Mobile's Rise in the Future of the Internet

Thursday, November 20, 2014
Jonathan Spalter

Chairman, Mobile Future, Inc. 

Dominique Lazanski

Director, Public Policy, Groupe Speciale Mobile Association 

Robert Pepper

Vice President, Global Technology Policy, Cisco Systems, Ltd. 

Craig Mundie

Senior Advisor to the Chief Executive Officer, Microsoft

Dominique Lazanski, director of Public Policy at GSMA, Robert Pepper, vice president of global technology policy at Cisco Systems, and Jonathan Spalter, chairman at Mobile Future, join Craig Mundie, senior advisor to the chief executive officer at Microsoft, to discuss technological change and the move to mobile devices. The panel converses about the growth of mobile Internet, regulations and standardization, the implications of shared spectrum, and security issues.

This meeting is part of the Internet Governance After Busan Symposium, made possible by Google, Inc., with additional support from Intel.

MUNDIE: I'm Craig Mundie. I'm senior advisor to the CEO of Microsoft, although I'm in the final stages of my retirement from there. Today we have three people with me — Jonathan Spalter, who's the chairman of Mobile Future; Dominique Lazanski, who's director of public policy at the GSMA; and Bob Pepper, the vice president of global technology policy for Cisco.

And the topic that we've been given is essentially the issues around mobile Internet governance. And so I thought the first thing we ought to do is start off and ask the question, you know, what's different about mobile? You know, the Internet we know is connecting many, many things, and that'll continue to extend.

So what is the mobile Internet? And in what way is it different? So perhaps, Bob, let's start with you.

PEPPER: Sure. So, like everybody else, I want to thank CFR for pulling this together. It's very timely. So I joke sometimes when you talk to some of the companies in GSMA, right, and they talk about their mobile network. And as far as I know, there's no such thing, right? Networks are not mobile. We are.

The networks are wireless. I mean, think about it. The networks don't move around, but they enable us to be untethered. And we do a study every year called the Visual Networking Index. We've been doing it now for nine years. It's a rolling five-year forecast, so we actually go back and check the forecasts, and it's usually within 10 percent on the conservative side.

And in our latest study, on a global basis, so how do people connect to the Internet? Ask the question. How do people connect to the Internet, in terms of the traffic, right, in terms of the data transmissions? So globally, 76 percent of the traffic is wireless traffic. It's not fixed, right? So it's not this Ethernet plug that goes into the back of the device.

Of the 76 percent, 15 percent is mobile. Definition. It's a device with a SIM card, right, so it has a SIM, and it goes to some macro cell someplace, a cellular network connection. But there's another — that's 15 percent, but there's another 17 percent that could go to this device that actually is offloaded onto WiFi, whether it's public WiFi outdoors or it's WiFi in the room like here or at home or at your office. And that's done at your convenience or the operator's convenience.

Then there's pure WiFi, like my laptop doesn't have a SIM, so that's just WiFi. That globally is another 17 percent. The fixed network, right, is only 24 percent. Now that's global. And it's a global trend, but let's take a look at around the world.

In the U.S., 75 percent of total traffic is wireless. Between mobile, it'd be 11 percent. WiFi offload in the — in the U.S. is going to be 20 percent. In other words, of the traffic to that mobile device, that device with the SIM, more than 50 — actually, two-thirds of that traffic is going — this is by 2018 in the forecast — is going to be WiFi.

Europe, 80 percent of the total traffic is going to be wireless. India, 87 percent of the traffic is going to be wireless, because their fixed networks, frankly, can't hack it. And to the SIM device, it's going to be 22 percent. You know, it's a global phenomenon. So when we talk about the Internet, it is fundamentally wireless. The way people connect their devices to the Internet is wireless.

Second, you know, we talked about a lot in the first session, and this is post-Busan, about the ITU, and there were passing references to the broader agenda at the plenipotentiary. And the broad agenda at the plenipotentiary is to set out the roadmap for the workstream for the ITU for the next four years. You know, it's not just about the telecom and the regulation and the Internet debates, which were very important, but there is one of the — the bureaus within the ITU is the development bureau.

And Eric talked about the importance of capacity-building and development, connecting the unconnected, and the agenda tying what we do in the sector, the tech sector, to the Millennium Development Goals or whether — which, by the way, it was end of 2015, it's the next batch are going to be the Sustainable Development Goals, the SDGs. And I think it's extremely important to think about, if the Internet is important, how people get access, how people in the emerging markets get access where they don't have it today, what are those networks going to look like? And, again, there's going to be a big, big piece of that, which is going to be related to spectrum, which is the third bureau, which is the radio bureau.

And I think most people would agree that the radio bureau at the ITU is doing a pretty good job. If we didn't have it, it would have to be created, because spectrum is global. And it provides spectrum on a coordinated basis, a harmonized basis. And even there, they're beginning to look at and begin workstreams on managing spectrum and making spectrum available in ways they never would have in the past, right?

So including unlicensed, not just licensed, including, they're having conversations now about sharing and spectrum sharing, which is something that they'd never talked about. So there are some good things happening there, and particularly in mind on the notion of connecting the unconnected, especially in emerging markets, and the other parts of the ITU, and that is all going to be spectrum-related, or for the vast, vast majority of it.

MUNDIE: Thanks, Bob. So let's go back a little bit more to this question, though, of what's different about mobile. And as it relates to the future of regulation, standardization, or anything else, are there anything unique about it? You mentioned the Internet of things. So, Dominique, I'm interested to know, when you think about the Internet, everything, as some people talk about it, or at least the connection of all of these other classes of devices, what kind of issues does that bring forward that you think might lead us to think that the mobile aspect, whether it's mobile because people are mobile or because we're being observed in a mobile environment, you know, what are — what are the things that you think are going to call for special consideration in this area of regulation or standardization?

LAZANSKI: Well, I think you touched on a lot of things that I wanted to mention, but just to highlight a theme that seems to be emerging — and that is connectivity, right? So for the GSMA, which is the mobile phone trade association, over 800 mobile operators are members, this is a priority for us and we have goals to connect the unconnected. And not just mobile Internet, which is a high priority, but also just connectivity in general for communications.

So I just wanted to build on your stats, just really briefly. We at the GSMA also have a lot of intelligence, and we do a lot of work internationally beyond developed countries. We do a lot of work in developing countries on capacity-building, spectrum management, regulatory best practices. And so out of that, we have a lot of wonderful information.

And just to highlight that today, the mobile Internet subscribers numbers is 2.2 billion, looking to grow to about 3.8 billion in 2020. And the primary growth will be in developing countries. So another issue about connectivity is the fact that people will experience the Internet, whatever you want to call that, through apps or through websites or however you want to call that, on mobile for the first time and only that way, and on mobile as you've described it, not just cellular mobile, but WiFi, as well.

And so it sounds very basic. Connectivity is really a big priority, you know, is really the big focus and should be, because we want to connect as many people as possible as quickly as possible. And I also wanted to highlight what you said. The ITUR, the radio development bureau — or the radio bureau in the ITU is a major focus for us. There's a conference coming up next year in about a year called the World Radio Conference next December.

