The Future of Telecommunications Starts with Autonomous Networks

‼️ Autonomous networks = the next big thing? Will Townsend, VP & Principal Analyst at Moor Insights & Strategy gets the download in his conversation with Vishal Singh, Nokia‘s Senior VP, on how AI and machine learning are revolutionizing network efficiency and the power of autonomous networks in the telecommunications sector.

‍

📺 Tune in as they cover:

‍

🔹What’s driving the industry’s shift towards autonomous networks

‍

🔹How AI and ML are boosting network efficiency and autonomy

‍

🔹The challenges and opportunities companies face in adopting these technologies

‍

🔹Nokia for service providers’ strategy for supporting telecom operators in this evolution

‍

🔹The future impact of autonomous networks on the telecommunications sector Learn more at Nokia.

‍

Watch the full video at Six Five Media, and be sure to subscribe to our YouTube channel, so you never miss an episode.

Will Townsend:
Welcome to another Six Five Media virtual webcast. Today, I’m spending time with Nokia and we’re talking about the company’s journey with autonomous networks. Joining me today is Vishal Singh. He’s the Senior Vice-President of Cloud and Network Services for Nokia. And I’m Will Townsend. I manage the networking, telecom, and security practices at Moor Insights & Strategy. Vishal, it’s great to talk to you today.

Vishal Singh:
Thanks a lot, Will. Great to be here with you. Look forward to the conversation.

Will Townsend:
Well, let’s start with a little bit of context here. So, autonomy certainly drives efficiency, but from your perspective, what does autonomous networks do in the context of telecommunications and why is it so crucial?

Vishal Singh:
So, we have been in this journey of, I would call it driving automation in the networks for a long, long time since the inception of I would call it 4G World. And even before, people have talked about driving automation. There are two aspects to it. One is obviously the aspect of efficiency, which automation naturally brings to the table. But more importantly, it’s really the customer experience. And as we progressed into the digital world, this idea of putting customer in the center of the universe in everything that you do as a service provider became really a central idea. And you can’t really achieve that without actually driving automation at the core of everything that you do. And when we evolved into the cloud native journey from the 4G world into 5G and look at 5G SA and the rest of the softwarization of the network, or I would say cloud native journey, or cloudification of the network, autonomous became very critical because you can’t really do that degree of cloudification or, I would say, become hyper digital without actually going autonomous on the network, which is basically, for all practical purpose, becoming hands free. And therefore, this becomes a central idea now. As we evolve from 5G, 5G advance into 6G world, as we evolve rapidly into a digital world and cloudification of the network, being autonomous, or driving autonomous network idea, or the journey of autonomous network has become a central theme in many of our conversations with our customers.

Will Townsend:
From my perspective, autonomy can drive higher degrees of reliability, and availability and network assurance. In the telecom industry… I mean, it’s a brutal industry. I mean, operators have to deal with a lot of challenges. From my perspective, monetization has been a big challenge beyond access. But I’d love to get your perspective on what could autonomous networks do for telecom operators beyond operational efficiency? And what are some of the challenges that it can address for them?

Vishal Singh:
So, let’s talk about first what it can do. If you think about it, monetization you touched upon, we are putting in all this investment into building these networks. And if we don’t have a path to make money, there’s no point doing it. This has been a very big topic across the industry. We lost the bus, or we couldn’t get on with the bus in 4G world, but 5G provides a unique opportunity for customers, our customers, to make money. And how do you do that? If you think about it, the central idea of 5G was to monetize through B2B. When you go down the path of B2B, the use cases that you talk about, the ability to drive low latency, high speed networks, it is not possible without actually going fully autonomous and giving the control to the enterprises to be able to provision bandwidth on demand, for example, or quality of service, for example. All this needs seamless execution in the network. You cannot have manual intervention if you want to go live in a stadium with an NFL game, with all the cameras and everything else coming into play because you need really, really very, very high bandwidth network, low latency, high speed network provision for that particular game.

And there are many such examples. The gaming industry is big in terms of what it needs to do. And then robotics and drones. I can rattle use case after use case of what is possible. And then the question really is about the challenges. The challenges fundamentally in realizing that vision or realizing that opportunity is really this network, which is still very hybrid. We still have 4G into the network. We still have 2G and 3G. The networks are not truly cloud native 100%. And somewhere or the other, there is a lot of friction in the network. And operators are doing their best, because there’s a lot of transformation going on on the BSS, OSS side of the world. But where we come into play as Nokia is to help our customers reduce the friction between the digital channel or the digital engagement layer of the operators to the network itself. That’s where our ability to stitch all the capabilities that comes from the network bottoms-up with the front end of engagement or digital channels with our autonomous network solution to help them address the challenge and actually pivot towards executing things like slicing, which can unleash the B2B opportunity that I just talked about.

