Today’s episode features an interview between Matt Trifiro and Iyad Tarazi, President, CEO, and Co-Founder of Federated Wireless In this interview, Iyad discusses CBRS, shared spectrum, and the massive potential for disruption and innovation they represent, as well as the interesting intersections between CBRS and edge computing.
Today’s episode features an interview between Matt Trifiro and Iyad Tarazi, President, CEO, and Co-Founder of Federated Wireless
Iyad is a technology industry trailblazer with experience in successful commercialization of disruptive cloud native SaaS technologies. Prior to Federated Wireless, he served as Vice President of Network Development and led the Network Vision modernization project at Sprint Corp.
In this interview, Iyad discusses CBRS, shared spectrum, and the massive potential for disruption and innovation they represent, as well as the interesting intersections between CBRS and edge computing.
Key Quotes
“The most thrilling part of CBRS is that it truly is free. It's really democratizing wireless access.”
“This is like Airbnb. Don't sell the house, I'll rent the room for you tomorrow. That's basically what we're doing. We make the spectrum available immediately, which is great for economics and great for innovation.”
“When we go from a few people building technology on spectrum to thousands and thousands if not millions of deployments, the amount of innovation we’re going to get in the software technology ecosystem is going to go through the roof…Innovation is going to take off. That's democratization in the best way possible."
“The U.S. will become the most innovative, by far, in creating solutions and applications and technology for wireless, because they’ll be able to get a thousand enterprises or a thousand companies all innovating at the same time.”
“At the end of the day, we're a software company that sits on top of an open ecosystem with a bunch of partners built around a very, very innovative spectrum model that's enabled by a really innovative government. The FCC and the DOD and NTIA and the White House--what they've done here over the last 10+ years in terms of creating this innovation model has really been first in the world.”
“All we really do right now is facilitate the management and enable the system. We pass more than 90% of the value of what's happening and what the spectrum offers over to our customers. We're more focused on enabling the ecosystem and enabling our customers than anything else. We're an enablement platform at the end of the day.”
Sponsors
Over the Edge is brought to you by the generous sponsorship of Catchpoint, NetFoundry, Ori Industries, Packet, Seagate, Vapor IO, and Zenlayer.
The featured sponsor of this episode of Over the Edge is Seagate Technology. Seagate’s new CORTX Intelligent Object Storage Software is 100% open source. It enables efficient capture and consolidation of massive, unstructured data sets for the lowest cost per petabyte. Learn more and join the community at seagate.com
Links
[00:00:00] Matt: [00:00:00] Hi this is Matt Trifiro, CMO of edge infrastructure company, vapor IO, and co-chair of the Linux foundatio, state of the edge project. Today. I'm here with Iyad Tarazi , president CEO and co-founder of federated wireless. We're going to talk about Iyad's background as a technology executive, the history of federator wireless and the future of 5G edge technology and a CBRS.
[00:00:21] Yeah. And how are you doing this morning?
[00:00:23] Iyad: [00:00:23] Great, Matt. Thank you for inviting me.
[00:00:25] Matt: [00:00:25] Well, you're very welcome. It's always a pleasure to talk to you. you know, even though I feel, I know you fairly well, at least in, business circles, I don't know much about your background. Like how did you even get into technology?
[00:00:37] Iyad: [00:00:37] Sure. probably the story goes way back when I was a kid, I grew up. out in the middle East, my grandfather was one of the first engineers out in, in mandate Palestine. It was part of the British empire at the time. And I used to sit down with him at the kitchen table. when he will bring his drafting diagrams and [00:01:00] pencils and so things, yeah.
[00:01:02] Slide rolls down and talk for hours about, you know, what a great thing is to do, to be an engineer. and along the way, I said, okay, that's what I want to be. And that was back when I was eight years old, I decided I want to be an engineer and I never looked back. It's the best decision I've ever made. I love it.
[00:01:19] Wow.
[00:01:20] Matt: [00:01:20] That's great. And do you get to do much engineering?
[00:01:22] Iyad: [00:01:22] no, I try. I pretend. No. I mean, I do a little bit, obviously running a technical company requires you to know the basics. Sure. Back when I was about 16 or so, I let them at least came here and I did all my degrees at university of Maryland and SMU and took a bunch of different jobs with different companies.
[00:01:39] I basically alternated between startups and big companies. That was my strategy from day one. Because startups, you learn at a massive speed, but you don't have enough resources ever. Big companies. You have all the resources in the world, but nobody will allow you to run at the speed you want, but you have all the resources to learn as much as possible.
[00:01:58] And so I kept alternating [00:02:00] and I'm still doing that. I'm really enjoying it. Yeah,
[00:02:03] Matt: [00:02:03] that sounds like a lot of fun. you know, this is a show about edge. And so I normally ask people, you know, how did you figure out about edge and what are you doing? But you're really about CBRS and there's an intersection of edge.
[00:02:12] And we'll talk about that, but why don't you tell me. Like how you even got involved in CBRS.
[00:02:18] Iyad: [00:02:18] Sure. And by the way, edge, everything is edge these days. So it doesn't matter what you do. You're part of the
[00:02:23] Matt: [00:02:23] edge
[00:02:25] it's true. It's true. I mean, I think in five years, we're not going to say it anymore anyway, it's just going to be the internet
[00:02:31] Iyad: [00:02:31] yeah, that's right.
[00:02:32] I do. I do absolutely believe that. But back to CBRS, when I was at sprint. we worked really hard with SoftBank, to figure out how to create real mass scale in the wireless business. the way the discussion got started is, SoftBank was trying to get some headsets provisioned with a certain spectrum band.
