Guy Daniels, TelecomTV (00:00:10):
Right. Hello everyone. Welcome back to Debate five, the final debate of today, the final debate of this year. So please come and take a seat, bring your teas and coffees and cakes with you, sit yourself down and we'll start this final debate and welcome back to our online viewers as well, the business case for 5G Standalone. This is our last session. We've left, we've kept the best to last. Obviously I'm pausing slightly because we have a rambunctious, is that a word? They're rowdy. A rowdy. We have a rowdy audience, which is great. I'd rather have a rowdy audience than a silent audience. So thanks very much for staying with us. This final debate looks at the business case for 5G standalone. We know that the narrative for of the telecom ecosystem has for a long time being that 5G standalone is a necessary bridge to what comes next, the 6G, yet others claim it's too late, too complex, too expensive, there's lots of arguments against.

(00:01:19):
So should operators continue to invest in 5G standalone or even start to invest in 5G standalone to extend connectivity and support more services, thereby leaving 6G to be more focused perhaps on efficiency gains Or is it time to change the narrative and place big bets on a clean slate approach to 6G? So there's a lot to talk about. And to help us decide, we are joined again by a panel of experts who are going to be delivering their expert witness statements. Then we'll have a motion, then we'll have a debate. So if I could first of all ask our first guest, Itsuma Tanaka, who is president and CEO of Docomo Communications Laboratories Europe, to deliver our opening expert witness statement.

Itsuma Tanaka, DOCOMO (00:02:14):
Good afternoon, London. And please take a good look at me because after this, vendors, operators might consider inviting me here again. Okay,

(00:02:29):
So let's start with the fact I believe the business case for 5G SA exists already. We have connected cars, government funded projects to build resilient, more connected societies and they pay for efficient, reliable, and predictable network already. And according to three GP specifications, if you want to use AI fully in your network, you need 5G SA. But the real problem, let me talk about the real problem from here. The real problem about this ROI question is about scale. Look around the world again, let's say nationwide deployment, it's still rare and not because SA doesn't have a business case, but because too many operators treat cloud native, just let's a shopping list, give me cloud native 5G check. And they learn afterwards. It demands entirely new lifecycle and defined skillsets. And I have witnessed operated spent millions hand over everything to the vendors and end up living with really complex, really costly system, which they cannot even evolve without writing another check.

(00:04:08):
So they're locked in from maintenance, from support and dependency. So let me provoke here. When you hear 5G SA prerequisite, it's 6G, listen, whose voice that is, it's usually those who already built sa, those who aim to draw operators into this look in business model because nobody wants to admits some cost, but capability for the future is really clear. The recent geopolitical developments, the industry demands clearly require resiliency, ai, energy efficiency and SA is the minimum ticket for this. But 6G will elevate, should raise the bar for this. And those who already done sa, listen to me, you have done great work already. All the lessons, the painful lessons around the automation of observability AI ops that already gives you competitive edge. But for the future, because many of them are still on 4G NSA, let's try to find the new G generational pressure point where cloud native economics and operator capability finally click. Okay, so thank you very much. Thank you. Thank you.

Guy Daniels, TelecomTV (00:05:40):
Lovely. Great. And can we have our next expert witness please? Anita Döhler Anita is of course CEO of the NGMN alliance,

Anita Döhler, NGMN Alliance (00:05:49):
So representing NGMN here, operator driven alliance. Of course I will not speak about specific business cases and I'm also not formally allowed to disclose the motion yet, but I can tell you that my case is supporting the no and I also believe also believe that actually this motion can be answered with a yes and with no with almost the same arguments because I think what we need to have in mind is that everything we do in our industry intends to deliver value to our customers. So consumers and businesses and value in the future was always difficult to predict. So what does value mean to our customers and we believe that. So in NGMN, we believe that it's probably even more difficult to predict today because of the dynamics we have, we have software driven networks, we have all different kind of new use cases may be appearing, which maybe can be delivered with 5G standalone but we don't know yet.

(00:06:59):
So there will be probably additional requirements for the uplink. We see a change in the traffic patterns coming from agen AI conversations. So therefore the main sentiment is that the networks need to be open. Modular network operations need to be simplified clearly and therefore of course 5G standalone is paving the way for the future and it paves the way through setting the ground for cloud native networks. So fully interoperable cloud native networks, which again are important to be able to fully automate networks and also for embracing native AI networks and some of those features 5G standalone most likely will not be able to deliver. So we will need also to continue of course after 5G is a what we do already of course in the industry and with the first studies on 60 items in 3D ppp therefore, so we need to move forward and we also need to learn.

(00:08:20):
So it soon already mentioned also that it demands the change of processes at network operators demands a complete new culture to operate networks and we need to move forward because we need to learn, we need to experiment much more. So the culture of being quite a conservative, which is understandable in our industry because we have a high responsibility for resilient networks, needs to be also changed. And for this we need to learn. So therefore my sentiment for the motion is the no and I'm sure that we will get explained the motion later in more detail.

Guy Daniels, TelecomTV (00:09:06):
Anita, thank you. Thank you. I can't wait to find out what the motion is. I'm looking forward Anita, I can't believe you've already given away. Oh my god. Can we get a drink now? Thanks Anita. Can we get our next expert witness please? Thierry van de Velde who is global core networks, pre-sales leader cloud and network services at Nokia.

Thierry van de Velde, Nokia (00:09:30):
Thanks. So 5G is an enormous success. In fact during my speech 3,500 5G subscribers will be added. We are adding this year 600 million subscribers. We are going to reach 2.9 billion subscribers on 5G 8 billion by 29. It's an amazing success. It's probably the most successful technology, although I will sound like Trump now. But who of you?

