Enjoy the 5G Vista: Inside the successes and challenges of the 5G Broadcast in-stadium project led by the DTG

The DTG has been the leader of a consortium named 5G Video In Stadia Technical Architecture, or 5G Vista for short, which aimed to develop a 5G-based platform to broadcast live sports events and enhanced content in stadiums and similar locations.

The project, which was focused on using further evolved multimedia broadcast multicast service (FeMBMS) – also known as 5G Broadcast – is one project incubator within 5G Create, run by the UK government’s Department for Digital, Culture, Media & Sport (DCMS). It won £1.3 million funding just over one year ago towards the project’s £2.3 million total cost to find out about how to create and enable the 5G backhaul that will be integrated into today’s broadcast systems – from glass to glass – to deliver exciting content for in-stadium viewers of live sporting events.

Recently in February 2022 5G Vista completed a successful trial at a live football match at the 30,400-seat MK Dons football stadium in Milton Keynes, UK. A group of fans at the match between MK Dons and Ipswich Town on 12 February were able to try out a prototype of the 5G Vista app, which provided content for them over a 5G Broadcast network integrated into the TV broadcast network.

Here, we speak to Alex Buchan, strategic technologist from the DTG and project lead, about the successes and challenges for Vista, and what comes next.

Your original objective was in two parts: to test and demonstrate the technical feasibility of using 5G Broadcast technology, and to also work out the consumer benefits. Talking about the technical side firstly, did you achieve that?

Yes. I’ve been involved in various working groups and meetings about this technology for a long time, and it was always focused on how public service broadcasters like the BBC could use this to deliver BBC1 to TV’s in people’s houses. But there were never any mobile network operators involved, and yet there were obviously a lot of benefits for using it in a mobile network, for example because [network operators] could really save on capacity and costs for serving places like stadiums.

So this whole project was about showing the commercial use cases for that for a mobile network operator, and then also driving mobile phone uptake [enabled with FeMBMS], because there’s no mobile phones that support it; or there wasn’t before the project started.

Further evolved multimedia broadcast multicast service (FeMBMS) is a further development of the LTE broadcast mode eMBMS in 3GPP Release 14. It enables 100% of the transmission capacity to be used for broadcasting services

From a technical side what we were trying to achieve was to work out how we integrate camera solutions into the 5G Broadcast network, and then we needed to understand how we would install all that at the stadium.

So [the question was] how would we actually broadcast the content out to the fans in the stadium? [To find that out] we needed to create a whole end to end network from [broadcast] cameras through to the [end users’] phones, glass to glass.

So we did that. We brought in [additional] partners [to the consortium] from TV production and from camera coding, which were Imaginary Pictures and Ateme; we added them into the project and then we integrated their side of it into the Rohde & Schwarz broadcast side of it, and then we used O2’s spectrum to send it out and we installed it all in the stadium.

So technically speaking, we created the end to end solution, which was what we wanted to do.

How about the consumer-facing side of the project? How has that gone?

The final part was the commercial side of it, so what do fans think of it? What would they pay for it? How would they pay for it? Do they see themselves using it? What else would they use it for? If they weren’t in the stadium, would they use it outside the stadium? And what scenarios could they see themselves using it in?

At the start of the project, we convened a sports fans focus group. We asked them what they thought of this as a concept. We didn’t have any app we could show them – it was right at the start of the project – and they were a bit like, “No, I don’t want to watch something else while I’m trying to watch the game, it would distract from what I’m there to see, and I’ve paid the money for”. But by the end of the project, once we had them in the stadium and we showed them the app and they were watching it there, they really got it and they understood what this adds to the game.

When things like a shot [at the goal] happened, they could look at it from a different [camera] view. Or when someone got tackled off the ball or something like that, they could see it and they could find out what happened. They all really understood that when you’re at a match, there’s things that happen and you miss them, and then you’ve missed them and you can’t see them again.

This is about combining that sofa experience with the live experience as well, so they really got it. It was a real achievement for the project, I think, that they really understood it.

So all round, really good successes for the project.

We’ve written all our findings into a report that’s going to be ready by the end of March.

Your original idea was to test the technology, work out how to create the glass to glass network, and then to also demonstrate the consumer benefits of using 5G Broadcast, but you’ve actually also developed – as a by-product of this research – a content-based app for consumers?

Yes. We developed an app, and then the technology [for the backhaul] provides a means to do that without having to spend millions of pounds on [increasing] stadium capacity, which you would do normally with a mobile network solution because you’d have to put in a lot of base stations, but if it’s 5G Broadcast you don’t need all that infrastructure. You just have a couple of base stations and everyone can get it.

