The 2013 CTO Roundtable
Several of the cable industry’s top technology execs provide a snapshot of networks that never seem to stop migrating toward something new
As we’ve done for years, we’ve interviewed the CTOs of several cable operators. We’re pleased and proud to once again be able to provide the insights of several of the most knowledgeable CTOs in the cable industry.
Time Warner Cable’s Mike LaJoie, Cablevision’s Yvette Kanouff, Cox Communications’ Kevin Hart, and Buckeye Cablesystem’s Joe Jensen share their thoughts on some of the most prominent technological challenges they are dealing with today, and a few they might have to contend with tomorrow.
[This online version includes CTO responses regarding UltraHD (4K) TV and on how their own TV viewing habits have changed in response to new technology, which were omitted from the print edition because of space considerations. -BRS]
CED: What is your current thinking on multiscreen? What is missing?
Yvette Kanouff: Cablevision has been very aggressive with our multiscreen strategy across multiple devices including iPads, iPhones, Mac, Windows, PCs, Android and the Kindle Fire.
We have more than 200 channels of linear programming available with our Optimum App and have nearly 50 TV to Go offerings available for Optimum TV customers.
When thinking about multiscreen, going forward, there are several important factors including: (1) Authentication remains an important part of our multiscreen strategy. We want to ensure that our customers have a seamless and easy method in accessing their content on all devices.
(2) Ensuring our customers have an easy to use interface for accessing content, we have implemented a unified experience across devices so that our customers can easily shift from a PC to a TV to a mobile device, for example.
(3) Quality of experience – this is extremely important.
Multiscreen is not just about download speeds and ABR [adaptive bit rates]; it’s about ensuring our customers have a consistent and quality viewing experience. There are many technologies that go into optimizing QoE on Wi-Fi and in home networks that we constantly work to improve.
(4) When it comes to mobile devices, we find it’s very important to constantly work on evaluation of behavior, crash logs, and usage to ensure that we optimize the customer experience.
Joe Jensen: We’re moving cautiously forward with that.
Implicit is that we have solid encryption and a digital rights management solution, that we can deliver the content in a way that is error-free from a customer’s perspective, that we have sufficient server capacity to manage simultaneous streaming.
Once you have the rights, you have to stand up a system that addresses the security, the capacity, and the quality of signal – hopefully we’ll be able to package that up and present it to our customers over the next intervening months, hopefully not years.
There’s a significant capital cost to stand up the platform, the transcoding capabilities, the origin servers, the authentication – all of this is certainly a challenge for smaller operators. We feel we’re at least large enough to put the energy and resources into making something like this happen, and as we scouted the landscape we thought Azuki would be a reasonably good fit for what we wanted to accomplish.
We’re cautious, of course, because it’s an entrée for a direct solution from programmers to our customers. We want to do something in the best interest of our customers but that also maintains our ability to migrate our offer to be the aggregator of choice. One aspect of that is TV Everywhere. We’re having some challenges because sometimes it’s hard to get programmers agree to go through the process of providing authentication.
I’ve had more than one programmer tell me, “Well, we’re working with everyone who has a million subs or more and we’ll get around to you sometime in the future.” That’s been the biggest obstacle to this point.
Kevin Hart: On multiscreen we’ve been very successful with our launch of Cox TV Connect, which we rolled out a little over a year ago on the Apple platform, and we recently have gone live on Android devices.
We’ve had probably half a million downloads to date. We’ve got about 90 channels live on the platform.
The challenges are just like the other folks’, it’s filling out the infrastructure, the CDN environment, the encoding, and then getting the local encoding contracts in place with some of the providers. It’s been fairly straightforward for the most part, and we’ve got good response from our customers.
CED: Time Warner Cable has just extended multiscreen delivery to subscribers viewing outside their homes. Are there any particular technological challenges associated with that?
Mike LaJoie: We’ve just launched our TV Everywhere app, and there’s a nice set of channels for which we have the rights.
