Your browser is not supported. Please update it.

15 November 2006

China cancels mobile TV licenses, waits for homegrown system

One thing that has perhaps not been made clear with the emergence of China’s own mobile TV standard is that it has moved to actually shut down the other two mobile TV trials.

Samsung was previously helping with a Guangdong Province T-DMB trial, which was at advanced field test level, and Nokia had received permission for technical DVB-H trials. Those trial government licenses we are now told by suppliers operating in China, were actually withdrawn. This is so that the cellular community cannot move forward in a direction that would oppose the broadcasting regulator.

Broadcasting is regulated in China by the State Administration of Radio, Film and TV (SARFT) while the Ministry of Information Industry regulates telecommunications.

The two have found themselves struggling over the regulation of both IPTV services and now mobile TV. It’s not that either would do anything very different from the other, but their increasing overlaps make it harder to gauge if what they say is law or not. This is especially difficult where the regulator is not just looking out for local Chinese industry, but also has to be sure that it can set up a unified supervisory agency for content control and for censorship.

Censorship is tough in a market that is fragmenting and beginning to fire video at Chinese people from all over the place ‘ with DVRs, internet video, mobile TV, IPTV and a resurgence in cable all happening in China as much as they are happening elsewhere.

The cellular community would definitely like to take the easy option where mobile TV is concerned, and take on an international standard, where volume production and competition is driving down equipment pricing and where they know how the equipment performs in the real world. Instead they are faced with a new standard that is compromised in that it supports satellite but also has terrestrial transmitters, which is likely also to make it more expensive, and of course it is totally untried. And Beijing only gets one shot at holding the Olympics, so this system needs to be up and running by mid-2008.

Until this August it was expected to be a straight forward fight between DVB-H and T-DMB, with the smart money on T-DMB because of the more advanced level of trials. As it was, if both trials had been allowed to continue, then when the home grown technology looked like failing it would have been a simple matter of letting the mobile carriers off the hook and allowing them their own choice of technology.

But there is also the issue that the DVB-H spectrum which is favored the world over is in the UHF bandwidth up to 700 MHz which is occupied by analog TV services. In China analog TV will not get turned off until 2015 at the earliest, so very little, if any, of this could be made available in time for the Olympics.

What is now favored for China’s new mobile TV services are likely to be what they describe as satellite band or S-band. Again there is complexity here in that the traditional S-Band is in 2.0 GHz to 2.2 GHz, but China is likely to allocate that spectrum to 3G, leaving mobile TV from a satellite to come somewhere in the 2.5 to 2.7 GHz range, which takes us neatly into the area of TU Media, and its experience setting up S-DMB in nearby Korea, still the most successful mobile TV service in the world, which uses virtually identical spectrum.

Obviously a Chinese wide TV network would almost certainly have to have some portion of delivery from a satellite, in order to hit such a broad footprint that is not built up. And the only other company from outside China that has experience of this type of arrangement is Alcatel, which also has a great advantage of having supplied most of the country’s DSLAMs. The Alcatel proposed mobile TV service uses S-Band and is a combination of satellite and gap in fillers to get indoors coverage.

China could have taken the view that each city might offer different TV services locally, and national coverage wasn’t perhaps a necessity, but it hasn’t. So both Alcatel and Samsung are likely to treat this news as encouraging and once a Chinese standard is settled, may find adjusting to the requirements of the standard easier than other suppliers.

But what’s crazy about this supposed standard, is that it has yet to be defined and won’t be until very late this year, which makes stopping existing trials all the more shocking. The technology of choice that most people think will be at the heart of the standard seems to be called STiMi although there is a lot of confusion over whether this has actually been mandated or is just a candidate.

STiMi hails from the R&D arm of SARFT itself and stands for satellite and terrestrial interactive multimedia. It is said to be related to the Eureka 147 DAB standard and is OFDM based with its own coding. STiMi is said to support 7.5 MHz bandwidth slices and data rates up to 12 Mbps. More or less the same as DVB-H.

Whereas other reports suggest that there are in fact two systems, a digital terrestrial TV portion called ADTB-T (Advanced Digital Television Broadcast-Terrestrial) which uses a single carrier frequency and the mobile TV version using time domain synchronous OFDM with 4,096 carriers.
SARFT is clear that whatever the standard is, it must be homegrown.

China cannot invent totally new technology to do all the jobs in modern society. It will inevitably tread on a few toes in terms of intellectual property, and while its manufacturers may enjoy unencumbered use of technology inside China, when it comes to taking this technology to overseas markets, it will be almost impossible to avoid patent litigation.

When we have previously addressed this topic, and before it was obvious that the trial licenses had actually been cancelled, we thought that the normal Chinese outcome would be the result.

A Chinese only standard would be mandated and either fail or be too expensive to buy, and then companies inside China would get permission to ‘not’ use them, resulting in a two or three way stand off. We had thought that DVB-H and T-DMB would be in pole position once this happened.

Now the outcome may be somewhat different. Once the 3G licenses in China are sorted out, it may make a lot of sense for the cellular operators to investigate using multicasting on 3G spectrum. The content would still need to be regulated and censored by SARFT, and not MII, but the technology will be contingent only on the 3G technology actually working.

To follow a strategy of that type it would make sense for Chinese cellular operators to begin offering unicast services now, which they are already doing, and developing the interface and usage patterns. This traffic can then be pushed over onto advanced data services as they arrive and finally onto an efficient multicast broadcast system.

But it seems likely now that the sales window of the Olympics WILL be missed, and all that will be on offer come 2008 is likely to be unicast video, unless China breaks long ingrained habits and speeds its new technology to market. It would need trials over the next 6 months and then a year to build and test a commercial system. But how can that be done when there are no chips yet in production that support the new standard. That alone will mean a 12 month wait, and the Olympics is already missed.

In the meantime, both China Mobile and China Unicom are supposed to be on the verge of signing agreements with handset makers to facilitate the new standard, which is completely dependent upon chips being available, and so cannot happen. Non-Chinese manufacturers will almost certainly be needed to make this new technology work anyway and they will need time, probably a full two years.

One thing that we might see emerging, is a multi-standard chip, based on existing designs for DVB-H, MediaFLO, Japan’s ISDB-T and T-DMB, but with a new Chinese format added. All of this might be run in software on a DSP.

While it might burden the chip running that many standards and require various types of software or firmware, it could be done in time for the Olympics, but it’s still cutting it close. If that was managed, it can’t be with Chinese chip companies, unless they buy a license to existing DSP designs, and since avoiding such licenses is the whole idea, there would be no point.