Mobile broadcast

The graveyard of mobile communication history is littered with failed broadcast standards, even predating cellular telephony itself, culminating with Qualcomm's abandonment of MediaFLO technology in the United States, along with the failure of DVB-H services elsewhere in the world. Against this background, it is natural to question why any other technology for mobile TV on the table now should be any more successful.

Yet there are growing grounds for hope that the Integrated Mobile Broadcast (IMB) standard, defined as part of release 8 of the global 3GPP initiative for mobile communications evolving from GSM, will succeed. This is partly a matter of timing in that IMB is coming along when at last, with tablets and especially the iPad, there are portable wireless devices on which people actually want to watch longer form content and linear programming. But IMB also reflects lessons learned from the failure of earlier standards and addresses some of the problems that have not been tackled fully before, notably scalability, spectral efficiency and above all the ability to integrate broadcast and unicast services seamlessly. The latter is essential for two reasons.

IMB integrates unicast and broadcast

First, IMB enables subscribers to switch between unicast and broadcast on the fly to optimize use of the spectrum. This is desirable when an increasing number of people start watching some live content unicast at the same time, in which case it would be switched to broadcast so that it is only transmitted once rather than many times. Second, such integration would facilitate roaming, which itself can take two forms. First, an operator may not provide complete coverage across the service area of its IMB technology, in which case it must be possible to provide the same set of IMB-based services via unicast. Second, via a similar mechanism, if a customer roams to another operator's network, whether in the same country or internationally, IMB services that the user subscribes to should be available via unicast in those territories. Of course, this raises the thorny issue of content rights, but that is another matter beyond the scope of this article.

IMB requires device cooperation

Operators should be aware that while IMB ticks the right boxes, it is not a total panacea and cannot deliver a successful mobile TV service on its own. The cooperation of device manufacturers will be required to ensure that there is adequate onboard processing and memory in place to support the content caching that will be required to make mobile broadcast really work and to ensure that unicast/multicast handover is sufficiently fast. The GSM Association in its IMB white paper “Integrated Mobile Broadcast (IMB) Service Scenarios and System Requirements” has called for handover from broadcast to unicast to be performed with minimal loss of frames and for the respective streams to be synchronized to within a second to avoid degradation of user experience during the process. But this is not within the gift of the standard itself and requires cooperation between handset and the service, since the former has to recognize that handover is taking place through appropriate notification from the network. Billing has to be properly integrated with the two modes as well, so operators will still need to make significant up-front investment to make IMB work.

How IMB saves spectrum and allows incremental deployment

Nonetheless, IMB does promise to make large mobile TV deployment affordable, first by broadcasting over time-division duplexing (TDD) channels, which in some countries at least includes spectrum currently unused. Second, by virtue of being a GSM-based Universal Mobile Telecommunications System (UMTS), IMB can be deployed over existing cellular sites without requiring the up-front infrastructure investment of DVB-H. This is a huge benefit, especially as some operators made that DVB-H investment and have had their fingers burned. With IMB, operators can introduce mobile broadcast gradually and ramp it up as demand increases or suitable content becomes available.

These factors, combined with the promise of being able to ease the pressure on congested radio spectrum by offloading data onto the TDD circuits, have encouraged some operators to conduct trials. In the UK for example, two major operators, Orange and Vodafone, have been trialing IMB.

Operators should support FDD/TDD integration

Most mobile unicast services use frequency-division duplexing (FDD), in which two-way duplex communication over a link that only allows traffic to go one way at a time (i.e. half duplex) is enabled by sending upstream traffic and downstream traffic over different frequencies. TDD achieves the same effect at a single frequency by allocating the slot to upstream and downstream traffic alternately, and has the advantage of being able to do this on demand to suit varying traffic profiles. In the case of IMB, it can be configured to deliver downstream broadcast only.

Given the potential of TDD channels, it is important that operators that have favored the FDD variant so far should ensure their networks can accommodate FDD/TDD integration, irrespective of whether they plan to deploy IMB and mobile TV right away. This is simply because, given the rapid rise of data and video traffic, all spectrum is precious and valuable regardless of what specific services are deployed. Then if, as seems quite likely, TDD becomes widely adopted as part of the 3GPP Long Term Evolution (LTE) project by say 2014, operators will be well-placed to leverage LTE TDD cost benefits to add extra capacity to their networks.

IMB will drive up bandwidth consumption

Another point about IMB, which is quite the opposite of earlier mobile broadcast standards, is that it could end up being a victim of its success. It is worth remembering that a major driver for mobile TV will come from the increasing population of tablet devices, which have much larger screens than smart phones — 9.7in in the case of the iPad. This means that mobile TV services capable of producing good SD or even HD quality pictures on such devices will consume more bandwidth. Recent mobile broadcast standards such as DVB-H evolved from a DTT background but recognizing that bandwidth could be saved by allowing for the limited display capacity of mobile devices.

Typically, devices supporting DVB-H displayed pictures at QVGA resolution, which is 320 × 240 pixels at 15fps, consuming around 300kb/s. Some handsets only display at QCIF, which is 176 × 144 pixels also at 15fps, requiring only 100kb/s to 150kb/s. But in an IMB world, a range of resolutions would have to be supported, extending up to SD, at 720 × 576 pixels, perhaps at 30fps, consuming up to 2Mb/s and perhaps slightly more. In the future, HD would come into the equation as well. It can be seen then that while IMB may save bandwidth now, it may in the future stimulate significant further growth in traffic, so operators will have to continue investing in measures to offload bandwidth. It's possible that Wi-Fi and femtocells will come to the rescue in the end, exploiting the fact that most mobile TV is likely to be consumed within and around the home, often in effect as secondary TV sets. Wi-Fi could be used to deliver mobile broadcast services to devices when they are within range, while femtocells may also be deployed as home cells, partly to ensure universal cellular coverage within the home but also to offload unicast traffic from the broader external macro cellular network.

The business case for IMB

The business case for IMB is stacking up, since it promises to derive incremental revenue from new services, improve customer experience because existing unicast services will work better when broadcast, and reduce data transport costs by offloading unicast traffic. It will also support new content distribution models, including predownloading near live content such as news or sports before users wish to see it. It could also download content that is less time-sensitive, such as music, games or electronic books. But this again assumes that the device has the required capacity, which will be true for tablets but probably not for smaller smart phones. All in all, IMB brings plenty of food for thought.

Philip Hunter writes the Beyond the Headlines Europe e-newsletter.