Your browser is not supported. Please update it.

21 May 2008

News Corp could be the hidden agenda for Qualcomm’s UK spectrum buy

All the coverage surrounding Qualcomm’s £8.3m ($16.4m) successful bid for a 40MHz chunk of UK L-Band spectrum (1452MHz-1492MHz) has the casual observer baffled. No-one seems to understand why the company would not only spend $16m buying the spectrum – when no other major company was even prepared to bid – but exactly what it can do with it.

Qualcomm watchers will not only be aware that the company has waged an obsessive war against DVB-H, pushing its own MediaFLO mobile TV system, a similar but updated design, but will also be aware that Qualcomm has historically proven entirely new forms of communication by buying spectrum, creating the technology that will use it best, and then using that as a demonstration platform to go and conquer the market. That is the basis of how CDMA came to market, establishing the company’s gigantic intellectual property business in cellular.

In the context of the billions of dollars that Qualcomm hopes to make $16m is a tiny amount of money, especially if for that, the chip giant can defend MediaFLO and give it a basis in Europe ‘ if that’s all that it has to spend. But of course it now has to decide whether or not to build a MediaFLO network and decide on the size and scope of such
a project. Qualcomm is reported to be spending $800m on MediaFLO in the US, and last quarter spent some $76m in the US, building out more and more markets for a service that is supposed to be taken to market by both AT&T and Verizon, the top two cellular operators there.

The big problem with that service is none of Qualcomm’s fault. It is not well marketed by Verizon (AT&T has only just launched) and the content offering that it has been allowed by the major US content owners leaves a lot to be desired, with Fox and Disney’s ESPN and other properties all allowing only their pre-digested, made-for-mobile content profiles on the system. MediaFLO in the US would take off tomorrow if it were a free-to-air service with better content, but it is a paid service with short form content ‘ as we say, it’s not the technology’s fault. AT&T has been able to convince Sony to launch a film
channel on MediaFLO, which may make some difference for its own offering.

What Qualcomm did in the US was take the pain of getting involved in mobile TV away from the cellcos by building out a network, negotiating for content and sourcing the devices. It has conducted two trials in the UK of MediaFLO technology, working with
News Corp’s BSkyB, and there are obvious strategies that it might adopt to cement that partnership. For instance, it might build out a major city like London, or even several
Cities, with MediaFLO transmitters and then use them to deliver a huge chunk of BSkyB programming in mobile TV. Cellular operators, frustrated that there is no 700MHz spectrum, thought to be ideal for DVB-H, available in the UK, might join in as a route to market ‘ especially Vodafone and O2. The other large UK operators, Orange and T-Mobile, have merged their TDD spectrum and entered a commercial mobile TV trial using TDtv from Nextwave.

The important thing would be to offer full programming from Sky, not made-for-mobile snippets, which although fine for cellular streaming snacking, are known not to take off in high enough numbers for full mobile TV. Sky might pay for much of this in reduced TV churn, and choose only to charge non-Sky subscribers. The prospect of picking up more device business from Vodafone and others, might lead Qualcomm to
keep device costs down, in partnership with its usual Korean allies, Samsung and LG, which both have MediaFLO handsets. BSkyB has worked with all the major consumer electronics manufacturers in providing set-tops, and has its subsidiary NDS to solve any
insurmountable technology issues such as digital rights management protection.

The bouquet could be enormous. The 40MHz spectrum slice is supposed to be split into one high power 12MHz segment and multiple 1.7MHz slices in slightly lower power, with two of these segments used as a guard band between them. But since this spectrum has not gone to multiple players, MediaFLO could now apply for a waiver to allow high power transmission across the board, perhaps shifting those guard bands to the outer edges of the spectrum. That would have the effect of there being over 36MHz of usable spectrum, which it can carve up any way it likes to fit current transmitter design ‘ say into three 12MHz chunks. Each of these might allow MediaFLO to deliver 40 TV channels, and many of the existing Sky TV bouquet could eventually be accommodated.