Will Townsend:
Yeah. Network slicing is exciting. Once these operators get to 5G standalone networks, they’ll be able to do that. And to your point, automation is critical to get to that nirvana of network slicing and steering, to your point, latency and throughput, and tailor that to applications and use cases and workloads. But hey, our conversation would be remiss if we didn’t touch on modern AI. And AIOps is something that’s been around for quite some time. Machine learning and artificial intelligence have been around for decades, but now we have generative AI, we have agentic AI that’s fast following that. I know this is an area that Nokia invests in quite a bit. So, from your perspective, when you look at modern AI and machine learning, how is that driving and accelerating this whole notion of autonomous networks?

Vishal Singh:
We talked about this topic of classical AI, ML for quite some time. We’ve been doing quite a bit of work. And just to take a step back, data has been at the core of this whole subject. And service providers have always claimed that they know their customers better. They have so much data about the customer’s experience, their usage of the network, but it is the over-the-top operators who actually took the cream away and made more money while there has always been so much information available in the network with the service providers. And the maximum they could do is put that data into a data lake, spend a lot of money trying to just put things here and there, but they could never really unleash the potential of that data. To me, generative AI is actually the first real big change that allows the service providers to do so much more because with classical AI and ML, we could actually do a lot of learning. So, we could embark on making data available and make the models learn what you can do with that data. But learning can only be possible based on what algorithms you are building based on what you want to achieve. Correct? It was still reactive as classical software engineering, but with generative AI and agentic AI for that matter, now you can do some crazy stuff.

So, just talking about generative AI, we have data, just to give you an example, where we know what happened on a call, or a video call, or simply a voice call as a subscriber or a customer moves from one part of, let’s say, going to office in the car, driving, making a call. And the experience goes up and down. But it is impossible for customers to stitch that information, because the data that gets generated over different networks from different vendors, different protocols, different standards, it’s just very, very fragmented. So, the opportunity that we have is to, first of all, help our customers to create that coherent data of that experience by stitching and harmonizing it across different vendors, multi-vendors, across different domains. When I say domain, we are talking about transport core and radio networks. When I say multi-vendor, we’re talking about Ericsson, Nokia and others across the board. Correct? And therefore, as CNS as my business, we have amazing ability to help our customers really harmonize and stitch that data, and create that coherent record, and then make that available to be consumed by any northbound application.

Now, when you do that, or I would call it that becomes the basic hygiene for applying generative AI large language models, because you need large language models to be able to consume that data all along, so that it is so rich that somebody can just ask a question that, “What happened with this customer? He is a high-value customer. Tell me how has that experience been for this particular high-value customer in Manhattan, New York today between 9:
00 and 10:
00?” Correct? Now, when you ask that question, you need a response, which is based on all kinds of possibility. And you can’t build a software for that question to be asked. So, that’s why generative AI is slightly more different from classical AI or machine learning because this question is just generic. Correct? Now, you apply agentic AI, you can do crazy stuff because now if you apply the human context to that generic question, then somebody… If you asked me that question as a human, I would say, “Okay, you asked me that question.

Let me tell you what all you can do with that. You can actually see this person was on breakfast watching a video, and what video he watched. Was he watching an NFL game, or was he on Netflix, or was he watching news? What news channel was he watching?” That’s what agentic AI does with that. It just applies the classical cognitive, or I would say augments that with human logic beyond the generic question. The generic question is a generic question, but then you can do so much more. And then you start throwing that to a model and the model says, “Okay. Now, we have an opportunity to offer a service to this customer without human intervention.” Correct? That’s what you can potentially do. So, now you’ve got a sales and marketing teams to actually be able to get opportunity to create a product to be offered to this high-value customer, which is just, for all practical purpose, a snap away in that sense because you can actually create that product in real time for that customer to be able to subscribe that immediately. You know, to do all this autonomous network becomes foundational. You cannot do this kind of… You apply the generative artificial intelligence, apply the large language model, train it, make it available to an agentic AI, but if the network is not frictionless, how do you get to the outcome? How do you get to the monetization?

Will Townsend:
Exactly. And all of that infrastructure also supports, in general, AI workloads. Because you mentioned the multi-domain, large language models hosted in the cloud, AI is moving hybrid, smaller language models, network edges, and you need the resilience of an autonomous network to make that all happen. But equally important is security. And there’s been a lot of news. I’m not going to pick on anyone in particular, but wireless networks, telecommunication networks are critical infrastructure. Bad actors attack them to extort ransom and do terrible things. At the same time, attack surfaces are increasing because of this disaggregation that we’re talking about. Cloud, on-prem, you’ve got Converge Core hosted in the cloud, you’ve got virtualized RAN, and so it creates this huge attack surface. So, from your perspective, what is Nokia doing to ensure the high degrees of security in addition to autonomy with the network infrastructure?

