Gabriel Brown, Heavy Reading (00:00):
So the session here towards open ran deployments at scale come, so we have right now we have three of five. We're two missing. Let's see, let's see who we're looking for. I think we're looking for Mark Atkinson of you are here and Rob Soni as well.

(00:42):
Now here we are and Mark is coming. Okay, fabulous. We have everybody. Welcome Rob, thanks for joining us. So the session here, well Mark, oh, here he comes. Luck. What an entrance. The playing your tune Mark, please have a seat. Okay gentlemen, please take a seat. So the session here is towards open ran deployments at scale, applying operational experience. I think the kind of work in hypothesis is as operators get more experience using an operating O ran that will enable larger deployments over the next half of the decade. The session format's going to be a little bit different here. There are no slides. I'm going to ask each person to do a few opening statements and then we're going to move a little bit more to a discussion type format. As before we're going to take some q and a please with your name and affiliation will be very helpful as well. So it's a terrific panel. I'm not going to introduce everybody individually, but just ask you to do so as I bring you in, starting with Abeta from N NTT Doomo and Oex Sirki, please go ahead tell us, you have I think one of the, if not the biggest multi-vendor base station deployments in the world. I guess a good starting point, which just tell us a bit about your operational experience so far.

Sadayuki Abeta, NTT DOCOMO (02:16):
Thank you. My name is Sadayuki Abeta, chief Open RAN strategy of ntt docomo and also the CTO of OREX SAI. First of all, thank you very much for inviting me. So I like to share that NTTDocomo experience first, so NTT dot develop 5G Network 2020 and we use that or interface open interface and we have the day one, we have the one CUD vendor and two RU vendors. So that is the day one which we use. But thanks to the open interface and now we have the four CUDU vendors and four RU vendors. A number of the RU itself, I think we have the more than 30 type D type of are used that is already installed. Our network and the network itself already widely deployed. So nation widely we deployed 5G network and the number of customers for 5G exceed, I guess it's 50 million, so many 5G customer already enjoyed our network and we also introduced at the SA and the NSC.

(03:36):
So the one thing I like to say that our network fully much vendor interoperable network, so even the different vendors equipment, we supported Interv in front fall, but not only the ful but for the X two interface. Because we support N-S-A-N-S-A, we need to support the interability between the 4G and 5G through the X two interface and we have the 4G layer chain network, but on the top of our 4G legacy network we deploy 5G network. That mean that we support exit office without the partners. I think also the part is here so that believe that this is the biggest much vendor network, but why say that the people worry about how the much vendor network does work? Well we have already done and we didn't see that senior problem to introduce the much vendor network and of course that we need some test to introduce the mass vendor network for example, we need to interpret this between the DU and R RU and first test back to the 2019. Honestly it takes more than six months because specification or specification itself affects in 2019. So we use that one but through the vendors interpretation is slightly different so it takes the time.

(05:28):
But the 2020 we did some new CUD vendor and RDU vendors but that case best case is we takes only one week actually that a company next to me they are using that we took only one week to complete one core and the pixel put, so tion is or alliance is now stable and we can use that to different vendors ORU.

Gabriel Brown, Heavy Reading (06:05):
Okay, so let's leave it there. I'll come back to you in a moment, but that's kind of good evidence that multivendor system can work and work very well. Obviously you have a demanding customer base as well, so that should be clear. We'll come back to you in a moment. Soki, I'm going to go to Rob, Soni at t. Rob, perhaps you could bring us a little bit up to date on where you are with ran and then if I could just give a little follow up. I guess my question for you is why does RAN scale specifically matter to you when at and t is this huge network already you have good scale even before O RAN came along.

Rob Soni, AT&T (06:40):
So first thank you, thank you for inviting me and allowing me to be together with these esteemed colleagues and partners. Enjoyed working with them for the last few years on O ran technology in a variety of different roles and a variety of different companies. So I'm only two years in a little bit more than that at and t, but at the same time in the last year, I think we can report rather significant progress in at t's march towards availability of O RAN technologies. I think our approach is a little bit different perhaps than what I hear a lot of times it's not open FRONTHAUL driven, it's open network management driven. So for us, the fundamental foundation actually to be behaving open is to create a platform, an environment where others can participate and we can manage in a single way. One of the biggest problems we have downstream is finally managing all of these partners.

(07:28):
So stability of the platforms themselves that actually allow for other partners to enter and exit the network are really important. Without that you'll fail. So from our perspective, yeah, we are now in the middle of launching our SMO platform itself. It will be widely available in our network later this year we've announced a number of application partners and we're working closely with Ericsson, particularly in support of those partners and developing an ecosystem, developing a plan, a promise and a capability for those app partners to be able to innovate so that they can gain scale. Not just working specifically on the Ericsson SMO platform but also working through the lab. We established in Dallas, the center of the universe. Anyone from Dallas in here? Yay. No, nobody agrees. Dallas is the center of the universe. There's always someone from Dallas. I just moved to Dallas in September.