(00:10:00):
So who of you remembers the triangle of 5G standalone when we started 10 years ago? Evolve mobile broadband, massive machine type communications and ultra reliable low latency communications. Exactly. We have delivered one third my friends only the enhanced mobile broadband. Yes, 5G standalone is slightly better in the uplink than 5G non-ST standalone. So we have so much work to do, not just in 5G but also in 6G. Thanks to your great ecosystem of partners, we have been able to learn so much while deploying these 5G core networks. We have learned about containers, about CICD, we still have problems to fix. We have learned what a chip plate is. Okay, at Nokia we have learned to discern I would say the better Kubernetes distributions from the less evident ones. We have learned to work in full vertical stack for tier two, tier three operators and then in the horizontal mode where multiple CNFs of multiple vendors can coexist in the same network.

(00:11:08):
So we have learned so much. 5G SA has tremendous difficulties today due to the standards, due to the way the standards were designed. For example, the release 16, release 17 value added services of 5G SA stop at the edge of the 5G coverage. There is no interworking specified to 4G in this room. We will probably see less than five 5G SA users. Although all your telephones are capable of 5G sa I'm sure because international roaming is a headache and is an absolute pain to set up with 5G standalone and it works like a charm in NSA mode in non-ST standalone mode. So how can we promise these consistent services in that triangle if we're not able to realize a experience that will cross the borders? We have a lot to fix in about 6G. There is tremendous, tremendous potential in 6G. We will have an AI driven physical layer that will allow us to extend the coverage to provide tremendous throughput increases in 6G and in 5G advance also by using AI to interpret the I diagram that we all remember from our time at uni. So let's march jointly towards a energy containerized network that will satisfy the needs of 5G S-A-N-S-A and 6G users. Terry, thank you very much indeed. Wonder applause with Terry?

Guy Daniels, TelecomTV (00:12:49):
Sure. We'll pick up on some of those comments during our debate and please can we get our final expert witness of the year, Beth Cohen and Beth is returning to the stage telco industry analyst at Luth Computer, Beth?

Beth Cohen, Luth Computer (00:13:00):
Yes. So 5G SA is critical to the telecom industry. So the telecom industry has benefited greatly. What's ironic about 5G in general is that it's been a tremendous success in terms of solving the last mile problem cutting costs, delivering more efficient networks, higher bandwidth. But ironically the average user look at their cell phone, they have no idea, they don't care that it's running on 5G, 4G 6G, they just want to make sure that their Korean soap operas are running so they can see 'em on the subway on the way into work. However, 6G is still kind of a pipe dream, let's be honest here. So I think that 5G needs to really become more robust and add network slicing, which we've been working on for I don't know how many years and we still don't have a standard

(00:14:10):
And so we need to be more robust about 5G infrastructure that the telecoms have invested literally billions of dollars into. I should point out the telecoms tend to be kind of conservative about their investments. They want to get maximum value out of their investments and so those billions of dollars have not yet paid off let's say. And so I think it makes sense for the tech refresh cycle to get more out of the 5G and then wait for the 6G pie in the sky when it comes if we will to do that next check refresh cycle that we need. And I will catch up because everybody else went over and I didn't thank you very

Guy Daniels, TelecomTV (00:15:05):
Much. Oh yes, it was noted. What a great collection of expert witness statements there. Fantastic. That's good even there's a lot of food for thought, but before we start discussing what you've had to say, I think it's time for motion.

Graham Wilde, Three Group Solutions (00:15:23):
Guy, you're absolutely right. So it is and the motion is the success of 5G standalone will postpone the need for 6G and the toin costs, the coin toss. I can't say that can you? I can't say it. No. Try again. The coin toss said I am in favor of this and Chris is again, so let me stand up and tell you

(00:15:45):
About why you should be voting for this, right? So where do I start? Okay, let's start with my favorite subject, which is myself. So my job title ladies and gentlemen, inside Hutchison is the head of 5G business development and the more observer amongst you may notice that there is a slight problem with that title. So I don't want 6G, I don't want 6G for the next 20 years. I need to keep going in my job, but I'm reminded of a time, well I used to live in Hong Kong and my wife and I moved back to the UK in 1999 and we moved to this small town in Herefordshire. We had a house and we needed to furnish the house and we went to an electrical retailer in Herefordshire and we bought a washing machine and it wasn't an expensive one, it was like a candy or something like that. We got this washing machine home and then about three weeks later it went wrong. So we went back to the retailer and said, this washing machine's gone wrong. And he said, yeah, they're shit, aren't they those?

(00:16:56):
And I'm like, when I was listening to Thierry just there, I was like, actually do you know what you Nokia kind of virtually invented 5G SA, right? And now you're kind of going, well there's a lot of problems with it. We need to go, oh, forget that we need to go onto to 6G. It reminded me of this bloke is stopped with the washing machine. I was like, oh my god, look, we are only 9% of all mobile customers in the world are connected to a five GSA network at the moment. Okay, 9%, 91% of people do not have five Gs a and we're already talking about throwing it away and going on to 6G. Come on, we've got to recover the return on investment on 5G, otherwise this is just ridiculous, right? We cannot progress as an industry unless we get the money back for the investments that we made.

