Now On Demand: Leveraging Cloud Native, in the Americas- CET 2021
The commercialization of Open RAN 4G and 5G platforms for enabling operators to take advantage of fully virtualized cloud-native network architectures is now in the Americas and should advance quickly. This executive session will offer the state of play for Open RAN 4G and 5G platforms and how Rakuten Mobile, who already disrupted the cellular market in Japan, plans to do the same in the Americas.
Executive Speakers:
Azita Arvani – GM, Americas, Rakuten Mobile
Ian Hood – Chief Strategist, Red Hat
Patrick Lopez – Global VP, Product Management, 5G, NEC
Eric Stonestrom – President & CEO, Airspan
Partha Seetala – Founder & CEO, Robin.io
Thierry Maupile – EVP Strategy & Product Management, Altiostar
Transcription Details
Time: 50:48
SPEAKERS
Thierry Maupilé, Patrick Lopez, Ian Hood, Eric Stonestrom, Host, Azita Arvani, Partha Seetala
Host:
The commercialization of Oran, 4G, and 5G platforms for enabling operators to take advantage of a fully virtualized cloud-native network architecture is now in the Americas and should advance quickly. This executive session will offer the state of play for Oran, 4G, and 5G platforms, and how Rakuten Mobile already disrupted the cellular market in Japan, plans to do the same in the Americas. Joining us are Azita Arvani, General Manager, Americas at Rakuten Mobile. We also have Ian Hood, he's chief strategist, at Red Hat. We also have Patrick Lopez, Product Management at NEC. Also, with us as Eric Stonestrom, he's President and CEO of Airspan. Also, with us is Partha Seetala, he's founder and Chief Executive Officer at Robbin.io. And last, we have Thierry Maupile, Executive Vice President of strategy and Product Management at Altiostar.
Speakers welcome to the program.
Azita Arvani:
Thank you
Host:
Thanks for being here. I hope I forgot everybody's name more or less pronounced correctly and if I didn't, I sincerely apologize. If you don't mind, Azita, I'd like to go right to you on the first question for the session. What is the state of play for accelerating the commercialization of Oran, 4G, and 5G platforms, that to enable operators across the world to take advantage of the first really fully virtualized cloud-native network architecture?
Azita Arvani:
Hi everyone, thanks for having this panel and having me on the panel. I would say that it is state of the art now, it's definitely much better than it was two and a half years ago, where we embark on this journey at Rakuten Mobile. But I also wanted all of us to kind of pause and reflect back on what has happened in the last several years. So, if you remember we had this seminal paper have never [unintelligible: 02:07] virtualization and software-defined networking back in October of 2012, almost a decade ago. And while we have operators that envision the cloud, and we had some suppliers that could envision some cloud solution, there was really not much going on in terms of deployment. And even for those who couldn't envision it, it was more for the less real-time network functions in the network, and not for the radio aspect. Being said, we looked more like I've seen and enterprise [unintelligible: 02:45] So, it was this notion of yes, we could need to do cloud but there was no real compelling reason, it was just virtualizing [unintelligible: 02:53-02:56] not giving us a big compelling reason. So, what Rakuten Mobile did was to say, we want to do a cloudify mobile infrastructure, not because it's an interesting intellectual exercise but because we have many cloud services, like e-commerce, travel, payments, insurance, logistics, TV, honjo, many more. And what we want to do is create a connectivity substrate that we could provide all these services on top of. And what we want is that because all of our Rakuten services, they reside in the cloud, we work in the cloud, cloud is in our DNA, our question was; how could we come up with this kind of activity substrate that gives us maximum flexibility in terms of adding new services on top, is very, very affordable and accessible to all kinds of subscribers and is also very secure? And because resided in the clouds, it was how do we do this in the cloud? And so, we had the compelling reason why we should go into Cloud and we had a compelling reason how we could do it in the cloud. Because again, the cloud is something that they've done with all the other services. So, in a span of a couple of years, with very strong commitment, and blood, sweat, and tears of ours and our partners, we went from this; we could do cloud, to we get cloud. And we all know that radio access network or our beloved vRan where costs go to party, that's where the 80% of the cost of the network is. So, we started with that rather than starting with the non-real-time parts of the system. And so, we worked with Red Hat, we worked with Altiostar, later we work with NEC, with Robin, with Airspan. And so, I would say that as an industry with all these distinguished partners we have here, we've traveled much further in the last two and a half years than we did in the decade prior. So, it is a great place to be in, we're ready for takeoff but at the same time, you need to be diligent because the industry this movement is still young. There are still a lot of naysayers, they're still all kinds of things that are being thrown out so we need to really be careful to cultivate it, to nourish it, whether it's by more operators deploying it, whether it's by governments giving incentives to it. So, I would say that we need to go from oh, yeah, one day in the future we will go to cloud and Open vRan to know, this is our present. We have to go from could do cloud, to we're doing the cloud.
