angle-left null Episode 2: 5G: Big Change Is In the Air
Double Take podcast

Episode 2: 5G: Big Change Is In the Air

00:43:39
Double Take podcast Audio Thematic Equity
January 2020
Episode 2: 5G: Big Change Is In the Air

Telecom carriers are touting much-hyped 5G networks, which promise blazing speeds and massive bandwidth. But it will take many years, an enormous build-out of infrastructure, and a lot more radio spectrum real estate before it changes how business gets done. In our second “Double Take” podcast, Mellon’s Head of Global Equity Research Matt Griffin and outside wireless expert and patent holder John Dooley demystify 5G from both an investment and technology perspective.

Transcript

Rafe Lewis: Welcome to Double Take. I'm your co-host, Rafe, Director of Investigative Research here at Mellon.

Jack Encarnacao: And I'm your other co-host and Investigative Researcher, Jack Encarnacao. You've seen the endless advertising from mobile phone carriers touting it. You've seen endless media stories breathlessly fawning over it, and you've grumbled and begged for it as you tried in vain to access your email in a subway tunnel. We're talking about 5G. This episode of Double Take will attempt to help you separate the hype from reality, helping our listeners understand what's evolutionary and what's revolutionary in this next generation of mobile communications.

Rafe: To dissect and discuss this great topic, we have two guests, Mellon's, Matt Griffin and John Dooley, managing director of Jarvinian Advisors and a pioneer in using satellite wireless spectrum to address global holes and wifi coverage. Up first, Matt Griffin. Matt is Mellon's director of equity research, a tech guru who covers the semiconductor technology hardware and telecom media industries and now the lead portfolio manager of Mellon's relatively new internet of things strategy. Which is a concentrated equity portfolio invested in the firms that will build and benefit from the proliferation and eventually ubiquity of IOT. Matt previously worked as a tech analyst at a long short hedge fund. He was a portfolio manager at Putnam investments and held several other positions prior to that too. He holds a BA in economics from Duke University. Matt, welcome to Double-Take.

Matt Griffin: Thanks for having me.

Jack: It's great to have you Matt. So high level, separate the signal from the noise here on 5G. How transformational do you think it will truly be?

Matt: I think it depends on the context, so I would start there first, you have to think about what the applications are going to be. A part of the issue of thinking of 5G is that it's optimized for IOT and machine to machine communication. That's something that's hard to imagine and picture what that ultimately looks like. What we're used to is enhanced mobile broadband that was the real killer app for 4G. That's what created companies like Facebook and Uber and Google or made them extremely wealthy. Really the scale businesses riding on this enhanced mobile broadband platform, otherwise known as 4G. What 5G does is take us in a similar direction for consumer. In other words, more speed, more bandwidth, lower latency. So the cell phone experience or I should say smartphone experience will continue to progress, but that's not what this is about. This is about machine to machine communication on a massive scale. So you can't talk about 5G without talking about the internet of things. It's really the critical communication layer for the IOT and it is really designed for the IOT.

Matt: So I think as we look forward there's a lot of things that we can't even see yet, because it really becomes about the internet of things, machine to machine, and not about the consumer experiences that we try to put these into context.

Rafe: Got it. But one of the questions that comes up a lot is why wasn't this possible with high speed broadband? Right. I have a factory full of fixed equipment, I've got ethernet cables everywhere. The highest possible speed connections you can imagine. Why is this going to change my life as a factory owner?

Matt: Well, there's a few things there. First is mobility. Sure you want, you can wire everything but that's a lot of wire and it doesn't give you much flexibility. So 5G is obviously wireless. It's also a standard platform. What you have right now, if you look at automated factory for example, is a lot of different connectivity protocols. You may have ethernet, you may have wifi, you may have all these different other ZigBee, Z-Wave, things like that. What you have to look at for 5G is it's one platform, ubiquitous and more important it's optimized to address every type of application, connectivity application there is. So it could be a one slice of the network can be designed for very high bandwidth, low latency type of applications. Another slice of the network could be designed for very low power, long distance, but very low bandwidth and everything in between. And so the ability for 5G to slice and dice a network like that now serves a purpose of all those different protocols that you mentioned in the past.

Rafe: And just to clarify, it sounds like the slicing and dicing is a key feature here. Was that not possible with LTE and the fourth generation of wireless?