Now that Busan's over, we at the GSMA are focusing on that, because it's an opportunity to really try to be very aggressive and to discuss all these emerging issues around spectrum, spectrum management, the release of more spectrum, harmonization of spectrum, and the ways to get people online that way.

MUNDIE: So this spectrum question I think is an interesting one. There's only one spectrum out there. We can't really make more of it. And as a result, you know, we're going to have to change the model. That seems to be a foregone conclusion. One could ask, you know, why do we see as much offloading, to use the term Bob used, to WiFi? It's largely an architectural question. I mean, the mobile telephone started out with a wide area architecture, wide area communications still going to be important, but increasingly very local area communication is important. You know, I've got this fitness band that talks Bluetooth, not WiFi, not cellular, but it's an Internet protocol on that radio, too.

So it seems like we're going to have a lot of different radios at a lot of different ranges, and the shared spectrum is going to be important. So let's talk a little bit about, what are the implications of that? You know, if you look in the past for example, in the TV space, we had satellites. And when satellites happened to, you know, spill over borders, it was — caused consternation, I'll say.

You know, it seems like we have a huge problem looming now, not just because radios don't stop at the border, either, but because the devices that people carry are going to move with them. You know, the average device now has, let's say, many gigabytes of data on it. That will only get bigger. So what does it mean for data portability, if you can — you know, carry a few gigabytes across the border in your pocket, you know, how do you think this is going to be dealt with in these areas? Where should those discussions happen?

SPALTER: Well, I believe — let me just sort of give you a little bit of more numerical color to that point — first of all, we know that globally roughly — in terms of global capacity of spectrum use, it's about 65 percent. We're at about 80 percent of capacity here in the United States. Those numbers...

MUNDIE: Which spectrum? Cellular?

SPALTER: Cellular spectrum, electromagnetic...

MUNDIE: OK. But that was a choice of their architecture.

SPALTER: It is a choice of architecture, but the point is, is that we' are ascending towards greater and greater capacity constraints across spectrum, which logically suggests that there has to be a set of not only technological and architectural enhancements in evolutions of the kind that you've discussed, Craig, in many different fora, spectrum sharing, spectral optimization, spectrum efficiency technologies, many of which are being developed in R&D labs here in other institutions around the world.

But we also have to, I think, understand the ascendancy and the importance of understanding how we actually deal with licensed spectrum, the evolution of new spectrum-based technologies like WiMAX and WiFi, and the role that they play not only in developed, but developing societies.

You know, an interesting onramp to understanding the complexity of the issue in terms of mobile broadband spectrum, a single — single strand of fiber optic cable carries about 2,000 times the spectrum, all of the mobile spectrum that we had actually here in the United States. This is an issue that transcends our own border. It is a global and globalizing issue and one that has to be joined, I don't think only by governments, but by the multilateral institutions that are charged with thinking about these things, including the RWC.

I would also say that I think that there's significant — there is a significant role that is being advanced currently by forward-thinking companies, many of which are based in the part of the country that I live in on the West Coast, that are also thinking about adding their voice, their pocketbooks, and their programmatic work like Facebook's to thinking carefully about this problem and constructively about this problem.

I think that this is even more pressing, because though Bob rightly suggested that the mobile Internet is based on these phones that we have that we've all turned off that have SIM cards, but as Cisco I think very, very elegantly and eloquently has been helping us understand, the Internet of everything, which is also a spectrum-reliant technology, and has the potential to add hundreds of billions of dollars in economic productivity globally, will transit not between human beings, but between things, things that are actually going to be the key catalysts of economic growth and development and commerce and transaction and health care and education in so many societies around the world.

As you say, all of these wireless sensors, hundreds of millions of them that are going to be deployed globally, are not going to be static. They're going to be very mobile. They're going to transit borders, facilitating greater commerce, greater efficiency of commerce, greater efficiency of manufacturing, we know, but in order to do that, we have to begin to also engage thoughtfully on how and where standardization, security, privacy issues will relate to the Internet of everything.

I'd also make one more point. I chair an organization called Mobile Future. Our remit is to think about the future. I recently was able to spend some time at Stanford University's Biomechanical Engineering Lab, and I saw an extraordinary advancement in the Internet of everything, where we now have scientists that are deploying nanoscale wireless sensors that can transverse our vascular system to actually bring drugs and biometric analysis to specifically targeted organisms, organs of our body, and that, in fact, are powered by electromagnetic spectrum that transverses our epidermis. They don't have to have batteries. This also will be an ineluctable part of the Internet of everything.

I think that ultimately we're not actually going to be calling it the Internet of everything. It'll simply be the Internet. And these two will be wireless spectrum dependent aspects of our future life. If we don't engage the Internet of everything in this context with greater concentration on how licensing models might still add value, but also in parallel the advancement and the enhancement of spectral sharing, optimization, and efficiency technologies, then we have a lot to worry about.

MUNDIE: One of the reasons I think this question of standardization is so challenging — and we increasingly so — is that people who thought about these things historically in the ITU and other places, spectrum, for example, have always been dealing with something that's a physical system. And yet more and more, the Internet — and, of course, we've seen the movie before in computing, per se — is these things are defined by software. And in the world of software, everything almost always evolves as a de facto standard that comes just because practice has already been established by products.

And I think as we think about whether it's the ITU or the ITF or any of these other standard-setting organizations in the future, it seems like we're going to have to draw a brighter line between the things that ought to be de jure standards in order to promote, you know, rapid growth of the market, sort of like GSM did for phones at one point, versus software standardization, where any attempt at a priori standardization has historically failed in computing and I predict will fail in the, you know, mobile Internet and every other aspect of the Internet.

So I'm curious, you know, to get any of you to comment on, you know, how do you think we should manage this division of labor, if you will, or division of architecture, between the things that software is going to define in the Internet and the things that physics are going to have to define for the Internet?

PEPPER: Yeah, so, you know, I think the answer is yes. It becomes...

MUNDIE: But — but where?

PEPPER: But where. So in order to answer that, let me just pick up a little bit on what Jonathan talked about. The fastest-growing category of devices that we see being connected are machine to machine. This was something that we didn't even sort of pick up a couple of years ago, but we're now forecasting by 2018 five years in terms of Internet protocol-enabled devices, right? Globally, it's going to be 35 percent of all the things connected. In the U.S., it's going to be 45 percent. In Japan, it's going to be 65 percent.

And the heterogeneity of those devices is really quite dramatic. There are going to be some, like your, you know, wristband that are going to have a requirement to only go a couple of inches from, you know, using something like Bluetooth, and it's going to not — so in terms of the characteristics, some devices are going to communicate over short distance, some over long distance, some over — some are going to just be very bursty, just going to, you know, burst a little data, they're going to chirp and die.

Some are going to have continuous power. Some are not. Some are going to use ambient power. Some are going to — like you talked about, Jonathan. Some are going to be broadband, some narrowband.

