 Okay, let's get started. There's still some folks are getting food out there, but they can probably hear me or drift in in any event for the main attractions, which are the speakers. So welcome to Wireless Fiber, the quest for affordable gigabit fast broadband everywhere. I am Michael Calabrese. I direct the Wireless Future Project here, which is part of New America's Open Technology Institute. And as I think most of you, so many familiar faces, most of you know, we have lots of events concerning mobile broadband, spectrum reform and related topics. But interestingly, when we sent this invite out, a tech reporter asked me, isn't wireless fiber and oxymoron? You know, are you talking about wire line or wireless? And well, you know, my response was it's purposefully a provocative oxymoron, because in fact, we're talking about both wire line and wireless. Wireless fiber is telecom shorthand for what is really a hybrid fiber and wireless solution to getting high capacity, very high capacity connections, you know, that last 1,000 feet even longer without trenching, which is very expensive. It is fixed wireless, not mobile, that Yahweh and some other companies are calling wireless to the home. But wireless to the home sounds like a couple of my friends who use, you know, 4G instead of DSL, and that's not what we're talking about. We're talking about gigabit fast connectivity. This technology highlights how wired and wireless networks are complementary. Access to fiber running under a street or on the rooftop of a fiber fed building becomes the jumping off point for extending fiber fast internet access to other locations using point to point or point to multi-point radio technologies, which you'll be seeing and hearing more about shortly. Today we have a good mix of almost kind of half and half of startups and big high tech companies, all of which see enormous potential in substituting wireless for wire line over the last 1,000 feet or even further. What stood out to me and what was really the impetus for this event is that these companies are using Spectrum with very different propagation characteristics for this purpose. For example, as you'll hear Mimosa is using unlicensed mid-band Spectrum at five gigahertz while Verizon is relying on sort of what you might call the lower portion of the millimeter wave, 28 gigahertz. Starry is at 37 and others are in the unlicensed and likely licensed bands at 60 gigahertz and higher up into the 80s. The companies have also advocated for very different Spectrum access rules. Some unlicensed, some exclusively licensed, some lightly licensed and some based on the sort of open dynamic sharing that the FCC adopted in its final order this year for the citizens broadband radio service at 3.5 gigahertz. Of course, none of these frequencies or access regimes are necessarily incompatible but we'll hopefully hear more about the trade-offs that our speakers believe will either facilitate or deter deployment and affordability. Because wireless fiber is a rather novel concept and because these companies are all approaching this challenge very differently, we'll start with instead of having just kind of an interactive discussion from the get-go, we'll start with five or six or so minutes of opening remarks from each of our speakers. Then they'll come up together to the panel table and we'll have a more interactive discussion and close with your questions and comments. So please be thinking about what sort of, whether it's a clarification or a provocation that you wanna throw at us during the last segment. So to lead off and the speakers will go in the order they are on the bio handout, which is on the table if you didn't get it. In fact, please check in if you did not just even slip out to let them know you're here because we like to keep attendance and make sure you, and especially if you're not on our, if you didn't get this directly, that you're on our list for future events. So first up is, and I'm not gonna make long introductions because you have the bio, they're bios, but first up is Boris Mazel. Boris is the director of business development at SICLU, a startup. He has been involved from the outset in all aspects of SICLU and particularly with working with carriers, service providers, cities and municipalities to get it rolling. So Boris. Okay, thank you, Michael. I'm really glad to be here and to share our perspective on the millimeter wave. My name is Boris Mazel. I'm the director of business development in SICLU. We are vendor of equipment in the millimeter wave in the V-band and E-band. We'll get to that in a moment. I'll start with a short story. I participated at the conference, V-SPAPALUSA about two months ago. It was in Las Vegas and this is a conference of mainly focused on the V-SPAP market, wireless internet service providers. Very interesting market, very open to new technology. And the conference ended and I had a flight at 6 a.m. from Vegas. I took Uber and 4 a.m. I'm sitting with the cherry Uber driver and he starts asking me questions. What are you doing here? What's the purpose of your visit? And well, I participated at the conference. What was it about? And well, it was about internet, about connectivity. And he says, you know, there is a new thing in the connectivity and 4 a.m. Vegas Uber driver. And I'm saying, what is that? And he says, millimeter wave. So this is what we are going to talk about here. And I'll start with fiber. So the fiber is an essential part of any communication network. And I just took a picture here that is publicly available at the site of the city of San Leandro in California. About 100,000 residents there. And you can see that they have plenty of fiber, several tens of miles. But the question is whether this fiber is inclusive, whether it solves all the city needs, whether it connects all the parks, all the anchor institutions, all the SMBs, all the long income neighborhoods. And the answer is probably, well, it is there, but it's not dense enough. And this is how we envision millimeter wave. Millimeter wave technology is not a fiber replacement. It's a fiber, it's a complementary to fiber. This is the densification technology that enables fiber reach. And in some cases, it's fiber enabled. So in this hybrid fiber wireless topology, fiber can go to one of the building, it can be commercial building, it can be MDU, it can be anchor institution. And there are plenty of different solutions in the millimeter wave that can be used as an aggregation, can be used as an access point to point, point to multi point. It can be different capacities. Today, there are systems available, general available in the market that can provide multi gigabit, five gigabit capacity. And can be installed at the street level, can be installed at the rooftop. But the bottom line is this technology doesn't fight fiber. This technology complements fiber. Some of the technical aspects of this technology, basically there are three main things that can describe a millimeter wave. First, there is virtually no interference. Second, this technology can provide really fiber like multi gigabit capacity. And it can be rapidly deployed due to the regulation and some of the equipment properties. In terms of the spectrum, we mainly focus on two main frequency bands inside the millimeter wave frequency. One that is called V-Band, which is the 60 gigahertz band. And the second one is the E-Band. There are two bands. The V-Band frequency is, it was 57 to 64 until FCC released additional seven gigahertz of spectrum there. So now it's 57 to 71. It has 14 gigahertz of unlicensed free spectrum, which is more than all other bands, licensed, unlicensed, even more than all other bands put together. This frequency with today available technology can provide 56 gigabits of capacity. And we are just at the beginning. The millimeter wave is pretty new technology. It's just developing. E-Band is 78 gigahertz. It has only just 10 gigahertz of lightly licensed available spectrum. And with the currently available technology today, it can provide 20 gigabit. So in terms of comparing this technology, this spectrum to other frequency bands, it has 10 times wider channels than, for example, sub six frequencies, for example, 5.8 Wi-Fi. And the antenna being with is 10 times more narrow. And this what gives this technology immunity to interference. You have a lot of spectrum. You have a lot of channels to choose from. You have wide channels. So you can use lower modulation to get to gigabit capacities. And antennas are very narrow. So the spectrum you use actually becomes very simple, very simple task. And this is what is available in this frequency band. So this slide should illustrate the rapid deployment and the number of links it should have an animation, but I assume it's PDF, right? So you don't see it here, but we have from 2012, which is this picture. We didn't have any unit installed in San Francisco. Now it has thousands of units deployed. And one of the, probably one of the service provider that you heard about web pass was acquired by Google. They have thousands of customers served by this technology in San Francisco area. Another project that we've done together, and this is an example of public-private partnership that we've done with the city, is deployment in Santa Cruz. So within just below 10 weeks, we connected working together with the city and the service provider, local service provider there. We connected less, within less than 10 weeks, we connected 18 locations with providing gigabit capacity. There were SMBs, there were community centers, there were MDUs. So this is an example of how this technology can be used in public-private partnership. And this is one of many other examples that we've done together. So just to sum up, this technology, millimeter wave, can provide a competitive advantage to newcomers. For example, we see it coming back to our Uber driver and Wisp. We see as Wisp wireless internet service provider use this technology to provide very fast, very competitive internet connection. It complements fiber, it enables fiber. It lowers the investment. Now when you use this technology, you actually start serving customers from the existing fiber pop. Now you know exactly what the demand for the broadband is. So when you need to put additional fiber, you don't need to guess what the penetration rate will be because you already serve customers with the fiber-like capacity. It's a future-proof, multi-gigabit speed, fast deployment. And we have tens of thousands of units deployed worldwide and here in America as well. Thank you. All right, thank you Boris. Next up is Milo Medin, who is the Vice President of Access Services at Google, where he initiated Google fiber. He served, among many things, served under President's Council of Advisors of Science and Technology, the working group on Spectrum that led to the recommendation on 3.5 and is here in fact for the FCC's Technical Advisory Committee meeting, which is tomorrow. Milo. Thank you. So I actually don't work for Google, I work for Alphabet, which is one of those funny kind of things. I work in the access piece of Alphabet, where our mission is to provide abundant access for all. So what does that mean? That means higher speeds, more availability of faster broadband. And I think when we talk about Spectrum, it's important to recognize that at the end of almost every broadband network, there's wireless, because we have Wi-Fi. That's the dominant mechanism for how people connect to the internet in their homes, as well as their businesses. It's funny, I used to work when I was in college at Livermore Lab on nuclear weapons design, and which is a fun industry to work in, by the way. I once asked a system analyst at Livermore, I said, what language do you think we're gonna be programming in in five years? And he said, I don't know, but it'll be called Fortran. And for those of you who are old, we laugh at that, most of you don't know what Fortran is. But the point is we use the same words while the underlying technology changes. When we built Google Fiber and launched it in Kansas City, we worked very hard to deliver a real gigabit to the home, but we had to work even harder to deliver a gigabit through the home. And because in the end, the devices inside the home are connected wirelessly as well. The kind of spectrum you use depends on what job you're trying to, what service you're trying to provide. So if you're trying to provide large-scale coverage, a low-frequency spectrum is best