To embed our video on your website copy and paste the code below:
<iframe src="https://www.youtube.com/embed/LteCbuUjGfM?modestbranding=1&rel=0" width="970" height="546" frameborder="0" scrolling="auto" allowfullscreen></iframe>
Guy Daniels, TelecomTV (00:24):
Hello, you are watching the Open Ran Summit part of our year-Round DSP Leaders Coverage. I'm Guy Daniels and today's discussion considers the future of Open ran. How might its development change to reflect commercial realities? How will it be integrated into the generational evolution of mobile systems and what is the future roadmap for Open ran? Well, to answer these questions, I'm joined on the program today by Ian Hood, CTO, chief Strategist Global Industries for Red Hat Petr Ledl, VP head of Network Trials and Integration Lab, and Chief architect of Access Desegregation program at Deutsche Telekom. Matteo Fiorani, head of distributed Unit and Infrastructure at Ericsson. Rick Mostaert, vice president of Product Management ran for Avenir, Beth Cohen, SDN Network Product Strategy at Verizon Business Group and Patrick Lopez, founder and CEO of Core Analysis. Hello everyone. It's good to see you all. What a great panel we've got today. No, as I mentioned, we are looking at the future of Open ran, so let me first of all ask how have our expectations changed in what's been a relatively short lifetime for Open ran, what hasn't quite worked out and what unexpected new ideas have come to the fore? Ian, perhaps we could come across to you for your thoughts first.
Ian Hood, Red Hat (02:07):
I think what we've got to do first is think about Open ran as a tool in the path on the journey to an open software-defined autonomous system. That's kind of the first thing is it's a tool on a path and if we sort of look at the things that have changed over the last couple of years, some of the actual commercial parties have changed in terms of what's up first because initially TCO would kind of at the top and that was one of the things people looked at along the performance and latency, things like that. But now kind of the focus is evolving to how do we take advantage of this and enable enterprise revenues and also increase consumer market share and go after that quality of experience. So that's kind one of the changes there. And while they didn't go away, now TCO is still high on the list, but now we've got automation to improve operation efficiency and the whole lifecycle management.
(02:58):
These are still sort of the key factors and so the things that have taken us longer in the industry is to get to the point where we sort of say, you know what? We now can deploy this at scale, meet the performance, and the economics are kind where they need to be over time. So we've kind evolved to that point and we've also got a very strong ecosystem. We've got multiple partners supporting it and this is kind of where we're headed to actually go do this and we can do this without compromise on the performance and also on the front H costs.
Guy Daniels, TelecomTV (03:35):
Well thanks very much Ian. Great summary of where we are at the moment, but let's get some other views on this. And Rick, let me come across to you first. What's your perception? How have expectations changed? What's been good, what's been bad?
Rick Mostaert, Mavenir (03:50):
I think from my perspective, some of the good things are, I'll start with the good is the ecosystem has exploded. Companies that are, I mean the amount of engineering, talent, innovation, if you look at companies like Nvidia, Intel, Amazon, everybody's in, we're publishing I think on LinkedIn you can follow John Baker and see the ecosystem and that is new. I mean we need that fresh influx of ideas and product to come in. So that's been fantastic. I think another thing that's been more prevalent lately, and I think that the prior speaker mentioned operational efficiencies and things like that, and now with AI Open Interface is looking northbound critical for again other companies coming and innovate on top of existing infrastructure, existing deployments. What I would say probably hasn't gone and it's getting better but it's been a little slow, is I think the radios ecosystem on the radio side started off very slow. In fact, to the point where we had to at Avenir actually create radio to tilt gap and hopefully we see all the vendors come in with open radios because that'll really enable true diversity in the supply chain, true diversity in the vendor ecosystem on the ran side, which is what I think the operators are really looking forward to.
Guy Daniels, TelecomTV (05:14):
Great. Thanks so much indeed, Rick, for your comments there. And I know all of our guests want to come in on this opening question. So let me first of all go to Matteo. Matteo, what are your thoughts of how expectations have changed?
Matteo Fiorani, Ericsson (05:29):
Thanks Guy. I think that open run is maturing as the previous speaker also said. I think definitely we see a very strong uptake. We see a number of things happening. As also mentioned before, cloud infrastructure are maturing a lot, so they go from being optimized for data center as also being optimized for one. For example, we start seeing strong solution for deploying single server system. That is one of the big challenges, the beginning for cloud infrastructure to scale down to distributed run deployment, which is fundamentally new for that type of applications. Then we see hardware platforms coming in. Of course Intel has a very mature platform, but then we see it now that there are also other challenges coming in with very strong solution such as for example, Nvidia and many others. And a lot of focus also on the automation aspects. Of course, as mentioned before with a lot of AOPs vendor for example, tried to enter the business with their own application to try to make things work. Of course this is complex so it will still take time for this ecosystem to fully mature, but we see a strong progress in line with the expectation. Definitely.
Guy Daniels, TelecomTV (06:39):
Great, thanks very much Matteo. Patrick, let's come across to you because it's only been a couple of months since you published your state of the state of the market report on Open ran. So what are your thoughts?
Patrick Lopez, Core Analysis (06:50):
Thanks guy. I think I'll just talk about things that haven't gone fast enough. Obviously we heard about a lot of progress and a lot of innovation on the open and run field. From my perspective, what hasn't progressed fast enough has been actually operators and carriers posture towards Open ran in the sense that, I mean open ran is a transformative project for network operator in the sense that you have to have a cloud native infrastructure, you have to have the processes, the tools, training in place in order to be able to deploy that. And also, I mean one of the desires of open run was the ecosystem multiplicity of vendors, the ability to pick and choose a solution from various vendors. And in order to do that, you need to be able to integrate them. And I think that open run is not expected to sport, but not all, but many operators have been waiting for others to actually do the hard work of doing the integration and the deployments before they can jump in. And I think that it's probably not the best approach because in order to benefit the most from Conran, you need to operate it and to operate it, you need to actually integrate it yourself probably.
Guy Daniels, TelecomTV (08:24):
Thanks very much Patrick. Well luckily we have two operators on the panel. So first of all, Petr, let me come across to you. Integration is very close to your heart and your job. What are your thoughts on how expectations have changed?
Petr Ledl, Deutsche Telekom (08:38):
Thanks. I think that I agree to the picture that that was outlined by predecessors here. You are right in terms of integration that this is a challenge that we need to tackle. And from BT perspective, we are also believing in the industry cooperation in this space to develop a framework, providing a certification and batching regime that would allow to decrease the integration cost that is related to necessary testing and alignment. So to a large extent we can do it once in operation and then maybe remaining 20% relief for the individual operator efforts that is pretty much attributed to their individual needs. So that's one thing challenge. The other thing that I would like to call out here as well is that from our perspective, architecturally we are really calling for the horizontally layered architecture and implementation. We really want to move away from the vertically integrated silos, even if we adopt approaches like Cloud run. So for that, what is needed is also acceleration of management interfaces. So management interfaces not only to network functions, but also to cloud infrastructure as an enabler. And that will allow us to adopt really independent management plane. In this sense, this is something where we are also investing to explore own development in this phase of SMO solutions perhaps. And this is really an opportunity that we see, that we see new bringing potentially will potentially benefit benefits.
Guy Daniels, TelecomTV (11:06):
Great. Thanks so much Petr and Beth, let's come across to you and hear your thoughts for this opening question.
Beth Cohen, Verizon (11:15):
So I want to add, I mean six, right? One of the things that I'm very hopeful about is that it's a cooperative effort between the vendors as well as the operators. So that always, I mean that's really what needs to happen to make it success because if the vendors are getting together and not talking to the operators or the operators are doing it separately, which the operators really can. But I do know that the operators are participating actively in providing the requirements and providing the support to make sure that the open room initiative is successful. And I think there's a high degree of motivation to make it as taking advantage of open source as much as possible and the disaggregation that cloud architectures and infrastructures facilitate with and incorporate all the newer technologies to make sure that Open ran is successful. Obviously the interface with the radios is always the tricky part because at some point you need to touch on the hardware, but I'm very hopeful that this will be a successful project. I do be aware that we need to make sure that everything's secured. We need to have hooks to allow us to incorporate into our backend systems, our OSS systems, our BSS systems, as well as make sure that the architecture is compatible with our underlying other cloud architectures.
Guy Daniels, TelecomTV (13:04):
Thanks very much Beth. And thank you everyone for those comments. And Ian, I'm going to quickly come back to you because we started with you. Do you want to quickly come back with a comment there?
