Service providers worldwide are gearing up to offer mobile TV, but they face a plethora of technology choices that are ultimately going to be determined more by market forces and spectrum availability than by technological merit
Mobile TV services may still be in the early adopter phase, but service providers around the world are scrambling to map plans to capitalize on the latest developing revenue opportunity.
At the vanguard of the mobile TV push is South Korea, where satellite DMB services - launched last May by SK Telecom-owned TU Media - have generated around half a million users watching about 60 minutes of programming a day, including a 74%-26% mix of video and audio programming, notes Vikrant Gandhi, an industry analyst for Frost & Sullivan.
S-DMB is one of two competitive offerings in Korea, where regulators helped push deployment as a means of supporting local manufacturers. The S-DMB service costs about $13 a month and provides 14 channels of video, 22 audio channels and three data channels, according to information supplied by Samsung, which offers end-user devices to support the offering. Subscribers also must purchase a special phone or hand-held device. Models offered by SK Telecom retail for as much as $800. According to an estimate from the government's Electronics and Telecommunications Research Institute (ETRI), S-DMB should have about 630,000 users in 2006.
However, that's well short of initial targets - TU Media had originally projected one million subscribers by the end of 2005. Meanwhile, S-DMB has begun to lose ground to a competitive offering based on terrestrial DMB (T-DMB) launched by another industry consortium including several telecom network operators in late 2005. Although this offering provides fewer channels (six video, 18 audio and one data channel, according to Samsung), it is offered at no charge to end-users and - more crucially, as it turned out - has support from free-to-air broadcasters that produce the country's most popular TV shows. Here, too, users must purchase a phone, which runs in the range of $600. But the steep price hasn't prevented users from purchasing devices to support the offering, which should have 1.45 million users in 2006, according to ETRI's estimate.
The T-DMB offering, which operates in the 200-MHz spectrum range, is less costly to deliver than the satellite service, and the goal is to make it advertiser-supported. But as Gandhi notes, 'Service providers are finding it difficult to sustain service and fear that it will be long before they turn profitable.' As a result, he anticipates that the South Korean government may allow service providers to charge a small fee for the service.
Mobile TV is also on trial in Europe, where industry consortiums of broadcasters, telecom network operators and others have been trialing both T-DMB and DVB-H, a rival standard favored heavily in the EU. According to data supplied by Nokia, average viewership in three key trials in the UK, Spain and France was about 20 minutes per day per participant. In each of these trials, the majority of participants said they were willing to pay in the range of 10 euros a month for the service.
The US, meanwhile, seems to be leaning toward a wholesale approach to mobile TV. Qualcomm, Modeo (a unit of cellular tower operator Crown Castle) and Aloha Partners are each planning to deploy competitive networks throughout the US that will broadcast popular television programming to mobile phones, with cellular network operators acting as resellers of the service.
The upshot is that as operators assess the various technology choices in deploying mobile TV and subsequent business models, one thing they're discovering is that there's no one-size-fits-all solution, and any choice is loaded with challenges.
Offloading TV
Top of the list is the fact that the chief mobile TV technologies jockeying for mindshare - DMB, DVB-H, Japan's ISDB-T and MediaFLO - all require cellcos to deploy additional network infrastructure.
The reason, of course, is that doing mobile TV on 3G puts a serious load on the network, says Sean Koh, regional technical manager for multimedia and messaging solutions at Siemens.
'If you do streaming video over a 3G network, your cell shrinks,' says Koh. 'DVB-H lets you broadcast streaming video so it doesn't eat into your cellular bandwidth.'
David Glidden, director of television strategy and mobile television for mobile broadcast equipment manufacturer Harris Corp, offers this scenario: 'If you were to assume that subscriber levels get to the range of 15% to 20%, and that they're watching 20 minutes a day, on average, it would be a rather significant increase in the loading on a unicast cellular network,' he says.
That said, broadcast-based mobile TV presents its own set of challenges, says Mike Katz, product marketing director of video products for NMS Communications.
'If you go with a technology like DVB-H, you need new handsets, new spectrum, a new network, additional radio equipment, you have to have backhaul for the traffic, and you have to put all of the layering architecture in place, just before you can even think about offering a service,' he says.