for that, because it goes through things. If you're trying to provide higher and higher capacities, faster speeds, you need higher-frequency spectrum to be able to do that. You don't want it to go too far because you wanna reuse it often. And the faster you go, the more spectrum you need, and that forces you into the higher-frequency bands because that's where the spectrum is available. So at Alphabet, we have focused on a number of efforts on the spectrum side, partly in terms of the services we offer, not just through Google Fiber, but as the gentleman, Boris, mentioned from SICK Glue, we acquired WebPass this last year who uses wireless extensions to connect up apartment buildings. We've spent a lot of work on trying to empower sharing a spectrum, because the key thing that we need is spectrum. That is the raw material that all the wireless networks use. And one of the things that I, it's interesting, I sit on the Secretary of Defense's Defense Innovation Board, and it's interesting as I've gone around to military bases and talk to DOD folks, this by way is my personal opinion, not that of the DIB. They always say we need more bandwidth, right? They need more wireless bandwidth. So there is demand on both the government and the civil side for more spectrum. And the question is, can we actually advance policies that make spectrum more abundant, more usable for everybody? And I think we can do that. Part of the dynamics that we need to see are actually driving higher speeds. I think a lot of, if you think about capacity as an engineer, we solve that problem through reuse. But if you actually want real aggregate speeds, speeds increases, you need wider bands of spectrum. And so that's part of what we think the work at 3.5 has opened up 150 megahertz of spectrum for sharing down there. If you look below, there's additional radar spectrum that could be shared and above. There's the C-band downlink, which is also another block of spectrum that's relatively lightly used. So I think we've got a lot of bands that we can make available through policy and in the end, make speeds to consumers faster, give consumers more options for connectivity. We hope to be a part of that conversation, both in terms of delivering service and the general policy goal of making spectrum and broadband more abundant. So thank you. Thanks, Milo. Next up is Sharla Rath who is Vice President for Wireless Policy Development at Verizon. And Sharla's particularly focused on alternative spectrum auction mechanisms and developing sources of spectrum for future needs, including 5G. And she's also my colleague on this Commerce, Department of Commerce Spectrum Management Advisory Committee. Sharla. Thanks. Actually, I'm happy to be here. I think even if I weren't out on the panel, I'd wanna be here to listen to what everybody else has to say about this. It is interesting. One of the things that I'm gonna talk about is how our wireless fiber are what we're doing in 28 gigahertz is actually sort of a precursor in ways to work that we're doing to advance fifth generation networks. So more than a year ago, we formed the 5G Technology Forum, which now includes a number of our vendors, Apple, I always have to feel like they're no longer the seven dwarfs, but I always have to make sure that I name all of them. Apple, Cisco, Ericsson, Intel, LG, Nokia, Samsung, and Qualcomm. Plus we have a number of other venture capitalists that are helping us out with this. We did this about, it's almost a year and a half ago now because we wanted to move forward on full 5G connectivity. I mean, this is the nature of our business. If you're a wireless operator, you're always thinking about the next generation of technology. What's interesting here is what the definition is of 5G number one and number two is you will hear us always say is that 4G is, there's a lot of life left in 4G as well. And I think what we're gonna find as 5G develops that it's less the kind of network advancement that we've all become used to us. Those of us have been in cellular for a long time. And more about just a whole ecosystem that we'll develop and it will include parts of what we're a lot of what we're already doing but just more of it. So what we also wanted to do, and I mentioned that we have venture capitalists as part of the 5GTF. And it's because we knew that as part of what we were doing too for 4G, we would need to do the same thing in 5G which is really think about developing an ecosystem and create sandboxes for people to come in and actually work with the technology and see what kinds of services and products they could develop. So that's also part of how we view moving forward on this. So nobody yet has really talked about what their view of 5G is and I'm not gonna say this is a Verizon view but what you'll see as you look through the literature it is high capacity, very low latency. And I think most people in this room know what that is but what I like to just to describe it as is we think about latency now in terms of our visual interaction and visually we can actually see things fairly quickly but not as fast as when we touch something if it doesn't respond in a millisecond we perceive a lag. And so what this is about, if you wanna think about it, low latency is about touching a button here and having something move in a way that makes you feel like you've actually moved it. And so that's a key component. There are also things that we see clearly densification. That's occurring now in 4G but it'll occur in much greater degree in 5G. Also things like very long battery life. We'll see more of the ability, high capacity. I talked about but it's also not just high capacity it's high capacity in a very dense environment. So it's sort of like everybody is getting high capacity. So this is what we were viewing and basically we started thinking about okay what we wanna move forward, what frequencies? Well it had turned out that the timing was very good. We've already mentioned the technological advisory council. They had advised the FCC only about three years ago that they should start looking at millimeter wave technology for this new, this coming generation. And I know several people here have already got products in these bands. But the commission actually, what they wanted to do was take a look at the bands and see whether they could also be used more abundantly for terrestrial mobile. So right around the time that they were actually putting out an NPRM we were also looking at what frequencies do we wanna engage in. And I know Michael will probably talk about what about low frequencies. We've all talked already about the fact that the low frequencies have certain characteristics that make them very beneficial to certain types of services. But what we were actually seeing is a need to have very high bandwidth. And a couple of people have sort of alluded to this but let me say it again that what the commission did this July was actually provide in terms of licensed spectrum alone four times the spectrum, actually more than four times the spectrum that is available for use by wireless operators today in terms of the amount of bandwidth. So it's a huge amount of bandwidth that comes with some very interesting technical difficulties. I've been around long enough that I was actually at the commission when we first started looking at these frequencies for terrestrial use. And in fact the idea was is they were gonna be used for cellular like television. The idea that you would reuse the frequencies and provide a television service to a lot of and this goes back a ways. I'm not gonna tell you how long you can look it up and figure out how old I am. But it actually it really has been in the last several years that these frequencies have been able to use in ways that really make a lot of sense for terrestrial mobile. And that's because of a lot of the technology that's being developed that certain I'm sure people are gonna be talking about on the panel but also I have to give a shout out to people like Ted Rappaport at NYU just sort of pushing the envelope in terms of various pieces of the technology that actually will allow us to use these higher frequencies in ways that we really weren't able to before. So that said what we did is we actually did spectrum lease deal with one of the holders of those frequencies that we've now been, we did an experimental license before but this summer the commission actually gave us permission on the lease in 28 gigahertz which technically is actually not part of the millimeter waveband depending on who you talk to but it's close enough that I count it. And what we've done is in the last year we've done a lot of real world testing. There have been a lot of laboratory testing on this and then we've gone out and we have actually put it in apartments where we have multiple, you know when we talk about all of this there's a lot of it if in fact if you get on 5gtf.org we talk about some of it and we also talk about some of the specs that we've developed in this area so I just urge you to do that if you wanted to learn a little bit more but we have an apartment for example outside of Dallas in a place called Euless, Texas where we're actually trying to see what happens in different types of topography but also when you have multiple users and you might have something that blocks a portion of the stream that's coming from the node to the user. We're also incredibly interested in making sure that what you wind up with the customer is that it's very easy because one thing people talk about is that trenching fiber is actually expensive but what is also expensive is truck rolls to a customer. So what you wanna be sure of is that you've got a piece of equipment for the customer that will actually help them figure out the best placement in their house. So all of this, all of what we're doing right now is really as Michael alluded to is really about fixed services really what I would call more nomadic than fixed because we actually envision things like if you're lucky enough to have a second home and we had service in the area where the second home is you just pick up your device and bring it with you so it's not like it's fixed in your house you can actually bring it with you but true mobility, the idea that you can actually walk across and move from, and when we talk about mobility a lot of, if you're in a technical world you know what it means is really being able to see and slice through and actually see the person all along even when they switch from node to node. That's a ways away, we're working on that a lot of what we're doing is actually leading us to understand more about that piece but we do know that our pre-commercial deployment next year is actually gonna be, it will be fixed but what we're seeing in the mobility side at least from what I've heard from some of our network people is that it's raw but promising in these bands and again I wanna keep going back these bands are very high frequency and they don't lend themselves as easily to mobility but because of the technological developments in both the antenna arrays as well as being forming and massive MIMO which I think other people are probably gonna be talking about we're able to actually do some things in these bands that we weren't previously able to. So I think I'll stop there, I don't know whether there, I think a lot more will come up during the course of our conversation but I wanted to just give you kind of an overview of what Verizon was up to and what we're doing, thanks. Thank you, Sharla. And next up, Jamie Fink who is the co-founder and Chief Product Officer of Mimosa Networks which is out in San Jose, right? Yeah, right, okay, Santa Clara, near San Jose. I've been there but at Mimosa, Jamie has developed technologies such as massive MIMO and wireless spectrum reuse technologies designed to dramatically increase network capacity and scalability and we will talk about that. Okay, we're waiting for a few slides to come up but I'll start with a little bit of a story here. So we've been, we've just, we're about a four year old company and I think four years ago when we got into the business we were kind of called crazy to be thinking about using fixed wireless for internet access and that was only four years ago a startup was struggling to get funding, getting after, solving this problem of getting lower cost internet access. So fast forward a couple of years, one of my favorite things that I saw in the field with a client of mine out in Minneapolis, they were