The Network Media Group

View Original

On-Demand, April 26th: The Performance of Open vRAN Architectures - CET 2021

As Open vRAN has been challenged in recent times, this framework has time and again proven effectual for performance, cost savings, and optimized system integration. Open vRAN performs very well in high-density environments, widely proven with the Rakuten Mobile deployment, in Tokyo and other large cities in Japan. The cost savings of Open vRAN have also been shown to be effective by Rakuten Mobile, and system integration challenges are being resolved. Watch this executive session with leaders in the Open vRAN space on debunking the misconceptions around this architecture.

Executive Speakers:

  • Thierry Maupilé - EVP Strategy & Product Management, Altiostar Networks

  • Caroline Chan - VP and GM, Network Business Incubator Division, Intel Corporation

  • Juan Carlos Garcia Lopez - SVP Technology and Ecosystems, Telefónica

  • Honore LaBourdette - VP, Industry Verticals & Edge Telco, Media & Entertainment, Red Hat

  • Patrick Lopez - Global VP Product Management - 5G, NEC

  • Ihab Tarazi - SVP & CTO, Integrated Solutions & Ecosystems, Dell Technologies

Register To View

See this form in the original post

By Submitting This Form I Agree To Altiostar and Dell Technologies Privacy statement And Terms of Use.


Transcription:

Host:

As Open vRan has been challenged in recent times, this framework has time and again proven effectual for performance cost savings and optimized system integration. Open vRan performs very well in high-density environments, widely proven with rapid mobile deployment in Tokyo and other large cities in Japan. The cost savings of Open vRan have also been shown to be effective by Rakuten Mobile and System Integration. Challenges are now being resolved. On this executive session, we'll talk about Oran solution integration and debunking some of the misconceptions around this architecture. Joining us are, Thierry Maupilé, Executive Vice President of strategy and product management at Altiostar. Also, we have Caroline Chan, she's vice president and general manager of the Network Business Incubator Division at Intel. We also have Juan Carlos Garcia Lopez. He's senior vice president of Technology and Ecosystems at Telefonica. Additionally, we have Patrick Lopez, he's Global Vice President Product Management-5g at NEC. And we also have Honore LaBourdette, I hope I got that name right Vice President, Verticals & Edge Telco, Media & Entertainment, that at Red Hat. And lastly, I have Ehab Tarazi, he is Senior Vice President and Chief Technology Officer of Integrated Solutions Ecosystems at Dell Technologies. And gang, welcome to the program.

Caroline Chan:

Happy to be here. Thanks for having us.

Honore LaBourdette:

Hello, everyone. Thank you.

Host:

Absolutely. Thanks for being here. We know that everyone's under a bit of extenuating circumstances trying to do this virtually, so I think we're all learning here. So, we have the whole gang here, a really good cross-section of the ecosystem. Thierry, if you don't mind, I'm going to start with you on question number one for this session. At a very high level, explain the evolution and the growth of Open vRan, and what is the adoption rate of Open vRan right now, and has it really accelerated from last year now into 2021?

Thierry Maupilé:

Hello and thank you for putting this event together. I think it's a great opportunity to provide an update on the progress. Definitely, as an industry are making tremendous progress. I think maybe the best way to characterize what Open vRan about is, really it's a movement. It's a movement, which definitely is not stoppable, it's remarkable that being, as we all know, we are all challenged with a COVID situation that obviously continued much longer than we would like. But despite this challenging environment, we are making tremendous progress. As you mentioned, the fact that Rakuten has now been deployed at scale for 4g and 5g is a tremendous validation. Obviously, it is important for Rakuten itself, and for the Japanese market, and for the Japanese consumer because they are already seeing the benefits of the entrance of a new operator. But I think it was very clear that there were a lot of questions that need to be answered in terms of is it real? Is it working? What is the performance? So, all those questions started to be answered very positively and it gives definitely a confidence to the other operators. [unintelligible: 03:08] the broader ecosystem to invest in Open vRan because the benefits are such that it is definitely going to be the best architecture, the best platform for 5g. So, it is already applicable to 4g, as we see that most operators want to take advantage of this new architecture and the fact that you can choose different elements of the network now, it is a very diverse ecosystem. So, what we see is that besides what we expected, the Greenfield new entrance will be the early adopters at scale. So, Rakuten, Dish, United States and we can expect one or two more large-scale range deployments. But what is more important for us is in the last, I would say six to nine months, the operators who have existing networks are really not just considering, they have decided that they will start to insert Open vRan into a network. And besides Rakuten, the development of the ecosystem is such across all elements of the value chain, from Silicon to software, to the cloud, to radios, to the system integration. The ecosystem is at a point where definitely it gives the opportunity for the operators really to plan the deployment. And we see certainly the first large-scale deployment in legacy networks in the second half of 2021.