That said, with dozens of trials of DVB-H and DMB underway around the worlds, there's some debate over just how concerned operators really are about infrastructure costs.
'The much bigger question for them is what will it take to get a broadcast network up and running‾' asks Nick Pilbeam, Asia-Pacific Networks and Enterprise MD for Motorola's Strategic Business division. 'For example, you need as many channels as possible, because people won't just watch what you give them. So if that's the driver on the revenue side, then you need a technology that can support it.'
Competing standards
As for which technology has the strongest offer, that varies more by environment than actual technological merit. Even though DMB technology has the advantage of being proven in Korea - and has generated serious interest in China and a handful of European operators - it has stirred little interest in the US, perhaps because the satellite version is too expensive and the terrestrial version offers fewer channels than US consumers desire. Qualcomm is, of course, using its own MediaFLO technology, while both Modeo and Hiwire are going with DVB-H.
DVB-H is also building traction in Asia Pacific via a regional initiative, the DVB-H Asia Pacific Alliance (DAPA), founded in June by The Bridge Networks, Indonesia's MECA, Malaysia's MiTV, Nokia and Intel, which aims to establish a DVB-H regional forum to support regulatory preparation for DVB-H adoption, as well as keep members briefed on technological updates and best practices.
Indeed, DVB-H - ostensibly a mobile offshoot of the international DVB video standard - is expected to dominate in terms of subscriber numbers in the next five years, with 121 million users accounting for 57% of the global market, according to Informa Telecoms & Media. That could also give it the edge in terms of economies of scale, says Scott Ramke, Modeo's VP of marketing and business development.
'At the end of the day, it's about embracing what looks to become a global open standard that will have the support of bellwether technology companies, which will drive scale,' says Ramke.
Qualcomm VP of engineering and market development Rob Chandhok claims MediaFLO will have a substantial edge over DVB-H because 'a FLO tower can deliver twice the distance or two times the number of services.' At the National Association of Broadcasters show earlier this year, Qualcomm demonstrated 16 channels of video using MediaFLO.
MediaFLO has gained limited traction outside the US, though that could change if Qualcomm's technology edge proves to be as powerful as the company claims. Chandhok points to a recently announced trial in the UK as evidence of MediaFLO's international support.
Spectrum sweepstakes
In any event, the fate of any given mobile TV technology is going to depend in no small part on spectrum availability. The US is something of a special case in that the government made appropriate spectrum available in the early part of this decade when the telecom industry was in a slump, allowing companies like Aloha and Crown Castle to obtain the spectrum at bargain rates and making the concept of a dedicated mobile broadcast network operator a uniquely American phenomenon.
In other parts of the world, regulators have been happy to parcel out spectrum for trials. But in many cases those regulators have not yet determined how to allocate the appropriate spectrum on a long-term basis. In some cases, the more desirable 700-MHz spectrum is committed to analog television for quite some time.
'One thing that helped in the US is the knowledge that analog broadcasting is going away in 2007,' notes Glidden from Harris Corp. 'Very few places in the world have hard dates anywhere near that. Some in Europe have it scheduled for 2012.'
In some countries, there is unused spectrum available around 200 MHz or 1500 MHz, which was traditionally earmarked for digital audio broadcasts. But whichever spectrum new-market entrants decide to use, it may command a higher price than what was seen in the US simply because the business plan for using that spectrum now appears so much more promising. As a result, in competitive markets outside the US, perhaps only the wealthiest companies - traditional telecom service providers or broadcasters - will be able to own and operate their own networks.
As Glidden points out, 'In some environments, broadcasters and mobile operators have common ownership,' making them a natural choice for owning and operating mobile broadcast networks.
As an alternative to acquiring new spectrum, some wireless network operators in Europe and parts of Asia have unused spectrum that they obtained as part of the 3G spectrum allocation process. Some equipment developers, including IPWireless, are betting that network operators will choose to deploy an emerging technology alternative that can use that existing infrastructure (see sidebar, 'Using what you got', p.
18).