a fiber provider in the city and they had had a special deal with the city to get access to pole rights, all the things that you need to do it right and they had deployed fiber services in competition in a cable market so they had to be very low cost so they were out pushing $30 fiber services and giving 100 megabit plus to their customers and they found that that was still very, very hard for people to understand that $30 is it worth getting 100 megabits? So they started dabbling with wireless and they started going after a new architecture type that I'm gonna talk about today and they decided they were gonna offer a 50 megabit or sorry, a 25 megabit for $25. So 25 for $25 or $30 for a fiber service that costs them $1300 to deliver to that house. So if you think about that you go, oh well that's great, I'll take the $30 offer. Well, we're out of touch massively because 90% of his customers went for 25 for 25. So that tells us that the customer's demand is not always about speed, it's usually about experience, latency, I think as people have been talking about but cost and competition is critical and we're dealing with very different socioeconomics. So I'm gonna try to touch on a little different areas. I think most people have been talking more about millimeter wave today. Based on my experience with this, I think we started it from a very different perspective. And I call this kind of why is sub six gigahertz critical for wireless fiber? Because the last mile is mostly trees. This is where most people live when you're looking in suburban areas. I'm gonna differentiate between urban glass lined buildings versus suburban areas because I think that's a very easy way to look at it. But really if you articulate this problem and you're trying to get in close with your customers, the real problem is look at all the trees that we live amongst. How are we gonna get from point A to point B that's less than 300 meters away and you're having to penetrate through four or five sets of different trees and branches to get this done. This is a real life example of kind of that poll that you saw in the last example of what you have to deal with when you start to get into these environments. And this is the real challenge that we have to solve. And right now, if we look at the physics and we look at the math, the only thing that really makes it through trees is lower band technology unless you have lots of things to reflect off of. So the cities, the glass environments make perfect sense for a millimeter wave technology, probably not so much for suburban areas in our practical experience in the field. So this is where we kick off what we call at Mimosa a micropop or a micro point of presence where we're instead of doing tower based technology we're fixed wireless as most people remember back to the YMAX days 10, 15 years ago, it was all being done from towers. And what we were doing at towers was over radiating the population. I could never service all the people within range of the tower that might have wanted service because I just simply couldn't provide that much spectrum to get the job done. So it was over radiating the population density. So the goal of the micropop is actually to match effectively the transmit power levels and the coverage that you get from your radio technology to the population density. What does that mean? Well, the way that I like to look at it is we need to be roughly speaking between 250 to 500 meters away from customers with this architecture. And this allows us to make sure that there's plenty of loss in the link budgets and wireless to make it through some branches and trees. And oftentimes that's not even gonna be good enough and you need to have a dense enough constellation of access points in the suburban area to have alternate ways to aim and point. And I think Charlotte said it really well. Now you have to have this really great experience of I have to be able to teach customers and installers what's the best way to go. And it might not always be the shortest path to a base station that we're seeing in the field. But given a little bit more context to this, so what markets does this mean? Well, it's about 1,000 households to 2,000 households per square kilometer, excuse me, square mile. And what does that mean? Okay, Salt Lake City, for example, is right at about 900 to 1,000. Cities like Austin, Texas are about 1,000 homes per square mile. San Francisco, 2,000. That's why my friend over here from Siklu uses millimeter wave to go building to building because people live in multi-dwelling units in those markets. So what I'm trying to really address is these places that probably were old DSL neighborhoods that are not getting the love anymore that they used to and they're also not getting fiber. So how do we fill the gap with wireless to be able to deliver fixed internet access? So I'm gonna deviate a little bit here where with I think the general perspective of go up to get more spectrum, I'm gonna go focus on technology that we develop to try to get more out of the spectrum that we can and ever have. And this comes from techniques that we call spectral reuse where we synchronize all the base stations to eliminate our own self-interference. So technology is like Wi-Fi as we know them today. They interfere with each other. You, your neighbor, your next door neighbor beyond them, they all kind of can see each other's access points. And at some point, you just keep on taking up every channel and then if you put too many devices on them like in rooms like these, usually Wi-Fi kind of falls down. So then you keep putting up more access points and filling up the entire spectrum. That is in effect self-interference and to avoid self-interference, you take up the rest of the spectrum. This new technique that we've created effectively eliminates the self-interference problem by synchronizing the access points so that I can have adjacent cells coordinated with myself or with other ISPs. And this is something that's actually never been done in any kind of coordinated fashion with licensing, so to speak. So today I think with what the Google guys and the Microsoft and CBRS plan has done with the SAS system in 3.5 gigahertz, this is something that could greatly expand and multiply the bandwidth in those bands because now service providers could actually coordinate with each other and synchronize their base stations on the same frequencies and not interfere with each other. So I think it's pretty cool to be able to look at Silicon Valley and what technology is now doing, not just with MIMO technology, like we've been doing in Wi-Fi and cellular, but now also with spectral coordination and reuse. So really, where does this get us? I mentioned my friend in Minneapolis at $13, $1400 to deliver fiber. Really, we're looking at about a $250 to $300 spend and we're able to, in the low bands, take advantage of the chip technology developed in the Wi-Fi industry to keep it low cost. So we're now able to really change the game and set the bar much lower for the cost to get in to play and that's what's gonna stimulate this business, in my opinion, rather than having to come down from very expensive millimeter wave technology, we're already down in very low cost because we're down in these lower sub six gigahertz bands. So a couple things that I get asked all the time. So why did I start in five gigahertz? Well, it's a pretty simple answer. It was the only multipoint, point-to-multipoint band available for use that would work. There is literally no multipoint band unless I have a lot of money to spend and go up in frequency. I was kinda stuck with the Wi-Fi spectrum. So my entire job here was how do I coexist with Wi-Fi? One of the things that we've noticed that is pretty incredible when you're looking at outdoor versus indoor networks and suburban neighborhoods where people have single family homes, single family homes are great mufflers for five gigahertz technology. So they literally kind of build a fence around your wireless to make it such that when I'm operating in the neighborhood and I'm also only using one or two channels across the entire neighborhood, we actually don't see the noise from interference from inside the house. We see signals, but it's not enough to bother us. So we've done some pretty incredible technology to know what kinds of signals we're looking for and ignore Wi-Fi signals inside of the house. And we're seeing people use Wi-Fi channels that are the same ones we're using outdoors on their in-home networks with no interference. I think this is a really huge innovation that's gonna make it very easy for us to become better and better behaviors when we do see interferers in the neighborhood to be able to deliver really strong performance. So a couple of things I'd like to point out on the multi-point side. I kind of drew a quick comparison of different bands and the characteristics that we look for. So we look of course for path loss. The lower down you go in spectrum, of course, the further the signal is gonna go. We look at foliage penetration. Foliage is kind of not an exact science. It definitely is better the lower down you go, but we also see characteristics where some bands in the higher areas do a little bit better and it's not pure physics every step of the way. A lot of times it depends on what kind of deciduous pines you have in the neighborhood. It's gonna change things. But really also as we go up in spectrum, rain fade and things like that also limit the distance that you can go. So as you can kind of see in this, of course, being in the five gigahertz, we have big benefit, but we also have less spectrum than the millimeter wave band. We only have about 600 or so megahertz to deal with. So you can understand why 600 megahertz, I have to be really focused to recycle that bandwidth and do it in a very good behavior fashion. Now looking at kind of a sister band, as I would call it in the mid band, the 3.5 gigahertz, the CBRS, the citizens band is actually pretty close in characteristic to the five gigahertz, but it's now licensed and of course protected. So now that has 150 megahertz. So you can probably imagine my next line was, well, 150 megahertz, I better be really good about reusing it or not gonna get that much bandwidth out of 150 megahertz. I definitely echo Milo's comments around the other nearby bands to the 3.5 gigahertz. I think that the SAS database being an incredible innovation and sharing is something that we should be expanding into more and more bands because we do need multi-point in these lower spectrum. So I definitely would love to see continued action to push the SAS into more areas for multi-point bands. And also I'd like people to start considering taking the synchronization and coordination technology for not just planning around each other, but planning to allow you to reuse the same spectrum. So you're not having to fight over if somebody got an exclusive geographic area, so to speak. So last thing I wanna show here is, I talked about multi-point, but the reality is we are all dependent on fiber and backhaul technology to get into all of these tight woven locations and cities. And we're never gonna have fiber in all the places that we want. And this is where you really have to know your bands and whether or not we have enough spectrum to get the job done. So I've kind of drawn different mileage distances that you need for the different applications. I'm gonna start on the far right and looking at kind of getting back to tower locations if I can't get to fiber. And these times I need to get more than three miles. And this is really a case where I think if you look at the millimeter waveband, there's not a lot up there other than I think what Verizon's doing in the 28 gigahertz to be able to do long distances because having enough power, but also the traditional microwave bands, the old part 101 bands, can serve very, very nicely in these areas. So the 23 gigahertz band, for example, is a very widely open two gigahertz of spectrum that can go very long distances to fill voids in kind of suburban and mid-rural areas where we can't get to fiber but need to go distances. Moving tighter in, I think the sick blue guys nailed this with their products. We see a ton of their gear actually deploying with ours, which means people recognize that they have about a one kilometer kind of range problem to get to our access points. So it's a very collaborative effort to be able to take backhaul and feed these base stations. And then