Host:

Caroline.

Caroline Chan:

I think that besides what Terry said really well, the momentum both in Greenfield and Brownfield, we also here at Intel see another trend as really going beyond just Telco at the consumer side. We also see a lot of announcements as well as POC, even early pilots, that taking this, the Open vRan architecture, the fact that it has flexibility, adaptability, and also a very broad ecosystem allowing people to develop the application. we see that really started coming into the enterprise networks as well. I think that is a momentum that we should not ignore and that is a distinct advantage of having this open ecosystem, open interface at driving towards cloud-native, that allows a broader participation in this ecosystem, from application developers to the telcos, to the enterprise CIOs and to the cloud service providers. That for us, it's a very exciting expansion of our overall market space.

Host:

Juan Carlos.

Juan Carlos Garcia Lopez:

We have seen in as Thierry mentioned this year, a lot of great progress on Open vRan, with many announcements on multiple sites. The Open vRan Policy Coalition is to create awareness on this topic, on public administrations and the industry in general, the announcement of operators in Europe, the G-5 in supporting the development of Open vRan. Also, the [unintelligible: 06:59] with the specifications, a deep reinforcing the Open vRan working group and using Open vRan components has also improved. And then, as the result of the trials and also, early initial deployments are still limited in some Brownfields and [unintelligible: 17:16] that are quite promising. That gives us an indication of what 2022 and 23 will be, where you will see more in scaled deployment.

Host:

Patrick.

Patrick Lopez:

I can only echo the sentiments of my fellow panelists here. For NEC, 2019, 2020 was Open vRan positioning as really a disruptive technology, I think we'll see in 2021, 2022 Open vRan becoming a mainstream technology. We're proud with some of our colleagues here to have deployed some of the most advanced networks in Japan, that have demonstrated that Open vRan can be deployed in high-density urban environments with high capacity, with performance on high Massive MIMO Antenna that provides some of the world's leading performance. And we're seeing as well that at the same time because of the flexibility of the technology, because of the disaggregation and the openness of the technology. As Caroline mentioned, you can actually have the same technology being applied to different use cases and different topologies such as enterprise environment, where for the first time really, you can decide that for vRan one size does not fit all, you can actually adapt the footprint, you can adapt the vendors, you can adapt the technology to your use case and really optimize both the operational model and the cost structure associated with that. And obviously, the last part of it from our perspective, we're especially proud that [unintelligible: 09:16] obviously, disaggregation means more vendors, even if the interfaces are open and it requires strong Systems Integration capabilities. And we've been able to demonstrate also that there are now blueprints and there are now capabilities to deploy Open vRan at scale in a true multi-vendor environment, really at every stage of the value chain, every element, every domain of Open vRan. And we've deployed those technologies with a variety of partners, so I think that's a strong indication of maturity.

Host:

Perfect, Honore as well.