Matt: Correct. 4G really was designed for one thing which was dumping massive amounts of data to a smartphone. So think about video apps, all the things that we know in the smartphone, 5G is designed for these different types of services. The LTE network was designed in mind for consumer applications where you were paying $40 a month for that unlimited connectivity. The 5G network will be designed to enable that application as well, but you'll have more bandwidth cost per PIP be lower. So the price will be lower for consumers. But there's all kinds of other applications where you may not need that amount of bandwidth, you may not need all that and you don't want to pay for it. So if you think about a sensor network where you have millions of sensors out there gathering data wirelessly, you don't want to pay a lot, you need to pay pennies per sensor for connectivity.

Matt: You can't do that in 4G or there are some protocols that are late stage 4G where they're trying to do some of that, but 5G efficiently does this. So that you can tier the network to enable all of these new types of applications and services that are very different than enhanced mobile broadband.

Jack: You talk Matt about companies that sprouted into existence because of 4G LTE, like Uber and I know you said we can't necessarily see right now what use cases 5G will enable, but we can make out the edges, right, of some things that wouldn't have been possible or actually aren't possible as we sit here that might be possible in two to three years. You mentioned machine to machine communication. Can you make that more real maybe for our listeners, what do you think of, what's your mental picture when you talk about machine to machine in 5G?

Matt: Sure. I think the important context here is the nature of the businesses that we saw in 4G and what they will look like in 5G. So 4G LTE, enhanced mobile broadband enabled massive scale businesses, advertising based businesses like Facebook or Google, video-based businesses like Netflix, subscription based, apps, all these things. And there were few winners, you've had massive value creation across big platform companies. Scale is the key to those businesses, first mover advantage, getting user scale. If you think about the internet of things and not the consumer internet, which is really what we were just talking about, but the internet of things and overlay 5G on that, it's very fragmented. It's not going to be a winner take all in 5G like it was and consumer 4G, because there are different domains of expertise.

Matt: Factory automation is IOT application that will ride on 5G that really has nothing to do with the consumer internet though those companies will not be the same. And so what is interesting to me is that this fragmentation creates actually a larger number of smaller winners, if that makes sense. So rather than a small handful of really big winners like we saw with the consumer internet, you end up with a large number of smaller winners. So that's exciting because it really lends itself to us mid cap approach and there will be a lot of players that create incredibly strong businesses around the IOT with very strong moats and competitive advantage. But they won't be trillion dollar companies.

Jack: In the calls Rafe and I do, we often, of course talk about 5G particularly any tech name or any tech firm is seeing that as their vector of growth. And one of the things I'm thinking about as you're talking about industrial IOT and 5G was one point of contact who is outlining for me a vision where, let's talk about like water meters, okay in your house that measure how much water you're using in real time. That can only be such an intensive data flow right now as far as every second of water usage can be sent back to the billing agency for example, or the municipality to figure out what's happening in your house and then apportion resources accordingly. So am I thinking of this right in that 5G would allow such a more intensive exchange of these data points and connectivity to be so much smaller and tactile. That things like that will be just so much more powerful that the intelligence of the machines all around us will increase tenfold.

Matt: Yes, IOT exists today and has existed for several years and that's without 5G. So as I said, there are other communication protocols optimized for IOT, low power things like ZigBee, Z wave, those are wifi variants, wifi obviously and then 4G. But 5G comes at it in all those areas with more sophistication but also lower latency and higher bandwidth. But some of that also depends on the spectrum. 5G will allow you to use newer spectrum or high frequency spectrum that was not usable before where there does exist a lot of bandwidth. So 5G really does open up many more applications than all the current communication protocols we have combined. 5G will do that and more all on one platform. So if you think about the IOT and what we're trying to do here, it's really automate many things in the background. So the smart metering example, or smart grit or think about environmental and social goals, how do we reduce emissions and achieve all these things we want to achieve.

Matt: IOT technologies are really central to that efficiency, better, more efficient transportation, less congestion, all these things, better power generation, more efficient power transmission. Same thing in terms of utility. So that is the promise, is better environment, better social good, and it's IOT and 5G combined that will enable a lot of these things.

Rafe: So Matt, it sounds like this is a much more democratized generation of communications because you're going to have a lot of winners and not just a concentrated few mega platforms that win here. But I wonder where you see risks for legacy business models in the new generation of communications here?

