IPNSIG Academy: How to Benefit Today from Future Interplanetary Internet Infrastructure: Why and How to Join IPNSIG's Project Working Group
9:59PM Jan 16, 2025
Speakers:
Mike Snell
Samo Grasic
Keywords:
IPNSIG Academy
interplanetary internet
DTN network
terrestrial applications
remote villages
data mule
helicopter delivery
Arctic deployment
email service
instant messaging
network monitoring
satellite internet
Starlink
LoRa convergence
global network.
I think I think it'd be great. Yeah, we're.
Thank you. Good afternoon, and welcome to another IPNSIG Academy
webinar. I have just a few housekeeping items to go over with you. First, a reminder that we are being recorded. And secondly, we ask that you use the Q and A tab located at the bottom of your screen to ask questions. They will all be answered at the end of the presentation. So I'd like to introduce our presenter samo grass, who's an old friend of IPNSIG and who leads the IPNSIG pilots and project working group. Samo got his PhD from Lulu University, which is just south of the Arctic Circle in northern Sweden. And when he graduated, he started doing terrestrial DTN network work just north of the Arctic Circle with a group of reindeer herders. And that first project used the data mule capacity of DTN, the the reindeer data and other data would be gathered and stored locally, and then twice a day, a helicopter would fly to these remote villages and drop off and pick up the digital mail, so to speak. And then I think, if I'm remembering correctly, about five years later, he started work on a project to improve that communication capacity for these reindeer herders. And that was a group of multiple ground stations, I think they were typically located on snowy mountain tops that provided BP over Laura to provide communications to these remote villages. And then his next phase of his career was to start late Lab A, B, that's a startup company developing DTN solutions for research and other industries. So he's an interesting guy because he's doing interplanetary networking related work, and his history has been primarily terrestrial DTN networks, but I think he has a compelling vision for DTN development and adoption in terrestrial applications and partnerships with a variety of industries. And so that's what he's going to be talking about, and then at the end, he'll talk about how you can contribute towards that goal by participating in IPNSIG pw G's global terrestrial network. So I will pass the microphone over to samo.
Thank you, Mike, start my slides.
PowerPoint show. Okay, can you see my slides?
Okay, perfect. Thank you,
Laura and thank you, Mike, for very for kind invitation to participate in this and very kind introduction. So I would like to start with what, what DTN actually is. So it's like, it's, it's, it's a building block of the interplanetary internet. It was actually discuss in many, many presentations previously done under the IPNSIG Academy. And what I'm going to do today is I'm going to actually focus primarily on terrestrial use of DTN, application, application, terrestrial application of DTN despite the fact that it's actually deriving from interplanetary internet, if you're interested more, if you're interested more into the space communication and networking, and how DTN is applicable there, please. I mean, just check the previous IPNSIG Academy where Alberto Montilla made a great presentation about space communication. So very shortly about DTN, just to kind of a repeat, DTN is going to overlay networking layer. And the key advantages of the DTN is actually it's a store in for a protocol. So that means that that we that in order to communicate with between two nodes in the network, we don't require that we have end to end connectivity. Instead, we try to push the data as far as we can when we when we have opportunistic or permanent connectivity available, available, and for that means we actually can get data from one place to another. So
um, and let me begin by looking back 15 years when
I've done my first DTN deployments that was done within the EMF project. It was done in the remote Arctic villages in northern Sweden. This is the background photo that you can see is actually taken from the from the Small Cottage that we were staying for almost a month during the summers during the three year period, project period, and it's fairly inaccessible. And up until now, this still lacks any kind of a modern connectivity. It's it lacks any it or power infrastructure, so it's accessible only by one hour helicopter flight or four days walk. So it's a pretty isolated place,
especially for European condition.
