Season 1, Episode 4 – The Impacts of 5G and WiFi6 on the IoT

IoT Wavelength Podcast Season 1 Episode 4 – Impacts of 5G and WiFi6 on the Internet of Things with Stuart Strickland, Ph.D.

Vaughn:
You’re tuned into the IoT wavelength where we look in depth at the people in technologies that are stretching the boundaries of the Internet of Things and transforming the world we live in. I’m your host, Vaughn Amann, and I’d like to welcome you to this fourth episode in the IoT wavelengths podcast series.

Wireless networks have greatly expanded the market for IoT devices and enabled IoT applications in areas that wired networks never could. The 5G mobile network will open up new IoT applications with its higher speed, lower latency and network slicing capabilities. At the same time, the i56 standard along with the newly approved 6 gigahertz spectrum allocation will greatly increase Wi-Fi network speeds and reduce network congestion, in areas with a large number of devices. So what of those changes mean for network owners, operators, and users?

Well, my guest today will help answer some of those questions. He’s a Distinguished Technologist at Aruba Networks, a Hewlett-Packard Enterprise company, where he leads the technical development of Aruba’s 5G strategy. He drives initiatives on spectrum allocation, Wi-Fi cellular co-existence, edge computing, and heterogeneous network architectures. And he contributes to 3GPP. IEEE, and the Wireless Broadband Alliance. He earned his Bachelor’s degree from Columbia University, and his PhD from Harvard University.

So let me introduce my guest, Stuart Strickland. Welcome to the podcast, Stuart.

Stuart:
Thank you

Vaughn:
5G promises a lot of benefits in terms of increased speed and capacity, and reduced latency and that’s especially helpful in IoT devices. And there are a lot of applications that aren’t really IoT-enabled today that could be in the future with 5G. Applications in the automotive industry for instance, vehicle-to-vehicle communications and infrastructure, but there are a lot of manufacturing applications that could be enhanced through the use of 5G. So, are you seeing your customers inquiring about those kind of applications already?

Stuart:
Yes, and the two sorts of the applications that you just described, sort of fall nicely into two categories, one like the vehicle to vehicle communications are links that don’t exist at all today. So that’s a new opportunity. And the other category and manufacturing is one where in a lot of cases, they have traditionally been hard-wired because of their latency requirements, and we’ve already been seeing for some time a desire to move those to wireless connections and 5G is certainly going to be a part of that, although some are doing it with other technologies already already today.

Vaughn:
But it helps to have the entire infrastructure – basically everywhere you go is going to be 5G enabled, at some point in the future.

Stuart:
Yeah, I think that there’s a lot of momentum around rolling out 5G services. It’s going to mean different things in different areas. I think if we put the IoT applications in context, around 2015 when the requirements were first being drawn up for 5G, those of us were involved in that imagined a couple of major transformations in communications. One of them was certainly a proliferation of all sorts of different devices that don’t normally communicate with each other, or they communicate with each other over wires today. So IoT devices, not just handsets that people walk around caring in their hands. The other one was an undeniable increase in the appetite for data consumption that would have overwhelmed the existing cellular networks. And then there are specific expectations about ultra-low latency machine-to-machine communication that 5G is designed to address.

And there’s also changes in the network architecture – decoupling the radio technologies from the core network technologies because a lot of the devices that want to connect to a network might connect to one kind of network in one environment and another to another. So most of these devices are going to be multi-mode whether that means they do WiFi as well as 5G or whether they do 4G as well as 5G, they’re going to be moving between environments where they’ll have to use one technology or the other. Whatever is available in the environment. And the management of that requires a lot of architectural thought.

Vaughn:
Right. So the software-defined radios that are part of the 5G equation really do give you a lot more flexibility, but still you have to have specific antennas for certain bandwidths and things of that nature that are hardware-specific.