Matt: Sure. It's a little hard to envision that, but I guess one example would be think of the current smart home setup today. The way your smart home works is you likely have a cable coming into the house, a cable modem, a wifi router, and that's running all of your TVs and smart home devices, smart speakers, et cetera. And that is using basically radio frequency over co-ax cable and then you're using wifi for the end point connection. If you think about what 5G can do, it can bypass all those hops. So one protocol I.e 5G can go right from say the media broadcaster, directly to a TV that has 5G connectivity natively and it bypasses that cable wifi, cable modem connection. So you can see right there that 5G could be a threat for the cable broadband business, because the bandwidth will be similar, but it will be able to be deployed device by device.

Matt: And I could envision device manufacturers or service providers partnering with carriers to say embed conductivities for a certain number of years, sell a TV that has five years of 5G connectivity, things like that.

Rafe: No wifi router needed?

Matt: Exactly. So all that Kluge of wifi and all this stuff and it's expensive as well. So that's really the promise and there will be disruption around that.

Jack: Great. You mentioned spectrum a couple of times. I'm going to ask you a question about that because as you probably aware, one of the limiting factors we hear about on 5G is just how ubiquitous these base stations are going to have to be all over metropolitan areas. Just so much more of an intensive demand to make sure 5G is accessible and widespread. So at the same time it seems like spectrum is a way around that that there's this maybe misunderstanding of how far a 5G signal can stretch relative to 4G. Could you lay out your thoughts on that for us?

Matt: Sure. There's a lot of confusion I find around people thinking about spectrum and protocols. So 5G, 4G, these are the protocols riding on the physical spectrum, but the characteristics of the spectrum vary based on the bandwidth, I'm sorry, based on the frequency. So very low frequency, what you would typically think of as FM radio or UHF, old TV, 600, 700 megahertz, 800 megahertz. Those signals are long wavelengths and they travel over long distances. So they're ideal for radio broadcast, TV broadcast, and the early days of cellular where you needed broad coverage, coverage was really the goal in cellular for the initial growth phase of the industry. But there's not a lot of that spectrum available, so that's all been allocated it's hard for a carrier, carriers can't get more than 10 to 20 megahertz slices. Okay. If you think about millimeter wave spectrum, which is much higher frequency, so I'm talking about 600 megahertz versus say 2.8 gigahertz, right? So really microwaves. That spectrum has not really been usable except for maybe things like satellite communications in the past, 5G makes it usable.

Matt: Why that's important is because there is not crowding in that spectrum and there's much more available spectrum. So instead of getting maybe 10 to 20 megahertz, now you can get a full gigahertz. Well that's massive bandwidth. So for very high bandwidth applications, that's a great spectrum area to use. However, it has a lot of fallbacks which primarily is it needs generally line of sight and can only travel short distances. Because these are very short wavelengths disrupted by glass, buildings, leaves, foliage, rain, whatever. So Verizon has been ramping this up, millimeter wave for broadband replacement, they've had pretty good success getting about a kilometer. So that's pretty good. But again it limits the distance you can offer that service over millimeter wave. So where millimeter wave will play is more dense urban areas, where you can divide up that spectrum into smaller pieces with short reach.

Matt: Think about fiber, a fiber run down a center of a street in an urban setting and then that fiber is driving streetlights that have small cells on them that are basically small 5G base stations that are only powered to do a such a small area, so that you're not overlapping with other ones. But in doing so, you can divide up the spectrum and get very large bandwidth, high speed in these dense urban areas. Or think of a stadium for example, places where there's a lot of people and using a lot of bandwidth at once. If you think about rural 5G, that's a different use case, different spectrum. So T-Mobile is rolling out nationwide 5G on 600 megahertz, right? That traveled a long distance. Not a lot of bandwidth though. So what they'll be able to offer for broadband replacement is not going to be the several gigabytes that Verizon can do in millimeter wave, but maybe it's more like 500 megabytes or 300 megabytes. However-

Jack: So slower downloads or whatever you want to-

Matt: Correct.

Jack: But it reaches further than we think of 5G being able to reach.

Matt: Correct. And the important point there is even at those lower speeds, that's still a viable broadband replacement. And so again, the application there could be driving more broadband into rural areas, which right now is very expensive to do with traditional DSL or cable modem.

Rafe: It's funny because what I'm hearing from you I guess is that this is revolutionary on the kind of commercial and maybe the public sector end of the economy. But for consumers it's more incremental and evolutionary, which makes me wonder, why am I seeing so many ads as a consumer from wireless carriers about 5G? When A, it's not really there yet, and B, it's not going to change my life much.