Host:
So, let's talk about it and I'll go around the room here. Let's talk about the stakeholders in this cloud-native sort of space if you will. Let's start with Eric at Airspan, as you recently became a partner with the Rakuten Communication Platform called RCP. What is the significance of that as Oran is now really commercially viable?
Eric Stonestrom:
Well, I think, as Azita, the key is we have done it now. So, the first-mover advantage will be accrued with folks like Airspan and the others on this call, who've been able to actually make it happen in real life. Our collaboration with Rakuten in Japan clearly dismisses the concerns about whether Open vRan is commercially ready. I see it and know it every day, and we see the deployment, statistics, the traffic, the success, and it is. So, that's really a great position at this moment to now take these ideas to the rest of the world, there's been a lot of learning and we're very honored to be part of the group that has earned its stripes through the learning and look forward to welcoming more into the ecosystem. But the nice thing is, this is now based in reality, and it's happened now and it is happening now every day.
Host:
I want to go to Partha next, again, your role, sort of in a cloud-native ecosystem. So, how does Rakuten Mobile create an ecosystem around you to help with the requirements for these underlying cloud architectures?
Partha Seetala:
So, Robin is the underlying cloud platform on which the Rakuten 5G network runs, all the way from vRan, to their Mac clusters, to their centralized data centers, and so on. Now, when it comes to cloud-native deployment, there are a few things that need to come into play for the ecosystem to actually take benefit from. So, number one is, of course, the platform and a standardization around how network function vendors can essentially put their network functions on top of this platform. So, Robin has done quite a bit of extensive work in running, for example, a complex network function like NRAM. As Azita said, 80% of the complexity, and the spend, and the network is in NRAM. Now, with that standardization, and the world that we have done in pushing the Kubernetes, let's say platform power to be able to run these things. Rakuten has been able to go and onboard a lot of network function vendors on top of this platform in a very standardized manner. So, it's not like each one is going about doing their own way of onboarding and an own way of doing lifecycle management. That enabled Rakuten to very quickly go and build a large ecosystem that can interoperate with each other and that can be onboarded onto the network very quickly.
Host:
Patrick, I want to go to you next. Again, along the same lines, Rakuten Mobile headed core network relationship with NEC, and now really has a deepen that relationship to racket in specifications. Can you explain that to us?
Patrick Lopez:
Sure, I think I'm old enough to remember when we first saw Linux coming into telecoms network, and probably some of my panelists here will reflect on the same situation where everything was on [unintelligible: 09:00] and how to be hardened, and how to be very proprietary to be efficient. And everybody said, listen, this is how telecoms networks are built because you're never going to be able to find the level of performance or level of availability and scalability in an open-source operating system. I think those naysayer has been proven wrong, and I think that when I'm looking at Open vRan and I'm looking at cloud-native core networks, I see a lot of similarities where a lot of people not so long ago and still now, say that basically, you cannot run an efficient high-performance network with cloud-native telecom functions. Rakuten is proving them wrong, and all of us I think, are proving them wrong on a daily basis. And so, to the question that you had, we started on this journey with Rakuten quite some time ago with our traditional products in the radio field. NEC has been an early advocate and leader in the Open vRan field, and based on the success in this relationship Rakuten has asked us to work with them on the core network. We've been deploying core networks in Japan for the last 20 years but it was our first built and designed in a cloud-native way microservice architecture, based on container core network that was built from the ground up for a cloud-native network. And we built it together with Rakuten, and it's deployed in a commercial network. I mean, everyone looking at the Rakuten network knows that it's one of the most advanced networks in the world, but not only that, it is in one of the most competitive environment in the world in terms of being high density, high-performance markets. I think that we can all on this panel, look at what we've accomplished collectively so far and say, yeah, it is possible, but not only it's possible to have a cloud-native network, a cloud-native network actually behaves better, has better performance, and better cost structure than a traditional network when you're looking at 5G.