You quickly run out of capacity, especially if you’re watching video services, in a unicast world; you have to put in a lot of infrastructure to support video content. Whereas with 5G Broadcast [multicast] you send it once and then you get it, the person next to you gets it, the person next to them would get it. It’s not adding on any extra capacity. You send it out once and then everyone gets it. So you only need to have a small number of base stations and then everyone’s covered, and that’s what makes it commercially viable.

The concept [for our app’s content is it] would be a curated set of services that would go to everyone [in the stadium] so everyone would see the same thing. It gets pushed to them; there’s no uplink [on the 5G Broadcast network].

There could be all sorts of a different personalisation there [for viewers suing the app]; because you’ve got quite a lot of capacity to play with [using this network] you can send them multiple audio services, so there’s accessibility type solutions there, and the user can also select what they want to see.

What is the actual capacity on this network?

We had about 25MBps channel capacity to play with [during the MK Dons trial] to send out the video services. That allowed us to send out six HD video services to the customer trial list, plus some audio; the audio was minimal data rates so you could add loads of audio on top of that.

To put it into context [the capacity at the trial was equivalent to] an HD TV multiplex for sending out digital terrestrial TV, which would be around 40MBps for maybe 20 or 30 channels of HD content. Playing around with the modulation and the frame rates etc, you can potentially send out loads and loads and loads of TV services, like a traditional TV service basically.

It gets around that issue where you are in a stadium and you can’t access the network because everyone’s trying to access it at the same time. That’s an issue because you could solve that by just putting more capacity, throwing money at it, but for a stadium that’s only used once every other week, the finances don’t stack up for that; that’s the problem. This [solution with 5G Broadcast] gets around that, because it’s very cheap to install and even for a stadium that’s not used that often, you can get the money back on it quite quickly.

How has the project opened up doors for adding FeMBMS into smartphone chipsets with Qualcomm, because they weren’t part of the project originally. How did they come into it?

Now we’ve got to a point where Qualcomm has agreed to put [FeMBMS] into a chipset, and they’ve produced a first prototype phone as well, so we’ve got to that stage as well.

There’s always been this chicken and egg problem, where it’s like, “well, we’re not going to put in the handsets, because there’s no demand for it”, then “there’s no demand for it, because there’s no handsets”. You just go around and around in circles like that. This whole project was designed to unlock that hiatus.

Throughout this project we’ve been able to have parallel discussions with Qualcomm, which we started right at the beginning. At the start they were saying, “look, well, we’re not really sure why we would put this in our handsets”, and now fast forward 18 months, they’ve released a test handset with their Snapdragon [chipset built into it]. It supports the direct [FeMBMS] signal being sent to it for a broadcast.

At Milton Keynes we had to use a specialist receiver to be able to receive [the signal as we didn’t have phones with embedded FeMBMS], and then that went by a cable to all of the phones. [Now Qualcomm have built a prototype FeMBMS smartphone], for further trials, we could just send it straight to the phones. So that’s a massive achievement.

We’ve been working on this [at the DTG] for probably seven years in various working groups throughout Europe, and there’s never been a handset that supports [FeMBMS] directly. So it’s real achievement.

Originally when we spoke early on in the project you said about challenges with cameras, edge computing, broadcast technologies on site that needed to be integrated into the mobile network, and the issue of latency. What are the challenges that you faced and also maybe discovered during the project, that need to be solved before you can move on?

The cameras were positioned around the pitch by Imaginary Pictures, then what you need is some kind of encoding equipment, which you need specialist hardware for, [but that then] increases the difficulty for a stadium to install [5G Broadcast] because they need some specialist encoding equipment on site, which they might not use all the time.

So we worked with Ori Industries, which provides edge computing stuff, and it created a kind of containerised encoder, like an app. The app is a virtual machine running on a server, like a service which runs the attending coding. So instead of need specialist hardware, you use this generic hardware with a virtualised service on it.

Initially we had a bit of trouble integrating all of that, but through the process of R&D [which is the point of this project] we’ve got that all working really slick, so now a football stadium, for example, could have a generic bit of equipment – a server or whatever – which they could use for other stuff like analytics or whatever they wanted, and then they’d have an app on there. When they were doing a game and they needed some content encoding, they could just spin up the app and then they could use it for that two hours and then switch it off again.

