ARTIFICIAL INTELLIGENCE

Christina Cardoza: Hello and welcome to the IoT Chat, where we explore the latest developments in the Internet of Things. I’m your host, Christina Cardoza, Editorial Director of insight.tech. And today we’re going to be talking about the rise of intelligent networks thanks to 5G and AI with Ian Fogg from CCS Insights. Hey, Ian, how are you doing?

Ian Fogg: Hey, doing well.

Christina Cardoza: Thanks for joining us today. Before we dive into the conversation, can you tell our audience a little bit more about yourself and what you do at CCS Insight?

Ian Fogg: Sure. So, we’re a research and analysis company. I lead the networks or the network-technology research at CCS Insight. Before CCS Insight I’ve worked with operators, analytics companies, other analyst firms. So, really many, many years of experience in network technology.

Christina Cardoza: Awesome. Looking forward to dive into it. We just had your colleagues Martin Garner and Bola Rotibi on the podcast. We were talking about IoT protections for 2024, specifically in the edge and AI space. But as part of that report—and I encourage all our listeners to go take a look at that report to see what’s coming over the next couple of years—but as part of that report there was a lot about 5G and networks in there. So I wanted to chat with you today a little bit more about what can we expect in that space, and these things sort of run parallel sometimes.

So before we jump into what we have coming, let’s take a look about where we are today, or what got us here today. So can we just start off explaining how has the 5G- and IoT-network landscape changed over the last few years? What have you been seeing?

Ian Fogg: Sure. Well, I mean 5G first launched back in 2019, but that was really very early versions of the 5G standards. And what we’ve seen more recently is tremendous enhancements in what 5G is capable of. We’ve seen 5G being used for more and more different things.

So one of those is, for example, private networks, and we’re tracking the number of private networks that are launching or are being announced. For example, in 2023 up to Q3, we saw 1,279 private networks announced with revenues of over €100,000. That’s up from 1,081 in 2022 and 761 in 2021. 5G was deployed in 45% of those networks we saw announced in 2023.

We’re seeing other things happen too. We’re seeing non-terrestrial networks arrive. We saw a lot of activity in that last year around this time, with announcements from, at the time, Qualcomm, and that’s part of the upcoming release 17 5G standard as well. We’ve seen more focus around REDCap—reduced capability—on 5G starting to be talked about by the vendors, but not really out there yet in the market.

One of the other things we’ve seen happen last year is the operators announce OpenAPI initiative, something the GSMA—the operators organization—has been very big on. And that includes network APIs too, to manage network quality and other network-type settings. So we’re seeing all kinds of things happening over that last year.

Christina Cardoza: Absolutely. A lot happening and, like you mentioned, 5G started to come around, maybe around 2019. The conversation started well before that, and the conversations are still happening. It always amazes me how much 5G is being adopted or deployed, and how much there is still to come and how much we still have work in this space.

So, what would you say, looking to 2024 or even beyond that, what would be the next step for 5G in these IoT networks coming up?

Ian Fogg: Well, this takes us back to the predictions that we put together at the end of last year. So, some of these are related to what I just said, some of them are looking a bit beyond that.

So, one of the predictions we had was that by 2025 a digital marketplace for app-based network functionality offers more than a hundred versions of network capabilities and APIs. We’ve seen some small initiatives, but we expect that to expand tremendously over the next 18 months or so.

We’re also expecting that hybrid private and public 5G, through network slicing, emerge as the dominant option for private networks by 2030. So what that is, is that a standalone private network is using a dedicated network just for that offering. But a hybrid solution is using some of the new capabilities of standalone 5G, network slicing to give you a quality of service that is different to other people using the wider cellular 5G network. And that then bridges the gap between a location where an enterprise may have the dedicated private network, but they maybe want people or devices to move between those dedicated sites. And the macro network, the regular 5G network with a network slice, bridges the gap—a hybrid solution. So that’s something we see becoming very dominant by 2030. And obviously it’ll happen not just suddenly in 2029; it’ll happen gradually over time as we see something happening.

Something else we see—although there’s been some negativity around NTN—we see that continue to grow. It’s valuable for the IoT space. If you think about container tracking, it doesn’t require very high bandwidth services. But there are other things that we see happening very near term, like the smartphone space, where we see that’s still growing despite some of the negativity; there’s a lot of activity in that space. It’s part of the upcoming 5G standards, part of release 17 to have a non-terrestrial network capability. And by 2027 we expect 15% of smartphone users have satellite-enabled devices.