The users of this platform are nomadic Sammy population that live in these remote villages during the summer periods, where they have their herds as well during the summer time. And the another kind of focus group in this project were actually as well tourists that are usually hiking on this lonely mountain trails and between these villages as this is some 15 years ago. It's like one of the most obvious service back then was to actually offer web cache pages, something that it's fairly I mean, that it's really not doable today due to so many connectivity needed in order to just load one simple web page. We were also offering subscription to the audio post podcast. We were tested some of the email implementation. And I think despite that, emails were kind of working, they were not used that much. And for the very practical reason that we haven't really been through, I mean, thinking about and is basically people didn't know exact email addresses from their colleagues that they wanted to send from there. So one of the most popular surveys during that time actually became we called it not so instant messaging. It was a time when, yeah, instant messaging was a big thing. We called it not so instant messaging. And this service was that was offered, was actually allow mess, I mean, messages to be passed from between the users within the village, or actually, as well, even outside, outside the villages by sending SMS. It was just kind of possible to send them SMS outside when that SMS actually reached the gateway, or actually send the
in emails as well. So
it was actually very, very popular service used during that time. So how everything works. So Michael mentioned that we use the helicopters as a data mill. So in this photo, I mean, in this diagram in the middle, you actually see that we had a we had a border node that was actually connected toward the mobile internet, to the internet, and that border also was actually located in this small shed next to the big fuel tank that was actually used to refuel the helicopters. Because this is kind of one of is kind of one of the less spots that you can still reach with a car and a truck, and then usually this was kind of starting point for for the helicopter pilots to fly in. What was interesting here is like so we that the bandwidth of the jet fuel line were actually was actually much higher than bandwidth of the that we have over the Wi Fi. So we actually were sometimes not actually able to get all the bundles there on time. So helicopters managed to, I mean, pilots managed to refi their machines way too quicker than we thought. We actually stashed the this DTN nodes that we built under the seats, which is not the best idea. I would not do it again, because we later on found out that actually most of these seats are crashed, so don't do that. So and then this helicopter flew twice per day into one of these remote villages. And during that time, they basically just drop people and pick up new people that sometimes they they brought this Wilson goods for the people living in the village. So they their stop was usually around 15 minutes, depending on the day. And this data was then actually collected and offloaded to the one of these base stations. Some of them were like, more like improvised that you can see here. It was basically a box with a big lead acid battery, solar panels and high gain Wi Fi antenna. So yeah, this is how everything kind of worked during that time. So why am I kind of talking today about this? I think there are some, there were some lesson that we learned during that time that are still valid today. And I would kind of want to go and it's as well, one of the reasons why PVG does things that we do, actually, so, so
first, maybe kind of describe, basically,
this was a typical day up in the mountains, so the project manager actually made a good decision, and actually send it us. Send it up. Send us up to this mountain isolate mount air for over a month. So DTN was only lifeline. So it was we actually were highly motivated to get the DTN and services running because it was kind of an only way to kind of keep in contact with our families and loved ones.
So and as you see here, like we're this is
one of the mornings where actually we heard a helicopter flying in. So the first thing that we do, we just pull out our laptops and see, you know if we will get any updates, some fresh bundles during the rest of the day. We more or less try to get the the everything charge, and we were moving the solar panels, and try to get the battery charge. Try to get all the equipment charge, as we fully rely on that. So I think one of the first thing is that despite a quite extensive lab testing that we've done before, moving equipment up the world, I mean, this real world test secret should expose quite some severe both softer bugs then, as well some protocol flows that I mean, we actually were not able to detect in numerous, numerous simulation that we done, for instance, I mean that my personal kind of experience was that that I worked with was actually profit routing protocol that I managed to kind of improve only after we've done this real world test. So I think it's really, really important to actually test things in real world scenarios. Because sometimes it could be just, you know, the driver crashing down, and it crashes in a particular way that kind of locks the other the implementation of DTN and things like. So it was actually many unforeseen kind of situation happen. It was important to optimize implementation for for each scenario. I think that's as well valid today. So it's like, I mean, this DTN, that kind of we built, it was, we tried to preserve as much power as possible. Kind of used a lot of sleep time as well, and things will relatively well until they didn't. So it's like, what happened once is that actually we just kind of stopped to receive bundles, fresh bundles, when the helicopters were landing in we're still kind of offloading. We didn't know what's actually happening on other side. And as we were completely isolated of the other part of the network, we didn't really know what's actually what's happening. So I just kind of one interesting thing is that you know when, when you how we perceive them, when information is old, like after three, four days of having no news, you know what's happening in the outside the world. We were quite excited to actually read two, two days old newspaper that one of the tourists brought in. So we're really excited. So I think it makes you, I mean, I mean, reconsider, you know what actually, what information is actually old
at the time, we also learned that there's actually quite
advantages to have decentralized DTN services. So for instance, during this time when things were down, we actually still keep on using the not so instant message service to community within the village, and as well, to send it between the villages, despite the fact that we had connectivity with the rest of the internet in at the end, in order to fix this problem that we had, we actually had to hire he copter flew to the back to this border node and actually examine, you know, what actually was the problem? Funny enough was actually that the main problem during that time was actually the one of the time, local times in one of the machines, or one of the nodes actually skewed to an extent that it actually started to drop bundles. Because, yeah, consider it to be kind of cold. So it's kind of a timing is kind of a quite a big, an important issue, I think, in dtns. So after this, we decided to do the net even, to do to implement even network monitoring over the bundle protocol. So basically, in principle, we're just basically sending out log files over the over the DTN, and that actually allow us to basically monitor and check as well if there were some anomalies on the network. So, and I think this last point, I think it's something that I'm really trying to push in our in our project working group, because it's like, I think it's so crucial, important to start to use DTN. And I think sometimes it can really, really be beneficial, because it's, it's it's often, when you do the especially in the lab environment, it's often to do the bit of shortcuts and actually use the IP connectivity and just try to check and things like so try to use it in as much as possible for any service. That's kind of one of the one of my motto. So, so, a change for DTN. So if we Yeah, it's 2025, and, you know, I mean, clearly the many things has changed during this time in this 15 years. And I think the big elephant in the room is, of course, the access to satellite internet constellations, for instance, like Starlink. And it's from the people that I work and I think people that I kind of meet as well from other regions. I mean, it's a really, really big game changer for these remote communities. It's amazing how people lives can be actually transformed just because of this. It's it has, it offers quite a competitive pricing. And it's true, like instant broadband internet connectivity is something that, yeah, we can actually enjoy in more urban areas. However, I would still say that there is still some issue with it. So it's like, I mean, and the first one is that it has a relatively high power consumption, clearly, because it needs to send the signal quite far away, and partly because CI, in many, many cases, you want to have the antenna in the line of sight. So you want to kind of melt the snow. So this is, this is challenging. If you go into truly off grid setups as especially in the polar regions, you will have really hard time to keep even starting kind of up and running throughout the entire winter when you don't get basically any power for the maybe four months.