There are issues. If someone’s outside the home and they’re connected to some other network, I have no ability to influence the quality of somebody else’s network, right? If they’re in their home and they’re connected to a Wi-Fi router that we’ve provisioned, they’re going to get a pristine experience. Not to say it’s not going to be a good quality experience on someone else’s network, we don’t have the ability to make sure it’s as good as it can be.
CED: You can’t control the QoE?
LaJoie: Right. I don’t want that misinterpreted. It’s not that we need to do quality of service or control bandwidth in that way, but in the home, it’s a managed service. It’s not using the same bit rate that you have for your high speed data service. So if you have a 15 megabit data service and you’re maxing that out, it will still work in your house, as long as there are still bits in your service group, it will still work. But if you’re off our network, we don’t have the ability to deal with all those vagaries.
CED: Is the growth in streaming video consumption likely to present any technological challenge?
Hart: Right now, it’s in single-digit percentage of our capacity in terms of the infrastructure we’ve built out. Today we’ve got great headroom in capacity and quality of service. As it becomes more mainstream, and as we continue to integrate second screen apps – as we’ve done with the primary screen for recommendation engine and additional content, we’ll continue to have to invest in infrastructure to scale, but we’re in good shape.
We announced at CES a personal entertainment experience – PVE – sort of the next generation of our second-screen tablet, with live linear, with access to our VOD library, as well as pre-populated over the top applications. There’s a really very quick, intuitive, easy to use interface. We’ve gotten great reviews. We’re in the final stages of development and getting ready for deployment and launch later this year. It’s the next coming of the second screen, with personalized recommendation integrated into the primary screen as well.
As far as I know, I haven’t seen the integration of the live linear content as well as the VOD access in addition to the over-the-top applications, all integrated with the personalized recommendation engine, and integrated with the main screen. Having the integration of those three types of content, assuming a rich VOD library, that should move the needle.
CED: Wi-Fi is becoming a default connectivity option for more and more devices, including dual-mode cellular handsets. Not too long ago it was considered merely adequate for home networks, and now MSOs are building public Wi-Fi networks. How has Wi-Fi been peforming?
LaJoie: In general, you get a better broadband experience with Wi-Fi; there’s less contention. Every time you’re connected to a public radio network you’re subjected to a lot more contention than you are on a wired network, generally speaking. If you’re talking about entertainment quality of video over a wireless network, it’s probably less reliable than over a wired one.
CED: But as long as the subscriber is on a Wi-Fi connection, the experience will be at worst acceptable for subscribers?
LaJoie: Yeah. It will be as good or better than anything else you can get on a wireless connected device. More and more, Wi-Fi is becoming the broadband network of choice, from a mobile user’s perspective.
If they have a Wi-Fi network, they would rather use that, more and more, than their cellular network. Wi-Fi is generally better, and usually less expensive. Wi-Fi is becoming the wireless broadband network of choice, and LTE is becoming the fallback.
You certainly get better experience.
Jensen: Wi-Fi coverage within a home can be spotty. I’ve tried to understand how we could build a home network that provides seamless viewing and surfing, and it’s a real challenge. There’ll be continued work there. I have more comfort with wired networks because they’re not so susceptible to interference. As more and more Wi-Fi enabled devices pop up, that just adds to the congestion. Moving from 2.4 gig to 5 gig will only be delaying for a short time because the congestion will simply migrate to wherever the most efficient means of delivering content are.
CED: CCAP – the converged cable access platform – lays out the path to distributing video and data over a largely unified network, which should be less expensive and more efficient. What are your plans for adopting the technology, and what might be the first advantages from doing so?
Kanouff: CCAP brings a significant amount of flexibility to our networks, both for optimizing bandwidth and the management of MPEG over QAM vs. DOSCIS/IP delivery.
As we continue to evolve our bandwidth planning, CCAP plays a role in this. We look forward to CCAP being widely available.
Hart: CCAP is a big part of our plans.