Host:
Thierry at Altiostar, along the same lines of your role in the value chain here, but really more so, just to add on; if you can set the stage a bit for the discussion today. Why is one of the biggest benefits of an open architecture, the flexibility to choose the optimum supplier for each element of the network?
Thierry Maupilé:
Thank you. So, before we talk about who can be the optimum or the best supplier for each network element, we have to acknowledge that unfortunately, we have been evolving from 2G to 3G, to 4G and now 5G, the industry was going in the wrong direction. And the wrong direction means that not enough investments, limited innovation, obviously, a number of suppliers has been shrinking, and probably an evolution where the benchmark was good enough networks. Where allowing companies coming from another part of the world to come in to compete, obviously, on price, [unintelligible: 12:47] was fair. Fair leveling playing field or not, I think we ended up in a situation where the operators don't have enough choice. So first, opening the architecture, allowing this innovation to take place in fact has had two very important effects. One is that you have companies who are very well known, very well established, who didn't participate. [unintelligible 13:13] in this space and you can take the companies who have built the network in Japan, Cisco, Red Hat, and a few others. We are participating more on the IT side of the network. So, now they are able to bring the cloud and the web-scale experience, as we see now even the public hyperscalers are able to participate. So, large companies with a lot of r&d power are able to participate in this part of the network domain. And also, companies like Altiostar, so emerging aspirants, companies who are here to bring innovation are allowed now to participate, because this is a much more fair environment, where you can rely on open interfaces. So, the rules are clear, there is no protection, no bias to entry. So, I think first of all most for the operators is choice and then, it's all about collaboration. Every network that we have been working on, it's a very, very close collaboration between the operator and the ecosystem partners. And it starts by co-designing, co-designing what is the right architecture because there are flexibilities. With architecture, the beauty of Open vRan when you virtualize and when you are disaggregating, separating the hardware and software, you can see some very elegant things in terms of how you can design the network based on what the use cases are going to require. But also, the network is becoming more programmable, it is software-defined and it's very attractive in terms of how you can leverage automation. So, when all those elements are defined, which are becoming the requirements of the operator, then obviously they have benefit to choose who they believe is the best supplier for this particular network element, which element will have been in a true world of open, fully developed, fully defined interfaces when there is also the confidence that those network elements can work between each other and it's quite remarkable. I just give you one data point, many companies here are participating, obviously, in the network in Japan, but you've just to take on the radio side, between 4G and 5G Rakuten has been able to select and to deploy six radio vendors. You prove that this Open vRan promise is real and can be implemented by the operator. And if I take the other projects we have, same. The operators are at the table selecting who they believe is going to be the best vendor based on their need, and they know, they have the confidence that if things are not going according to plan, they will be able to select someone else. So, it's going to be very dynamic and it does really give the confidence not just to the operator but again, for the companies who have to invest in technology, that they're going to have a real market where they can compete, not just on price but really on the merits of their innovation.
Patrick Lopez:
I can jump in on what Thierry just said, I think there's another aspect of it which is, in the past operators were looking for a one size fits all in terms of their radio solutions, they were looking for an optimum, basically, a vendor that could accommodate most of their geography, most of their deployment and in some cases because they didn't want to be locked in, they will select two vendors and give one geography to one, one geography to the other but that was the extent of competition at the end of the day. Right now, with Open vRan, because you have more vendors, but also because you have vendors that are specializing in specific aspects of solutions, you can actually start to design your network in a way that makes sense to optimize not for price, but optimize for performance, optimize for use case, and therefore being able to look at, for private network, for automotive, for industry [unintelligible:17:22] for augmented reality, for dense urban environment, for rural environment. You can start thinking about, I don't need to have the same vendor, and I don't need to have the same solution, and I can actually design a network that's fit for purpose. That's relatively new in the radio aspect.
Host:
And finally, Ian, I want to get to you on Red Hat's role in the cloud-native value chain, but more so or additionally, the benefits in containerization of network functions.
Ian Hood:
Thanks, Abe. The rest of the speakers kind of already talked about the ecosystem nature and the choices of the architecture. But really, when it comes to why containers, and really now it's not just containers, it's really about the application being built in microservices, in containers, on bare metal, to extract the best value out of the underlying platform. Now, [unintelligible: 18:15] been used for initially 4G and 5G, and also make every piece of the architecture be something that can be scaled up independently and securely, so these containers can attack each other on the fly, and do this in such a way that they can really deliver services much faster, and really combine that with a consistent way to do things. Partha talked about it earlier, in terms of a consistent way to actually onboard these applications and do this quickly, that's really quite important. And really, a lot of our customers really know, as you can see by their partnership here, that they can't do this alone. They want this freedom and flexibility and they want this advantage of best-in-class capabilities to deliver the economics they need and do it in such a way that they can expand their market share with that strong ecosystem we have.