Antenna sizes are one of these issues as well, like so I know that. I mean
that, for instance, when it comes to sensor networks, and antenna size is actually quite a huge challenge. So for instance, if you want to develop a GPS kind of color that will actually upload the data to the LEO satellites, these colors are usually quite big and bulky. They have quite a relatively high power consumption, so you get maybe one or two updates. So it is a bit of an issue. So and then, as well, Leo is actually getting really crowded. So truly, we need to address, probably environmental impact just from launching all of these satellites. And probably it's going to be interesting to see what happens in five years when these satellites get derailed and then start to kind of fall back, if, because, if I heard it right, it's gonna we can expect, like, up to 40 satellites is going to burn in atmosphere every week. So there's going to be interesting time. So it's a bit like unknown, I would say. And then we, of course, we have a geopolitical constraints where, you know where these services can be applied to where they cannot be, although from what I heard, I mean, it's like I heard people using Starlink, even in many, many places where sterling is not supported. So, so. And now I'd like to kind of try to go into maybe some characteristics of of DTN that are very, very kind of a vital for, maybe for terrestrial application. And now I'm playing a bit with words, so it's like, I mean, first thing is that I would like to kind of point out is that is that when we talk about delay and torrent, maybe we should, you know, consider d as well as for deep sleep. And one very practical example is basically the project that I just recently finished, and it's a normal track platform that, as well Mike mentioned. So it's a kind of project I've done together with the rainy crew. Together with the reindeers. And the idea is to basically track their reindeer, their reindeer, and as well, they are able to send the messages. So here on the right, you see, basically, this is the application that they have. You have a positions, and then if they click on sitecon, they can actually send geo tech, location, and this, all this kind of works in a completely offline or online situation. So basically, if you're connected to online, you will actually get, you know, the updates for a server. Otherwise you can use it in offline mode. So they have small devices that they can just have there in a pocket. So it's basically a Bluetooth interface. It's an interface between the Bluetooth on a phone and Laura network. And the key
feature of this, of this, no matter platform, is that
it's it's actually most of the time, it's actually in a very deep sleep mode. So it's actually you shut down the entire network, you wait for 15 minutes, you wake up, you basically share and exchange the data that you want to exchange, and then you put it back to sleep. So it's like, by doing this, we actually got two years of battery life for one of the base stations that we accident tried, although there could be the solar panel, we forgot to plug it in on a couple of base stations, and the base station still work for two years before I actually had to go up and check what's actually the problem. So it is widely deployed in northern Sweden in the moment. So it's going to have a couple of 100 nodes, if you count in colors, as well as the nodes so, but in general, I would say it's probably around 4050, nodes when it comes to base stations, and then the rest are basically colors on the reindeer. It's based on the 430 megahertz Lora radius, and we use really high gain antenna. So it's like, basically, we're not, we're kind of still, like, within the limit that we're actually sending. But we're using the really high gain antenna is kind of crucial to actually collect the data. So and we're getting links like this base station can talk with each other on like it happens quite often that, you know, we get links up to from 60 to 80 kilometers without bigger issues.