We’ll have a variety of vendors and their CCAP technology in our labs probably early next year and beyond, and then hopefully things pan out to deploy in 2015 and beyond.
DOCSIS 3.1 is a critical component in terms of higher speeds and more efficient throughput. Just the general blocking and tackling around those technologies and extending our network are well under way this year, as well as enabling some of the CMTS upgrades as well.
LaJoie: We’ll deploy it very quickly, as soon as products are available. Probably the first application will be the ability to more conveniently bond larger channel groups, to go beyond eight bonded channels without having to introduce a lot of inter-chassis switching fabric. We can go beyond eight bonded channels today, but it requires a lot of extra work. With CCAP, with some of the products with some of the vendors, we’ll be able to bond more channels, which then means we can offer broader-based products.
The other thing that’s interesting is the ability down the road to combine different kinds of transport. You can manage an Ethernet product out of the CCAP chassis and ride it over PON. And we’re deploying some EPON today in two commercial tests.
So the ability to use DOCSIS provisioning and manage DPOE [DOCSIS provisioning over EPON] through the CCAP platform is going to be a real boon for us.
The ability to manage multiple access platforms using the same chassis and use DOCSIS provisioning and management will be really great.
CED: Does that suggest some of the first products based on CCAP deployments might be aimed at business customers?
LaJoie: Probably not, no. Bonding multiple channels – the application of that lets you offer higher speed tiers. Or let you have a broader set of channels that your modems can range against so that you can have larger service groups, depending on what service tiers we offer where.
CED: The trend has been to smaller service groups. Would that cut against that?
LaJoie: There are a couple of different levers. You can go to smaller service groups with the same number of channels. Fewer modems on a service group means you have less contention, so you can offer higher speeds up to a certain level.
As I go broader in channels – say I bonded 16 bonded channels together for the sake of argument – I can offer peak product across those 16 channels that might have a certain level of uptake, but I can also still be offering lower-tiered products that could be offered across a broader paid base of customers. In that scenario, depending on the uptake of the broader channels I have, the more options I have for how I build my service groups; it gives me more flexibility.
Kanouff: We are looking at all of options available to us including service group splits, DOCSIS 3.1, node splits, 1 GHz, MPEG 4, HEVC, switched digital, etc.
Certainly service groups are low hanging fruit for addressing bandwidth needs.
Jensen: This is perhaps a naïve view of what was supposed to be accomplished by CCAP, which was to significantly reduce the per-port costs for the delivery of DOCSIS-based services. We’ve got an interesting perfect storm developing where CCAP is heading, as well as DOCSIS 3.1, as well as the IP migration. How do these play together, and how do operators manage the network for a graceful transition?
As in the past, when we reach some threshold, we’ll start deploying the new technology, and redeploying the older technology into hubs or other areas as needed for growth. We’ll probably see a similar thing happen with the introduction of a CCAP router and edgeQAM. Then on top of that we migrate to higher bits-per-Hertz with DOCSIS 3.1.
All of this is wrapped up in trying to manage the transition of technologies. As a smaller operator, it’s just something to manage through.
We don’t have the luxury from the capital perspective, nor do we have the human resources necessary to make that sort of quick transition.
It’s going to be a long term process.
Hart: Particularly with a lot of the second screen applications, but leveraging the IP technology, building out a more efficient transport mechanism on the IP layer, leveraging some of the new technologies is imperative with broadband growth – with 50 percent peak usage growth year over year – it all just makes it imperative to continue to invest in technologies to enable the IP traffic stream. Convergence has been going on for years, and it’s starting to increase particularly in our footprint because the over the top video streaming is driving growth.
LaJoie: Internet protocol and related technologies are almost there, they’re almost good enough to support a quality entertainment video experience. A lot of people say, ‘gee, we got to talk about when is television is going to be ready for IP?’ When the real question is: when is IP going to be ready for television?
