Network enhancements revolutionize the world, but at what cost?
IEEE Future Networks Podcasts with the Experts
Network enhancements revolutionize the world, but at what cost?
We delve into the transformative power of cellular networks and their hidden cost on our connected world. While these networks have revolutionized access to information and bridged distances with instant communication, they come with a significant environmental footprint, consuming about 4% of our electric grid capacity. As we evolve from 4G to 5G, we strive to expand coverage. Listen to find out how this demand on our grid is set to rise and how we might address it.
On this episode, IEEE Future Networks podcast host, David Witkowski discusses with Doug Kirkpatrick from Eridan Communications this important topic.
Click here to listen.
Click here to download.
Subscribe to our feed on Apple Podcasts, Google Podcasts, or Spotify
IEEE Future Networks Podcast Host
David Witkowski
IEEE Senior Member, co-chair of the INGR Deployment Working Group for IEEE Future Networks
Subject Matter Experts
Doug Kirkpatrick
Eridan Communications
Podcast Transcript
Welcome to the IEEE Future Networks podcast series, featuring interviews with industry experts. Participation in the IEEE Future Networks technical community. community is open to active IEEE members,
including student members. Learn more about the IEEE Future Network's technical community by visiting our website at futurenetworks .itriplee .org. And now, on to today's show.
On today's episode, we're going to talk about the IEEE Future Network's technical community. we delve into the transformative power of cellular networks and their hidden cost on our connected world. While these networks have revolutionized access to information and bridged distances with instant communication,
they come with a significant environmental footprint, consuming about 4 % of our electric grid capacity. As we evolve from 4G to 5G, we strive to expand coverage. This demand on our grid is set to rise.
On today's episode, we delve into the transformative power of cellular networks and their hidden episode, David Wittkowski, chair of the Deployment Working Group for the IEEE Future Networks Technical Community, talks to Doug Kirkpatrick, CEO of Airden Communications.
Together, they'll explore how a groundbreaking advancement in radio engineering promises to enable enhanced cellular coverage, while significantly reducing environmental impact. Hello, my name is David Wittkowski,
IEEE senior member, co -chair of the Deployment Working Group for IEEE. Future Networks. On today's podcast, we are going to hear about the efficiency of the cellular network, the efficiency of the technologies that are used in the cellular network,
some of the challenges that are facing the industry, governments, and telecommunications companies around the world as we seek to grow the nation's and the world's networks from 4G to 5G.
I'm very pleased to have with me today the CEO of ARADAN Communications, Doug Kirkpatrick. Kirk -Pactrick. Doug and I have known each other for a few years. We've had some interactions in the real world,
and I thought that his technology from Aroden was very interesting, relevant to many of the things that we are hearing in the news now about the power consumption of cellular networks,
how efficient they are. And we're going to talk more about that in detail, but first let me welcome Doug to the podcast. Hello. Thanks for joining us. It's a pleasure to be here. be here, David. Thank you for having me. - So let's kick it off right into something that's very topical,
which has been in the news lately, there have been stories about how cell carriers are actually turning off cell sites at certain times of the day,
usually in the middle of the night to try to save electricity because of the amount of power that they consume, which turns out to be of course, into a cost for them. Just immediately,
why is this happening? Tell our audience why they're doing this and what does it mean? Why they're doing this, they're doing this to save money. They're doing it to save money because it finally came home to roost how much money they were spending for the electricity to drive their 5G networks.
when those 5G networks were being at best marginally used. It highlighted a real challenge in that the networks have now crossed a chasm to where the electricity usage that they've always been pulling on was a inconvenience to where now it's affecting bottom line profits by a large amount.
There was an announcement by the CEO of AT &T. I would be stunned if it's just AT &T being transparent. You're probably going to be seeing this type of behavior across most of the major mobile network operators around the world because the amount of electricity that you're talking about.
about is measured in billions of dollars for each of them. It's not a small number. That's a huge number, especially given inflationary pressures and other things that it's going on in the economy,
both domestically and around the world. That's huge. It's a huge number. It's getting bigger. And there's been a little bit of a wake up call.
