 get going. So my name is Ben Hilburn, I run the Gunradio project and so there's I think the significant portion of the talks in this room are Gunradio based which is awesome. Part of my job obviously is project lead is to evangelize the project so I'm gonna try to convince everyone here that you should be using Gunradio. Actually it also is part of my evangelism duties. I did bring a whole bunch of stickers so if you do not have a sticker yet please grab one. As a quick survey I'm trying to get a feel for what the audience is like. I recognize a lot of people here there's a lot of the core Gunradio developers in here in the room. Can I get a show of hands for people who have never used Gunradio before? Wow okay this is awesome. Alright how about a show of hands for people who've never used a software-defined radio before? Okay alright that's an interesting mix. So I'm gonna give, this is one of the first talks of the dev room, so I'm gonna give a kind of a brief intro to software radio, a brief intro to the Gunradio and then it's actually it's not a very technical talk. My goal is to kind of give you an overview of some of the cool stuff that's happening in the community, talk about some of the stuff we're doing and hopefully convince you to get involved and play with some of the awesome Gunradio applications that have been made recently. So with that I'm gonna get going. So real quickly what a software radio? One of the two of the most prominent standards professionals bodies in our in our industry are IEEE and the Wireless Innovation Forum. This is the official definition accepted by everybody radio in which some or all of the physical layer functions are software-defined. What that actually means is instead of sitting at a breadboard and designing hardware that does math or you know fabricating a PC or a CCA that does math, you write software that does math right. So you create programmable radios and you write software that defines your radio algorithms for you. For the record I consider software radio and software-defined radio to be interchangeable. I've never heard any definition that distinguished between the two in a way that made sense. So Gunradio itself consists of a bunch of tools, a scheduler runtime and a lot of block IP that lets you create applications that do really really cool stuff and there's a lot of people who are really involved with the project who make really really awesome blocks and it's everything from really basic blocks like adders, signal generators to things that are extremely complex, polyphase filters. I see I have some examples up here. OFDM cyclic prefectors. For those of you who are not familiar with wireless communication standards, OFDM for example is what's used in 802.11 or Wi-Fi, LTE and that sort of thing. And the variety of blocks that's in the project is fantastic and it's only extended by all the work that's done in the community, what's called out-of-tree modules which I'll cover in a second. So you chain these blocks together and what we call flow graphs. So as an example here this is a very basic flow graph that's taking RTL-SDR source which is you're seeing a little RTL dongle like $20 little receivers, passes it through a basic flow graph and dumps it to the audio. So this is actually a very basic FM receiver that you can make with GUNI radio. So my high-level pitch for GUNI radio is you can do basically anything you want, the software radio, you don't have to use hardware, you don't need hardware to use it, you can do simulation, you can do hardware and loop prototyping as you're building stuff up and then in the end you can actually put it out and deploy it whether that's using, attaching a radio like this little one I have here to your PC or actually sticking out on an embedded device and there's a significant chunk of the community who actually uses GUNI radio on embedded devices. So with that I'm going to actually, nope, out-of-tree modules. So the one thing I really want to point out before I start jumping into specific applications, we have this concept of out-of-tree modules in GUNI radio. We try to keep the primary distribution as clean as possible and that means that every new application that comes out, if we try to pull in everything that's being done with every new wireless standard the GUNI radio distribution would just explode, right? So every time somebody wrote a new version of 802.11 standard or a new cellular standard or a new module to track airplanes or a new module to track ships, a new satellite communication standard, the GUNI radio would get very, very big. So we have this concept of out-of-tree modules. As people create these they post them on GitHub and then they show up here on SeaGran. So SeaGran is our version of Pearl's archive network. So it stands for the GUNI radio, the companies of GUNI radio archive network. If you go to SeaGran.org this is the website you'll find. It's actually auto-populated from what's called PiBombs. And PiBombs is our out-of-tree, or it's a package maintenance system