(00:17:51):
And I'm reminded of when Three was making the case to merge with Vodafone, Three CEO Robert Finnigan said to support the merger, he said literally we could have taken all the money that we've invested in 5G stuck in a bank account and we'd have more money than we would having invested it all in 5G. So we've got to let this run, we've got to let it run. There's a lot of capabilities that come with 5G SA, which I think consumers and businesses want. Network slicing at the enterprise level is really interesting. It's a lot cheaper and more capable than private networks. And then at the consumer level you've got fixed wireless access where you can guarantee throughput and we've got gaming services and even social media services that could use that. So I think the customers want some of the capabilities of 5G SA. I think that the industry wants a payback, so that's why you should vote for the motion. Thank you very much.

Chris Lewis, Lewis Insight (00:18:48):
Well I think Varley showed what he thought of that while turning on his back. Now of course there's a perfect solution to your problem Graham, which is titular inflation, which you've never been shy of in the 40 years I've known. You just change your business cards, you don't even have business cards, right? It's a software issue and I think the view of 6G is that, and when I used to draw 5G diagrams, when I used to use diagrams, I always put 5G i'd, I'd put IOT in there, I'd put wifi in there, I put fixed networks in there. And I think the fact of the matter is that 5G has, as we've said, we're sort of disillusioned with 5G. We don't necessarily need a technology refresh, but we do need a refresh in terms of the positioning of all of the elements. A lot of the discussion we've had today has been about how the relationship, not why are we still talking about fixed and mobile separately, we must get more towards the solution view of the world and as we think about the things that we will have to connect and support in the future, the uplink has been mentioned several times and quite a few people have mentioned smart glasses, these meta ray raybans, if you're living in the US I could use it for a live analysis of what's been done in the orders in front of me.

(00:20:06):
It doesn't work here yet. So I could actually work out whether guy was lying when he said he was in favor of the greens in the second debate. But that requires a major, major improvement in the uplink. It requires obviously connectivity through to whatever GPU has been processing in the background if we're going to deliver. And in fact Nokia the capital market, they recently talked about not even the response times, we talked about generally within 5G, but even sub millisecond response times if we're going to get that automation, that factor, automation, the sort of things that Graham's ports need to deliver to make sure the containers are analyzed and everything, especially video. And if we think about the markets we've struggled to address and not only is it 10 years from 5G, it's also 10 years since the CEO of Erickson talked about 50 billion connected devices in the IOT world.

(00:20:56):
I think the latest figures are we're at nine for that figure. So we do need to refresh, we need to rethink, we need to have a framework within which all of the connectivity, not just the cellular, but we bring wifi into that discussion, we bring the fix into the discussion because if we don't do that we're going to miss out on those opportunities. And of course the beauty of this as we go through the automation sequences that we talked about in the last debate as we move that way, we do not need physical changes in the infrastructure. It's a software led development, so actually moving towards the 6G and we know from historical that every other G actually delivers a lot better value. So for goodness sake, let's forget 5G, I don't care about it. As you said, let's look at a framework that actually puts all of the connectivity that we need for the future into the right structure that works with each other. That gives us a chance as an industry to support all of those demands that we know are there and we can address it in a way that delivers commercial response but also gives that technical support. Thank you very much.

Guy Daniels, TelecomTV (00:21:55):
Thank you. So there you go. There you go. Two strong arguments, four against there. I can't believe you're calling my vote calling into question. You do know I'm colorblind don't you? I said that earlier. I did tell you that I raised that earlier. Okay, very good. Great. Thanks very much Chris and Graham, let's chat about this now and we'll go to the room first for any questions. Is there anyone who hasn't asked a question yet today who we'd like to ask a question because now's your chance. Anyone in the back of the room there? Only if they're using 5G SA they can. Yeah, that limits us a little bit, doesn't it? Any questions? Anyone over here? Over here? Let's get these two over here. Either either bro.

Dean Bubley, Disruptive Analysis (00:22:39):
I've asked a question.

Guy Daniels, TelecomTV (00:22:41):
So my name is

Dean Bubley, Disruptive Analysis (00:22:42):
Dean, no,

Guy Daniels, TelecomTV (00:22:43):
Was volunteering another question so we're going to have to go to some repeats I'm afraid. Dean,

Dean Bubley, Disruptive Analysis (00:22:46):
My name's Dean and I'm a slice denier but I also one of the founders of the 6G reset initiative and one of the parts of that question that was phrased the motion was about the need for 6G and whether 5G standalone obviates that. So I'd quite like the interest in the panel's view on what they see as the needs for 6G because I think that that hasn't been articulated very well and that actually the sort of magic hexagon needs some serious editing that ITU put out because that's a government document not actually about users and customers. If we assume if we assert that 6G has a bunch of problems that it needs to solve, some of them are to do with better coverage, whether that is via satellite in remote areas or via network sharing and if for indoor there's lower power, there's a bunch of things to do with inclusivity, there's stuff to do with security maybe post quantum. I would say that an awful lot of the needs for 6G are completely orthogonal to whether we have 5G standalone or not.

Guy Daniels, TelecomTV (00:24:03):
Right. Great. We're going to go into more detail on 5G as saying route to 6G, but let's jump straight into this point now. What is the need for 6G? What is it addressing? Why RI

Thierry van de Velde, Nokia (00:24:17):
Dean, you and I remember the I diagram in our studies, right? I mean this 256 QAM the way we are communicating, you're sending me eight bits in one symbol but you're not doing that when you are going to the cell edge today because I will have a blurred reception here, I will no longer be able to distinguish the eight bits. I will probably fall back to 64 QAM. If you go to the cell edge today with 5G with 6G, we are putting AI in the run to distinguish your eight bits despite you sending me a weak signal from the edge. So we believe a lot and it'll happen in 5G advanced already for the uplink because we will deploy AI technology in our base stations in 6G. I hope that you in your smartphone will have the AI receiver to be able to distinguish more dots on the I diagram in the complex plane. Is that a good hint? So you will be able to reduce your power to me, right? To a point where I can barely distinguish the eight bits in the I diagram. You'll be saving probably two hours of battery time per day with 6G.