(08:23):
It's been fun to transition there and one day wake up and it'd be 80 degrees and the next morning wake up and it's 13 degrees. I'm not kidding. That was within three days of each other and that was in January. It's a different place, an exciting place and it's a place where it has a strong telco heritage. And for those of you've been in the telco world for a while, no, Dallas is an important place. A lot of companies centered there and of course we're centered there. At and t is centered there. We worked with NTIA and through the public wireless innovation together with our partners, Verizon, doomo and Geo. We built a lab 22 ecosystem partners, some of which are in our network, some of which are not to accelerate the availability of O ran technologies and to demonstrate to the world that it's no longer an r and d problem.

(09:12):
It's an iot problem and this is huge because for us to be able to demonstrate that we need scale, otherwise what happens is everything becomes a one-off an R and d project, which is not good for the industry and it's not good for us. We also announced our ability to consume both Fujitsu and Avenir radios and our network and we're planning on introducing them later this year in our network. This will be the first time we've had radio and baseband in a significant volume or scale from a different baseband and radio provider. We're already though anyways have introduced a number of O ran elements together with our partners, Corning and CommScope, who are now Amphenol and Airs span. I'm very excited to see them also be managed now specifically by the same platform. So this enables us to manage radio das small cells and ran under a single platform environment but also enable third party applications to be successful in this.

(10:09):
Of course we're looking forward to introducing cloud ran later this year in our network itself. Commercially, we announced about a year ago we did our first call on the Erickson platform for cloud Ran again not far from the center of the universe, red Oak, Texas. We did that roughly a year ago and so we've made some progress and we eventually the lab with a commercial solution specifically for them now and driving towards scale. So yes, scale absolutely matters for us. We have worked not just within our ecosystem and I think the challenge is if you stay a hundred percent focused on your ecosystem, the reality is the other operators don't necessarily benefit from that. They also don't get the scale and we've been able to leverage the work done by the Doomo team. They've been instrumental in helping us move forward in the Accord lab and move the ecosystem partners.

(10:59):
But we're also working with a variety of operators, not just in through the O Ran alliance and through the Accord lab, but also now through the telecommunications infrastructure project. Reestablishing that as a place where O ran work can be done and I think it will be at least known fairly soon that the focus of that will shift on the O ran piece towards SMO and really about enabling multi-vendor SMO. That should be something that we expect to see. And then again, it comes back to we need to create the scale for the disruptors. We need to create the scale for the incumbents to be able to enter our network, but we also need to create the environments where they can plug, and I hate that word but I'm going to say it anyways, plug and play. It's not a plug and play environment today, but we need to get to a space where finally we can get reasonable timeframes, reasonable durations for when we introduce new hardware, new software capabilities in the network, whether it's at the application layer or the SMO, whether it's at the cloud infrastructure layer in particular, enabling other companies, new entrants like Wind River or Red Hat to participate and support the ran.

(12:05):
We see good positive stories around them. And then finally to provide the scale opportunities for others to enter our network on the radio side. And we have lots of opportunities to do that going forward as we introduce new frequencies in our network.

Gabriel Brown, Heavy Reading (12:19):
Fabulous, thank you. Rob. Let me turn to Bernard at TELUS. Bernard scale of a different problem I guess in Canada, a vast country, but I think I'm not unfair saying TELUS is on global terms, kind of a mid-size operator in terms of population and revenue. You've got a lot to cover there.

Bernard Bureau, TELUS (12:37):
Yeah, that's right. Gabriel, just for context, in Canada the government asked us to remove equipment from a Chinese supplier. And so we are about halfway done with traditional infrastructure and we're doing the other half with open run and V run 4G and 5G without any proprietary equipment. And you're right, Gabriel, Canada is as big as Western Europe. We only have 40 million population and we cover 15% of the landmass because the rest is the wild and cold areas. But ultimately we still cover 99% of the population. We don't have the same site grid as in many Asian countries or even in Europe. And so when we started to deploy oan and vRAN, we were a little bit worried about the performance at the edge because we have a lot of edge, our site grid, our sites are far apart and you need to have really performing network functions to be able to provide the type of performance that our customers require.

(13:50):
So good news, as soon as we started deploying Oran and vRAN, we were already reaching our targets in terms of performance, in terms of availability, so very, very happy about that. When it comes to scale, as you said, we're not very large. We're about 10 to 15% of the size of the biggest US operators to give you an idea of where we are and yet we have been able to move forward quite significantly on Oran and vRAN as it stands today, our deployment is going really well. We've been able to ramp up to our scale, which is not the same scale as very large operators. We have 5% of our network that is Oran and vRAN and we're going to have 18% by the end of the year. We're going to be at 50% by the end of 2027. And in 20 28, 20 29, we're going to remove all of the traditional base band that we have and at that point in 2029 we will be 100% Oran and vRAN for 4G and 5G.

