NEC’s Patrick Lopez, global vice president of product management, lays out the four elements of Open RAN
Editor’s note: This Q&A is adapted from an interview conducted at the Open RAN European Forum. It has been edited for length and clarity.
Q: Could we start with a bit of a retrospective look at why NEC has decided to pursue the Open RAN opportunity and what are some of your high level goals in the space?
A:NEC has been a prominent and reputable vendor in the telecom space for a long time, but it’s been focusing mostly for telecoms in its domestic Japanese market. From that perspective, very early on NEC embraced the trend of open and disaggregated networks. NEC was actually the first company to commercialize an OpenFlow-based SDN controllers about six, seven years ago. And that kind of opened the gate to a new generation of networks, networks that are more cloud capable. And those first implementations actually took place in Japan with customers like NTT DOCOMO, for instance.
Naturally, when Open RAN emerged and became an interesting trend to follow and NTT DOCOMO was one of the first companies to investigate it and then adopted it. And elected NEC to deploy what we believe were the first open radio units in a commercial network at scale.And for those who know the Japanese market, you understand that it’s a very high density, very urban environment. So the requirements in terms of performance, and in terms of stability, and in terms of radio frequency interference management are very high.
Being successful there first NTT DOCOMO and Rakuten Mobile have basically projected the NEC brand outside of Japan in the telecoms world. And as a result, NEC was invited to a number of evaluations from vendors that are interested in Open RAN and we’ve been able to use our experience of deploying the world’s first massive MIMO Open RAN [network] in a dense urban environment. And that has proven quite successful with other operators as well outside of Japan, particularly Western Europe, but also in other parts of Asia Pacific and in North America.
Q: From your perspective, from NEC’s perspective, what makes a radio system open?
Open RAN “is about enabling choice and innovation”
A: There’s a lot of talk about open and it might mean different things for different people. Open RAN means that basically you’re breaking down the radio access network into separate elements. So what used to be essentially one appliance connected to an antenna that was provided by a single vendor with proprietary interfaces and optimization is now broken down in at least three elements, which is the radio unit, the centralized unit, and the distributed unit.
We’ll talk about RIC maybe a little later. That’s the fourth element. But basically, the concept of Open RAN is that you break those elements so that it can scale independently from each other. You separate the hardware from the software and you deploy on commercial off the shelf hardware. And most importantly, you have open API and open interfaces between each of those elements. That allows you to scale more efficiently.
And it allows you also to mix and match in terms of vendors for those elements. So multi vendor is a true important part of Open RAN. And that’s where the limit is between Open RAN and virtual RAN, if you want. There are a lot of vendors that have announced virtual RAN capabilities, which is basically the separation of the hardware and the software and the virtualization of the software to maybe a container or virtual machine-based environment. But that does not guarantee that you have open interfaces between the element and that you can deploy multi-vendor environment. Actually, a lot of virtualized RAN solutions are still from a single vendor. From our perspective, Open RAN is about enabling choice and innovation and that comes from opening up with interfaces and allowing different vendors to participate.
Q: Tell us a little bit about some of the work you’re doing in Western Europe. I’m particularly interested in the projects you’ve got going with Telefónica in several of their markets. And can you discuss the process of going from lab-based interoperability testing and verification to putting that into a commercial network and then scaling it up?
Telefónica as proof that NEC Open RAN can scale
A: Open RAN, like any new technology, requires maturation because essentially you’re taking one system that was proprietary and tightly integrated by a single vendor and you break it down and you’re implementing new interfaces, but also you bring in new vendors. There’s a certain level of complexity that comes with that. And Open RAN is fairly new. I think four or five years ago, the term Open RAN emerged and now we see it commercially deployed at scale. So it’s been really fast going, but there’s still maturation that needs to happen. And a large part of that maturation is in the integration between the different elements and the different vendors. A lot of operators have different aspirations towards that. You have some vendors that would like to do it themselves and take the role of integration of all those vendors and all those elements.You have some vendors that would like to do it, but realize that they might not be able to do it just now either because of skillset, because of capacity, or because they need to learn. And then you have operators that are just not interested to do that at all. They want an integrator to come in and to do it all for them. So across that spectrum basically, our experience has been to deploy those systems and to help you to integrate an end-to-end system with multi-vendor environments.
This is why we launched NEC Open Networks at Mobile World Congress. NEC Open Networks is basically a market promise, which is that we will deliver radically open systems, multi vendor, but no strings attached and no compromise in the sense that a system, an Open RAN system, can be open and multi vendor, but there’s no sacrifice to performance, to stability, to availability. We are able to guarantee end-to-end the same or better level of performance and capability than an integrated proprietary environment.