Dean Bubley, Disruptive Analysis (00:25:24):
So in other words, we care about the radio not the core.

Thierry van de Velde, Nokia (00:25:28):
That's right, that's right. That's very rightly said.

Beth Cohen, Luth Computer (00:25:31):
But does the customer care?

Thierry van de Velde, Nokia (00:25:34):
Well

Beth Cohen, Luth Computer (00:25:34):
I don't think they do

(00:25:35):
better battery life. Not at all.

Guy Daniels, TelecomTV (00:25:38):
I had every time I was told I'd get more battery power. Anita, do you want to come in this one?

Anita Döhler, NGMN Alliance (00:25:44):
Yeah, I would like maybe to, first of all, Dean, you answered your question partly by yourself because coverage and those type of topics are important. I would like to maybe decouple it a bit from the term 6G because I think what we are discussing is whether we need or whether there will be innovation in future. And then the second question is how we frame this innovations through 3G PP releases, maybe open source software driven components of innovation. So therefore I personally and no matter that I'm also just representing NGMN here, but I personally believe that innovation will be always there and the question is how we as an industry can prepare ourselves to be ready for innovation, which we maybe cannot even foresee at the moment in 25 for 2030 and by owned. And this is why it's so important that when we go forward that we think about how we can even better design the network so that they can be ready for AI native but maybe also still need to define in more detail what it means and all the other challenges we have in front of us. We can imagine some of the use cases we have been discussed today already, like additional uplink requirements or requirements when it comes to traffic, not only volume but also traffic patterns from agen AI and the like. But I think the main sentiment is really, and that's what we are also manding from an operator perspective and NGMN that we need to achieve flexible networks and then of course network operations need to be simplified and this is I think the hygiene factor we need to tackle also as an industry.

Guy Daniels, TelecomTV (00:27:35):
Great Thanks, Itsuma

Itsuma Tanaka, DOCOMO (00:27:36):
Jumping as well. Thank you. And just adding another angle, one of the reasons why we think we need 6G is, well this is about another uncomfortable fact but there are lots of new features around 5G that are still not deployed by the operators and why it's not just about the business case but technically many of them are backward incompatible. So we need some good excuse to introduce opportunity to unleash the full potential while it's in the standards and technology so that those requirements like energy efficiency, AI native can be fully

Guy Daniels, TelecomTV (00:28:19):
Deployed. Thank you so much

Beth Cohen, Luth Computer (00:28:22):
Beth. So in the US we successfully turned off 3G. It was kind of brutal. I will have to say my Luddite relatives got sent little tiny flip phones from Verizon for free to get them off 3G and that was so company that the industry could start using the 5G capabilities and features because they got rid of all the 3G devices. Six G to me is still so amorphous that I am struggling. I mean the companies are still trying to figure out beyond FWA and private 5G what to do with 6G. I mean we're not even finished going through all the great things that we can bring to 5G, so why should we just jump into something even more amorphous that hasn't come yet?

Guy Daniels, TelecomTV (00:29:24):
Yeah, thanks Beth and Thierry, you want to jump up on that one? Well, well I

Thierry van de Velde, Nokia (00:29:27):
Wanted to say part of the response on the core right dean, because you said my answer was about radio, but in core what should we do different in 6G from 5G? If you look at the 5G standards today, they have been defined in a perfect standard silo in a single system silo. So we have the EPS standards today and we have the five GS and indeed the working is very badly specified. We have this for the connoisseur, right? N 26 U dcom protocols that were in fact never deployed. Why? Because the vendors, we took those two specs together and we created a dual system product. Everyone deployed these dual system products with proprietary interworkings inside that CNF and that has to change for 6G. We have to create a 6G core that will from day one speak 5G access stratum and 6G access into the same standard well-defined core network function. We can no longer live here with these proprietary implementations where it's not possible. For example today, right to compliment your 4G core vendor with a new fresh 5G core vendor in 6G, we want to give you the ability to compliment your 5G core vendor with a new fresh dual system core vendor so that your 6G subscribers can go to a multi-access score from day one that's totally independent from your 5G standalone core. Go

Guy Daniels, TelecomTV (00:31:01):
Ahead Thierry, That's great. Hold that fault Neil. We'll come to you in a sec because Mark's been waiting their patient late so get Mark in and we'll come back to that. But just to remind everyone in the room when we're talking 6G, I mean we are like what, four and a bit years away from estimated freeze of that which is what 16, 17 quarterly meetings to go where there is an awful lot of discussion and compromise still to go. So we've got a long way to go. Mark.

Mark Gilmour, ConnectiviTree (Europe) (00:31:31):
I suppose, this is where my question comes in because we could have rebranded 5G NSA as just an evolution of the 4G air interface and it would've been a roaring success then up until now if we'd kept the 5G moniker away for a little bit. And now at this point when we start introducing the core, the standalone core, I mean it was only called standalone core because we decided well 20 18, 20 19 that we would split it up so we could get 5G in early. So in reality we haven't actually really deployed the 5G experience end to end yet.

Thierry van de Velde, Nokia (00:32:18):
No.

Mark Gilmour, ConnectiviTree (Europe) (00:32:20):
And so my question then is if we ditch the 5G core element of it, is that then putting a higher burden on the 6G core or whatever that's going to look like solve those problems? Surely 5G standalone as we call it now should be the path to that innovation that's going to come in that 6G era, the 2030 onwards and should be software driven and not necessarily should be a complete hardware rip out again. So it seems to me that perhaps there should be a, well what is that path? Is the motion somewhat arbitrary because actually the path to that should be via

Guy Daniels, TelecomTV (00:33:15):
That path you mentioned there is certainly requested by a good number of stakeholders equally to those. Don't anyone want to feel that?

