So, is the correct way to look at this, then, that the private 5G market, when it talks about private 5G, is obsessed with a future where Volkswagen is animating its production lines using Release 18-level URLLC 5G networks? But that the wider enterprise market does not care about any of that? That this VW ideal is a drop in the ocean? “I believe so, yes.” And that actually, the future of private networks is not private networks – but a global network of radio networks with shared core infrastructure? “The future is going to be hybrid networks, absolutely.”
Some interviews are better than others; this is a snapshot, with truncated answers, of a conversation with Kathiravan Kandasamy, vice president of product management at US-based carrier services company Syniverse. The Syniverse proposition in the private networks market is not new; lots of self-styled IoT MVNOs, working in the same way, are promising private-to-public network roaming, mostly for mobility use cases in the supply chain. But Kandasamy’s offer sounds bigger, and more ambitious, and his answers about the wider market are candid.
So what about the argument that factories will buy and deploy cellular in the same way they deploy Wi-Fi? That it will not use centralised cloud infrastructure, often; it will be on-prem, and it will become such an easy box-sale that you can almost run a dealer channel to sell it to enterprises? “Yes, I know, that is the dream. But I also think a lot of people oversimplify the complexity of cellular,” says Kandasamy. These are big questions, about the digital transformation of more than just a few blue-riband industrial sectors, and the role of telecoms in all of it; and they are unflinching answers.
The full transcript, worth printing (in two parts), is available below. The Syniverse vision – which rivals what the likes of BICS, Telit, Eseye, and Twilio are doing with cellular IoT roaming – gets lost, arguably, in the noisy narrative around straight all-edge private 5G, being offered by big-brand equipment vendors and upstart network providers, in combination with system integrators and cloud providers. But its horizon is wider, also showing the (crucial but uncertain) place of national operators in sharper relief – to the point traditional Industry 4.0 looks like a niche pursuit.
It is worth a read; all the answers below are from Kandasamy.
…
Explain the Syniverse role in the telecoms ecosystem, and its role as a consequence in the private networking space, vis-a-vis carriers and enterprises.
“So, Syniverse started as a carrier to the carriers, if you like. They use Syniverse as a middle third-party hub whenever they need to transact – for signalling or data clearing, or whatever. So a lot of it is [about] IP exchange (IPX) services, which is like a private network for MNOs to connect and transact. A large part of our business is in the messaging space, as well. And we are selling a lot of these capabilities in IoT and private networks, especially, to regional tier-two and -three service providers and resellers.
“So we are almost like an MVNE in that space. We are an aggregator; we package connectivity with SIM management, and with policy and security, for smaller regional carriers, as well as enterprises and resellers. We provide a lot of services to companies in Asia Pacific; manufacturing in China and shipping globally – doing logistics, transportation, connected vehicles. They need connectivity when they ship these devices, and Syniverse is the one-throat-to-choke – so they don’t have to work with multiple operators, region by region, or country by country.
“One Syniverse SIM provides access everywhere. And we do eSIM as well, allowing for all types of API support – so all these transactions can be done in software. There is no physical exchange of IMSI profiles, meaning customers can order, provision, block, suspend, and apply data through our portal or through APIs. We are basically extending carrier features to enterprises.”
And the private networks piece…
“Yes, so our focus is to enable those use cases where devices leave the private network and latch onto a macro network and still have connectivity. The other angle is how to keep all of these IoT devices secure, and still have that control and management policy built on top. Which is our role; that is what we do. And the other angle is we give enterprises the chance of redundancy. We are working with some utility companies with private networks, for example, which want a fallback to a public network if, for whatever reason, their private network fails – which is what our SIM does. A lot of utilities, especially in the US, are targeted by bad actors, and so it offers a safety net.
“We don’t do devices. We don’t do radios, or access points, or core networks. We do have our own hosted core, so we can provide cloud-hosted core services, but a lot of the enterprises today want on-prem deployments. We provide interoperability between their private network and the public network – for mobility use cases. We rarely go direct to enterprises. We work with SIs because they tend to be the boots-on-the-ground in the private networks space. And we provide them with the SIM connectivity, a hosted cloud service if required, and all the SIM management. And we work more or less the same way with carriers, as well – to route traffic onto their macro networks. We are happy in the middle, between carriers and enterprises.”
So does the enterprise have control over those carrier preferences – over where you route their traffic, according to the kinds of performance guarantees and service level agreements their use cases demand?
“Yes. Syniverse aggregates all of that [roaming onto the public network] – so the enterprise only has to deal with Syniverse, which provides access into all of these other networks. So when your devices leave your network, they automatically get access to any of these others, and we can apply policies for whitelisting and blacklisting, and which network to access depending on signal strength or cost, and so on.
“There are so many ways to connect to this network and not to this one, for example, or to only cap data usage, and so on; there are so many parameters. But the most important point is that, if you have a hundred private networks, you don’t want to have to strike access deals for every device for every site for every carrier – you just want to connect to Syniverse, which does all that for you.”