Host:
So, Azita it's back to you. I know it took a while but let's get to it now. I know it's a very broad question I'm about to ask you and I know it's not going to be an easy answer, but do your best. So Rakuten Mobile has already disrupted the cellular market in Japan, clearly, how do you plan to do this in the US leveraging cloud-native deployments in Japan, but in the Americas?
Azita Arvani:
When we were doing this work in Japan and people got to see the demos at the World [unintelligible: 19:39] Congress 2019, we got a lot of operators coming to us and saying; we really like what you're doing, we believe in this cloud, we can see that these things are coming in, how can we learn from what you're doing? And so, we were very open with them, we invited them to come to the lab, kick the tires for themselves to see how this thing worked. But after a while, you realize that what they really needed is how did they take advantage of all the challenges that we've gone through, the problems we've solved the lessons learned? So, what we decided to do is create a blueprint of what we have done in Japan, and take that with our intellectual property, with our know-how, and lessons learned and wrapped up into this platform called Rakuten [unintelligible: 20:29] platform and offer that to other operators outside of Japan. And this question you asked about the US, [unintelligible 20:37-20:40] is to say, how do we then make it easier for people to take the Rakuten communications platform and build a network based on that, would be much faster, less costly, and they don't have to go through the same challenges that we had to go through back in summer of 2018. So, that I call the easy button if you want to do the cloud-native, Open vRan virtualized and take advantage of all the benefits that the panel and I talked about earlier. This isn't easy, but you can do that, it's an open platform, its multi-vendor, it has three layers, we start with the most advanced unified telco cloud layer and it's containerized, the microservices architecture, and it's sitting on top of the hardware that's optimized for telco cloud workloads. And then on top of that, we have the network function layer, this is where we get the best of the partners providing various network functions, so that's that. And then on top of that, you have the intelligent and secure operations layer that we can think of as the next generation OSS. This is where we really take pride in all the intellectual property that we've done to make it easier for us to [unintelligible: 20:03] a lot of automation, even zero-touch operations. The productivity of our people is like one service rather than engineering to everyone, 20,000 subscribers versus one technician for every 1000 subscribers, [unintelligible: 22:19] so you can see that 20x productivity improvement. We aspire to be a level four autonomous network in a couple of years. So, that that layer of how do we do end-to-end orchestration, how do we onboard various network functions, so they're not just stitched together as point products, but they work holistically as part of the entire system? So, all of that is in that layer. So, all three layers are needed but of course, depending on the operator, somebody wants to do end to end, someone wants to do just a virtual core solution or the virtual vRan solution, so it comes in either you can take as much or as little as you need for the network, and I think that really helps operators get to this cloud-native Open vRan, virtual vRan environment much faster.
Ian Hood:
Just wanted to add a couple of comments to what Azita said here because we definitely all learned and the experience we had with Rakuten over the other couple of years we've been working together. We've been able to take those blueprints across the entire architecture and apply that to our customers and other partners. There's a couple of key things that I think are important in terms of that zero-touch automation, and that's a consistent way in which you onboard things, scale them up and scale them down. And that's where the operator framework and Kubernetes have come to the light, to allow us to do that. And more recently, the openness project for actually consistently deploying the right hardware. Again, picking that right platform for your hardware to do these things. Those are a couple of other key areas that are important to ensure that we can replicate this capability, not just in the Americas but around the world.
Host:
I want to move to Eric at Airspan and a little bit from Altiostar, but then I want to get to Partha as well. So, Eric, I'll start with you. You proved already that a disaggregated deployment model is feasible at scale. How has that experience of working with Greenfield and disruptive players helped really push the envelope even further?