Ian Hood, Red Hat (13:12):
Yeah, I did. One of the things that Patrick mentioned I think is really important and that is that operators really need to kind of control their own destiny here and become sort of the integrators along with their partners in building this out in this multi-vendor world. That kind of the first thing. And I think the other thing that because we've learned so much from building our LT networks and our 5G core networks in this same manner and built reference architects forward and kind of gone through that process, gone through the automation, the lifecycle management efforts, the complexity that we think we have is not as difficult as people have been making it out to be. So for those who haven't actually started yet and kind of wait for the others, they really need to take a page out of those who've actually gone ahead and are driving this who've actually kind of solved some of these things. So that's where I see kind the ones who've kind of said, I'm going to start now. I can't really wait, have I got the advantage of what they're going to go do? And those are kind of sitting on the sidelines. They really need to just get started.
Guy Daniels, TelecomTV (14:13):
Great. Thanks very much Ian. Well let's kick on because there's a lot of ground we want to cover today. My next point is whether or not this has been technically more complex than we originally thought. Have we perhaps over-engineered Open, ran, and Patrick, let me come across to you first.
Patrick Lopez, Core Analysis (14:33):
Thanks Ray. I don't know, I've been working almost 25 years in telecoms and Open RAN doesn't feel really more complex than other technologies that have been introduced. What's about Open RAN is that a lot of that technology used to be closed proprietary. So the complexity was there but it was hidden because it was basically managed by the vendors. But it's always been there now by creating open and disaggregated network for the first time in durran, basically you are exposing that complexity but with the benefit of having many vendors and many operators looking at it at the same time and proposing an architecture, a set of interfaces and protocols that are open, which actually removes complexity in the sense that well now they're better understood. And if you look at the journey of Open ran, I think it's remarkable to see that particularly the Fronthaul interface, which was probably one of the first world work items of the technology.
(15:50):
And that was at the beginning, Harley contested and there were a lot of doubt. It's not even something that we talk about anymore, or at least in its current version of category in the sense that now everyone takes more or less for granted that integration between virtual CODU vendors and R vendor following the specification of the or N Alliance. Well you are going to do testing, but the integration itself, so heavy lifting of that, that's pretty much come. So I'm just looking forward to see that same level of maturity going through each of the layer of open ran the back hole, the mid hole and interfaces with the SMO and near real time non real time rate. But I think it's progressing in the right direction. So answer your question again, I don't think it's been that it's been engineered properly because now you have a variety of vendors that are able to compete on the same field with the same open interfaces and we're seeing that the integration efforts are accelerating.
Guy Daniels, TelecomTV (17:01):
Great. Thanks very much Patrick. Yeah, maybe the questions really is whether or not we expect too much or we still expect too much rather than the over-engineering side of it, but let's get some other views on this and Beth, lemme come across to you next.
Beth Cohen, Verizon (17:17):
So where I see, I wouldn't say it's over-engineered, but I think it's different engineered and it's something that I think both the vendors and the telecoms have needed to overcome the hurdle of having a different skillset. So working with open software, working with disaggregation, working with cloud architectures and infrastructure is a different mindset, requires a different set of skills. And I think particularly in the beginning there was a lot of need to pivot certainly within the telecom industry, but I can see that the telecom industry has embraced that approach. And I know Open RAN is based on the container architecture rather than the virtual machine architecture and the containers themselves and bernet still have significant gaps that need to be addressed in terms of supporting multi interfaces and other types of basic telecom requirements. But the good news is that the CNCF is in fact supporting these new architectures and understand the importance to the telecom industry. I just attended the one summit last week and I attended an entire day of workshop activities around open ran, but there was a lot of cross-pollination with a number of the other projects, so it was very heartened to see that. So more to come.
Guy Daniels, TelecomTV (18:55):
Thanks beta. I like that. Engineered differently. We've got some more views. So Ian, let's come to you next and then we'll go around over the guests. So Ian first?
Ian Hood, Red Hat (19:06):
Yeah, I want to kind of follow on best comments there that it requires this different set of skills and that's really kind what has really kind of made it difficult for the operators to kind of upskill their teams and become confident in deploying these technologies. Learning how to go deploy in this, I'll call it differently engineered world, but we have seen that's an over, but I have to say we've kind of had what I would call feature creep over time of what the goalposts are and what is kind of our first okay four into what we're going to get done so we can actually get further down this path versus adding other things that yes, they'll become useful to add more value to get to that enterprise services and AI and those things. But I think we've hardened enough of it now to kind of know where those things come and with the work, as Beth mentioned as well with the other working groups, the security groups and the other aspects of these things, I think we're closing in on a lot of those gaps that people were concerned about and I'm seeing I'll call it less feature group now that we had early on.
Guy Daniels, TelecomTV (20:15):
Good to hear. Thanks very much Ian. And we'll come to Rick in a moment and also to Matteo, but Petr come to you next over-engineered or feature creep or just differently engineered, what's your thoughts?
Petr Ledl, Deutsche Telekom (20:29):
So I would say also that it is not really over engineered. I would say it's more complex. It is also more complex for us as operators to integrate and deploy and manage. But I would also like to highlight, and this is also why I would say that it's not over-engineered, I would highlight the work that was done in the OR alliance because this community has been working diligently to simplify the process. So by providing architecture design interfaces, specifications for key interfaces, test specifications for potential compliance and interoperability. So I think this was great help to navigate us to the complexity as much as possible.
Guy Daniels, TelecomTV (21:26):
Interesting. Thanks very much. Petr and Matteo, what are your thoughts? Yeah, look, actually
Matteo Fiorani, Ericsson (21:32):
What I would like to do is a bit summarize the key takeaway from the previous speakers because I agree basically with fundamentally everything they said, and I think one takeaway is fundamentally that though it has not been over-engineered, but rather it has exposed the complexity of building telecom systems that before was not fully exposed because was handled within single vendor deployments. So that is fundamental, the fundamental aspects. Then maybe the thing I would like to add is that the integration comes in different shapes and forms because in the OR alliance architecture there are multiple interfaces and multiple touch points. So there is one aspect being the vertical integration of the software on the cloud and hardware platform that I think is the one that has been tackled the most and I think it's the one that is getting more mature. Second point would be the integration more on the horizontal domain for example, on the software toward a radio on the frontal interface. And that is what is also maturing, especially on the CAT and the remote radios, maybe some more work to do on the massive milo but on the remote and the cat as Patrick mentioned before, it is mature. The last piece that I think is very important and we need to see some maturation on is on the management interfaces though are still in early days and I think we have not seen so much in multi-vendor in those aspects. So that is where we would like to see more maturation going forward.
Guy Daniels, TelecomTV (22:56):
So Rick, has this been more complex than we originally thought? Have we perhaps over-engineered open run?
Rick Mostaert, Mavenir (23:06):
No, I don't think it's been over-engineered at all. In fact, I think a lot of people thought it would be complex, but through experience we've actually discovered that the interfaces are very well defined and we've had a lot of success doing integrations and in fact we've seen large scale deployments now where multiple vendors have come together. The KPIs are very good, it works very well. So absolutely not. And just one more point I'd like to add is it's not just about do open fronthaul, we've seen a lot of innovation and openness within the whole RAN stack. For example, we have multiple vendors now coming in at layer one and by using the fpi small cell forum FPI interface, we've been able to integrate multiple layer one chip sets into our DU software and we see that happening across the industry as well. So the complexity has always been there. Like somebody else mentioned, we've opened it and exposed it and by doing that we've actually made it quite achievable to see the multi vendors come into the ecosystem here. So absolutely not over-engineered, it's been working very well in our experience.
Guy Daniels, TelecomTV (24:11):
Thanks very much Rick and thanks everyone for those comments. Now might we see more mainstream governance and specification development activities? For example, moving more work into 3G PP perhaps Matteo, can I come across to you for your thoughts on this one first?
Matteo Fiorani, Ericsson (24:34):
Yeah, that is an excellent question actually. I would start by saying a little bit about the OR alliance and the tremendous progress that the OR Alliance have done in the past years. Actually if I go back in time, I still remember one of the very first or alliance meeting in 2018 and I remember that the entire OR alliance community was fitting in a single room and we didn't even feel all the chairs. But now the community has enlarged a lot and this has grown a lot, or Alliance has a standardization body, it has matured a lot in the quality of the produced specification and in the participation of different vendors and operators. So there is a lot of work and tremendous work that OR Alliance has done. And another aspect that is very important is that now the OM alliance is becoming to some extent more global by submitting specification through Etsy Pass process so that they actually go into front terms and have broader reach because more companies are actually part of Etsy as a standard organization.