So, yeah, this is, as well, one of the, this is one of the,
this is how, basically, this base stations got deployed. We just kind of change them and we actually pull them, because it's a national park, you're not allowed to actually install anything permanently, but as long as it's, I mean, I mean, if it's actually used for any purposes, then they have, kind of Sammy population, have exclusive right to actually bring the equipment that they need there to do their work. The second D that I would like to kind of expose is actually doing it over everything. So here I would like to really, really stress the importance that DTN is an overlay network architecture. So that means it can be actually applied on any kind of a communication infrastructure. And then we can talk about, you know, internet that we today. We can use it on fiber, optical links, radio, even storage devices. I mean, it's kind of incredible how much data you can actually move. If you actually take, you know, one of the nest with a couple of 100 terabytes hard drives, and actually, you know, on a plane, and, you know, send it to another continent, like it's, it's, the bandwidth is just enormous. So it's, uh, it's kind of big potential you can you can go then all the way to a bit more obscure like, um, convergence layers that you can as well. You can use a simple paper, you can use the core codes or even smoke signal. That's something I'm still kind of want to do. Like, basically, I would like to modulate the smoke and actually have a like a magnifying kind of glass that you can actually focus just on one spot, and then actually try to get one way link using that, just for illustrative purpose. Of course, it's going to be slow and not practical, but it's something I want to try. But here are actually some examples that we try previously, and I would like to, like to use this example just to show how seamlessly is for users to use. DTN. So for instance, if you develop a service, let's say, in this case, I'm trying to kind of imagine, let's say we, you are a climate researcher, and you would like to collect data from different places. I mean, most often you need to kind of fiddle with different kind of a connectivities with different ways so how you collect the data. But for instance, if you adopt DTN to to or if you adopt your your software which you use to analyze and collect data to DTN, you can actually basically just leverage from all the underlying convergence layers that DTN is offering. So for instance, I mean, it's like, the first case that I want to show is, like, it's actually one of the base station. This is actually a real world example that was as well tested during the EMF project in Slovenia. There was actually a very remote metrological station that had no connectivity. And how this was actually implemented. There was actually one person going to this base station every day, or to this metrological station, plug in a USB key, send ship it over the normal mail. Then was actually collected and just plugged in a computer. We had some scripts that actually collect everything and send it over the DTN to the kind of this back end were actually process and stored this data. Second example that as well was used was actually underwater boils that actually I worked on. And there we use this Lora connectivity, Lora editing connectivity to collect the data. It actually worked under up to two meters under the water, surprisingly well. A third example is as well, one of this example that was used in rural scenario as well. That's, again, from sloven It's like it's one of the meteorological station located inside the rather long, I think it's a kilometers long cave, and it's actually quite a big tourist attraction. So they actually have a train that runs every hour, and actually passes by this base station. So what maze, kind of the company maze did in sloven is actually they mounted a small note on this train, and actually that train, when it goes in, it collects the data and brings it actually out from the cave. And when it comes out, it actually offloads it over the internet again, over the DTN and to their system that they use. So and, for instance, you know, if you push this really far in a bit into the future, for instance, you know if you want to kind of deploy some sort of sensor network in the on the moon, you will be able to actually use in the same way, DTN to collect the data. So using exactly the same software suit that you actually adopted DTN to, you can actually very seamlessly use and and collect the data. So I think this is one of the, quite a big advantage of of DTN that actually we can actually leverage from some already today.
So, and the last did, I would like to you know, kind of
to discuss, is actually for democratizing access to space. And this has to do a lot about the current commercialization of space industry. I mean, we see just at the very moment there's actually, yeah, I think it's just being like one of the launches that that I think probably many of you are watching at the very moment. So, and I think the World Mission communication model was basically that there was actually used previously by the space agency. So basically you had one mission and one communication model. So each mission actually had their own way to actually get the data back to Earth. It's kind of outdated, and it's getting very crowded. The resources, the radio resources, are getting quite scared. For instance, Deep Space Network is getting oversubscribed, so we clearly need some new models to actually use, and this is, again, like the idea of interplanetary internet. So, but what's, I think, crucial here, there will be new commercial actors that will entering this space industry, and probably most of them will not be interested or skilled or to actually, for instance, you know, like work on the space communication, they will just basically want to get their data to and from whatever they're deploying. We can probably expect many commercial links. They will be probably very diverse. There are some, most often, these commercial links might actually have kind of a, not open, but some proprietary technologies that we're not aware but what's good with DTN here is as well, dt can run on any of this kind of knowledges, and it can actually serve as a common interface in future. So one practical example that I would like to kind of make here is like, for instance, you know, like, you are a small startup company that you know you're actually finding out how to autonomously grow plants and harvest them, something that you know, it could be of an interest for for future space missions. If you adopt to this technology today, you can actually use it on on day to day basis, over the internet, you just get a bit more reliable connectivity. So that means, you know, we can actually have advantage maybe in future, when there will be it for this, for your product, that we actually get advantage to to have a technology that's already hard and used, so you don't need to start from scratch. And I think it's kind of similar for, for instance, for the mining industry. It could be similar for the for the transport industry, so and logistics. So, yeah, because it's like, there are so many actors into this, so it's good to have a kind of a common ground where you can actually work there.
