How We See CE

Last week I was attending a conference on IPTV in San Jose, and one of the keynote speakers, Charles "Kip" Compton, at Cisco, told an interesting story about how kids today experience TV. The story mirrors my own kid's experience, so I'll borrow it and pretend that I actually made it up.

My son, who is two and half years old, has no idea about the concept of TV broadcasting schedules; he is so used to always having his favorite TV shows and movies available on demand (from our Tivo, for example) that it could never occur to him that TV shows are broadcast at specific dates and times. This is not a problem at home, but sometimes he really does not understand why he cannot watch Bob the Builder when we are at our in-laws' home in Spain.

Now that today´s toddlers cannot imagine living in a TiVo-less world, it's easy to imagine a not-so-distant future where consumers won't watch live TV anymore (apart from the Super Bowl) and will always experience TV through a DVR devices. And once you get used to using TiVo, you need it in every TV in your home.

So, taking this into account, what will the DVRs of the future look like? I see three possible scenarios:

• One DVR per TV (i.e: multiple DVRs per home).
• One DVR per home, sharing content with "disk-less/tuner-less DVRs" (thin DVRs) through a home network.
• A "network-side DVR" (no physical DVR in the home), managed by the Service Provider, reachable from every "thin DVR" through the home network.

Of these three scenarios, the first one probably does not make much sense, for several reasons: storage capacity is wasted, more TV tuners are required and managing recording schedules in several DVRs would be a nightmare anyway. That's not to say that some storage won't be there to provide a "pause live TV effect", but the TV does not seem like the right place to keep your movie library. What's harder to foresee is which of the latter two models will be more prevalent, but in any case, both will put a tremendous load on the user's home network (whatever it is based on: wireless, powerline, phoneline or coaxial cable).

The amount of bandwidth that will be required to stream recorded High Definition TV shows to each individual disk-less DVR, 24 hours a day, will be huge. My prediction is that wireless networks won't be able to sustain this amount of bandwidth, specially as they become even more popular and the number of frequencies available for non-overlapping wireless networks is therefore reduced. Imagine having 5 of your neighbors each using his 802.11n wireless network to simultaneously stream TV shows to 3 networked TiVos. The interference from your neighbour will be strong and continuous; it simply won't work. And the problem will get worse as new extended-range wireless access points penetrate the market, increasing the area of interference by the access point.

Fortunately, several high-speed wired options (based on powerline, phoneline or coaxial cable) exist today, which can offload bandwidth-intensive streams from the wireless networks. Most likely, hybrid networks that combine some form of wired and wireless technology will become popular: video-related content will make use of the wired network, while non-video content (Web browsing, email, VoIP, etc) will use the wireless network.

Once I started writing this blog post, I remembered that Maury Wright at EDN had recently published an article on the same topic, which mentioned networked DVRs as one of the main drivers for home-networking technologies:

If there is a single device that has prompted the demand for video-capable networks, it’s the DVR.

He reaches a different conclusion, though: he thinks that

The 802.11n standard has perhaps the most promise, but standards skirmishes continue to burden it.

I think that even when the 802.11n standards battle is finished, wireless still won't be the best option for scenarios where a high percentage of the population has networked DVRs at home, due to the interference reasons I mentioned earlier (at least in the scenarios where the Service Provider has to take the decision about which technology to use for the home network).

There is one thing on which Maury and I agree:

It’s increasingly looking as though all of the home networking candidates might get a piece of the market. In fact, Motorola’s Chakalos says, “We would like to see them all happen.” Chakalos, for instance, believes that 802.11n will prove useful in video distribution in North America but that brick-and-concrete construction in many European homes may limit the use of wireless in video distribution.

In fact, it looks like in Europe the solution that seems to enjoy more popularity among IPTV Service Providers is indeed wired networking. The good thing for companies like the one I work for is that of all the wired options (powerline, phoneline and coaxial cable), the number of phone lines and coaxial cables available in the average European home is too low to even be considered as a serious option by IPTV Service Providers.

Chano Gómez
DS2