 Okay, the next talk is SDR Makerspace by Alessandra Chetain who has also done GQRX. Okay, thank you. Hello everyone and thanks for joining me for this talk. And thanks to the organizers for giving us the opportunity to present this project called SDR Makerspace, which is a collaboration between the European Space Agency and Libre Space Foundation. This is my first time at FOSSTEM and also actually the first contact with the GNU radio community. So those of you who don't know me yet, I work as a software engineer in the SETCOM industry, mostly with control systems and embedded systems. But I have been in contact with, I had my first contact with open source already back in 1995. So yeah, I'm actually quite old. And the way it happened back then was that I wanted to get into programming, learn to program computers. But those of you who are old enough to remember those days, you know we had like DOS 6.0 and Windows 3.1 and getting the development tools weren't quite so easy. So I got to university and I got exposed to a very early version of Linux, which was quite good because we had all development tools available for using building software. And I also had the opportunity to use the internet suddenly and get a lot of open source software already back then and see how other people were writing code and do what today is known, the copy and paste thing from Stack Overflow. It existed already back then. So that's how I got into software engineering. And clearly I realized the value of open source for me personally and that's how I started to write open source software in the beginning. And eventually I got involved with Libre Space Foundation, which is a nonprofit organization dedicated for promoting open source in the space industry and the space communities in general. You will probably know them from the SETNOX project, which is a global network of ground stations available. In principle for everyone who participates in the community and they have also built a UPSET together with the University of Patras, I believe. And UPSET was the first open source satellite to be built and launched. It was launched in 2017. It worked for a while, but as often it happens, CubeSets have a very short lifetime. And the kind of launch, just briefly about the SETNOX network, for those of you who don't know, it's a global network of satellite ground stations. We are not talking about big 20-meter parabolic arrays and things like that. No, just some small amateur level ground stations for now. They are distributed all over the world. And the idea is that users who are either interested in satellites or operating satellites can schedule observations over several ground stations and get received telemetry or enhance their coverage instead of just having one or two ground stations. And already today the network is, I think we can safely say it's global. This screenshot of this map was taken about one month ago, so I'm pretty sure the network has grown since then. The green dots represent active ground stations that are actively producing data. The orange ones are also receiving data, but they are in what we call the testing phase. So they are being tested before they are deployed into production. And I don't have the numbers, but I think it's an insane amount of data is being collected every single day. So it's working quite well. How much? 2,000 satellite observations per day. So, yeah, that means 2,000 satellite passes over a ground station. Yeah, so, yeah, it's working pretty well and it's still a work in progress. Of course, it's built using open source technologies and all the software that has been built for this infrastructure is available on the Internet. Now, last year I had the opportunity to be at the open source CubeSat workshop, which was the second CubeSat workshop organized by some very nice people from the European Space Agency and hosted by the European Space Astronomy Center in Madrid. It was undoubtedly one of the best conferences I've been to or workshops I've been to. And if you look at the group photo, you can see there are a lot of young people and the age distribution is maybe a little broader than you normally see at science conferences. Let's see what I used to see. It was a great experience to meet all these young people who were still at university being passionate about both space and open source because the space industry used to be one of the... or is an industry with a lot of friction and takes time to have things changed. So, but all these young people will eventually graduate and get out into industry and they will influence their future employers to be more open. So, there was a lot of GNU Radio talk. You can see on this first picture, Michel talking about the AMSET GNU Radio DVBS receiver project. And on the other picture, it was just a typical lunch break where a 16-year-old Julian, who has his own company building open source satellites, is presenting his first bucket cube satellite to Danny and Peter N4IP, who I think many of you know as a quite famous person in the SDR community. So, it was a great... the community is being built up around that event. We also saw the cube set ground station at the European Space Astronomy Center, which I think is a quite... or I'd like to think that it's a typical cube set ground station setup. They have two antennas for two band directional antennas on a two-axis rotor, some rack equipment for the actual radio and power supplies. And you cannot see, but there are actually two HEC-RF devices there, open source hardware, for those of you who know. And they are running GPredict and GQRX as the software for the ground station. So, yeah, there's a lot of potential for us to make a lot of good open source software and hardware and just technology knowledge in general and bring it into the space community. And I think those of you who work in space or just SDR in general know that they are quite difficult and complex topics. There's a lot of engineering disciplines that need to work together. And you need to do a lot of testing and design and integration. And so in that area, both space and SDR has a lot in common. I mean, just to design a software-defined radio system, you need to know RF and analog and digital electronics, math, DSP, software engineering, and probably a little more that I can think about. So there's a lot of... I mean, quite risky to use these technologies without having sufficient understanding. And that's exactly what we see quite often in CubeSat missions, that the communication subsystem is just left as a... Kind of like, well, it's got to work. I mean, we have cell phones and they work all