looking further down towards what we call the micropop architecture, we need to be at links that are reliable at a half mile and that generally means we need line of sight. It's very unlikely that we will find an online of sight technology to serve at that distance that's gonna feed access points and that means we're gonna need to rely on different bands to get that done. I think the 70 gigahertz band and the unlicensed 24 gigahertz band are very good examples of those that can get the job done as well. So with that, I will hand it back off. Thank you. Thanks, Jamie, that rounded out the perspective quite a bit. Next we have Alan Norman, who is a public policy director at Facebook. And Alan fairly recently joined Facebook. He came from Google where we worked together actually many years. Alan led Google's ambitious Wi-Fi initiative in India called Google Station, among other things. Alan? Thank you, Milo. Excuse me. Thank you, Michael. Glad to be here and appreciate the opportunity. The first thing I wanna mention at Facebook, the company's mission is to give people the power to share and make the world more open and connected. And it really fits into this exciting time that we're in now of what's gonna happen in terms of how we really deliver broadband, really broadly. The particular role that I have is supporting Facebook's connectivity lab and connectivity efforts, especially around delivering broadband. And the company does not wanna be a service provider, but is working hard on developing a lot of enabling technology to reduce the barriers to delivering connectivity. And we've really broken the problem down into a couple areas. One is infrastructure, another is affordability, and another is relevance. I'm gonna talk about some of our infrastructure initiatives which we'll tie directly in. And just like you saw some of the earlier presentations, in fact, everybody's, people are thinking about this across the board of different densities and different topographies. And it's not one size fits all when it comes to spectrum or how you think about the problem. So in dense urban, really the challenge is really about capacity and quality of connectivity. And so there I will talk a little bit about what we're doing in 60 GHz with Telegraph. But there are other, as the density goes down, there are other challenges. The economics are different, the propagation matters. And so Facebook is doing work on high altitude platforms with a program called Akila. There's another program I haven't mentioned here, but it's very relevant called Aries, using massive MIMO to do backhaul and really lower the cost. And then for very rural areas or very spread out areas where other things can't deliver, we're doing work in the satellite arena. So if you look at Telegraph, it's using the 60 GHz band to rapidly deliver connectivity in dense urban. And it's not a replacement for fiber, it's a compliment, just like I mentioned previously, and fiber is fundamentally the lowest cost, but if you can get there less expensively and sooner and offer something that has huge benefits, not just here in the US but around the world. So the idea is we have small nodes on street furniture. We're gonna use basically wifi chipsets to solve one of the major cost barriers. And then using a lot of Facebook labs work around using the cloud, we're able to drive down a lot of the costs and solve some of the problems around interference, around deployment and delivering capacity. So the model is you have light poles based 50 to 100 meters with nodes that are meshed together, and then off of those nodes, there's either connectivity provided to buildings or could be used to provide wifi in the outdoor environment. And it looks like ethernet access to the buildings, although it's actually wireless. So what are some of the problems? One of the problems is that in this band, you get very low short distance, and so the way that's solved is by meshing together light poles. Another is you don't always get line of sight in the city, so we're taking advantage of reflections off the side of the buildings. And then there's the issue of interference and we're using the cloud to plan a lot of the network frequency use so it looks like a single frequency network. You know, each band has its characteristics. They can either be disadvantages or advantages, depending on how you think about it. One of the huge advantages here is that in the V band, there's 14 gigahertz of spectrum, as mentioned previously. That makes a huge difference spectrum and those quantities solves lots of ills. For example, having done a lot of work in the TV band, I'm thinking, oh, that'll help wifi improve coverage in the home. Well, it turned out that having a lot more spectrum at five gigahertz, you could actually get more coverage than an eight or six megahertz channel in the TV band. Just by using intended technology and different modulation schemes. With 14 gigahertz, you can definitely do meshing and deliver multi gigabit capacity to buildings throughout a dense urban environment. That's significant. The second is that this particular band has high oxygen attenuation. So you don't get much propagation. Solve that by having nearby connectivity through the light poles. The benefit is you can get very high frequency reuse and you don't have an interference problem. Another issue is that it takes a long time to get an ecosystem going in low cost. Unless you're Verizon and can drive all sorts of participants. And for example, at this conference, we don't have a chip vendor represented. One of the advantages of the 60 gigahertz band is there's already a wifi like standard. You know, IEEE and there are people developing chip sets. It's questionable at this point what the timing will be, but it's clear that in the long term there's the standard and there are folks that seem to be committed to developing those. That's important to getting low cost. And then finally, it helps to have global scale. Over 20 countries around the world have already adopted the B band in a licensed form and making that available. And these are major markets. So the timeline is that there's already a testbed up in Facebook's campus in Menlo Park. The company is working with a service provider in the city of San Jose to deploy a test network in San Jose. And then working with operators around the world there'll be trials in the next calendar year. Thank you. Look forward to the panel discussion. In our final speaker, we'll go next to the, everyone is gonna come up for the panel discussion shortly, but first Virginia Lam Abrams who is Senior Vice President for Communications and Government Relations at Starrie. And Virginia has, I'll let her tell you about Starrie, but I know that prior to this she served the same role essentially, Senior Vice President for Communications and GR at ARIO. So two different startups. Thank you, Michael. I promise I'm the last speaker. I'm gonna be quick so we can just get to the panel portion. So again, I'm Virginia Lam Abrams. I head of communications and government relations for Starrie and many of you may ask, well, what is Starrie? Well, we'll start by saying our mission is to bring high speed wireless internet to billions of people worldwide. And so how do we plan to do that? Well, we are a company of approximately 100 people based in Boston and New York. We are a startup and we have developed a fixed wireless, gigabit capable broadband technology. What we like to term as wireless last mile. It's a point to multi-point technology that utilizes millimeter wave spectrum. We are in between the 37 and 40 gigahertz band. We built the entire technology stack in-house from the network level node, our base station known as Starrie Beam, to Starrie Point, which is the transceiver that can sit atop an apartment building or an MDU, or in the window of a single family home, to Starrie Station, which is our in-home WiFi hub that sits in the consumer's home. What's great about our technology is that it does not require direct line of sight. The technology is cost efficient, easy to scale when you compare it to fixed wire line. We're about 100x less the cost of fixed wire line and we are designed to serve urban and suburban environments. So think in the 1,000 to 1,500 homes per square mile. We are backed by world-class investors here. First Mark Capital, KKR, IAC, Tiger Global, Soros and Highline Venture Partners. So again, our Starrie architecture, it's Starrie Beam, to Starrie Point, to Starrie Station in the home. And here are some images of our base station which was specifically designed to be easy to implement and it's small, it's compact, it's 16x16, therefore it doesn't require a whole lot of permitting at a local municipal level. The home transceiver is even smaller. Here you see the MDU version as well as the single family home. And then of course our touchscreen Wi-Fi router which is Starrie Station. We are currently in Abedah in Boston which we plan to open up further in the Q1 of next year and by the end of 2017, look to expand to a handful of additional cities across the U.S. So that's it, thank you. We'll go right into our discussion while all the speakers are having a seat. Actually I meant to say this at the beginning, I wanted to thank them for coming all the way to Washington. This is unusual, actually five of six are from out of town, four from California. So people came a long way for this event. Also there's some seats, front row now is completely empty if anybody in the back wants to come up and take a seat, that's fine. So as I mentioned at the outset we'll have some, I'll ask some questions just to get the conversation going but everyone should obviously jump in with anything you think needs to be said and then we'll go to the audience later for some questions and comments. So to start, I find it interesting that the most visible first movers in this space such as for example, Starry and Siklu who we've heard about web pass I think would be another one have been using wireless and core urban areas where conventional wisdom says that fiber is readily available up and down maybe every street in some of the cities where this is happening, Boston, San Francisco and so on. So why is that, why is this starting as an urban phenomenon when the lack of internet connectivity is probably disproportionately an ex-urban and rural phenomenon and can we expect this technology wireless fiber to have an impact on rural and ex-urban connectivity as well down the road? Boris, if you wanna start. Okay, I'll start. Well the sweet spot of millimeter wave and V-band and E-band is actually denser. The need for capacity there. So the density of population is there. The density of MDUs, SMBs are located there. Yeah, the fiber is somewhere in the cities but we've seen a picture, a snapshot of San Leandro and this is a typical situation here. So the fiber is there but it's not dense enough to get to all the SMBs, to get to all the MDUs and fiber extension using this technology can provide definitely fiber-like capacity at very low investments. So this is I think why there is a good match between the needs in terms of capacity and what this technology can deliver. That being said, we have some deployments in suburban areas as well and we have a product that is in development for point-to-multipoint radio that is going to be available in Q1 that is actually targeting suburban areas, single-family homes. So we are looking at these places as well but the sweet spot of E-band and I mentioned that we are not fighting fiber and we are not fighting physics as well. So the sweet spot is really dense urban. This is where we see the greatest demand and greatest match for the technology. Myla? I'll add that, you know, when you actually build a bunch of fiber networks, you would expect that as density goes up, the cost per unit of connectivity goes down. That's true to a point but it's not a curve that looks like this, it's more like a smile. As the density continues to increase, costs start going up again because the cost to actually do underground construction in very complicated environments like San Francisco, New York, Chicago, you're all underground. You don't have a lot of common duct space that you can get at and so it turns out that the big cities and the rural markets have a lot more in common in terms of fiber construction prices than you might expect. Some of that can be alleviated if cities take proper steps to have, you know, build once, dig once where you've got conduits