December 15, 2007

JOHN DAHLQUIST

Communications Executive Profile (sponsor)

John Dahlquist is vice president of marketing for Aurora Networks. His 40 years of experience in broadband communications products and services include senior positions at Harmonic, Philips Broadband Networks and General Instrument.

What has been driving optical infrastructure in the access network over the past several years?

Our customers compete in a very competitive world, with direct-to-home satellite providing high-definition capacity and telephone companies expanding the capacity of their networks to provide high-speed data and video services. Cable operators have the ability to deliver all of these types of services and deliver more capacity in total. What's really driving our product development is how to provide them with the tools to create that capacity out of their existing infrastructure.

Do you think operators need to pull fiber to the home?

Cable operators already have 1 GHz of bandwidth going into the home. They don't need to build fiber to the home today. It may useful in certain areas, but they don't need it everywhere. The question is, how do you take advantage of what you have today and leverage that investment to be able to provide the bandwidth that's needed for tomorrow's services?

The HFC network is known for its flexibility. How does fiber in the access network play into that equation?

In the coax world, you had a single transmission path. With fiber, you have multiple paths: from the headend to node sites and from nodes to the home, if you take fiber that far. You have a much broader highway, and you can utilize that fiber over and over again to deliver unique, targeted services to different segments of the system without having to go back and pull new fibers and spend money on very expensive construction.

There's also the ability to easily move fiber deeper into the network, which eliminates more of the RF amplifiers and increases the system's reliability and narrowcasting capacity. It also reduces maintenance costs because when you do fiber deep, in comparison with standard HFC, you eliminate 75 percent of the actives. Maintenance goes down, and the cost of powering the system goes way down.

Do you think there are any misconceptions about the fiber-deep approach?

The biggest misconception is that it costs a lot more to build it than to build an HFC plant. What we've found on every greenfield application is that costs are within percentage points of each other when you look at total cost of the build. In comparison with a total rebuild, which becomes very much like a greenfield, fiber deep may be a little more expensive if an operator can salvage some of the existing equipment. Fiber deep also may not be as cost effective when operators are just doing minor tweaks to the system.

The benefits, however, are additional capacity, fewer actives, less maintenance and savings on power bills. The opex side of the business gets much better, and you have a much more reliable system. If you do have a failure, you can isolate it immediately because this node is only servicing 100 people.

Apart from deep fiber, you advocate the digital return. Why so?

We digitize the whole 5 to 40 MHz band. By doing that, we have a higher quality signal, and we don't have the distance limitations that you have with analog return. We use digital return 100 percent of the time, which has helped us drive costs down. It's a little more expensive than analog, but you get Ethernet capability, status monitoring built into it, as well as a lot more capacity than you would have with analog return. Additional techniques - such as concatenation and bringing two segments back from one node on a single laser - get the cost even lower. Digital return is in our opinion the only way to go.