(15:04):
So we're really happy about where we are. And if I may, I'm going to just talk a little bit more about what you say about scale because we're not very large as we said. I think what's most important is the attitude of your partners. If you've got partners that are really there to support you there to fix issues as soon as they arise, this is what leads to success in an Oran and vRAN deployment. And thankfully we have good partners and I'm just going to name them very rapidly because they've been awesome in supporting TELUS in our deployment. It's Samsung Networks, it's Intel Safari Rapids platform running on an HP server. We're using Wind River and we have radios from both Samsung and a small company that's called BMI. And the goal, like I said, we're probably not as, how could I say it? Rigorous as some of the very, very large operators, our attitude is to move forward and resolve issues as they arise. And up to now it's worked out very well for us. Our partners and ourselves were very involved. As soon as there is a problem, we resolve it and we move on and we continue with our deployment and our deployment is actually going to be pretty good. And keep in mind, TELUS is operating nine channels of LTE per sector three of NR and soon to be five and our performance is totally meeting our requirements.

Gabriel Brown, Heavy Reading (16:42):
Fabulous, thank you Bernard. Before I go on to Gerardo, do we have a burning question from anybody and keep the mic person on his or her toes? Okay, we'll start thinking 'em up because we're coming back to the audience in a minute. Gerardo, at Qualcomm, give us your point of view then on scale in the O ran ecosystem. I mean let's just leave it at that and then I'll drill in a little.

Gerardo Giaretta, Qualcomm (17:07):
Yeah, thanks Gabriel and good morning everybody. So I think one aspect that maybe it's a little bit unique for us is you don't have scale. If you don't have semiconductor and chip solution then can support or run protocols at the right performance at the right power consumption, right? That's a key component. It's a foundational aspect. You can define very mature specification and so forth, but if you don't have the platform, again from a chipset software perspective, which is at the right price point at the right TCO at the right power consumption, then operator not deploy open ram because it's simply too expensive and too power hungry and so forth. So that has been our focus and our investment in the last five years I guess. And we are very happy with what we achieved so far. I think we have talking about scale last fall, we have launched a commercial service in Vietnam based on our platform with vie a great partner and that's really a full open run massive MIMO deployment at scale because there's going to be a nationwide new 5G network in Vietnam.

(18:29):
So we're talking about 20,000 plus sites and this also speaks to some critics that they say open run is not good for massive o. Well open run is good for massive IMO and we have proven that and people can travel to Vietnam and prove out. So that's an example of how if you work with key partners like Vitel and if you invest in the right technology from a silicon perspective, then you can enable high scale ecosystem on that perspective. I think the other deployment that they're working towards is with DoCoMo in Japan for their VRA project and also with Dox with a besan. So on that front we are working with the NEC with HP with AWS. And again, if you think about these two networks are very different in the sense that you see Vietnam, it's a network where of course as I said, high performance because Massimo, but we're of course it's a new network, it's a developing country and so very cost sensitive and so we demonstrated that open run can be deployed in such a network with those targets and then this year we're going to deploy in DoCoMo, which is I think known to everybody as the toughest network to be approved for.

(20:03):
So that's demonstrated to the quality of the product and the performance that an open run solution can achieve. So these are just two real example, meaning we have deployed last year, we are deploying this year of scale deployment that with the right partnership with the right technology, I think open rank can cover every single use case that we can think of at the right cost structure, at the right performance. I think I want to go back one for us, one key foundational technology is making sure that every newcomer, every disruptor has also access to the best silicon technology because if somebody wants to enter the round market and has to rely on all the silicon technology, they're never going to be able to compete with incumbent. So that's how we are opening up from our point of view the ecosystem and we have great partners in NEC, Fujitsu MAs and other.

Gabriel Brown, Heavy Reading (21:03):
Okay, good stuff Gerardo. When you announced vitel, I followed up with them. We do sometimes when hear a vendor story, it took me a good while actually to get through it had to go up the chain, eventually went to the board to send back some comments but they did confirm. Yeah, they're very ambitious and they have a big rollout plan soly. It took like two months for me to, you

Gerardo Giaretta, Qualcomm (21:24):
Should have asked me,

Gabriel Brown, Heavy Reading (21:25):
I did ask you but I thought well I'll double check. No doubt. No

Gerardo Giaretta, Qualcomm (21:28):
I would've put in contact with,

Gabriel Brown, Heavy Reading (21:30):
Yeah, it's a very big deployment coming there given the time in the cycle. Mark Nokia so open ran for you. I mean I've been really nice to everybody. I was a little bit sharper with you. Is it a case at Nokia? You are already one of the big scale vendors, you have huge product volume, you have a great supply chain, you're nimble on that side with O Ran coming in. Is this not essentially fragmenting the market a little bit and taking some of your volume away?