'Most 3G operators own at least 5 MHz of TDD spectrum that they're not using,' says Jon Hambidge, marketing VP for IPWireless. 'If you look at this from the operators' point of view, how would you feel if after all that 3G spectrum you bought, the killer app for 3G requires you to buy new spectrum‾'
IPWireless expects to have equipment available in mid- to late-2007 to support the multimedia broadcast multicast service (MBMS) standard that the 3GPP has defined for W-CDMA Release 6, which can use TDD spectrum and is well suited for operators that already have deployed 3G. The IPWireless offering, dubbed TDtv, can be deployed on a network operator's existing tower infrastructure, and purportedly deliver up to 50 channels per 5-MHz block of TDD spectrum (assuming 100 kbps per stream).
The other benefit, Hambidge adds, is that it doesn't require users to buy all new handsets. 'You don't want to have to force the subscriber to buy a new handset just for this service. With TDTv, you're looking at a $10 maximum upgrade in handset costs.'
Although MBMS has been defined, it's not expected to be finalized before the end of 2007, although vendors like IPWireless and Huawei Technologies are selling prestandard MBMS solutions. In May this year, Hong Kong's PCCW launched a mobile TV service using Huawei's cell media broadcast (CMB) technology, which is based on MBMS.
Fractured landscape
With many technological choices facing operators, an inevitable question is: to what extent should cellcos be worried about interoperability‾
The prospect of a harmonized mobile TV standard - at least at the RAN layer - is decidedly slim, according to Toni Paila, broadcast chair for the Open Mobile Alliance.
'There's not likely to be a harmonized RAN standard for mobile TV because of various regional issues like spectrum availability,' Paila says.
Consequently, the OMA is focusing more on developing standard specs at the service/application layer of mobile TV. 'The usage case we see is to combine DVB-H and MBMS at the service layer so that consuming mobile TV services is transparent to both the user and the network.'
Paila says the three critical components to look at in developing an apps-layer standard are the program guide, content/service protection and content delivery. Smaller items to look at include specs for purchasing services, roaming, device management and interaction.
Paila declined to give a timetable for the arrival of such a standard, as it's early days for mobile TV in general. In the meantime, the next few months are likely to see a flurry of activity as service providers weigh their options and place their bets on mobile TV.
Whatever technology operators choose, advises Motorola's Pilbeam, they should tailor the network's design and capacity around how users consume mobile TV content. 'Mobile TV was originally envisioned as a primarily outdoor service, but in fact trials suggest that people are more likely to use it indoors. Also, unlike traditional broadcasting, the antenna is moving much if not all of the time.'
Lara van Rooyen, marketing manager at Ericsson, agrees.
'Only a small percentage of users will watch the usual broadcast channels - the rest are more likely to watch more specific, on-demand content that's driven more by menus than schedules. It's a whole new way of watching TV, and operators must plan their networks for that.'
Mobile TV 30-second primer: unicast vs broadcast vs multicast
Unicast: point-to-point video streamed separately to every terminal in the cell requesting mobile TV. Whether video is delivered via the 3G data channel or the circuit-switched video call channel, unicast means each user eats up more bandwidth that can get eaten up quickly if too many viewers try to access the service at once - which also impacts access to other services on the same network.
Broadcast: point-to-multipoint video transmitted only once over the entire service area. Users view the same stream without requiring additional bandwidth per user. However, because mobile TV service requires a wide selection of channels that must be broadcast all at once, it requires a wide swath of bandwidth - one reason why broadcast-based technologies like DVB-H and DMB require separate spectrum. Broadcast is also one-way, which means interactivity must be done separately.
Multicast: similar to broadcast in terms of viewers sharing the same stream, but instead of sending the signals constantly over the entire coverage area, multicast delivers only to those parts of the network where users are tuning in, which makes for more efficient use of spectrum.
The lure of advertising
- Joan Engebretson
In addition to sorting out which technology or wholesaler to use to support their mobile broadcast television deployments, carriers and other stakeholders also will need to resolve how to pursue advertising opportunities and apportion any resulting revenue streams.