Dominantly, I would say, starting in Europe, because electricity prices are so much more expensive in Europe right now than they are here in the United States. But around the world,
where people are realizing, "Holy mother, what's going on here?" You're watching numbers that, for those of us that have been watching this for the last couple of years,
are going to be a lot more expensive than they are here in the United States. 10 years, we're like, you didn't do the math. It's one of those, you're amazed that companies that are this sophisticated didn't sit down and draw that straight line and kind of go,
we're going to have a problem boys and girls. And all of a sudden they realized they had a problem is, oh my god, what do we do? So So, yeah, it's a big problem.
It's a huge problem. And it's going to get a lot worse. And so a lot of this, I'm sure has to your point, this has a lot to do with the question of availability of power,
given the situation in Ukraine, energy shipments out of Russia prices going up. But in general, it sounds like the problem was there already.
And it wasn't even the situation situation in the Ukraine that caused it to occur, it was just the thing that made it very obvious. So why is 5G technology having this problem?
Why didn't we have this problem in 3G and 4G? You did. You did. You just, it was just, it was below the critical bleeding factor.
I was just, as we were getting ready to come on, I was reviewing some of the the materials that we had actually put together into some presentations back in 2016 and 2017.
And there's a thesis from, I want to say it's the University of Pittsburgh where a master student was looking at using 380 volt DC transmission process instead of the minus 48 volt DC and looking at where power consumption was and even in 2015,
we're in a 4G world. The 4G ran in the United States was consuming 2 .5 % of all U .S.
electricity. Now it's getting worse because 5G consumes more than 4G, and particularly if you talk millimeter wave,
the power consumption of millimeter wave is astronomical. astronomical. And for the amount of coverage that it provides. And as we realize that a lot of the 5G business applications,
how the mobile network operators need to monetize that investment, they can only monetize that investment if they push coverage out to areas that have not yet been covered.
I .e. out into warehouses. warehouses and large dock areas and agricultural areas, mining areas, timber areas,
all of these ancillary applications, even vehicle to vehicle, was something that people were talking about in 2015 -2016, although that's been tamped down. All of that requires massively expanded coverage,
which just exacerbates this situation tremendously. unless you look at the fundamental underlying causes of the inefficiency in the first place. You know,
and as you and I have discussed as when you and Earl and I were discussing these kinds of things five years ago, and this is not a surprise. Those of us who've been pushing the technology edge for what we know we need to get to for ubiquitous wireless without melting the planet,
this is not a surprise. This has been something that we've seen coming for a decade. It's very interesting that you bring this up because of course one of the promises of 5G,
one of the things that they've talked about is this idea of say autonomous vehicles. To your point that requires the network to be fully covered even in in places in the middle of the desert because you want to drive across the desert and have your car take you there.
That means that you have coverage in areas that you didn't have before. If you're going to have 5G enabled robotic surgery, which was one of the things that we always heard was a application area of 5G,
that implies that there is 5G in places that would require robotic surgery. Therefore, to your point, you're covering a lot. a much larger percentage of the planet. What is it about?
So we talked about the millimeter wave and you brought up our mutual friend and the late Dr. Earl McEwn, my schoolmate, your chief technology officer until 2020. He did a lot of work on the analysis of the efficiency of 5G and specifically,
I remember having lunch with him in 2019 and he... shared with me some of his papers that he was doing analysis of the inefficiency of millimeter wave gigahertz power amplifiers and I think anecdotally any of us who have 5G millimeter enabled phones know that when you're on millimeter wave you get a very,
very fast experience but you also deplete your battery very quickly. If you would talk about two things, one is, what did,
why is millimeter wave less efficient? And then in general, why is, even in mid band or low band, why is 5G less efficient than 4G?
