Last Thursday I attended a workshop organized by European Commission at Brussels. This workshop is for PLUM consulting to present their initial findings on
First stakeholder workshop for the study “Challenges and opportunities of broadcast-broadband convergence and its impact on spectrum and network use”
Overall the workshop presented few findings that are significantly new. But there are some interesting facts about the audio-visual market that can be found in the presentation slides linked here. Other than that perhaps the most interesting discussion is how Plum categorize their investigation scenarios, about which I will talk a bit more later. And the biggest surprise to me is that, even though the workshop title is about the convergence of broadcast and broadband, there is so much emphasize on broadcast and so little on broadband. It seems that the mind set of the whole discussion is still very much linear broadcasting oriented, as if the audio-visual service in the converged platform would not be very much different from today! The flexibility of the converged all IP system just seems irrelevant throughout the discussion.
The number of the attendees to the workshop are well over a hundred. Most of them are senior people from regulators, broadcasters, DTT and mobile operators and TV manufactures. It is a diverse group, but people from the broadcaster sector, such as France Television, EBU and etc, are the most vocal ones and dominate the discussion. My own theory is that they are aware of their precarious position amid this fast changing audio-visual ecosystem, and thus they would rather grab every opportunity to emphasize how efficient DTT is in terms of special efficiency and cost, and how indispensable DTT is as a public service at time of emergencies. For instance: even though Mr. Webb from Plum has been very careful when presenting the LTE-option for converged platform, and clearly states that the cost saving of a converged platform is minimal at best. The simple mentioning of ‘cost saving’ has attracted a lot of counter-argument. More than a few broadcasters did not hesitate to throw out facts claiming how costly it would be to have a large number of mobile transmitter sites than a few high power high tower TV transmitters. In another instance, a senior people from EBU pointed to the spectral efficiency comparison between LTE and DVB-T2 in rural area, and argued that how unreasonable to choose LTE when it is only half as efficient as DVB-T2 (2bps/Hz vs 4bps/Hz).
On the face of it, all these arguments may seem valid. But I think it is a very biased interpretation of the numbers, because most of these people are so used to the concept of broadcast. All these comparison and conclusion make sense only if we assume the only difference between a cellular based system and DTT is the height and power of the transmitter, and assume all the LTE transmitters will be broadcasting all the TV channels all the time everywhere. However, we should not forget the biggest advantage of LTE over DVB-T2 is its unicast capability and the ability to dynamically switch between unicast and broadcast! So, unlike DTT, not all LTE transmitters need to be always broadcasting, and not all the TV channels are needed to be distributed to everyone all the time.
In my opinion, we will never reach a fair comparison between DTT and LTE if we stick to metrics like spectral efficiency that does not capture the difference between unicast and broadcast. Let’s face it, DVB-T2 is THE MOST spectral efficient transmission technology for distributing content of high demand in a large homogenous area. However, the reason we are having this workshop is because the demand nowadays has become more in-homogenous, both in spatial and temporal distribution and in the content popularity. Spectral efficiency alone won’t be able to show the fact that broadcast cannot adapt to these inhomogenity in demand very well. Perhaps a more relevant evaluation metrics should be ’spectral efficiency in satisfying demand’, defined by number of viewers served per Hertz per base station, instead of spectral efficiency in delivery bits, defined simply as bps per Hertz per base station. With this metric, it is now possible to directly show how inefficient DTT is when it is broadcasting a genre-specific content, e.g, a fishing channel, to a tiny group of viewer spread out in the country, while LTE system can unicast the content only to this group of viewers at a few base stations. Sadly, the presentation did not emphases on the flexibility advantage of LTE, and the main argument for LTE is its smaller foot print that minimize cross-border issue and facilitates national SFNs.
Now, back to the investigation scenarios. After presenting various trends and aspects in audio-visual content consumption, the proposed scenarios are categorized in a very broad way, which I think is very good approach to keep a coherent view on the overall picture and avoid the risk of dive into dozens of either uncorrected or overlapping small scenarios. The scenarios are defined by two dimensions: one the impact of OTT on audio-visual consumption, another is the viewing habit being either in home or outside the home. The impact of OTT is easier to understand, as it primarily reflect the on-dmand trends. While some people argue that a mobile/fixed dimension would be better than in or out of home. But I think mobile is a too general term, when in many case people could be using mobile at home but connected to local Wi-Fi. Therefore, in/out of home could better capture the different requirements for local or wide area connectivity. By combining these two dimensions, Plum defined three scenarios (one of the combination—low OTT impact and high out of home trend— is discarded as it does not seem likely to happen), Anywhere and now, Home Hybrid and Little change (see the attached figure).
These simple scenarios are very interesting definitions and approaches to the otherwise complex problem with so many different aspects to consider. Although not much findings were discussed in this first workshop as the study has just started, it would be very interesting to follow its progress and see more results in the future.
The next stakeholder meeting will be in this July and the final presentation would be by the end of this year. Some of the related studies are this subject could be found here.