And then the another thing that happened in this 15 years
is that DTN has really matured. I mean, it's like, if it's something that was done previously, you know, mainly within the research field. It's actually really, really coming to the industry. So DTM bundle protocol actually advanced now to version seven. So I know, if you think about the the internet protocol that we're using right now, it's probably IPv four, maybe six. But you know, here we have number seven, so and seven is more than six. So I would may want to make a case that, please, let's start to use bundle protocol as well for our everyday communication. So many additional components that were maybe missing 15 years ago are actually now standardized. You know, it's like, I mean, just now we've got, like, for instance, one of the latest graph is like, labor discovery and, I mean, all these different things that are being implemented then. So we have much, much bigger dtns, of things that we can actually use. We also have a growing number of open DTN implementation, which is really, really good. So it's like that because I'm during the time when I was kind of deploying this DTN for the first time. You basically have ion that was not open source at that time, not at least for for us. And then we have DTN to implementation English as well, IBR, DTN, that came a bit later. So, but it's nice to have entire suite and different kind of implementations, although I will have, I mean, kind of stress here, that many of these implementations are very tailored for certain scenarios. So sometimes it is nice to have, it is hard to actually find some universal VTn implementation that works in situation that you would want, in addition to what kind of convergence layer they're supporting. That's another challenge that I think we need to really address. So DTN is being adopted and plant in upcoming space mission. So I mean, it's really something that it's worth to invest so I think the time is really right to jump, to jump on this wagon. And I think there are many needed competences when it comes to DTN. And
then the last thing that actually changed, that is quite a big thing. I think we have
IPNSIG, something that as well we didn't have 15 years ago. So it's kind of a and within the ipns, we have a pilot project working group that I'm kind of leading. I would like to talk more about now. So it's like, what, what pilot project working group? I think it's like, most and foremost, it's a meeting place for detailed enthusiasts. So it's like, so we have a members from very diverse backgrounds. We have members from academia, just doing research. We have from various industries. Space changes, space changes in some certain individuals. And it's a really dynamic group. So it's like, although the group is kind of growing, you know, people that are active, I mean, they're kind of a they're really like, it goes like, I mean, people coming live as well. So it's very, very dynamic environment, and you have really big variety of people. So what, what we're trying to do in the group is to basically focus on testing and refining the protocol. That's kind of one of the main objective, and we do this by building a global detail network. And in practice, how this, this, this group works, is actually we have an online meeting please. Once per month, it's kind of a bigger meeting where we try to check in as well for some formal stuff, but then we have so called technical meetings that are much less formal, and we try to discuss different topics or issues that we want to discuss.
So a couple of words about our current operation
network. So as I said, like, we're trying to build a global network. So it's like, I mean, on this map under here, like on the bottom right, you can actually see the most of these kind of a nodes are located. I mean, even if it's just one pin, it could be several nodes. And on top of it is basically contact graph presentation. This is how we these nodes are actually connected logically So, and we we want to run the bundle protocol on diverse platform. So it's like, so we are using all sorts kind of devices, I mean, that people have. So it's like, we want to use different different operating system, different platform, different platform, different convergence layers. In order to kind of test, for instance, interoperability among the implementations, we want to do the long term testing, because that's something that quite clearly. We figure that many things happen after a couple of days, when you keep things running, if you have some memory leakage or something and something that's quite vital. Then for, for instance, for space,
deployments where you're willing to have a harden technologies, we've tried
to kind of set them performance under web loads. So this is something that we just kind of started. We tried to employ heterogenous network links. So that means we're kind of using different kind of convergence layers just to test them. It's like, primarily we use the IP base link, so it's like either UDP or TCP. So we like to as well test now when networks start to grow, to just see how resilient this network is. So I can maybe say that this is actually kind of second generation of network, because the first one that we built was actually quite, you know, it had a star formation. So that meant that, you know, if this central node was down, basically entire network was down. Now we actually try to get every node that is kind of a core of this network, at least, to be connected with two other nodes. So to assure that if one of these nodes goes down, we still have, you know, connectivity over the entire network. We kind of faced some scalability challenges, and for that, we actually had to implement some new tools. So for instance, you know, like when, when we started, we're just like, up until 1020, 30 nodes. It was quite easy to we all configured nodes and we manually added the configuration file. But that's something clearly not possible. When the network start to grow, we're still kind of looking for, you know, what are the best way for network participation? How? Jon, so, I mean, we're preparing some tutorials, some some even kind of workshops that, you know, it will actually help people to join these networks. And while doing this, we actually quite clearly addressed some we had to address some of the governance and policy issues we actually tried to push for to get, actually our own valid IPN numbers, which is kind of equivalent of the IP number on the internet. So, and what we're trying to kind of focus bit more now is to actually implement DTM based services. So we're not just, you know, running some tests, but we actually start