CED: Cost/bit has been a metric often cited when comparing video networks and data networks; it is still less expensive to distribute on the video channel than on DOCSIS channel. How does that come into play when considering when to start relying more on the data channel?
LaJoie: There’s a broad set of considerations.
The cost per bit argument is still operative. It’s probably still true that it’s less expensive to wiggle a bit over native QAM transport instead of going through the DOCSIS infrastructure. But remember that DOCSIS runs over QAM anyway. It’s all another layer of protocols that go over the native QAM stuff.
So, yes, it’s more expensive. But the benefits of using native packet switching are approaching the point where it’s making that choice close to a wash. The issue around TCP/IP is still that it’s a best-effort protocol. The reason we use MPEG-2 transport is because you can sync it to get reliable delivery of packets. The packets all arrive when they’re supposed to arrive, and they’re synced and they’re clocked, so you get reliable delivery of video. Video is not very forgiving.
That said, technology on IP is starting to get to the point where it is almost reliable enough. It can almost give you an entertainment quality video experience. With quite a bit of effort we can be there, but we want to make sure the quality is what people expect for an entertainment quality video experience in the home. It’s still a long way before it will be, flat out of the box just as good as what you expect out of television.
Certainly we have the ability to do it, but it still requires a lot of work.
CED: What are you evaluating in terms of customer premise equipment?
Hart: We’re looking at the G8 gateway [Cisco’s proposed hybrid system designed for the migration from today’s systems to IP video]. Six-tuner capability, 2 terabytes of storage – that all facilitates more of the gateway concept. We’re in testing currently, hoping to get ready for deployment later this quarter or in the next quarter.
We’re also looking to testing DTAs – a universal DTA, and that would be part of our network strategy to reclaim some analog channels where it makes sense.
Jensen: We’re still offering 65 channels of analog because our customers like it.
We’re moving into a realm where we may displace some of those analog channels so that we can reclaim that spectrum and make it available to some of the new technologies in the next couple of years.
We are anxiously looking for an IP alternative to the DTA. Something that we could deploy today that would allow us to reclaim those channels and for all intents and purposes operate like a DTA, but could be in a later phase upgraded to support interactive services – VOD, streaming, and so forth – on that same device.
We’d like to have an evolutionary platform so that we can pull the sophistication back into the network so that our devices don’t have to be so expensive and intelligent just to make it all work.
Kanouff: There are many different approaches to home gateways, with the core feature being the bandwidth and delivery flexibility into the home. Gateways provide a good sliding-scale of IP vs. QAM delivery.
TR-69 gives us a lot of provisioning flexibility inside the home.
Our DVR strategy is in the cloud. Other smart devices are a part of our multi-screen strategy, such as tablets, mobile phones and laptops, allowing customers to have Optimum across many devices. So, there are many pieces to the home evolution and gateways are one piece of the strategy.
The cloud plays a big role for Cablevision. Our DVR cloud solution provides us with the flexibility of adding storage, tuners, and services without any in home limitations. Similarly, software applications have benefited with cloud based services for needs like geo-diversity.
We have been very aggressive in moving services and management into the cloud.
LaJoie: I don’t know if you’ve seen the latest Roku box – the Roku 3 – but if you look at our application on that box, you select a channel and you get that channel very quickly – under 2 seconds.
Jensen: Retails models might be a hard sell.
Maybe it will be easier if we could see retail devices that are sub-$80, or sub-$100, because that will change the equation. When we first launched high definition, we were offering the HD converters for sale. We got a response of… what? Point-zero-one percent? It was low.
People just aren’t in the mode of buying a complex box and seeing obsoleted in two years. You see that with cellphones.
Somehow we should convince people that this is as good as a cellphone should be – it’s technology that’s good for two years, and then you move on. But I don’t think we’re there yet. I think we’re going to have to pursue a leased model, as well as position ourselves on retail boxes like Roku and Boxee and some of the others to provide our customers what they want, on the device they want, when they want.