Matt: Yes. I think like any, the early stage of any technology transition, we apply the past toward the future in terms of how we contextualize everything. And so yes you're thinking about, well, my phone on 4G already does everything I need it to do, why do I need this 5G thing? Well, the carriers are spending the money and they want to monetize it. So what you'll see is what we've always seen in the past is that as the technology begins to mature and the ecosystem around it begins to grow and mature, the new applications start to flood in. No one really accurately predicted wealth creation or who the winners were. So I think there's always a little bit of opaque view there at the beginning stages. But as that ecosystem matures, even on the consumer side, there will be a lot of new applications.

Matt: I'll give you an example. Next year Apple's iPhone is rumored to have what's called rear facing 3D sensing. So currently the face ID function on the front of the phone reads your face, that's called face ID, that function is actually 3D sensing uses Vixa lasers. Next year, it's likely that Apple puts that on the back of the phone. And why that is important is because now the phone already has a visual representation of the world, but now it will actually have a 3D vision of the real world. And then when you combine those two things, think about the type of application you could have in say a video game. Where you can be immersed in the real world with a virtual overlay because the phone knows what that real world both looks like and how it's measured, and where distances are, and where you are within it.

Matt: So now that phone has a three dimensional vision of where it is in the world. So think about immersive AR, VR type experiences just as an example.

Rafe: And you need a lot of connectivity for that.

Matt: You need a ton of bandwidth and more important low latency. Latency is a critical development in 5G, and I can't emphasize it enough, 4G latency is call it around 150 milliseconds depending on how you measure it, who you talk to. 5G in theory is going to be close to zero, latency in 4G doesn't really matter because all that means is maybe you're buffering a video or the app is a little slower. You don't really notice it. When you're talking about real world mission critical applications like cars or vehicles talking to each other and trying not to hit each other. That should happen in real time, low latency or else problems happen. But even in consumer applications, something like that, highly immersive video experience, a video game, video games latency is bad, right? You play a video game remotely and there's latency, you're very upset. So again, there's a consumer application where latency and bandwidth comes together. And so there's more stuff on the consumer side to come that's going to be very exciting, we just can't quite see it yet because the platform doesn't exist.

Matt: But as you said, I think the real promise of 5G is the commercial and business and IOT side, which really it's machine to machine communication on a massive scale. Trillions of connections versus billions. There's something like 130 devices being connected to the IOT every second right now. And when 5G is out there, that's going to even explode further. So it's really that machine to machine at massive scale that is the promise of 5G and IOT.

Jack: Trillions instead of billions, that's the take home point. Matt thanks so much for joining us on Double Take, It was really enjoyable.

Matt: Thanks for having me.

Rafe: Okay, Double Take listeners, we've heard from Matt Griffin about how 5G will revolutionize the worldwide economy well beyond the consumer experience of sharing memes and streaming video content. But now let's dive a lot more deeply into how 5G will ride the invisible radio wave spectrum all around us. Jack, who's our outside view coming from?

Jack: It's coming from John Dooley. John is the founder and a managing director of Jarvinian Spectrum Opportunity Fund and a recognized expert on radio frequency spectrum. Prior to founding Jarvinian, John established Jarvinian Spectrum Advisory practice. Through Jarvinian Advisors, John has developed regulatory and technological solutions for the reformation of previously unusable radio spectrum. John also helped create TLPS, a private WiFi band made possible with satellite spectrum. He's also the developer of a regulatory and engineering solution for TerraStar, which has opened up unanticipated new 4G spectrum and L-band. He has also created advanced nano materials that enable next generation wireless filters and antennas. John holds numerous patents in wireless and intelligent computing technologies. John, welcome.

John Dooley: Thank you.

Rafe: I don't know if you have the requisite experience here. Come on. We're talking about spectrum and all you've done for the last, what, decade is spectrum?
John: Yeah. I think spectrum is this very new physical resource which we have been using for about a century. But it's become of unique economic avoidance in just the last 20 years or so.

Rafe: It's kind of inherently abstract, right? It's this invisible stuff that's all around us, but let's try to make it tangible. I think maybe one of the best ways to do that is just cut to the chase and let's try to understand exactly how mobile carriers will be able to jam exponentially more data through the 5G pipe than they can today with 4G and LTE networks. Where is all this radio wave real estate coming from?