to use it. So some of the current
project that our group is working with. So as I said, as a Prem mentioned already, like we are expanding the network, we would like to get new nodes and new members into this network. So we're really open and welcoming anyone who is interested to participate, to just contact us, join our and just kind of join our network. So we're implementing, testing new convergence layers. I mean, like, what we've done recently was actually implement IPv six. That's something. And then now as well, like, we use Laura a lot. I mean, like this time person involved in some, experiments. Well, with Laura, you can run DTN over some UART interfaces, which kind of allows you to bring all sorts of kind of a from infrared, also kind of this legacy, old, Legacy communication devices that you can actually bring into and use it in the day. So we actually, last year, we actually ended up with two email over DTN implementations. They're both open source and available to check. We,
one of our colleagues, Scott Jon, who's kind of working with
the mail implementation, actually managed to push that will get IPN DNS records. So we that means that you can actually get email on, on your IPN based note as well. And when I mentioned previously scalability as well, as well testing the dtns, it's kind of a DTN context chain mechanism. So the idea here is that we try to kind of find, we try to find the smallest common denominator which in dtns, which is kind of contact, and then we want to basically exchange this contact with with your neighbors, and then this neighbor, then further exchanges contact same. By this we actually can grow a contact graph of entire network and and we can actually access nodes that, you know we haven't manually actually configure for for our node, it runs so exclusive OR bundle protocol, and currently we're actually checking first we need to implement this, because it's currently implementing the Bah, but we're actually checking to kind of a redo the implementation in C to and as well, we're adding some local policies and the concept of this exit nodes, because what we're trying, what we're kind of figure out, I mean, seeing as well, some certain people organization would want to keep their networks not exposed, you know, like how I mean, but they would still like to basically connect our networks, and we'll do this for this assign, no. So hopefully this will get, once it gets, you know, more maturity will get standardized and, of course, going to remain open source. Other of the project that kind of finished kind of last year, or at least test to start to test. Still, some work that needs to be done is actually implemented DTN on as Adrien library. So the code is already on a GitHub, but it's the code is there actually, I think you can only send out once I will try to update it so we can actually have both sending and receiving of bundles. And I think this really brings, like an opens, like a plateau, I mean, for plateau of new types of nodes. And I was actually surprised how this, I mean, implementation is relatively small, so it's not really heavy, it's rather simple. So basically, we get the bundle in, you basically decode it, and you try to then extract the data that you're interested so as you see, like, I mean, we have a bunch of this Lily go devices. Most of these devices are actually Lora. For instance, this rover as well can run. It's based on is Peter too. And for that one, I would like to, I need to add, as well the besides the LoRa convergence layers for Jon, I would like to add the IP, or I'm probably like UDP for it. So we can actually, if you connect with your ESP, to the to the to your Wi Fi, you will be able to connect them to the rest of the network. So it's a smaller, off the shelf rover that you can actually now control over. I mean, with using the DTN, which is kind of quite a nice project, and I've done it to actually as a part of the demonstration for one of the events. So some of the future projects that we are
actually looking into. So,
so we would like to first, like and foremost, like, focus on more detail services. So really detail native services. For instance, you know, the email application. We would like to, you know, check a bit like into DTN and video streaming, DTN chat. That's something that's probably going to be first one in this list as well. And then, of course, some network monitoring and management, maybe some simplified words, although they are kind of some standardized as well. But so there's a lot of work that needs to be done. The second thing is, as well DTN API, that's something that was done last year, but it's not finished, so hopefully it will get finished this year, so we can actually have a easier way to interface with DTN. We're looking into some plans to deploy DTN in Nepal, so we'll The idea is to kind of have a sort of communication back up with small devices that I showed you previously, and tracking IX, I'm kind of similar, as we've done with the some Iranian population. Similar thing. Hopefully we'll get some funds for this as well as we were planning to do in Tajikistan, where the idea is to kind of track the snow level parts. These two things are still work in progress, so we're trying to get funding for that. And then it's like one of these things, like, we, if we're allowed to dream, I mean, hopefully, you know, we can get some PVG and satellite up and running in some near future. Like, so, how to join project working group? So the first thing is, you know, visit our IPNSIG web page. I'm actually adding the core code here, if you want to kind of use it, where you have kind of structure how to join the PVG mailing list. So it's like, and the only thing it's free, it's open and free for anyone. The only thing that we can expect from you is like, when you join, we'd like you to introduce yourself, and you can do that by either attending one of the meetings and just, you know, say, like, Okay, well, this is, this is me, and this is what I would like to do with within this group. And if you're a bit of like a shy person, you can as well do this introduction on over the mailing list. So the second step would then be, if you want to join the operational network and