effectively. So you don't have to use PiBombs install GUNI radio if you're on any of the major Linux distributions you can have to get install GUNI radio or DNF install GUNI radio. If you're interested in using out-of-tree modules, PiBombs is a great way to do it or you can build it yourself. And there's a lot of really, really great stuff already in here. I already gave a few examples. That's interesting. So if you've seen any of like the big applications that made a lot of the tech news like tracking error traffic, mapping ships in real time over Google Maps, all that kind of stuff, all of that code, those applications are here on SeaGran. So you can just go grab those, connect your SDR and do really cool stuff. I do want to talk about GSOC and SOCUS real quickly. We've participated in GSOC for four years for the last five and done SOCUS for the last two. Some really, really awesome stuff has come out of these programs and we're generally considered this to be very successful for the project. And actually one of the presentations later today is from Sebastian Mueller, who is one of the successful GSOC students from last year. If you're interested in working on GENERATING and would like some funding, these are great ways to do it. We will be doing both or applying for both GSOC and SOCUS this year. Martin Braun is our community manager and he's leading this effort. For those of you who aren't familiar with the way GSOC works, we put together an ideas list and we submit it to Google and Google says yes they sound good or no they don't. We've had pretty good luck in getting accepted and we have a lot of really cool ideas for this year on the general side. It's everything from C plus plus code generation, Android work, QT graphics, DSP stuff, security stuff. And this kind of touches on one of the major topics that I want to talk about which is that software defined radio as a field I think is extremely interesting because it's very, to use a corporate buzzword, it's a very cross-domain. Just about everything you can think of in terms of, thank you Phillip, so everything you can think of applies to software radio whether it's you know kernel level OS implementation details, if it's networking, it's DSP or FPGAs, embedded graphics work. Pretty much every technical area somehow relates to software radio and thus to Gino Radio. So even if you're just getting started here you're not, you know, you couldn't build an LTE modem to receive cellular traffic, that's fine. You don't have to be a DSP expert. You don't have to be a wireless expert. You can contribute at pretty much any level and there's lots and lots of ways to get involved. Okay, so the Gino Radio conference. We've had six very successful years. Last year, GRCon 16 was awesome. We had 304 attendees and 20 sponsors. It was hosted at the University of Colorado Boulder. We typically do a full week, four days of talks, one day of a hack fest. And last year was our first year doing a hacking challenge which Bastille did, which was actually really cool. This is my plug for it. We're in the throes of finalizing San Diego for this year. So it will be in mid-September in San Diego, California. We'll have our call for papers and talks out pretty shortly. So please plan to come if you're interested. And then much like everything else, it's at all levels. If you've never used a software radio before, it'll be useful and interesting and you can come ask us questions. And if you are implementing satellite modems, you can come work with other people that are implementing satellite modems. One of the major things that we did for Gino Radio last year is we created a foundation. So, Gino Radio's gotten to the point where there is significant investment happening in lots of different areas. Industry, government, academics, hobbyists, and we needed a way to focus this. So, the foundation's current responsibilities are raising money to support the project. We've gotten to the point now where we have significant project costs. That includes a very significant AWS infrastructure for our CI systems and web systems. Managing the finances, we actually have a not insignificant amount of IP now that the Free Software Foundation does not hold for us. So, the Gino Radio Foundation holds that. And putting on GRCon, which is at this point a very expensive event. Yes? I have this question. Why is the FSF not holding or why do you not want them to hold? The FSF holds the copyright to all of our code. And that is what they protect. But the Grenadier as a project now has lots of IP that is not just code. The logo is the most basic example. This is the most basic example. More complex examples would be all of the code that is in our web back end. For future responsibilities is funding project development. So, and I think we're going to touch on this in the panel shortly. So, maybe a little bit based on the panel description. So, I don't want to go into too much detail there. But we've reached the point where a lot of the really, really interesting work is