John: Yeah, I think the difficulty with a 5G and one of the great opportunities with 5G is we don't have today a means of putting meaningfully more information in the same size channel, the existing radio-frequency footprint. That capability was really exhausted from a physical perspective in the 4G period. So in 5G we really have only two options. One, to continue densifying networks, basically making cells smaller and smaller and smaller, recycling the available radio frequencies spatially. And that obviously presents certain logistical challenges. And then we just need more real estate. We need more radio-frequency spectrum upon which to deliver services.

John: And one of the challenges we have is that radio frequency spectrum is finite. There is only a set amount of it that's physically usable. And then have that set amount, there's only a fraction of that, a very small fraction really, that can be allocated for commercial network purposes. Things that major carriers can use to deliver consumer broadband services. The rest is aviation and public safety and broadcast, that sort of thing.

John: So one of the challenges in 5G is if we were to really make use of the new standards that support fifth generation, we need very broad channels. And in the parts of the radio frequency spectrum that we've traditionally operated, we don't have very large channels. And so we're forced to move to increasingly short wavelengths, increasingly high frequencies. And that comes with a very great set of engineering challenges, many of which we don't have particularly good answers for yet.

Jack: Very interesting. So correct me if I'm wrong, but won't this take a quarter a century or more for the mobile carriers to build out truly nationwide 5G networks because of the limitations you've been talking about and the short distances that the waves can travel?

John: We certainly hope not. I think from a standards perspective, we're already looking at what 6G is going to be.

Jack: Wow.

John: Let's define what we're talking about. In the better part of a year or a little bit more, your phone is going to have access to 5G capability on its existing spectrum. Even this is starting for many carriers already. And so many functions are 5G related to latency and band combinations and security, they're coming.

John: But I think what we're talking about is when will I have these massive channels that can give me multiple hundred megabit or gigabit speeds? And carriers, some of the two largest, have that spectrum today. But it exists at radio frequencies that don't go very far. So if we could take a step back and look at this from a physical perspective, if we remember in high school, the chart of the electromagnetic spectrum, electromagnetic fields, it's all the same phenomena, whether we're talking about a radio wave or an infrared wave or a visible light wave or a gamma ray from space or radioactive material. It's all the same thing. It varies only an energy level and frequency.

John: And so radio waves operate at the very lowest end of that spectrum, but they're not all the same. Within that small, tiny little portion of the overall electromagnetic spectrum, radio waves of low frequencies move around environments very, very differently than radio waves of high frequencies. And it's because of the energy associated with the photon, the energy associated with the radio wave itself.

John: Higher frequencies have higher energies, which means they want to interact with matter. So many of the lowest frequencies that we've used since the 1980s, the world to them is transparent. When they interact with your body or with the air, the atmosphere, building structures, it's almost like it's not there. And so you can cover very large areas with them very easily. This becomes a challenge, however, because when we want to make capacity, we don't want to cover very large areas. We want to reuse those frequencies over and over again. And so what we've traditionally considered to be the highest frequencies that we use, around 2.5 gigahertz, for example, are very well suited to the kind of late 4G, early 5G network topologies that we've evolved.

John: Today, however, some of the biggest supply of radio frequencies is at 24 gigahertz and above. Some of the earliest systems being developed today are at 28 gigahertz. So we're going 10 times the frequency. 10 times the frequency is a lot more than 10 times the difficulty, because at those frequencies, the radio wave wants to interact with everything. It wants to interact with your body. It wants to interact with the building structures around you. It wants to interact with the atmosphere in between, water molecules, oxygen molecules, that sort of thing. And so it doesn't go very far. And it's not even quite a line of sight type operation. So the amount of infrastructure you need begins to grow exponentially.

John: Now, the reason we're tackling this great challenge is because a single channel at those frequencies can be larger than all of the carrier holdings at the conventional frequencies we use today put together. And that opens up the possibility of doing things that we've never contemplated before with wireless, things that look a lot more like what you can only do in a fiber network.

Rafe: But John, to Jack's question though, it sounds like you're going to need a ton of antennas, a ton of back haul, you name it. The amount of equipment that has to get deployed to make it possible for me to use my phone or to have industrial IOT or you name it, where there are buildings, where there is precipitation. Is this not a massive, massive multi-year effort?