So when you look at that, the spectrum requirements are going to be very, very heterogeneous. Short distance, long distance, support propagation that's dense, others are going to be very dispersed. Some is going to have to support broadband, other narrowband, and so on, and when you begin to map those, some of the requirements will be baked in, but a lot of that is going to be in the application side, software driven.

So one of the things that I've done with a colleague is looked at, so what are some of the policy issues related to the Internet of everything, right, and technology issues? So there's a group of technology issues, and some of those are pure technology, things like, you know, power questions, you know, reliability, scaling, and so on. Then there are some pure — some really important policy questions which we do need to talk about, because it's how do you take these things across borders. It's the transborder data flow. It's the data localization. It's things like open data, intellectual property.

But then in the middle where the technology and policy intersects, that's where it gets really interesting and potentially more difficult, because that is interoperability standards, where are they made? Are they software or are they legacy telco standards? And I think they're increasingly going to be — based on software models, which are going to be industry-led. By the way, there are — two years ago, there were two industry consortia looking at standards in the Internet of things, Internet of everything space, there are now seven, just to give you a benchmark. There are eleven global standards groups in automotive, so seven is probably not too many. It's probably not going to stop.

The most recent one is a brand-new one called the Industrial Internet Consortium. It already as of about a month ago had eighty-five companies. It's probably more than that now. And they're working precisely on industry approaches — instead of traditional government standards. Some of the other issues which are critical — trust. We heard trust in the first panel. Without trust, people aren't going to use these things, and that's both privacy and security, and then the other issue that is — this intersection in that Venn diagram of policy and technology is the spectrum. What kinds where? Is it going to be opportunistic use? Is it going to be dedicated? Is it going to be licensed? Is it going to be unlicensed? And the short answer is it's going to be all of the above. We're going to need more of all of the spectrum, and the only way we're going to get more is through some of the new techniques.

But it's about architecture and the interoperability that comes from the traditional approach on software standards as opposed to the traditional hardware standards.

SPALTER: I'd also just add one important point, that in addition to those categories that you've mentioned, that network management practices also are going to be much more software-reliant, much more software-dependent, and much more software-advanced. And already we've seen just in the last — if we take the United States as an example, where we would update our network configurations at one of our major national wireless ISPs one time a day, more advanced software-based approaches to network management, which will have its implications for the Internet of everything and the management of those interconnections, is now six times a minute, 6,000 times a day.

And we are still at the infancy, as we're evolving networks towards 5G, of explosive complexity, but also advancement of network management practices, which will also have global and cross-border implications.

PEPPER: Well, when you have the heterogeneity devices and not all the devices have the same requirements in terms of either time-criticality or other, you know, dimensions, and that all, you know, bits, you know, are not created equal, because they have different requirements of when they get where, and queuing and streaming is different than download and et cetera, yeah, how you manage that is going to be increasingly important. And the management's going to be software-driven.

SPALTER: Absolutely.

MUNDIE: Dominique, let's just use GSMA as an example of, how's that group of 800 companies going to drive the government standardization process in the world we just described? Let's just say that today, you know — well, to be very precise, so ITUR, which everybody commends today, has operated on a 100-year-old model of basically the way we manage interference is we give everybody channels. And this was just a big place for the world to negotiate channel by channel, you know, what was — what happened in each little chunk of the spectrum.

So now we know that that's over time got to go away, for all the reasons we've just said. And it's probably going to be replaced with a much more dynamic, software-driven, you know, access system, and which really converts the spectrum allocation process increasingly into a market-driven process probably, or at least for large segments of it.

And the devices will be more like digital computers that are wideband, as opposed to, you know, channelized, and in that environment, you know, the whole process of negotiating, you know, will be reduced to a few sort of super-channels, and then a bunch of, you know, market and software-driven processes. I'll just speculate that that's going to happen.

So does GSMA agree with that kind of analysis? I mean, they're already facing this challenge of — you know, to use, you know, the WiFi world, for example, as a complement, why are they doing that? Well, because that world's nanocellular, or small cell, and the big cells, you know, don't get the kind of packing that you need and reuse.

So, you know, you can see all the trends. How does an organization like GSMA, you know, come out on that issue? You know, how is it going to provide guidance to its governments in that multilateral process of allocation? And, you know, and are people starting to think really about the long-term future as opposed to sort of fine-tuning, you know, the last stages of the world we've known for 100 years?

LAZANSKI: Well, I think you raise a really interesting question. And I want to take sort of a 1,000-foot-view of it, in the sense that what we...

MUNDIE: A thousand, that's pretty close.


From a pilot's perspective, 1,000 feet...

LAZANSKI: Well, but — but sort of stepping back from it, one thing we haven't discussed yet is sort of the regulatory best practice, country to country and internationally, and what mobile operators need in order to ensure that everything we've been talking about so far is able to happen in a very competitive, market-driven, often industry-led kind of way.

And so one of the things that we talk a lot about at GSMA is looking at, you know, regulation and country by country and how to do best practices on spectrum management, how to ensure that regulation isn't onerous or overly burdensome, and these are issues that I know have come up time and time again, but it's really important that companies that are providing and using licensed spectrum and providing users in primarily, again, developing countries where we're going to see the growth of most users, have the opportunity and the ability to be very agile and to kind of react to the market and deal with things like geographical problems, with backhaul, with putting up masks, things like that. And part...

MUNDIE: I still put that in the category of, you know, sort of tweaking the old world, not really defining the new world.

LAZANSKI: But it's important for the new world. It's important to have that flexibility. And I think that's — that — to me, that's one of the most important things. And level playing field, as well. So...

SPALTER: May I just add a thought?

LAZANSKI: Yes, absolutely.

SPALTER: And I think it's important, Craig, that I note that you as a member of President Obama's council on advisers for science and technology asked all of us in the United States to raise our vision in thinking about our broadband future and imagine a future where potentially we might even be able to create 1,000 megahertz of new spectrum by transiting to new models, new software-based models, new approaches to the exquisitely complicated processes of licensed spectrum that we've been using.

You also challenged us — and there's beginning to be implementation around that challenge to begin to advance within our government, but also between government and industry, new forms, new habits of cooperation procedurally, but also technologically around sharing, not just geographic and temporal sharing of spectrum, but also dynamic sharing, sharing based on cognitive radio standards. And these all are, indeed, very promising. They portend well, I think, for the mobile future here and globally.

I, however, have to I think stick with Dominique's perspective that as we transition to that brave new world and promising new world, that there still is wisdom and validity to existing licensing models, to existing reallocation models that amongst many other things provide an ecosystem and a climate for certainty that provides for more, not less CAPEX or investment, very, very important investment in the extension and elaboration of broadband infrastructures that are going to contribute greatly to continuing to close the gap for the next billion who will be onramping into the Internet and also maintaining the viability and dynamism of existing networks.

MUNDIE: I'll just say...

SPALTER: I don't think it's one or the other. I think we have proceed on both paths.

MUNDIE: Well, we have to manage the transition, you know, but — you know, I'll just say I'm on the other side of that, you know, position, because my view is if you know that this new future's going to happen, every single new allocation of any spectrum for any purpose that is on the old model is just another big piece of legacy that you've got to figure out how to manage in the conversion.

