10:45 AM -- I recently had the chance to chat with Patrick Ostiguy, founding president and CEO of Accedian Networks. Patrick’s company is a leading provider of Performance Assured Ethernet solutions, which we reviewed in a recent Heavy Reading white paper. Patrick will be participating in a session on the topic during Ethernet Expo Americas 2012, which will be held from Nov. 6 to 8 in New York City.
Below are highlights from my interview with Patrick.
Heavy Reading: Can you share a bit about the growth you have experienced in 2012 and highlight some recent deployment activity related to various applications?
Patrick Ostiguy: We’re in the apex of our growth. We’re certainly continuing good growth. We recently added another Tier 1 mobile operator. We also added our first volume-consuming AAV (alternative access vendor) in Europe. The AAV phenomenon is starting to get outside North America and is getting into Europe. That AAV is serving a large Tier 1 mobile operator.
We also won another Tier 1 business Ethernet service provider. They are looking to offer performance assured cloud services. The operator is leveraging their own transport and building their own data centers. They are not only offering hosted and cloud services, but they also are offering the performance assured Ethernet lines that link the enterprise premises to the performance assured cloud. As bandwidth is getting commoditized, that’s a means for a service provider to truly differentiate themselves.
HR: What does your revenue mix look like in terms of sales related to mobile backhaul or business services? What trend do you see there?
Ostiguy: Accedian has an installed base of approximately 120,000 customer sites, including 75,000+ cell sites. The sheer size of the mobile backhaul projects dwarfs the Ethernet business projects, but mobile backhaul projects tend to be lumpy and they come in phases. The backhaul part that is more steady is what is related to the AAVs. Even though they are individually lumpy, they sort of compensate as a whole.
Business Ethernet has been more organic and usually has accounted for less than 50 percent of our revenue per quarter. Recently, business Ethernet has picked up more in the form of new account wins. That should translate into increased volume in the months to come. We have started seeds in Asia-Pac, Latin America, and the Middle East, but the bulk of revenue is from North America and Europe.
HR: Heavy Reading survey feedback and conversations with many senior service provider experts indicate increased availability and expanded use of end-to-end service management solutions to be one of most important developments in the Ethernet market. Accedian has been very active in this space with your performance assured Ethernet solution. Can you comment on how PAE has been received in the market?
Ostiguy: The PAE story has been very well received. The general comment is that adding visibility allows them to reduce mean time to repair (MTTR). Business Ethernet is a service that typically is hard to maintain and hard to repair. PAE helps eliminate uncertainty about whether an operator is going to get an anticipated ROI (return on investment).
HR: Operational cost savings associated with the use of service OAM technology appear to vary widely from operator to operator, partly depending on how extensively OAM technology is deployed across their networks in support of various services. Do you have data points that you can share in terms of the opex impact for customers using your PAE solution?
Ostiguy: Each one that we've seen that has tried some form of PAE has quickly recognized the benefits of it. What goes above and beyond that opex savings is it gives them the ability to differentiate themselves. Especially with business Ethernet, there has been widespread commoditization. PAE gives them an ability to take business away from competitors and reduce churn.
One of our customers -- a leader in global Ethernet services -- experienced an 87 percent reduction in their truck rolls. Obviously, that’s on the high end of what we've seen. If you look at some of the surveys we’ve done internally, we usually see a 40 percent to 50 percent reduction in truck rolls. The opex savings also come in reduced provisioning time, reduced SLA violations, and faster trouble ticket resolution. Usually PAE allows them a six-to-seven-month ROI.
HR: What about capex benefits?
Ostiguy: In terms of capex savings -- that's often neglected -- there is savings associated with the use of test equipment or more expensive access platforms that are often used for business Ethernet. You don’t need test sets as much because PAE generates birth certificates when a service is installed and it centralizes and automates a lot of that -- accelerating deployment altogether. Sometimes you see them using a switch/router when all they are doing is handing off Ethernet at the client site. They do that because it's the only gear that they can manage. PAE, in general, supports all the edge traffic conditioning you will ever need and more, at a fraction of the cost of a switch/router.
