Your browser is not supported. Please update it.

1 November 2006

Alcatel pushes case for S-Band TV with Samsung, satellite partners

Alcatel gathered powerful partners around it for its late play into mobile TV this week, with Samsung agreeing to make DVB-H style handsets which work in the S-Band satellite spectrum that Alcatel favors, and with the announcement that partners SES Astra and Eutelsat will launch a satellite especially to push out services across Europe.

The €130 million ($166 million) deal for the satellite and associated equipment means that Alcatel and partners will be able to offer services right across the European footprint from early 2009. The joint venture company will be set up by SES Astra and Eutelsat following approval from relevant regulatory agencies that in-fill transmitters can be used on the ground in the satellite frequencies, something that is thought to be a formality.

This is a system very much like the marginally successful S-DMB from TU Media in Korea. TU Media is a joint venture company and operates in a 50 MHz band from 2.605 to 2.655GHz sprayed from a satellite that sits halfway between Japan and Korea and available to serve both countries. It is run by South Korea Telecom (30%), Toshiba (10%) and a mixture of local Broadcasting companies, Samsung, and Renault-Samsung Motors and has 11 video channels, 25 audio, and three data channels.

It is on track to win roughly 650,000 subscribers in Korea by the end of 2006, remaining ahead of the Government supported standards based T-DMB system, which has more like 400,000. However the S-DMB service has fallen foul of politics and so has less content on it than perhaps a similar Alcatel-inspired service would have.

The Korean government backed the terrestrial T-DMB as a standard and offered up appropriate 700 MHz spectrum for this which makes it far cheaper to install, but imposed the condition that it had to offer ‘free to air’ content. As a result broadcasters love T-DMB, grant it content and look to sell advertising on it, and operators prefer instead the S-DMB because it give them an opportunity to raise their ARPU.

S-DMB was supposed to win 700,000 customers by the end of 2005, and 8.5 million by 2010, so it is just over a year behind schedule and likely never to reach the 2010 target. The reason is that the service was expensive to build, which makes it expensive to buy, and TU Media is thought to need around 5 million customers to break even.

But the S-Band services which Alcatel plans, though building up a head of steam, means that the Korean experiment will give pointers to the performance and cost of building out S-Band services in 2.0 and 2.2 GHz, which will behave more or less the same as S-DMB.

The big problem with anything from a satellite is that signals in this spectrum cannot get through more than a couple of walls, and if it is being beamed from above, might make it to the top floor of a building and perhaps the one below that, but will soon run out of penetration inside buildings that have more than two floors.

Permission needs to be given by regulators to build ground stations that will offer in-fill radio that runs parallel to the ground and which can therefore enter buildings like cellular services, through windows or one outside wall.

Alcatel and SES were claiming this week that S-Band would make mobile TV networks between 40% and 50% cheaper to build than comparable terrestrial only services.

Olivier Coste, Alcatel’s mobile broadcast activities president said this week ‘We have done in-depth analyses with mobile operators and terrestrial broadcasters to compare the cost of rolling out a good mobile TV solution ‘ that is, one with good indoor penetration in most urban areas as well as nationwide coverage ‘ using UHF and the S-band,’ says Coste. ‘Because you have to build out a new network from scratch with UHF, the capex cost per TV channel can be 40-50 per cent lower with the S-band approach.’

We think that the Korea experiences shows that he is largely wrong and estimates in Korea suggest that the entire S-DMB service costs could go way over $500 million, with something like 8,000 gap filler transmitters contributing about $340 million of this cost, the lion’s share.

There are two advantages to using S-Band, and the main one is that there is a unified piece of spectrum available across Europe in the 2.2 MHz range. The other is that when any given European cellco or broadcaster wants to launch services in their country, they can start selling phones right away for outdoor usage, and add the terrestrial repeaters slowly, town by town, as they gain revenues. These can almost certainly be placed on existing 3G towers, because the cell sizes will be similar to cellular services. DVB-H, depending on the spectrum used, can use some cellular towers, but its cell size is so much bigger, and therefore the network costs so much smaller, that it won’t necessarily fit in tightly with existing cellular tower locations.

The reason that Samsung is the obvious choice for the phones is that it has already contributed S-DMB phones, along with LG Electronics and about 16 other suppliers, to the TU Media service in Korea and would only have to alter the internal radio and antenna. This should be simple for Samsung, since during the past 18 months it has promised or delivered S-DMB, T-DMB, DVB-H and MediaFLO phones. Samsung is as agnostic as you can get and will partner for phones with anyone it can, standard or no standard, to gain an edge in supplying handsets.

Alcatel and Samsung have agreed to collaborate in testing their technologies for interoperability and will push for S-Band to be part of the DVB set of standards as the DVB-SSP (Satellite Services for Portable devices) Ad-Hoc Group. They will then market their combined solution, first targeting Europe. This might be seen as something of a turnabout for Samsung that has pushed hard for T-DMB to be a European standard, but it is consistent with its totally technology agnostic nature.

The first satellite, not due up until 2009, will actually be provided by Alcatel Alenia Space Systems, a company that was, in the most part acquired by both Italian Finmecanica and French Thales last year, so no longer controlled by Alcatel.

This new satellite, the W2A will be operated at 10 degrees East and can be used bi-directionally and doesn’t only have to be used for mobile TV services. In fact if no-one in Europe launches an S-Band mobile TV service after all, this satellite is unlikely to be a financial disaster, because it will end up being used as a general purpose wireless distribution system for any portable device, right across Europe including communication with phones, PDAs, laptops and vehicle receivers.

It could probably justify its cost purely on specialist vehicle TV receivers and could be used to send the same signals that are on a DVB-H service, to in-car devices (although DVB-H proponents have gone to great pains to prove that DVB-H can also handle this). In fact it could even be used as part of the advertising and program distribution network for DVB-H and other services, with S-Band receivers placed at DVB-H transmitter sites.

The satellite companies were careful to say that the S-band payload has been optimized for a broad range of business applications such as security surveillance and other commercial data services including two-way communications. So the spare capacity that does not get used up for mobile TV, can be used in the same way existing Ku and C band satellites are used.

Alcatel says it is in discussion with several operators’ to start trials of what it calls Unlimited Mobile TV during 2007. We’re not quite sure what kind of trials these are likely to be, largely terrestrial ones, we suspect, since the last time Alcatel showed off its system it had to use a helicopter to simulate the satellite signal because there is no S-Band satellite it can rent space on yet. This trial was in Toulouse with Orange and the French Space Agency CNES.

By the time the European S-Band satellite is in operation there are likely to be established DVB-H, T-DMB, DAB IP, Cellular multicasting and even MediaFLO services already in operation, with more about to be launched using freed up analog TV spectrum, so it really will need to have a big cost advantage over all of them as it will be very late to the party.