Honore LaBourdette:

I think there's a consensus sits across the panel that Oran great achievements for the industry in the past 12 to 18 months. They've helped define smaller groups of industry partners like all of us on the panel who can offer customers leading technologies, and this focus work is helping to accelerate adoption and speed up real-world deployments. Today, Red Hat is engaged with the majority of operators around the world. We're in various stages of exploring, testing and deploying, cloud-native vRan based on Open vRan architecture, and I think this has been iterated across the panel in the context of the various stages. Caroline also mentioned the Greenfield versus Brownfield. We're also seeing how Greenfield operators are taking Open vRan head-on and achieving great progress and in many instances, good success. While Brownfield operators are taking a kind of a multi-pronged approach. Some of them are hitting it head-on, while others are experimenting with Open vRan on a smaller scale, in different sandbox environments like small cell. So, Bharti Airtel, for example, is one operator that's looking at small cell. These Brownfield operators are looking at different deployment scenarios and they're experimenting; they're learning and then they're rolling out to a more mainstream strategy. If you look back, it's only been 12 months ago, where there were little to no conversations about Open vRan. And now, it's clearly a must-do in the industry for the telco operators. And we think that open source has played a key role in accelerating the adoption that we're seeing today.

Host:

Ihab, I'm going to transition to you over at Dell Technologies and again, sort of setting the stage for the conversation. Why are solutions integration so important and open and disaggregated networks? And could multiple integrations involved with 5G deployments really slow the speed of innovation?

Ihab Tarazi:

Very good question. So, I'm going to build on all the comments that the speakers so far have talked about. The benefits of Open vRan very significant, there are tremendous advantages and both efficiency, power, speed, innovation, flexibility. At the same time, not every operator has all the capabilities to do that in house, they don't have maybe cloud skills, they don't have enough software developers, nor should they be because we have so many new operators also entering this world, where they are taking advantage of new spectrum, new capabilities. A lot of these new operators have no plans to hire a vast technical team, a lot of them don't want to build a lab. So, the model is evolving to take advantage of all the capabilities that we all talked about. I really liked Thierry’s comments that this is a movement, I'm going to use it from now on Thierry if it's okay. But long story short is that, if we don't provide solutions, it becomes very hard for the users, for the operators to take full advantage of those capabilities. Solutions mean, we give them outcome-based, so they don't have to worry about, is the software compatible with that software? What exactly is the configuration? What CPU chipset I need on my server? How do I set up my BIOS? All these things are taking care of a solution because that's pre-determined, validated, and continuously updated. So, as new versions and new ecosystem partners come up, it can be updated. I don't think it's slowed down to innovation; I really think it speeds it up significantly because that matches the model that operators are used to today. Which is, they have a vertically integrated model with outcome-based and they don't have to build all the capabilities in-house to be able to deploy it. So, I think it speeds up innovation because it creates ways for new players to come in, integrate with the ecosystem, have somebody like an OEM be able to bring in the scale and capabilities out there, and provide the support that's needed for the operators and give them all the different solutions?

Host:

Thank you. Thierry, I want to go back to you. Energy consumption is a key issue that's been raised when it comes to Open vRan architectures. How is this being addressed by the Oran or Open vRan ecosystem from your perspective?

Thierry Maupilé:

Well, firstly I will characterize this topic about energy consumption or energy efficiency as the issue of the week or the issue of the month. There's been a lot of skepticism about Open vRan by companies who are not ready to have their Open vRan solution or believe that they need to protect their current business model. So, what I can tell you that everyone knows in a radio network that the energy consumption, which at least 40% or more is coming from the cell site. So, inherently, if you have all the equipment, which is the traditional model, this monolithic architecture where you have to put everything at the site, where you have not just a radio but the baseband unit, obviously is very power-hungry. So, having an architecture where you're moving some very important elements to the cloud, [unintelligible: 15:44 its edge and centralized cloud is definitely bringing a lot of efficiencies. There's also, as you pull resources in a data center, you're able to achieve some multiplexing efficiency. So, there are many benefits that indirectly also come like Ihab was alluding to, when you adopt things which have been proven in the IT world, in the data center, in the cloud, where the tremendous achievement evolution of performance, I think it's at least in the last 10 years fivefold. The compute power that you can have at a data center is five times what it was 10 years ago, where the power consumption only increased by 6%. So, all those benefits are inherently going to be beneficial to the vRan, to the new Open vRan architecture. So, is it necessary to demonstrate what Intel has been demonstrating for the last 30 or 40 years, the benefit of Moore's law? So, we believe that this is going to continue and will also impact positively the overall cost. And we need to look at the overall cost of the network, not just pointing out to a single element, which maybe could have a different profile in terms of energy consumption. But overall, everything that we have seen in real deployment shows that you can achieve significant savings, including on the power. So, we strongly believe that this is the right architecture and the evolution will be continued to even generate more gains.