Now, what does that mean for IoT? Well, often what has happened in the past is the consumer space has driven innovation that then gets reused for other things because there’s a commonality. Once you’ve built that satellite capability, the satellite players can choose to support more than one type of customer.

Christina Cardoza: Great. Now, you mentioned more of a rise in hybrid private and public networks. When I think of these terms, I equate them to the cloud: hybrid cloud, private cloud, public cloud. Is this the same concept of using networks? So, for instance, when you have more mission-critical solutions or applications, those would be on a private network, much like they’d be on a private cloud. What’s the distinction there, or the similarities between the different types of cloud uses and then the network uses? Will these things be parallel? Also, if you’re using a hybrid cloud, would you likely be using a hybrid network?

Ian Fogg: I think it depends on the use case. You think about a hybrid network, the key thing that’s really driving that’s, that’s enabling that, is this switch from the initial versions of 5G that started back in 2019, which were using what’s called non-standalone access. They relied on a 4G core network, which means you didn’t get all the capabilities of 5G that were being hyped in the 2017, 2018, 2019 period. You need to have the 5G core network up and running at the mobile operator, at the service provider. And that enables a whole load of new network functions, and one of those is this thing called network slicing, where you can have essentially a dedicated quality of service. So that’s something that’s happening.

I think the cloud space is parallel, but I think it’s an interesting analogy to say, well, there’s more than one type of cloud; there’s more than one type of network that can offer a high quality of service.

Christina Cardoza: Great. And at the same time that all of this is happening, 5G is getting more advanced and we’re adding more capabilities to it. The demands and the needs of the Internet of Things is growing also; companies want to connect to the internet more. They want fast, reliable, real-time information, and that requires the network to work fast also. So how can the network and 5G keep up with this ever-growing demand of IoT and some of these more intelligent solutions and devices coming online?

Ian Fogg: Well, I think this hybrid capability is particularly important, because it does bridge the gap between maybe a dedicated network that’s maybe a dedicated private network that’s maybe in a port or in a manufacturing facility or in maybe a warehouse or a barn or something.

I mean, one of the areas we looked at here actually for hybrid private networks or hybrid 5G was around precision agriculture. And if you imagine agriculture, you have the farm, you have some farm buildings where you may have a dedicated capability, but you can’t put that over the whole farm area. It’s far too big, far too problematic. And that’s a classic example where a hybrid solution enables an agricultural offering to have a high quality of service right really across a wide range of areas.

One of the other things I think we see as an important thing at the moment, and increasing so, will be using AI in a whole range of different areas. We think AI is going to be important in a range of types of products in different parts of the network. But one of them is we see that AI enables 5G networks to improve their availability significantly, perhaps even to move beyond five-nines availability by managing the traffic patterns better, by making sure that the network isn’t just on, it’s offering a good enough quality of service and managing around outages or downtime issues to give a self-healing element to the experience. So that’s something else we see increasingly important on the network side.

Christina Cardoza: I love that example of the agriculture and in the farm, because that’s one of the examples that IoT is just growing and demand for it—farming, you wouldn’t really think becoming online or using these advanced capabilities, but that’s just the reality today. Every industry around the world is leveraging Internet of Things and and network.

I want to dig deeper a little bit into the role of AI. You mentioned that it’s going to be improving availability and it can do things like self-healing. So exactly what’s going on in this space? How can AI be utilized and be applied to do these things and enhance the 5G and IoT networks?

Ian Fogg: Well, the network guys are using AI on all kinds of areas. They’re using it to improve the RAN management, because as you’ve gone from original 4G to 4G Advanced or LT Advanced, and then onto 5G and then onto release 17 and onwards, the complexity of the RAN has got much greater, and there are more settings that need to be managed. The interaction between the base stations, between different frequency bands, is much more complex, and AI is a key way of enabling ongoing management of that RAN to improve the coverage and improve the performance.

We see it being very important in the Open RAN rollout too. So, Open RAN is something where, historically, service providers have bought a base station from a network vendor, and everything is basically integrated and included from the same vendor. The concept of Open RAN is that there are interfaces within that base station so that a service provider can mix and match different suppliers. And we think the complexity there is something that AI can help improve the Open RAN experience and help drive Open RAN adoption. So we think that that’s very important too.