(25:37):
So this is actually very important to enable a global adoption of this specification. So we should keep in mind that Door Alliance is playing a very important role here. Then of course going forward and especially when six G comes into play, we need to think how an alliance and 3G PP will work together. What we want to avoid of course is duplication of specification. We need to make sure that or Alliance and 3G PP continue to work in a complimentary way and not in a conflicting way. And that can also lead to the fact that some aspects could be moved from the OR alliance into three GP when it comes to six G specification, what aspect exactly needs to be discussed and agreed of course, but that would be one way to avoid fragmentation, which is extremely important.
Guy Daniels, TelecomTV (26:24):
Absolutely nicely said Matteo, and I think it's really important to say that the Owner Alliance has done a lot of extremely good work over the years and continues to do so.
(26:34):
But
(26:34):
Where's the next step? The future of Open ran, what sort of linkages is it making with not just 3G BP, but Etsy and other associations? Beth, let's come to you for your thoughts on this one and then we'll get some additional thoughts after.
Beth Cohen, Verizon (26:49):
So speaking as a person who's very active in the LFN Linux Foundation network, and I was very heartened by seeing that there's a number of projects that are in alliance. I'm actually going to be working on, there's O Cloud, which is the infrastructure piece, and ICUT is the project within LFN, which is the set of infrastructure. It's a reference model, reference architectures that support these types of workloads. So I'm going to be exploring the synergies there. And also I know that there's some work to move some of this stuff into GSMA, which is also a telco standards body if you will, that supports the requirements documentation. So I am very heartened to see that yes, that the O Ran alliance is not sort of a standalone A, we don't talk to anybody else kind of organization. It's really looking piggyback on all the great work that other standards bodies and open source organizations have done in this area already.
Guy Daniels, TelecomTV (28:02):
Yeah, thank you Beth. And it's not just about moving towards more work into three gpp, but it's also bringing on board all the other associations that are doing parallel activities and broadening out that community of work. Ian, let's get your thoughts please.
Ian Hood, Red Hat (28:17):
I really liked Matteo's comments here on the fact that this open approach and disaggregating things has exposed kind of the complexity that was hidden by the perform proprietary ones. I think Patrick said the same thing, but in terms of where do we do the work, what I'm also really happy about how the O Alliance works is that they actually look at, let's take advantage of some IT U specs that are really important, IEE specs for synchronization, these kinds of things that already exist rather than let's recreate the wheel that's already there. Then how do we connect the work in GSMA and IEE and the OR alliance? And of course, as Beth mentioned, make sure that we have connections into the cloud native world and deal with the other interfaces that we know we're going to need to do open lifecycle management and then apply benefits of analytics and AI for not just for performance, but also for the ability to drive down the power consumption through our sustainability efforts. And that's another piece that's going on, if you will, within different aspects of the GSMA.
Guy Daniels, TelecomTV (29:28):
Yeah, thank you very much Ian. And yeah, the more we work on this, more groups work on this, the more we'll get done. I'm going to expand this out in a second too because Matteo mentioned a six G earlier. But before I do that, I'm just going to go across to Rick for your thoughts on how the initial work and so much has been done by Owen Alliance, how this might get expanded and enhanced.
Rick Mostaert, Mavenir (29:53):
Well, I think it's maybe more of a request and where I think it needs to go, we hope we get there is in addition to the specifications and the technology and I think bringing in specifications and technology from other industries, especially cloud, and again it's just bringing in more capabilities into this industry, which is fantastic, but we really to get to the end state, I think we need some kind of a product certification predation type of lab situation where it could become just much more plug and play. That's going to take some time, but I think the standards bodies need to take that up and make sure that's kind of built in as part.
Guy Daniels, TelecomTV (30:34):
Great, thanks very much Rick. Well I'm going to come across to Patrick now, but I'm also going to expand the question and move on a little because I also like to ask about whether or not open rank could become more generational agnostic and what impact it might have on possible six G scenarios and candidate technologies because it all kind of flows together these two questions. So Patrick, what do you think?
Patrick Lopez, Core Analysis (30:58):
I think we would be remit if we didn't mention at least briefly the telecom infra product, which at the beginning at least of open ran was a great promoter before there was the O Ran alliance. I remember one of the team that was leading at Telefonica wrote the first RFI ran publicly with Vodafone at the time and published the result and started really the work and it was very different because it was not specification, it was actually testing, it was actually putting people together, companies together and creating the foundation. So even though the tip has moved down to other projects, it's still part of open run. I think they deserve a mention here. And that goes to your question, how do we see open run in a multi-generational activity going towards six G? And probably Open Run was one of the first technologies that was very early on desire to be multi-generational in the sense that it was creating originally for 5G, but very quickly operators were saying, well we're not going to deploy 5G in isolation or we're going to deploy 5G on top of our existing network.
(32:21):
And it could be really difficult to deploy 5G on open run and then keep two G, 3G 4G on existing technologies and some vendors to get on to themselves to create versions that are two G, 3G 4G and 5G capable. And from that perspective it was already multigenerational. And when we look forward at six G, I think that most of the people on that panel will agree that it's not a question of whether open run will be part of six g open run will be part of six G and will be one of the architecture and one of the deployment options within the six G environment just by the virtue that basically those interfaces, the philosophy of open and desegregated network is now permeating a lot of part of six G even outside of the run itself. So open run if you want is configured a lot of the percept of architecture for six G.
Guy Daniels, TelecomTV (33:29):
Thank you very much Patrick and thanks for referencing tip there for its early work. We had tip on one of our panels yesterday where we looked at the rear Ian, lemme come across to you, is open ran already multi-generational.
Ian Hood, Red Hat (33:46):
I think Rick said it was sort of because of it being a open kind of approach and being cloud native by nature and distributed six G is its evolution to bring by more value on top of this infrastructure to enable those enterprise capabilities, the AI capabilities that kind of just fits into how this is going to evolve in this journey and there's no end state if you will, but I do agree with Rick that we need more things like we have the Oex in Japan, which is testing variations of the multiple flavors of how we interconnect the puzzle pieces. But there's also another group that's been formed in the last couple of years also driven by our friends at NTT in Japan. I was at their members meeting just a couple of weeks ago in Vancouver and they're seeing an evolution that includes the mobile side of the world and so people in a working group now looking at use cases to actually improve the distances that Open Ran can actually support by putting all photonic networks into that backhaul and fronthaul scenario, right? So there's another expansion potential coming out of yet another form where people are looking at how do we improve the capabilities of delivering these kinds of services around the world using photonics to drive down latency between data centers and also out in the front hall and backhaul
Guy Daniels, TelecomTV (35:09):
Options are increasing there. That's interesting Ian, thanks for that update. And Beth, what are your thoughts on Open RAN and how it's positioned or positioned as generational agnostic and what it can do for, we hate to say six G but we're going to have to six G.
Beth Cohen, Verizon (35:24):
So I think that Open Ran is very definitely thinking about the future and really architected to support multi-generations. But six G is very definitely still, we wouldn't call it a Iest project, it's probably a little beyond that, but I don't think the requirements are really fixed enough to really say that Open RAN can work toward supporting it because it's not really crystallized enough. But I think Open RAN is architected in a way that allows it the flexibility to support new generations of technology and new generations of thinking ai, hey we had to bring it up sometime AI and machine learning. What I heard is that six G is going to incorporate those aspects of the technology and Open RAN has to be ready to support that. So I think it is or it will be, but again it's still very cloudy.
Guy Daniels, TelecomTV (36:39):
Okay, thanks very much Beth. Let's come across to Matteo. What are your thoughts on the Open RAN and how is position for Future? What lies beyond 5G and being generationally agnostic?
Matteo Fiorani, Ericsson (36:56):
I think at Open is definitely a generational agnostic to some extent. Of course today it's designed to support mostly 4G and 5G and the architecture is being optimized for that. When it comes to six G, I think one key question we need to address is what are the six G use cases? What are we really building six G for? And then based on that we need to understand how the open run architecture can evolve to meet those use cases that we want for six G, there need to be an evolution of the open run architecture to be able to tackle six G and all the relevant use cases. There is no question that brand will evolve in that direction, but to understand exactly how that would look like, I think the first key question we need to address is really what are we building six G for? What are the key six G use cases?
Guy Daniels, TelecomTV (37:43):
Yeah, absolutely. Thanks Matteo. And hopefully that's hopefully that I was going to say five years away but it might not be. Hopefully that's four or five years away looking more the medium term. A final question for you. Can we outline a likely and honest scenario for Open RAN in the years ahead maybe like two or three years ahead, what is the future roadmap of Open Ran and Rick could we start with you?
