Comment The Nokia Networking LTE Broadcast trial announced in Germany this week is all about building bridges between broadcasters and mobile operators, because it is about sending TV for televisions through an LTE Network, rather than TV for phones.
And for a change the trial was explained clearly in its release, which meant quite a few media outlets actually understood the distinction and did not confuse it with TV direct to portable devices.
We were keen to hear more and managed to get a few words with Arne Schaelicke, head of marketing for LTE and LTE Advanced for Nokia Networks.
Schaelicke's first point was about synchronisation: “Most trials of LTE Broadcast have involved a single base station, for instance in an arena. But synchronising multiple signals from more than one base station is not simple and other systems (in our view) do not do this well. So we wanted to prove that Nokia Networks could do this.”
In order to create a single-frequency network (SFN), signals from two base stations must both arrive within the guard interval of an Orthogonal Frequency Division Multiplexing (ODFM) signal, which is tiny, so synchronisation has to be precise. A key element of OFDM is a Guard Interval to allow the receiver to cope with strong multipath situations and Doppler effects. If data arrives a bit late, but before the guard interval has passed, an OFDM system gathers it all up into a single signal.
The trial is going to go on for some time, at least a year, possibly two, and permission to use “unsold” 700 Mhz spectrum was obtained using an experimental licence in Germany.
Schaelicke wanted to make it clear that the choice of the spectrum was not really important, he could have chosen 450 MHz or something else, but it gives us an idea of the scale – LTE versions using 800MHz are usually planned with a 10km cell radius. 700MHz propagates even further than 800MHz, but broadcasting adds some complexity, so by using four conventional Nokia LTE Flexi Multiradio 10 BaseStations they could be set roughly in 7km cells, and that more or less brings you the 200km2 that the trial reaches.
The politics of LTE
But the big thing here is to understand the politics. Nokia Networks is really doing this to bring broadcasters closer to cellular operators and to prove that LTE is spectrally efficient and it can be effective at sending SD and full HD TV signals to set-tops, whether that is direct through a wall or through an external antenna to the roof, and then down a cable, as well as direct to a smartphone or connected tablet.
Nokia Networks clearly does not want to burn all of its bridges here. It sees this trial very much a departure from where all the other LTE Broadcast trials are going, “Yes we may try to connect straight to a smartphone on this trial and we have done other lab demonstrations that are aimed directly at connecting to smartphones with LTE Broadcast, but that’s not what we are talking about here,” reminds Schaelicke.
So how does he see this panning out with broadcasters in Europe – do they eventually abandon their DVB-T networks and rent space on an LTE network? Schaelicke is very clear on this subject, it’s not his place to speculate, although he welcomes us to. “We need some clarity about the technological aspects before broadcasters can begin to talk to MNOs and resolve how they might work together,” he adds.
But the finances are clear. MNOs, and no one else, can afford 700 MHz and similar pieces of UHF spectrum, and broadcasters are already deserting them. We have seen broadcasters in Germany say they would leave the broadcast networks in favor of sending content OTT. That makes sense for the broadcasters, because price for delivery over the internet is cheaper than over broadcast multiplex. But does it make sense for a cellco?
Cosying up to the broadcasters
If the content arrives OTT and goes to portable devices, it will mostly be carried by Wi-Fi, but where it isn’t it could harm the performance of a cellular network. The cellco is taking no revenue out of OTT video delivery, and it also arrives unicast and in adaptive bit rate over its network, which it can have no control over, and in order to cope with it, the cellco has to buy expensive video optimisation systems to contain it.
It would be far better to court the broadcaster and deliver the broadcast content on its behalf, over the existing cellular broadcast network, and on the one hand get a delivery charge for those delivered to TVs, and on the other get a chance to sell a mobile version of the content direct to a phone.
Now to have a phone version and a set-top version on different codecs and resolutions, you would have to have two deliveries in parallel, but Schaelicke agrees that with a code that is split into two formats, one could be highly reliable with a strong FEC, for smaller in-home devices, and the other codec could fill out the content for larger screens, so that just a single split version could be sent.
This was part of the spec of DVB-H, but it was never implemented as far as we know, and it was also used in MediaFlo.
Schaelicke will admit that this is possible, but then brings us back to the longevity of this trial. It will be in place as these types of compromises are worked out.
We raised the idea that this might just be a special case for Germany, which barely uses DVB-T for terrestrial TV delivery, it’s much more a fan of satellite. Schaelicke says in other markets you would change some of the parameters, such as the spectrum choice, but essentially this is a trial whose outcomes could be used anywhere. The idea is to throw off lots of survey data, as each thing is tried on this testbed.
But there are other candidates for this in Germany. There have been trials of DVB-T2 Lite, a broadcast method for mobile TV delivery to portable and mobile devices. But that requires that the masts for DVB-T are kept and used. Another rival is a new technology called Tower Overlay, which operates on top of LTE, in effect, a way of recreating digital terrestrial transmission, but integrated with unicast cellular delivery.
The story of Tower Overlay is interesting in its own right, having been developed by Ulrich Reimers, and uses the carrier aggregation capabilities of LTE and LTE Advanced. Reimers is a well-known name in Digital TV and was one of the scientists responsible for foisting DVB-H on the unsuspecting world. He works at the Technical University in Braunschweig’s Institute for Communications Technology.
Reimers was one of the people that convinced old Nokia at the time to throw its weight behind DVB-H and we all know how that ended, and Nokia today is a very different fish from the one that dominated mobile devices. So perhaps Nokia is keen to push its eMBMS version first, and get in before Tower Overlay builds momentum. It would certainly also lead to Nokia selling a lot of its Flexi Multiradio 10 BaseStations in any country that adopted it.
The system does not lend any hope to UHD streams being accommodated on LTE, and Schaelicke talked about support at the top end of its range for 8Mbps HD, and support for 2Mbps SD at the lower end.
For our money this is a natural progression, and over perhaps a seven- to nine-year time frame, TV broadcasters may well drift into using LTE for delivery if the pricing is right and become customers of the cellco.
The rest of their delivery would be over fixed line internet, but this way they would not be excluded from reaching portable devices. Content would be a key differentiator – what if you could get ProSiebensat only on T-Mobile in Germany over LTE? Or that it came with guaranteed quality on another network?
That would lead inevitably to partnering with original content, or perhaps cellcos snapping up TV broadcasters and offering their service exclusively to drive sales. It could also mean an LTE router or set-top in a good many homes.
The Nokia Network system will create a Single Frequency Network (SFN) initially in four base stations, connected directly via fibre, to aid with that synchronisation issue.
TV delivery would not eat into the mobile data plan and broadcasters could drive advertising initially to the same TV sets, but later to reach mobile users as well and that might lead to interactive services.
The trial, which began transmissions in early July in Munich, Germany, uses evolved Multimedia Broadcast/Multicast Service (eMBMS) software at four sites of the Bavarian broadcast company, Bayerischer Rundfunk, in Northern Munich.
The sites are connected by a high performance optical transport network and the other partner is the Institut für Rundfunktechnik, the research institute of broadcasting companies in Austria, Germany and Switzerland – so we must not think of this purely as a Bavarian solution.
Copyright © 2014, Faultline
Faultline is published by Rethink Research, a London-based publishing and consulting firm. This weekly newsletter is an assessment of the impact of the week's events in the world of digital media. Faultline is where media meets technology. Subscription details here.