John: Yes. And the sites are of a different character. So if we're looking at traditionally what conventional tower companies have offered, which are these big monopoles you see on the side of the road that cover many kilometers in any given direction, that's not what this is. This is the type of thing that's starting on the street corner. This is the sort of thing that they'd like to move within the enterprise and within the home. And there are reasons why this can work in a practical way.

John: Most of your consumption is actually not when you're truly mobile, at least not today. Most of your consumption's when you're sedentary. You're at your office, you're at your home, you're in predictable environments where there's peak demand and that's the first place you want to do this. And indeed the major carriers that have this spectrum, that is the first place that they are doing it. But even in those environments you have to understand that when you're holding a device at 28 gigahertz, for example, most of the energy is being absorbed by the soft tissues in your hand converted to heat long before it ever gets into the environment. And that that challenge remains. And one of the reasons that, beyond our slowly cooking ourselves, one of the things that we're looking at doing is using lower frequencies for the uplink, for example. You can overcome some of the physics with raw energy.

John: So basically if your frequency is 10 to 100 times less readily propagating, use 10 or 100 times more power. And because the wavelength is so short, we actually can build fairly small antennas that act like giant antennas at lower frequencies. That's practical. Your phone can't fall a suit. Your phone is limited in its form factor, what it can do at these frequencies. And even if it could emit large amounts of effective radiated power, you wouldn't want it to. It would burn you. So we have this sort of hybrid network approach we're taking.

John: But I think in any, without getting too deep in the technical details, it is a network build of a very different character. It's much more viral, it's much denser. And it's going to take a lot longer I think to become mature.

Jack: John, when I think about what you said in terms of the most intense peak uses being in sedentary environments, where you're staying relatively stale and how that can help this infrastructure challenge ahead of us, one of the kind of long range visions is when cars are talking to cars using 5G connectivity and they're very much moving. The whole notion is that they're moving and keeping in touch with each other to prevent safety incidents or a million other applications you could think of. So how could you help us think about how close we are to that? To I guess for lack of a better term, 5G that's on the move as opposed to 5G that's static?

John: Well, in terms of lower frequency access, we're very close. That doesn't buy you a lot in terms of throughput, but it does buy you a lot in terms of reliability and latency. And that's something that automotive applications or any number of industrial applications that have never been compatible with wireless. Wireless has always been, in many use cases, in many industrial applications, peripheral. It's nice to have, but you can't rely on it for something that someone's life depends upon it. You cannot rely upon it for something that is requiring millisecond response times. The standard now supports that. And so that part of the connected car experience I think is covered by what we already have in the inventory and have built out in terms of conventional network models.

John: In terms of giving the car multi gigabit capability to talk to other cars, we're not there yet. And the predictability of where roads are make it somewhat easier, but it's not something ... I was listening to an automotive expert recently talking about it and they were kind of shocked by the idea that the car would rely upon connectivity. The whole idea of the car is that it's autonomous. I hope they're right.

John: I think from my perspective, one of the most transformative elements of the autonomous vehicle revolution is going to be not that the car needs to consume a lot of connectivity, but that you, the driver now are consuming a lot of connectivity because you're essentially in your living room. The car has become this third or fourth sedentary environment. So if most of your bandwidth consumption and most of your perception of service, the things that carriers rely upon to reduce churn, is currently formed in the home and in the office and maybe at the coffee shop, the car now becomes that third or fourth place. And it's not a place where you can rely on wifi or any other mechanism. That may be the most dangerous part of all this.

Jack: Sounds like a lot of crash detection to come because if my car is my new living room, my car better be stopping without me having to tell it to because that is tremendous distraction obviously. But in a world where your car is smart enough to know when you're getting into hazards, perhaps that's not as big of a deal.

John: Well, we're already seeing it. Whenever I see a Tesla, I'm always anxious to look at the driver, make sure they're not asleep. It's funny, these things go from being novel to frightening to mundane. And it happens much more quickly than we think. The question is can the resource keep up with that human tendency to adopt new things and overuse them very rapidly. And in wireline networks when that occurred in 20, 25 years ago, the solution was economic. It's logistics. When you need more capacity, you lay more fiber. In wireless networks as that's occurred, it was a little more challenging but it was also logistical. More cells, incremental spectrum. The 5G period presents us with a case in which it's uncertain that we have a resource that we can add to it that is readily scalable. That way we can throw more logs on the fire, but they might not work as well. And that's I think where there is opportunity.