Mark Atkinson, Nokia (22:01):
Wow, great question and good morning everyone. Delighted to be here on behalf of the network supplier community. As we know not all suppliers are the same when it comes to open ran, you've got the scale tier one suppliers, as you said some are doing more than others, we'll leave it there. Then you've got a lot of the smaller companies who either want to play in baseband or in radio or in both. One thing that we need to put on the table immediately is Open Brand is not a new technology, right? The foundation of everything we do is ran, ran needs to work, ran delivers performance, it delivers efficiency, scalability, security robustness, all of those things. And I don't think there's any operator in the world that would say I'm willing to go backwards or compromise in any of those things because of Open ran. So how do we do open ran then we open up interfaces around the RAN that works to enable multi supplier deployment scenarios and from the offset we had a very concrete plan is Nokia to focus on opening up on the baseband side.

(23:19):
Why? Because we believe there's a stronger opportunity to integrate third party radio suppliers than there is for third party baseband suppliers to succeed in this space. And we've seen many names pop up since the beginning of 5G. Many of them are still around in some marketing context, but most of them don't have a real business when it comes to baseband and it's because it's pretty damn difficult. What we do, we spend a few billions a year in r and d Soda, Huawei, soda Erickson, maybe the others a little bit less, but you need to have a lot of scale in revenue to be able to fund a big r and d machinery and do everything which the operators need to be competitive to monetize their networks. So we got to a point about a couple of years ago where we were able to integrate third party radios with our products and we have some very successful customer deployments ongoing right now where we have a multi-vendor scenario of Nokia baseband combined with third party rf.

(24:30):
Deutsche Telecom in Germany is one example we are super proud of which was announced earlier in the year DoCoMo in Japan, which was a perfect example of where Open ran makes sense. DoCoMo deployed at the beginning of 5G with an incumbent supplier, then decided over time that they needed more performance and they were able to get more performance by replacing the base band whilst keeping all of those radios on the tower. And of course everyone knows the expensive thing in terms of swaps or modernization is when you need to touch a physical tower site and climb and disrupt the network performance and users and so on. To come back to your question that we believe Open ran is a vehicle for us to actually grow our business rather than being concerned that it will constrain our business and as a whole we see that the opportunity is there to cooperate with different suppliers both in the macro radio as well as indoor solutions that however has a problem in itself that there are not that many radio companies around.

(25:41):
When you look in the kind of global view, there are some companies where everything points back to one or two companies in China. Then there are NEC and Fujitsu who are extremely good in what they do but don't have the massive scale of the tier one suppliers and you don't need just r and d capability, you need scale of supply chain, you need to place orders to component suppliers one year in advance in order to get what you need. So this site ability to be able to not only develop good technology but be able to deliver good products needs to be taken in context as well. Maybe a last word and not all operators are equal in the open ran sense. There are some operators like DoCoMo that say I want to be the integrator. I'm ready to take base bands, I'm ready to take radios and then I'm ready to take accountability myself for making the network perform. There are others who would rather say I want one supplier as a prime but I would like sub-suppliers behind them and you take over that work and so we for instance have Fujitsu radios in Nokia r and d processes labs and we treat them like Nokia products because that's the way every three months you get a new release out of the door and ensure that the release delivers the performance, scalability, all of those things I spoke about

Gabriel Brown, Heavy Reading (27:11):
Good stuff. Very good answer. I was too pointed I but I think the point you make about scale and r and d is obviously critical. I mean it's very hard. We've got Gerardo as well, no doubt humongous r and d budget but behind it and I think your point on the supply chain, I remember back to Covid when actually networks carried on and were built and it was a phenomenal success and it was because we had these big vendors with supply chain and the inventory and logistics really to get it done. Do you feel Mark that you've had obviously that answer to me is like you've embraced it, you see an opportunity to grow your business. Do you feel like you've had a clear enough steer from the customer base about what particularly they want? What's priority for them? Is it very diverse or are you now clear in your mind? What are the priorities?

Mark Atkinson, Nokia (28:03):
I think so we have different flavors of customer. We have some customers who want to do it because you can do it so there's not a technical or business rationale. There are others that want diversity of supply chain. There are others who think they can get to better cost points by having different suppliers and so once you have the technology then it becomes a business discussion and of course we have different business discussions with different players in the industry. I was actually very, very interested to hear from Rob and the focus on the SMO as well that typically we try to push the RAN discussion as an industry to open fronthaul but actually open ran is a lot more than open. We're talking about virtualization, we're talking about SMO and we need to embrace openness in each of those things. Are we the poster child in every area? For sure not. So we've taken our time in some of these versus others, but we have clear plans where we are going with all of them.