In the US a small city, relying on just a couple of 5 kW 700MHz transmitters plus 10 to 20 smaller gap fillers, could reach around 1,600 square kilometers. The equipment might cost $1m and the cost of civil permissions and truck roll might double this. In LBand
the price can immediately double, because the spectrum has less reach. However, fitting this out in low power transmitters, although making the problem more complex, might actually reduce the cost. Instead of one big powerful transmitter and gap fillers which
repeat the signal, a satellite system (that’s what Sky does best) could be set up to feed lots of synchronized transmitters at the rooftop level. The roof rights for this particular type of transmitter cost far less, but there are some penalties in continued opex.

One of the areas of supremacy that Qualcomm boasts of over DVB-H is that it has a 4dB advantage, which means that for a given power level, at an adequate signal-to-noise ratio, it can get by with transmitters being further apart. The DVB project disputes this,
but the benefit is delivered due to the turbo codes used in the outer forward error correction, which are more modern than those used in DVB-H. If we accept it for a minute then there is every chance that a strong in-building MediaFLO signal might be achieved with only $3m or $4m of cost for each town, even in L-Band. So while it might
cost $400m-$500m to completely cover the UK with an L-Band MediaFLO network, reaching perhaps 50% of the population might be achievable for a fraction of that, concentrated as they are in the a handful of cities.

If anyone knows how to do this in the UK, Arqiva and National Grid Wireless, which ran the two MediaFLO trials and which run most of the broadcast transmission networks in the UK, should. There are no actual power limits attached to the license, only a provision for putting no more than 150 transmitters in any 50km x 50km area, which we think MediaFLO could get by with.
A statement from Qualcomm said that the spectrum would allow it to experiment with a number of innovative wireless technologies and, given that UK regulator Ofcom has made this spectrum tradable and technology agnostic, it can try out lots of things with the parts of spectrum that it isn’t using at any point in time. It could use MediaFLO
as a basis for music services ‘ or run a US technology like HD radio or a European technology like DAB. One of the other bidders was known to want to offer satellite radio
services and the UK spectrum is roughly the same that Dish in the US is using from a satellite for a trial of DVB-SH mobile TV with Alcatel-Lucent.

The reason this may work well for Sky is that it has virtually saturated he UK with almost 9m households out of 24m taking its paid satellite TV service. Its quarterly additions have shrunk and shrunk and it has now had to embark on bundling with broadband and VoIP services to continue to deliver growth. The next step might be to take on an MVNO and it has a close relationship with Vodafone and could offer a device of its own, complete with a Sky TV subscription on the handset. Such a move would give it a huge value add, and a full quadruple play ‘ and we should never be surprised at the audacity of a News Corp-owned company, especially in a territory where it is threatened by breakthroughs like Freesat (the new free satellite TV service in the UK which threatens to have 200 TV channels by Christmas).

There are many other scenarios possible, and even a major breakthrough in the UK for MediaFLO would not necessarily see it flourish elsewhere in Europe. There would certainly be no TV roaming, since this spectrum just is not available in the type of size elsewhere, which would mean that to repeat the strategy at Sky Italia or at its other properties in South America and Asia Pacific, the entire service might have to be thought out again from scratch.

Just what EU telecoms commissioner Viviane Reding will make of a technology which is a direct rival to Europe’s recommended standard, DVB-H, being deployed in the UK, is anybody’s business. But the key to this is will she or can she legislate against it? We don’t think so. There is a trend towards technology neutral spectrum allocation and Ofcom would have to repay the auction price if this spectrum were not genuinely neutral. But then again, if making DVB-H a preferred standard fails at the first challenge, then it bodes badly for the rest of Europe in mobile TV. In the end, it will be commercial interests which Qualcomm manages to leverage, both among broadcasters like Sky, and among cellular operators, that will decide whether this spectrum is played around with a little and then sold off, or becomes a springboard for broadcasting and an element in a whole new formula for quad play for satellite TV suppliers.