Thierry van de Velde, Nokia (00:33:26):
Okay, question. I would like to take lift here. One important innovation out of 5G standalone that we have indeed not realized. I happened to have contributed to the standards in a domain called internet PD type layer two over 5G. The military love it because they can keep their military grade routers. We don't need to ask them for IP address space. We can compliment their wifi networks with 5G today. Not deployed. Not deployed because we have not got an eye open on our main competitor wifi. Wifi is a layer two network. Ethernet is a layer two network. When we go to a Volkswagen factory and they have connected robots there, but they want us to stretch that wifi network with 5G base stations. They don't want us to provide an IT service there. They want us to provide a metro internet forum service over 5G. Of course we have to provide that in that combined core dealing with 5G and 6G access. Of course it'll be a new chance to provide this 5G loan service to both 5G and improved 6G air interface users. Yes, we have to continue trying to show the market the added value of 5G standalone.

Guy Daniels, TelecomTV (00:34:53):
Well all of you, Anita, first

Anita Döhler, NGMN Alliance (00:34:56):
Just maybe to go back to the comment of Mark that we introduced 5G non-ST standalone first as an industry, I would like to add that a lot of options came with regards to migrations to standalone from non standalone but made it extremely complex. And one of the requirements we put out already last year is that we want it to be backwards compatible and most important also that there should be really no such complexity which we had was migrating from 5G non standalone to standalone when it comes to the future a generation of

Guy Daniels, TelecomTV (00:35:38):
Yeah, absolutely. Thanks very much Anita. And that letter is on your website, isn't it? To the 3GPP? Itsuma?

Itsuma Tanaka, DOCOMO (00:35:45):
Thank you. Rebranding to me 5G SA. The concept was built 10 years ago and now today it's a different world and towards 6G it's going to be even more different world because all the change around ai, the speed is incredible. So I don't think heap staying with the paradigm 10 years ago doesn't make sense today. And from my r and d perspective especially, we definitely need something better. Something will take all the lessons from the past operating 5G SA and also consider leapfrogging for those 4G NSA camp towards the better 6G. That's something we definitely need to work on specifically in 3G ppp,

Guy Daniels, TelecomTV (00:36:35):
Thanks so much. And Beth,

Beth Cohen, Luth Computer (00:36:36):
So I want to address a much more practical perspective on this. If we're talking about deploying 6G in the 2030 timeframe, that will be enough time that the radios we'll need a hardware refresh at that point anyhow, so it seems to make much more logical sense to just kind of get all the capital expense paid out. So that, I'm

Mark Gilmour, ConnectiviTree (Europe) (00:37:07):
Not sure that Ian's doing right now in four and a half years time,

Beth Cohen, Luth Computer (00:37:12):
But other companies have invested four and five years ago, so that's like a 10 year cycle. So that's kind of closer to a typical refresh cycle is seven to 10 or 15 years. So that's just finance. So you have to take that into consideration when thinking about when you're going to deploy the network. I mean let's be honest, 5G was put in for the benefit of the telcos because it saved them a ton of money and it solved the last mile problem. It wasn't because the consumers were rushing out to get new fancy phones, they weren't.

Guy Daniels, TelecomTV (00:37:57):
Thank you. Thank you Beth. We could develop that argument, but I wonder maybe come across to Neil who's also been waiting patiently to come in on something earlier.

Neil McRae, HPE Networking (00:38:06):
I think we have industry amnesia. Why is 5G SA not start? Well because someone decided they were going to define the architecture for how we run 5G and they decided, hey, we're going to do it on virtualization. There was one massive problem with that. No one could make it work. So then there's a panic, oh wait a minute, this 5G core thing is going to slow us down because we've not just defined the standard, we've defined how you must build it, right? First time ever in a standards forum where they stepped beyond just what the standard needed to achieve. So we have a panic, oh my god we need to roll out 5G because everyone, this is what's going on. Verizon didn't help because they wanted to do this other wacky thing that basically nearly killed Nokia, I think.

(00:38:59):
So what's the lesson from that, right? The lesson is, and I see the same mistakes happening with all the AI stuff that's being defined as part of the solution. If I use AI or something else to deliver something, why should anyone care at the end of the day? So we've just spent the whole last 10 years ripping CPUs out of radios. Now we're going to stick GPUs back into them to make them work. That feels super challenging to me. It feels cost prohibitive and certainly the benefits that we're going to be able to monetize from that feels incredibly challenging. So I wrote the business case at BT for 5G and our CEO stood up at MWC and said the case was awful. It was an awful case. It was terrible because the only thing that the case based upon was, oh, if we don't do it and someone else does, we might lose customers.

(00:39:59):
What an awful way to think about it. But that was the killer app was to keep customers. So what's different about 6G or 5G SA? So 5G SA, I think there is opportunity in that, but it's not in our traditional marketplace part. What I just said earlier on, it's about leveraging the capabilities to go into business, to go into things like smart cities and all these kind of more, I'm trying to think of a word, but industrial solutions, right? But that again requires a different engagement with customers and a different engagement with how you deploy it. It's not just sticking radios on the top of big masks and hoping that everything works and if we don't learn that lesson, if we make it too complex, I think the complexity arguments is very well said. We just find ourselves thinking why didn't we learn from the last time?