Stuart:
Yeah, I think it’s helpful to think about the roll out of 5G in a couple of layers. There is, of course, the new radio technology which has a lot of innovation in it, both in terms of moving into new spectrum where cellular hasn’t operated before. That has some characteristics we can get into. Also in terms of its use of more sophisticated MIMO intended technologies, but then there’s also the layer of the core architecture which you alluded to with software-defined functionality, and that’s partly driven by an effort to control costs. To use general purpose hardware rather than dedicated telecommunications hardware for the core network, but it’s also partly aimed at being more flexible, and in particular rather than offering the sort in a monolithic services that that cellular operator offers today to its subscribers, who all basically want the same thing to be able to text and watch videos and wander around, to being able to offer more finely tuned services to the enterprise. The technical term for that within the 5G specifications as network slicing. But it basically means being able, through software to define a part of the network resource pool, that’s specific to a particular application or a particular customer or user, and so that’s a second level and then the third level, and I won’t go further than three to the… We could go much deeper but a third level is around the use of edge computing that is putting resources closer to the place where is being consumed and generated. So rather than having to go back to some central data warehouse, to store or retrieve data you can get it locally on-premise. So there’s been a kind of proliferation of these micro data centers at the edge of the cellular Network, where they haven’t lived before in order to serve particularly latency sensitive requirements or privacy sensitive requirements.

Vaughn:
So that’s sort of the end game is having all these pieces in place. The very, the high frequency, low latency part of the 5G networks along with edge computing. And that really enables a lot of these really exciting applications like augmented reality and virtual reality, where we can see telemedicine being improved and first responder support. People having… in ambulances, having kits, where they can communicate directly with a doctor in real-time or physical therapy, remotely in rural areas or even remote surgery with haptic feedback is one application that’s really exciting. So your surgeon could be doing the surgery from across the world.

Stuart:
Yeah, and all of those are pieces of 5G. And in some cases, they will… It’s an evolving standard. And we can talk about the timelines in a moment, but some of those are pieces that will come together over time. Some of them are actually mutually exclusive use cases. So the use of 5G and millimeter wave to deliver very high bandwidth is not going to be the same flavor of 5G that you use to connect to the Internet while you’re driving down the highway, for the use in an industrial environment, with very low latency has some different hardware requirements which might not be cost-effective for an application that’s basically doing streaming of video monitoring of equipment. So when we look at the promises of 5G, it’s important to realize that the realization of those is going to adapt to the trade-offs of each of each particular application.

And so, the kind of 5G that you see blanket coverage of is going to be very different than the kind of 5G that you see on an assembly line or in an operating room. In order to meet the requirements of the specialized environments, there’s going to need to be a lot of local investment that doesn’t make sense to have spread out over white geographic areas.

Vaughn:
So, 5G what we’re seeing today is just the beginning of the evolution of several different flavors of networks I guess, that are merged together under the umbrella of 5G.

Stuart:
Yeah, I think that’s a good way to think about it. And it also highlights that while you’ll have different constraints under 5G radio depending on whether it’s operating millimeter wave or whether it’s operating in the LTE band, you also have some applications where devices are handing off from a WiFi coverage over a hospital campus to a 5G radio coverage in Millimeter-Wave within maybe robotic operating theater – sort of fantasizing there – in coverage outside of the campus once you leave. And one of the really key things to make 5G work as a kind of seamless experience is going to be the coordination of networks that allows devices to move between those areas without the user having to having to intervene or be aware of it.

Vaughn:
Well even today there’s the capability of seamlessly moving from a WiFi network into the 5G network on a telephone call, for instance.

Stuart:
Yeah, so what you’re referring to is WiFi calling, which has become a very popular thing. I think they’re probably now hundreds of operators in almost every country in the world to support WiFi calling. And the reason is that it’s a convenient and inexpensive way for them to fill coverage gaps. It’s a very positive experience when you’re able to get a call inside of the building that might not have gotten before. It’s still a bit kludgey in terms of the seamlessness of the hand over, but that’s one of the things that will also improve as the cellular networks upgrade to 5G, their visibility of local WiFi networks will become much better, or I should say a interfaces are opening up to allow local WiFi networks to share information that will enable the 5G core to just see them as an extension of the cellular network. And that’s going to be particularly important because if you think about gaps in coverage today, with LTE as you go into a highly efficient building, those are just going to get more acute in the context of 5G, because 5G on the whole will operate on higher frequencies that penetrate less well. It has other technical requirements that require a higher signal-to-noise ratio that are going to make it harder to get into buildings and the buildings are becoming more efficient, so we think the outside-in model where you basically have cell towers outside reaching indoors, is going to become more difficult in the future. And so you’re going to have to rely upon indoor networks and those indoor networks are going to be a mixture of WiFi and 5G. So being able … and it may be that the 5G part is actually deep inside the building. So it’s a bridge; that trajectory from wandering around outside to being well within the WiFi coverage area of an outside of cell coverage to being in the morsel of 5G goodness in a special place that will be necessary and also possible within the IT architecture.