Eric Stonestrom:
We are very good at scalability. In 4G, we've deployed hundreds of 1000s of cells for tier-one MNO's and that experience gave us a great roadmap to replicate that success was Rakuten with 10s of 1000s of units deployed now and growing. What was really good with Rakuten and what we're bringing into RCP is that it's done in a way that can be really replicated where the 4G tier-one vendors that tended to be, each one had its own roadmap and it took years. What we've achieved in Japan, in months, took years in that less organized way. So, that's a really powerful thing that Open vRan brings to all of us and brings to the industry. The proof of our success with scalability as an example, in Open vRan and with Rakuten, is demand for new products. We've just introduced a millimeter-wave indoor product to help get into the buildings. And I should add Rakuten is just miles ahead of the US carriers and carriers in America in terms of what has been done with millimeter-wave. To that end, we're developing a series of products, we're also developing a lot of wideband 5G to take advantage of the C-Block and CBRS spectrum that hit the market. But it's really important that the people that build-out and realize on that spectrum are going to have to use the techniques that have been put in place now and scaled in Japan. So, I think it's really a Lodestone for the whole industry as we move forward. But the nice thing is, the timelines are just compressing, we're all getting good at what we do in our roles and a lot of heavy lifting, going in from a lot of companies to make it all happen a lot faster. That's very antithetical to the way the big OEM equipment providers like to work previously, they worked at their own pace, someone referenced earlier this geographic monopoly that they each got once they've managed to win a contract. Open vRan and Rakuten's innovation has just blown that out of the water for the better.
Host:
Partha, robin.io has done some really interesting things in the cloud-native space, as everyone on this panel is probably aware. Can you explain for the people out there watching this that may not have their head wrapped around this topic as best or as much as you guys do... Can you explain the transition from virtual machine-based vRan deployed for 4G to a cloud-native containerized vRan software capability?
Partha Seetala:
Yeah, certainly. It actually is a very important question because one school of thought basically says that, why not just use virtualization that has been tested in the 4G world? Why not just extend that to the 5g world? And some beyond just the obvious reasons that you want to leverage bare-metal performance and all of those things, there are some very interesting observations that we made and the benefits that we derived out of the architecture that Robin worked upon. Number one is that... Azita talked about RCP as a completely qualified, well-tested platform that you could go and deploy at other parts, not just in Rakuten, Japan, but even other places. And that architecture required that you have a Lego block kind of building model, where you could take out certain pieces and replace it with something else and so on. Now, what it means is that beyond just an application runtime environment, like a virtual machine or a container, you also have to think about a few other things. For example, these network functions, how do they scale? You want to have a standardization when it comes to scaling. How do you upgrade? You need to have a standardization when it comes to upgrading. In the virtualization world, people essentially went about doing their own way of doing Lifecycle Management, [unintelligible: 28:00] network function vendor would come up with their own automation, their own way of doing scaling, healing. and things like that. right. With the Robin platform, what we have been able to accomplish is regardless of what network function you bring in, whether it is vRan, or it is core, or it is any of the edge applications, each one has a different way in which you need to onboard it. Each one has a way in which you need to scale it but we brought a standardization which goes beyond just Kubernetes, which allows any network function vendor to easily go and get onboarded onto the RCP platform, and then easily be managed by the OSS layer that Azita was talking about. I think these are very important innovations that are required if you truly want to make this a multi-vendor, ecosystem-driven platform. And that's where I think moving away from virtualization to containers, that's one big benefit. So, that's one point. The other point that I want to bring up is, and this is a very important one especially when it comes to vRan is that, when it comes to cloud-native architectures, one of the design principles of cloud-native is to completely abstract away the underlying hardware. Essentially the application vendors should not know nor care about the underlying hardware. And that's actually great because that essentially allows the operator to go and replace underlying hardware and all that. However, we all know that when it comes to run, it's easier said than done, because vRan is very sensitive to the underlying hardware, it's very sensitive to the BIOS version. They're sensitive to the underlying FPGA that are available and so on [unintelligible: 29:37] and all of those things. So how do you reconcile these two things? On one hand, you have to abstract away the hardware. On the other hand, you have to provide all these capabilities and expose them to the vRan application. That's again another place where Robin has innovated, and what we have provided is a way in which we could again provide a policy-based way in which, first of all, the Robin platform discovers all these things. hardware capabilities. and it provides a policy language that can be now leveraged by the central function vendors without really now having to worry about the underlying hardware. But these are some of these things that are possible because of the way the container technology has evolved and how Robin has [distortion: 30:17] to bring these innovations.
Host:
Patrick, I'm going to go to you. Sorry, Partha, I think you were cutting out just a bit towards the end there.