Gabriel Brown, Heavy Reading (29:09):
Yeah. Okay, good stuff. Do we have a question from the audience? There's a couple. There's a gentleman in the middle here. Go there first and they're going to bring you a microphone I think I'll tell you. We'll come back to you. Why they bring the microphone out? Let me pick up on the cloud or VRA point and let's go to Bernard and then to Abeta-san Bernard, just to clarify your comments, I think I understood well, but you was talking about by 2029 or let's give you till 2030, you're going to be a hundred percent. Is that the objective?

Bernard Bureau, TELUS (29:44):
Yes, and we're going to keep the radios of the tower that are traditional radios and we're going to have a C3 to E three converter or starting our first few sites this year and everything is going to be vRAN and o ran. Any new addition on those sites that our traditional run are going to trigger automatically conversion to V run and O run from the baseband with these converters

Gabriel Brown, Heavy Reading (30:08):
And maintaining the radios. I think that's a little bit like yes,

Bernard Bureau, TELUS (30:11):
Exactly mark's right. Vast majority of the cost goes when you need to send crews of the tower and unfortunately we've had to go through a few of them in the last decade and a half at Telus. So we are really allergic now to any vendor lock-in. Not that we want to change vendors all the time, it's absolutely not the case. We are committed to long-term relationships with our suppliers. The key is really having a choice and in many places throughout the building blocks of OAN today we have choices but we need to make sure that the choices are a little bit more vibrant, a little bit more healthy.

Gabriel Brown, Heavy Reading (30:53):
Okay, so the gentlemen, we're going to get your question in just a moment, but I want to bring in a better San on vRAN. Obviously you've done the front hall part. Give us an update on your, you mentioned it briefly but give us an update on where you are there and obviously Gerardo gave a little preemptory comment.

Sadayuki Abeta, NTT DOCOMO (31:10):
The question is focusing on only the vRAN well open around

Gabriel Brown, Heavy Reading (31:13):
Well on VA.

Sadayuki Abeta, NTT DOCOMO (31:14):
Oh okay vm? Yes. Yes. We have already deployed the VAM but it still is a small area and we are also adding that one more vendor this year and we are planning to deploy it but you know that we have already have the 5G network nationwide and so we gradually include that. So the William part and also keeping that traditional part because the operators how we is important. So not to know that learning is at some advantage but also we need to consider total cost. So based on our deployment strategy we use both the torrential hardware and also the Arbitraries

Gabriel Brown, Heavy Reading (32:12):
Hardware. Okay, good stuff. So let's have the question in the gentleman in the middle here. She's right there with you now.

Audience Question #1 (32:22):
Alright, thanks. My name is Manish Gange. I'm from HFCL in India. I've been on the supplier side and also on the operator side. In the past, one of the goals of Oran was not only to do the CapEx efficiencies but the operational efficiencies as well. Opex cost really. How do you see as the operators where we are heading, are we really realizing some of the operational efficiency goals with the open ran deployment and mostly because of the ization and thanks for bringing the V ran into the game.

Gabriel Brown, Heavy Reading (33:00):
We didn't want to go. Rob, do you want to have a go and then maybe Bernard,

Rob Soni, AT&T (33:02):
It's actually a really good thing to bring up. I think one of the things that we should be very direct about, about what's working and what's not working in O ran. So I think there's naturally a temptation to say, oh look, we stood up five vendors. Oh we set up this lab with 25 people. I'm making fun of myself actually it's bullshit. I'm sorry. The reality is it's about automation, it's about security. We introduce all these vendors into the network, we say to ourselves, oh yeah, we can do it. It's simple. It's not O RAN is transformational both for the vendors as well as for the operators. It changes their way of working, it changes our way of working. We actually have more complexity, which means we have potentially higher operational costs. So people often challenge, why can't you have, now you're open, can't you have 15 baseband vendors in the network?

(33:56):
Can't you have 15 ship set vendors in the network? Can't you have 20 radio vendors in the network? Can't you pick and choose, okay, New York City wants this radio, Los Angeles wants that radio but we're not there as an industry and you know why oh one and oh two are just not mature and we still very much are in a space where O RAN is not yet focused completely on maturing the O one and the O2 specifications. Yes, a hundred percent focused on open fronthaul but even that lifecycle Mark made that point. They decided to bring Fujitsu inside of the kimono to say, yes, that's a Nokia product. Why did they do that? Because they wanted to solve a problem for us. They wanted to be able to say, hey, you don't have to deal with the fact that software from here and software from there may not be aligned.