actually join with your note is to is to maybe try to get first IP number, a valid IP number that you can actually use then to identify your note. And we have a we can actually work as well on this service. So if you actually have, if you go on to open ipn.org you can actually request up to 10 free or probably even more, just contact us. IPN address, detect and you can actually use for your notes again. Here's a quick code that you can use to access it. Step three would be, then to build your own dtno, so you can choose, like, any kind of hardware you like. I mean, most often we use, you know, we try to, kind of use some old laptop, or we actually, quite often now use the Raspberry Pi devices. Now, when it got more accessible, I mean, before we use the LA potato I'm running, I have a couple of nodes that actually run really well running implementation on the the raspberry why as well. So it's, it's actually quite, quite a good it actually offers quite a good performance despite of very scarce resources. So most of the time we actually use the Linux OS, although there's some implementation, it can run even on windows that it's most often preferred. Then you need to choose and install one of the detail implementation. We'll kind of continue to use I on E but there's, we're kind of slowly as well, moving to the agtn, because it's quite user friendly and offers offer some nice graphic interfaces as well. After that, you need to kind of decide, you know, how you going to connect to this network. So that means, you know, like, will that be over the internet? Will you actually use some other technologies? So it's like, that would kind of, what I would recommend here is to console the PVG group. For those who are not having the static IP addresses or or are actually behind the carrier grade net, you can actually as well use the zero tier. We have a VPN number to use if you want to join and partake in this. Then you need to configure and test the ion. So we can actually do that together. So and in order to then get access to entire network, but not just your directly connected neighbor, you can actually then install the DTN X service. We have the instruction as well available if you, if you scan this thing so, and after it's just the one thing is kind of left is start to build your own DTN services, and then start to use DTN as well. With this, I will conclude, thank you, and now we're going back to
Okay, thanks very much, Sam. Oh, so we do have a question
from Rodrigo Munoz. He says, Have you checked DTN over, Laura, when and regarding DTN over, Laura, do you have any reference to review? Um, the I am actually checking a bit. I mean, I
was personally kind of checking with DTN over, Laura, and I haven't used it, I haven't seen the implementation on that, but it's a, it's a really good question, because, but what we what, what I would like to do is this, actually, to have, like, a DTN over the so it's basically the same as we have IoT I would like to have DTN, yeah. So, so we'd like to use the DTN for the sensor networks, so you can run basically just plain DTN for your sensor network. But of course, if you want to kind of leverage for all the LoRa N infrastructure that's available out there, it would be a great experience. I haven't tried, but it's a really good idea, and I think it's feasible to do it the problem that we face, I mean, it's like because during this reindeer project, actually, we did look into, deeply into the LoRa event. The problem with the LoRa is actually that it allows, I mean, only, to collect the data. So it's more like one way communication, which is not really a problem, because, you know, depending on your application, but you can use DT anyone on one way, but it is one of this, I would say, quite a big advantage, that with DTN, you can actually pushing data into the network, not just get it out as in a typical sense network. So
okay, there's no more questions from the audience, but I have some questions for you, yep.
So when I listen to your 2024 stint presentation, somebody asked a question that I would have asked, which is, you know, when you go down the list of DTN implementations, whatever happened to IVR DTN and Dr Alberto Montez? Response, if I understood it correctly, was essentially that the the mobile network software changes so rapidly that it quickly becomes unworkable to update any DTN application running on Android. And so I'm, you know, did I understand him correctly? Yeah, despair of any kind of smartphone version of DTN, yeah, I'm not like, so there is actually,
so my answer would be, like, it's like, it's to keep the anything running on a mobile platform. It's actually quite a challenge. I mean, that's something that because this custom built application I built for the raining car. It's actually using the QPI framework, and it's actually cross platform framework that allows you to, basically, yeah, you write, you write application in in CC plus plus and Q ml, and then basically, you can deploy it to an embedded devices or any it supports the iOS and supports the Google platform. But I do have to say it's, it's really, really a challenge, just despite the fact you don't need to maybe update the code, but you know, you need to update the package because of there's so many security and so many changes happening all the time that it's actually rather costly, and it's always you need to basically rebuild the application from scratch over and over again. So the what, actually, what? One thing that we did last year, it's actually Bush, young Russia, from maze, from from this company dimension in Slovenia. They actually, they maintain this old implementation that we've done with a profit. And they added, actually bundle protocol v7 into it as a stack. So right now it's a bit patch, and I think the entire implementation would need to be redone. But one of my ideas is to maybe to work more towards the end of the years, to actually have a dkn based mobile application that we've probably written in Qt, just to kind of ease in the entire process of actually keeping on a mobile platform. But it is actually really, really hard to keep them. I know there were some attempts to as well, to bring in, into the Android core, you know, just into the but there was as well, so many issues. So it's actually quite unfortunate that we haven't managed to actually get kind of a native DTN support like, at least in Google, like raspberry pi is the obvious answer.