Thierry Maupilé:
If you don't mind, just one data point for us validate what you said about your question. In the Rakuten network, obviously, we started with 4G, and already It was a tremendous improvement, leveraging virtualization [unintelligible: 30:44] the partners who are on this panel. We're able to activate a cell site in just slightly more than eight minutes versus what took us to take days or weeks. But what is interesting is that when we move from VM to now, this CM, with the containerization platform, we could buy just half this time from eight minutes plus to less than four minutes for 5G. This just shows the velocity of the innovation, and how this next-generation evolution of the architecture is going to yield other benefits. So, I just wanted to share this data point.
Partha Seetala:
[crosstalk: 31:25] performance theory, if I'm may add [unintelligible: 30:28] even the performance improvement from the 4g to the 5g network, there has been a pretty sizable improvement in the performance going from virtualization to containers.
Host:
Great. Patrick, I wanted to go to you and it's a pretty good segue in terms of throughput. And again, for people that are not as familiar with the Rakuten network as we may be, in terms of throughput, explain how the density and capacity of Rakuten mobile's network means. It is performing actually better than the conventional 4G network. Again, in terms of throughput.
Patrick Lopez:
I think Azita will probably have a lot more specific details and I don't want to give specific data point that she may not be comfortable disclosing. But at a high level, I think the architectural framework allows a better distribution and use of the spectrum. So, we spoke at length about moving from virtualized to containerized. What that allows is a number of things, obviously, you separate hardware from software but also in the cloud-native architecture, you separate statefulness from statelessness, and you also separate in 5G, the control plane from you the user plane. The combination of all these aspects allows you to have a lot more granular connectivity configuration. And when it translates into Open vRan and it translates into the radio, you have advanced techniques now such as Massive MIMO and digital beamforming that allow basically to radiate the signal in a much more efficient manner than in 4G or in non-Massive MIMO configuration, which allows you provide a lot more capacity to a lot more people without having dissipation of energy and without having interference. And obviously, in order to do that well, in a multi-vendor environment, I think that's the ultimate proof point, that Open vRan works. If we are able to deploy in the middle of Tokyo, a massive MIMO Radio with sudo vendor that's different on third-party hardware, with an orchestration system and a cloud system that comes from another vendor, and all of that delivers basically superior performance compared to 4G and compared to other 5G networks, I think that it shows that not only it works, but it performs the way we would like. Azita do you want to share more details on the performance?
Azita Arvani:
Yeah, thanks Patrick, you had a great start. I would just also add that talking about some of the things that Patrick was talking about in terms of Massive MIMO, and digital beamforming, and so forth. Those are some of the things that come with 5G in addition with network slicing but we can do it much better in an environment where it's cloud-native and much more flexible. So, you could potentially do it with the legacy stuff as well, but it's just not as flexible and it's going to be much more costly, you're more likely to over-provision for things. So, that's on one end that you could do 5G advanced capabilities much more flexibly, at much better price. But in addition to that, we can bring in, if we have urban areas, we can have a different RU vendor for that, we have suburban areas, we have indoors, outdoors, all of that combination, you don't have to think about buying this stove pipe vRan tower for each of those use cases. We 3can just find the ideal radio unit from an innovative radio unit vendor and you can mix and match and they will all terminate in the same software. So, that is very, very powerful.
Host:
Azita, quick question. And again, not an easy question and you could probably talk about this for hours, but give us a realistic view. What have you seen in Japan at this point in terms of leveraging Oran for 4G and 5G?
Azita Arvani:
So, it's actually a great continuation of the thing we were talking about. So, we have almost 10 different radio unit types that we're using right now, 4G and 5G. And if you think about our disaggregation of the vRan, we're doing it at three layers. One is, you go from this [unintelligible: 36:13] to Open vRan which the members of this panel are all believers in that. The next one is to go from purpose-built hardware and software to virtualized. You could do all Oran with your own legacy purpose-built stuff, but it'd be much more expensive. So, if you can run these things on cots hardware and do it in a much cost-efficient way, that really reduces the cost. And then the other thing is with the radio units itself, the radio units came in as this sort of black box that you couldn't touch and now, we are redesigning and making it as easy to install as an enterprise Wi-Fi radio unit. So, for example, the work that Eric was talking about with the millimeter-wave, so that's something that we partnered together to build and create. So, The 4g network is where we started with the virtualization and took advantage of Open vRan and virtualized vRan, we've moved over to the 5G environment that's containerized based on Kubernetes and it's giving us much more efficiencies, and cost efficiencies and flexibility there. And whatever else it comes in the future; we're going to make sure you know that we'll be on top of it. The point is, since the architecture is divided into much more granular levels, we can bring in the best of breed much easier from a broader supply chain.