happening in the out-of-tree modules, right? And we need a way to fund development of the project core to keep everything moving there. And the best way to do that is to pay people. So, funding development within the project is one of the goals. We just covered some of the awesome stuff that's happened recently. And I actually talked about this one at GRCon a little bit. All the other ones are mostly new. So, Virginia Tech recently ran an experiment where they put an embedded system in a sounding rocket. This is a picture of the rocket taking off from a NASA center. And then they had a software radio connected to the ground station. This is the antenna array. They actually had the radio running at the ground station and in the rocket. And so, the radio was their comms link for the rocket that they launched. And they maintained the link throughout the course of the event and gathered experimental data, which is really awesome. Unfortunately, NASA lost the rocket. It landed somewhere on the ocean and they never recovered it. But they did get useful data before it disappeared. This one is actually really cool. So, can I get a show of hands for people who know about OuterNet? Okay, so not, I'll give you a brief breakdown. So, OuterNet is, actually I can't remember if OuterNet's also the name of the company. OuterNet is a project to try to provide Internet to the world, right? It's like Facebook's internet.org and Google's Project Loon. The goal is to provide Internet to the world through satellites. And their eventual goal is to do this through CubeSats. So, CubeSats are one of the areas that the generative is getting used more and more heavily, which is really cool. Right now, OuterNet does not have its own satellites. So, it's using MRSAT, low-orth orbit satellites that are already there to broadcast, this is where they kind of got in trouble. Their original tagline, we will provide Internet to everybody. And what it actually turns out is they will provide access to very specific websites at certain times. So, they would provide access, for example, to Amazon.com or certain parts of Wikipedia. And one of their major sales pitches early on also was that everything was going to be open and you could build a receiver. They would release the specifications for a receiver and they encouraged everyone to build a receiver and then share how you built it. As it turns out, what's actually open is the hardware and the actual receiver software codecs are all closed source. But that's okay, because we have green radio. And so it is really awesome work that was done here. The complete walkthrough is on the Green Radio blog now, reverse engineering the satellite link. And it takes you from, okay, I see the satellite signal. You can do this with something as basic as these little SDR, right? To decoding it in Green Radio, this is a Green Radio graphical sync. And seeing the web pages that they're broadcasting, right? And so there's a complete walkthrough on the Green Radio blog. This is a neat example. So I had this example here. This is from Mike Osman. This is actually a couple years old, but I have it here because I want to use this as kind of a launching pad. This was first presented at the GRCon 14. He subsequently presented it at DEFCON. One of the areas that we've seen Green Radio really start to take off is in the cybersecurity community. So for a long time, the cybersecurity world really focused on link layer and above, right? They assumed that they were getting bits, right? It's like you give me a black box of bits and I'll figure out how to reverse engineer it or hack it or whatever. But the actual comms link itself was always just assumed to be there. And one of the really interesting things that we're starting to see with Green Radio is the cybersecurity communities adopting it and starting to do really, really cool stuff at the layer below. So this was one of the first examples that really, it's kind of edgy. It got a lot of attention. Mike Osman basically took something that was leaked by Snowden as part of the NSA, what he called the NSA play set and implemented it in Green Radio. And what it was, was the ability to sniff signals traveling over VGA cord with a software radio and then see it on your computer. And so what you see here is his radio setup. So this is a hack RF is his basic little cheap software radio. I don't mean cheap as in poor quality. I mean cheap as in these are relatively affordable just to be clear. This is his radio setup. And this is the output of his green radio flow graph, which as you can see is him booting Pentu on a machine looking at a VGA cable. So since then, so that was in GRCon 14, we now have three years of lots of interesting cybersecurity work and it's not just reverse engineering stuff. In some cases it's actually resurrecting stuff. So at Schmucon just last month, as an example, Brandon Creighton recreated the AMP cellular network. Can I get a show of hands for people who know what AMP's is? Okay, so AMP stands for the advanced mobile phone