[00:02:52] They owned and they didn't have enough spectrum. So it took a while to get Apple to come on board. So as part of the plan, they said, Hey, if we team up [00:03:00] with sprint, we create a big ecosystem with a lot of 2.5 spectrum. We can get. A better cost, better technology, better everything. And I was part of that project.
[00:03:10] And at the end of it, I sort of learned a very basic rule that a spectrum now is a scale business. the more spectrum you have and if you can organize it the right way, then you can create just, you know, a step function in cost and performance. And this is was way before people are even talking about 5g.
[00:03:28] A lot of these concepts are now built into 5g. And so after I left sprint, I said, where can I get my hands on some spectrum that has all of that scale and, cost advantages and performance advantages. And maybe I can figure out a way to build an ecosystem without being tied to anybody specific. Randy, excuse me.
[00:03:48] Matt: [00:03:48] When did you leave sprint? Just
[00:03:49] Iyad: [00:03:49] I left sprint. I was in 14. Yeah. So 2014, I ran into a couple of scientists who have been working on, you know, very smart [00:04:00] scientists. Who've been working on a white papers, and beginnings of proof of concept of how to create, this sort of shared spectrum. And the minute I looked at it, I was like, wow, that's it.
[00:04:12] This has a lot of spectrum that's configured in the right way. If somebody can build a real production systems around it and an ecosystem, you get all the advantages that a carrier typically today in terms of scale and performance, but you can give it to anybody that wants it. and, I stuck with it.
[00:04:27] And when we S when I started was two engineers, and now what about a hundred? We have an operational system. We have over 50 customers direct, and about another 70 indirect customers. And we're just building.
[00:04:40] Matt: [00:04:40] That's awesome. There's a lot to unpack there. So there's a lot to unpack there. So first off, what. w what is CBRS and how should I think of that in terms of spectrum?
[00:04:51] Iyad: [00:04:51] sure. Well then maybe we'll start with spectrum. Is spectrum is really the right to use the airwaves in a certain area, to put signal on airwaves in an [00:05:00] area that's really reserved to you.
[00:05:02] And there are different models of how you do that. There's sort of unlicensed. Spectrum means go out there. Throw out your wifi access point, use the airwaves and hopefully nobody else's there. And if you run into somebody else, so you guys can fight it over, and everybody gets the best they can. Right?
[00:05:18] And you have licensed spectrum, which is what Carriers use that says, Hey, I need the FCC, the regulatory body to make sure nobody else touches my spectrum in the place I'm in. So I can get all of the best quality I can squeeze into it. So spectrum ultimately is the right to use the airwaves to transmit signals on it.
[00:05:35] That's how people talk for wireless and how to transmit data. CBRS is a piece of the airwaves in the U S it is a very large amount of spectrum. it is almost what a carrier typically used to have in the 4g world. And that spectrum today is used very little, but by very important people like the Navy [00:06:00] and.
[00:06:00]other DOD entities. And so the challenge that was put in front of us and we've solved it is Heidi actually create, computations and sensors that will calculate out the very small amount of the usage and take the, all of the other usage and make it available to the public, you know, in a simpler software way.
[00:06:21] So think of it as, the equivalent of the Pentagon had a. At 10 apartment building, they were only using one apartment and they needed someone to write the software to be able to rent the other nine. and except that the Pentagon in this case every day will change apartments because they don't like to stay in the same apartment.
[00:06:39]and so we had to build software to be able to constantly detect where they are and, rent or make available the rest of the spectrum in a way that was applicable to the industry. Well, that's what CBRS is. It's what sharing means here in this model, sharing here means that we're able to make the spectrum that is not used, but would have been [00:07:00] reserved and would have been unusable.
[00:07:02] To the public in a software way. That's why I like the Airbnb example because a lot of, the sort of rooms that people had or houses that they weren't renting there just weren't available commercially for people. And then once you got the software tools and the internet and the application on the handsets and the business models, you're able now to make, you know, a lot more, vacation properties available in, you know, in a city that weren't available before.
[00:07:28] Because of the software and the tools. very similar model to what we did in spectrum here.
[00:07:33] Matt: [00:07:33] Yeah. So, so let me, try to say back what I think I heard, but also elaborate on it a little bit. So, when I have a phone, like the one I've got sitting right next to me on my desk, it has a radio in it, a receiver transmitter in it.
[00:07:45] And that receiver transmitter is capable of tuning too. Often many parts of spectrum, and this is just the RF spectrum, the same spectrum, that visible lights on that radio waves, TV waves are on. And in the past, the [00:08:00] FCC has taken blocks of licensed spectrum to blocks and they've auctioned it off and companies like Verizon and at and T and sprint and so on bought those.
[00:08:07] And they had exclusive access to those, so that they could offer a quality cellular service to their customers without colliding with. Verizon's or at and T signal. And what you're saying is CBRS is a big piece of spectrum that was largely unused except for the Navy. and what I mean by largely unused is like you said, it's like, it's like a building where there's a very important tenant and he, or she always wants to get.
[00:08:33] To their office, but you got all these extra rooms and you can use them as long as the CEO, isn't trying to get him to his or her office. Right. And then all the people that are putting in these other rooms just need to not bother each other. You know, it's a little bit like the software that you wrote is maybe analogous to the software that like VMware and Amazon provide in this sense, in the sense that you have a server.
[00:08:56] It has a lot of resources. I'm running one, one. You know, one [00:09:00] easy, two instance on it and it's not taking up the whole server. So, Hey, wait, I've got all that empty space. I've got all these extra rooms. Why don't I stick a couple more servers on their virtual servers? And you're just doing that with spectrum.
[00:09:10] Is that essentially right?