So if I go to the emerging countries, who actually have far weaker — you know, overall utilization, I could say, wow, they're so lucky, because as long as they don't screw it up, you know, they could get to the new world faster, and the new world will actually guarantee them a lot more access at a lot lower cost than the classical models.

So that's why I think this is a tough challenge, that if you just continue to try to project the old world in to get connectivity for Africa or anywhere else, you may essentially be kind of — I'll say where the landline guys were, you know, for decades, you know, saying, well, just wait, we're going to get there with the copper wire. And eventually, you know, the wireless world came along and the emerging country said forget that copper wire thing, give me that wireless thing.

And so I think, as we move to the Internet of everything and we're going to have so much more demand for connectivity, that there's really an opportunity to say, no, no, those people should lead us to the new promised land, too. They should get on with it and not, you know, build more legacy. And it's clear — as you say, I mean, the — you know, yes, we've encouraged this in the United States. A lot of the technology flows from here. But, you know, if people are aggressive adopters of it, as opposed to, you know, classicists in the way they approach this, they might get a different outcome.

So let's open the floor to the audience now. I'll remind you to — the man right there with the red and blue tie, please stand up, introduce yourself, and ask a question.


...It should, yeah.

QUESTION: Hi, I'm Alan McQuinn with the Information Technology and Innovation Foundation. I just wanted to ask, we've seen a lot of — the trend of a lot of countries erecting these kind of barriers to cross data — cross-border data flows with a lot of, like, data localization policies. What does this disturbing trend kind of show or lead to going forward with the Internet of everything and adoption of mobile networks? I just wanted to see your feedback on that.

MUNDIE: So, Dominique, you want to take that?

SPALTER: I'll give it thumbs-down.

LAZANSKI: Exactly.

PEPPER: I think — frankly, if you think about, you know, something — as the world moves to a world of Internet of everything and we're looking — and most of that data will be analyzed and stored in the cloud or, you know, if there are — some people call them cloudlets, we call it fog, at the edge of the network, right? Because not everything has to come back to the center.

When you think about that, and we're talking about global networks, it doesn't work. Let me just give you one concrete example. You know, an aircraft engine today is generating terabits of data, and that — and that — those data are actually going back to the engine manufacturer, right, who are monitoring that engine in real time and know more about that engine than the pilot, and if a part needs to be replaced, that part can be forward shipped and it can be waiting for the plane when it lands.

Well, you know, plane takes off in Tokyo on the way to London and it crosses, you know, China, and then Russia, and then Germany, and then, you know, France, Belgium, and lands in the U.K., and if there's data — you know, what does — what does data localization mean in that world? It makes no sense.

And, frankly — and, you know, there are a lot of other examples, but, you know, I think that the efforts for mandated data localization for all data within a country actually would be — it's, number one, anti-innovation, it's anti-investment, it's going to hold back a whole host of huge economic and social benefits globally.

Now, individual, you know, customers, especially governments, a government may want to keep its, you know, national security data within its country. That's fine. That is very different. Any customer that, you know, that's, you know, using a Microsoft data center that wants to keep, you know, data in — you know, it's in a particular country, that is the choice of a customer for, you know, buying capacity in a data center. That's very different than a government-mandated requirement for data localization, especially in the global world of global enterprise, global networks, global business, global travelers. Data localization I think is completely counterproductive.

MUNDIE: But in the spirit of the governance question here I'll just say, you know, you have to look at what's driving this local requirement. And I think — in the sense, you probably have to put the data into three buckets, all right? There's — you could say — as you said — national security-related specific data or government operations. There's personally sensitive information. That's sort of the privacy bucket. And there's everything else.

So, you know, the governments probably don't care that much about the turbine engine data, you know, from a privacy point of view or a national security point of view, so — but today, because they don't have a taxonomy like this, when you have the conversation, that's where I think we end up in trouble, where, you know, you're worried about privacy and you end up sweeping in the turbine blade data. You know, then that's where the problems come.

So I think, you know, the technologists are going to have to help, you know, the governments who worry about this on behalf of the people to separate these things out, if they're going to pursue it at all.

PEPPER: Craig, let me just — in terms of the privacy data, again, some conversations I've had very recently, some very practical implementation questions about, you know, there are some countries that say all personal data on any citizen has to be kept within that country, cannot be shared.

MUNDIE: So I agree that's bad, too. I'm going to look — I'm well on the record about...

PEPPER: Oh, no, I know you are.

MUNDIE: ... you shouldn't be controlling the data at all, you should be controlling usage.

PEPPER: No, I know that. OK.

MUNDIE: Let's go right here — just to keep going. Sure.

QUESTION: Hi, John Mallery, MIT. So could you speak to the question of the security issues in mobile and lots of computing devices, particularly since we'll be increasing the attack surfaces, and so how that might be handled, any ideas there.

And then in specific, there are cases now where we have sort of power grids and other kinds of critical infrastructures relying on wireless for sensor data and control, which brings in a set of wireless vulnerabilities, and it's being done because of, you know, the cost is low, so any reactions in that area and where we're going? Thank you.

LAZANSKI: I can do this...

MUNDIE: Want to take security? OK.

LAZANSKI: I'll do security. Or you actually can do it, as well. But, I mean, I can...

PEPPER: Do you want to start it?

LAZANSKI: I'll just say a little bit about the security issues. So this is obviously a concern at GSMA and something that we — a lot of our members are working on. We have a security group that continues to meet very regularly to deal with these issues, because as the Internet evolves very, very quickly, these issues evolve extremely quickly, and they are tied to privacy and identity, mobile identity, and issues like that. And the SIM card, of course, being yet unhackable at the moment, but possibly we don't know about that in the future.

So I think one — I think the key takeaway for the security issue from a top level perspective is that the industry continues to be leading on these issues and continues to be working towards a place where they know users entrust is a big part of that, and that's the goal, right, because — and if you don't have a secure network or a secure device, your users lose trust and that's really, really difficult.

So I was just — I just wanted to highlight that in terms of a very industry-led, very internationally industry-led kind of work that's being done.

SPALTER: Can I just offer one perspective to add to Dominique's? And, by the way, I just read from a Twitter feed that the head of Germany's domestic intelligence agency just reported yesterday that they are receiving at least one hack a day from Russia, which speaks to the general anxiety and insecurity of networks, not only fixed, but wireless.

From an innovator's perspective, what is interesting is how security is actually leading to convergence — industry convergence in ways that one would never have imagined. We now have a major aerospace company that's become a telephone company. Which one? Boeing has actually deployed its own telephone, Black, which is now commercially available. And it's — the entire business model for Black is to provide a mobile apparatus. It actually will provide a level of security that heretofore, at least on a commercial enterprise basis, has never been imagined.

Similarly, we're seeing parallel Internet communities that are being evolved in much so the way that social media has around platforms like Wickr and others, that are based on the anonymization of identity and the security of identity in ways that we've never seen before.