system. It was one of the first cellular networks deployed in the United States. And that's what you used if you had one of these. I should have had like a size reference here. And this is like the size of your head, right? It's huge. But he rebuilt the AMP spec in Green Radio. And if you have a software radio, you can now, as an out-of-tree module, you can go grab GRAMPS and run an AMP space station. Android is an area that's actually really exciting. You might have seen it on my slide for GSOC. If you're interested in Android work, please come talk to us because we'd really like to get some people working on this over the summer especially. So a lot of the early work is actually done by my predecessor Tom. These are images I shamelessly stole from your blog. So in this example, you can actually see this is a software radio that's connected by USB to an Android phone that's running. And increasingly what we're seeing are mobile handsets that would allow you to actually use hardware within the phone so you wouldn't actually have to cable a software radio to you. You could use your phone's software radios instead of having to cable one yourself to it. So an example of this is the new Motorola, I think it's the Moto Z where you can have like arbitrary hardware packs, you stick onto the back of it and you have what's effectively either USB or PCIe connection to your hardware. So I mean, you effectively could carry around a radio programmable software radio in your phone which is simultaneously very exciting and probably very terrifying for anyone in infosec. So this was one that was demonstrated at GRCon 16. It's worth watching the video if you haven't, anytime I actually, my slides are all, I've uploaded all my slides. I tried to put links everywhere in the slides in case you want to download them and look at them. So using your new radio, actually Alexander are you in the room? No? Yeah, I think he's standing outside. So Alexander and his team at Fairwaves hijacked drones, using GUNI radio. And so this is a picture at GRCon this last year of, and Alexander is like standing right here. He's holding a PlayStation controller and it's controlling this drone and the PlayStation controller's input is going to GUNI radio and then it's, and he's injecting it into the control stream of the drone's wireless comms. He subsequently flies it into the ceiling and it crashes. But it's really cool. Again, this is that last year's GRCon, you can get the slides, you can see the video, this is all open source. Another area I want to talk about is radio astronomy. So to me, radio astronomy is super awesome. I mean, there's radio waves all around us, right? So the ability to actually use radio waves to study the cosmos, I think is, it's fascinating. And so these are examples of huge radio astronomy dishes, right? So this is Arecibo, and these dishes are kilometers wide, right? It's absolutely huge. And these images are all from U-Haugh Vernon's presentation at GRCon. But what we're seeing is actually GUNI radio getting picked up and used very broadly in radio astronomy, which is really, really cool to me. So U-Haugh Vernon actually, who's at the Haystack Observatory, had a presentation at GRCon about some of his work. And U-Haugh's unbelievably brilliant is actually mind blowing, like using the RTL dongles that I mentioned, he like mapped the surface of the moon with radio waves and the resolution on those things is horrible. So he does incredible work. But more recently at, so the NRAO, which is National Radio Astronomy Observatory in the United States, started integrating GRN radio into not only their own work, but their curriculum. And there's now a separate project called the Open Source Radio Telescope Project. And their goal is to create a community for people who want to build your own radio astronomy telescopes at home cheaply. You don't need something like an Arecibo dish to receive radio waves from other galaxies and study them. You can actually do it relatively cheaply at home. So the Open Source Radio Telescope Project is a, they just kick this off. If you go to the website, you'll notice it's still somewhat under construction. The idea is to post designs, schematics, tips, GRN radio flow graphs to help you get into this kind of thing. Another thing I want to mention here is the C-SERRA program. So anybody who's actually been on the GRN radio discuss mailing list, you've probably seen the name Marcus Leach. He's one of the most active people in our community. He used to work at ESPRAC, which was a Canadian Radio Astronomy Group. This is his satellite, or not his, yet. But they're, so here's this, they're running this project C-SERRA and their goal is actually to make dishes like this broadly available to everybody. So that you could just upload your GRN radio flow graph. So you stick it on this dish and get access to those kinds of resources. So if you're interested in this, this is all happening in the open source world. And it's gonna be awesome. I see me in two minutes until