Telefónica, they have trusted us with the prime integration of their Open RAN project in four different countries, Spain, Germany, U.K., Brazil. And Telefónica is a large operator and each of those networks is a different network with different conditions, different spectrum, different vendors. And it’s really, I think, the proof for the entire industry that Open RAN is ready and is mature. If you can deploy that in all those different networks, under all those different conditions, with a different variety of vendors, and it works equally across all those, it means that Open RAN is ready to become a technology that can be deployed in all environments moving forward. That’s really in that case, the capacity to demonstrate that not only Open RAN works in the greenfield environment, we’ve seen some announcement that some operators have done that, where it’s relatively less complex if you don’t have to integrate with an existing infrastructure, but here what we are doing is demonstrating that you can deploy and industrialize Open RAN in the brownfield environment that has existing legacy systems.
Q: You characterized it as a maturing technology and it’s been interesting for me to watch the way that the value proposition is articulated has also matured. It used to be all about supplier diversification and lower TCO, but now the focus seems to have gone over to Open RAN as this vector for innovation, particularly around the RAN Intelligent Controller. Could you maybe share for us that vision of a programmable RAN, what it’ll mean for operators and what it’ll mean for even the end user?
Programmable networks, Open RAN and the promise of 5G
A: Programmable networks have been an important part of our value proposition and our strategy at NEC. Some might not know but Netcracker is a wholly-owned subsidiary of NEC and it’s one of the leaders in the service management orchestration and automation field end-to-end. Let’s take step back from Open RAN for a second. What is the promise of 5G? The promise of 5G is to create connectivity products that are going to be adapted for different use cases, different verticals, different industries, different devices. Whereas today, we all have the same connectivity and we all share that connectivity and it’s kind of the best effort and the only difference from one operator and another is how much you pay. You might have a different performance profile on a different cell site, but it’s all the same.
Now going forward, we all understand that an autonomous car, that collaborative robots, that an application for remote controlling a construction system, they all have different needs from the connectivity and from the network. And it’s not possible to satisfy all those needs simultaneously with a best effort network, which is everything that we’ve done up to 4G and even 5G NSA. In order to satisfy those needs, you need to be able to do slicing and you need to do orchestration and automation across all domains, core, transport, and radio. And now if we go back into Open RAN, an exciting part of the innovation is the radio controller because it allows that RIC basically to have a management system across all the RAN in a multi-vendor environment and for the first time orchestrate the experience of RAN.
What does it mean? The precursor of that technology was a self organizing network, but it was really centered around specific vendors and it didn’t work very well in the multi-vendor environment. And it was basic. Now, you can imagine a network that is Open RAN, and that has RIC, and you can imagine that for instance that network is composed of massive MIMO antennas. And at the network level, for the first time you’re probably going to be able to detect interferences between the beams themselves of massive MIMO and in real time to reorient those beams so that they don’t interfere with each other. And what we found from our studies…is that in some cases more is not better in the sense that more beams have capacity to create more interference, to actually degrade the user experience. But if you’re able to focus those beams where you need them, and even to shut down some beams that might be adjacent and might be interfering with it, you can actually achieve a better user experience and better overall network health. So I think that’s what’s exciting about the RIC is that it opens up a number of new optimization capabilities to reduce power consumption, to increase user experience, or to create very specific connectivity products for specific industries, and use cases, and devices.
Q: You made the comment about how the way we have built networks results in not much differentiation. When you think big picture about running RAN and other network workloads in a multi-cloud environment, any thoughts on what operators need to do organizationally with their people and their workflows to fully take advantage of the flexibility and the speed that would come with investments in cloud-native networks?
‘Cloud is a revolution’
A: I think cloud is a revolution, right? And you can argue that 5G networks, particularly 5G Standalone network and advanced 5G networks, they’re more cloud than telco. And the difficulty is probably not in the connectivity itself…The difficulty is the orchestration and automation of that system at launch, because all of sudden, if a large part of your network is composed of virtual elements and those virtual elements basically pop in and out of existence or scale in and scale out based on demand and based on specific conditions, well, it’s no longer a network that you can manage manually like we’ve done before.
Before, we know that to do any change into a mobile network, it’s 18 months, and $5 million, and it involves somebody physically going into a data center and physically accessing a console on a server and making changes. That’s been telecoms for the last 20 years. Now with the cloud, all of a sudden, well, you don’t need somebody to physically go somewhere, that you can with a single pane of glass actually have a good visibility of the network, understand how the different workloads articulate, and where they are based, and you can programmatically manage your network experience. But obviously in order to do that, it requires different skills than just pure telco engineering. And I think most network operators have recognized that and they’re scaling up and they are hiring specialized skillsets in order to do that, but it’s a process, it’s a journey, so it’s going to take a little bit of time. So in the meantime, they’re relying on a lot of companies to help them along this journey. And best case, at NEC, we’re not only a vendor of telecom products, but we’re also a full SI company and we are helping many of our customers basically scaling up…in order to make their network more programmable.
