Skip to main content

Cox’s Rolls Kicks Up Bandwidth

Cox Communications, like other privately held MSOs, has been aggressive in upgrading its existing systems to 1-Gigahertz operation. Basically, it doesn’t have to worry about spooking Wall Street with its capital-spending budgets. And that additional space provides valuable real estate for new services, including higher-speed broadband based on DOCSIS 3.0 technology. Jay Rolls, Cox’s senior vice president of technology, said he believes cable is on the cusp of solving the “bandwidth problem” for at least the next 10 years. He recently spoke with Multichannel News technology editor Todd Spangler.

MCN: Cox has upgraded many of its systems to 1 GHz. How broadly has that been deployed?

Jay Rolls: It’s pretty widely deployed. It’s the majority of our markets. We’re well along the way. We’re not going to be finished next week, but we’re well past the halfway point.

Obviously, there’s a sheer level of logistics and contracts and everything it takes to do a rebuild like that. And then there’s another element that has to do with testing your coax network just to make sure it can pass 1 GHz. There’s just a little bit of cable replacement you need to do — you have to go through and find weak spots. It’s percentage points, not 20% or 30%.

MCN: Does moving to 1 GHz reduce the pressure to retire analog channels?

JR: Absolutely. Originally, none of our [consumer-premises equipment] could tune to 1 GHz. Now we have equipment from our vendors to do that.

We are just getting close to the point where we need to put things above 870 [MHz]. DOCSIS 3.0 is kind of the poster child for placing above 870. It requires new CPE anyway, so you have this new capability you can introduce without impacting any of your existing services.

MCN: Are DOCSIS 3.0 services going to appeal primarily to business customers or high-end consumers?

JR: I think the answer is both. Obviously, [deployment] becomes driven by business requirements, not technology. At what point and in what markets do you feel like DOCSIS 2.0 runs out of steam? That provides the answer for when I think I need to go to 3.0. There’s no nice clean answer for you about when, where, how — we have lots of demand across all our markets. There are multiple dynamics that will dictate when we go to 3.0 in a certain place. You’ll see it roll out the next few years.

MCN: What’s after DOCSIS 3.0? What’s the next big area of technology development for cable?

JR: We’re already seeing a shift toward products and services in the industry, and away from “how do we get the bits down the wire.” I think the “what’s next” is more of a product focus.

You could argue that if your node sizes are a reasonable size, and you’re doing switched digital video and DOCSIS 3.0, and CPE is moving toward eight tuners — that’s gobs of bandwidth — you could argue that you’ve largely solved the bandwidth problem for at least a good 10 years. You’re going to focus on what you’re delivering, instead of how you’re delivering it.

MCN: How significant is the Oct. 6 ruling in Cox’s favor in the patent-infringement lawsuit filed by Verizon Communications?

JR: I think it was pretty much of a key outcome for the cable industry, for sure. Largely, what Verizon was trying to do was go after PacketCable. Not only were we found to not have infringed, but we invalidated two patents — the same two patents that were successfully litigated against Vonage for. One of the learning experiences is that cable operators are going to have to get a little more savvy about intellectual property. My observation is that as an industry, we’re not doing a very good job of capturing and protecting the inventions we’re coming up with.

MCN: How far along is Cox’s backbone buildout, which will cover 13,000 route miles of fiber? What will that allow you to do that you couldn’t do previously?

JR: If all goes well, we’ll be largely done by the end of the year. It does a couple of things: It gives us owner economics on transport, which are very advantageous — it’s owning versus renting. Trust me, for our scale the math works.

The other thing you get is, you get operational improvements on just-in-time bandwidth growth. Previously, when we needed more bandwidth we would have to lease more circuits and work with a third party. Now I just stick line cards into optoelectronics and turn on new wavelengths. You massively simplify it from a time, labor and third-party dependency standpoint.