(34:44):
So we don't have yet mature processes for automation, but more importantly for security, the more people you have tripping around in your network, the more danger there is overall. I'm not trying to say the boogeyman is here he is. So the reality is we need to actually be very cognizant of the fact that this is really what's changing about ran and this is why operators are moving slowly and I know it's hard for the vendors, especially a lot of the disruptors and the incumbents to say, hey, how do I jump into this network and how do I gain scale? You have to find a partner and I'm seeing this consistently with all of the vendors. They find a specific operator partner to gain their own scale in o ran technology and then grow outward. It's impossible even for them to run in parallel 10 different scale deployments with operators.

Gabriel Brown, Heavy Reading (35:32):
Good stuff. Anyone want to add to that or we go to another question fella at the back.

Audience Question #2 (35:40):
Hi, that actually answered a lot of my first question but just introduce myself. My name is Dmitri De, I'm a PhD student from Virginia Tech in the United States and I am actually looking into security of O ran, especially how you're going from a single vendor configuration to a multi-vendor configuration. Obviously as you're including a lot more vendors, a lot more mobile network operators, non teleco vertical industries, your expectations are going to change, especially in the United States, you're including a lot of the rules and regulations that have to be considered regionally. So I guess my first question to you would be what efforts are being done to include I guess the standardization of all these security requirements for all of these different entities? And then my second question is including the introduction of quantum. So I've looked at your working group 11 documents quite a bit surrounding security over the new interfaces and everything and I'm noticing that a lot of them, much like most of the 5G network use asymmetric encryption, quantum is expected to break asymmetric encryption. So I guess what are you guys doing in preparation for the advent of quantum and quantum attacks? Okay,

Gabriel Brown, Heavy Reading (37:07):
So two really good questions. Both very difficult. It's fair to say thank you nra. I think he's still here so you can also try and catch him in the break which is in 10 minutes, but Rob, let's land this on you because you kind of queued him up and then we'll have whoever wants to join in on the security side. I mean

Rob Soni, AT&T (37:26):
I want to respect the other folks here on the panel and be all really short with my answer, but definitely Virginia Tech is a great partner for at and t. We work closely with the folks there, so definitely appreciate follow up if you want to on this topic itself. But in particular for me, security is not about the interfaces, it's about, it's actually about disaggregating hardware with software and creating new threat surfaces, new vulnerability services. So the challenge we have is often the Oren Alliance has been focused on security on the interfaces themselves. Look, we're not separating radio from baseband widely yet and we don't have across untrusted front hall. That doesn't happen. So worrying extensively about that interface and insecurity, it's something we do but we don't over rotate. At the same time worrying about our app's going to make an entry point of insecurity in the network.

(38:21):
Again, another focal point for the O Ran alliance, it's it's an important problem, but look, the apps run in our environment. We're going to scan them, we're going to look at them, we're going to secure them. So I don't think that's something that an SDO or a standards team needs to focus on. But yeah, there's a lot of opportunity I think overall to look carefully at what does a disaggregated solution really mean and how do you automate the process of determining whether they're vulnerabilities or not. We're doing a lot of this manually today and it's painful so we need to move forward quickly to automated, not just prior detection but also for intrusion detection because we live in a different world now and there are bad actors all over the place. So being able to automate that and look for and recognize when things are happening and have forms of anomaly detection. Again, it's a great app. Our app use case, it's a great piece. PhD dissertation topic. I think there's a lot of opportunity there.

Gabriel Brown, Heavy Reading (39:14):
Great, great. Good stuff. Another question, we've got time for maybe one or two more. There's a gentleman in the front and then this fella here, probably not going to go to Rob first for this one, but can come in later.

Audience Question #3 (39:31):
Hi, good morning. Hong down from Iowa State University, so actually at the center of AgTech. And so my question relates to how to bring high capacity connectivity to the rural areas and agriculture farms. So one of the challenges there is actually how to provide connectivity at affordable cost. My question is, given that there are some discussion the or around community, look at shared ORU spectrum sharing, just curious about your take beacon, maybe medium-sized carriers about the potential needs, how you see how that may potentially help or is there something worth to look at in the commercial deployment like shared ram and spectrum sharing?

Bernard Bureau, TELUS (40:19):
Bernard that's what's got your neighborhood I think mean. I think we're already in an environment where there's a lot of complementary solutions, whether it's home internet through satellite or through fixed wireless access from a terrestrial network. I think the difficulty is always going to remain that we've got spectrum that covers really, really far but with very narrow bandwidth and then we've got covers that the spectrum that covers is not as far but with a lot more bandwidth and so on and so forth. The difficulty is making a good business case while not having to densify the grid. That is really difficult for that. You need solutions, you need solutions where you're going to give as much room in your capacity for people that are at the edge from low band spectrum and for that you need to be able to offload low band spectrum as much as you can.