and other things that can be used. Also the entrance facilities into buildings can become very problematic because you may not be able to get permission from a real estate owner to get onto a building or into their building because of various exclusive contracts that are present. And so that makes life way more complicated than you would expect. And I think that's part of the reason why there's such interest in fixed wireless in the major urban core. Yeah, and I was just gonna, Milo basically said everything that I was gonna say is that it's, you know, as a, you know, we actually obviously have fiber to the home and it's in cities, there are just a lot of difficult situations where you just may not be able to reach all of the, you know, potential customers. So, you know, we definitely view our 28 gigahertz wireless fiber is we will also be deploying in areas probably where we already have fios to reach some of the, some of the customers that we can't reach now. So it's not so much the distance of the dig of the trench, it's the complications of being in the city, the locations of the fiber. Right, and we're probably a little bit in a different situation than others here because others here, that is the mode of getting to the customer. But we also will be doing, in some cases, side by side, and it's not, you know, I didn't talk at all about what our deployment plan is and, you know, I can't really say what cities, but it's, you know, it would be both within where our fiber footprint currently is as well as in other cities where we, you know, don't have that as a, as a, you know, possibility. Jamie, do you have a different perspective in the sense that, you know, because I think when we think about the economics of fiber to the home, meaning the wire line fiber, you know, I think that, you know, most of us fret about, oh my gosh, when you get, as the density goes down, this population density, these homes spread out small towns, we think that's where you would think it would be most expensive. Now we're hearing, you know, there's these cost problems or practicality problems on either side, but is that also a spectrum issue? I mean, is that where you're, you were talking about mid-band that gives you some greater, there's no doubt that millimetre wave, millimetre wave has taken off because the short distance that you need to get to land buildings that you need to get to and they're all line of sight are super easy. It's a super easy slam dunk to use that technology and make perfect sense in those environments. But when we move into the suburban environment, you just, you absolutely run into the situation that, you know, obviously trees block almost everything. Now it doesn't mean every area is gonna be like that, but the sub six gigahertz is pretty much the only thing that you're gonna be able to run. And in many cases, even my five gigahertz stuff, you know, at the density levels dropping below what I was mentioning earlier. But, you know, if you go beyond a mile, you're gonna have a very great difficulty unless you're shooting from a tower as far distances. So that again goes back to an old model which we know doesn't work that well with firing from towers and we're going from rooftops on multi-dwelling units works perfectly. So it's that thing in the middle that's somewhere between, you know, a hundred meters out to probably a half mile that's super, super hard. Happens to be the lowest cost thing for fiber, but fiber's still not even barely getting it done at the right cost points even in those markets and would be my argument. Right, and that's where your micropop. Yeah, poise. Right, exactly. Okay. Good. So that, it takes me to an even more general question which is, which I think a couple of you touched on indirectly is, you know, what do the economics look like in comparison to trenching wireline fiber all the way to the home or business? In other words, you know, is there a rule of thumb or a goal in terms of the cost saving? And I imagine it does differ, obviously urban, suburban, rural, but I mean, is this a major cost saving? Does this make it more affordable to do this? I think, just speaking for alphabet, there are places where if you can make the economics of fiber cheap enough, you will always build fiber to the home. I think everybody would choose that approach, but that takes time as well as CapEx to deliver. And part of the appeal of wireless is there are certainly areas where the economics of fiber don't work but would work for wireless. There are other areas where you may not, it's not just the cost of doing the build, it's actually how quickly you can do the build. And to the extent that you can use wireless as an extension of that, you will be able to build faster, deliver services to users quicker, to get to revenue faster with a wireless approach. And I think the question is really sort of speed and latency, right? What can you do in terms of really driving a competitive product? Because I think my old friends in the cable industry are being quite aggressive about deploying Doxus 3 and Doxus 3.1, which will be pretty good in the downstream. And so I think if you're really gonna be on a competitive basis, it's not just about delivering faster speeds but also being able to deliver services faster. And so we constantly are looking at that mix about how can we drive the cost of fiber construction down but also using wireless as an extension to drive those services to users faster. And I suppose in addition to potential or cost less disruption in terms of all the permitting and all that stuff. It depends, right? There's a lot of variability. Cities can do a lot to make it easier to build. And but I think that also is true for wireless, right? The lot of cities can make things harder to put up wireless infrastructure or easier. And those cities, the ones that make it easier will see better infrastructure to get deployed. Right, any other? Yeah, I certainly think you're seeing a different perspective exactly as Milo brought up about how can you get into a city and now poll rights, especially with Title II, are available to virtually anybody, equally depending on the size of the ISP, doesn't matter anymore. So if you're good about going after a city, the things that cities want, they still want good old muni Wi-Fi. They still want Wi-Fi across the city that's faster with lower latencies. And I think that if you combine the efforts around both connectivity gap and Wi-Fi through the city and use that as a backbone to also support private entities doing broadband, that's a great combination in the, that's really gonna benefit cities to want to work with you as an ISP. So the success stories that we're seeing are when you early on approach cities and say, what is it that you need? How can we help you? And then how can we get equal fair access and make this a win-win for everybody? And that's gonna dramatically reduce the cost. But I mean, in general, we're seeing costs just on the fixed broadband side for three to five times lower cost than fiber. Yeah, you were saying on your slide, it's about $250 to $300. About $250 to $300, yeah. Yeah, which I imagine must be, somebody else mentioned trenching is at least 1,300 or more. It depends. Yeah, yeah, Ellen? Yeah, I just like to echo Milo's point, but just think about it a little differently. I think that the, it's really important for people to think about this systemically, that in the long run, I think fiber is actually lower cost. If you amortize it over enough time and enough people, less power, you don't have to maintain it as much and things like that. But because of the time advantages, the reduced CapEx improving out market demand, if you start with something wireless and get sooner, it can have huge competitive advantages. So rather than thinking about just one piece at one time, think about this over time, how you're going to deploy from one and move to the other and watch how people do that, I think is actually very important. I agree with the Milo and Allen about the time to market is incredibly important. And I think we're at a time in the innovation cycle where all these technologies are nascent and you can actually build them to address a lot of these hurdles that we see in local markets, which is why we specifically built our base stations and our nodes to require the least amount of permitting when we go into a market. And I think there's a lot of opportunities as these technologies are going to develop for us to look down the road and say, hey, these are the issues we know we're going to encounter. How do we build our technology to address that upfront? Rather than having to say, oh, we didn't realize it was going to take us an hour of time in order to build out this network. I also think that with the bands that we're working in the millimeter wave bands with the propagation and the fact that, especially for storage technology, we're about a kilometer, a kilometer and a half radius that urban areas, it lends to building out in urban and suburban areas. And that, again, 1,000 to 1,500 home density per square mile. Yeah, I suppose you have the multi-points within range of the one access point. Okay, I'll add a different perspective. We see a lot of customers that are not comparing wireless business case to a fiber business case, but rather looking at the business case from, even from consumer perspective. For example, we have a lot of deployments in the safe city, which is becoming the smart city market. Our devices are connecting cameras, for example. So it's not the business case from doing fiber to this pole that will connect the cameras. It's just, you can't put or doing wireless to this pole. You can't put a wireless device that will cost several thousands of dollars and just connect wifi, uni-wifi or a camera that costs a few hundreds of dollars. So it's the business case, it's not always comparing us to the fiber build but to other technologies as well. And for the fiber to the home, Google said the benchmark of $70 per connection per month. So you can't put equipment that will cost $10,000 on the rooftop and hope that it will make, maybe it will be less than fiber, but from the other perspective, from the customer's perspective, the business case will not work. I feel like I need to say something too, but I don't, not to talk about the specifics of comparing fiber to the wireless build, but we have actually touched on a very critical issue which is working very closely with communities that that can make or break your deployment in terms of how expensive it is. And therefore I think, I can't remember who said it, but somebody on the panel just said that you wanna do, you wanna make it a least cost environment but you also wanna work very closely with the communities and have them want you to come in and actually provide the kinds of services that you're providing. And I think instead of looking at the various providers as being a revenue basis for them on an individual like location by location, more to think about that this is a way to potentially bring more business into the areas to do a lot of really good things for a community. And I give as an example what we're doing in Boston with them in terms of laying new fiber that actually eventually probably will support what we're also doing with millimeter wave technologies, but it's really about working with a community well to get the kind of permitting and do the kinds of things. And frankly, it can affect the cost of deployment very much so, but we also look at it from a wireless point of view where we are not the local provider, we've with 4G and other things. In some cases we don't have a choice that we have to densify the networks because it's actually funny as I don't know if any of us have really mentioned the underlying component here is there is tremendous increase in demand for what all of us sitting here are doing. And so as an operator, we're out there just saying we gotta be able to meet the demand and the way we meet the demand is actually trying, getting new spectrum, getting better technologies, working with cities so that we can densify our networks, all sorts of things, but that implicit in all of this is that there is a tremendous amount of consumer demand for the product. Okay, well we heard from your opening remarks and presentations that your companies are using a wide range of spectrum bands from sub six gigahertz, this mid-band spectrum, all the way up to 70 and 80 gigahertz for this purpose. And so this is very