Bullpen: Multi-Platform TV

More people are watching video on personal computers, portable media players, game consoles, Blu-ray players, mobile phones and so on. This is being driven by changes in TV viewing habits. Just how much are they changing? Well, quite a lot, according to Jeff Cole, who is director for the Center for the Digital Future at USC Annenberg School. To quote their recent study (see www.digitalcenter.org):

“A revealing finding about the perceived importance of the Internet for entertainment is the dramatic increase over a short period of three years. In 2006, only 21 percent of Internet users aged 17 and older thought of it as an important or very important source of entertainment for them, but by 2008 that number had climbed to 64.2 percent. That represents a threefold increase. The dramatic increase probably has to do with the growing popularity of video-sharing sites and the ever-increasing availability of free entertainment content online.”

This trend is fueled by the growing number of “digital natives,” consumers who have grown up with computers, easy access to the Internet, and, most recently, YouTube and Hulu. The term digital native has been coined by an interdisciplinary collaboration of the Berkman Center for Internet & Society at Harvard University and the Research Center for Information Law at the University of St. Gallen. (See www.digitalnative.org.)

 Digital natives are characterized as the 12-24 age group, who:

 Will never read a newspaper
 Will never own a landline phone
 Will not watch television on someone else’s schedule
 Want everything on the go (mobile)
 Watch less television than previous generations

Watching TV

This last one depends entirely on our definition of “watching television.” Conventionally, we have limited our definition of watching television to include linear or on-demand content viewed on the traditional TV set. This content could be received via broadcast TV (that is, over-the-air) or via TV service delivered by a cable, satellite or telco operator.

Today, there are more and more ways to watch television without going anywhere near a TV set, and digital natives are watching a lot of video by:

Sharing video online
Visiting entertainment video sites
Watching television on mobile phones

Although viewing on the TV set may be declining because of leakage to the Internet, TV viewing as a whole is increasing. And that means operators have a great opportunity to increase the value and stickiness of their services by offering TV entertainment across multiple screens. But, far from being a replacement for MPEG-2 over quadrature amplitude modulation (QAM), this should be viewed as an additional delivery mechanism to support these new devices. Meanwhile, MPEG-2 over QAM will continue to function as a wonderfully cost-optimized architecture for video delivery to the millions of deployed set-top boxes.

We need a term for TV that can show up on any device—from a traditional set-top box to a netbook to a Wi-Fi-enabled iPod Touch. I’m going to use multi-platform TV (mpTV) to describe the union of all these delivery mechanisms, device alternatives and viewing modes:

Delivery mechanisms—broadcast, satellite, cable, IPTV, broadband and wireless
Viewing devices—TV, PC, mobile phones and media players
Viewing modes—linear, on-demand, digital video recorder (DVR), time-shifted and interactive

Differences

What is technically different about delivering television to a broadband-enabled device, compared to an MPEG-2 set-top box? Well, just about every layer in the protocol stack is different from currently deployed digital cable.

Transport protocol: There is already a lot of Internet protocol (IP)-delivered video in the headend in the form of IP-encapsulated MPEG-2 transport streams. However, native IP clients, with more buffering, do not require a constant delay network and a constant bit rate (CBR) channel. They use IP delivery protocols (for example, HTTP, RTMP) that were first developed for the Internet (and actually work even better over a managed IP network). Recently, Adobe and Microsoft have made announcements around Flash (smooth streaming) and Silverlight (adaptive streaming), respectively, that support on-the-fly changes in bit rate without the need for re-buffering.

CODEC: Obviously, satellite has moved to MPEG-4 aggressively as it was more bandwidth constrained than cable (no switched digital). Newer devices are more likely to support MPEG-4, especially since Adobe added support for it (and AAC) to its Flash player. In general, the processing vs. bandwidth equation has tipped toward more-processing/less-bandwidth because of the inevitable march of Moore’s Law. (While it is possible to find the VLC plug-in to play MPEG-2 on your PC, most subscribers aren’t going to jump through that hoop.)

Signaling (for VOD): DSMCC LSCP is replaced by RTSP.

Encryption: Broadcast methods based on a hard token (secure micro) are replaced by Internet-based security that relies on a two-way, challenge-based scheme. A two-way connection also tells an operator how many instances of a single account are active at any time. This is a requirement when the service isn’t tied to a physical access port in the network.

Presentation: Usually, we think of the electronic program guide (EPG) when talking about the presentation layer, yet the user interface is much less constrained when you have a keyboard and mouse. Also, sophisticated, multi-media authoring tools make creating the user interface a whole lot faster and cheaper. This is epitomized by the example of Adobe Flash and the Creative Suite 4 authoring environment.

Most of these options are driven by a mix of open (for example, IEEE) and proprietary (for example, Microsoft and Adobe) standards. To play in this environment, the service provider typically has to pick a platform. For example, Hulu and YouTube chose Flash, and Netflix and Amazon chose Windows Media Player (Silverlight).

Options

What technical options are available to operators to get into mpTV? The two main options are to (A) leverage the existing on-demand infrastructure and (B) build a new silo to supply the new IP-enabled devices. I’m going to make the argument for option A, which has a great number of advantages, including:

Re-use existing content distribution and asset management systems
Leverage low-cost server infrastructure
Benefit from server load-sharing across multiple platforms
Leverage existing back-office interfaces
Use existing subscriber management and billing systems
Support session-shifting, allowing you to pick up where you left off when switching from one device to another

To leverage the on-demand infrastructure, it needs to be made “device-aware”—essentially the on-demand back-office needs to be able to support the correct format and delivery mode required by each requesting device.

So to recap, multi-platform TV is already big among digital natives, and it will only increase among the consumer population at large. This provides an opportunity for service providers to increase the value of their offerings. There is no reason to build new silos and flush what works down the drain. Instead, by leveraging the existing on-demand infrastructure, service providers can expand into mpTV today.

Michael Adams is vice president, applications software strategy, for Tandberg Television, Part of the Ericsson Group. Reach him at [email protected].