the time, send signals. So I mean, why wouldn't a CubeSat radio just work? And unfortunately, many missions fail because people don't put enough effort in it or because it's just too difficult to them and come as a surprise. So one of the objectives with the SDR Makerspace project is to simply create a set of open-source hardware and software and technology tools or knowledge and make it available as kind of like off-the-shelf projects and components that people can use in the future. Both CubeSats and also the space industry in general if they are interested. So as I mentioned, it's a collaboration between the European Space Agency and Libra Space Foundation. Libra Space Foundation is sort of like the prime contractor in the project and goes out scouting for people who implement the various things or... We try to focus or we are running short duration projects, many small projects and if a project cannot do everything in one round, then we just maybe consider taking a second round instead of having one very long project which may never finish. I try to illustrate how the actual workflow is running with these small projects. So Libra Space Foundation starts by scouting and presenting some areas of interest where we want to explore. When people potentially implement this, they write a proposal and I write... When I say write a proposal, it's quite informal so we don't want a 200-page proposal and things like that. Just describe the project that you want to do and what we are interested in and after a few iterations that proposal becomes a statement of work to be executed. Then the work is executed in several phases. I illustrated three tasks or three phases but it can be any number of phases that fit into the timeframe allocated and there are milestones in between and weekly activity meetings, mostly using online collaboration tools. Clearly some people live at the same place and work at the same place and are really trying to just use the tools we have available. Once the project is concluded, we usually have some source code that is available on GitLab and a final report for the project. Currently, or I try to group the sub-activities we have at the moment or have planned already now for the future, group them in three groups. As you can see, we have some GNU radio contributions or GNU radio work, some testing and more R&D-like work. The GNU radio contributions currently have the GR SOPE which we had a presentation about earlier today. GRLEO which is a satellite communications channel simulator. GRCCSDS which is a transceiver implementation on the IQ storage, which I will talk a little bit more about later. So, yeah, these are the projects that either have concluded or being worked on or we know is going to be starting very soon, but since we have about, I don't know, six months left in the project, we will probably get more activities as time goes. So, yeah, now I will talk a little bit about some of the activities that have been concluded and we'll try to leave some good time for questions and discussions at the end. GR SOPE has been presented earlier this morning, so I'm not going to talk a lot about it, but it was intended to bring a new hardware protection layer into GNU radio using the SOPE SDR API. And the main difference between GR SOPE and the GR OSMO SDR, which we already have today, is that it has a plugin-based architecture, so it loads the driver backends as plugins, which will make it much easier to distribute to the end users at the end and we will be able to support new hardware much faster and much easier than we had been able to with GR OSMO SDR. And as the author of GQRX, I can tell you that has been quite a challenge, so I'm looking forward to what we can do with this eventually. But already today it is available and is out of three module. The project has concluded and it works. If you install it, you will get a signal and a source block in your GRC, GNU radio companion, and you just drag and drop the blocks and it should work out of the box. You need to know the device strings that you would use in SOPE SDR, but other than that, it's just really simple plug-and-play, like in GRC. The second project that has concluded is GRLEO, which is a satellite channel simulator for communication links. And what it does is to simply simulate the free space loss and all the losses that we get when we send signals through the atmosphere. And there are some ITU recommendations that describe or provide some mathematical models. I think we also use some other literature for implementing it, so it was basically just implementing it and making it available as a GNU radio block. Probably you can't see what's going on here, but it was just to illustrate how relatively simple it is to use. This is one of the examples that comes with the out-of-tree module, and first we have, this is just to generate a baseband signal, a frequent, an FSK baseband signal, and we have the actual channel simulator block there, and the rest here is just GNU radio visualization tools. These blocks here are the parameters, are the simulation parameters required to configure, to, you need to simulate the satellite orbit. It actually simulates the passage of the satellite in real time, so you can simulate your communication link as if the satellite was passing over a ground station. You can configure different type of antennas for both the satellite and the ground station, and you can actually also use different channel models. Currently, I think we only have the Leo channel model, but it has been made so that in the future it can be expanded with other models as well. If you run this simulation, you will get a nice real-time running application in GNU radio, and what you can see here is actually the Doppler shift of the signal as the satellite gets close to the time of closest approach. So one thing that will be really interesting with this GR Leo simulation is that it can also be used for testing with hardware in the loop. On this picture you can see here is an electrical mock-up of a satellite. It's a very simple one. There is a power system with a solar cell simulator. There's some command and data handling unit, some GPS, and the actual communication unit. Now you could connect this communication unit to a good SDR receiver you have, probably through some attenuation as necessary, and then run the baseband through the GR Leo block, of course with proper mission parameters, and then run your software-defined radio receiver that you intend to use for the mission and get the mission data out of it. So I think this will be very useful, both for people who build satellites and also for