(00:40:51):
And that to me is super scary, especially when I see wifi literally selling hotcakes right now. There's never been a more competitive aggressive wifi deployment and also internal networks and actually driven because people realize they need it for AI solutions. So how do we ensure that we don't come up with a solution that's driven by almost a fantasy of reality as opposed to actually what can we deliver realistically? How can we make money out of it so that we can afford the next investment? Because right now Anita's members are saying we can't afford the next investment so please make it really cheap. That's ultimately what her letter says because why the business case was awful. Someone said we needed 5G for capacity. Hell no we did not. I did the maths on it. 4G would've bought us at least three more years if we just kept running it. The only driver for 5G was volume of devices because the 4G core could not cope with the signaling and that was proven time and time and time again. So we need to do a much better job of leading the industry and it starts with the standards. Please don't define the solution for us because if you do that one, it's really hard to generate competition and two, you're assuming that this capabilities all telcos across the world are the same when we know for sure that isn't the case.

Guy Daniels, TelecomTV (00:42:17):
There were so many extra questions that can come out with this. There really are, but let's address what we can at the moment. Thierry, do you want to come in here? Well, yes

Thierry van de Velde, Nokia (00:42:25):
In the core first, right, because my specialty, okay, so you have now two schools of top Neil and we are in one camp and there is another one. We are in the camp where standards should indeed refrain strictly from imposing any implementation format, whether it will be c NF 1.0 without operators CNF 2.0 CNS dipping into GPUs, CNS running on arm. It should absolutely not be in the 3GPP standard. Now machine learning is not the same as GPU. We have deployed 10 use cases of machine learning. For example, smart paging, predictive rebalancing the load in our core networks. We're doing that on X 86 without any GPU and we're doing that energy. So three BP should indeed reframe from in the standards specifying agents inside the three BP core network network functions, agentic ai, you know what I'm talking about? If those agents start to talk to each other in a core network, it'll destroy the compatibility between an AMF from Nokia and a SMF from another vendor because we will no longer understand their language to us. It'll destroy roaming as well. It'll destroy the agreements between MVN Os such as Sky UK in this country with VM O2 as hosting MNO. They will no longer understand each other's signaling if we allow that second school of thought to happen in the

Guy Daniels, TelecomTV (00:43:53):
Standards. Okay, thanks Thierry. Anyone comments on this one? Because of course 3GPP will say they're doing nothing. It's just what their stakeholders are telling them to do.

Neil McRae, HPE Networking (00:44:06):
Remember last year I said if you want to control this, get involved.

Guy Daniels, TelecomTV (00:44:09):
Yeah,

Neil McRae, HPE Networking (00:44:10):
That was the message.

Beth Cohen, Luth Computer (00:44:11):
Yeah, if you're not in you've got no say.

(00:44:14):
That's true for ORAN too. Absolutely.

Thierry van de Velde, Nokia (00:44:16):
If operators do not express an opinion on this very essential aspect AI agent inside the three P network function or outside as an agent configuring that network function, of course if we have an agent configuring outside, I think we can fix our 5G standalone roaming problem. I think we can industrialize the way we're going to configure these 5G standalone roaming agents. I think we can give that task to robots to do and next year I hope to be here with much more 5G standalone users in the room as a result of AI but let's not march into that pollution of the standards with what today. Sounds like a fashionable technology.

Guy Daniels, TelecomTV (00:44:59):
Thierry, thank you very much indeed. Anyone else want to comment at this point or we'll go to another question? You okay, let's go another question because one at the back. Let's go straight to the back.

Sherif Sedkey, Virgin Media O2 (00:45:07):
Thanks, My name Sherif Sedkey from Virgin Media O2, ORAN architecture. I have a question if I would like to remind us few data sets. One last August that 3GPP TSG has finished their work for release 19 and the study items for release 21 of the major outcome 6G will adopt the same waveform as 5G as one of the outcomes minimize the major change. So this is one solid fact then AI work is continuing and whatever work items and study items for 5G advance will move to release 20 and 21 and then we move to the core some changes and there is a two camps, as you said the major reel was significant learning only was in the devices when we said we start with the modularity of the device and the capability is then add on not from there what we used to do before. Now my question is why we don't call this 6G as a retrofit of a 5G system and give it several more years till maybe release 2022 and then we think about a real 6G with a more capability and more giving more breaths for this 5G investment. By the way, ran vendors just my own observation are to some extent saying that they will try to give a live extension for their BB use to support this 6G as a basic configuration of what is installed 26 or 27. So again, what's the idea here?

Guy Daniels, TelecomTV (00:46:49):
Are you saying we're artificially accelerating the need for 6G?

Sherif Sedkey, Virgin Media O2 (00:46:54):
I don't want to say, but it seems to me like this

Guy Daniels, TelecomTV (00:46:57):
Hypothetical scenario we might, yeah, right. Okay. Thank you very much for that question. Anyone want to address this Beth?

Beth Cohen, Luth Computer (00:47:04):
Yeah, I'll talk a little bit about that. Yeah, there's a feeling that there was some stuff that wasn't in 5G that basically got pushed into 6G and so there's a sense that oh well we need these features so we need to accelerate 6G. But again, it gets back to we haven't realized the investment, the ROI on 5G yet, so why are we rushing into 6G?

Thierry van de Velde, Nokia (00:47:34):
I tend to disagree. 5G is tremendously successful and when we are here in the debate also in the thesis, there is a blur in this debate happening right now on this stage between the success of 5G standalone, which is modest indeed confined to fixed wireless access and to some low

Beth Cohen, Luth Computer (00:47:53):
Band coverage private networks, private 5G

Thierry van de Velde, Nokia (00:47:56):
Private 5G, yes, it is happening in some segments, but then 5G is tremendously successful. The investments have been very profitable. The eight bits per second per hertz spectral efficiency is realized today in NSA, all your handsets are benefiting from that. Our customers love us for that spectral efficiency

Beth Cohen, Luth Computer (00:48:15):
Except the real customers, the end users don't give a crap.