(41:20):
How do you do that? You extend the coverage of mid band so you can offload the low band consumption and how do you do that? You do that, there's more and more and you can see it from many different suppliers at the show here, more and more massive myo in the FDD bands of the mid band. We are also seeing some suppliers that are able to create products with maybe eight 16 TRX in the low band that essentially with the multi-user MIMO increases the capacity. So it's all of these things because there won't be more low-band spectrum in the future. So we need to rely on technology. Densifying sites is too expensive even if orran is a little bit less expensive. I think what's exciting I find is that hopefully in a few more years we're going to have more choices with radio suppliers to come up with these products because the evolution needs to keep on going. In our case in Canada, our traffic grows by about 30% people that are at the edge where only low band spectrum can reach them still grows by 30% every year and there is no more spectrum. So we need to rely on innovative antenna technologies to be able to get the best out of the coverage for those people that are far from the sites.

Rob Soni, AT&T (42:55):
Just want to add a little bit and it's good to see you shared. ORU is an academic problem. Why we'd rather share spectrum. That's what Bernard does. They share spectrum, they don't show radios, segmenting the radio, creating two different operational planes for two different operators to be on the same radio at the same time. It's an operational nightmare. Try to think about how it would be rationalized in the end-to-end in network. It doesn't seem actually something we're going to see in this decade. I'd love to be proven wrong. I think you can do a lot more by sharing the radio and creating roaming arrangements and roaming agreements between the operators and sharing the spectrum than sharing the radio. This is what we did with a ST space mobile. Look, we are direct competitors with Verizon but we are sharing spectrum with them for space access because we need low band spectrum, we needed available. If you look at even the model with starlink with D 2D, it's also a shared spectrum model. So as much as I love the idea, I'm a former academic myself, love the idea of creating all these problems for myself of saying how am I going to secure this? How am I going to manage this? I honestly don't believe the right approach should be to share the radio. It should be to share the spectrum.

Gabriel Brown, Heavy Reading (44:09):
Okay, big topic we can go more into but let's have a last question from here please.

Audience Question #4 (44:14):
Hello, my name is Ben, she, I'm from one Hungary, which is a service provider in Hungary. My question is, as I understood you are moving towards multi-vendor model and scaling up and as I counted you have more than eight vendors in your radio network. What's the main reason behind that? Is it bringing in competition or is it because you're innovative and want to test more and more vendors with their own technologies? Because it brings in serious operational questions. As my colleagues asked that

Sadayuki Abeta, NTT DOCOMO (44:47):
Ai, one of the motivation is that the completion, so we deploy the different type of network according to that our strategy. So some cases we needed the small transmission power for indoor, some cases the integrated type, some cases the machine al. So according to the deployment scenario, we asked the vendors to provide it and we selected the best solution. Of course we also need to consider the test cost. So based on the total cost we decide which vendor we will use and the other reason that avoid that supply chain risk. So as you may know that two to three years ago there the semiconductor shortage and some company had to provide solution on time. We can keep that our strategy deployment strategy plan and since we can use it other vendors so avoiding that it's a spread chain risk and also the competition.

Gabriel Brown, Heavy Reading (46:01):
Good stuff. They've padded us on the time. We've got a couple of minutes although it's now flashing red. Lemme say is there a last question out there? Chilling in front row.

Audience Question #4 (Chih-Lin I) (46:16):
Well we have three leaders up there and also appreciate one year ago the big news of the 14 billion move into the orran deployment even though some criticized that it's not multi-vendor version. However, I think it's very clear that as Orran technologies multifaceted, it takes time to mature and also it will need a system integrator. The question is whether the operator itself wants to be the system integrator or partner, a professional system integration house to do the job or I think in some cases Ericsson, Nokia and this traditional manufacturers are actually in the best position to be the system integrator. Of course we pray that they will be open-minded with open arm to be willing to integrate many other third parties, modules or subsystems into the system. So I'm just wondering if we could have some sort of experience comparison between at t's experience and the Domo experience because DoCoMo from day one seems to took up the system integrators role and how is that experience in compared to at and t having a major group vendor playing the seat? Great role. I think someone says I have a prime partner, but it's basically the CI integrator. I would say

Gabriel Brown, Heavy Reading (47:55):
Let's do that. But could I perhaps suggest we asked RAO just to give us a little bit of a perspective on the support you had to give out to Vieta. I mean it's quite an interesting case, so not saying you're a system integrator, but tell us how you're involved in

Gerardo Giaretta, Qualcomm (48:08):
Supporting it. No, I think the question is very good and I can provide the two different example of the network that we deployed so far and the tel case is a case where the operator two ownership of the system integration, right? So tel with the VT, which is Vieta high tech, which is how the vendor inside, inside via the operator, they took the ownership of doing system integration. It's an open run single vendor solution. In that case, even though VIETA provides the layer two, layer three software and Qualcomm layer one software. So there is some diversification in the software stack, but that is the case where the system integration is done in via the case of the work that we're doing with DoCoMo. I think it's a combination of NEC and DoCoMo being the system integrator. So it's a kind of a case by case scenario. I think it will depend eventually by, we have as you said, different kind of operators. Some tend to have, I don't want to say abandoned, but stop investing in the system integrator capabilities. So they will have to rely more on prime vendors or third party system integrator. Some I think we were part with the millimeter wave deployment of Rakuten a few years ago and even in that case Rakuten mobile was the system integrator. So some operators still have that very, very strong skill and I think they're great leaders on that.