Patrick Lopez:

I'll add on to that if we can. So, sorry, just to jump in. I think an important point is that Open vRan is not only open but the large majority, if not all vendors Open vRan are virtualized. And as we move into a more cloud-native architecture, in a cloud, you centralize what you can and you decentralize what you must. And that allows for economies of scales, including economy of power consumption because all of a sudden, when you used to have one baseband for one radio unit, whether it was used or not, now you can [unintelligible: 18:04] function, and you can have separate distributed units serving different cell sites and separate centralized units serving different distributed unit. And what that means also is that when there are sites that sustain less traffic than others, you can actually start to granularly ramp down the power emission per site, which reduces substantially the amount of power that you can, in the end, use to serve the same user experience. So, Open vRan and virtualized vRan, allow for a much more granular energy consumption framework.

Caroline Chan:

In addition to what Patrick and Thierry said, I just want to add on a point that from Intel's perspective, we are working with ecosystems to really enhance and improve energy consumption. Like what Thierry said, what's happening in the cloud has really proven that as generation [unintelligible: 19:08] CPU improvement far outpace the power consumption, that's number one. Number two is, there's a lot of telemetries that the cloud has been leveraging to run down different power states within the platform as the workload decreases. As the network AI really improves, Open vRan has a lot of network AI architecture, such as RIC, ASAP, it does allow us at the silicon level to predict and forecast what workloads going to be, we'll be able to actively manage the power station. That's one secret that a lot of people probably do not realize unless you're intimately involved in the cloud. What we do in the cloud and data center, enterprise cloud will absolutely be leveraged, and to move into the Open vRan world and to really enhance the power consumption.

Honore LaBourdette:

I think it's also important to note that software can extend efficiency when it comes to energy and I think Patrick was speaking to this point as well when he talked about virtualization. Container technology, especially on bare [unintelligible: 20:19] tends to be lighter by nature and software enables the ability to activate and deactivate capacity based on traffic loads, resulting in better energy utilization. So, with the combination of all these things, I think we're actually going to see improved energy utilization and conservation as a result of the implementation and deployments of Open vRan.

Caroline Chan:

Absolutely, couldn't agree more Honore.

Host:

Ihab, I want to go back to you, as we move out of the proof of concepts and into commercial deployments of Oran around the world, Will the radio ecosystem, in your opinion be able to really keep up with that demand, particularly when it comes to advanced technologies like Massive MIMO?

Ihab Tarazi:

Yeah, absolutely, I think the best way to think about it, we can't think of this as a replacement of the old vertically integrated model. This is a whole new world; this is a cloud model. Similar to the power discussion we just had that this is consumption-based and can be very fine-tuned and optimized, the same with managing spectrum capacity and be able to debut different radio technology. So, there are multiple factors here. First, the performance of x86 and the new capabilities with instruction sets for data pipelines, from Intel has been doubling almost every year in our experience. So, we see tremendous growth, this is Moore's Law at work. Now, it stopped working in some other areas, maybe, but it's definitely going to work and continue to work for the next many, many years. And then you’re adding acceleration on demand, which was already developed for AI and for a lot of other capabilities, and you have a whole new fleet of capabilities with smart NICs that can be plugged into those servers. So, we have tools we've never had before and a combination, everything is multiplying, doubling every year and performance. Then you add to that the fact that you can distribute the containers on bare metal, as Honore was saying, you really can shift the workload. So, before we get stuck to say you have one tower, figured out how to get all this bandwidth and do it. Now, you have a whole cloud design that works and you can optimize for it. So, from my experience, and we're working with many customers, not only can we keep up, I think it's starting to exceed in our view performance in some specific deployments. When we move to enterprise deployments, which are at early stages, this is going to be a significant advantage for enterprises to be able to go there.