Thierry van de Velde, Nokia (00:48:21):
Well they do because in Scandinavia they have been able to switch off wifi. The youth is switching off wifi to save battery. They're relying on 5G only. We see that they prefer 5G to wifi. If you give them the bandwidth, if you give them the throughput. I see youngsters, before I took the train to come here to London, I was amazed to see the number of people watching these pre downloaded Netflix movies. They download them on the platform because they know that on the train it won't be so good. I see people on the plane watching these pre-download movies, they downloaded that probably in 10 seconds on the platform they

Beth Cohen, Luth Computer (00:49:00):
Want. They had the ultra wide,

Thierry van de Velde, Nokia (00:49:01):
They want this super high speed from us. They're so satisfied, they love us.

Guy Daniels, TelecomTV (00:49:07):
So we have some differences of opinion here, which is what a debate's all about, isn't it? We've got different viewpoints. Did you want to come in on this

Itsuma Tanaka, DOCOMO (00:49:15):
One? Well I would. Well let's put whether 5G was successful. Okay, put that aside for now. Well as I said, we are dealing with the new paradigm, new problems around our society around the world today, right? And look at this as an engineering problem and this new problem, we need to solve it and if existing 5G SA could solve it, good. If it not, we need something new. So that's how we frame discussion in three GPP today. The two camps fighting hard each other every day.

Guy Daniels, TelecomTV (00:49:54):
Thank you Itsuma and Anita, I wonder what the problem is. What are the driving factors, the driving forces towards 5G SA that you are seeing from your members?

Anita Döhler, NGMN Alliance (00:50:04):
5G SA,

Guy Daniels, TelecomTV (00:50:06):
What's the top priorities?

Anita Döhler, NGMN Alliance (00:50:07):
Can I answer it a bit more broadly?

(00:50:10):
Of course to evolve the networks with regards to cloud nativeness with regards to modularity, with regards to AI assisted operations. But I would like to take a little bit different, that's why because I think I'm a bit lost with different feature discussions here as I think that I mean three GPP and NGMN of course we are also discussing and the entire industry is discussing and do we need to say 6G? So what does 6G means? So we link it to specific 3G VP releases and it's important to of course when we move forward as an industry to decide what is part of those releases. And this is where we as NGMN are active, consolidating operator opinions, working with vendors, working with academia. We are providing input to 3G VP. Of course it's not done by just NGMN providing an input. It also demands, and that's backing up what Neil said last year and today of course it demands also the operators and all the different participants of the ecosystem contributing in 3G PP to the next releases.

(00:51:27):
But we are doing this and so AI for networks, networks for ai, I know it's a bit of a buzzword, but we are really breaking this down to specific requirements and we are just submitting license statements with content on this to through GPP for the meetings next week. So we have meetings through GPP meetings next week and I think that's the important driver to think about from a value perspective and what is needed to support innovation and the different trends we have in the industry, which we already mentioned I believe today in the next releases. And then of course the question is what is really needed to be part of the releases and what comes with software driven innovation with open source and other topics. So when we look at the development of ai, the question is can we cope with this as an industry which is really dependent on standards. So which parts need to be standardized and which parts will maybe not be standardized by 3G VP but will be important also to enable interoperable solutions.

Guy Daniels, TelecomTV (00:52:40):
Thanks very much Anita. That's great. That just reminds me because the meetings are next week if anybody wants some end of year reading, there's about a thousand PDFs they can download next week to keep up to date with what's going on there. Thanks very is a question at the bat. Andrew.

Andrew Collinson, Connective Insight (00:52:55):
Thank you. I'm really sorry because I'm going to go back to the ultimate arbiter of truth and justice in my world, which is my wife who will say to me, Andrew, why isn't the wifi working on my phone when I'm on the train? So my point of raising this is who cares 6G 5G, who cares and what is this problem? What is this AI problem? The only credible answer I've heard about difference of AI use on the network is in the core networks, which I've heard from Dean saying, yeah, moving traffic around between data centers. What is it that's so important about AI that we really know about today in 2025 that's going to be in existence in two or three years time? That makes a huge difference to the use of mobile networks that aren't Wi-Fi. I'm not saying it doesn't exist, I'm just saying I haven't yet heard it.

Guy Daniels, TelecomTV (00:53:47):
Okay. Okay Andrew, and if anyone can answer that in context of this, please do. Thanks Beth.

Beth Cohen, Luth Computer (00:53:55):
You're spot on. Actually, the average Joe Schmo consumer does not see the value at this point in how it's going to affect their use of their device. I mean we're talking consumers obviously at the enterprise level. There's a whole lot of use cases that are being developed. I wrote a white paper that was published a couple months ago about edge AI and how the telco is affected by that, but that's all on the industry side. The average consumer doesn't see this at all and I don't see that they'll ever see it because it will just be hidden.

Guy Daniels, TelecomTV (00:54:37):
Hang a second Andrew, and let's move across because I'm aware of the time and I really want to get some more viewpoints in. Anita, did you put your hand up there?

Anita Döhler, NGMN Alliance (00:54:44):
Yeah, I think we need to also to go away a bit from the average consumer because there will be also a physical ai. There will be different use cases which just watch hand, which are not related to the, let's say a classic average consumer and that's something we need to record. I my wife

Andrew Collinson, Connective Insight (00:55:03):
A narrative device. What industry use cases are there demand and use solution that can't be met currently.