different propagation characteristics. One thing I'm wondering, and Jamie, I think, focused on this the most, if substantially wide channels of mid-band spectrum were made available to you, that is between say three and six gigahertz, or even in the six gigahertz band, would that be preferable to millimeter wave spectrum for this purpose? Or? Not always in my answer. When would it be? Or, yeah, I'm sure it's a line of sight and distance issue partly, but it would be good to explain that because there'll be debates about opening different bands and so on and it's helpful to know which have some particular value added. I mean, because it seems, as Jamie was saying, the propagation is better in the mid-band range. I think people are already nailed where millimeter wave should be used. I mean, there's no way you should ever go lower in spectrum because the availability's up there and it does the job. So we should only be using spectrum where it does its best and where it has its biggest advantage. So I would definitely advocate that the lower spectrum and if we are able to open up more three gig spectrum, that would probably be in that kind of suburban to mid-rural. There's no sense in trying to, it's gonna probably get too far, so to speak, and over cover the population density that you can serve and that's where millimeter wave being so short and so high capacity is perfect for lighting up buildings because there's 50, 100 people in those buildings. So that's kind of my view of it, but we need it all. I look at it, I look at millimeter wave and connecting to a building more as infrastructure than I do access to a home. It's feeding 40 people. That's not very different than 40 homes being served by an access point that's fed by millimeter wave. Milo, do you have? I would just say we've done Andy Clegg is here and we've done a collective millions of points of propagation data in the three, five band. I think we have a very good understanding now of where both millimeter wave as well as sort of sub six spectrum have sweet spots. I think the suburban less dense areas where you've got a lot of foliage, the sub six gigahertz will give you better penetration. If you can get enough spectrum aggregated, then I think you've got the ability to actually operate at higher coding rates. You get beamforming and MIMO that can actually help with reuse even in those bands. And so we think we can get to a path of delivering gigabit services in those bands, whether that's called 5G or not, 5G is more a marketing term right now, I think than an engineering term. But I think there's work going on in both millimeter wave and sub six. In Europe, the EU has standardized on the three gigahertz band for a bunch of their 5G activities. A lot of the trade offs depend on can you get fiber close enough to the user where the millimeter wave propagation is not a problem. You know, there are in a lot of Japan, you can get dark fiber for less than 50 bucks a month in the city, in almost all the cities there. And that makes designing a network very different than designing a network here in North America or in Europe where you don't have that kind of really, really, really deep, affordable, ubiquitous fiber. I don't know if you disagree with me on that. No, I think that's true. And just to your point, I think 5G is not gonna be, you know, I think it's a mistake to think of it as just the high frequency bands. I think it's gonna involve, you know, as it evolves, as you think about it, sort of IOT types of applications that could become much more ubiquitous. You might need, really, you can probably use much lower bands for that. You know, in terms of what we deal with here is that one of the issues that you have, it's a great mind game or, you know, thought exercise to think about that. But those frequencies actually, you know, we have to figure out the sharing in the, because they, whether they're being used or not is sometimes a question, but they certainly are allocated and assigned to users. So it becomes, you know, as I said, it's great thought exercise. And I think all of us here would say, yeah, that would be great if we could have access to that spectrum. And, you know, maybe by some of the things that we're doing in 3.5, we'll be able to expand that and make it more usable. And I think that would be good. I think the really important that we focus on making the five-figure expand is a whole more useful, not just for the important application that Jamie's talked about, but also in the home, just being able to deliver truly a gigabit. Look at some applications like one that Facebook has Oculus. You know, someday it'd be wonderful to use it on Verizon's network, for example, and get the low latency and the feedback, which may happen someday, hopefully. But certainly, you know, more near-term is getting it to work over five gigahertz in the home and get people to understand what you can really do when you have remote presence and that kind of virtual reality experience. Okay, let me come back to what Sharla mentioned, which was something at the outset. You know, I suggested folks pay attention to is not just that you, your companies are operating at this whole wide variety of frequencies, you know, different tools in a kit. It's kind of where we ended up on that. But there was also different access policies governing, you know, that spectrum. So, you know, a number of your companies are relying on unlicensed spectrum, whether it's, you know, five gigahertz and mid-band or 60 gigahertz. Then there's lightly licensed, 70 and 80. The 28 gigahertz band that Sharla mentions is exclusively licensed. And then, you know, we've also talked about, you know, there's an opportunity perhaps to extend the shared, the spectrum access system, the dynamic sharing of the 3.5 band to more of the three gigahertz band, for example. Jamie mentioned below or above. So, I guess my question would be from a business and regulatory standpoint, is the shared access approach to spectrum feasible for this? I mean, we know it's been feasible for best efforts, Wi-Fi, but is shared access, is the shared access approach feasible for, for delivering high capacity to the home or business or our exclusive licenses necessary by and large? And I'm assuming when you say that, you're talking about like a SAS, a spectrum access system, because in fact, 28 gigahertz is shared with satellite and we will have to go through a process of working with them to share that spectrum. So it's, you know, yes, it's coordination, you know, for sure, but it is absolutely shared. So, yes, I mean, the answer to that question is absolutely, it can be shared and used for that purpose. You know, whether you need a SAS in some of these bands, I don't know whether you would or not, but it can be shared. And to sort of answer your first part of your question, all of those different kinds of approaches are gonna be necessary for us to be able to make the best use out of the spectrum. You know, and I think that's become quite obvious over the last several years, just in terms of, you know, our involvement with the 3.5, you know, what we've been doing with LTEU, 5 gigahertz, you know, what other parties are doing on other bands, there's just a lot going on, and no single approach is the right way. So. Yeah, I'm just, I got a little one for you, I think. No one approach actually solves the gamut of all of the wireless access problems that we need to solve in order for us to really empower the future to be and the one where bandwidth is abundant and unconstrained. I mean, I think that is the goal here, and I think sharing is an expedient because realistically, part of the dynamic is even if you wanted to go to clear out all the incumbent users, it may take a decade to do that, whereas you can share the spectrum a lot faster, but it's not to say that there's not some bands where you can, it works, lessivity would work, but to echo Charlotte's point earlier, if you look at the spectrum map in the United States or in almost any country, it's all taken. All of those little slices are assigned to somebody. Changing that allocation is a long, complex political process, not primarily an engineering process. And so, to the extent that that's the problem, you've got viscosity and the inability to have flexible use. We're gonna have to find ways of making that process take less time, and sharing is the primary way to do that, not just between civil users, but between federal users and civil users. I mean, the key here is gonna be identifying the different bands, different incumbent uses because today we know obviously there's military use that we need to protect in the 3.5, and there's many other different bands that are highly underutilized that are easy to protect, and there is some engineering that we need to go through in many cases detect like we're doing with the ESCs in the 3.5, but there will be other types of radar, other types of satellite systems that are in place that whether they be in a database that exists by the FCC today, or something more nomadic, those are the things we have to take into account. SAS as a generic mechanism works so long as we've solved the incumbency problem. So there's a lot of work to be done to help that sharing go, but I think the mechanisms are now in place from a policy perspective that we could drive forward. Now we need to come up with technical and new rules that are gonna be amenable to the other parties. I think just to add on to Charla and Milo's comments, I mean, I absolutely agree with what they said in terms of the approach and having a flexible approach in terms of unlicensed, exclusively licensed and shared, but another benefit of sharing, and this is going to represent the view for startups is that it really enables new entrants to come in and work in these bands. And I think that if you went with a exclusively licensed model that that would really foreclose opportunity for a lot of smaller players, and particularly folks like us who are new and while we're well funded, are not obviously as well funded as some other folks. And so that's another massive benefit, I think. I don't think anyone can really predict where we're going to be five, 10 years from now. And so the ability for folks to innovate in these bands and to experiment in these bands makes a lot of sense. I think the high order did here is still more efficient use of the spectrum and in particular more spectrum for broadband. And SAS and other tools, coordination, sensing, database, I think are understood, but they're not that simple yet that people can say, oh, we can just use it and be confident that it works. But the other side of that is getting the political equation right so that people who are using the band understand that, look, if we really have to accept that the spectrum is going to be used more efficiently and broadband's an important use, let's find a way to share rather than block and then we can both use it. And that mindset might actually accelerate things a lot more, but I don't think we're there yet. So if we look at the millimeter wave bands that most of you are using, the FCCs, and this was mentioned by someone, that the FCC's spectrum frontiers order in July made available a massive amount of new spectrum that's good for this purpose. So 3,250 megahertz for license, 600 megahertz for shared access, which is still to be defined in detail, and 7,000 megahertz, 7 gigahertz more of unlicensed up in this 60 gigahertz range. And it asked about additional bands in the further notice. So my question is, should the new FCC do anything substantially different about access to the high frequency bands to the millimeter wave, or was the current approach on the right track? I think that from our perspective, and we see it in different countries, in other countries, where these bands are open, either unlicensed or lightly licensed, it really drives forward the telecommunication infrastructure. We see it in the developing countries, and we see it in Western Europe as well. So I think that the current model, and it also has to do with the physics of this band. So in millimeter wave, in 60 gigahertz, and in e-band, 78 gigahertz, there is no need to coordinate because of the propagation of those frequency bands, because of the directivity of the antennas, because of the large number of available channels. So we see that it will be beneficial that these bands will still be unlicensed in the B-band frequency, and we agree with comments proposed by Google and by Facebook. So