Host:

Honore, I'm going to go to you next and I'd like to have one Carlos respond if that's okay. So, how should a service provider really start planning for cloud-native microservices-based on a virtual and Open vRan infrastructure?

Honore LaBourdette:

So, I think it's widely recognized in the industry that Open vRan is about building applications faster and more efficiently, but to the comment about it's a movement, it's also a mindset, and it's a set of disciplines and principles and practices. When any organization starts out planning for cloud-native applications, it's important to consider the entire lifecycle, from selecting the development platform, all the way until an application is truly production-grade and ready for delivery in the cloud. I think this planning needs to consider three primary things. First, a platform that's based on open source, and ensuring the platform has the right capabilities to address ram requirements is critical. You also want the most secure platform that's available, and one that offers low latency processing, explores or ensures timing and synchronization, provides hardware acceleration, automation, requires a small footprint. And as we've spoken about before, runs VMs on bare metal. And lastly, one that's built to take advantage of hybrid cloud environments. The second element or key element to consider during the planning process is selecting the Altiostar workloads. Just running containers is not enough. You need a true microservices architecture to achieve the benefits of a cloud-native approach. You must select and support the right workloads that have been architected in a truly cloud fashion or in a truly cloud-native fashion. Like the work, we're doing with Altiostar and Intel building on open shift. By standard standardizing on an open platform like open shift, you can reduce development learning, maximize agility and enable applications to scale up without requiring significant changes or modifications. And then third, this is something that Ihab spoke about it, please, I think the third most important part of the planning is to consider the integrator or the OEM. If the operator doesn't have the capabilities, the capacity, the skills, or the resources to take this on themselves, selecting the right SI partner, or the right vendors that can come together to bring it all together becomes a very important last step.

Host:

And Juan Carlos, anything to?

Juan Carlos Garcia Lopez:

Well, I agree with Honore, with most of what Honore said there. So, mindset is extremely relevant, so the new built on an open architecture is totally different to the traditional one in terms of how you deploy it, how you integrate it, how you test it, and how you manage it. And this means that you do need to build new skills, so do you need to plan for this in skill training. You need also to change your processes in adding all the IT practices of CICD, CVCT- Continuous Verification, Continuous Testing, all the DevOps nematology. So, this is extremely relevant. And the operators have not always [unintelligible: 26:40] the network departments ready to that, so you need this training on these tools. There is another important factor that Honore mentioned, that is the integration partners. So, you have to count on partnership to support you on this task, at least at the beginning. And finally, I think there are two aspects that are relevant. So first, is the infrastructure itself. So, you need to create this cloud environment in positions that are close to the [unintelligible: 27:10] stations on the far right, and this means that you will have to prepare the aeration nodes with the right environment, and this is something that takes some time and you have to plan it in advance. So, I'm converting these [unintelligible: 27:16] nodes into small data centers to host this cloud environment is something that you need to plan ahead and start working on soon. And finally, thinking about day two, so the moment in which you have the open architecture, the Open vRan architecture in place, you need to think about planning also for the orchestration and automation in environment that you will need to manage all that complexity that you have built. So, think that the traditional vRan has a beauty, that you have black boxes that come with this solution for management, and now you have an architecture based on components that are distributed, that also is dynamically changing and scaling up and down, evolving and moving with the traffic analysis and complexity of [unintelligible: 28:13] in a particular part of systems. It was mentioned already in the chat, the artificial intelligence you will need to learn how the traffic patterns change and how you have to scale up and down the capacity in the vRan, and all that is these things that you need to plan ahead for having a successful [unintelligible 28:37]

Host:

Thierry, I want to go to you and now and Juan Carlos touched on this just a bit. Maybe in a little more granular way can you explain the levels of integration required when integrating Oran networks and really, what are the options for Oran integration.