Vishal Singh:
So, to me, there is no autonomous network if you don’t secure by design. This is just basic stuff, correct? It’s just, to me, hygiene. You have to build your networks in such a way that they’re secure by design, and especially critical infrastructure we are talking about. Why so? Because two things. One is obviously the stuff that’s going on is putting a lot of pressure, or the regulators are actually really, really going very hard at putting a lot of conditions on our customers to be able to build these networks, which are secure by design. So, there is a clear push from the regulatory environment to consume that. But that, to me, is absolutely reactive work we will do and industry is going to do. But just the idea of what you just mentioned, that the disaggregation and the distributed deployment of these networks is so crazy that the attack surface is virtually everywhere. And the idea of Zero Trust is just so much more relevant from a classical IT world into this world. And today, the exposure to the bad actors, especially if I may bluntly say the nation state actors, who are actually going after our customers here, let’s say in North America, is just mind-boggling.

So, Nokia actually is, in my view, with the experience we have in the telecom industry, the knowledge of the network and our ability to understand what happens in the network, we are in a very unique position compared to all our other counterparts from the IT world to really help our customers be in that journey, to secure the network, to secure the critical infrastructure. But nobody can do what we know about the network and what we can do. So, we have a portfolio of products from basic things like certificate management all the way into identity access, to network detection and response. So, our focus on critical infrastructure security has been on endpoint detection, on network security, as well as in XDR, extended detection and response. We are actually, for all practical purposes, in a very unique position to help our customers in that journey. And we are already seeing some very good results.

Our security business is growing in 20, 30% year-over-year. It could easily double year-over-year. And somehow this is also timing. As I said, the demand is there, the pressure is there from regulators. There is a lot happening in the network. But if you just bring this back to this idea of autonomous network, you have to build networks which are secure by design to be in the journey of autonomous network. So, if you think about it, TMF has got a autonomous network methodology or a framework, where all operators are today at 1.5 to 2 in the rating in terms… This is from one to five. Out of that, most of my customers are 1.5 to 2. And we’re helping these customers to, first of all, do an assessment of their maturity, in that sense, on the TMF maturity model.

And we find security is a big question mark, big topic of discussion in that. Therefore, we are also helping our customers to secure the network by design, both in terms of the portfolio we offer, the products that we bring into the network to build the network, but also in terms of helping them build their safe posture, security posture, ensuring that they can actually take on a lot more, and making them more proactive through our products in security, things like what we do in Cyber Dome, which is helping them actually have an understanding of the threat score at every network function level, so that they can take specific action and apply playbooks. So, it’s a lot of exciting stuff actually. I’m very, very excited. I’m sure you can hear that in my voice.

Will Townsend:
Oh, yeah. Secure by design, it’s been a big focus for companies. And from my perspective, security is a team sport. You have to involve every stakeholder within the organization, not just push it on the NetOps team or the SecOps team to manage. But Vishal, as we wind our conversation up, I’d like to spend a little bit of time with you to understand your perspective and your recommendations that you would make to telecom companies and operators that want to begin this journey towards a more autonomous network infrastructure.

Vishal Singh:
See, first of all, my recommendation number one is that you need to start the journey today. And you have to start it with a building block. You don’t have to really embark on a Big Bang transformation. As we know, these networks are live networks, they have hybrid networks, all kinds of technologies in the play. So it cannot be that one fine day you can just embark on it like a big bang program. You have to have a strategy to take you from level 1 or 1.5 or 2 to 5. And there it involves a certain understanding of what that journey looks like. So that’s number one. Start it through a building block approach, one step at a time. Second point is that you have to look at a partner who understands your network very deeply. This journey of autonomous network cannot be made through someone who does not really understand your network.

Understanding the network is not a vendor-specific approach. It is a multi-vendor approach. So therefore you also need to select a partner, or engage with a partner, who has a multi-vendor credential. So, that’s my number two recommendation.

And last, number three is, it’s a digital journey. Irrespective of whoever you select, the partner needs to really understand and create value for you in terms of the business impact. Ultimately, it is about the business, it’s about your customers as service providers. And if we, as Nokia, for example, do not understand whatever we do, how does it create business value, how does it create business impact, how does it help you monetize your network make money, there is really no point. So, these are three fundamental recommendations, I would say.

Will Townsend:
Yeah. No, I think those are solid recommendations. And Vishal, this has been a very insightful conversation. I also want to let our viewers know that I’ve written on this subject, Nokia’s vision and journey with autonomous networks. There is a research paper that is posted to the Moor Insights & Strategy website, as well as some additional more informal blog style articles that go into a little more depth around AIOps, and security and some of the things that Vishal and I have talked about today. But Vishal, thank you for your time.

Vishal Singh:
Thank you. My pleasure. And I really appreciate having this conversation.