Vaughn:
And you mentioned the availability of more bandwidth of higher frequency bandwidth it gives you the ability to have lower latency, and more concentration of devices in one specific area. And I guess the FCC has just recently opened up some bandwidth for 5G applications.

Stuart:
Yeah, so it’s not specific to 5G, but yes, you’re referring to the 6 gigahertz band, which there was just… We’ve been working on this now for a couple of years to work out the rules under which this would be possible… We’ve opened up about 1.2 GHz which is more than WiFi has today, in the 6GHz band for general use. That’s shared spectrum under some particular rules, but let me put that in context.

Right now 5G networks, if they want to roll out, they really need to reuse the spectrum that the cellular operators have today, and that tends to be pretty limited. It’s about 50 megahertz of downlink spectrum per operator per major region. And I’m going to refer to the bandwidth MHz because the amount of data that you can push through is really proportional to them.

The other place where 5G can operate today is in shared spectrum with WiFi in the 5 GHz band, which there’s maybe, depending upon where you are, maybe about 500 to 700 MHz of spectrum available. So basically 10 times as much spectrum. So you could push through more data, but that’s shared spectrum. So, 5G operating there has to co-exist with WiFi so if what you’re looking for in 5G is clean spectrum, like you’re used to getting from a cellular connection, you can get that in the LTE bands, but it has to be shared with LTE, so there’s not going to be that much, or you can do it in the shared 5GHz band, which is a lot more spectrum but it’s shared, so subject to interference and other things.

Or you can do it in the millimeter wave spectrum, where there’s huge amounts of spectrum available. We’re talking about maybe 400 MHz channels for 5G and millimeter wave, but it has some very peculiar properties. Once you get up to frequencies that high, and we’re talking 25 to maybe 60 GHz.
Once you get up to frequencies that high the signals don’t penetrate buildings at all; they barely penetrate bodies. So, that’s really sort of room-level coverage that you’re talking about, not campus-wide coverage.

So the networks have to become more dense. That’s why I was saying earlier that some of those deployments might be more expensive, but the exciting thing you just referred to is that back down near where WiFi lives today, in fact immediately adjacent to that in the 6 GHz band, which is used in some areas for point-to-point backhaul communications for cable signaling or for television signals to go back to the home stations, so there’s some restrictions on where that can be used. But a scheme has been now laid out by the FCC, and we expect that the first products to use that spectrum will probably be WiFi6 products towards the end of this year, but it’s open to any technology. So we expect there to be 5G radios operating under the same rules in that spectrum as well. And that, initially, this is a ruling of the FCC, so just in the United States, but similar efforts are under way with the CETC, the equivalent body in Europe and Asia and the Pacific, as well, so we expect that there will be … the bottleneck of spectrum is going to open up in the next year or two, and that conveniently, coincides with the time frame for the development and availability of 5G products.

Vaughn:
So, 5G technology has the capacity or will have the capability in the future of this very highly increased bandwidth available to users plus high concentration of end points but today it’s not there because we’re not able to use that spectrum yet and it it’s going to require things like micro-cells, many micro-cells in some areas to support that full benefit.

Stuart:
Yeah, so maybe it’s worth talking a little bit about the timeline, both in terms of how 5G as a concept or standard has developed, but also in terms of where we think products are going to come out and what people are going to actually experience. So I mentioned earlier that initial requirements were defined around to 2015 by the IMT. These requirements were really based around expectations or benchmarks of throughput and efficiency and latency, the number of clients a node should be able to support things like that.