[00:09:12] Iyad: [00:09:12] That is the essentially right
[00:09:14] Matt: [00:09:14] okay. and this is done through software, but you also mentioned sensors and things like that. how, so th this thing that you built, it's called unfortunately, SAS, S a S not S a, well, it actually is SAS. It's SAS as SAS. So what is, what does SAS in that sense?
[00:09:30] Iyad: [00:09:30] It's spectrum access system. And that is a regulatory term, but I'll probably the best way industry-wise to describe it. It's we call it a spectrum controller. And it's a software as a service product. It's a set of API APIs. If anybody's familiar with software as a service product, where if you exercise them, you get to our system, you ask for a spectrum assignment and we give it back to you in a hundred milliseconds.
[00:09:52] So it is really sort of, part of the, cloud system that's accessible, using a set of API APIs, and peoples can [00:10:00] embed them into their applications, any applications, which typically are, say a private network. Or a fixed wireless network or a broadband network. They need access to airwaves or spectrum, and they can embed, our product software as a service type API APIs, and they'll in their code.
[00:10:18] And they're able to get to us as for assignments when they need it.
[00:10:22] Matt: [00:10:22] Yeah. So let's do a fun little exercise. let's build a private wireless network that uses CBRS. Okay. So let's imagine a scenario. I'm a mining company and, I'm. I want to, you know, connect all of my assets, all of my workers, my machines, my robotics whatever's on my thing.
[00:10:41] There's no seller provider. that I can necessarily hook up to no cell towers near me. So I think I want a private network. I mean, there's other reasons to have a private network, even in a city, but out here I want a private network. Like where would I start? What's the first thing I need to do to get a
[00:10:53] Iyad: [00:10:53] private network.
[00:10:54] Yeah. So, I think it's a fantastic example, especially for a mining company, because it's not [00:11:00] just the availability of a network, but you also would be concerned about people's safety. You'd be concerned about privacy of data. You'd be concerned about the. Ability to control that network, especially if you're beginning to roll out some automation.
[00:11:12] I mean, I could see applications like early warning systems to be able to make sure that if I have danger, I'm able to detect it. You're putting it if I'm at the
[00:11:19] Matt: [00:11:19] explosives or yeah.
[00:11:21] Iyad: [00:11:21] Yeah. Camera's to see our people safe. you're basically maybe putting some, communication systems on your machinery. you know, Above ground below ground, to be able to get data out of it.
[00:11:32] How much of your hall, where are you going? Where are you located all of that data. And you want to be able to automate and manage that data in a secure and controlled way. So what do you do? I think the first, typically the first thing people start with is the application themselves. So like, Hey, I like these cameras or I like these.
[00:11:49]terminals that I can put in the trucks or I like these sensors that would tell me what kind of components and what's the composition of the gas, you know, in certain chambers. And once you find, [00:12:00] the end application you do, then the next thing typically you ask for, Hey, what does it run on? A lot of these things run on wifi, but started from a mining application.
[00:12:08] That's probably quite difficult to be able to build a wifi network that reliable, and then. Well, because the range of wifi typically is not as predictable, as you would with a wireless carrier. The advantage of having dedicated or reserved spectrum is that you're able to have a very predictable deterministic network.
[00:12:29] You know how much spectrum you're not fighting for it. You don't know if anybody else will turn up a system next to you. You don't know if somebody has a handset with a hotspot that's turned on, that would begin to use the same AirWave. You'll have, you know, much more advanced algorithms and 4g and 5g that are built around security and power and reliability.
[00:12:47] Again, nothing against wifi, just that, typically wifi is used for, opportunistic usage, opportunistic phone, download opportunistic access to the internet, where, you know, some variability on the performance, [00:13:00] isn't gonna kill someone. and a lot of the sort of very, reliable. secure networks that people are building for industry for Datto type automation, like in a mining operation.
[00:13:10] They want to know that camera is always going to work. The speed of transmission is always going to be high, that the sensor will have a five nines availability that nobody's able to going to be able to listen in or break in and try to figure out where their trucks are or. What they're hauling and what the health records of their people are, or what's happening underground.
[00:13:29] Or if they have an advanced warning system, nobody wants that to be out within the public internet. So that creates another set of the next set of requirements in terms of the reliability of the network, the capability, the capacity, and typically you'd need some dedicated airwaves or spectrum to be able to say, it's all mine.
[00:13:47] And I can actually close it off and manage it, then put the right edge compute on it and the right, you know, wireless systems on it and all of that. So once you figured out what the applications are, you basically go to the manufacturers that you're [00:14:00] buying these cameras from and say, what sort of systems do you have support?
[00:14:03] And right now, a very large number of them support, CBRS or shared spectrum, either directly or through dongles that you can connect to. And the next step would be for you is to figure out, okay, now how do I build a network in order to be able to connect these devices back, you have typically have to do two things.
[00:14:21]you need the set of servers or the edge cloud type complex to be able to store all that data and run algorithms on it and be able to make decisions on it and display it and trend it and all of that.
[00:14:32] So, you know, a portion of your team would be concerned about. Hey, how do I find that edge compute box and how do I connect it to fiber and make sure that it's in my secure location and all of that, another set of people would be thinking about how do I put the radios or the access points that will actually be strategically placed to be able to talk to all these cameras and sensors, collect over the airwaves from them, what they need, then be able to pipe it back into this edge compute.
[00:14:57] And so, typically people call that, you [00:15:00] know, wireless connectivity network for the latter part. And an edge compute, slash fiber for the first part I described, the good news about these connectivity and networks for wireless is that, in the old days you used to be able to use, to have to be an expert at it, hire wireless engineers that have things like 4g and 5g in their title.
[00:15:20] And. You know, a lot of three letter words, you know, PhDs, etc. Yeah, we don't need that anymore. I think the model has evolved quite a bit right now into managed service. So there's a lot of people will come installed, managed, deploy these sort of wireless networks, connectivity as a service solutions, similar to manage Wi-Fi.