Rick Mostaert, Mavenir (38:13):
Sure, that'd be fantastic. Yeah, I'm going to try to set aside my open ran Kool-Aid and be pragmatic about this. But I think if you look at the situation today we have open RAN has been very successful in standalone networks where there's not any existing barriers to entry. You look at Dish network in the US 20,000 sites, Japan, lots of SA enterprise networks around the world. Radio availability is good, a variety of RAN vendors, it's been there but the majority of the deployments that could go to open ran are in the future. So if you consider what's happening with pilots, lots of pilots announced across Europe, north America, Asia, those are in progress. We start to see this brownfield scale out happening really ramping up probably next year as these pilots get completed and multiple companies are involved with those types of pilots. And so as we get through 25, 26, 27 and start to scale this out, we will see a significant share of modernization projects.
(39:22):
Replacement projects will be all ran and that's been mandated. All the vendors are aligned and if so, as the vendors come through with their roadmaps or the ones that are already there, we will see that in the next five years really start to get some significant volume in the market and we are well on our way to get there. So good start. We're scaling out, we don't talk about is it possible it's not a size project anymore and now it's just we still have some barriers in entry. We have a lot of installed base that's not open. RAN is open, ran capable, but it hasn't been integrated. But as we do modernization, we'll see those types of networks will convert to open. I think majority of the industry is behind that now we're pushing forwards.
Guy Daniels, TelecomTV (40:12):
Fantastic. Thanks very much Rick. And we covered Brownfield commercial scenarios in a panel yesterday and there's a lot more to learn on that. Let's go to Petr next. You've been so closely involved with Oprah ran, what's your thoughts of the next two or three years? What's your roadmap?
Petr Ledl, Deutsche Telekom (40:33):
Yes, I would definitely build up on what was said heated before, so I would definitely subscribe to that. We should be seeing increased disruption and deployments as mentioned at the moment we see already scaled deployment, especially from the greenfield deployments, but we already have also committed from the major brownfield carriers. So we will see the deployments scaling in the Brownfield environment as the renewal cycles are coming. And of course by then it is important that we are able to move the horizontal model of architecture and break the silos and of course solve also the integration challenge. What we will be also seeing is the advancements in hardware and software developments that will again improve on performance and efficiency parameters that we need in the large scaled deployments and in the complex configuration and also especially in the software development to see the adoption that is allowing cloud run approaches and required automation as well as mentioned SMO and management layers with the evolved management interfaces and task but least of course ai, ML based optimization solutions that will help us to contribute to all the maturity, performance and efficiency that parameters that we need to achieve.
Guy Daniels, TelecomTV (42:44):
Thanks very much Petr. Lots to do still but lots to look out for as we're on the years ahead. We've got a few more views to get in before we close the panel. So Ian, let me come across to you next.
Ian Hood, Red Hat (42:55):
So I think I agree with what Rick and Matteo and team were talking about here in terms of where things are going in terms of brand itself. But as we got started, the next place for open ran is kind of actually in the private wireless kind of world where that's how I'm going to go deliver these enterprise services on that same horizontal cloud platform that we talked about. So that's kind of the next sort of thing we need to get moving on and get those use cases that Matto talked about. How do we go solve for those? But there's a couple of other things that are actually in flight we kind need to make sure we work on, which is as we all know, the amount of power being consumed in the network is majority in the radios but also in the DES and sees that are part of the architecture.
(43:41):
So Brett had an intel came up with an open source project to go look at the power consumption by the processors in Linux and cloud native platforms and to kind of drive down that approach. But in the advent of opening up the world, we have AI capabilities today, predictive analytics that actually speak to the back channels of the radios today to actually go adjust them for performance et cetera on the fly. And we need to kind of open up how we do that to drive down power consumption for sustainability purposes with more of open interfaces for that purpose that go not just to the CU and S that happen to run on standard COS hardware, but all the way to the radios who are much more customized hardware at this point as that evolves sustainability kind of another big piece of the puzzle that we need to work through. But this all leads to okay, we're building this highway, making it open, expanding it out, but if we actually want to get to those enterprise deliverable services on an open platform, we know we can do that with traditional ran. We need to actually put all these things cu do use IUs, ups and other applications on a common platform at the edge to go deliver those services and open RAN is helping drive to that path.
Guy Daniels, TelecomTV (45:03):
Very interesting. Thanks very much Ian for those observations. Beth, I'll come across to you next please.
Beth Cohen, Verizon (45:11):
So I want to pick up on several things. One that Ian said and then Rick also several people have mentioned the tech refresh. I think particularly the larger telecoms where there's Brownfield, I think we're going to see more open ran coming in as tech refresh kicks in. We have to replace the equipment every so many every so often and we saw this with SD-WAN with AS companies had to do tech refresh on the routers, they brought in SD wan. So I expect to see sort of similar types of incorporation of O ran into our infrastructures, particularly with the larger telecoms that have the more difficult brownfield. Another area that I see ORN really stepping up and being a really good fit for is the PIX wireless access, which is of course growing and that includes the private wireless as well as just a subscription based fixed wireless access. So there's great opportunity for those O ran boxes to scale it down to be the single router antenna boxes that we need to support the FWA implementation. So I see great opportunity for O ran to really be a really good fit for those use cases.
Guy Daniels, TelecomTV (46:54):
Thank you very much Beth and Patrick, based on what you
Beth Cohen, Verizon (46:58):
See and allies
Guy Daniels, TelecomTV (47:00):
In the market, how do you extrapolate this? Where do you see the roadmap and the path ahead for open run over the next few years?
Patrick Lopez, Core Analysis (47:08):
I think there are a couple of possible scenarios for the evolution of open run and pragmatically. If you are an operator and you're looking at introducing a new technology such as open ran and possibly a new vendor, this is a huge undertaking just because it's producing a new technology in Duran is expensive and lengthy and introducing new vendors is also expensive and lengthy, generally speaking in telecoms. So we started to see a trend let's say in Raiders wanting to deploy something that is let's say open ran, capable and even open ran, but with maybe not a lot of variation in terms of number of vendors, maybe either a single vendor option or when I say single vendor, I mean for the radio and the CODU or maybe CODU and radio from a single vendor with maybe another vendor for other radios. Then I think that's fine pragmatically to get started to get deployed, right.
(48:23):
I think that we're going to see over the next few years though possibly for the first time vendors that are not traditional radio vendors are going to bring something that is unique or that is very different from other vendors. And we mentioned the Rick earlier, this is probably where we see the most innovation, the most efforts regarding the integration of machine learning, deep learning and AI into networks generally speaking. And we're starting to see signs of companies that are not necessarily radio vendors traditionally that come up with Algorithmics that can actually change quite radically the way we deploy it, the way we manage radio networks, either from a spectrum efficiency standpoint or from a power consumption standpoint or from a traffic management standpoint. And I think that as those technologies and new vendors start to mature in the market, there is a chance here that really accelerates and change a lot the way we think about deploying networks in the near future.
Guy Daniels, TelecomTV (49:43):
Right. Thanks very much Patrick for those insights and we've just got time to go across to Matteo for your thoughts on the future direction of Open Ran.
Matteo Fiorani, Ericsson (49:52):
Yeah, thanks. I think a lot have already been said. So what I maybe want to highlight is just three quick things. First is that definitely adoption will continue to grow. We are in trials and soon in deployments some of them will become large scale. Also I expect with Brownfield operators, the two things that I think will help in that to go in that direction is one is what we define as pre-integrated solution. So especially in the early days of open run to mu a little bit of the complexity of the system integration, having vendors to partner and doing pre-integration is one great solution to be able to accelerate solution that are ready for the customer and ready for deployment. The second important aspect to make this deployment successful is actually that we need to be able to meet the performance and KPIs of the additional networks also with the cloud native and open network. That is something that is challenging overall because especially for large operators, this can involve the coordination of many different frequency layers that are deployed in different ways. So this could be a challenge that need to be addressed to be able to deploy at scale, but we do see that this is the direction we are going.
Guy Daniels, TelecomTV (51:04):
Great. Thank you very much, Matteo. We must leave it there though, although I am sure we will continue this debate during our live q and a show later. For now, thank you to all our guests for taking part in our discussion. If you're watching this on day two of our open Ran summit, then please send us your questions and we'll answer them in our live q and a show, which is coming up later. The full schedule of programs and speakers can be found on the telecom TV website, which is where you'll also find the q and a form and our poll question For now though, thanks for watching and goodbye.