One is physics. The other is the current technology. approach and one of the reasons that I really enjoyed working with Earl is we came at things from such a complementary point of view.
He always started by focusing on the device physics and what it would take to generate the RF. And I'm a vacuum electromagnetist. I always looked at it from the standpoint of,
so I have the RF here and I need to get it over there. what's the difference between millimeter waves and 800 megahertz versus 2 gigahertz and what are those characteristic differences and I'm going through a window or I'm going through a tree or one of those pieces.
In millimeter waves you have both of those playing against you. On the one hand the sizes of the structures that you're talking about exciting in order to generate the RF are much much much smaller.
smaller, and the contribution of the electrical resistance in those devices as a fraction of the amount of power that you can push around just goes up linearly with increasing frequency.
Therefore, your efficiency goes down. It's not magic, it's physics. The concomitant piece in the millimeter wave, and even worse now that we're talking about submillimeter waves for for 6G,
is that the propagation through the environment is equally worse. I did my PhD thesis in millimeter and sub -millimeter waves. I have been literally doing this for four decades,
and the propagation characteristics, once you get above three to five gigahertz, except for a few windows,
where the X -band radars work in seven to 12 gigahertz, the KU -band satellite systems in the 13 to 18 gigahertz, then you just start going off a hill.
So once you get up to 29 or 35, if you're trying to propagate more than a few hundred meters, you're really having to put a lot of power out there. And if you're trying to go through leaves or moisture or anything,
anything like that, you just lose your signal tremendously. The flip of that is if you just look at 5G, now I'm just talking the things that I'm doing in 5G.
The 5G OFDM signal is not significantly different than the 4G OFDM signal. The fundamental underlying inefficiency in that signal is an inefficiency in that signal.
how we're trying to do it. Because of legacy ways that we go to market today, how the MNOs actually have their networks laid out,
there is a massive incentive for them to do what we call carrier aggregation and to put very wide signals through the amplifier. Well, I don't have to channel my inner Earl McCune to be tell you that is also a very very inefficient thing to do.
And as the costs of 5G have gone up, the economic incentives to drive those bands together have also gone up, which drives the efficiency even further down.
All of a sudden all of those chickens are coming home to roost and it's like somebody finally turned around and said, "Ma 'am, it's hot out there." and this is not a good thing." So it's a situation where in order to change it requires a fundamental re -examination of three things.
A, how do you generate the RF in the first place? That is clearly what Eridan has been about since our founding 10 years ago. It's a fundamentally different way to build a radio.
Two, how do you use the RF in the first place? spectrum? What are the better parts of the spectrum to be using for terrestrial communications, and what are the best uses for the other parts of the spectrum?
And then finally, what's the best way to lay out a network? Because some parts of the network clearly you want to have a very long reach, i .e. if you're driving on a highway you don't want to constantly be going cell to cell to cell to cell to cell to overhead.
and doing that would be tremendous but at the same point in time it's pretty easy for the network to figure out that I'm walking down Castro Street or Market Street and I'm going to be on this particular node for a long time so you can shift me off to a much smaller domain cell and use then have this heterogeneous layered network topology that allows you to be a lot more efficient in terms of how the signal gets
from A to B. All of that is physics. It's not negotiable. It's not even arguable. But we're not there. And there's a lot of going backwards that's going to have to happen before we can go forwards.
