21 August 2006

Mobile Digital TV

Mobile TV is widely tipped to be one of the killer mobile services that will appear on handsets in the next few years. Mounting evidence demonstrates that consumers are prepared to watch and pay for video content on mobile devices. Already, the 5G video iPod is the best-selling MP3 player in the US.  What’s more, over 4 million Disney video clips have been downloaded from iTunes. 

Mobile operator 3 Italia secured over 100,000 clients for their new DVB-H service within just five weeks of its launch by making sure it was available across 2,000 Italian towns and cities. The uptake in Italy is even faster than that the S-DMB service in Korea, where over 600,000 receivers were sold in the first half of 2006.

Forecasts for growth in mobile TV vary considerably, though one analyst suggests that around 75 million TV-enabled handsets will be sold during 2009, up from around half a million units in 2005.

Improving Mobile Broadcast Standards
Mobile TV on the handset has progressed rapidly from the initial video delivery services that were envisaged on the first 3G networks. Low bitrate (less than 100KBps) and low resolution (QCIF - 176x144) standards have advanced to high quality 30fps QVGA services that are available today.

The rollout of mobile TV is being enabled by a number of technologies. Improved video encoding algorithms, such as H.264, allow video to be encoded at a bitrate that is 30-50 percent smaller than that which could be achieved with MPEG-4 Simple Profile. High quality 2-inch (and larger) QVGA screens with low-power LED backlighting are an important part of the system, as are broadcast technologies such as DVB-H and Mediaflow, which allow efficient use of spectrum to enable multicast TV to many devices. Finally, system on chip (SoC) technology that delivers the performance needed to receive and decode mobile TV is critical in providing the battery life and handset cost-profile that the consumer expects.

Content Delivery Models
As well as having the right technology, the success of mobile TV depends on having content and delivery mechanisms that meet the needs of the consumer. There are three primary delivery mechanisms.

Video streaming: Because this approach requires a dedicated link to each user, scaling the service to a large number of users rapidly consumes the operator’s valuable 3G and 3.5G bandwidth. This model is most appropriate for very high-value time-sensitive content, or for initial rollout of mobile TV services.

DMB (digital mobile broadcast): There are a number of standards for DMB including DVB-H and Mediaflow, but they all have in common that a number of channels are broadcast/multicast without requiring a dedicated data connection to each user (unlike streaming).

Video podcasting: With this model, the content is delivered via download over a PC or direct to the handset, which could be during off-peak times. This model is appropriate for content that is not time-sensitive and for consumers who may not have good DMB reception – for example commuters on underground trains.

Technology Challenges
Despite the progress made, there are still a number of technology challenges for mobile TV – especially DMB. Battery life is a major issue: the power-hungry components include the display, DMB receiver, and application/video decoder SoC. Higher levels of integration, combined with more efficient implementation, will help to reduce power consumption within the DMB receiver and applications processor. The issue of display power is being tackled with more power-efficient backlighting, as well as new display technologies such as OLED.

For DMB technologies, one of the major challenges will be to ensure that the user has a consistent experience as they change location and, when data is lost, that the degradation is minimized. Ensuring adequate reception and playback quality will require judicious DMB receiver and antenna design, combined with smart decode algorithms that can hide the loss of data for short periods of time.

Will Mobile TV Be Successful?
The underlying technology for enabling mobile TV in handsets is available today. Over the next two to three years, the technology will develop in such a way that the cost and power profile will be suitable for mass market handsets. This means that, by default, handsets will become capable of receiving mobile TV. The key challenge will be availability of ‘must see’ content, and a business model that is acceptable to the content provider, carrier, and consumer. Thanks to technologies such as DMB and video podcasting, the actual cost of delivering the content will be very low compared to using the 3G networks. For successful DMB deployment, both spectrum allocation and the initial infrastructure will be critical factors.

Delivery of content will be driven by the structure of national TV markets. For example, in the UK where broadcast rights are held by national broadcasters such as the BBC, it is very easy to deploy new content delivery models suitable for mobile TV, especially with the strong digital rights management schemes that will be supported on handsets. Other markets are much more fragmented. For example, in the USA the major broadcast rights are held by local ‘city’ broadcasters. This arrangement will make mobile TV deployment more difficult.

Research has shown that consumers are only willing to pay $10 to $20 per month for a mobile TV service. The content and service providers will have to base their business models on this level of revenue per user. This optimum price point must be combined with a good user experience, as demonstrated by the iPod/iTunes experience.

Given that the initial consumer reaction to mobile TV has been a positive one, and that there are signs of significant market potential, it is likely that the right technology, content and economics will be in place to ensure that mobile TV is a success.

Implementation Technology
The essence of an optimal mobile TV solution is the ability to deliver very good performance at the right price point, with an outstanding power profile. These are the fundamental qualities that have been at the heart of the ARM architecture and continually improved over successive generations.

Implementation technologies must be matched to specific requirements within the system. For example, a DMB front end performing OFDM demodulation, forward error correction and stream decoding/processing, matches the performance profile of embedded processors like the ARM968E-S™ and Cortex™-M3. The requirements of the application would better fit the capabilities of the ARM1176JZ(F)-S™, or Cortex™-A8 processor. These processors support ARM TrustZone® technology, which provides integrated software and hardware security – a vital component to enable digital rights management (DRM). Additionally, the Cortex-A8 processor also supports the ARM NEON™ architectural extension. This is a 64/128-bit hybrid SIMD architecture that accelerates the performance of multimedia and signal processing applications, including video encode/decode.

Besides processor cores, there are other ARM technologies that are highly relevant to mobile TV implementation. For example, OptimoDE™ – ARM configurable data engine, offers designers more flexibility to address specific processing challenges posed by high-performance mobile TV design.