And just explain the proposition for the carriers, as well – which have national and international roaming agreements of their own, which they could aggregate on behalf of the enterprise, as a single-throat-to-choke.
“Because I am now allowing traffic onto their networks which is not from their subscribers. I am selling the coverage – which is revenue-add for them. But [that negotiation] is already done. I already have all of these roaming agreements – with 700 networks in 200 countries. And for the enterprise, we are giving access to all of that from a single SIM – so instead of talking to 200 networks, or however many you need for your global tracking devices, or whatever, you have this one-throat-to-choke.”
Talk about the primary verticals and use cases for this private-to-public IoT roaming, and vice versa.
“It started with transportation and logistics – like airports, rail yards, shipping ports – and it has moved into education, into university campuses; and in the past six months, utilities have become very hot. When private networks started, we thought it would be billions and billions of IoT devices, but really in the past two years, it has been consumer smartphones and laptops that have primarily been driving private network growth. And we also see increased awareness and education for enterprises that their devices can still work when they leave the private network.”
So give me a couple of scenarios of IoT devices…
“Distribution centres and trucks.”
But that implies vehicles and pallets. Is that not the domain of smaller LPWA trackers, on licensed or unlicensed cellular? Is that a market for private networks, even?
“Well, it could be anything. We have worked on airport use cases where any of the fleet and baggage, any of the vehicles, close to the gate and tarmac, are covered by the private network, but when they move into a much wider area, they just lose coverage. The use case for the private network is always about these more advanced use cases they want, or just as a replacement for Wi-Fi. But within that context, there’s always a subset of use cases that require mobility. Not every device on a private network needs to travel or room; a lot of them will be stationary. But, as we have found as well, in critical industries like the utility space, there is great value that devices can latch onto a macro network in the event a private network fails.”
So in the short term, what are you most excited by?
“There are two things. The first is that, when a device leaves an enterprise, at some point, because of the radio signal, it will detach from the private network and scan and attach to a macro network. Now that transition is easy. But the other way around is much harder – for a device to detach from the macro network and roam back onto a private network. How do you get a device to recognize the private network and forget about the macro network?
“And we have developed some intellectual property, called Automated Network Reselection, so the device recognises it is in range of a private network and detaches from the macro and attaches back to the private network. And this is getting a lot of traction – because if the device is continuously attached to the macro network, you are actually paying a carrier money, and, when it is back in the private network, you want it to offload.
“From private to public is easy, because the signal fades and the device rescans. It’s not an issue. When it’s in reverse, coming back the other way, then it is more complex. And most private networks are built where there is overlapping macro coverage. So the device is always going to hold onto the last active profile. If it’s a UE, you can manually do the network reconnection; but if it’s an IoT device, that is not going to happen. Just because it doesn’t have a screen.”
And is that piece of IP something BICS and Transatel and 1NCE do not have?
“We don’t believe they have it. They have the ability to do the roaming part, but not the automatic network re-selection, when it comes back. For the IP today, all we need is the signalling; the data can stay where it is. The data can be home-routed; it can be locally broken out. It is only signalling, so it’s not a major latency concern.”
Is the plan that eventually to license that technology? Do you aim to build a lead in the market first?
“Today, we want to sell it like a software-as-a-service. And yes; the idea is definitely to build a lead in the market first. Absolutely.”
Right; and what is the second thing?
“So the second part is that we think private networks will need a simplified way to connect into carriers. So we are also developing what we are today loosely calling, Private Network Hub, where we can allow private networks to connect to carriers more easily – with Syniverse – instead of approaching every one of them separately. If you want AT&T and T-Mobile subscribers, say, to be able to roam onto and use your private network, you still need some sort of agreement with those carriers. We want to be the hub in the middle. We want to organise all of these agreements with carriers, to allow all of their subscribers to roam onto your private network – almost like a neutral host model.”
And the application for that is really handsets, right? Like a corporate privilege, afforded to staff – to say, look, I’ve got a new 5G network and it’s my network, but you can use it here, too?
“Correct. It will mainly be UEs – smartphones and laptops. These are use cases mainly in office buildings, university campuses, stadiums, and so forth.”
And when you mentioned earlier about the security of this private/public network roaming solution; just explain what you mean by that, and how you are delivering that?
“So private networks were initially built-out because companies wanted to keep everything within their control. So these island private networks emerged. But today, because of the applications for most of these use cases reside in the cloud, such as when there are five factories running off the same cloud. You want the best private network from a security perspective, but some data has to leave the private network – because everyone is hosting applications with cloud service providers. That is what we are addressing – with private UE IP pools, private APNs, everything. It’s an SD-WAN construct. We can route your traffic without ever touching the public internet. So there are a lot of things we can do to keep your data private even while it leaves your private network.”
This article is continued here…