Rafe: Well, speaking of that, that's a perfect segue to my question in a way because what I'm wondering is if the government in the United States, and for that matter other geographies, if the governments don't free up more spectrum than is currently going to be available, is there enough real estate for 5G to work the way folks like you envision it?

John: It's difficult because for the last three decades, the government has reallocated spectrum that largely was under its own domain. Since the cold war we've had all of this spectrum that was allocated for military purposes and a lot more than we needed. And so most of it was fallow. And I'm sure ... I don't want to understate the difficulty, but you are basically reaching into your own piggy bank as a government, reallocating for commercial use, holding an auction. People came to that auction, spent very large amounts of money for that real estate and then it was of commercial use.

John: We're largely out of resources that can be re-farmed, as they say, in that way. And so most of what has to come to market is coming from existing commercial holders, folks who might be using that spectrum for satellite purposes or for microwave backhaul purposes, things that are not mobile 5G. And incentivizing them properly, getting through regulatory processes that make sense and are enduring is a real challenge. There are a number of ... C-band for example, which is happening right now, refarming of satellite spectrum at three gigahertz, which is incredibly interesting. Spectrum is an example of an early casualty of that, something that was to happen very quickly and now will probably be delayed a good amount of time.

John: That is I think going to be the standard set of issues that we deal with. So I think we don't have enough in the inventory today and we're almost certainly not going to have enough to get us where we need to go when we need to go there.

Jack: I feel like it was only a few years ago that four G LTE was the end all be all of what we could achieve with our phones and cellular communications. How long will 5G then be king? You've already mentioned 6G in our talks here. How many years will we get out of this generation and what would 6G even look like?

John: It's hard to say. My position has been that we passed a pretty important milestone that's not often talked about in 4G. In the 1G period, those are the 800 megahertz systems of the 1980s, the luxury period when that was something that was a novelty, you'd have a cellular phone in your car.

Rafe: The size of a shoe box.

John: The size of a shoe box if you were lucky.

Rafe: That's Wall Street.

John: Right, and the little curly Q antenna on the roof and that sort of thing. Those were mono-band analog systems. So those are effectively full duplex walkie talkies, FM radios. Very, very simple. In the 2G period, we took those radios, we made them digital. And that gave us an eight to one capacity advantage.

John: So suddenly the analog system that was so expensive, now I could put eight people on the same frequency effectively. That's when cellular technology became more egalitarian. That's when you know the average person could afford a cell phone. That's what started this whole revolution in the mid-nineties. Then in 3G we added the capability to put more and more data on that same size channel. And 4G with LTE, yet more.

John: So essentially beyond all the other features, all the other really complex things that we do at standards bodies to make and define each generation, the consumer and the industry has associated every generational transition with an ability to stick more bits in the same channel. The 4G period became very close to the theoretical limit. It's called the Shannon limit. It's an information theory constant that says in any given practical noise limited environment, this is how much data I can effectively stick in a channel.

John: And so from 4G onward, unless we have any fundamental revolution in our physics and mathematics, which we don't expect in the near term, each generational transition isn't really about being more spectrally efficient on the channel itself. And that's so important to understand that. And so all the gains that we're going to see in terms of capacity are no longer achievable by just sort of flipping a switch. They're only achievable through some much, much harder work in terms of network intensification and much harder work in terms of getting just more real estate to put this stuff on. So the stakes are so much higher. So I think when does 5G exhaust itself? Not until we can all get our hands on a lot more raw material than we have today.

Rafe: So 5G to a certain extent is the last G for quite a while it sounds like?

John: I think in terms of providing you the user with something that you feel in terms of effective throughput. 6G, I don't even want to start talking about it, but making assumptions about our ability to use extremely high frequencies and other features that really get beyond point of how many bits can you as a carrier deliver to me. And I think you're going to find that we are going to be somewhat constrained in that for a good while.

John: The analogy I often give people, they ask why is the system not collapsing? My phone still works. Everything's fine. It's more like the delivery of water. I might want infinite water because I want to create a water park in my backyard or something. I can't have it because the municipal water system, the pipes are of a finite, fixed diameter. That's essentially what we're dealing with in wireless. We're not going to be able to give people necessarily what they want when they want it without a great deal of infrastructure investment. But we can keep the system running because it's self-limiting. Basically, channels are a fixed size. I'm sort of rationing what I'm effectively able to give any individual user.

Rafe: John Dooley, thank you so much. This was fantastic.

Jack: See you next time.

John: Thanks.