[00:15:40] But there's, that's probably one of the fastest growing concepts right now, or offerings in, from carriers and from SSI partners and from equipment makers is managed wireless services and managed wireless networks. And then you also still have to go find people that will help you figure out what edge compute to deploy.
[00:15:59] And luckily as [00:16:00] well, there are a lot of people out there now. you know, at the innovating on the innovation side that are willing to come and deploy an edge compute complex right next to you, and be able to support you and collect that data and connect you to the fiber and help you manage it. So that's sort of how this evolves.
[00:16:15]and then assuming you find the right partners for connectivity and edge compute, you would spend all your time now into refining your application. Automating it. Put the right algorithms on it, be able to predict what's going to happen. Be able to make your operations more efficient, remove some of the excess resources that you didn't need.
[00:16:34] Be able to figure out how to apply, you know, data and science to it, to understand, Oh yeah. W we deployed it. We really didn't know that this portion of the system wasn't working well until we collected all the data. And that's the vision of where everybody's going. People want to own their data. They want to own their operation.
[00:16:55] They want to be experts at running their operation and digital model, and they [00:17:00] want to find the right partners to help them build the applications, the networks and the edge boxes.
[00:17:05] Matt: [00:17:05] Yeah, that makes a lot of sense. a couple of questions about that. So let's say I, I hire this, the system integrator and the system integrator, works with a bunch of vendors that sell the radios and things.
[00:17:17]let's say that I have, you know, four acres. And I'm just I'm oversimplifying this, right. Because I said it was a mine and stuff's underground. Let's say I'm just doing the surface area for four acres. And when I think of cellular communications, I think of like these giant towers, I know that there are small cells and things, but so what am I actually installing on this four, eight?
[00:17:35] Like how many radios, how big are they? Are they on poles? Are they on rooftops? Are they sitting in the ground? what? And I realized there's a lot of choices there, but what would be a typical.
[00:17:43] Iyad: [00:17:43] a lot of choices. I think maybe the typical deployment these days is that a warehouse on these four acres, you have a big warehouse where the trucks come out and in a corner you have some office and office location.
[00:17:59] And then [00:18:00] maybe also this is where you're basically monitoring and maybe do an assembly and all of that. So in that typical warehouse, you typically need maybe, anywhere between one and 10. Different. I think of it as ceiling type installations of, little rectangular boxes that looked like the wifi boxes and the ceiling.
[00:18:18]the systems vary between one that can have a lot of power and can cover all four acres, but it can be very expensive and very hard to design and optimize because it needs to do a lot of things into really simple, cheap, 10 boxes that are. One 10th, the cost and you don't care where they are and you can sort of, you know, I'd use the term loosely here, but sprinkle them across the location and they can work with each other, talk to each other and figure out how to split the load between them until support you and, you know, different people have different sort of
[00:18:51] Matt: [00:18:51] yeah.
[00:18:51] And your system integrator is gonna help you figure out what's optimal for your system and your budget
[00:18:54] Iyad: [00:18:54] and sprinkled connected to your ether net network, wired into the ceilings [00:19:00] wire and into the walls. And then, connected back into you, whatever ethernet network that gets into that warehouse, more or less, that's what it's going to look like.
[00:19:08] Matt: [00:19:08] Okay. Okay. That makes sense. now, when I talk to people in the mobile world, right, they talk about, you know, the mobile core and things like that. Like w what else do I need?
[00:19:19]Iyad: [00:19:19] so for us, the good news is a lot of these concepts that are slowly sort of melting away, right? it used to be that the systems had to be quite structured.
[00:19:28] There is a core, and there is a. Software on the radio. And there is a mobility manager and an identity manager and all of that, these days, that's not what you need. You just need the radios, connect them and you need some sort of controller for that system. And it has a bunch of software and that controller is a software product, and that's typically that fits inside the edge compute because the most efficient way to do it right now is to turn that software controller for your wireless system.
[00:19:59] Into a piece [00:20:00] of code that fits into the same box, where you collect your data and manage it and manipulate it because the world has sort of got inverted Dana 10 years ago, managing the mobility and connectivity over wireless was the most important thing to do when we first launching iPhones right now, completely the opposite.
[00:20:17] Now it's actually the volume of data and videos and all of the data you collect is a hundred times the volume. Of whatever software you need just to connect to these wireless systems. So now Ben, that's how edge computed was born because we ended up with infrared wireless networks because you ended up with the, Hey, I don't really want all that connectivity boxes.
[00:20:36] I just need the box that has the data and all the logic that is my edge compute. Can you shove whatever software controller for wireless you need on it? So I don't have to buy a second box and that it really is sort of a simplest way. I can describe the status of the technology right now. There's a bunch of access points.
[00:20:53] They connect over ethernet. They make it back into this edge compute box. The edge compute box now is shoved with all sorts of [00:21:00] software to manage control and be aware of what your wireless network is, what your cameras are doing, what your sensors are doing and the real smarts. And what do you do with that data?
[00:21:08] How do you start a tidy replicated? How do you analyze it? How do you put AI on it? How do you present it? That's the real work right now.
[00:21:16] Matt: [00:21:16] Yeah. and just to be clear, this, controller box you're talking about is literally a server that you could stick in a rack somewhere. I can have multiple servers
[00:21:26] Iyad: [00:21:26] where on the, on somebody else's server, like for example, if you already have a server, which is an edge compute, that controller is nothing but another software layer on the same exact box, you don't even need a separate rack.
[00:21:37] Interesting. Okay.