I wonder, as those business models advance and as investment fuels, catalyzes those models, what that will actually mean for this increasingly polluted Internet that we have that is subjected daily to more security breaches. And if — where the interface of the governmental process is thinking about security will be with this new innovative approach to business model creation and technological advancement in the space.

MUNDIE: Let me just say I had one thought myself on this. I think, just like the privacy comment we made a moment ago, this is another one that has — you know, has — suffers from a lack of good taxonomy across the set of issues, because even in these remarks, you know, you hear people talk about different aspects.

I think one of the big challenges now is that at some level every device that's going to connect to the Internet, the Internet of everything, no matter what it is, that device is a computer. You know, if isn't a computer, it isn't connecting. And the big challenge we have is that the lessons learned, you could say, starting with the personal computing and server kind of world of computing, has not yet really migrated down into all the other computers. They're kind of where these other ones were a decade or more ago.

And so to the extent that it's a hack kind of security question, these things are really easily tipped over right now, and I think that, you know, there's going to have to be a bigger call for just, you know, improved engineering practice in the computing environment of every one of these connected devices.

To your question about, you know, wireless critical infrastructures, I mean, I put this in the category of whoever is designing these things, as like in any other business kind of consideration, has to think about, you know, continuity of operations and what the risk-reward tradeoff is.

I mean, if you say, hey, you know, this thing's got to keep running, what risks are you worried about? At some level, you could say, wow, how about an EMP, you know, weapon and it just wipes out all of the radios, you know, permanently? Well, then a lot of things are going down.

You know, if you go down a click and you say, do I worry that some bad guy gets near your critical infrastructure and has a jammer and just will locally jam this thing, well, that's a risk. You know, do you think that's a risk you have to worry about? And is that the right risk-reward tradeoff?

Today I just think our big problem is that, you know, we've got a lot of people bringing these Internet technologies forward. You know, they're lower cost, you know, they're really easy to deploy, and people are always under cost pressures.

We have not yet really evolved the standards of engineering practice. You know, in a sense, it's like having zoning rules or building, you know, codes. We don't have building codes for the Internet applications. And I think that that's, you know, a much more localized question, but is one of the things that's still going to have to come if we don't want to wake up one day and realize, wow, there were a whole class of risks we never thought about.

PEPPER: Yeah, so, Craig, a positive note on that — because I agree with you that that's the issue — last month, we and a number of other companies convened the second Internet of Things world forum. It was in Chicago. The year before, it was in Barcelona. About 1,500 people there, across all of the various industries. It was, you know, smart grids, you know, manufacturing, transportation, aircraft, mining. There's a big, you know, application for this, transportation, public sector.

When we asked the 1,500 people there what was their number-one issue that they saw as a potential barrier to the adoption — deployment and adoption of Internet of everything — it was security. You know, you look at the industrial Internet consortium, these eighty-five-plus companies, that's the top of the agenda. So this is sort of top of radar now and people are working about it. You know, it's easy to say security by design. It's much more difficult to do, especially when all of these devices have compute...

MUNDIE: I know that.

PEPPER: But, yeah — yeah, exactly. But the fact is, it is the number-one issue within these industry consortia and standards groups, right, because of the recognizing the vulnerability, by the way, being led by the electric utilities who have critical infrastructure and you talk to some of them like, you know, BC Hydro in British Columbia, you know, this is what they're doing and they have whole systems put in place addressing specifically the issue that you raised. So it's now an industry conversation.

MUNDIE: Over here?

QUESTION: Jeff Moss. So I'm curious to get a feedback from the panel on the idea of the — since we're talking about wireless and Internet of things and future innovation, I believe there's going to be a huge amount of innovation around this movement. But it seems we've got a long tail issue where these devices that get put in online or wherever, some of them will last, we predict, ten to fifteen years, will be their half-life.

So there's no way for these devices to update themselves, if they don't have a defined half-life or shelf life, where if they don't get an update, they just kill themselves, we're going to be dealing with fifteen-year-old vulnerabilities in the future. And this means we're going to have billions of connected devices online that have not been updated and are vulnerable. And so the security person in me is worried that, you know, with IPv6, we'll have the address space to put everything online. Once everything's online, if these devices aren't updatable, we're going to have a larger botnet than, you know, we've ever fantasized about.

So now with this manufacturing stage right now, how do we think about these issues? How do we — do we make sure that all Internet of things devices have to be updatable? They have to...

SPALTER: Talking taxonomy, about taxonomy, what do you mean by device? Are you talking about mobile platform, mobile sensors...

QUESTION: Anything that's — yeah, anything that's a compute device, as Mr. Pepper's been referring to them...

MUNDIE: Really anything that's connected to the network.

QUESTION: Right, it's essentially running software. That software is going to have a bug in it one, two, five years from now. Bad guys could use it against us. How do we make sure we're not building this infrastructure that is going to be installed and forgotten about?

MUNDIE: Do you want to take it?

SPALTER: Well, I — not being a product designer nor an engineer, I think that...

MUNDIE: OK, we'll give it to Bob.


SPALTER: But having managed technology companies and product design...

MUNDIE: Actually, I think I'm the only one who's a product designer and engineer.

SPALTER: Well, then you take it.


SPALTER: ... how you actually develop intelligent hardware and adaptable hardware and hardware that could be iteratively programmed where you can actually do status updates on the — on the hardware, not just the software. And it's an interesting question. I think it's a business model question for many of the OEMs.

MUNDIE: Now we know why you're not an engineer. His question is a really good one. And I'll say that the short answer is that, you know, the marketplace has not been demanding that capability. You know, and so it's a lot — you know, if you're two guys in a garage and you want to start your company quick, the last thing you want to stop and do is figure out how I'm going to build up the infrastructure to service this thing.

You know, if I can just stick it out there and people will buy it and they don't demand any of this from me, then it's not a problem. Today, look, for example, you know, at television sets. You know, I mean, in the last few years have been a rush to put — all the television sets have an Internet connection on the back.

But, you know, the reality is the TV set manufacturers of the world did not really stop and think, hmm, you know, have I done a lot to secure this thing? Have I really given it a great, you know, automatic update mechanism or anything like that? The answer was typically no.

And so, you know, the problem you're describing is not new. I mean, there are billions of noises out there already that don't update themselves or don't have a good facility for that.

So there's only two ways that gets fixed. Either, you know, a consortia come together and say, look, we're just not going to do this anymore, and it becomes sort of a voluntary or market-driven thing, or the society decides, wow, you know, it's really bad when we let somebody attach a device that can't be maintained to the public network, and then you put a regulation in. And then that kind of moves things along.

And I expect that we'll go through that process where, you know, once a really bad thing happens, you know, enough times, then government will come forward. The challenge here in this globally interconnected environment, again, per this conference is, how does the planet agree on a minimum level of serviceability for the software in every device that's connected? I mean, it's a really hard problem. Even GSMA, you know, has not been able to achieve this within the two levels of the software stack for the phone itself, all right?

And so, you know, I think whether it's sophisticated organizations or others, you know, it's going to be a big challenge. And when you try to project out into a global regulatory environment, I think it's going to be a big challenge.