Something else we think is interesting and important is around green issues: hitting carbon targets, managing energy costs on the network. And, again, what you need to do there, if you’re a service provider, is you want to keep the performance for your user base good, but minimize the energy usage. And so you want to drop the energy usage but still maintain the network experience. So how far can you cut back network resources and yet still offer the experience that users need.

And that’s something where a machine learning tool can help with that RAN optimization. And we feel by 2025 a combination of intelligent radio access network–technology, automation, and AI-driven power-down techniques enables at least three leading operators to bring forward their carbon-neutral targets by several years. It’s one of the predictions in our report.

Christina Cardoza: Yeah, I was just going to ask that. It’s interesting because I think over the last couple of years sustainability and green issues have been a top concern for industries all over the place and being able to use AI and even these 5G networks to hit some of these goals. So I was just going to ask what you predict for the next couple of years: is it going to be more important, and do you see organizations and industries actually being able to reach some of their goals, or those goals really becoming a reality over the next couple of years with some of the capabilities happening in 5G and IoT networks?

Ian Fogg: One of the sustainability angles here is around smart grid. There’s obviously a massive drive to remove coal, oil, and gas carbon technologies from power generation. One of the options to replace there is obviously nuclear. Nuclear has enormous capital cost, long lead times, a lot of complexity. So a lot of people are looking to solar and wind power as a cost-effective and versatile way of generating green electricity.

The challenge there is that solar and wind power generation is not always predictable—depends on cloud cover, on the solar side, and time of year. And on the wind power, it just depends on what the weather systems are doing at the time things are happening. And you could have night times where there’s too much power being generated and you want to encourage end users to consume some of that electricity. There may be other times you want users to drop their energy consumption because it’s too expensive or there isn’t enough green power being generated.

So we think greater use of solar and wind power necessitates that smart grid technology to manage the supply and demand. And we expect smart grid technology will become widely adopted in most advanced economies from 2028, if not before. But we can see some signs of that happening even now.

Something else we see as important is power-as-a-service, an integral part of tower infrastructure services by 2025, because one of the challenges or one of the difficulties of getting power to tower sites: it’s difficult, it can be time consuming, it can depend on permits. But what we see happening there potentially is the tower company managing the power to the site. So, taking a greater role in the site management than they have done before. And we think that’ll become a cornerstone of tower companies’ offerings by 2025.

Christina Cardoza: One thing that interests me about the smart grid is obviously we have all of these new and intelligent devices connecting to the network and going online, and we need to prepare the grid to be able to handle all of this at the same time we’re trying to be more sustainable. And then as part of being sustainable people are using more—electric vehicle technology or electric technology and plugging that into the grid.

So is there anything happening in the 5G or network space that is going to be able to help that demand or help the smart grid really become smart and become more sustainable at the same time?

Ian Fogg: So, if you take electric car EV charging, typically you want to do that overnight if you can, if it’s on a residential solution. So you can see a range of ways that network technology come in there. Many EVs have cellular capability, so the user can remotely control their car and set the power-saving modes, tell it where and when to schedule the charge. Many EV chargers in the home have also a similar remote control—often that’s on Wi-Fi or something else.

But then you have smart meters in the home, which typically have a cellular connectivity. So the power company can monitor what’s being used and charge people—in some cases charge people in very granular ways. There’s a power company in the UK, for example, which has a tariff which has half-hourly pricing. So the pricing varies dynamically during the day based on what the power generation and the overall consumption is. So you can see it in different places—whether it’s in the EV, you have connectivity; whether it’s in the EV charger; whether it’s in the smart meter—three different places just involved in the end-user EV charging process where you could see network technology becoming very important.

Christina Cardoza: Great. Yeah, it’s definitely very interesting to see. We always talk about these elements different from one another, as silos: this is what AI’s doing in this space, this is what network’s doing in this space. But it’s really interesting to see it as an end-to-end solution, how we’re using all of these technologies to really meet these goals and to make some of these ideas as a reality.

One thing I wanted to talk about is, obviously from this conversation it sounds like there’s still a lot of work to go in 5G. 5G is still going to be around for a while and keep improving itself and improving industries and businesses and other technologies. But we’re already hearing some rumblings about 6G come up. So I’m not sure if it’s too early to start talking about 6G—what’s the reality there? People are getting hyped about this, but where are we with 6G? When should we prepare for 6G, and how long will 5G still be dominant?