Thierry Maupilé:

So, from our point of view and from what we hear with our interaction with operators, the definition of system integration is going to change and is evolving. There's no doubt that some operators, why some of them want to do the system integration because now they have the opportunity even more to have control, control of their network. They want to be able to design the network the way they would like, they want to be able to select the best of breed, elements from multiple vendors. So, it's extremely important also, even from a security point of view. So, when they want to do a zero-trust, end to end fully secure network, they need to have control of the architecture, over the network elements. So, it's a bit counter-intuitive, but by opening up, having an open architecture, in fact, is the best way to have a secure network. But that's also a discussion on its own. But we also believe that when I say system integration is evolving, that a lot of things can be done, you know, up front or before you deploy. The fact that you have open interfaces that are fully defined, allows the ecosystem to, by themself make sure that those elements can be really interoperable. So, this is going to be before you consider deploying in any network, that those elements are proven that they can work together. So, that's a benefit, obviously, of open interfaces. The other aspect, which is going to be to have a tremendous impact is the automation. Because as you're able to automate a lot of network functions, you are minimizing the intervention of physical resources, of manpower, reducing the number of errors, and all the benefits of automation, where you can do self-diagnosing, and self-healing. Even just before we talk about fixing issues, if you see what we have achieved in Japan, where now for 4G to get the site up and running, it's eight minutes. And just from 4G to 5G with further automation and application of AI techniques, it is now less than four minutes to get to one site up for 4G. So, that's really a tangible example of how all those things that we can do as an industry, including this certification of Oran compliant elements is going to minimize, or it is going to change the heavy lifting system integration that you had to do because this is why you have to go to the vendor, this mono vendor. Because they were controlling all the elements, so they were the only ones who could do the integration. So, I think things are really evolving in a positive way, where we are giving the [unintelligible: 31:51]to the operators to have more visibility, flexibility, and control of the network.

Patrick Lopez:

I'd like to add a little something because we've taken the systems integration role in a number of deployments for Open vRan, we have a little bit of experience in the domain. I think, as was mentioned, the role is changing in the sense that systems integration in an Open vRan world has become a lot more about configuration management. Because at the end of the day, these interfaces are open, that's the concept of Open vRan. So, a lot of people might be thinking, well, why do you need integration at all, it should be plug and play. And to some degree, obviously, it is. But what is necessary is an intimate knowledge of the radio and DSI associated with the connectivity in order to optimize the network deployment. And specifically, here, there are a couple of examples, but in order to take the best advantage of a Massive MIMO network with 182 antenna elements, 16 layers, 64 transmission transmitter, 64 receivers, you need to be able to have, and to connect and to synchronize the management of the Massive MIMO, the management of the beamforming, together with the software of the radio. But not only that but the underlying cloud fabric all the way to the core. And I think that we've demonstrated in a couple of networks, we had a press release recently with [unintelligible: 31:41] where Netcracker, which is an NDC company was doing the end-to-end slicing of an Open vRan with an open core network end to end. Being able to orchestrate that in an automated fashion programmatically, that's where the true system integration efforts take its first dimension if you will. So, as mentioned a lot of the SI in the past were SI because they were delivering maybe 90% of the solution, and they were plugging some missing elements. Now, the SI work is more taking an ecosystem and optimizing it for the best use cases for the operator, and not necessarily actually delivering most of the solutions or even any element of the solution of the technical product that are being deployed.

Host:

So, going forward and maybe a bit of a futurist question and I'll start with Caroline and go around the room and finished with Altiostar, Thierry at Altiostar. What can the ecosystem that includes operators and the supplier and solution provider community as well, what can they do better to manage the multivendor integrations for efficient and fast deployment of functionality into 5G Oran? Carolyn, I'll start with you.

Caroline Chan:

We found that if a more we have seen so far and [unintelligible 35:12] summarize what folks said. The ecosystem much bigger now, every part of the ecosystem had to do its part. So, even just for integration, Intel setting up a lab in New Mexico to allow all our ecosystem parts to remotely log in, many of our partners here have actually taken advantage of that. So, we do some pre-integration work, we integrate with NEC open radio, with the openshift cloud environment. A lot of pre-integration work to be done by the time we even ship our reference software, reference hardware, and be able to really drive the cloud economics, this is what cloud does today. We must adapt, again, about the Telco mindset, the monolithic mindset, I think that's really important to try, especially if you want to get into the enterprise world.