So that's really interesting. You talk about carrier aggregation, the need for linearity and the power amplifier, of course, drives against efficiency. The wider the amplifier, the more linear it needs to be,
at least in in the current technology. What is Aroden doing to address these problems? How is your company's technology changing the way that you do radio in such a way that it solves for that 5G power consumption,
power efficiency, and ultimately cost problem? Well, the easiest way to start with that is something that you and I, and hopefully a significant number of your listeners will recall from our childhood when we saw the first car phones and we saw those gargantuan plumber size toolboxes that went in the trunk to be the power supply for that car phone because that was a linear power supply and today your charger is the size
of a stack of 8 or 12 quarters and it got a heck of a lot smaller. What changed? What changed is you went from that linear linear power supply to a switching power supply,
and you went from a theoretical efficiency limit of about 60 % for the linear power supply to about 99 .6 % for that switcher. And as a consequence,
your inefficiency went from 40 % to 0 .4%, and since the amount of heat you can dissipate per unit surface area is constant, that means your surface area could go down by a factor.
of 100, which is how you went from the toolbox down to the thing that's a stack of quarters. At its core, what Eridan does is just that, except 10 ,000 times higher in frequency. And I'm throwing a whole lot of babies out with that bathwater when I make that statement,
but just so that everybody understands that what we're talking about is a fundamental change in the circuit. I can't tell you how many people have come up and said, "Can we please get your power amplifier?" And I just rubbed shoulders and I go,
"That's going to be difficult. We don't have one." The problem that we solve is how do you make the fundamental radio piece more efficient?
When you do that, there are a bunch of ancillary pieces that come with that. The first thing that comes with that is you can't do it with silicon. Part of what we do is we do it in gallium nitride.
That's another whole layer of invention and development that has been going on for the last 10 years. The next thing that you run into is that there's a different sweet spot for this technology and where it's going to grow and grow from rather than the legacy technologies.
Legacy power amplifier technologies are more efficient than... bigger you make them, the higher the power you make them, because there is a fixed overhead that they are amortizing across a higher output power.
In the switching technology capability, today there's a limit of about somewhere around 5 or 10 watts per module.
Now you can stack those modules together to get to higher power. if that's desired for something like a macro cell, but there's going to be a very native efficiency point,
somewhere's in that one to ten watt average output power range, where this technology just absolutely excels. It's going to be a lot like the transition that I went through as a young graduate student,
where we see using mainframe computers and we went over to large numbers of PCs. Think of those big RF systems that have those massive cables that climb up the towers up to the antennas as your mainframes and think of the Eridan radio technology as the PC.
It will require more of them, but you will have an inherent far more powerful and far more flexible and far more dynamically tunable in terms of what you use to what you need system using that type of much more Atomic scale technology right you're driving things down Disaggregating it so that everything can be flexible and move around as as the capacity requirements move,
the radio access network can move. This is something that a group in the IEEE has been looking at. I've been trying to help them and make sure they understand, at least from the radio perspective,
what's going on. But the minute you unlock a radio that has this type of dynamic performance capability, capability, all of a sudden the back end,
what people have talked about as the ran intelligent controller, all of a sudden can have all kinds of dynamism in it that previously linear amplifier systems,
which are fixed in frequency and fixed in characteristics, cannot do. That is fundamentally what this new radio technology is going to bring, and hopefully we will see it.
it as everybody realizes how important efficiency is, efficiency will be a cornerstone of 5G and 6G and beyond. - And normally, this would be something that would,
frankly, be hard to get the carriers to care about. But to your point, with the consumption of electricity driving against cost, against profits,
they have to do this. It's getting that, it feels like to me, me, like they've not only realized that the mainframe doesn't work for what they want to do, but also that the mainframe cost them a lot of money.
And by going to PCs, they now have to, they can avoid that, right? Absolutely. And in some cases, there will be things that look very much like an old -fashioned macro cell,
but there might be two dozen errant radios on the inside. that you don't see how they're working as a network. Very much like today, you go in and you look at a server and you don't see 24 CPU PCs,
there's 24 CPUs sitting inside that server. And that is very much the way that this is going to go over the course of the next 10 years. And ARIDIN will be at the core of that.