'The real wild card is advertising,' says Bob Shallow, director of multimedia experiences for Nokia. Shallow notes that most television programming in the US includes a local commercial every half-hour that is treated in different ways on different platforms. Cable television operators, for example, use it to sell interconnects, while DirecTV uses it to promote itself.
'It's not clear how this will be handled in the mobile space,' says Shallow. 'It's safe to presume that terms will be negotiated between the broadcaster and the cellular operator about who controls what portion of the investment.'
Bryan McGuirk, president of the media solutions group at SES Americom, which is providing infrastructure to support Aloha Partners' Hiwire mobile television network, had a previous career in advertising and is particularly enthusiastic about the possibilities. He notes that most consumer package goods manufacturers today spend the majority of their advertising budgets on tailored localized advertising and promotions.
'That's what we'll be able to tap into,' he says. 'The mobile device is perfect to make advertising relevant and timely. Advertisers will want to take advantage and even pay a premium for that.'
Because mobile broadcast television will be deployed over dual-mode mobile handsets, it will be relatively easy to add an interactive component, such as the ability to immediately contact an advertiser in response to a commercial.
'The technology will enable advertisers to target a specific kind of advertising to a limited number of users in a way that a normal broadcast network wouldn't be able to do,' notes David Glidden, director of television strategy and mobile television for Harris Corp.
Using what you got
- John C. Tanner
For operators pondering mobile TV, the options for deploying services are increasingly splitting into two camps - those that require all-new infrastructure and spectrum, and those that utilize existing 3G networks.
Mobile TV technologies like DVB-H, DMB and MediaFLO have been touted as more efficient ways to deliver mobile broadcast video because of the impact of high-quality video on 3G bandwidth capacity. However, with operators looking for less expensive options than building a new network for mobile TV, solutions are emerging that leverage existing network infrastructure.
Ericsson, for example, offers a 'Unicast' solution that allows 3G networks to run streaming video at data throughput rates of 100 kbps. Alcatel is proposing a hybrid satellite/terrestrial solution similar to S-DMB, except that it uses the S-band on existing satellites rather than requiring a new one to be launched.
One of the more innovative solutions being touted by companies like NMS Communications is using the 3G-324M standard for circuit-switched video calls to deliver mobile TV. Hong Kong CSL, for example, is offering 25 channels of 64-kbps TV using a 3G-324M-based solution from NMS and Golden Dynamic.
The advantage of using 3G-324M, says Mike Katz, product marketing director of video products for NMS, is that it uses existing capacity and - crucially - doesn't require a new handset.
'Existing 3G handsets are already compatible with 3G-324M,' he says. 'Menus are created on the fly, so you don't need a new client on the handset, and you don't need a second antenna like you do with DVB-H.'
Just as crucial, he adds, is that 3G-324M supports interactivity, one of the key elements of any new TV service.
'Interactivity allows you to build a viewing community around programs, to include allowing people to upload their own videos,' says Katz. 'We're already seeing that with YouTube, YouSeeIt, RocketBoom - that long tail of content where people upload their own videos and viewers vote on the ones they love or hate.'
However, such mobile TV solutions tend to be unicast, which is bandwidth-intensive and faces serious scalability issues in the future as subscriber bases grow, says Nick Green, proposition manager for mobile TV and broadcast at LogicaCMG.
'That's damaging not only because it could deny mobile TV service to paying subscribers, but it also impacts other 3G services that may actually be higher-value services than mobile TV,' he says.
That's why the 3GPP is developing MBMS (multimedia broadcast multicast service), which will give W-CDMA networks armed with HSDPA the ability to run multicast video. Multicast is widely viewed as the most efficient way to run mobile TV in the long run, says Jon Hambidge, marketing VP for IPWireless, because it 'allows users to share bandwidth - say, a 300 kbps video stream - so that not every user is stealing bandwidth from each other the way unicast does'.
Jeff Belk, senior VP of Qualcomm, agrees that the future of mobile TV belongs to multicast. 'If you're in London, that Arsenal game is going to want to be watched by a lot of folks, so you'd better be multicasting that or it's an inefficient way of delivering it. However anyone does it, you still need to do it.'