The 3GPP, Third Generation Partnership project is the standards body that is responsible for figuring out how to realize those requirements, and they work on a release schedule. That’s the same body that has standardized LTE and before that, 3G. They work on a release schedule that has about a 15-month cadence. The first release that treated 5G was called Release 15, which was finished around the middle of 2019, or the middle of last year, and it defined the 5G radio and the basic architecture so that if you were a manufacturer, either of user equipment or network equipment, you could go out and make something that followed the the 5G vision, although it wasn’t fully-featured and it wouldn’t meet all of the performance requirements.

We’re just finishing, and I say “we” because I represent HPE at 3GPP standards, but we’re just finishing Release 16, the next one in that cycle. In fact, the work on it was really finished at the beginning of this year, and the final crossed Ts and dotted Is are going to complete by the middle of this year, and that defines … it’s the first release in which the products that come out of it are going to meet the IMT’s 5G performance standards in terms of throughput and latency, etcetera. It defines the ability of 5G radios to operate in unlicensed spectrum. It defines networks slicing and a couple of other key features. It leaves some things out because it’s a limited … the organization can only do so much at once. So what it left out are mostly these issues around mobility between radio technology, between WiFi and cellular. It left out a lot of IoT control and machine-to-machine and ultra low latency aspects and those are going to be part of Release 17. And we’re just starting the process of defining what’s going to be in Release 17, and that’s always a negotiating process. And those things that are left out will be in Release 18. But we’re expecting that to be finished as a specification towards the end of 2021. But in the meantime products are going to start coming to market, and the products that we expect to first see in the market are really internal to the operators – their wireless backhaul. So the connections from a building to the Internet rather than being carried over a cable will be carried by, in some cases, 5G radio signal to a cellar core network. And that’s partly motivated by the fact that that’s a closed end-to-end system where an individual operator can define what they want to be on the top of the building and what they want to be in the core network, and they don’t have to coordinate that with all of the handset manufacturers for example, or all the IoT equipment manufacturers. They can do it at their own time frame. But it’s partly also an early stage development because we do expect the public facing cellular towers or nodes to be much more densely deployed, and that means that those nodes themselves need to have some backhaul to the core network.

So investing in what the cellular industry calls fixed wireless access, that is where the nodes don’t move, but the communication between them is wireless. Using 5G for that is probably the very first thing that will happen. And for most people, it will be invisible.

The second thing that’s likely to happen, is the beginning of handset support for 5G. You already see this today. There’s a lot of debate over what actually counts is 5G and what an operator decides to label as 5G may not necessarily be with another operator considers to really be 5G, but in any case, at least nominal compliance with the standards to the point that 5G can appear as a symbol on your phone, that’s already happening now. And we expect it to become more widespread, but it will on the whole be done by carving out a little bit of the existing LTE spectrum, so because of the bandwidth constraints even though you’ll be able to see 5G as a consumer walking around town, you probably won’t experience much difference between the LTE signals and the 5G signals. They’re going to be determined mostly by how close you are to the base station, but over time, and I think the next round, once features are part of The 5G spec, that now are of interest to industry and enterprise we’ll start to see operators offering managed services to industrial and enterprise campuses, and that will really be the part of 5G that was both most interesting to the operators because this is a new market for them. It’s not just providing unlimited data plans to consumers it’s offering higher value services to enterprise, and also will be probably the most interesting phase of 5G from the point of view of innovation because now people who are in the position to figure out what cool things to do with it will finally be able to get their hands on something in their native creative environment. But that’s probably something we won’t see this year, but we’ll start to see in 2021.

Vaughn:
So for now, the carriers are the ones that are most affected with having to change their infrastructure and to do investments in the infrastructure to provide all of this capability in the near future, and quality of service is one benefit of 5G the great improvement in control over quality of service on your calls, and then the ability to do the network slicing gives you the ability to have virtual private networks that you can’t do now over the cellular network.

Yeah, so in in that is coming in the short run or in the medium run?