Hello, you are watching the Open Ran Summit part of our year-Round DSP Leaders Coverage. I'm Guy Daniels and today's discussion considers the future of Open ran. How might its development change to reflect commercial realities? How will it be integrated into the generational evolution of mobile systems and what is the future roadmap for Open ran? Well, to answer these questions, I'm joined on the program today by Ian Hood, CTO, chief Strategist Global Industries for Red Hat Petr Ledl, VP head of Network Trials and Integration Lab, and Chief architect of Access Desegregation program at Deutsche Telekom. Matteo Fiorani, head of distributed Unit and Infrastructure at Ericsson. Rick Mostaert, vice president of Product Management ran for Avenir, Beth Cohen, SDN Network Product Strategy at Verizon Business Group and Patrick Lopez, founder and CEO of Core Analysis. Hello everyone. It's good to see you all. What a great panel we've got today. No, as I mentioned, we are looking at the future of Open ran, so let me first of all ask how have our expectations changed in what's been a relatively short lifetime for Open ran, what hasn't quite worked out and what unexpected new ideas have come to the fore? Ian, perhaps we could come across to you for your thoughts first.
Ian Hood, Red Hat (02:07):
I think what we've got to do first is think about Open ran as a tool in the path on the journey to an open software-defined autonomous system. That's kind of the first thing is it's a tool on a path and if we sort of look at the things that have changed over the last couple of years, some of the actual commercial parties have changed in terms of what's up first because initially TCO would kind of at the top and that was one of the things people looked at along the performance and latency, things like that. But now kind of the focus is evolving to how do we take advantage of this and enable enterprise revenues and also increase consumer market share and go after that quality of experience. So that's kind one of the changes there. And while they didn't go away, now TCO is still high on the list, but now we've got automation to improve operation efficiency and the whole lifecycle management.
(02:58):
These are still sort of the key factors and so the things that have taken us longer in the industry is to get to the point where we sort of say, you know what? We now can deploy this at scale, meet the performance, and the economics are kind where they need to be over time. So we've kind evolved to that point and we've also got a very strong ecosystem. We've got multiple partners supporting it and this is kind of where we're headed to actually go do this and we can do this without compromise on the performance and also on the front H costs.
Guy Daniels, TelecomTV (03:35):
Well thanks very much Ian. Great summary of where we are at the moment, but let's get some other views on this. And Rick, let me come across to you first. What's your perception? How have expectations changed? What's been good, what's been bad?
Rick Mostaert, Mavenir (03:50):
I think from my perspective, some of the good things are, I'll start with the good is the ecosystem has exploded. Companies that are, I mean the amount of engineering, talent, innovation, if you look at companies like Nvidia, Intel, Amazon, everybody's in, we're publishing I think on LinkedIn you can follow John Baker and see the ecosystem and that is new. I mean we need that fresh influx of ideas and product to come in. So that's been fantastic. I think another thing that's been more prevalent lately, and I think that the prior speaker mentioned operational efficiencies and things like that, and now with AI Open Interface is looking northbound critical for again other companies coming and innovate on top of existing infrastructure, existing deployments. What I would say probably hasn't gone and it's getting better but it's been a little slow, is I think the radios ecosystem on the radio side started off very slow. In fact, to the point where we had to at Avenir actually create radio to tilt gap and hopefully we see all the vendors come in with open radios because that'll really enable true diversity in the supply chain, true diversity in the vendor ecosystem on the ran side, which is what I think the operators are really looking forward to.
Guy Daniels, TelecomTV (05:14):
Great. Thanks so much indeed, Rick, for your comments there. And I know all of our guests want to come in on this opening question. So let me first of all go to Matteo. Matteo, what are your thoughts of how expectations have changed?
Matteo Fiorani, Ericsson (05:29):
Thanks Guy. I think that open run is maturing as the previous speaker also said. I think definitely we see a very strong uptake. We see a number of things happening. As also mentioned before, cloud infrastructure are maturing a lot, so they go from being optimized for data center as also being optimized for one. For example, we start seeing strong solution for deploying single server system. That is one of the big challenges, the beginning for cloud infrastructure to scale down to distributed run deployment, which is fundamentally new for that type of applications. Then we see hardware platforms coming in. Of course Intel has a very mature platform, but then we see it now that there are also other challenges coming in with very strong solution such as for example, Nvidia and many others. And a lot of focus also on the automation aspects. Of course, as mentioned before with a lot of AOPs vendor for example, tried to enter the business with their own application to try to make things work. Of course this is complex so it will still take time for this ecosystem to fully mature, but we see a strong progress in line with the expectation. Definitely.
Guy Daniels, TelecomTV (06:39):
Great, thanks very much Matteo. Patrick, let's come across to you because it's only been a couple of months since you published your state of the state of the market report on Open ran. So what are your thoughts?
Patrick Lopez, Core Analysis (06:50):
Thanks guy. I think I'll just talk about things that haven't gone fast enough. Obviously we heard about a lot of progress and a lot of innovation on the open and run field. From my perspective, what hasn't progressed fast enough has been actually operators and carriers posture towards Open ran in the sense that, I mean open ran is a transformative project for network operator in the sense that you have to have a cloud native infrastructure, you have to have the processes, the tools, training in place in order to be able to deploy that. And also, I mean one of the desires of open run was the ecosystem multiplicity of vendors, the ability to pick and choose a solution from various vendors. And in order to do that, you need to be able to integrate them. And I think that open run is not expected to sport, but not all, but many operators have been waiting for others to actually do the hard work of doing the integration and the deployments before they can jump in. And I think that it's probably not the best approach because in order to benefit the most from Conran, you need to operate it and to operate it, you need to actually integrate it yourself probably.
Guy Daniels, TelecomTV (08:24):
Thanks very much Patrick. Well luckily we have two operators on the panel. So first of all, Petr, let me come across to you. Integration is very close to your heart and your job. What are your thoughts on how expectations have changed?
Petr Ledl, Deutsche Telekom (08:38):
Thanks. I think that I agree to the picture that that was outlined by predecessors here. You are right in terms of integration that this is a challenge that we need to tackle. And from BT perspective, we are also believing in the industry cooperation in this space to develop a framework, providing a certification and batching regime that would allow to decrease the integration cost that is related to necessary testing and alignment. So to a large extent we can do it once in operation and then maybe remaining 20% relief for the individual operator efforts that is pretty much attributed to their individual needs. So that's one thing challenge. The other thing that I would like to call out here as well is that from our perspective, architecturally we are really calling for the horizontally layered architecture and implementation. We really want to move away from the vertically integrated silos, even if we adopt approaches like Cloud run. So for that, what is needed is also acceleration of management interfaces. So management interfaces not only to network functions, but also to cloud infrastructure as an enabler. And that will allow us to adopt really independent management plane. In this sense, this is something where we are also investing to explore own development in this phase of SMO solutions perhaps. And this is really an opportunity that we see, that we see new bringing potentially will potentially benefit benefits.
Guy Daniels, TelecomTV (11:06):
Great. Thanks so much Petr and Beth, let's come across to you and hear your thoughts for this opening question.
Beth Cohen, Verizon (11:15):
So I want to add, I mean six, right? One of the things that I'm very hopeful about is that it's a cooperative effort between the vendors as well as the operators. So that always, I mean that's really what needs to happen to make it success because if the vendors are getting together and not talking to the operators or the operators are doing it separately, which the operators really can. But I do know that the operators are participating actively in providing the requirements and providing the support to make sure that the open room initiative is successful. And I think there's a high degree of motivation to make it as taking advantage of open source as much as possible and the disaggregation that cloud architectures and infrastructures facilitate with and incorporate all the newer technologies to make sure that Open ran is successful. Obviously the interface with the radios is always the tricky part because at some point you need to touch on the hardware, but I'm very hopeful that this will be a successful project. I do be aware that we need to make sure that everything's secured. We need to have hooks to allow us to incorporate into our backend systems, our OSS systems, our BSS systems, as well as make sure that the architecture is compatible with our underlying other cloud architectures.
Guy Daniels, TelecomTV (13:04):
Thanks very much Beth. And thank you everyone for those comments. And Ian, I'm going to quickly come back to you because we started with you. Do you want to quickly come back with a comment there?
Ian Hood, Red Hat (13:12):
Yeah, I did. One of the things that Patrick mentioned I think is really important and that is that operators really need to kind of control their own destiny here and become sort of the integrators along with their partners in building this out in this multi-vendor world. That kind of the first thing. And I think the other thing that because we've learned so much from building our LT networks and our 5G core networks in this same manner and built reference architects forward and kind of gone through that process, gone through the automation, the lifecycle management efforts, the complexity that we think we have is not as difficult as people have been making it out to be. So for those who haven't actually started yet and kind of wait for the others, they really need to take a page out of those who've actually gone ahead and are driving this who've actually kind of solved some of these things. So that's where I see kind the ones who've kind of said, I'm going to start now. I can't really wait, have I got the advantage of what they're going to go do? And those are kind of sitting on the sidelines. They really need to just get started.