Host:
Another good segue and Eric, I was going to go to you next anyway. So, how are you working towards more live 5G sites from Rakuten mobile's Greenfield launch that started last year, and how does that really translate to the US legacy Brownfield market now?
Eric Stonestrom:
So, first, and as Azita said, we got super innovation going on in the actual design and now that we put the foundation in place with the others on this call, it's quick to add new radios. We have seven different projects going on, including five different radios and the get started time on each is getting shorter and shorter, which is great. So, we can bring specific innovation on a specific use case, in the most cost-effective way and we're not reinventing the wheel every time it's very quick, and that's getting quicker and quicker in Japan. In the US, obviously, the US carriers are already moving towards 5G, it's been rather lackluster so far. If you look at the Brownfield carriers, they advertise a slightly higher speed. My analogy is, it's a five-lane highway that has been moved up to six or seven lanes with 5G. So, the promise of 5g hasn't really been realized at all, and that's where standalone will simplify and allow Open vRan adoption and we're seeing a lot of green shoots in the private 5G space. We're seeing green shoots even with the bigger carriers saying, I'd have to do things differently. Otherwise, it's really bringing the wrong weapons into 5G. So, it's happening and Open vRan has as I said earlier, made it a process that everybody has to pay attention to, it's no longer something a couple of really super innovative brilliant groups in Japan are doing, it's now something that is really necessary to make 5G realize its potential in the US. So, we're confident with partnerships like RCP will all bring a lot of value to the table.
Host:
I wanted to get your perspective on the adoption of Oran globally from Ian, from Thierry, and from Patrick Lopez, but before I do that, I want to go to Partha. So, as robin.io is the first 5G containerized platform to go live in Japan back in 2020, for hosting ran and container form and in production today as well; what does this mean for operators globally?
Partha Seetala:
It's interesting, in 2019 I was at a GSMA hosted event and there was a lot of skepticism from the audience, as well as some of the panelists, around Rakuten's vision of doing fully containerized 5G. In fact, [unintelligible: 41:31] was one of the panelists in that event and there's a lot of skepticism, people were saying, well, it's never going to happen, it's never going to be launched in time, it's not even possible, you will not get the performance and all of those things. Now, I think collectively, we have all put that behind us, it's now known that it is possible, and it's not just possible the benefits are so huge that others cannot ignore it anymore. And what we are seeing now with this going live at Rakuten in Japan is a lot of operators are now looking at; how do I replicate this? How do I get the same benefits that Rakuten has seen and are there any blueprints that we can essentially adopt? And of course, RCP is one trade initiative and creating that blueprint that can be now replicated across other places and other operators. So, what we are seeing is there's a big shift in mindset, a big shift, and believability, if you will from the other operators. We're getting a lot of inbound interest from other operators are basically asking about, how do we do this, how do we replicate this success that Rakuten has seen? So, I think that's been an amazing thing to see.
Thierry Maupilé:
I think if I may add, we need to accelerate and to give enough kudos to Rakuten leadership because it was not suddenly an easy decision. If it is you had the motivation, the vision, and the motivation, they wanted to do something different, they didn't want just to build a mobile network for building a mobile network, it's a cool topic. It was because they are an e-commerce company and they believe that as part of their business strategy, connectivity to their consumers, to their end-user, it was very, very important. and that they will not get the benefits of providing this level of experience by just what do we are doing, continue to do some extent until the network is completely built out, which is to rely on an existing mobile network operator to provide the service. So, not only is it, they took the decision, they deployed those two networks, which is one network because this is the same software fabric, for example, that we have deployed in for them. But they have deployed a 4G network and a 5G network in record time, there is no network in the world at this scale that has been deployed so quickly. And again, this is been possible because of the fact that it's a software cloud architecture. So, obviously, it has given a lot of confidence to the rest of the ecosystem, to the other operators, that it is real, it is working. So, a strong and huge validation. So, again, for us, it was extremely important. And the beauty of it, if I look at all the panelists, my peers here, we're all working on multiple projects outside of Japan, so the collaboration model that was also put in place for Rakuten in Japan is itself replicable, and we are able to take a lot of knowledge, expertise, that obviously lessons learned from what we have done in Japan, and now to the benefits of the other operators. So, that has been no doubt, a necessary step. Eric mentioned about the first-mover advantage, but we all know that no one wants to be the first one to deploy such innovative, disruptive technology. So, this is behind us as Partha said, and I think now we have an ecosystem that is really able to deploy at scale, with a very diverse, already diverse, and growing ecosystem. So, all the fundamentals, all parts of the foundation, why we are building a new industry, including the incumbents, including, as you know, I say that [unintelligible: 44:33] was broken when even the skeptics decided that now, Open vRan is probably the best thing that will happen for the company. So, some of them are already announced that they will have opened one solution, so I think it's very encouraging that we are able to now help other operators leveraging what we have done in Japan so they can also start their own transformation and open a platform for having access to innovation.