Stuart:
Probably more in the medium to longer term in terms of the slices of the networks that operators will offer up to a specific customer and there’s some question marks still around the scalability of that and whether operators will offer each individual customer that they have with their own bespoke service or whether or not they will have say a slice for hospitality, a slice for manufacturing, a slice for finance and banking, a slice for defense and those will have certain characteristics. But basically you get to choose your flavor, rather than getting from all of the operator services. I think it’s unlikely that an operator would intervene and do all the things that the IT staff of a major manufacturing facility does today even though a lot of the work that’s going into the transition from LTE 5G from the operator’s perspective, is trying to make the operator network look and behave a lot more like enterprises are used to their IT systems, behaving today that is make them more flexible and bespoke. There’s probably some middle ground that will have to be found there between how much customization and individual enterprise wants and how much appetite an operator has for supporting that. And then on the other side, how much an IT staff that’s used to switches and WiFi access points wants to learn about how to run a cellular network ’cause that’s a whole… That’s a whole area of expertise that’s outside of the typical domain of of most enterprises themselves, very well understood within the operators, but not well understood it enterprise. So there needs to be some sort of negotiation over the next few years about who needs what, who knows what, who has the resources to scale delivery.

Vaughn:
So you’d see probably rather than reserved bandwidth being supplied or being offered by the carriers that they would give you quality of service offerings?

Stuart:
So I think one of the things that’s going… So right now basically every consumer of a major US operator gets the same thing. Or you may have different plans that you pay for it differently, but basically everyone wants mobile broadband, and that’s what they get. And there may be some quality of service expectations about where it’s available and how much you get that those are kind of hard coded into the network, basically around the initial deployment and design of where the cell towers will put up.

I think what we are going to see in the future, how far this makes its way down to the consumer level, and how far it ends up being a proposition for the enterprise is up in the air.

But I think what we will see is operators being able to say, “look, for these individual devices, or this class of devices, or the devices that are on-premise on this site, we can guarantee them a certain bandwidth, all of them, and we will guarantee them a certain maximum latency and they will all get that” Or we know that this device say is an emergency stop button and this other device is a video camera. They have very different quality of service requirements — a stop button doesn’t care about bandwidth, but doesn’t wanna wait — a video camera doesn’t care how long it takes but needs gobs of bandwidth. So that kind of differentiation, I think we will see, and I’m sure there will be pricing models built around that.

Vaughn:
From your customers’ standpoint are there any concerns about the 5G roll out? I know security is a big concern on everybody’s mind.

Stuart:
So there are a lot of concerns and most of it, our customers or people who I end up speaking with mostly these are what we call enterprises, and that has a pretty wide range from the IT directors at university campuses to hotels and conference centers, casinos, so hospitality venues to industrial manufacturing customers, to federal customers, DOD-type customers to finance, and they all have different areas of anxiety, and concern. Mostly around uncertainty of timing and features of 5G networks, and uncertainty about whether they need to be thinking about this themselves or whether it’s something that operators are going to take care of in the macro network, the outside network world that they always had before. Does it affect their investment in their existing IT infrastructure, whether that’s wired or WiFi – how do they prepare for that? So the first level is what’s coming, what’s this wave and do I need to get out of the way or ride it?

And so for those one of the really good things about the 5G architecture is that it is designed to accommodate multiple radio technologies. So, basically if they continue on the path that they’re on now, the WiFi6 products that they are probably lining up to buy will interface with the 5G core network. So if their operator services that involve networks slicing, that can be delivered regardless of whether they have a 5G radio indoors or not. In terms of issues around security that large… So there are two levels to think about that. One is the air interface and encryption of data, and that’s become a pretty well-known problem domain both within WiFi or IEEE communities and within cellular or 3GPP communities, and the level of security and encryption is pretty much the same. And I think from that point of view there shouldn’t be a lot new to worry about. And the expectations that the market has shown are met in in both cases, but there are other areas of security around things like the visibility of data and control of data once it leaves the air and ends up in the network.