Guy Daniels, TelecomTV (14:13):
Great. Thanks very much Ian. Well let's kick on because there's a lot of ground we want to cover today. My next point is whether or not this has been technically more complex than we originally thought. Have we perhaps over-engineered Open, ran, and Patrick, let me come across to you first.
Patrick Lopez, Core Analysis (14:33):
Thanks Ray. I don't know, I've been working almost 25 years in telecoms and Open RAN doesn't feel really more complex than other technologies that have been introduced. What's about Open RAN is that a lot of that technology used to be closed proprietary. So the complexity was there but it was hidden because it was basically managed by the vendors. But it's always been there now by creating open and disaggregated network for the first time in durran, basically you are exposing that complexity but with the benefit of having many vendors and many operators looking at it at the same time and proposing an architecture, a set of interfaces and protocols that are open, which actually removes complexity in the sense that well now they're better understood. And if you look at the journey of Open ran, I think it's remarkable to see that particularly the Fronthaul interface, which was probably one of the first world work items of the technology.
(15:50):
And that was at the beginning, Harley contested and there were a lot of doubt. It's not even something that we talk about anymore, or at least in its current version of category in the sense that now everyone takes more or less for granted that integration between virtual CODU vendors and R vendor following the specification of the or N Alliance. Well you are going to do testing, but the integration itself, so heavy lifting of that, that's pretty much come. So I'm just looking forward to see that same level of maturity going through each of the layer of open ran the back hole, the mid hole and interfaces with the SMO and near real time non real time rate. But I think it's progressing in the right direction. So answer your question again, I don't think it's been that it's been engineered properly because now you have a variety of vendors that are able to compete on the same field with the same open interfaces and we're seeing that the integration efforts are accelerating.
Guy Daniels, TelecomTV (17:01):
Great. Thanks very much Patrick. Yeah, maybe the questions really is whether or not we expect too much or we still expect too much rather than the over-engineering side of it, but let's get some other views on this and Beth, lemme come across to you next.
Beth Cohen, Verizon (17:17):
So where I see, I wouldn't say it's over-engineered, but I think it's different engineered and it's something that I think both the vendors and the telecoms have needed to overcome the hurdle of having a different skillset. So working with open software, working with disaggregation, working with cloud architectures and infrastructure is a different mindset, requires a different set of skills. And I think particularly in the beginning there was a lot of need to pivot certainly within the telecom industry, but I can see that the telecom industry has embraced that approach. And I know Open RAN is based on the container architecture rather than the virtual machine architecture and the containers themselves and bernet still have significant gaps that need to be addressed in terms of supporting multi interfaces and other types of basic telecom requirements. But the good news is that the CNCF is in fact supporting these new architectures and understand the importance to the telecom industry. I just attended the one summit last week and I attended an entire day of workshop activities around open ran, but there was a lot of cross-pollination with a number of the other projects, so it was very heartened to see that. So more to come.
Guy Daniels, TelecomTV (18:55):
Thanks beta. I like that. Engineered differently. We've got some more views. So Ian, let's come to you next and then we'll go around over the guests. So Ian first?
Ian Hood, Red Hat (19:06):
Yeah, I want to kind of follow on best comments there that it requires this different set of skills and that's really kind what has really kind of made it difficult for the operators to kind of upskill their teams and become confident in deploying these technologies. Learning how to go deploy in this, I'll call it differently engineered world, but we have seen that's an over, but I have to say we've kind of had what I would call feature creep over time of what the goalposts are and what is kind of our first okay four into what we're going to get done so we can actually get further down this path versus adding other things that yes, they'll become useful to add more value to get to that enterprise services and AI and those things. But I think we've hardened enough of it now to kind of know where those things come and with the work, as Beth mentioned as well with the other working groups, the security groups and the other aspects of these things, I think we're closing in on a lot of those gaps that people were concerned about and I'm seeing I'll call it less feature group now that we had early on.
Guy Daniels, TelecomTV (20:15):
Good to hear. Thanks very much Ian. And we'll come to Rick in a moment and also to Matteo, but Petr come to you next over-engineered or feature creep or just differently engineered, what's your thoughts?
Petr Ledl, Deutsche Telekom (20:29):
So I would say also that it is not really over engineered. I would say it's more complex. It is also more complex for us as operators to integrate and deploy and manage. But I would also like to highlight, and this is also why I would say that it's not over-engineered, I would highlight the work that was done in the OR alliance because this community has been working diligently to simplify the process. So by providing architecture design interfaces, specifications for key interfaces, test specifications for potential compliance and interoperability. So I think this was great help to navigate us to the complexity as much as possible.
Guy Daniels, TelecomTV (21:26):
Interesting. Thanks very much. Petr and Matteo, what are your thoughts? Yeah, look, actually
Matteo Fiorani, Ericsson (21:32):
What I would like to do is a bit summarize the key takeaway from the previous speakers because I agree basically with fundamentally everything they said, and I think one takeaway is fundamentally that though it has not been over-engineered, but rather it has exposed the complexity of building telecom systems that before was not fully exposed because was handled within single vendor deployments. So that is fundamental, the fundamental aspects. Then maybe the thing I would like to add is that the integration comes in different shapes and forms because in the OR alliance architecture there are multiple interfaces and multiple touch points. So there is one aspect being the vertical integration of the software on the cloud and hardware platform that I think is the one that has been tackled the most and I think it's the one that is getting more mature. Second point would be the integration more on the horizontal domain for example, on the software toward a radio on the frontal interface. And that is what is also maturing, especially on the CAT and the remote radios, maybe some more work to do on the massive milo but on the remote and the cat as Patrick mentioned before, it is mature. The last piece that I think is very important and we need to see some maturation on is on the management interfaces though are still in early days and I think we have not seen so much in multi-vendor in those aspects. So that is where we would like to see more maturation going forward.
Guy Daniels, TelecomTV (22:56):
So Rick, has this been more complex than we originally thought? Have we perhaps over-engineered open run?
Rick Mostaert, Mavenir (23:06):
No, I don't think it's been over-engineered at all. In fact, I think a lot of people thought it would be complex, but through experience we've actually discovered that the interfaces are very well defined and we've had a lot of success doing integrations and in fact we've seen large scale deployments now where multiple vendors have come together. The KPIs are very good, it works very well. So absolutely not. And just one more point I'd like to add is it's not just about do open fronthaul, we've seen a lot of innovation and openness within the whole RAN stack. For example, we have multiple vendors now coming in at layer one and by using the fpi small cell forum FPI interface, we've been able to integrate multiple layer one chip sets into our DU software and we see that happening across the industry as well. So the complexity has always been there. Like somebody else mentioned, we've opened it and exposed it and by doing that we've actually made it quite achievable to see the multi vendors come into the ecosystem here. So absolutely not over-engineered, it's been working very well in our experience.
Guy Daniels, TelecomTV (24:11):
Thanks very much Rick and thanks everyone for those comments. Now might we see more mainstream governance and specification development activities? For example, moving more work into 3G PP perhaps Matteo, can I come across to you for your thoughts on this one first?
Matteo Fiorani, Ericsson (24:34):
Yeah, that is an excellent question actually. I would start by saying a little bit about the OR alliance and the tremendous progress that the OR Alliance have done in the past years. Actually if I go back in time, I still remember one of the very first or alliance meeting in 2018 and I remember that the entire OR alliance community was fitting in a single room and we didn't even feel all the chairs. But now the community has enlarged a lot and this has grown a lot, or Alliance has a standardization body, it has matured a lot in the quality of the produced specification and in the participation of different vendors and operators. So there is a lot of work and tremendous work that OR Alliance has done. And another aspect that is very important is that now the OM alliance is becoming to some extent more global by submitting specification through Etsy Pass process so that they actually go into front terms and have broader reach because more companies are actually part of Etsy as a standard organization.
(25:37):
So this is actually very important to enable a global adoption of this specification. So we should keep in mind that Door Alliance is playing a very important role here. Then of course going forward and especially when six G comes into play, we need to think how an alliance and 3G PP will work together. What we want to avoid of course is duplication of specification. We need to make sure that or Alliance and 3G PP continue to work in a complimentary way and not in a conflicting way. And that can also lead to the fact that some aspects could be moved from the OR alliance into three GP when it comes to six G specification, what aspect exactly needs to be discussed and agreed of course, but that would be one way to avoid fragmentation, which is extremely important.
Guy Daniels, TelecomTV (26:24):
Absolutely nicely said Matteo, and I think it's really important to say that the Owner Alliance has done a lot of extremely good work over the years and continues to do so.