Host:

Ihab.

Ihab Tarazi:

I think I will add to what Caroline was saying, I think what most people may or may not realize is that as Juan Carlos was saying, this is about automation, CICD, new skills, all cloud-based model. Therefore, the testing, integration, validation, cannot be the old way, it has to be based on a cloud model. If you're building a cloud network, then every piece of software you build needs to have API's open interfaces, it needs to be shippable, you should be able to have it as containers and provide access and some kind of a lab or GitHub repository. I think that the industry is slowly making its way there, all of us, and there are enough labs between all of us, I think, for people to experiment. But the idea that this software is a consumable application as part of a cloud with automation and tools, is what we need to do better. And over time, all the lab integration is becoming more software-based, so that thinking, that skill set is what we have to build for.

Host:

Honore.

Honore LaBourdette:

I think that the work that everyone on this panel is doing together and collaboratively to minimize the complexity, ensure the integration, the certification, the testing, is really a proof point to how we can help telco operators to accelerate successful deployments. There's subject matter expertise, engineering resources, lab resources, open labs, and joint labs, all being contributed by all of us on the panel, and we're working together to try to solve for some of that integration requirement and doing it upfront before the telco operators are having to confront some of the challenges that they might have to confront. And again, I think that's a proof point to how it can work and it's a critical component to this whole integration piece. I think it was Patrick that said, the integrator is then looking to bring it all together to drive business outcomes, and that's a very important component. As much as we can take the complexity out of it up front, I think the better the entire industry will be in terms of accelerating deployments.

Host:

Patrick.

Patrick Lopez:

I think we're seeing the emergence basically, of a new operating model. And it's an operating model where the operators, depending on their size, depending on their use case, depending on their geography, will be able to pick basically what role they want to play in deploying and managing, and operating Open vRan networks. But it's not only Open vRan, it's really a 5G cloud-native network, it goes beyond Open vRan, it is the core, it is the transport. If you want to be able to do end-to-end slicing, you need to be able to orchestrate from the core to the transport, to access resources and capabilities, and to do that in a programmatic fashion. And to do that well, the good news is that you need a lot less domain expertise than you needed did in the past. In the past, you needed to have specialists in radio, and within the radio, a specialist in hardware specialists in software, and specialists in core, specialist in transport, and specialists in underlay and overlay, all sorts of domains show a specialty. As we are seeing more and more vendor moving towards a cloud-native architecture in all those domains, basically, you need cloud-native capabilities, you need to be able to instantiate private cloud, you need to be able to integrate that private cloud with your far edge, and maybe hyper scalar far edge and maybe hyper scalar public cloud. And then you have a number of vendors that are coming with products and solutions that are compatible with all those environments. And all of a sudden, you don't necessarily have to have the deep domain expertise to understand within the vRan, how to configure, we were talking about beamforming the beam weights of a specific solution for a specific use case. All of a sudden, what you need is the cloud expertise to understand how to instantiate different containers, where to place them, and how to scale them up and down. And your vendors come with a number of blueprints and a number of solutions for specific use cases. So, from that perspective, I think there is a new domain, which is a cloud that enters the telco world, which requires skill enhancement, and we spoke about it see CICD, etc. but one that is well understood and internalized, actually, the level of complexity, operational complexity reduces because of automation and because of orchestration. So, I think that the role of the vendor like an NEC is right now to help any way we can with a systems integration for those operators that do not necessarily have the skillset or do not want to take that role until the level of maturity of the ecosystem allows those operators to take themselves that responsibility they desire to. Or until you have cookie-cutter solutions for specific use case that emerge and that requires low or little integration altogether.

Host:

Juan Carlos.