And there most enterprises are used to having data centers under their control or on their premise, and the coordination that will be required with operators and the visibility that operators will have of some of that data is an area that needs to be worked out. The second area is that right now there’s kind of a gap between the way that mobile network operators recognize and authenticate subscribers and the way that a typical enterprise gives credentials to a user and then decides what that individual user gets access to or what role they play in the local network.

So right now, basically, in the cellular world there is a SIM card and every subscriber gets a SIM card, and that uniquely identifies them and their subscription, and they get on to the network. And once they’re in the network, they’re basically just on the Internet. With an enterprise environment a certificate or user name and password will be given to an employee, but it may also be given to a printer or to a piece of machinery or to a sensor, and the IT department in an enterprise has relatively granular control over how they group different categories of users — what access they give, which ones can see each other, which ones can reach what resources. Part of the techie for mobile network operators and offering 5G services into enterprise will be to duplicate that level of control and flexibility so that that aspect of security, not just is the data encrypted over the air or did I let the right person onto the network? But once they get on to the network, what are they able to do? That’s an area that involves a lot of new work for the cellular community.

There’s a template for, and a pretty clear set of expectations in the enterprise.

Vaughn:
More roles-based authorization.

Stuart:
Yeah, and more control and flexibility over how those roles are granted, what’s the timeframe within which they’re valid, even where the user has to be in order for them to be valid, that’s something well understood and expected within the enterprise IT world. But it’s something that say LTE was not able to do. So that’s an expected innovation within the time frame of 5G for the cellular networks.

Vaughn:
So these are a lot of feature s that are envisioned, but haven’t been solidified yet, in terms of the network. So if I’m a CIO or a CTO, and I’m looking to future proof my purchases today, if I need to buy a network equipment, what kind of things should I be thinking about or what kind of equipment should I be leaning towards?

Stuart:
So, well, first of all, you should be thinking about what you need in terms of use cases and services that you want to provide, but if you think about the sort of three pillars of 5G at we talked about a bit earlier about increased bandwidth, about the variety of devices that would be on the network, and about low latency, there are a couple of concrete things that you can do today.

One is whenever the 5G radio and core networks become available, they’re going to realize their performance potential, they’re going to need local compute resources. So whatever you’re thinking about, in your current architecture, if it makes it possible to put the local data and computing resources on campus with what the cellular industry calls MAEC or multi access edge computing, or what HPE calls just edge computing, those sort of resources on premise are going to be usable both today and in the 5G world, that’s a core part of the architecture.

The second thing I would say is, and of course this needs to be thought of in terms of what their specific requirements are. But we have a new generation of WiFi coming out now. It’s available now, and those radio technologies will both be able to take advantage of the new 6 GHz band that’s opening up, and once 5G cores are in place we’ll be able to interface with the 5G core. So I think probably the least risky thing to do in the near term is to invest in those WiFi6 radios on the ground, knowing that they can interface to the 5G core and services when they become available. And we see a lot of operators even thinking about this is the strategy that there’ll be a 5G radio outside, but WiFi coverage inside and then maybe some pockets of 5G millimeter wave for specific things more deeply inside of the building. So I think that ends up being a pretty future-proof approach. As long as you make sure that you have basically you’re following the upgrade path that you’re probably already… Now, another aspect of that is that even if you were to decide in say two years time, that you want to completely cover your entire campus with some sort of private 5G service it’s quite unlikely that you would get rid of the WiFi because you’re going to have a lot of devices, whether that’s student laptops or whether that’s client tablets or whether that’s any number of devices that are already in your environments today that are still going to need to connect via WiFi. Not every device is going to have a 5G modem for some long-time, so that’s certainly not going to be a wasted investment. And it’s an environment that the 5G architecture invasions co-existing with 5G. So, my advice really, for someone who’s thinking about what they should be doing to prepare for 5G is to think not about the uncertainty around the technology but just think around what they need to accomplish in their enterprise and what tools are available today and are those… Do those tools have a roadmap that converges with 5G or are those tools a dead end?