[00:21:39] Matt: [00:21:39] It depends. Depends on what you're trying to do. I mean, you've got a bunch of Nvidia GPU's and you've got, you might have a full micro module data center. If you've just got a rack, you might have an equipment closet and it might have two, two servers in it.
[00:21:51] Iyad: [00:21:51] Amazing part right now is that edge compute boxes.
[00:21:53] Now can do everything a hundred times better than any server you could buy. So it becomes, if you pick your edge [00:22:00] compute properly, you never have to put another server rack next to it.
[00:22:04] Matt: [00:22:04] Tell me more about that?
[00:22:06] Iyad: [00:22:06] Well, because, it's computers all about specializations of processors. You have processors that do CPU, you have smart Knicks.
[00:22:13] Now that can do a lot of processing, right? that connectivity area you have GPU's that does a lot of processing for either AI or video. Or this is just another application. If you have the horsepower to process high speed and high volume video. Oh,
[00:22:29] Matt: [00:22:29] she was saying, I was going to say you don't need another server for the controller.
[00:22:32] Cause it's just compared to everything else. It's a tiny piece of software. Gotcha. Got it. Okay. You're saying you don't need sir, or the edge compute and I'm like, I know deep neural network I know is going to run on a small server box, but okay. But I totally understand you're saying, okay, so now, I need some kind of activity.
[00:22:49] Back to the internet because I'm going to purchase, I need to allocate, I need to have access to spectrum, right. Because I don't have any specifics. So let's now tie that into what federated does, because I didn't [00:23:00] buy any spectrum. I didn't buy it at auction. I didn't go to a big carrier to get some, what do I do now?
[00:23:05] Iyad: [00:23:05] Well, our software does two things. Number one is that it's embedded in that wireless network. We described it talks to us directly get spectrum assignment directly from us and we'll make it available in that location for you. And we'll track it. We'll monitor it. We'll make sure that, so think of it as we're giving you access to the airwaves in that location.
[00:23:23] You're in that four acre and we're putting your name on it. We're letting everybody know that you're the one using it. And if anybody else tries to use that same spectrum, we'll do something to protect you. And so we're basically giving you a virtual spectrum just for you. We give you the assignment, we'll make your system, use it.
[00:23:40] The radios. Know what part of the spectrum it is. The handsets know what part of the spectrum it is. the, all of that's done through software to tell your cameras with part of the spectrum to use. They know it's your respect for them. They get to use it. And, we, meanwhile don't give that spectrum to anybody else.
[00:23:55] We'll say that piece of the spectrum we gave away to the mining company. You don't touch it. [00:24:00] in terms of connecting back to the internet from the edge box. You know, I think the two most popular things is that if you have fiber connectivity, obviously that's preferred into that location. And if you're doing a lot of volumes of data, you may do that anyway.
[00:24:13] But if you don't, the other way we use our spectrum is what's called wireless back haul that wireless broadband, we can actually give you another set of spectrum and put a wireless link back from this box into the nearest location where the fiber is. And that's another very popular. Now wireless broadband is becoming very popular.
[00:24:31] Matt: [00:24:31] Yeah. Okay. That makes a lot of sense. and then, so, so you mentioned that, okay, I've got this, slice of spectrum that I've, been allocated by you. And, I had to pay you a little bit to do the figuring on that and say that I've got dedicated, but I didn't have to pay for that spectrum.
[00:24:46] Right. Like it's just mine free. I can use
[00:24:48] Iyad: [00:24:48] it. Yes. It's the most, it's the most thrilling part of CBRS is that it truly is free
[00:24:54] it's really , democratizing
[00:24:56] Matt: [00:24:56] wireless access. That is really fabulous. [00:25:00] They picked a horrible name for
[00:25:01] it. Yeah, no, I guess under any
[00:25:05] names, radios,
[00:25:08] Iyad: [00:25:08] CBRS is sort of is citizens' broadband.
[00:25:12]radio or something. Yeah. Right. So you don't even know is that it's free spectrum. It's available to you and we'll manage it for
[00:25:20] Matt: [00:25:20] you. Yeah, that's awesome. Okay. So, so now what happens if the Navy wants to use it?
[00:25:26] Iyad: [00:25:26] Oh, the spectrum, actually, we can move you around within a range. We have like 15 different slices of the spectrum right now.
[00:25:34] So if we give you one and the Navy was trying to get into the same slice, we will seamlessly. Without even you knowing, move you to another slice by telling all your equipment, your radios, your cameras, what to do to move them to another slice and it'll happen instantaneously without anybody knowing or noticing.
[00:25:52] And, because it's all software defined. That's software managed. We move your total layer that the Navy is not in that's part of what managing means is we [00:26:00] move you if we need to, in order to protect you from the Navy or protect you from your neighbor that's factory too. And so we're protecting you from that too.
[00:26:09] Matt: [00:26:09] Yeah, that's really great. And then, so, so are there other companies that provide, spectrum as a service services? And I imagine
[00:26:17] Iyad: [00:26:17] there are, yeah, it's cool. I mean, we're about 52% market share right now and we hope to continue to do well. The, our top competitor is the small company called Google. So,
[00:26:31] Matt: [00:26:31] and more market share than Google.
[00:26:33] I think you're the only person on the planet that can say that.
[00:26:36] Iyad: [00:26:36] Well, I've been seeing, how long can you get this ground? Well,
[00:26:40] Matt: [00:26:40] the reason I asked that I wasn't trying to figure out who your competitors are. I was asking you, I should say like, so you're both allocating the same spectrum. Is there some master database that the government runs?
[00:26:49] Like, how does that work?
[00:26:50]Iyad: [00:26:50] we actually, it's more of a cloud. it was also democratized if you and we have a wave to connect each other and negotiate to make sure we all have access to all the [00:27:00] spectrum. And when one of us wants to assign it, we tell the other that we're about to sign it. And, you know, so that update your database.