Juan Carlos Garcia Lopez:

If I may take the chance to disagree and just to change a bit. While I agree with Patrick on understanding that the telecom operators need to embrace the cloud technologies and practices, and that will require a change in mindset and skills, I do not agree that we have to forget about them, question some technologies like radio, transport or core. I think that while [unintelligible 42:33] and I guess this is something that you can get in the market and find partners to support you with a lot of expertise. It is the case of many of the [unintelligible: 42:42] here on the call. On the network side, I think the operator has a view and a lot experience that can be applied to design systems that are differential. And I think this is something that operators need to embrace, that is the capacity to contribute and participate in the design of the elements, in a way that they can introduce differentiating factors that can help them compete in the market. And really, in areas like a public network you need to have a more standard solution in [unintelligible 43:24] areas that are now becoming very relevant, like the [unintelligible 43:27] networks or the specific solution for certain sectors. And this capability of designing systems, on radio, on core, on transport, is specifically fit for purpose, this is a place where the operator can apply a lot of this practice that has been gained along the years and will still keep a good amount of expertise. If I may add to that, something about what we need to do to progress and improve on this Open vRan adoption, I think we have as an operator plenty of room for progress and improvement. So, I think we will have to still try to go beyond procurement for instance, when we're having access to the technology and increase the partnership abilities to find the right components for the systems they want to implement and also, to find adequate support for integration and technical support when they [unintelligible 44:22] this on the field. And here I must say that we have a good representation in this chart of companies that are great partners of Telefonica, in this journey of Open vRan and I would like to thank all of you, Intel, NEC, Altiostar, in that you have done together with us. And this is very important, that is what we have tried to keep on reinforcing our skills in working with parties to create the solution that will make a difference. We already mentioned also everything we do to adapt processes and tools and manage the new complexity. So, what we are building is much more complex than what we had in the past. We have a new domain that Patrick mentioned, the cloud domain that was not in our [unintelligible 45:08], and this is a complex one. So, we need to change the mindset and also learn how to manage that and embrace again, everything that has been learned on the IT industry for years now and this is a big challenge.

Host:

And Thierry, you get to finish yourself or if you can respond even to Juan Carlos.

Thierry Maupilé:

Well, thank you also, Juan Carlos, for you and Telefonica to be a great partner and also, someone who is really driving Open vRan as being the right direction for the evolution of networking. I think I'm going to finish with where I started., this is a very strong movement, unstoppable movement. And what is the strength behind a movement? This is a number of companies and most of them are well represented on this panel, where we share the same vision, we know there's a need, there's a need to have a strong very diverse ecosystem. 5G, there's nothing more exciting, every G is getting more exciting but this is very prevalent, this is going to be important for every country, every economy, every industry sector, it is absolutely critical. This is a responsibility to have a strong LC ecosystem. And it is more obvious that the collaboration model is working. We have operators with industry, innovators, which are working well together. They are not just industry peers, we are all friends because we know it's not going to be easy, we know that we have challenges, but we are heading in the same direction. Sometimes I tell you, it feels very lonely to be a startup and that needs to be changed. We don't want to be one of the few startups in the space but it has been the consequence of the industry structure for too long. So, now we are not just building a network or the future network, which are going to be open, it's going to be a new open platform where people will be incentivized to invest, invest on the top, we are just talking a lot about the infrastructure layer, virtualization, containerization, automation, but there's much more coming. It's about making the network intelligent, all the data that has been, in fact, sitting there naturally used as its potential whenever you're talking data for making the network more efficient or making the network more and more autonomous as we are moving forward. But I'm thinking with all these open platforms, open API, now, companies can add their value on top of this platform, and that's the vision we have and that's what we want to realize with all my friends here, which are operators and industry contributors.

Host:

Well, we had the whole gang today. I'm glad everyone sort of played well together and I think everyone got some good perspective and in the discussion, we want to thank again Altiostar, Intel, Dell, Red Hat, NEC, and Telefonica for joining us on this session the performance of ORAN architecture. To our audience out there for this session on-demand, which will be on April 26th, please log on to thenetworkmediagroup.com.



For any inquiries, please email anejad@thenetworkmediagroup.com