(26:34):
But
(26:34):
Where's the next step? The future of Open ran, what sort of linkages is it making with not just 3G BP, but Etsy and other associations? Beth, let's come to you for your thoughts on this one and then we'll get some additional thoughts after.
Beth Cohen, Verizon (26:49):
So speaking as a person who's very active in the LFN Linux Foundation network, and I was very heartened by seeing that there's a number of projects that are in alliance. I'm actually going to be working on, there's O Cloud, which is the infrastructure piece, and ICUT is the project within LFN, which is the set of infrastructure. It's a reference model, reference architectures that support these types of workloads. So I'm going to be exploring the synergies there. And also I know that there's some work to move some of this stuff into GSMA, which is also a telco standards body if you will, that supports the requirements documentation. So I am very heartened to see that yes, that the O Ran alliance is not sort of a standalone A, we don't talk to anybody else kind of organization. It's really looking piggyback on all the great work that other standards bodies and open source organizations have done in this area already.
Guy Daniels, TelecomTV (28:02):
Yeah, thank you Beth. And it's not just about moving towards more work into three gpp, but it's also bringing on board all the other associations that are doing parallel activities and broadening out that community of work. Ian, let's get your thoughts please.
Ian Hood, Red Hat (28:17):
I really liked Matteo's comments here on the fact that this open approach and disaggregating things has exposed kind of the complexity that was hidden by the perform proprietary ones. I think Patrick said the same thing, but in terms of where do we do the work, what I'm also really happy about how the O Alliance works is that they actually look at, let's take advantage of some IT U specs that are really important, IEE specs for synchronization, these kinds of things that already exist rather than let's recreate the wheel that's already there. Then how do we connect the work in GSMA and IEE and the OR alliance? And of course, as Beth mentioned, make sure that we have connections into the cloud native world and deal with the other interfaces that we know we're going to need to do open lifecycle management and then apply benefits of analytics and AI for not just for performance, but also for the ability to drive down the power consumption through our sustainability efforts. And that's another piece that's going on, if you will, within different aspects of the GSMA.
Guy Daniels, TelecomTV (29:28):
Yeah, thank you very much Ian. And yeah, the more we work on this, more groups work on this, the more we'll get done. I'm going to expand this out in a second too because Matteo mentioned a six G earlier. But before I do that, I'm just going to go across to Rick for your thoughts on how the initial work and so much has been done by Owen Alliance, how this might get expanded and enhanced.
Rick Mostaert, Mavenir (29:53):
Well, I think it's maybe more of a request and where I think it needs to go, we hope we get there is in addition to the specifications and the technology and I think bringing in specifications and technology from other industries, especially cloud, and again it's just bringing in more capabilities into this industry, which is fantastic, but we really to get to the end state, I think we need some kind of a product certification predation type of lab situation where it could become just much more plug and play. That's going to take some time, but I think the standards bodies need to take that up and make sure that's kind of built in as part.
Guy Daniels, TelecomTV (30:34):
Great, thanks very much Rick. Well I'm going to come across to Patrick now, but I'm also going to expand the question and move on a little because I also like to ask about whether or not open rank could become more generational agnostic and what impact it might have on possible six G scenarios and candidate technologies because it all kind of flows together these two questions. So Patrick, what do you think?
Patrick Lopez, Core Analysis (30:58):
I think we would be remit if we didn't mention at least briefly the telecom infra product, which at the beginning at least of open ran was a great promoter before there was the O Ran alliance. I remember one of the team that was leading at Telefonica wrote the first RFI ran publicly with Vodafone at the time and published the result and started really the work and it was very different because it was not specification, it was actually testing, it was actually putting people together, companies together and creating the foundation. So even though the tip has moved down to other projects, it's still part of open run. I think they deserve a mention here. And that goes to your question, how do we see open run in a multi-generational activity going towards six G? And probably Open Run was one of the first technologies that was very early on desire to be multi-generational in the sense that it was creating originally for 5G, but very quickly operators were saying, well we're not going to deploy 5G in isolation or we're going to deploy 5G on top of our existing network.
(32:21):
And it could be really difficult to deploy 5G on open run and then keep two G, 3G 4G on existing technologies and some vendors to get on to themselves to create versions that are two G, 3G 4G and 5G capable. And from that perspective it was already multigenerational. And when we look forward at six G, I think that most of the people on that panel will agree that it's not a question of whether open run will be part of six g open run will be part of six G and will be one of the architecture and one of the deployment options within the six G environment just by the virtue that basically those interfaces, the philosophy of open and desegregated network is now permeating a lot of part of six G even outside of the run itself. So open run if you want is configured a lot of the percept of architecture for six G.
Guy Daniels, TelecomTV (33:29):
Thank you very much Patrick and thanks for referencing tip there for its early work. We had tip on one of our panels yesterday where we looked at the rear Ian, lemme come across to you, is open ran already multi-generational.
Ian Hood, Red Hat (33:46):
I think Rick said it was sort of because of it being a open kind of approach and being cloud native by nature and distributed six G is its evolution to bring by more value on top of this infrastructure to enable those enterprise capabilities, the AI capabilities that kind of just fits into how this is going to evolve in this journey and there's no end state if you will, but I do agree with Rick that we need more things like we have the Oex in Japan, which is testing variations of the multiple flavors of how we interconnect the puzzle pieces. But there's also another group that's been formed in the last couple of years also driven by our friends at NTT in Japan. I was at their members meeting just a couple of weeks ago in Vancouver and they're seeing an evolution that includes the mobile side of the world and so people in a working group now looking at use cases to actually improve the distances that Open Ran can actually support by putting all photonic networks into that backhaul and fronthaul scenario, right? So there's another expansion potential coming out of yet another form where people are looking at how do we improve the capabilities of delivering these kinds of services around the world using photonics to drive down latency between data centers and also out in the front hall and backhaul
Guy Daniels, TelecomTV (35:09):
Options are increasing there. That's interesting Ian, thanks for that update. And Beth, what are your thoughts on Open RAN and how it's positioned or positioned as generational agnostic and what it can do for, we hate to say six G but we're going to have to six G.
Beth Cohen, Verizon (35:24):
So I think that Open Ran is very definitely thinking about the future and really architected to support multi-generations. But six G is very definitely still, we wouldn't call it a Iest project, it's probably a little beyond that, but I don't think the requirements are really fixed enough to really say that Open RAN can work toward supporting it because it's not really crystallized enough. But I think Open RAN is architected in a way that allows it the flexibility to support new generations of technology and new generations of thinking ai, hey we had to bring it up sometime AI and machine learning. What I heard is that six G is going to incorporate those aspects of the technology and Open RAN has to be ready to support that. So I think it is or it will be, but again it's still very cloudy.
Guy Daniels, TelecomTV (36:39):
Okay, thanks very much Beth. Let's come across to Matteo. What are your thoughts on the Open RAN and how is position for Future? What lies beyond 5G and being generationally agnostic?
Matteo Fiorani, Ericsson (36:56):
I think at Open is definitely a generational agnostic to some extent. Of course today it's designed to support mostly 4G and 5G and the architecture is being optimized for that. When it comes to six G, I think one key question we need to address is what are the six G use cases? What are we really building six G for? And then based on that we need to understand how the open run architecture can evolve to meet those use cases that we want for six G, there need to be an evolution of the open run architecture to be able to tackle six G and all the relevant use cases. There is no question that brand will evolve in that direction, but to understand exactly how that would look like, I think the first key question we need to address is really what are we building six G for? What are the key six G use cases?
Guy Daniels, TelecomTV (37:43):
Yeah, absolutely. Thanks Matteo. And hopefully that's hopefully that I was going to say five years away but it might not be. Hopefully that's four or five years away looking more the medium term. A final question for you. Can we outline a likely and honest scenario for Open RAN in the years ahead maybe like two or three years ahead, what is the future roadmap of Open Ran and Rick could we start with you?
Rick Mostaert, Mavenir (38:13):
Sure, that'd be fantastic. Yeah, I'm going to try to set aside my open ran Kool-Aid and be pragmatic about this. But I think if you look at the situation today we have open RAN has been very successful in standalone networks where there's not any existing barriers to entry. You look at Dish network in the US 20,000 sites, Japan, lots of SA enterprise networks around the world. Radio availability is good, a variety of RAN vendors, it's been there but the majority of the deployments that could go to open ran are in the future. So if you consider what's happening with pilots, lots of pilots announced across Europe, north America, Asia, those are in progress. We start to see this brownfield scale out happening really ramping up probably next year as these pilots get completed and multiple companies are involved with those types of pilots. And so as we get through 25, 26, 27 and start to scale this out, we will see a significant share of modernization projects.