Host:
And just to close, Ian, Patrick and I'll let Azita finish. How do you see the pace of the adoption of Oran and traditional operators both in the US and globally?
Patrick Lopez:
I think we see announcements mostly every month, if not faster than that. Over the last few years, it was trialing and now, we see commercial deployments. The first commercial deployments were in rural environments, and they're moving more and more into urban environments as well, there is an interesting trend. I think everyone acknowledges that 5G business, 5G growth is going to come not necessarily from consumer traffic, but from enterprise traffic, and not only from the traffic but basically, the management of connectivity for enterprises. I think that what Open vRan is actually bringing to the fore is the ability for enterprises, for the first time maybe to be able to create a connectivity experience for their employees, for their systems, for their things, that they can manage themselves. And in the past, in order to do that, it was well, it was building a network, the same as a network operators, it was very expensive, time-consuming and it required very specialized skill set. What Open vRan doing, but generally speaking, what the cloudification of [unintelligible: 46:40] networks is doing is democratizing the ability to deliver connectivity experiences for a variety of organizations.
Host:
Ian, anything to add?
Ian Hood:
I just wanted to kind of cover a couple of topics, we have definitely made a lot of progress in this Open vRan picture, there's still work to be done, we're making lots of progress in opening up these interfaces. There are still some situations that we're finding as people try to innovate in certain places where they'll claim to be open, but they will actually go off and change things on the fly and not actually share that with the rest of the industry. You have to kind of stay on top of that as we drive this larger ecosystem. Because it's okay to do that in the lab but eventually, you got to share with the rest of the industry as we all can move forward. So that's one point I wanted to make in terms of the progress we're making and on top of that, there's still a bit of a challenge in open hardware underneath to actually meet the scale and performance we'd love to have for Massive MIMO butt we're making all that work. The comments that Partha made about being able to drive to the BIOS, and the hardware and the accelerators, that's really key. But within all of this, Thierry mentioned it as well, we as people on the panel here, working with each other and with other partners are involved in driving this forward in the vRan across all of the regions, things in Europe, things in North American, Latin America, this is all moving forward. But I think that's allowing us to move to the next phase of this picture, we've gone from the core to the vRan and now, really what we're going to do is go and drive for what we call the private 5G edge for the operational transformation and open up that network. That's kind of our next iteration because if we can actually do that, how can we actually pay for building these big highways, if we can actually get the enterprise to run their workloads, either in private or on provider edges in this architecture? So, that's kind of the next phase of the world we're going to see.
Host:
Azita can I get one quick comment from you about the adoption of Oran globally, of course, but also, tangential to the session topic today, in the US as well?
Azita Arvani:
Yeah, you're seeing a lot of interest. As I mentioned, the reason why we started this Rakuten communications platform is because we were already getting a lot of inquiries about how to do this, so those opportunities and growing set of those opportunities are becoming opportunities or Rakuten communications platform, or RCP as we abbreviated to. So, I see a lot of that now. I also think that one thing to realize is that right now, we have proven that this model works but this is not the peak, this is just the beginning of it and things will iterate as my colleagues here mentioned, things will get better and better from here versus some legacy system. When you buy it, you open the box, even though they don't want any boxes, but that's where the peak is and it kind of goes down from there. Here, you have something that's good out of the box, and it's less expensive, more flexible, and more secure. But then it gets better because as partners iterate and technology gets better, it just gets better and better so I expect to see more opportunities coming for RCP, and we'll make more announcements very shortly.
Host:
That's good to hear Azita. So, we did our best to get sort of a Rakuten mobile ecosystem if you will, or at least a cross-section of it on one session at one time. So, we hope we did a rather good job there. And hopefully, we can do this again sometime, hopefully in person and not virtual. Although both seem to work okay in their own respects. So, to all of our viewers out there again, we'd like to thank all of our speakers on leveraging cloud-native in the Americas for this session on-demand, April 26. Please go to thenetworkmediagroup.com. So long