[00:27:07] We give it
[00:27:07] Matt: [00:27:07] away. that's amazing. Now, one of the things that you and I are both familiar with and bringing a new industry, To life is you can't do it alone. You need an ecosystem and you need competitors, frankly, to grow it. Right? So, so you and your company, federated have put a lot of energy into building it ecosystem.
[00:27:27] There's the CBRS Alliance and other things. Can you tell me, what initiatives you've created or participated in and what those were about?
[00:27:34] Iyad: [00:27:34] Sure. I mean, and that's sort of where the relationship with Google can get maybe a little bit of an asterisk there. Yes. We're compete. But at the same time, we collaborate and creating a very open ecosystem and it's not just us, but a few others as well.
[00:27:46]we have over 200 companies that join. So when we first started, we made a very conscious decision that this has to be an open system, because success in spectrum is really about, making sure that there are cameras that use it and [00:28:00] there's phones that use it. It's less about. The spectrum itself, but about, the end application that are configured for it.
[00:28:06] And the only way you can build an ecosystem is to create users and you bring as many partners as possible. So about 50% of what I spent doing in the last five, six years as the recruiting partners in the ecosystem, more than even building a product, you know, carriers are part of it. Cable companies are part of it.
[00:28:24] They just went through the auction and spent over a billion dollars, equipment makers, almost all wireless equipment makers are in it. Over 40 equipment makers build now commercial systems, all the handset makers are part of it. I iPhones, Androids, Samsungs, LGS Motorolas and a few others. There are tablets on it, their laptops, about to come in now from some of the biggest manufacturers, wireline companies are in it.
[00:28:48] They went to the auctions as well to try to get some spectrum. and I'll explain the auction in a minute. And, people like, you know, small wireless companies, you know, called wireless ISBNs that are [00:29:00] scattered across the U S there's hundreds. If not thousands of them are also part of it. enterprises are part of it, utilities, educational facilities, some universities, hospitals, industrials, oil, and gas.
[00:29:15] A lot of people are part of it. And, ultimately it's the only way to do it. we build it in a way that you can come participate. We're not going to charge you, but you have to come in and be willing to embed it in your systems, edit your phones, edit your chip sets, make sure that you're committed to build the ecosystem.
[00:29:32] And all we really do right now is facilitate the management and enable the system. We pass more than 90% plus the value of what's happening and what the spectrum offers. Over to our customers. we're more focused on enabling the ecosystem, enabling our customers than anything else where didn't enablement platform in the end.
[00:29:52] Matt: [00:29:52] Yeah. Millions of small transactions that allow other people to build businesses inexpensively on top of it. I think makes a lot of sense. And this organization you're talking about, [00:30:00] were you referring to the CBRS Alliance?
[00:30:01] Iyad: [00:30:01] Yes. CBRS Alliance anybody can join and a lot of really good people
[00:30:07] Matt: [00:30:07] join.
[00:30:08] Yeah. Yeah. I've been to some of their events. It's really good.
[00:30:10]
[00:30:10] So, so you mentioned auctions and I'm a little confused because I thought the spectrum was free. Why would I have to buy it? If I can just use it for free?
[00:30:18] Iyad: [00:30:18] Okay, this is quite complex. So I'll try to explain it. Half of the spectrum is free all the time at a minimum, right? So you don't actually have to do anything.
[00:30:27] The other half was set up in a way that you can actually get a priority ahead of other people to use the other half. So it's chaired, it's still open, but if you come in and the spectrum is all used up, and if you had paid for, the priority license, which are the priorities placement, We will make room for you ahead of others because you paid money to do that.
[00:30:50] So people ended up paying 4.6 billion for the, to be able to get a priority ahead of others. For half of them, that's a lot of [00:31:00] money just for priority. It tells you how valuable it is. That's a reflection of the value of spectrum in general, but it's a reflection of the application value that they're going to build on it.
[00:31:09] And it's also the value of the ecosystem and all the things people want to build on it. Yeah.
[00:31:15]Matt: [00:31:15] that's really interesting. Now, another question I have that I'm curious about is, you know, you mentioned at the top of the interview that, you started federated wireless after a meeting, a couple of these PhDs at the Pentagon.
[00:31:27] And so I'm thinking about, you know, pretty some hardcore RF engineering involved in this. And, I'm connecting that to your statement that Google is your biggest competitor. And I'm wondering if. Like part of the way that you and Google are competing against each other. And obviously the other folks that are in this business is in the magic of how you allocate the spectrum or do you all use the same algorithms?
[00:31:51]Iyad: [00:31:51] some algorithms are standard. And some algorithms are, individual. I would say the, we have about 20 PhDs on staff just to give you an [00:32:00] idea. Wow, this is really, it's an algorithm business in the end of the day. That's actually why Google is good at it in terms of being part of it is because they're good algorithm.
[00:32:08] It's also you also, so the science that went into it are four pieces. First is you have to create the algorithms that I think of it as a heat map, where you're always watching where the signal is being used by whom in order to know what part of the. The map is open. It's like watching a traffic network.
[00:32:23] Matt: [00:32:23] How
[00:32:24] Iyad: [00:32:24] do you get that? We get a lot of data from the access points themselves that are turned on, but we also get a lot of other data databases that we feel in about terrain, about the kind of structure of buildings. We get some data from the FCC, we get a lot of data thrown into it. so that's the first piece is how do you build these algorithms to know that your system is available?
[00:32:45] The other piece of it, that was really a lot of the innovation we drove was. To build a set of sensors. think of these as sort of listening devices, specialized listening devices, highly secure with some machine learning that sits on the coasts and look at what the [00:33:00] Navy is doing. So we know at any given time what they're doing and where in order to make sure we never really put them in a tough spot.