Yeah, for exactly so. But for instance, you know, like, because, I mean,
what I'm trying to do in my, my company for, you know, for the customers as well, that I would like to kind of get this actually, I want to have the entire platform built over the DTN. And I will not lie, there are still many, many kind of components missing. So it's great. For instance, you know, for for, like, for, maybe for this very tailor made specifications. But when it comes to general usability, I think we still have some work to do. It's the same with when you try to use the implementations, most of them are made for certain scenario, and as soon as you know, like, if you misconfigure, you know your your storage, if you misconfigure things, things start to break apart. And for newcomers, this could be quite it comes a bit as a cold shower. So I hope that thing that we can probably all try to get better, to have more user friendly implementation and as well bring them to the mobile platform. So okay,
just
for kind of marketing purposes, I know it's probably in flux, but what do you think would be a good estimate of how many DTN nodes are in pwgs global network?
You mean kind of a to to because, I mean, I guess the more, the better. I think it's the what. What's kind of interesting is, though, like, which of these nodes will be permanently on? Because, I mean, what, what I would want is probably at least, you know, like, like, maybe 20 to 40 nodes globally spread that are there and really assured that they're always on. Because I see that as soon as, you know, after a couple of, you know, we kind of fix the network we have, kind of, we it grows, and then it happens quite quickly that, you know, parts of the network start fall apart. And that's probably because, you know, it's like people tried, we run some experience, and then there needs to be, I would say there needs to be more services that we can actually run over this. So I think that's why, I think one of the focus for this year would be to bring more services, like, from chats to this, very like practical things like that we can actually then use and we'll motivate people to keep these networks running. So Okay,
and
how would you characterize your take on the biggest challenges confronting PWG? Is it scalability, network administration? That's actually a really good question. So it's
like, one,
one very practical thing. I think right now it's like, I think it would be good to get more people, like new people in, because I see that many people that join this group end up with, you know, some new projects, and especially people that are, you know, involved in either, you know, defense or commercial project. They we kind of lose them, or they're on a side. They can, we can still contact them and everything. But so I feel like the success of the group is sometimes the biggest challenge. So it's like, I would say, so that's why I think it's really good to have, like, new people, more students coming in so and then I would then I would probably argue then as well, this thing about making it more user friendly, just for, you know, newcomers that would like to focus more on services, applications, we don't want to break their legs, you know, by feeling and, you know, manually configuring notes that can be quite tough, even for us. Like that worked for it for quite some time. Yep, yep.
Early has recognized for years that it's just too damn complicated.
Okay. Well, I want to thank you very much, Sam. Oh, thanks for joining us. It was a very interesting presentation with a unique perspective, and I hope to come back again. Thank you. Okay, Lord, can you bring up the summary slides? Okay, so this is the announcement that Laura referred to in the chat. This is something that she and I have been working on for all months. And so March 24 the 25th 28th of this year, we are planning a virtual online event, and it will have a number of on demand videos, and some keynote speakers, some technical presentations. And I think one of the most exciting things is hands on Raspberry Pi workshops. So if you want to learn how to set up your PI five to run DTN. This will step by step guide you through it, and I'm going to personally be a guinea pig and configure my pipe five during these workshops. So we're really excited about this. I think you'll find it to be a pretty compelling event, lots of useful, interesting, valuable stuff and registering it's really simple. If you just go to ipnsig.org, there is a tab for the event space, I think it's called Space summit at the top of the home page, and that will take you to the registration site. Cost is free. We're using this as a learning experience so we learn how to set these things up and also to generate interest in follow on events that will be have more of a fundraising nature to them. So you can, do you have anything to add? Laura,
guess not. Okay. You can go to the next slide then.
Okay,
as usual, this event will be archived on the ipnsig.org website. The video will be available there, and I'm assuming Sam will let us use his presentation materials, and you can do a lot of things on the IPM SIG website, and one of them is donate. We're a global organization of over 1100 members. We have a relatively small volunteer army to administer that organization and further the mission, and we would really appreciate donations. So if you're not a member of IPNSIG, you can join and the ipnsig.org website. We are truly global. You can see from this chart that we're from every continent on set, Antarctica, that was a lie, and many, many countries. And as a backup, if you can't figure out how to use the website to join, you can drop a message to membership@ipnsig.org so that's it, folks. Thank you very, very much. Please take the survey that will really help us in planning future events and thanks for your attention. Thanks again, Sam. Oh, great job. Thank you.