Guy Daniels, TelecomTV (00:55:12):
Right? That that's great. That's great. What is that use case? This really does address Thierry

Thierry van de Velde, Nokia (00:55:18):
Let me take it. Okay, so going back to this promising layer two over 5G and 6G, what Volkswagen told us when we wanted to install that private 5G network in their factory in Burg is they have these robots connected today to a local area network lamb and they're using a special technique on that called time-sensitive networking. In fact, every ethernet frame arrives at a fixed pace in time, okay? We have that very same technology possible on 5G standalone and so we were able to replace these wired connections that they had to these robots in these cages with a private 5G network delivering ethernet PDO type to those robots from now on, these robots are receiving one frame every 10 milliseconds and they didn't want that to be 9 11, 17. You're right, us on

Beth Cohen, Luth Computer (00:56:15):
4G, but that's 5G. That's

Thierry van de Velde, Nokia (00:56:18):
Not

Beth Cohen, Luth Computer (00:56:18):
60, it's existing.

Thierry van de Velde, Nokia (00:56:19):
5G.

Itsuma Tanaka, DOCOMO (00:56:20):
Absolutely. That's a great use case. Use case. Let's say prolonged battery life, your kids will love it and better coverage integrating all the different accesses and so on That think that's for 6G I'm sure

Andrew Collinson, Connective Insight (00:56:35):
They'd like,

Guy Daniels, TelecomTV (00:56:35):
Right? Why is that? We're going to skip just a few more because we have got literally about one minute left to wrap this up in. Unfortunately. Andrew, I'm very sorry about this. Do you want another example please?

Thierry van de Velde, Nokia (00:56:48):
Consumer. Okay, not in industrial chromecasting. My kids do it, right? They Chromecast on the wifi network to the TV, to the telly. I would wish that I would be able to chromecast my presentation from my phone to that screen over a 5G network. For that, I need to discover it. I need this layer two environment to appear. We call that a 5G line. And who's going to set up, who's going to detect the fact that I want to discover these devices? Ai. AI is going to configure these mini environments which are location based that are going to allow us to finally do broadcast multicast in a cellular environment and not just on wifi because our competitors have been doing it all the time on wifi cheap. That's right. We need to compete with them, Neil, we can no longer pass on that. Otherwise we will not complete the triangle that I was talking about, right? Which is E-M-B-B-U-R-L-L-C and massive machine type communication. This MMTC is exactly about that, about discovering more devices that are around you.

Guy Daniels, TelecomTV (00:57:57):
Those governmental documents are the worst things ever for the industry. I'm sure of it. I'm sorry, we cannot do another question now because we are literally 30 seconds away from the close, so we've got to end the discussion there and it was a great question. Thank you very much indeed. I'm sorry we curtailed it a little bit, but we could talk about this book and another hour we've had a discussion. Let's remind ourselves way back in time, it was only 30 minutes ago what the motion was. Let's have a reminder and we'd be very careful yet the success of 5G standalone will postpone the need for 6G. You heard the arguments for and against, so now's the time to vote on the final motion of the year green for red against. Please raise your paddle's now. I think you'll get it. Oh, it's tight, right? I'm going to need some help here. I'm going to need adjudication here. There's two paddles in the air there. I saw that trying just because of that. I think that's a green. Do you think that's a green? That's really close to the call, but I think we're having it as a green, so motion carried. Thank you very much.

(00:59:08):
It was more of a dead heat I've got to say, but again, yeah, it was really, really close. Don't really close. Thank you very much. To our panelists, thank you to everyone because we are at the end of the event now. Almost at the end of the event, I would like to thank our sponsors again because it takes a bit to put these events on, so to all our sponsors there, thank you so much indeed. Also to the TelecomTV production crew, the team that's in here, the team that's not with us today and the people that are not with us today who have done a great job all year. As always, thank you very much indeed. Thank you to our online audience for watching and participating in the polls and if you haven't participated in those online motions and please do so. I shouldn't have looked into that light there, should I?

(00:59:54):
Because all I can see now is a great big white circle in front of my eyes. Thank you to all of you here in London as well and for getting into the spirit of this debate because you really make this. Okay, we want to recap our emotions if we can and if Phil's ready with the recap, but let's have a look at our recap because this is a reminder online viewers, what emotions were motion. One was a Graham success and Graham, that one with quite a tight four for the Telcos must become AI infrastructure players. Second motion was also very tight. That went against, we just didn't believe in the motion there. No,

Graham Wilde, Three Group Solutions (01:00:34):
We didn't do that.

Guy Daniels, TelecomTV (01:00:35):
No, it is bad motion. Motion three was a bit more of a landslide. I think network APIs are fundamental to telco business model success. Motion four. Yeah, against network automation is a Trojan horse. No, we didn't believe that one and we've just had motion five and if by magic it shows that it was just about succeeded and we carried that just about. I'm not sure though. It was quite close. So if you're watching online and you disagree, now's your chance to vote and make a difference, right? That's almost it. Please join us downstairs in the networking room downstairs. There's all sorts going on there. Some hale, and hearty networking. You are welcome to join us. Please follow Sean out. He knows the way by now and he's also got the money. Don't follow Ray though. Whatever you do, you'll end up in the coffee shop now. We'll be back again next year right here for another edition of the Great Telco debate. It's in December. Again, I do hope you can join us. I really do, and just send you on your way a reminder of our next event, which is of course DSP leaders World Forum in May in Windsor. So for now from all the team, thank you very much indeed.