[00:33:07]and that requires a lot of optimization as well, because the key for it is to be able to see them coming. W how they're switching signal, be able to react in real time to their switch at the same time, be able to move people around them and do all that without anybody knowing that you did it. So all the software around it, all the algorithms.
[00:33:25] So that's a, it took us at least two years to deploy that network. We launched it a year and a half ago and it's taken us at least a year to optimize it. that's really a source of pride for us, for sure. And then the third piece is that this is a massive system in terms of the amount of data and the response time you need.
[00:33:42] So a lot of cloud scaling and we're proud of that relationship with AWS. What we will be working with them on cloud scaling auto scaling, distributed algorithms. How do you actually split up the data? How do you manage it on partition it and put it back in so you can offer people, you know, sub-second type response.
[00:34:00] [00:33:59] Consistently at five nines availability, no matter what they throw at you, a lot of innovation went into that. And then the last piece of sort of what goes into it is to be able to support your customers. at the front end, everything from customer management to tools, to help them plan, to help them analyze, to help them predictive analytics, you know, all the things they need to do to partition their spectrum, resell it, all of that, all the feature engine on top of that.
[00:34:26] So these are. That's a lot of innovation that's at the end of the day, like I said, we're a software company that sits on top of an open ecosystem with a bunch of partners built around a very innovative spectrum model. That's enabled by a very, really innovative government. the FCC and the Dov and NTIA and the white house.
[00:34:45] What they've done here over the last, you know, multiple years now, 10 years plus in terms of creating this innovation model has been just a. Really a first in the world and hopefully more of that to come. we are working now on an extension of it. The white house [00:35:00] just opened up another, or the FCC has just opened up another hundred megahertz, which think of it as another big chunk adjacent to it that, we're beginning to define that looks like an extension.
[00:35:11] We'll keep adding to it, to continue to give people more access to spectrum and, enable them so they can build the, as many of these private wireless and edge compute solutions as they want.
[00:35:22] Matt: [00:35:22] Now, if we had this idea in this technology back when we first started selling spectrum to the carriers, what would the world be like if it was all shared?
[00:35:35]Iyad: [00:35:35] you will, we, at a minimum, we can double the amount of spectrum. So at a minimum you may be able to do 10 X depending on the location and the congestion. More importantly, the things we wanted to accomplish here it is number one, why participation? You know, you had to have billions to be able to get into the business of wireless.
[00:35:53] Cause you had to build really big amounts of spectrum nationwide. Now you can come in for free, right? Literally you have [00:36:00] one building one room and we can get you into the business and you can own your own spectrum. And we can give you a free ecosystem and I-phones will work on your spectrum. And all of that, these are these things were completely unheard of.
[00:36:10] It's the white participation. That's really important because there's a lot of businesses now that need to transform their business to be wireless as well. They can't just basically outsource the technology. They need to be technology companies. I think the second piece of it is the speed by which we're able to make the spectrum available.
[00:36:28] Typically when you're doing it on exclusive basis. It takes years and years to completely clear it, move all the existing users have all the Navy reprogram all their ships, which is like 20 years. Right. And $20 billion. Then it's available in an explicit basis here. We're saying, I don't care. This is like Airbnb.
[00:36:45] Don't sell the house. I'll rent the room for you tomorrow. Right. And that's basically what we're doing. We make the spectrum available immediately. That's great for economics. That's great for innovation. And then the third piece of it is that just the, all the new ownership models that didn't exist. [00:37:00] Right.
[00:37:00]you know, it used to be, you used to be organized around, a wireless service as sort of what enterprises go by, what? And then they basically build applications on top of it. What was probably most. Exciting for us. As we saw people coming to CBRS is that there's now a retail spectrum model. There are now enterprises that value getting access to spectrum, and then they get to choose what technology to build around it.
[00:37:26] They get to choose the radio partners. They get to choose the solutions. They get to optimize it. When we go from few people, building technology on spectrum to thousands and thousands, if not millions of deployments. The amount of innovation or you're getting into the software technology ecosystem is just going to go through the roof because that diversity of users mean diversity of ecosystem in diversity of suppliers.
[00:37:47] I mean, innovation is just going to take off that's the democratization in the best way possible. The us will become the most innovative by far in creating solutions and applications and technology for wireless, [00:38:00] because there'll be able to get a thousand enterprises on a thousand companies, all innovating at the same time.
[00:38:05] If you look at the enterprise space, anything that's really open to the enterprise space, security, routing storage edge, the amount of innovation there is just breathtaking and very fast because it's a really big market with a lot of people. We're about to do the same for wireless. It's amazing.
[00:38:23] Matt: [00:38:23] W when did it first become available in such a turnkey fashion?
[00:38:28] Cause I remember as recently as a year ago, like if you wanted to do so we were trying to do something in Chicago. We actually had to go get some special, so that doesn't exist anymore. So tell me, what does it first been like? It's turnkey, as you just said, where literally you put the radios up, you hook it up, you turn it on.
[00:38:42] Yeah, right? Yeah.
[00:38:43] Iyad: [00:38:43] In January it became completely, you didn't have to talk to anyone we have now total. In the system close to 50,000 nodes that are deployed right now, and growing right then this is all the industry, not just us. I think within a year from now, you'll [00:39:00] be talking hundreds of thousands of nodes and.
[00:39:02] Four or five years from now, you'll be talking millions and millions of notes. that there's still a lot of enablement and simplicity being added to it. It's not a hundred percent there. Obviously there's a lot of innovation to come. All of the things we talked about earlier, making the nodes wireless nodes, simpler, making the controller simpler, making the integration with edge compute simpler.