So one area where it becomes acutely interesting is in the area around private networks. So I’ve been talking mostly about networks under operator control and moving from… where basically any subscriber can get on to it moving from a micro-network into an enterprise environment. But some enterprises choose to deploy their own private cellular network, maybe at an oil refinery or a nuclear processing, nuclear power plant. And in those contexts it’s a cellular technology, but it’s completely isolated from the the public cellular network. For people who are thinking about that they have some harder decisions to make because there is some early 5G equipment that they could use to start down that path, but there are some risks involved in whether or not they’re going to have to replace or upgrade that equipment along the way in order to keep up with the technology. Most of the customers that I’m seeing who have a need that is very specific to a private cellular network are pretty much doing that with LTE now and are basically saying, look, I know why I need a private network, I want control over it, I want clean spectrum, I think I can get a deal with operators who own that spectrum to make it available to me, but I’m going to use tried and true technology of LTE.

I’ll invest in that knowing that I will have to rip and replace it if I upgrade to 5G, because what I’m looking for there is not necessarily a performance gain. What I’m looking for is control, and the sort of balance between control and performances. It’s something that every enterprise grapples with.

Vaughn:
So for the vast majority of customers that are heavily reliant on WiFi today for those kind of features going with WiFi 6 compliant equipment at this point for the local area, and then relying on the 5G network food connectivity to the wide area later would be a good road map.

Stuart:
I think so, and they have to have some faith in the equipment vendors and the operators for the moment, because we haven’t worked through all these problems of mobility between those two networks, but there’s a lot of motivation among both of those communities to make that work.

So I think that in the long-term that’s going to be the most effective model for most enterprise customers who don’t have an acute need because of the control issues I mentioned earlier, to deploy their own private network right away. But if they just want to remain at the leading edge of technology and continue to provide whatever is the highest currently available level of service they can with a pretty high degree of confidence, invest in WiFi6 for their indoor environments and keep in contact with the operators about what their roadmaps are for the outdoor environments and what their roadmaps are for interfacing their core network to provide 5G services over WiFi indoors.

Vaughn:
So you see this symbIoTic relationship with 56 and 5G going on into the near future. At least I guess the important thing is to pick vendors that are working closely with the carriers and in the standards bodies that will be able to be flexible as the 5G roll outs are coming through.

Stuart:
Yeah, and not to turn this into an advertisement for our own company, but that’s something that we do. But it’s something that most of the leading vendors in this area are very active in. Whenever you’re working in the standards bodies the conditions or success of success or that an industry as a whole, ends up going on. We don’t want a proliferation of 20 different standards because they have to be 20 different vendors. We’ve been through that sort of thing before with DVDs and video tape etcetera. But the WiFi community and the cellular community, although they sometimes have been at odds with each other, are pretty well-aligned right now on what we think the future roadmap is, and the complementarity of WiFi and cellular networks, and the relationship between public and private networks.

And I think the positive thing from the enterprise point of view should be that really for the first time that I can remember, and I’ve been doing this since 20 some years, this is the first time that I see the cellular operator community really thinking seriously about what are the requirements of enterprise IT. And I think that’s a good sign. And there’s a lot of expertise around that in the IEEE or WiFi or 6 networking community, and I think there’s a lot of positive lessons being learned within the cellular community, about how to do that.

So I’m actually really quite optimistic about the interaction between private and public networks, and between licensed and unlicensed cellular and WiFi technologies going forward. And I think that, although there’s a lot of talk around the disruptive-ness of new technologies, I think that there’s a pretty clear roadmap of what people can do now to make themselves ready for future services when they become available.

Vaughn:
And in terms of new equipment, I guess if you do purchase something today, make sure that you get something that that has the software-defined radios that has the capability of moving with the standards as they’re being developed, and increasing your flexibility in the future.

Stuart:
Yeah, and I think most products that are available today are following a similar roadmap to the one that we are following where we are providing an option to control local networks remotely. Because we have a lot of customers who maybe they have a kind of franchise relationship between the central headquarters and lots of local branches, and being able to put the control and management of a network in some central location, but being able to put the resources, the computing resources and the data resources locally, that’s becoming a much more common architecture in the enterprise WiFi, not just from among HPE’s customers, but in the industry as a whole, and that’s an architecture that’s very much aligned with the direction that the 5G networks want to go, where control plane as we call it in the 3GPP world is off-site but user plane where the data is all dealt with invisible on-site. That’s a common model and one that the products you can find today support.