Ian Fogg: So there’s absolutely work happening on 6G, and I’ll talk about that in a second. But I think you are absolutely right: that 4G is still important in the 5G era. You think about that EV scenario I mentioned a second ago: most EVs, they have cellular; it’s a 4G cellular radio. You have a smart meter: it might even have a 2G radio in it, let alone 3G or 4G. Although they’re obviously progressing and upgrading them to 4G to enable a 2G switch-off to happen. So these network eras tend to overlap.

So even when 6G arrives, 5G will continue to be important. 4G will still be probably around as well, because these things are interoperable and they will continue to exist. The reason 6G is interesting and the things we can already see happening are we can see spectrum discussions being advanced.

We saw WRC—the international conference that happens every few years to coordinate spectrum alignments globally—so there are large economies of scale in any of these offerings. Discussion happened there around 6G bands—typically in the cellular space—people looking at the 7 to 16 gigahertz range. But there’s also some terahertz capacity, which is very, very high frequency. If you’re familiar with some of the 5G offerings using millimeter wave, this is even higher frequency than that, which is also being discussed—very line of sight, very, very high capacity, probably very short range. So there’s work happening.

There’s also work happening on the use cases for 6G. One of the use cases, which is different to what we’ve seen before, is using the cellular network to sense what’s happening. It could be sensing how much traffic there is in the roads, sensing people walking down a pavement. And that is something which is—you can see examples of that are happening now. Some of the—in the home, for example, some of the Wi-Fi access points can sense whether people are at home and use that as an alarm system, like a crude alarm system. But that’s quite a crude offering at the moment.

But part of the thinking around 6G is the 6G network would have a wider sense of capability to sense things happening across cities, across large areas. And then of course that means you can then draw analytics from it and make actions on the back of that. And that’s one of the key new use cases that’s being discussed around 6G.

In terms of timescale, which you asked about, is stuff’s happening now. I guess it depends who you are. If you are a company looking to deploy something today, 6G isn’t probably relevant. If your product roadmap is much longer, if you’re looking at kind of around 2029, 2030 or onwards time, and if you are starting to work on those sorts of things in your roadmap now, then 6G should be something that’s in that roadmap. They’re obviously not imminent, but back off. If you are a network vendor, then your R&D labs are fast furiously working on 6G things at the moment. And you are much more advanced in your 6G thinking.

Christina Cardoza: I think that point you made about there will still be some overlap and these technologies are interoperable, so I think that’s a really important takeaway for listeners. Also, obviously you just mentioned all of these benefits and capabilities and new use cases that are coming with 6G that can make the industry very excited to hop on it, but it doesn’t have to be a “6G is here now; move everything to this technology.” There is overlap, there is interoperability, so you can move things over slowly, what makes sense and with time, and be more purposeful about these changes coming, rather than going in all or nothing. So I think that that’s a great point. This has been a great conversation.

We’re running a little bit out of time, but I’m just curious, in this space, are there any final key thoughts or any predictions you want to leave us with, what we can expect to come?

Ian Fogg: I think just one last one on the 6G side. So, one of the predictions we had around that was that by 2030 the first 6G-powered, massive twin city is announced. Digital twinning is this idea of having a replica of the physical world in the digital world. And we see 6G being important in that, partly because what we just talked about around sensing and around smart city–type uses, 6G will have a lot more capacity; it’ll probably have much lower latency, more consistency, and it’ll have this sensing capability too.

I mean, one of the classic areas I think this might get deployed is perhaps in the Middle East, in the Gulf region, where there’s enormous efforts to build a number of massive new cities basically from scratch. Greenfield—well, not greenfield, because it’s the Middle East—but in new locations, a planned city. And if you are building something from scratch, you are able to do things differently, and it may well be that the first massive digital twin city 6G enabled happens in that region for those reasons.

Christina Cardoza: Wow, 2030 sounds like a long ways away, but we’re already in 2024, so, yeah. It’s going to be here before you know it. So I’m excited to see how this space progresses to get us there over the next six years, where else 5G and IoT networks are going.

But it’s been a pleasure talking to you today, Ian, about this. Again, it’s been an insightful conversation, and for our listeners who want to learn more about the CCS Insight predictions, we will have that ready for you. You could dig into the edge AI and this network space also. So thank you again for joining us, and, until next time, this has been the IoT Chat.

The preceding transcript is provided to ensure accessibility and is intended to accurately capture an informal conversation. The transcript may contain improper uses of trademarked terms and as such should not be used for any other purposes. For more information, please see the Intel^® trademark information.

This transcript was edited by Erin Noble, copy editor.