IPTV – when will it all come together?

Exhibitions feature it; magazines write about it; whole conferences are dedicated to it. But not a lot of people are doing it. Indeed, we cannot even agree on what it is. Yann Courqueux of Grass Valley looks at IPTV

So let me start by trying to define IPTV. If we separate the IP and TV, then we get television delivered over internet protocol. As broadband capacities increase then technically we probably could deliver television over IP, but why would anyone want to?

Terrestrial, satellite and cable broadcasting have been around for a long while, work well and are well understood by consumers. There are closed communities where pure IPTV works – Cornell University in the USA has a service for its students on campus rather than install television cable into a historic building – but there has to be something more for it to become an attractive proposition for general audiences. What is this, and what are the barriers to implementation?

There are technological limitations at present, and these need to be addressed before anything can happen. Most obviously the bandwidth issue needs to be addressed, and linked to this is a quality of service (QoS) guarantee. As we move to HD, audiences are getting ever more appreciative of image and sound quality, and they are certainly not going to accept lower standards just to move to a different platform - even if it is potentially more accessible,

To deliver consistently good quality images and sound requires the transmission circuit to be reliable. Grass Valley, working with others including the Pro MPEG Forum, has developed a standardised two-dimensional forward error correction system which is tuned to the needs of delivering high bandwidth, time critical data streams – such as IPTV – over long and potentially noisy IPTV circuits.

Combine that with an efficient compression codec such as MPEG-4 (which is capable of delivering premium quality standard definition at under a megabit a second) and the potential for underwriting reliable QoS becomes much stronger.

Equally important is content. Audiences are notoriously resistant to change: the only way to drag them out of their inertia is to offer premium content on the new platform that is unavailable any other way. That content has to be offered in a way not currently available on any other platform – video on demand is the obvious example.

So, to make IPTV happen will require major investments in IP infrastructure, including final mile connections capable of consistent high speeds; plus a new sort of head-end; plus the development of a new set-top box; plus premium content to tempt large numbers of switchers. Who has this sort of money?

The most obvious answer is the telco. They are seeing revenues challenged by voice over IP (VoIP) and mobile telephony: could this be their salvation? There are a lot of attractions, particularly if they share infrastructure (and customers) across both fixed and mobile content delivery. Telcos also have customer relationship management systems in place, and are best equipped to deal with the new, interactive possibilities of IPTV and its linked services.

A potentially attractive deal is to supply a single gateway that bundles IPTV and VoIP. The inclusion of a telephony option helps to reduce churn.

For telcos looking to move into television delivery, one of the most attractive options is the interactivity made possible by the two-way nature of IPTV over telephony networks. It means that the operator knows precisely what each subscriber is watching, and whether it is scheduled television, timeshifted content using either a conventional personal video recorder in the gateway STB or a network PVR, or video on demand,.

There is a very precise return on the number of people watching any given programme, and when they choose to watch it, which means equally precise payments to content owners and a great deal of valuable information for advertisers. That in turn can be used to target advertising precisely, to minimise skipping of commercials. A commercial for, say, a car could be followed by a short and localised spot for the dealer nearest the viewer, with a contact name to call, what colours are in stock, special offers and so on.

Once a profile of the individual viewers is built up, advertising can be targeted to the individual STB level. Viewers who make a point of watching nature and environment documentaries would not be good targets for 4 x 4 car advertising, for example. Equally, compulsive viewing of children’s programmes would be ideal for toy advertising.

This is fine for new entrants to the market, but what about existing broadcasters. Can they use IPTV to extend their offerings, introduce new services and improve their own business model? Yes: very many of the interactive benefits available to telcos are also available to conventional broadcasters.

The solution proposed by Thomson is the hybrid STB. This appears to the consumer as a single device but internally it combines a traditional broadcast receiver – terrestrial, satellite or even cable – with special services over IP. Research suggests that even with a full IPTV service available 80% of viewers will watch a broadcast channel as scheduled, so it makes sense for this to be delivered using conventional broadcast techniques.

Most on-demand content purchases are on blockbusters which, with good customer tracking, can be predicted. You could pre-empt user demand by pushing likely purchases into the hybrid STB in advance, either by trickling the content over IP or by bit-shaving a broadcast channel to create space for the advance download. All of this minimises the load on the DSLAM (digital subscriber line access multiplexer) and controls bandwidth.

Targetted advertising can also be sent in the same way, with the added benefit that you are not trying to fit it into bandwidth reclaimed from the main programme. Local spots work in exactly the same way, and advertising linked to the programme can also be delivered: a hybrid STB tuned to the first half of a live football match could be collecting beer commercials over IPTV in the background.

It also makes sense for the hybrid STB to include a web browser so that it can be used to access internet TV services in the same way as true IPTV.

There is one fundamental challenge to the hybrid STB, or indeed any way of accessing IPTV that is not a closed and proprietary system: standardisation. For economies of scale, any new system has to be firmly based on open standards. It minimises the risks and financial exposure of both the manufacturer and the platform operator.

The Digital Video Broadcasting (DVB) model has proved successful around the world in delivering reliable digital television. It has evolved subsets of its standards to serve the terrestrial, satellite, cable and handheld markets.

An important part of the DVB system is to make the mechanics of tuning and channel selection invisible to the user. The STB scans all available frequencies and picks up the signalling information which is presented in a standardised form. On connection for the first time, a new STB very quickly identifies all available services and presents them to the viewer.

That does not work for IPTV: you cannot just scan the internet for signalling information. There needs to be a way of identifying server addresses and performing the log-in and selection in just the same transparent way. DVB is working on a set of standards to accomplish this: DVB-IP. This will produce normalised information such that the STB can provide the same seamless service that the audience expects.

Thomson is taking a leading role in the DVB-IP Forum, as the creation of standardised hybrid solutions, and standalone IPTV STBs, depends on open and accepted formats. Across Thomson there is proven expertise in developing all of the elements of the new service: broadcast equipment including signal routing, compression, and playout and VOD servers; IP devices and communications networks; customer management systems for IPTV; and set-top boxes and receivers.

IPTV has immense potential. This potential is available to telcos and others who seek to enter the market with innovative triple play and quad play (voice, data, media and mobile) services which can each build demand for the others. But it is equally attractive to today’s broadcasters – terrestrial, cable and satellite – who need to offer new interactive services and create new income streams, to thrive in today’s highly competitive market. Thomson is working with many interested parties – telcos, broadcasters and the rest of the supply industry – to create open standards and deliver services for the future.