and it restores around all major outages. The second thing is latency optimization because control plane technology allows you to optimize your restoration and provisioning based on latency. Then, the third thing is we can offer QoS on optical so you can have different Classes of Service (CoS), which is not something people do today. The industry-specific VFN is a perfect way to show QoS because you're guaranteeing high performance and a lot of key premium quality metrics over an optical network. You can do that once you use control plane and other technologies. It's a combination of investments of 100G to scale and get improved performance plus control plane on top so you can optimize for latency, restoration as well as for QoS.
FT: You mentioned lot about low latency a few times. Do you see benefits for low latency for customers outside of the financial trading industry?
IT: Financial companies have the biggest business case for low latency, but just about every company benefits from improved latency. I think latency improvement is something everyone desires, but not everybody is willing to pay for it as a premium service. When we improve latency it improves the performance for everybody on our network. That's one of those quality metrics that people see in their day-to-day applications. All of the applications improve when you have lower latency. I guess there are some applications that are not real time, but real time applications will see a benefit. The financials have such a benefit from it that economically it makes sense for them to make it a premium product.
FT: Verizon and others continue to leverage Packet-Optical Transport Systems (P-OTS) in their networks. What is the value of these systems, and how have they served Verizon in its ongoing next-gen optical network transition?
IT: I think I would say two things. First of all, optical technology has been evolving extremely well based on our needs and it didn't happen by coincidence. We're very active to work closely with our key partners like Ciena and some of the component suppliers to get where we need to. We're starting to see the benefit of 100G and coherent technology that we all expected. Optical by itself is producing good results and the fact that we're going to extend it to the metro means there's enough scale and good economies where there's good pricing to be able to put it into the metro. Now the next-generation of optical has become mainstream in my opinion. In terms of packet into optical, there's a number of benefits there we have been able to achieve. It has a much more simplified architecture where you can have packets like Ethernet go directly over optical waves. You don't have to have another layer in the middle. I would say the combination of enhancements in optical plus the fact you can have packet directly over optical waves both streamline the economies of the network.
FT: The reemergence of coherent technology has been heralded as a great breakthrough to enable 100G. Would you agree?
IT: Coherent is clearly one of the key enablers of all these benefits. What coherent technology allows is the following. One, it allows you to have much lower latency and allow the signal to go longer distances without a lot of repeaters in the middle. The second thing is it gives us the improved operational efficiencies because now all you have to do is put cards at the end and not have to go to all of the sites to activate capacity. The third thing is it allows you to go 10X from the capacity to 10G to 100G. During the recent OFC show, we illustrated how you can go to 20 Tbps on a single fiber system. We think you can do even more with things like superchannels and other capabilities.
Coherent technology is clearly the enabler of expanding the capacity on the optics, simplifying the model, and improving latency and quality. The combination of those things makes it an efficient technology. If you don't have the scale in optics, you can't do all of the other stuff.