questions, right? Yeah, so the last thing I really want to talk about here is a single metadata format, which is what Tom mentioned. So at a high level, it doesn't sound like an especially sexy problem to try and solve, but it's actually incredibly important. So this is something we've just kicked off and my pitch to you is if this is something that you think you might be interested in, please, please come get involved with us. This is much bigger than just GRN radio and I really like for everyone who might be interested to come give us feedback. So the goal is, you know, your core samples, you end up with these huge data sets, right? Gigabytes or terabytes in size. And it's very difficult to make those data sets useful to anyone else, right? There's an incredible amount of information that you need in order to make use of these kinds of things. And it's basic stuff like I recorded it at this latitude, longitude, assuming that's a static variable, the static field. This is the creator. This is the date. The things like this was the sample rate. This was my center frequency. This is how I filtered the signal. This is how I scaled the samples down to annotations. Like, hey, I was running an automatic modulation classifier and I detected BPSK starting at this sample at this frequency. To even basic stuff like, you know, the cat jumped on the antenna and messed up my signal at this point and I fixed it starting at this sample. All that kind of stuff is really, really important and it's important for lots of reasons, especially as the two major ones want to point out here are scientific rigor, right? The ability to reproduce other people's work. We're really big on publishing code and making data sets available. And making data sets available doesn't mean anything if nobody else can actually use your data set. And the other piece of it is right now there's no way to collaboratively work on radio signals, right? If you have a giant data set, let's say, of a signal you gather from another galaxy, there's no way to annotate that and share it with everybody else and say, hey, take a look and add your own annotations. That's what we're trying to do with SIGMF. So SIGMF stands for Signal Metadata Format. The spec is very much a work in progress, which you can currently find on the, I'm struggling with the resolution. All right, well, if you go to the Green Radio GitHub, you'll find the SIGMF. Yeah, that's not terribly readable. Anyways, you'll find the SIGMF repository. Have a look at the spec. It's not at the point yet where you should edit it and be like, this comma makes this sentence confusing. It's at the point where it would be great if you went and read it and said, hey, I have this particular use case. I'm developing this modem and I want to be able to take a data set from Green Radio to a static offline tool like InSpectrum. And I think you lack the ability to do this. It would be great if you supported this. So please have a look, give us some feedback, and let us know. So with that, I have five minutes left in the slot. So Martin, I'll give it back to you for questions. Other questions? What's up? In scientific community, where the research on particles, like elementary physics, they also run these problems. I think human data sets and multiple scientists want to investigate it. Yeah. Maybe you could use some of the open stuff in there. Have you looked at that? Is that relevant or not? Yeah, so that's a really interesting point. And it is the thing that's unique about what we're trying to do with SIGMF is it's, you know, you chose the word signal very specifically. So what we've tried to do is everything else that we've seen doesn't allow you to describe features of analog or digital signals uniquely. And that's the specific problem that we're trying to solve. So like the particles, this would not be useful for it and doesn't replicate anything that's been done at the particle level, for example. But wave equations, yes. Close the door until we've done the Q and A. It's actually, it's time to change it. We have one minute and 20 seconds. Okay. Unless there's no more questions. Sorry? Is there a document, is there a book for Gini Radio? There is not a formal textbook for Gini Radio, no. Yeah, no, so it, oh, you're saying, yeah, okay. So, yes, I agree. And there have been a number of discussions about trying to write a textbook. Obviously writing a textbook is a tremendous amount of work. So right now, what we're really trying to push in terms of documentation efforts is just getting more documentation at the project level so that users who want to come and get involved and start doing something have what they need to get started. I would be, if you're interested in talking about how we might get something like a textbook, like a what? Like an O'Reilly book? I agree. That would be awesome. I don't know how to make that happen. I think that's a good, like, the fact that that would be awesome is a good final way. So I'd like to thank you guys for your time. It's time to go in and out.