That's fundamentally the building block that we bring to the table. - And that leads me to a question, which is early, you mentioned a power output earlier that that was,
it's fairly low, at least it is, I presume, right now. So typically when I'm working with deployment and I'm working with SiteBuild and I'm looking at EME reports and things like that,
at the power level. levels are above a couple of watts, which is, I think, what you mentioned. So how do you, what's the roadmap to get there? Because it sounds like right now you're looking a little bit more at what we would have called a Pico cell or a small,
small cell, as opposed to a macro. So how do you transition what Eridan is doing from that world into covering a mile,
two miles out. So there's two components to that. The first is that because we are an old digital switch system, the signal that we generate is infinitely cleaner than a linear amplifier.
And it's no secret that clean signals are demodulable at significantly lower power than a dirty signal is by a handset. handset. And we get about a 3 dB advantage out of that.
The other piece that's really important is that because we're an old digital system, an all -switching digital system, we are phase locked to the clock. As a consequence,
it's very easy to add multiple units together at very, very high efficiency of doing that addition to get to... to the 2040 60 watt output that you might want in a conventional macrocell.
All the while avoiding the overhead of having to do that in a building that requires air conditioning that requires 200 feet of very expensive and lossy cable to go between that hundred hundred and fifty watt amplifier up to the antenna.
So there's there's a win, there's another win, and there's another win. When you actually get all the way to that point four or five years from now there will be things that look like macro cells,
but they'll be very very different macro cell. It feels like with the clock accuracy that you're talking about, the precision timing, that that would also drive towards realization of the Moo MIMO,
right, where you would have different sites serving. a Subscriber from two different possibly even more directions to build that so talk talk about what you're doing in the context of MIMO if right now one of the ways that the carriers are saving money is Turning sites off which is a pretty blunt instrument Because at the end of the day,
you don't know that somebody didn't want to use that site in the middle of the night. If you're able to do MIMO and you're able to do things that allow the site to be available when it needs to be available,
but not to be turned off effectively when it's not, then I think that it sounds like that solves that problem. Can you expand on that? >> Sure. Because we are phase locked to the clock, what you're describing,
I'll put it in my physicist's term of a phased array. You can do what we call a long baseline baseline disaggregated phased array. You can have elements of your antenna at different physical locations,
coordinating with each other in order that that signal arrives at the intended destination in a coherent and aggregated way. That's one of the things that we've been talking about in MIMO,
but it's one of the things that people in the radar community and the DOD have been talking about for 30 years. We We enable that. The second aspect of that that you're getting to is sleep modes.
One of the benefits of the old digital technology is that it has three layers of very rapidly accessible sleep.
First and foremost, when we're off, we're off. On the RF side, we're not pulling any power. The receiver will still be there because the receiver has to... listen, but the receiver is maybe 5 % of our overall power budget total.
There is the back end that has to continue running. We have some methodologies that we are implementing over the course of the next couple of years in our product evolution ramp,
and I anticipate that we will find that we can relatively rapidly save somewheres between 90 % and 90 % of our power. and 95 % of the power. And when I say relatively rapidly,
I mean in the space of a single 5G frame, 10 milliseconds or less. We can either come awake or we can go back to sleep. That is a huge benefit.
The other power savings that you get is for the big wide open spaces. If we think about the long baseline massive MIMO, realize on it. a subcarrier by subcarrier basis that type of radio slash integrated antenna structure allows you to put a set of subcarriers with very,
very high gain where those subcarriers are being used, i .e. where there's a customer and not irradiate a whole bunch of farm fields that just have cows.
Now, if they're IoT cows, maybe you're gonna sweep them and check their collar. to see if they need milking, but then you're going to come right back again. The whole point here is that you don't have to just be blasting RF everywhere.
And that is your biggest power savings. Use the RF when you need it, where you need it. And that's how you can actually ultimately end up saving a factor of 100 or more.
It also seems like this would drive towards a network slicing business model. model where if a customer wants to purchase a slice of the 5G network, they would be able to turn it on or not as the case may be.