Vaughn:
So for enterprise customers, they could expect even more control remotely, from a centralized control center, of the entire network.

Stuart:
Right. Without having to send the data itself off to some remote location. Sometimes that’s necessary, but often it’s not. And we were talking about a little bit earlier for especially these new low-latency requirements. The further you have to send data before it’s processed and you get a response, the higher the latency is going to be, so more limited some use cases will be. In this architecture where there are edge compute resources on-premise but cloud-based management of systems which allows you to have a coherent network throughout your whole organization, those are more easily realized now and aligned with the 5G architecture.

Vaughn:
That centralized control becomes even more important as they’re pushing more compute out towards the edge.

Stuart:
Yeah, and presumably if you have a lot of local branches to manage you want that to be a consistent experience, so you don’t want it to be configured. If you have 5000 sites you don’t wanna be the network to be configured 5000 different ways. So having some central control is useful.

Vaughn:
Now your data center is going to be spread all over creation rather than in one central area.

Stuart:
Yeah, and that might sound a little bit scary, but it has some advantages. And it’s an architecture that the operators are moving towards today. It’s really sort of shocking how few of what the operators call packet gateways there are, but although there are many, many cell towers, all the data in the LTE networks is going back to a relatively few places before it gets deposited onto the Internet. Which has a lot of efficiency issues. And one of the things that we’re seeing, even within the operator’s deployment, is a push of data management to the edge. So that it might even be at some point that the cell tower that you talk to has a packet gateway to put your data directly onto the Internet there rather than having to go back to the core network where all the control plane operation of the network is going on.

So it’s a shared vision of moving data and moving processing resources closer to the end user, and that’s being done for the same reasons. And so is not surprising that it’s being done in similar ways by the traditional IT architectures and by the mobile network architectures.

Vaughn:
Well, this has been great information, really appreciate you sharing with us today. And if people want to know more about this over there some resources that you can point us to.

Stuart:
Yeah, absolutely, so it really depends how deep in the weeds they want to get and how much they want to know. So I would say that for those who really have an appetite for detail, the places to go are the 3GPP, that’s the Third Generation Partnership project that’s 3GPP.org.

But if you are coming at this more from a commercial perspective or maybe you just want an overview and don’t want to read 5000 pages of technical specifications on radio interfaces, the places I would go are the Wireless Broadband Alliance. This is an organization, it’s mostly driven by mobile network operators, but it deals with II and others as well, and they’ve done a lot of white papers on the implications of 5G. They are at Wireless Broadband Alliance, and the W in the B are not spelled out. So WB, Alliance all one word, no hyphens, or anything dot com.

And they’ve done in particular, a couple of white papers in coordination with an organization called the NGMN, which is sort of similar to the Wireless Broadband Alliance, but more from a licensed technology perspective. That’s NGMN.org, and they have some really interesting material on their website about use cases for 5G.

And then finally another good resource for knowing what the… In some cases, this is sort of a rosy picture, but what the future looks like for the cellular industry, and 5G – GSMA, the GSM Association which is GSMA dot com. There’s also a good source and I think if you were to spend half a day browsing through the material on those sites, you come away with pretty solid foundation in both the expected business models and the technology road maps of 5G.

Vaughn:
Great, okay, well, we’ll put those links on the IoT wavelength website as well, so that our listeners can follow those. And again, thank you very much for sharing all of your knowledge and your insights into the future of 5G, and we really appreciate you being here today.

Stuart:
Well, thanks very much for having me, it’s been my pleasure.

Vaughn:
My guest today has been Stuart Strickland, Distinguished Technologist with Aruba Networks and the leader of Aruba’s technical 5G strategy. If you’d like to get more information about the resources we discussed or see the transcript of this episode, visit our website at http://iotwavelength.com/podcast and thanks for listening.