it will drive, it will serve, as I mentioned, it will serve as a competitive advantage for the small service provider to provide really high capacity services, and it will drive the economy for them. And for folks who don't know, e-band is 7080, which is lightly licensed. Which is lightly licensed, yeah. So it's coordinated, and there is online mechanism, $75 for 10 years. There is a registration process, and if you see the data is publicly available, and the B-band, which is around 60 gigahertz, is unlicensed here in the US. Any other thoughts on this one? Okay. I guess one last question, then we'll open it to the audience, which is, I guess this is sort of, I'll start it with Sharla, but then to the others, which is, I'm wondering, Sharla, if you could say a bit more about Verizon's plans for fixed wireless at 28 gigahertz and whether that's, and I think you were sort of suggesting this in your opening remarks, whether you see that primarily as a platform for the eventual, quote, 5G mobile, as that develops 2020 and later, or whether these are really two separate things. No, they're definitely not two separate things. At least we hope not, because we hope all the work that we're doing here will actually inevitably lead to full 5G range of services. And in fact, as I mentioned earlier, we have done a lot of work with our vendors on coming up with the beginning set of specs that's actually feeding into the network provider's 5G forum, which is 3GPP. I know there are other groups that are doing things that'll probably feed into the standards process as well, but we really do see this as a path that I don't necessarily see that the device that we may be using will become a mobile device. That's not what we're talking about, but what it's really is that the pathway to all the network work that we're doing behind the scenes and all the specs and talking about the RF layer and all those layers will actually feed directly into the 3GP piece and has been feeding into that. Primarily with a densifying capacity. You're just bringing high capacity into all these places. Yeah, into all these places, and hopefully we will be able to make mobility work in those bands. That's primarily what we're talking about here, whereas there will be other elements of 5G that may not actually be, and we've already talked about that, that might not be in millimeter wave bands, but will be considered 5G, but not in those bands. But in those bands, we actually, playing off the work of our vendors, the work of Ted Rappaport and others, we're, we are actually moving toward what we hope will be mobility in those bands as well. How about for the rest of you, do you see your networks becoming, seems the focus right now is on fixed wireless, providing the high capacity to the home or business for use in their home or business, but do you see your network then ultimately as part of the 5G ecosystem for a broader connectivity or? God, no. I look at it 100% as how do we get Wi-Fi to consumers? So if you look at the amount of traffic that goes over both cell phones today, as well as just general Cisco data that we all look at, 90% of the internet goes over Wi-Fi and Ethernet. So really what this is, is we're out of low band spectrum for cellular. We have to go somewhere else to get connectivity closer to consumers. And some of that may be 5G spectrum up there if we can sort out the physics issues and all the problems. But what most people have is what they want, what they have today and how can we make that better is get faster speeds to the point at which you connect your Wi-Fi router. And that's it. And that's gonna be the same thing in cities and outdoor areas. People prefer to be on Wi-Fi if they have a choice because it frees up costs. It weren't for unlicensed spectrum. I'd hesitate to guess that my company wouldn't be here. Probably Siklu working up in the lightly licensed with really, really good favorable rules wouldn't be here. So I think that's, we look at it and inspired by what the 5Ghz has been done. I prefer not to be using the 5Ghz for fixed wireless but that's how we got here. That's what's available to us. Ellen? This reminds me so much of 15 years ago when people were saying Wi-Fi versus working for three years. Turns out both have been wildly successful. And I really, we're working with both. It's Facebook with both service providers who are very supportive of Wi-Fi and doing a lot and others that are strictly pursuing licensed paths and they're both looking at how they address this last mile access and both are there. And I think it's great because you get competition between different ecosystems. You get use cases to get proven out and different cost structures are appropriate in different places. So it's really not, yes they are competitive but they're also extremely complimentary in a good way. Yeah, I would just say a couple things. As I said earlier, we use the same terms but they mean different things. Wi-Fi today is not what Wi-Fi used to be, right? Cellular today is not what cellular used to be. In many ways, both are approaching the same, they're incorporating the same technical elements in them, wider channels, OFDM modulation techniques, beamforming, MIMO, et cetera. Even in the cellular network today, the dominant use of data is when devices are not mobile. So we talk about the mobile network but actually when you're watching a 1080p video you're probably not driving, right? At least you hope not. And so I think the dominant source of bandwidth consumption has typically been when devices are much more stationary and that from an engineering perspective when devices are relatively stationary, you have more degrees of freedom that you can use. Mobile devices have fundamental energy issues associated with them because battery technology is not moving at the speed of Moore's law. We've already seen situations where lithium batteries have the energy density of a mild explosive and if not managed properly cause problems. And so I think energy consumption is also one of those things that will drive the ecosystems maybe a little bit more separately than maybe we would like but I do think the dynamics about what does it mean for the mobile, what does the mobile infrastructure look like in five, 10 years, right? Where is bandwidth going to be consumed? We've already seen the mobile infrastructure get fast enough for at least some sets of users to move completely to mobile and not have fixed line communications. But I think bandwidth demand is continuing to grow and we're gonna have to find better ways of reusing that capacity. And yes, Wi-Fi for what it means today is actually going to be the dominant interface for devices going forward, including things like televisions and other things. The time that we use coax in the home for connecting things is that we've gone past peak coax. Let's put it that way. Okay, well we can I think open it up for some questions here in the, oh and actually when you, if you can let us know who you are and who you are with that would be great. Gladly, thanks. Rhonda Silva, I was responsible for engineering for the past 12 years at Time Warner Cable and exited the company in August following Charter's acquisition. So it was great to hear a lot of stuff that I'm very familiar with. I was responsible for the access technologies there whether that's coax or fiber or backhaul for Wi-Fi. So I was surprised. I was thinking this would be more political or public policy based. Thanks. Now I was supposed to be, I was supposed to be in Guadalajara this week at the, at the IGF. You got engineers, you got engineering talk. So it's good, it's good. I was supposed to be in Guadalajara this week for the internet governance form. Had to cancel for some personal commitments I couldn't get out. So that's the reason why I'm here, I'm glad. But one of the topics, one of the main themes for the IGF this year is around connecting the next billion. And when you think about where those are, you think of Africa, you think of India and China and maybe even some of the regulated parts of Europe where the same capital that we have at our disposal here in North America to make these investments to increase capacity, whether that's wireline or whether that's wireless. The return on investment is not the same in these markets. And so my question, I'm glad we touched on economics. I wanna press a little bit more and ask each of you to talk about in the current economics that you have, whether you're trenching 1300 bucks or whether you're dropping wifi hotspots for a few hundred bucks, how do you drop that in order of magnitude to get sub $100? What are the main components that are driving those expenses? Because if you think about success of getting that next billion on broadband so that Google works and people can get Facebook and even basic communications like email are accessible for folks. How can that, we're in the whole economic spectrum where the spectrum costs, operating costs, customer premise equipment, where would you most benefit from if we could get some innovation to fix the economics in those markets and then come back and benefit from it here in this market? And actually that was a question I didn't get to which I was going to direct first to Alan who has some experience of that as I mentioned drove the Google train station wifi project and Teregraph may have some relevance here. So I think there are a couple of things. First of all, a big component of doing infrastructure labor and it's much less expensive in emerging markets. So that affects things. The second thing is wherever possible take advantage of existing assets and existing creativity in the local markets. And the third one is wherever possible use leverage the device ecosystem. So for example, use wifi and things like that that people already have. So for example, at my prior employer Google we deployed wifi at the train stations that already had fiber and already had places to put the wifi equipment already had a massive amount of traffic and density of potential users. To our surprise, it didn't have great power. So that was actually something we had to learn a lot about. But through that process, I think with actually I'll let Milo if you could just talk for a moment about current users and things like that rather than have a former employee talk about it with public would be great but it's tremendous set of numbers. Yeah, I mean, I think the wifi I actually don't know the specific numbers there but traffic continues to grow as I think there's a couple hundred stations now connected and that footprint continues to build. In many developing world countries part of the challenge is not just the access network but how you get to the real internet get to the rest of the internet. How do you drive the cost down of the international connectivity segments which in many areas don't have a lot of competition. The Google does have activities. There are networks in Kampala and a couple in a couple other cities where we've actually deployed fiber not exclusively as middle mile not retail in Africa. And to try and reduce that friction to get bandwidth to users again cheaper primarily mobile network operators. I would just say we should not assume that the developing world will follow the same technology path that we did. So for example, there's a large city in Africa there's a large cable operator there halted their deployment of HFC infrastructure in 2013 and started deploying G-Pon fiber. Not because of capacity but because it was cheaper. The cost of the optics and the couple of the cost of copper and metal made it actually cheaper to deploy fiber. That was probably the last Greenfield Coax Network ever built was probably in that city. So I think the technologies change and in many cases fixed wireless coupled with low cost fiber construction. It's not uncommon in a very large city you could see 150 to 200 bucks a linear foot as a construction cost for fiber. And so when you've got it's actually much lower in suburban parts of the country than that. And so but in the developing world you know labor is cheaper you don't have the same constraints on right of way. You can build fiber very inexpensively and I know reliance in India is being very aggressive about deploying fiber infrastructure in that country. I've been in the customer premise equipment business for 17 years now and I've never seen it get below $30 or even the cheapest DSL modem in a fixed wireless that'll probably never get below $80 to $100 range. It's already