[00:39:21] The integration of the fiber simpler, making the application simpler. All of that with really the goal ultimately is that, you know, these things are plug and play networks that you can focus all your energy on the application and on growing your business, controlling your business, managing, automating, mining your data, becoming as good of a technology enterprise as you are as a mining company or as a warehousing company or a trucking company, that's really the goal.
[00:39:46] Matt: [00:39:46] Yeah. And you know, you mentioned the, you know, the example of the edge computing and the mining example. and I think that really the reason CBRS and edge computing go hand in hand is because the reason you put up [00:40:00] a private network, In these environments. isn't so that humans can pick up the phone.
[00:40:03] And I mean, sometimes it's so we can talk to each other, but it's more so that these devices that are on 24 seven can generate data. they don't have to be, have a fixed connection. they can be in motion because what are the things that sell your technology is way better than wifi is handoff.
[00:40:18] Sure. so I got all these advantages of a really reliable. discrete, I think is the word you use discreet SLA, network that I can count on for my critical data. and the value of that data, presumably, decays over time. And so some portion of it will benefit from processing, in near real time.
[00:40:38]and the premise of. Edge computing is that there are workloads that require that, either because you can't afford or don't want to ship all the data back, or you want to keep the privacy, you know, there's a lot of reasons for it. But so T tell me where you're seeing the most interesting intersections between CBRS and edge computing.
[00:40:54] Iyad: [00:40:54] Sure. I mean, I think the most basic concept is that the warehouse of the retail store now. [00:41:00] Ha processes as much data has access as much automation opportunities that we used to be a city worth of a network 10 years ago. Right? All the things you can do in a warehouse, all the, you know, barcoding, the robotics that moves boxes around the security cameras taught to me, the environmental sensors, all the things that you can do, there are all available today.
[00:41:20] All of these sort of edge IOT capabilities are there. But they also generate tremendous amount of data and they need to be operated on, you need algorithms to run them. You need automation, you need to be able to trend them. You need to be able to see the videos you need to be able to put in a machine, learning on them, all the things you need to do on them.
[00:41:36] Right. And so, basically what CBRS does is creates a connectivity bubble, right? In that location that connects in a secure, predictable way. These end points that are generating and acting with the box that has all the algorithms and data. And you're able to keep it all in the warehouse. So it's secure, controlled, and it's yours and available and low latency and high quality.
[00:41:58] And it's your [00:42:00] network.
[00:42:01] Matt: [00:42:01] Yeah. So you're, you mentioned your relationship with cloud providers and, can you help us understand what the, what that's for? Like why you have a relationship with cloud providers and what it provides to the end customer?
[00:42:11] Iyad: [00:42:11] Sure. I mean, I think. You know, our heritage has always been enable everybody.
[00:42:16] We want everybody to be able to use it. It's an open ecosystem. We want to create as much access to this as possible. so the two end points I described are actually are the top two investment areas for cloud companies. They've invested billions in the end points, the IOT management capabilities, identity management, all of that, but they also invested quite a bit in creating the edge, software and the edge solutions and the connection between the edge and the cloud.
[00:42:40] And, we fit neatly in the middle in terms of offering them a private connectivity and a manageable connectivity that fits right between the end points and the edge. And we do it in a way that is on by the enterprise, because that allows our partners to be able to wrap the whole thing in, into a complete private network, [00:43:00] including the edge devices, including the edge computing includes the connectivity itself into one.
[00:43:06]automation solution that can be at a hundred percent of the control of the enterprise.
[00:43:12] Matt: [00:43:12] Yeah. You know, what's interesting about that is, I think that's definitely a line of thinking there's a whole set of it, organizations that are like this subset of everything we do. It's got to stay on site.
[00:43:23] It's got to be in our world, but, there's also this reality that. Increasingly, workloads are moving to the cloud and, you know, Amazon makes a pretty convincing argument that they're way better at security than private it organizations, because they're protecting millions of people and they've got a staff that's gigantic and they can hire, you know, a hundred PhDs that do nothing but security.
[00:43:43]and so do you see that. Need to control onsite. Cause I mean, honestly having computers on sites, a pain in the ass, right. You just gotta have an it team and that's the cloud is great. Right. So do you see a world where I can provision all this in the cloud and just hook the radios up with a [00:44:00] fiber connection back to some edge cloud server and that's running.
[00:44:03] You know, all these workloads and even the controller. is that something you see coming?
[00:44:08] Iyad: [00:44:08] Yeah, so I think there'll be a lot of variation. So, and I have to drop off right after this. I have a customer called here and I can call back if. I think that, that decision of what to put in the cloud with the point of the edge is another function for your edge compute box.
[00:44:19]that's another way to control it. Yes. You can see variations where everything's controlled in the cloud. You can see it in places. All, everything is on the edge and most likely are going to see division of labor. And so that's another function you're going to see in the edge computer. You may have a really stripped down and edge compute to collect the data, but you don't have all of the backup capability and the storage and management throw that into the cloud.
[00:44:40] You might see it into the edge being just simply a router just around everything out of the box, out of the building. So these are all variations that you will see, and you'll see some places where it's a hundred percent contained. and everything is processed there and never read the cloud. I think these are decisions that get made as part of the edge compute strategies that people are deploying.
[00:45:00] [00:45:00] Matt: [00:45:00] Yeah. Thank you so much for sharing your knowledge and enthusiasm for this with us. I have one last question and that is if people want to find you online, want to find you personally, or federated where it's the place that you go.
[00:45:13]Iyad: [00:45:13] LinkedIn is the best place to go. Thank you. Okay.
[00:45:16] Matt: [00:45:16] You bet.
[00:45:16] Iyad: [00:45:16] Thanks. Very good.