HR: More than a half-dozen senior Ethernet service provider experts have told Heavy Reading over the past several quarters that they very much would like to see the MEF finalize a standard for hybrid NID technology -- or what is now being referred to as virtual NID technology. What is Accedian doing on that front?
Ostiguy: There are a couple of things. First, the obvious reason service providers are interested is that they often see that NIDs are being duplicated. At a handoff point, each service provider has its own NID. It's understandable that they would want to reduce equipment. The challenge is that it is very complex from a logistical point of view to resolve this.
In the meantime, we offer a V-NID. A wholesale provider would come with a NID. If you are a service provider and are extending out of franchise, you can have visibility through the use of software performance monitoring agents, called V-NID reflectors, that respond to measurement frames sent out from a V-NID accuator. This fulfills the need to reduce the number of back-to-back NIDs without having the conundrum of who owns access rights, etc.
Say, for example, Verizon Business wants to serve out-of-franchise over a leased line from AT&T Global. AT&T installs its own NID to terminate the Verizon service and hand it off to the end user. Classically, Verizon Business would ship their own NID and hang off it beyond the AT&T NID. They'd have their own eyes and ears. That's a classical back-to-back NID situation. Instead of Verizon shipping their own NID, they could use a V-NID accuator at the core of their network and send performance monitoring packets, bounce them off the AT&T NID at the egress port, and they’d see the performance that they're getting from AT&T.
Accedian is using standard reflectors such as TWAMP and Y.1731 integrated into the network elements.
HR: Do you have customers using the V-NID solution already?
Ostiguy: We have about a dozen live deployments using this today -- more on the international front.
HR: Carrier Ethernet has emerged as a technology of choice for the backhaul of 4G LTE mobile traffic. What key capabilities does Accedian offer to help mobile operators or wholesale service providers keep a lid on operational costs when deploying and managing large-scale mobile backhaul networks?
Ostiguy: One of the things that is really characteristic in cell tower upgrade projects is acceleration of deployment -- get to as many towers as fast as possible. They're using every trick in the book and using every alternative access provider possible. When upgrading, ideally you'd like to make just one trip to the cell site and use the content of the new base station cabinet being delivered by the eNodeB vendor to take receipt of the backhaul circuit and verify its performance. But coordinating the delivery of cabinet equipment to every cell site with the receipt of a circuit and the certification, testing and acceptance of a service all at one time is logistically impossible. This opens the door for the deployment of a NID at the cell site.
A NID can accelerate deployment via automated service turn-up. A NID can be drop-shipped to the cell site. It has factory-defaulted, out-of-the-box self-provisioning. There does not have to be an actual truck roll for the installation of the NID. It could be installed by a site attendant or by the AAV who is going to deliver the circuit anyway. This allows receipt, certification, acceptance and the birth certificate of the service without relying on coordinated delivery of cell site equipment.
Performance assured SLA is an important part as well. Digging into that, one of the performance parameters that is neglected but important is throughput. What's more important than throughput? We've come up with innovative ways to leverage Y.1731 and develop ways to assess throughput in a live manner beyond the birth certificate. We have developed means to re-verify and monitor throughput in deterministic way.
HR: Small cell backhaul has emerged as a hot topic in 2012 as operators look to expand RAN capacity. What are you doing specifically to address small cell backhaul applications?
Ostiguy: It's still in its infancy from a deployment standpoint. What's going to be a challenge is the form factor. We see direct applicability of the V-NID concept in the small cell application. We have a partnership with a large OEM that will not require them to install a dedicated box. For small cells, you may not be able to fit the NID there from a power, real estate, or heat standpoint. It just may not be practical. This is a perfect case for the V-NID product line.
HR: Patrick, thank you very much for your time. And I look forward to catching up with you in person at the Expo, if not sooner.
— Stan "EtherMan" Hubbard, Senior Analyst, Heavy Reading; Ethernet Executive Council Program Coordinator; and Chairman, Ethernet Americas 2012