Part of what we are obligated to do is – if that box receives any QAM linear channels, it’s required to have a CableCard in it. Charter just got a dispensation from the FCC to begin the migration without having to meet the separable security obligation.
I’m very excited to find out how something like that could be applied to other operators.
A gateway function doesn’t necessarily have to be the full-blow linear-capable box.
The gateway we’re looking at would simply be a means of distributing content around the home – a cable modem on steroids with MoCA, that can push that content out to devices, wired and wirelessly, wherever they are in the home.
We’d really like to find a solution that gets us over the hump of analog reclamation but with a platform that has some life in it. Putting out a one-way DTA is really scary, putting in a one-way device that will lock down that television for a long period of time and not give us the opportunity to provide additional services? I’d like to avoid that if we could.
CED: The latest next thing is Ultra HD, also known as 4K TV. What are the prospects for that? Are there any particular challenges associated with delivering 4K signals?
LaJoie: Not really. You need more bits for it, but it’s no more challenging than being able to support 3-D was. It will depend on the demand. If our customers end up buying a lot of 4K sets and 4K content shows up, it’s more of an opportunity than a challenge.
Hart: We currently have a couple of 4K TVs in the lab, and we’ve been doing some modeling in terms of network impact, but today there’s nothing that’s causing us any major concern. Hopefully it will become a great consumer benefit once it becomes mainstream.
Jensen: A 4K TV will have a lot of the same challenges as 3-D. It will be tough to get a lot of content available at first.
Will it provide that much of an improvement in user experience? When you go from 40- to 80 inches, pixelization becomes more apparent to some of the more astute viewers. So 4K TV is a natural migration, but the entire infrastructure is not there to really take advantage of it. Until, as with high definition, we have the whole food chain nailed down, with a 4K TV you might get a crisper picture, but it’s not going to buy you any more definition.
What’s the impact of moving from a 12 megabit high definition signal to a 20 megabit 4K HD signal? I could certainly swamp some networks.
CED: HEVC promises to provide about double the compression of current compression standards, notably MPEG-4. It isn’t required for 4K, but it should help, right?
Jensen: I hope it does get applied so that we can deal with some reasonable bit rates for those signals.
Kanouff: We see that HEVC provides a good codec for handling 4K content. It’s a similar situation to what we were in during the initial HD migration. We continue to evaluate HEVC and 4K as the technologies evolve and proliferate.
CED: With all the technological innovations in the last two or three years, have your TV viewing habits changed recently?
Jensen: I now record a lot more. I record things I might want to watch and allow them to disappear if I don’t watch them.
There was a show I wanted to watch and it was delayed because of a sporting event. I got 45 minutes of that event and 15 minutes of the show I wanted to watch. I fired up Hulu and caught the show I was interested in.
I find myself intrigued by non-traditional content. I’ve been watching TED. I find those talks enjoyable and entertaining.
Kanouff: I use my mobile devices extensively. A few years ago, I used my laptop and phone a lot for downloads. Now, I do much more streaming and live viewing on my tablet. I don’t even have a laptop anymore. Oh, and I love Wi-Fi!
LaJoie: I have less time so I utilize on-demand a lot more than I did before. I’m viewing a lot more content on my iPad. Oftentimes I’ll be watching something with my headphones on, it gives me the opportunity to be there with my wife while she’s watching “Housewives of Beverly Hills,” which is not my cup of tea.
I recently installed a Roku box in my office here and I use that exclusively for watching TV in my office. I have CNBC running in the background.
My viewing habits have broadened, I would say. The breadth of choice in how I can consume my video has really given me a lot more flexibility and I personally take advantage of it. I think it’s great.
Hart: I live in Atlanta; I’m a Comcast customer. Being able to access sports on the go has definitely been a big benefit, especially chasing my kids all over town on the weekends. Being able to leverage content on the iPad, the iPhone, Android devices. The more robust VOD libraries and DVR capabilities – definitely the time-shifting. There have definitely been changes in the Hart household.