SPALTER: But if your question, though, is how do you actually extend the shelf life of legacy products, which is what I think you were asking, I would counterpropose that given Moore's law, where things are, in fact, getting cheaper, smaller, and faster, that there might be disbenefits to thinking too seriously and long about that issue, in favor of actually advancing new product iteration, new product development, and new product advancement to markets that actually might need more affordable products, more affordable innovation. And, in fact, there's a lot of wisdom to the latter course than the former course.

MUNDIE: Yeah, I think he was actually also saying, though, that you're leaving behind this big tail. You can't — you don't have anybody who's forcing you to unplug the old ones when the new one comes along. And so, you know, you can't have it both ways, so you could say, hey, we'll go your model, and everything should be replaced with a new one because it's going to be better, faster and cheaper, but unless you forcibly take the other ones off, you're left with this legacy, you know, cybersecurity challenge.

Next question. Yes, sir?

QUESTION: Thanks. Fred Tipson, Institute of Peace. I'm having deja vu here with the Carterfone decision, where AT&T was fighting any attachment to the network because devices couldn't be vetted and made sure that they didn't do some damage, both with the customer and the network, and I think there's some throwbacks to that era. And Bob Pepper, who was at the FCC, may have some nostalgic reminiscences on that score.

But I wanted to raise this issue between standardization and intellectual property. And, Craig, you used the term de facto standardization, which I love, coming from Microsoft. But there is an issue that what the Chinese, for example, using standardization to advantage their homegrown version of technologies, to what extent are we facing a political battle here in the standards world to — where there's sort of a national competition to advantage certain kinds of companies and technologies, or can we hope that the ITU can rise to the occasion and avoid that kind of nationalistic competition?

LAZANSKI: So that's a really interesting question, because it brings up the issues of competitive standard development, as well. So within the ITU, ITUT is responsible for standards in technology, and there's a number of study groups that I participate in — I'm sorry you have, as well — on standards development.

And primarily, to answer your question directly, those people that participate — U.S., U.K., as well as China, Korea, Japan, right — so you see a lot of involvement in that particular area around standards development. Within the ITU, they often have study questions and they often discuss sort of future technologies and things like that.

So that that sits beside — as far as I see it — the IETF and other standards bodies that are more multi-stakeholder, collaborative, open, have developed from a grassroots, right? So the interesting thing is, China's probably going — I'm horribly generalizing, but China is probably using standards and going down the ITU standardization route, whereas a lot of the innovative technologies and the future technologies are really being pushed by innovators in developing standards that are, you know, bottom-up. And I think that's a really interesting point to kind of think about, right? It's not just one standards body. There are one standards body, but with the ITU, it's — you know, a lot of people don't look to the ITU for standards, right? So it's interesting

PEPPER: Well, I think you're raising a couple of different issues. One is sort of national standards versus global standards. And then, you know, when you have national standards, you have standards as non-tariff trade barrier, right, which is something that we've seen not just in this sector, in multiple sectors, and it's a big trade issue.

So there are some issues that are in this sector that are actually trade issues that, again, it's — where's the venue? Where do you deal with those? So — and there have been a couple of, you know, sort of China-unique standards questions, in a variety of the different technologies in this sector, that have been, frankly, real issues, as opposed to global standards.

When you get to this global standards process, you have the multiple standards organizations, but completely different processes, right? As Dominique pointed out, not only is the IETF, which develops the Internet for the — acronym, Internet Engineering Task Force, that develops the standards for the Internet, it's very agile, it's based upon what they call rough consensus. They reached rough consensus not through voting, but through humming. And when the humming in the room is sufficient, there's a rough consensus, then they move on.

You know, by contrast, within the — some of the traditional ITU study group standards groups, it took something like, what, seventeen years to develop the standard for ISDN. When they finished, nobody wanted it, and people said, ISDN? That stands for I Still Don't Know. Right?

Whereas in the IETF, the ability to have agile standards get it almost right, put it out there, and then fix it with iterations as you need to,is a much more dynamic agile model, and, by the way, it's completely global. So it's not only national versus global. It's sort of this broader consensus process versus a more traditional standards process that, frankly, you know, was built around traditional telephone standards — Carterfone's, you know, a good way to invoke that — you know, where most in the history, way back when the members of the ITU were state-owned telephone companies, the old PTTs, right?

And so it was a multilateral government PTT standards process around the telephone industry, which is very different — which is a very different process than the process today about Internet standards software, you know, agreements on software standards, interoperability, and so on. And that's where there's disconnect.

MUNDIE: It's closer to the issue of — I was saying earlier between sort of physics-based standards and software-based standards.


MUNDIE: And, you know, the older ones — I mean, Carterfone certainly, ISDN was all really more in the model of I'm fixing an electrical signaling standard, as opposed to I'm agreeing on a protocol by which two computers are going to talk. And I think that, you know, that's going to increasingly be one of these important issues.

I do want to add a comment on this issue of the non-tariff barrier kind of strategy. This is not new. I mean, you know, every government, including the U.S., Japan, and Europe, you know, at one point or another has essentially employed these strategies as ways, you know, to affect industry policy or to, you know, give some kind of favor to local manufacturers or whatever. I mean, television is a great example. How did we end up with PAL, NTSC and the Japanese standards, and then all derivatives of those?

I mean, they were each made locally, you know, for a variety of reasons. And, you know, the world suffered frankly for a long time with the fact that, you know, you couldn't move video economically from one to the next.

You know, the thing that's really interesting now is that the Internet, in my view, has taken the raison d'etre for most of those standards organizations, which was to resolve the natural incompatibilities that each region started with, and it basically managed to give the world sort of — while no one was looking, you know, a completely uniform way of communicating.

And now we're trying to figure out, well, what does that mean for those old things? Because we don't really kind of need them exactly for the reason we created them. So are they going to morph to the new problem, you know? Or are we going to — or is their nature going to be one that's going to try to sort of bring the new world back to the old world, and I think that is one of the biggest questions collectively that we face going forward, is to think about, you know, what are the new challenges?

Like these, you know, new spectrum models, for example, and the economic models around them. And, you know, they're very different and much broader questions than you get by looking at any one of the traditional organizations and deciding, you know, how are we going to subdivide the analysis or the — or the standardization activity?

SPALTER: I would just add that I think it's important and salutary that the Obama administration has decided to embed in its global trade agenda many of these issues regarding the Internet, as amongst many other positive impacts as a means of actions and gauging on these specific issues of tariff barriers in the virtual space. And that's an important evolution in the work of our trade agency.

MUNDIE: Let's go to the next question. ...Simple. OK, back there.

QUESTION: Hi, Jeff Farrah from the Senate Commerce Committee. Jonathan had mentioned spectrum-sharing earlier, but my question is for the entire panel. I wonder if you have any thoughts on how agencies can be — how they can be given the proper incentives to do either sharing or relinquishment of spectrum? And then given the theme of the day, what models do we see internationally that might be working better than what we currently have?

MUNDIE: Well, first, I'll take my prerogative to say you should probably read the PCAST report.