Gabriel Brown, Heavy Reading (49:44):
I beta, I'll come to you in a moment, but I want to bring Mark in first. Mark, thank

Mark Atkinson, Nokia (49:48):
You. Maybe a little bit of context from our side as well that there are two clear worlds that I see the world where the operator has r and d capabilities and is willing to integrate themselves but also take accountability for the outcome. So if field teams see performance issues in the network, it's who do I call? I call the global r and d team within the operator or the work of integrating is passed on to the suppliers and from Nokia's side we've got a very clear positioning in this that we are willing and able and happy to be the integrator as long as we are the baseband supplier. Why? Because the baseband is the heart and brain of the network. The radios are that it legs and feet. And so if you are having control and knowledge around the heart and brain, you're able to integrate the third parties inside. Whereas a third party IT house would very much struggle because they don't understand ram.

Gabriel Brown, Heavy Reading (50:52):
Okay, better. Sam, I'll give you a last word unless Rob you want to come in and address it?

Rob Soni, AT&T (50:57):
Yeah, do you want the last word? I do have some answers but I'll let you decide. Do you want the last word or the first word? Wait. Okay, go first. It was the last word.

(51:09):
So it's a good question. I think if we think of traditional RAN today we have turf integration, we have site integration, we have physical integration, we have logical integration, we have software integrated from our vendors with our legacy IT systems, with our technology development systems. We have collectors, we have probes that we integrate. Finally, even at the site we have to integrate the radio with the baseband once we separated the radio from the baseband. So that happens today and it happens either physically or it happens via software methods. I think as we see the RAN continue to disaggregate. Yeah, there are more integration fronts potentially. But as we've seen with a lot of the operators as they start to deploy ran with the notable exception, which is why he gets the last word of doomo, is they're starting with a single solution as the learning point, as the travel point and then saying, okay, let me go bring in potentially another server vendor partner.

(52:03):
So we use Dell for example today and we're benefiting the fact that Dell and Intel are preintegrated. In fact they've been preintegrated outside of the ran domain for years and now it's an integration question for Dell and Intel with Ericsson, both with their CAS layer as well as with the network function layer. So for sure we will bring integrators in. The question is for those interior integration points, do you need, should it be a Nokia, should it be an Ericsson, should it be a Samsung? Or if you don't have a dominant baseband player, somebody enabled like the model with Qualcomm, then you're probably going to bring a third party in to do it. So I think those models are going to continue just like we have today for every other part of the network, we already are in the space of either sometimes and we either decide we use our own resources, we bring in contract resources or we rely on our vendors. It's going to be just an extension of it. And what we really hope is finally the software matures to make the software integration seamless.

Sadayuki Abeta, NTT DOCOMO (53:05):
It's the last word. I better, thank you. So the answer is actually that up to operators. So our case, we are the system integrators, but the RXI actually is bringing the solution to the outside Japan and all we are talking was the tier two, tier three operators. They do not have enough capability to do the integration by themselves. So we bring that preintegrated solution to them ensuring that our integration secure result as much as possible to reduce the cost as much as possible. So yesterday we had saw a session as with how we shared or how we operate to create that ecosystem to sharing that result as much as possible. So for the operators, how to reduce a total cost is important. So then the approaches may be different because size of scale of that different scenario is different, but we could provide the best solution to that practice.

Audience Question #4 (Chih-Lin I) (54:11):
I have a request because we are trying to set up a ecosystem resource, which is the best practice experience sharing and we currently only have one operators in Canada that's already come forth and provided their experience sharing as our first best practice. I would like to plead to our leaders up here to consider, as long as it's not the proprietary information that you can share, I think would benefit a lot with many of the ecosystem.

Rob Soni, AT&T (54:49):
Which working group is this being shared under?

Audience Question #4 (Chih-Lin I) (54:53):
Actually it's the executive committee that's taking care of this right now.

Gabriel Brown, Heavy Reading (54:56):
So direct feedback to the exec committee.

Audience Question #4 (Chih-Lin I) (54:57):
Right, right.

Gabriel Brown, Heavy Reading (54:58):
Okay. Let's leave that plea in the air well made though. And joining me in thanking all of our panel. Thank you very much.