After that it was just looking at the end to end piece; the latency. Initially when we started, it was around 30 seconds. We couldn’t really have that because it was just way too long, so we reduced it down to two to three seconds, and it could be reduced down further with more specialist installations.

But in fact, when we did the customer focus group [at MK Dons], they said they wanted more latency in the system because they wanted more time to watch the live game and then go back and watch replays. [They asked if latency could be a] configurable thing, so in football you’d have it longer than maybe [you’d want it] in golf. That was interesting feedback.

Replays were the other thing the fans just said, “Yeah, we want replays. That’s what we want from this”. So what you’d have to have as a service, is you’d have to have some kind of curated production system where they’d have the live stuff which they’d push out, and then they’d have a replay bit which they could prepare and then could push that out as well on one of the channels.

So this would this work almost like a broadcast adjacent to the main broadcast from whichever rights holder is filming that match?

Yes, so at Milton Keynes [football stadium] we did that. We had our own production team there because Sky Sports and BBC Sport weren’t there, so we had to film it ourselves. But equally if [a rights holding broadcaster] was there, we could have just used their broadcast feeds and sent them out to everyone in the stadium, who could watch it, then you could add in your own broadcast feeds [over an app] to look at different angles, replays and stuff like that. For example, at Milton Keynes we had an interview with the manager which we sent out, and we had the replay channel, but you could have stats and data as well. Lots of options.

Lastly on next steps, you need to bring that that money in the chain – the coordination of team rights holders, sports organisations, venue owners, broadcasters and mobile network operators, and mobile handset manufacturers – together. How do you intend to move this project forward?

We’re in discussions about doing a trial with British Touring Cars in April. That trial would be the first time we would be able to use these new [Qualcomm] handsets that will be able to receive the [5G Broadcast] signal directly. So that’s one part of it.

The second part of the trial is part of the project was also looking at the security of the content, which is important for content owners. The University of Surrey has been doing a standalone piece in its labs about how you protect content in a mobile broadcast scenario. We would integrate that into the live solution as part of the trial [at British Touring Cars].

The third part of that trial would be using a slightly different sporting scenario. So football is one scenario where it’s fast paced, you can see everything that’s going on, you’re almost on top of the action. Motorsport is a lot different, where people can’t see what’s going on most of the time. They don’t know what’s happening. They don’t know who’s winning. And it’s very hard for the venue operators to keep them across the action. So this is a kind of a different use case which would be a good example, not just for motorsport, but for all sorts of distributed events like golf, athletics, and cycling, where you can’t see everything that’s happening, and when you do see them it’s for a very short period of time and you don’t really know what’s happening.

[The fourth aspect of this trial would be] showcasing it being integrated with the TV networks, because they would be there filming it live. We would say, “ok, we could use your content and push that out to the fans” and show them how that would work.

We’re aiming to do that in April at Donington park in Derbyshire, which is the opening race of the British Touring season. We’re in discussions with them at the minute. We’ve done kind of previous work with them in September last year where we used their recorded content but this would be live race.

Finally, why is FeMBMS – or 5G Broadcast – so important to the DTG and others?

Well, for DTG, it’s about being able to reach new audiences. Our members are broadcasters, and obviously they want to reach as many people as they can. This [network technology] opens up [the ability to] reach people on the go a lot easier, on trains and cars, in difficult to reach kind of scenarios, where previously it’s not been possible. [Using 5G Broadcast] potentially you could reach them with live linear content; you could send it straight out to people as they’re on the move to their phones.

So for DTG [this project] was just about exploring where the future direction of TV is, and 5G [Broadcast] is obviously an important part of that. It was just about understanding what role that will play in TV delivery of the future.

Mobile broadcast has been around for a long time; it’s not a new concept. MBMS has been around as a mobile standard for nearly 20 years, and it’s gone through various iterations, which are now at FeMBMS. [Three things have] driven it forward.

One is it’s easier to integrate into mobile networks now; previously it was quite a clunky solution, which was an add-on to the network and it meant sacrificing some other elements of the network and spectrum, which the mobile network operators didn’t want to do.

[The second thing is] now the TV spectrum’s have been sold off there’s a broadcast spectrum available to mobile network operators.

[The third thing is] people are used to watching stuff on their phones a lot more now than in 2006 or whenever [MBMS] first came out.

I think they’re the three things that have driven it to a point now where [5G Broadcast] feasible.

Alex Buchan took part in SVG Europe’s Football Summit 2022 on 10 March at the Tottenham Hotspur Stadium. A video recording of his seminar session will be made available to view next week.