(39:22):
Replacement projects will be all ran and that's been mandated. All the vendors are aligned and if so, as the vendors come through with their roadmaps or the ones that are already there, we will see that in the next five years really start to get some significant volume in the market and we are well on our way to get there. So good start. We're scaling out, we don't talk about is it possible it's not a size project anymore and now it's just we still have some barriers in entry. We have a lot of installed base that's not open. RAN is open, ran capable, but it hasn't been integrated. But as we do modernization, we'll see those types of networks will convert to open. I think majority of the industry is behind that now we're pushing forwards.
Guy Daniels, TelecomTV (40:12):
Fantastic. Thanks very much Rick. And we covered Brownfield commercial scenarios in a panel yesterday and there's a lot more to learn on that. Let's go to Petr next. You've been so closely involved with Oprah ran, what's your thoughts of the next two or three years? What's your roadmap?
Petr Ledl, Deutsche Telekom (40:33):
Yes, I would definitely build up on what was said heated before, so I would definitely subscribe to that. We should be seeing increased disruption and deployments as mentioned at the moment we see already scaled deployment, especially from the greenfield deployments, but we already have also committed from the major brownfield carriers. So we will see the deployments scaling in the Brownfield environment as the renewal cycles are coming. And of course by then it is important that we are able to move the horizontal model of architecture and break the silos and of course solve also the integration challenge. What we will be also seeing is the advancements in hardware and software developments that will again improve on performance and efficiency parameters that we need in the large scaled deployments and in the complex configuration and also especially in the software development to see the adoption that is allowing cloud run approaches and required automation as well as mentioned SMO and management layers with the evolved management interfaces and task but least of course ai, ML based optimization solutions that will help us to contribute to all the maturity, performance and efficiency that parameters that we need to achieve.
Guy Daniels, TelecomTV (42:44):
Thanks very much Petr. Lots to do still but lots to look out for as we're on the years ahead. We've got a few more views to get in before we close the panel. So Ian, let me come across to you next.
Ian Hood, Red Hat (42:55):
So I think I agree with what Rick and Matteo and team were talking about here in terms of where things are going in terms of brand itself. But as we got started, the next place for open ran is kind of actually in the private wireless kind of world where that's how I'm going to go deliver these enterprise services on that same horizontal cloud platform that we talked about. So that's kind of the next sort of thing we need to get moving on and get those use cases that Matto talked about. How do we go solve for those? But there's a couple of other things that are actually in flight we kind need to make sure we work on, which is as we all know, the amount of power being consumed in the network is majority in the radios but also in the DES and sees that are part of the architecture.
(43:41):
So Brett had an intel came up with an open source project to go look at the power consumption by the processors in Linux and cloud native platforms and to kind of drive down that approach. But in the advent of opening up the world, we have AI capabilities today, predictive analytics that actually speak to the back channels of the radios today to actually go adjust them for performance et cetera on the fly. And we need to kind of open up how we do that to drive down power consumption for sustainability purposes with more of open interfaces for that purpose that go not just to the CU and S that happen to run on standard COS hardware, but all the way to the radios who are much more customized hardware at this point as that evolves sustainability kind of another big piece of the puzzle that we need to work through. But this all leads to okay, we're building this highway, making it open, expanding it out, but if we actually want to get to those enterprise deliverable services on an open platform, we know we can do that with traditional ran. We need to actually put all these things cu do use IUs, ups and other applications on a common platform at the edge to go deliver those services and open RAN is helping drive to that path.
Guy Daniels, TelecomTV (45:03):
Very interesting. Thanks very much Ian for those observations. Beth, I'll come across to you next please.
Beth Cohen, Verizon (45:11):
So I want to pick up on several things. One that Ian said and then Rick also several people have mentioned the tech refresh. I think particularly the larger telecoms where there's Brownfield, I think we're going to see more open ran coming in as tech refresh kicks in. We have to replace the equipment every so many every so often and we saw this with SD-WAN with AS companies had to do tech refresh on the routers, they brought in SD wan. So I expect to see sort of similar types of incorporation of O ran into our infrastructures, particularly with the larger telecoms that have the more difficult brownfield. Another area that I see ORN really stepping up and being a really good fit for is the PIX wireless access, which is of course growing and that includes the private wireless as well as just a subscription based fixed wireless access. So there's great opportunity for those O ran boxes to scale it down to be the single router antenna boxes that we need to support the FWA implementation. So I see great opportunity for O ran to really be a really good fit for those use cases.
Guy Daniels, TelecomTV (46:54):
Thank you very much Beth and Patrick, based on what you
Beth Cohen, Verizon (46:58):
See and allies
Guy Daniels, TelecomTV (47:00):
In the market, how do you extrapolate this? Where do you see the roadmap and the path ahead for open run over the next few years?
Patrick Lopez, Core Analysis (47:08):
I think there are a couple of possible scenarios for the evolution of open run and pragmatically. If you are an operator and you're looking at introducing a new technology such as open ran and possibly a new vendor, this is a huge undertaking just because it's producing a new technology in Duran is expensive and lengthy and introducing new vendors is also expensive and lengthy, generally speaking in telecoms. So we started to see a trend let's say in Raiders wanting to deploy something that is let's say open ran, capable and even open ran, but with maybe not a lot of variation in terms of number of vendors, maybe either a single vendor option or when I say single vendor, I mean for the radio and the CODU or maybe CODU and radio from a single vendor with maybe another vendor for other radios. Then I think that's fine pragmatically to get started to get deployed, right.
(48:23):
I think that we're going to see over the next few years though possibly for the first time vendors that are not traditional radio vendors are going to bring something that is unique or that is very different from other vendors. And we mentioned the Rick earlier, this is probably where we see the most innovation, the most efforts regarding the integration of machine learning, deep learning and AI into networks generally speaking. And we're starting to see signs of companies that are not necessarily radio vendors traditionally that come up with Algorithmics that can actually change quite radically the way we deploy it, the way we manage radio networks, either from a spectrum efficiency standpoint or from a power consumption standpoint or from a traffic management standpoint. And I think that as those technologies and new vendors start to mature in the market, there is a chance here that really accelerates and change a lot the way we think about deploying networks in the near future.
Guy Daniels, TelecomTV (49:43):
Right. Thanks very much Patrick for those insights and we've just got time to go across to Matteo for your thoughts on the future direction of Open Ran.
Matteo Fiorani, Ericsson (49:52):
Yeah, thanks. I think a lot have already been said. So what I maybe want to highlight is just three quick things. First is that definitely adoption will continue to grow. We are in trials and soon in deployments some of them will become large scale. Also I expect with Brownfield operators, the two things that I think will help in that to go in that direction is one is what we define as pre-integrated solution. So especially in the early days of open run to mu a little bit of the complexity of the system integration, having vendors to partner and doing pre-integration is one great solution to be able to accelerate solution that are ready for the customer and ready for deployment. The second important aspect to make this deployment successful is actually that we need to be able to meet the performance and KPIs of the additional networks also with the cloud native and open network. That is something that is challenging overall because especially for large operators, this can involve the coordination of many different frequency layers that are deployed in different ways. So this could be a challenge that need to be addressed to be able to deploy at scale, but we do see that this is the direction we are going.
Guy Daniels, TelecomTV (51:04):
Great. Thank you very much, Matteo. We must leave it there though, although I am sure we will continue this debate during our live q and a show later. For now, thank you to all our guests for taking part in our discussion. If you're watching this on day two of our open Ran summit, then please send us your questions and we'll answer them in our live q and a show, which is coming up later. The full schedule of programs and speakers can be found on the telecom TV website, which is where you'll also find the q and a form and our poll question For now though, thanks for watching and goodbye.
Please note that video transcripts are provided for reference only – content may vary from the published video or contain inaccuracies.
Panel Discussion
The telecoms industry started to collaborate on Open RAN back in 2016, and telcos worldwide started initial commercial deployments only last year. Expectations remain high as rollouts accelerate. However, given the lengthy planning times and long lifecycle requirements of network technology, what is the future roadmap for Open RAN? How might its ongoing development change to reflect commercial realities? And how will it be integrated into the generational evolution of mobile systems?
Recorded May 2024
Speakers
Beth Cohen
SDN Network Product Strategy, Verizon Business Group
Ian Hood
Chief Strategist - Global Industries, Red Hat
Matteo Fiorani
Head of Distributed Unit and Infrastructure, Ericsson
Patrick Lopez
Founder & CEO, {Core Analysis}
Petr Lédl
VP, Head of Network Trials and Integration Lab and Chief Architect of Access Disaggregation program, Deutsche Telekom
Rick Mostaert
Vice President of Product Management, RAN, Mavenir
