We have said it before, but it is worth saying once more, that the Alcatel DVB-SH approach to implementing mobile TV may be late to the table, but it may indeed leap past all the other technologies that are chasing the nascent mobile TV market.
This week Alcatel got a cellular operator, the SFR French affiliate of the Vodafone group, controlled by Vivendi, to confirm publicly that all the assumptions about the cost of implementation of such a service are now borne out, after extended trials in France, which ended in December.
Alcatel claimed that a DVB-SH network was 40% cheaper to build than the equivalent DVB-H or other terrestrial only mobile TV network, making the claim when the concept was introduced 18 months ago.
Alcatel has already convinced satellite partners SES Astra and Eutelsat to get a satellite in the air by 2009 to push out services across Europe and despite the fact that this will cost some $166 million, and the fact that far more terrestrial repeaters per square kilometer will be required in S-Band 2.2 GHz rather than UHF 700 MHz, it now says that not only are its sums right, but that a prominent cellco agrees with the calculations.
The reasons are complex, but believable. The gap filling repeater transmitters will need to be installed not at new sites, nor at terrestrial TV towers, but at existing cellular towers because the behavior of 3G spectrum and S-Band is virtually identical in terms of in-building penetration, reach and propagation.
The system allows re-use not only of cellular sites, but also of the tower and existing antenna. The repeaters are low power and by using multiple internal antennas in the handset, which Alcatel says gives it 6 dB of gain, it can achieve the penetration it needs by merging both signals, the satellite signal blasted to a Europe wide footprint using a 12 meter antenna, and the terrestrial repeaters. It is the fact that new tower rights are not needed, and that power and antennas are already in place that make the savings, the company says.
The European satellite will target France, Germany, Italy, Poland, Spain, and the United Kingdom and terrestrial sites can be built ahead of the satellite signal going live in 2009.
There are 30 MHz of S-Band spectrum available right across Europe, licensed on a first come first served basis, and Alcatel has always preached that by taking half that, just 15 MHz of spectrum, it can deliver some 45 mobile TV channels, each at an aggregate rate of 256 Kbps.
Originally it said that the way it would do this was to break these channels into 3 different channels and encode and deliver them differently. Now it seems to have changed its tune slightly and it says that the satellite signal will be identical to the terrestrial repeaters and reinforce one another, but if that’s true it would still mean 20 plus mobile TV channels and there is another 15 MHz of spectrum still to use.
But by electing to use similar spectrum to cellular systems, there is virtually no radio planning required, as all the transmitters can be placed on the same towers, using the same cabling and same antenna systems that are already in place as those used by cellular operators. Alcatel is promising that each base station needs the addition of a €15,000 transmitter plus the monthly rental for the satellite which is a one off price per country.
Current rates would suggest that the entire satellite costs would be in the region of €50 to €60 million a year, and if this system is taken by 6 separate countries, it could mean €10 million a year for a single country, shared among 3 or 4 operators. In a country like France, where mobile TV penetration is likely to be 4 or 5 million handsets in just a few short years, this would amount to a cost per handset of a few Euros per customer per year, for the satellite costs, virtually nothing.
In dense urban areas there are far too many base stations, installed to cope with the large amount of voice traffic, so since DVB-SH would be a broadcast signal, only a few of these would actually require a transmitter. Whereas in rural areas, the satellite signal would be enough on its own.
The two companies jointly said that the tests validated fundamental assumptions on the performances of a DVB-SH network, notably the coverage, continuity and quality of service and they allow mobile TV coverage inside buildings identical to 3G coverage.
Now over the past year we have talked to several lifetime experienced ‘radio’ people who have told us that this was simply impossible and that DVB-SH is really like Satellite Radio, and would achieve no in-building penetration and was purely for cars with rooftop antennas. It appears that each of these people, some of them quite respected, were saying what they hoped was true, not what they knew to true, because each of them have a vested interest other areas of mobile TV.
Of all the potential target countries for the European satellite (there is also a US satellite going up) France, Germany, Italy, Poland, Spain, and the United Kingdom, this is a non-decision for the UK, where there is absolutely no UHF spectrum and where 40 MHz in 1.4 GHz is being auctioned off instead, with no UHF until 2012. Attempts to unearth one or more UHF channel will all take too long, including one to put DVB-H in Channel 36, and re-site the aircraft radar which currently uses it, because this would need to be re-housed and the process paid for.
But if Vodafone in the UK hears about the success of its DVB-SH pilot in France, it could be an early candidate, given that the only other mobile TV services likely to be launched in the UK before 2012 will be cellular based MBMS or in that odd L Band spectrum that the UK regulator wants to saddle on someone which would potentially double the cost of a network build.
Elsewhere in the satellite footprint, parts of rural France, Germany and Spain are unlikely to have any mobile TV before 2010, while Poland’s operators may be convinced to try DVB-SH alongside the more expensive DVB-H service which it intends to get working in major cities later this year. The software stack inside a DVB-H handset is likely to be identical to DVB-SH, and an additional radio and demodulator could mean the two systems could co-exist on a handset quite easily.
Tests using a helicopter to emulate the satellite part of the DVB-SH network also confirmed that the signals coming from the satellite combined well with those coming from the terrestrial repeaters, in a transparent manner, at the terminal.
The pilot used pre-commercial low-power terrestrial repeaters for the broadcast part, co-localized with several SFR 3G sites. The broadcast signal from a satellite was emulated by a transmitter located in a high-altitude at a fixed point. The mobile terminals used were pre-commercial terminals provided by Sagem and Samsung, using antenna diversity.