at about a hundred right now just for the equipment on site. So that's already gonna knock out every market you just talked about. So the only solution in my view in most of those spaces is there's not there's also not demand for premise connected homes as much as I just need something lower cost than a very, very limited cellular plan. So it typically is wifi and the Brazil is probably my largest growing market right now and it's mostly feeding places where they're bringing wifi to and so they're using even fixed wireless as backhaul as opposed to last mile in many of those cases and then wifi ends up being the primary mechanism but in Africa you're talking between 99 cents to $2.99 a month max capable spend by most of the population. So this doesn't make any sense for that market. I don't think it'll follow our trend but the tech feeding our homes here may feed their wifi in those areas. Yeah. The use case is also different. So one example is you go more rural. We're used to texting or sending texts and things like that. A lot more of the usage is actually video and photos. It's the way people communicate whether you have different languages or people aren't as literate. And so then the systems have to be built and deliver a different kind of experience which is actually a big learning as well. And just a quick commercial that we'll be announcing sending if you're signed up with us you'll receive an invitation for a forum actually a slightly bigger policy forum like this on getting the last three billion online so we'll bore into these issues and the innovations that are around the corner in a little more detail. Andy. Hi Andy Schwarzman from Georgetown Law School. I'm extrapolating from a single data point that I may be misinterpreting anyway and is inapposite in any event but. That just means we're in Washington. I'm struck by what is increasingly apparent that the demand for 600 megahertz TV spectrum in the FCC auction is coming in way down on the low end of projections. Everybody in this room is excited about the things that you're talking about and sees the benefits of connectivity and wants the world connected and sees all of the economic growth and all of the incredibly important things that can come from it. But to what extent are your business models vulnerable if demand turns out to be lower than we think takes longer to develop and comes slower? What if demand is 50% of what we may think demand is? Will this technology survive and be implemented just more slowly? Can this technology work with less demand? What if there really is less demand if we haven't done as good a job at educating the rest of the world on the value of connectivity? Well we won't, I don't think we want to. I know at least one of our panelists can't talk about the auction but this seems to be kind of a broader, just a broader question about demand in relation to business model here. I've been working in the internet, I don't know, 85, something like I'm old, right? Demand has grown anywhere from 20 to 40%, traffic has grown 20 to 40% compounded annually for I think the better part of a couple decades. So maybe we will go into a cooling phase where application, where data growth does not grow quite as much but I don't think so. There's a country in Western song, I like country in Western music, but you can't have too much fun, right? And I've never met someone who said their network is too fast or their network is too quick, too low latency. I would just say, you know, Wi-Fi, the Wi-Fi that we have today, right, with AC and where it's going with MIMO and the rest, is gonna unlock the barrier in the home to a lot more data consumption. Today, a video has remained, tend to be trapped in sort of the coax and is now escaping in ways that I think many people find unpleasant, but I think is the physics of the market are driving that. That's going to continue to increase. The number of devices in the home that are connected, I'm not typical, I'll just say that, but I did a check on my PF sense firewall in my house the other day and I had like 60 IP addresses a site, right? Now I have five, I have four kids, right, and a wife and a dog. The dog does not use the network, not yet, not yet. But if you just think about what that means, right? When you've got ubiquitous Wi-Fi distribution in homes at very high speeds, now that creates demand to feed those. And these devices, like my Pixel, continue to get much, much faster and can consume much more data by themselves. So I think it's a pretty safe bet to bet on more consumption. What I do think is the case is a bad bet to bet on higher ARPU, so we're going to end up delivering more bits at the same price. And that is what the photonics industry, what the wireless industry and what Moore's law has enabled us to do. And as long as those factors continue, then we will continue to be able to grow. There's another demand factor and that is outside the United States. There's huge latent demand just catching up to usage levels that are near where we are. So even if, which I don't expect will happen and don't anticipate in any way or see any evidence that things slow down here, there's just so much huge latent demand elsewhere to drive adoption that I just expect there's gonna be a large rising tide across the world. I've read about modems, it was in 1996 and it said that the barrier, the capacity barrier is 56K. Now, since then, and we talked to a lot of customers, we've worked with a lot of service providers here in the US and across the globe. I didn't hear, never, someone saying, I don't need more bandwidth. So the demand from bandwidth demand, they're radically different because, I mean, like take my story from earlier, 25 for 25 or 30 bucks for over 100 on fiber. People are gonna vote with their wallets and they will want decent connectivity. We still lack decent connectivity in most places and I'm not sure how much, different applications in different spectrum are gonna get you to decent and others are gonna put you up to a gigabit. Most people just wanna get the decent, let's start there. Take maybe one or two more questions if we have rapid fire answers here. So like all the way in the back, his, maybe his head or hand up. Sharon Bovat, voice of a moderate. When in Costa Rica looking for affordable healthcare, I actually bought coconut water that came in the coconut shell on a stand and they used their phone to my payment. So I'm seeing that connectivity is really the future. It's safer than cash. It's really the new society and anybody under 30 everywhere in the world, that's what's used. So I hear that people think that it might not happen. It will happen, but the question is, is that subsidizing it and America's role in that, what is your opinion? Because I do see it as an important issue, thank you. Oh, right, right. Are you talking about Americans having a role in subsidizing other parts of the world or within the US? I think one of the things, as engineers, it's harder for us to deal with the political dynamics, but I do think we can drive the cost of equipment and technology down to the point that the cost for deploying these networks becomes lower and lower. They do not have to follow the cost curve that we followed and pricing that we followed in the United States. That is both an opportunity for them and it's also a challenge, right? Because the same economic structures about revenue, right, or lack of ability for people to pay apply there. So I think we need to be really focused on how do we drive the cost of the technology down so that many of these countries need less subsidies if they get them at all. I think that's really the role of the tech industry and I think that's where all of us are focused. Yeah, and there's probably a fair amount of good information. If you look on the web for the Global Connect Initiative between State Department and World Bank, you know, it's aimed at this and probably has a lot of background here. Yeah. I'm Jeff Marks with Nokia. We, I'm gonna bring it back to policy and in the US a little bit to ask a question about, you know, Charlotte had brought up the issue of working with cities and one of the things that we've seen at Nokia as we've made base stations smaller and smaller and want thousands of them deployed, not every community is ready for that. It's very hard to get things deployed in cities. The federal government thus far has worked around the edges looking at like national parks and things like that. You know, it's Main Street that sometimes is the problem. So my question for you all as people who are looking at these higher frequencies and more and more base stations and more and more connectivity is, do you have examples that you'd think of who are good examples of cities to work with who other cities could learn from? Also, you know, I think we're gonna be seeing a lot of work on the federal level at encouraging cities to be better at this and if you haven't asked just I guess any of you who have ideas for what we should all be working toward in the coming year either looking at cities as examples or federal efforts that could be helpful. I would just say in our experience, you know, places like Kansas City have been incredibly flexible and accommodating in terms of the bills and more broadly, if you look at poll attachment policies, I think the state of Connecticut has done a tremendous amount of work there in terms of joint poll attachment. It's all about rights of way, right? It's all about getting access to rights of way and doing it in lower cost. And the folks in Connecticut have actually done that and added an additional chunk of comm space on polls in that state. There are other states that are doing interesting things as well as cities who are doing, we did once where anytime the ground is open, they'll add a conduit. We made available in many of our cities the option for cities to do that. I think that there's a latent uncertainty about what the value is that people got when they were looking at small sale and they were like, wait, wait, wait, all you're telling me isn't gonna get better brand X than here, but really a lot of people have really bad broadband or don't like their cable provider or whatever it is, that's the public perception and the constituencies of cities have to look at that and go, what are you doing for my population other than getting a little bit better wireless signal as far as their perception of it. And my view of it has been as they look at it and said, well, people are asking us for wifi in the city, people are asking us for these things, if you go to them with those solutions and also I think to the sick loop point earlier around and we can provide with municipal coverage for cameras, public safety, et cetera, that's a solution that people can buy into that's not rewarding one company, so to speak. So like the city of Minneapolis was a great example, I think they did a great job and they opened it up more broadly than working with one provider. So now we just say most true poles and pole attachment has been primarily a wire line issue, it's going to become a wire less issue. And so I think anything that can be done to drive lower cost and quicker access to poles will have benefits not just on the wire line side but wireless as well. There's a lot of, anything that you can put base stations on is something that's an asset in this game. And I think poles really need to be revisited because they're much more important for small sales and not just the fiber backhaul that powers them. We work with more than 60 cities here in US. Mostly it starts from providing security application connecting cameras at the street level and then they realize that this technology can virtually bring gigabit location to any spot in the dense urban areas. And then they see that there is enough demand to add additional applications like wifi. So they have the monetization and the value is coming from the security from backhauling the cameras but then they add wifi and sensors, et cetera. So it's kind of infrastructure for the smart city that starts from the safe city. So we are already a few minutes over. So wanted to thank you all for hanging in with us. Certainly more questions than I anticipated. I'm what I consider to be an opaque topic going in but I hope we learn something. So could you join me and I'm sure probably most of the speakers can stay a few minutes if you want to direct a question up here to them. But can you join me in thanking the speakers? We'll see you next time.