PEPPER: Which you wrote.

MUNDIE: ... which I helped write. But because — at least for the United States, you know, it specifically addresses this. You know, President Obama has issued an executive memorandum directing it, and U.S. government agencies are all operating under, you know, this, you know, movement toward a new incentive model for sharing.

So, you know, I would say, you know, there is a lot of work that's been done there in the United States. I think I'll just riff or add to your question and then ask the panel to address it. You know, what does the U.S.'s — I'll say — leadership on this, you know, mean for the rest of the world? You know, is everybody else looking at that and saying, wow, that's a great idea or it's a terrible idea? You know, how do you think that that is going to play out? Because it's clear that the U.S. is farther along both at the FCC regulatory level.

The reason it's an interesting question globally, as well as in the United States, is because in almost every country, huge part of the spectrum under the traditional allocation model went to the government. And then the rest went to the non-government.

And like in the United States, the FCC oversees the non-government part, but, you know, a different group, MTIA, basically oversees the U.S. federal uses of this. And in this new spectrum-sharing proposal that was put forward and embraced by the administration, and by the FCC, you know, it's really building a bridge where you get both government uses essentially being modified in order to accommodate sharing with commercial uses, and that's a way to get more effective utilization.

But it does require — you know, if that's going to become a global standard, that other organizations and other governments are going to similarly have to decide that they're willing to — you know, to take a similar step. So I'd be interested in people's thoughts on that.

SPALTER: Just last week, I was able to speak at an international spectrum management gathering here in Washington, D.C., and we really joined this issue very carefully. And I think there was a large group that was interested in what the U.K. specifically has been doing in Ofcom in providing an administrative fee, in a sense, where agencies actually have to pay to play for the spectrum that they're allocated, a model that would take quite a while to evolve here in the United States, but has merits.

There are a number of other incentive structures and opportunities including allowing any savings that might be accrued by an agency that uses or optimizes its spectrum more efficiently to not have that savings offset in future budget allocations, which often is a politicized process. We can do that — in fact, you can do that working with industry and — and the administration through your franchise at the Senate Commerce Committee.

But I would say that sharing of actual not only spectrum, but sharing of new thinking that's going on within our government I think portends extremely well for the idea that we're going to see progress. This last week — and we're in the midst of it still — there's a spectacular evolution in what's gone on in our spectrum management. We are now in our first auction, the Auction for Advanced Wireless Spectrum-3, since 2008. It offers potentially to bring sixty-five megahertz of new spectrum to American consumers. It has already been able to pay down the cost for an interoperable public safety network. But what it really represents is a tremendous achievement by the administration at the Defense Department, the Commerce Department, the FCC and the White House working with industry and Congress to show that, in fact, things can get done in Washington with regards to the idea that one can relinquish goods and assets that an agency has. I hope that that will serve as a template for going forward.

I'd just ask — add one last point. We think of incentives particularly in the United States on the conversations we've had on a macro budgetary level. How do you actually bring return monies to agencies that might risk losing it, if they — if the architecture for incentives are not done well.

I also think, as we saw in earlier attempts and earlier administrations here and elsewhere, to create incentives for greater performance in government agencies, for example, the reinventing government initiative that President Clinton and Vice President Gore undertook a few years back, that there is a profound and important human and personnel component, as well, that the Office of Personnel Management and creating specific incentives for civil servants to continue to perform and focus on this issue across administrations is something that also I think has to be added to the agenda, if we're going to have a sustainable path to creating viable incentives for sharing with things like — important things like spectrum.

PEPPER: Craig, if I could pick up a bit, so — especially on the U.K. piece, you know, I have the privilege to sit on the Commerce Spectrum Management Advisory Committee, so we're trying to implement the PCAST report, and we're making real progress under the leadership of Larry Strickling at NTIA.

I also sit on Ofcom's Spectrum Advisory Board. So I see sort of both sides of that. But specifically, picking up, Jonathan, on what you said, one of the things that Ofcom have done very effectively is an administrative pricing. So even the Ministry of Defense — other than for weapons systems — right, there are specific spectrum dedicated for weapons — but all other spectrum, MOD has to pay for. All of the other government agencies, there's a price they pay, so they actually look at the spectrum and what it's costing to use or, frankly, not use, what the tradeoffs are for becoming more efficient, and then return spectrum.

So that's not so much the sharing. That's the refarming, creating the economic incentives to refarm. And it's taken some time to actually get implemented, but government departments are actually returning spectrum, including Ministry of Defense, and that spectrum then can be repurposed for civilian use. So that's spectrum refarming.

On the sharing, you know, that — some of that is kind of a refarming, repacking, but some of it is the opportunistic use of spectrum that may have been allocated for one purpose, but it's not actually being used, so there's a difference between allocation, assignment and use, and the question is, with some of the newer technologies, can I use spectrum that might have been allocated for something that is only used occasionally or it's only used in certain geographies? Can I use it opportunistically now for commercial use so we get much more efficient utilization? And that's some of the things that we're looking at, at Commerce on the PCAST implementation.

One of the big questions that's come up that we're grappling with is the enforcement. How do you, you know, make sure that if there's priority service, it's only used at certain times by a government agency, but there's commercial use other times, how do you make sure that there's the enforcement so that you don't interfere when the government has a critical use and it needs it? That's not a showstopper. That can be done. But those are the kinds of conversations that are now ongoing.

MUNDIE: I'd just close the comments on this by two other thoughts. One, the issue of incentives — I'll just use the U.S. as an example — the spectrum change that we're trying to make here will take about twenty to thirty years to engineer, because you have a lot of these radars, like all the satellites, for example. You can't, you know, fly a mission up there and decide to change the radio out in the satellite, and then the satellite, you know, isn't coming down for a while. And so, you know, when we look at this, it requires a very systematic and very long-term view.

Part of the problem with incentives in this country is the kinds of things that they talked about in the U.K. are kind of — are legally prohibited in the United States. And reciprocally, the annual budgeting cycle doesn't allow you to give multiyear incentives, but you're looking at a many-year project for most of the government people to move from where they are to where you want them to be.

And so, you know, the report does actually propose a way to do this kind of incentives within the legal framework in the United States, but certainly I would encourage the Senate, you know, and the Congress at large to look at some of these long-term incentive changes.

I think these are issues that we're going to face in the military writ large, and this just turns out to be one essential feature of that, but it affects many other agencies of government, too.

SPALTER: Including offsetting budgets that actually can show savings and efficiencies, not offsetting those budgets.

MUNDIE: Yes, sir?

QUESTION: Pei Xu from Voice of America. Again, I want to ask a perhaps oversimplified, you could say essential question, that is, with the advancement of technology, some — quite a number of people — simply say that we are too connected. Our memory, or attention, is devastated by technology. We have the report saying that Steve Jobs banned his kids to use iPad, and we know that here in Georgetown University, a very famous professor banned his students using mobile device in class. So I wonder what your take — the panelists' take on this question?

MUNDIE: What's the governance issue? You want to know whether we should regulate that out of existence or...

QUESTION: No, I want to ask a very general question. That is...