And of course, they would realize savings from not turning it on and not having to pay a portion of the power bill that they're not using simply because they're buying a percentage of the network. So that's really interesting.
100 % correct. There is an end slur. to what you just said, and that is the entire power system. Requirements are so reduced that remote cell sites don't need to be grid connected.
They can be solar powered. And as a consequence, the savings for putting in those remote cell sites goes through the roof because to push a high voltage high power connection to a classic macro cell site.
It's about a million dollars a mile To be fair, you still need to connect those sites with some sort of backhaul Absolutely correct. You can do that wireless. You can particularly if it's a relatively Latency tolerant requirement That opens up the question of how you'd want to do that if you were actually doing the true connected car But I think there's a balance there between how connected the car is and how autonomous
the car is. And there may be a very middle ground there as to the right way to do that. That's really interesting. All this is really interesting. I'd like you to take it home as we close out in a way that I think our listeners can put it in the context.
If you were to transition the cellular network today to the technology that ARADAN is using, just if I could snap my fingers and suddenly ARADAN is the radio for all for the nation or for the nation in the world cellular networks what would that look like in terms of equivalent power savings carbon consumption how do you how would you put that into terms that the listeners can visualize there's a lot of ways to think
about that I mentioned earlier that two and a half percent in 2015 of the US electricity was used for the 4g ram network and 80 % of that was for the towers,
for the radios. So that's 2 % of all US electricity. It's gone up since then. 5G is making it far worse and at that point in time we only covered about 5 % of the roads in the United States.
If you looked at a full network optimization enabled by the Ayrton radio technology, you're talking about covering the United States with 5G and still using substantially less electricity to do that than we used in 2015.
The other way to think about that is if you tried to do that today, you'd use up almost all of the U .S. electricity, and 80 % of the electricity in the United States powers every major city in the United States.
So the difference between between aridine and not aridine is those cities are staying lit and cool and people are in a position to enjoy the fruits of this technological life that we live or those cities are going dark because your 5G network is sucking every ounce of electricity away from them.
Clearly that's not going to happen, we're not going to go there, but the bottom line here is that we can open up this path to comprehensive wireless coverage everywhere and we can do it with a budget that we lived with a decade ago.
- So that's really interesting because when you put it into those terms, the notion that we are going to broadband equity, access and deployment,
the BEAD program, looking a lot at wireless as a technology, it's a 5G fad. other efforts, government efforts to move 5G forward. If we were to achieve those goals,
it sounds like you would be saying that basically the 5G network would wind up taking up all the electricity, which would leave nothing for everything else. Not the least of which would be that the goals of some government state or national to transition everybody to electric cars.
And we've already talked about that. about how charging electric cars is going to be a challenge because the grid is not ready for it. So combine that transition to electric vehicles with a transition to 5G and you have a massive problem.
So we can't not solve this problem is what you're telling me. Exactly what I'm telling you. It is absolutely a requirement. Well I'm glad you're working on it and I'm glad you're continuing the work that Dr.
McEwen started and moving that forward. I know you have a fantastic team. I've had a chance to talk to many of them and they're all really brilliant people. And I hope that you will be successful in the future.
Thank you for being with us today. Again, this was Doug Kurt -Patrick from Eridan Communications, CEO. I'm David Wodkowski from the IEEE. Future Networks technical community co -chair of the Deployment Working Group,
and I will see you next time on the IEEE Future Networks podcast. Wherever you are, have a good day, good evening, good night. Thank you for listening to this episode of IEEE Future Networks.
the IEEE Future Networks podcast. Be sure to like, share, and subscribe for future episodes. Discover more about the IEEE Future Networks tentacle community, and inquire about participating in this effort by visiting our website at futurenetworks .iEEE .org.
This podcast was made possible by the IEEE Education Committee. Our music is by Alex Grawl, sound designed by David Wachowski, and voiceover is by Nora Wachowski. Wachowski.
----