us as one of the next projects where we actually want to create a testbed for testing SDR devices and SDR software with real hardware in the loop. So finally, the last, one of the last Guru radio projects was the GR CCSDS transceiver blocks. CCSDS is not even an attempt to guess what the acronym is for, but it's a body who defines a lot of standards for space communication, for example, over the whole, all layers of a space mission, including, for example, they had standards for how planetary rover has to send data to an orbiting relay station and how to compress images. And there are these three standards that define how to send telemetry, telecommand, and how to modulate the signals. And the GR CCSDS project, is to implement at least a good representative subset of these standards so that we can use them in the GRU radio applications. And so the way it works is we have mission data coming in, the CCSDS transceiver, an SDR device, and an RF front end. So it will be a complete transceiver for space applications. And to put it in context with the previous slide and the hardware in the loop simulation, the GR CCSDS transceiver would actually be this part here called GNU radio receiver, or transceiver. So it will be pluggable straight into this test setup. Now the final GNU radio work we are actually just started is IQ database, IQ storage and compression because as you can imagine, there was thousands of observations every day in the SATNOX network. It's actually a big challenge. We need to store data. The way it works is that the observer schedules an observation through the network. The network sends a request to the client which is the actual ground station. The ground station will then run the GNU radio applications and when the pass is over, they will send the data back to the network and make it available for the observer or the users. Currently we are saving waterfall images, audio recordings, that's for historical reasons, and demodulated data which is sent back. We would really like to store IQ data, at least low-rate IFIQ data for later reference and also with respect to when new satellites are launched, then our network has been proven quite valuable in the orbit determination of these satellites and for that it would be nice to have some IQ recordings to work with but that will require some efficient storage. This project will look both at compressing the IQ data and storing it efficiently and SIGMF is of course on the table for storing the metadata for it. Finally, we have a lot of test and evaluation activities going on and we evaluate both hardware and software. On the hardware side, we want to test the hardware under realizations which means we don't want to confirm what the datasheet is saying. I mean we trust that if they say a device has a 3 dB noise figure, we trust that's correct. But we also know that in practice you will never use a device in an environment where you actually get that 3 dB noise figure. You will always have to back down the gain to get maximum dynamic range and things like that. You always have some RF noise in your environment. There's no point in measuring it in an EMC room if you don't have it in reality. We also will perform radiation testing of selected devices with clearly the objective is to see can any of them meaningfully be used in maybe satellites in the future and look at FPGA tool chains, alternatives for the tool chains that the FPGA vendors give us and also the criterias like complexity of the devices of the drivers of the software of the tool chains and how open source friendly they are in general. That's basically all I had to say. Clearly thanks to the European Space Agency for believing in open source as a future and the Libre Space Foundation for the amazing work that has been done for years now. And of course the SETNOX community who unknowingly are the guinea pigs for testing all the stuff we are doing. So we're happy to answer any questions. Yes. This gives a kind of definition of what needs to be done but what do you do when the people do not publish the coding they keep it closed so you get the data but you cannot read the data. Yes so the question was what do we do when we receive satellites where we don't have to publish the data format. That's a big problem we are fighting. It's actually already a problem in the SETNOX network and currently we are only risk... So to answer the question the CCSDS for that project it's not a problem right because the CCSDS kind of tells you how you encapsulate the data. Correct me if I'm wrong Manolis? Yeah so the data comes out and then it's left to the users to interpret it. Can you use the ESA to pressure people to be more open? Yeah well so can we use some... Yes well it's clearly also the CCSDS network grows and people realize its values. We can put pressure on them and well I wouldn't call it put pressure on them but try to educate them and enter the dialogue before they launch the satellite because after they launch the satellite and realize that they cannot communicate with it of course they come to the amateur community and cry for it please help us but then it's almost too late. So yeah. Yeah so the question is whether the UP set is fully open source correct? Yeah it is fully open source and everything is in GitLab in the GitLab repository. So the software designs and the software. Yes that was... So the question was how to contribute to the SDR Makerspace project? Yeah so the SDR Makerspace project is sort of geographically limited because of the way this works but the set knocks itself can be joined by anyone and so at least through that anyway talk to Pierre or somebody who knows more about it how it works but I think currently it's... Yeah it's him. Yeah. Yes? Yeah it is... So the question was whether we could integrate GRCCSDS into GQRX so personally I always try to keep GQRX as simple as possible but having said that I've always also been open to make some sort of plug-in framework plug-in like framework that makes sense because clearly you want tools to be expandable so we'll see in the future. I think that was... Yes it's something we would like to... The question was about FPGA in the cloud it's something we would like to look into personally I don't know much about that or cloud in general I'm a low-level guy or low-level programmer but I mean already today you can... I think it's Amazon Amazon offers everything but they also offer that you can run some FPGA IP cores in their cloud service so yeah we would like to explore it There's one more question there Yeah but I think we need to start changing people over Thanks Feel free to keep talking