MUNDIE: No, but this panel is about governance, so I'm trying to make you be a bit more precise about the governance question you're asking.

QUESTION: Yeah, my question is that because the discussion is the technological change and the move to mobile, and some people say that we are too connected, perhaps such connection is not good for us. So what's your opinion about that question, if — yeah, this question is...

SPALTER: Well, there is domestic governance components to what you've addressed. There are school districts around our country that specifically prohibit the ability of students to use mobile platforms in classrooms. I think that it's — you ask a philosophical question, if you point it in a more global, multilateral dimension.

I would say, however, that I think that the advancements that we're seeing, particularly in mobility and innovation mobility, are transcending the traditional construct of what we've imagined these platforms are bringing us, which is human to human interaction only or engagement with entertainment or engagement with more conversational or recreational uses.

Every day, using a mobile platform, I get to have multiple reports of what my daughter's blood glucose levels are from a wireless sensor that she actually wears on her insulin pump. These are the kinds of implementations that we're going to be absolutely desperate for, not only here in the United States, but particularly in societies that have not been able to advance or engage with these spectacular steps forward, and a governance model that will accelerate and catalyze such advancements, I think, is exactly why we're gathered here today and why we need to focus on the alternative, new and potentially more innovative transformative, life-affirming uses of these platforms than what we've come to understand. What we need are not angrier birds, but healthier people.

MUNDIE: Anybody in the audience want the last question? OK, then I'll give the panelists a chance to each have one minute to give their final thoughts about what the biggest challenges are that we have related to mobile and governance.

PEPPER: Well, I — you know, we've talked about the need to — if everything's going to be connected and the majority of those things are going to be connected by wireless, it is about spectrum, it is about spectrum policy. It's going to be about licensed and unlicensed. We need all kinds. And it's moving to the new model of how are we going to better share where that's possible and where it's not, you know, how to get more, you know, big chunks of spectrum for, you know, the — where we actually need the big chunks that are dedicated? And...

MUNDIE: Well, since you and Dominique both mentioned this, I think maybe all of you did, about the licensed version unlicensed, I mean, isn't the generalization of that for the future just a question of priority, not — you know, licensed versus unlicensed. What does licensed give you? It's just...

PEPPER: Oh, the question is — yeah, the question is it — it gives you priority and it gives you...


MUNDIE: Absolute guarantee of priority.

PEPPER: It gives you a...

MUNDIE: But it has other costs.

PEPPER: ... virtual — virtual dedication. And that's the cost. So you pointed out that we're in a transition that's a twenty-year transition. And there may be technologies in the twenty-year timeframe that will allow us to aggregate, you know, if you want broadband, you need broadband, big chunks at a particular moment in time over, you know, a distance.

Between now and getting there, I do think that at the moment, given the current technologies, whether it's — you know, let's just talk about the LTE what they call 4G or some of the discussions about the 5G, we're still in a world where we're going to have dedicated chunks, and then opportunistic use of those dedicated chunks may be by other users negotiated with the primary licensees.

That's the next step of how you get more efficient use and licensed. And over time, twenty years from now, it may be a completely different world, but at the moment, the cognitive radios, the software-defined radios, the things that we look to for a sharing of big blocks of spectrum for the really high-speed broadband connections over long distances — also including back haul — we're not there yet.

And so I think it is about a transition, and I think the twenty-year timeframe is the right — probably the right time.

MUNDIE: But, Dominique, sorry, let's just — in your comment, address this thing, then. You know, just as a way — no, because we're trying to help people understand. You know, I mean, we're really trying to provide guidance to people who are not really expert in these areas. We're trying to help them think about the future.

So one of the things that I think is interesting in this regard is this issue of, how do we start to get the industries — all of them — you know, to move as rapidly as possible in this direction? So I mean, I'll just posit as a thought experiment.

If I went into all of your 800 companies and said, you know what, the governments of the world got together, they all had an enlightened moment, and they decided that they were going to increase the WiFi allocation under the existing rules by a factor of 100, OK, what would your 800 companies do?

LAZANSKI: Well, I think it's an interesting way that you're starting from. You're positing that they're not doing anything right now. I mean, that's — that's...

MUNDIE: Well, I'm not — no, I know they are. I'm trying to get you to help people understand.

LAZANSKI: I mean, and I think what — what they would do is do what they're doing now, move it quickly, innovate, you know, either work together or work competitively against each other...

MUNDIE: But they'd go use it.

LAZANSKI: Right, absolutely, but that's the whole — yeah.

PEPPER: That's why, by the way, we're proposing huge chunks of more WiFi spectrum up at five gigahertz, which is complementary to the licensed service providers and, in fact, the licensed service providers are using that spectrum for outdoor WiFi to extend what they're doing, so it's an architectural issue. So it's not either/or. It's both.

MUNDIE: No, of course. And the reason I framed the question for you that way was, if you just opened the newspapers around the world for the most part, and looked at what the regulatory discussion is, it's historically been about, "Give me more license spectrum for cell towers." Right? And — but the free part, which was essentially WiFi at 2.3 and 2.5 and all that stuff, well, there's like no discussion about it at all, right?

And so I'm just trying to make people understand that, yes, indeed, carriers globally are already, you know, moving to take advantage of these things, but it's not, you know, part of the regulatory public discussion very much.

SPALTER: Can I just offer one potential, modest concept that's actually beginning to get some traction? We know that change is hard. We know that it's particularly hard when there are sovereign prerogatives involved regarding spectrum. And also, successful business models that are anticipating disruptive technologies upending them.

But there is the concept, though, that as we proceed towards — in this transition and as potentially — and I say potentially — dynamic sharing, cognitive radio, and other types of technologies begin to get critical mass and scale, which they are not yet at, that we apportion in every — in every liquidity event around spectrum, every auction a portion of those proceeds for a dedicated R&D pool that would be used purposefully by our nations and other nations' R&D centers, academically, institutionally, and corporate world, to advance greater progress in the iteration and development of the new technologies that will define a software definition of the future.

MUNDIE: I think that would be a great thing to do. I think, you know, what we also know is that the thing that drives all these things the farthest, the fastest, is big commercial deployments, you know, and so whether it was WiFi that got a big push, you know, because people wanted local connectivity or the carriers, you know, that created the blanket of connectivity that covers the planet beneficially today, you know, using those engines to drive the world toward, you know, a new model of interconnection, and an Internet that's primarily going to be wireless, at least for its end point connectivity, I think would be a great thing to do.

So let me thank the panelists for their contributions.



Top Stories on CFR


Tens of thousands of unaccompanied children have been arriving at the U.S.-Mexico border each year, sparking debate over how to respond. A recent rise in their numbers has challenged the Biden administration.

Technology and Innovation

To avoid falling behind the rest of the world on internet speed and access, the United States should rapidly deploy fiber-optic internet nationwide. 

Pharmaceuticals and Vaccines

Only a year after the new coronavirus emerged, the first vaccines to protect against it are being administered. But production challenges, vaccine nationalism, and new virus strains are all presenting hurdles.