Disaster.radio: a disaster-resilient communications network powered by the sun
disaster.radio
disaster.radio
I wonder whether this project (if it's successful) would indicate that their numbers have declined past an inflection point, where their presence and availability is no longer assumed.
While this looks interesting I'm curious if they've done the numbers for Lora throughput. It's very, very low(which is also what makes it able to punch through everything).
Honestly I'm surprised they're not going for a p2p directional wifi. You've got a lot more hardware to choose from(pretty much all WISP hardware) and it has much better throughput with the benefit of dedicated links. Anything that's omnidirectional is going to have to deal with collisions which is not an easy thing to do.
Hams have the operational doctrine and also the flexibility in terms of tools to use. If the message doesn’t get through on one frequency band, change bands. If your transmitter modulation isn’t working, switch modes all the way downward to continuous wave Morse Code if you have to. The sun has been on a multi-week streak without sunspots so High Frequency communication will take operator skill to get messages through.
In a long range configuration, LoRa has a theoretical max throughput of about 140bps (bits) and that's assuming only one device is transmitting and no packet loss.
LoRa is built for low power, long range, low throughput sensor networks. It's not really suitable for a chatty mesh network. To get an idea of the intended use for it, take a look at LoRaWAN limits:
* An average of 30 seconds uplink time on air, per day, per device.
* At most 10 downlink messages per day, including the ACKs for confirmed uplinks.
* A good goal is to keep the application payload under 12 bytes, and the interval between messages at least several minutes.
... or a carrier pidgeon with an SD card, it would seem :-)
All these numbers are from google summaries, without further source checking.
Couriers are more reliable, though still not perfect.
African or European?
Human speech has an average rate of 39bps.
https://www.sciencemag.org/news/2019/09/human-speech-may-hav...
This magical 39 bit codec doesn't exist and probably won't ever exist. Even if it did exist, it loses information like timing, identity of the speaker (how they sound), and their tone. By the time you encode all of that, you'll be right back around the range of dmr. And if you're willing to discard all that, then just use digital text in the first place.
If you're talking about compression, Q-codes[1] were invented long before most of us were born.
Any increase in effective data rate of speech afforded by Q codes can also be used to increase the data rate of data transmission.
When people are talking, they're exchanging far more information than that at a far higher information density rate. Which is why everyone keeps telling you that what you're talking about is not really speech.
The study you linked is focused on linguistics and the effective 'symbol rate' of various spoken languages, by taking syllables spoken per minute and dividing that by the total number of possible syllables in each language.
It says nothing about how much information is actually exchanged between individuals doing the speaking. It doesn't factor tone, accent, pronunciation, mood, pacing, etc all of which are critical components of spoken communication and add up to a lot more than 39bps.
So when you say 'speech is 39bps and this thing does more (it mostly doesn't), therefore this thing is better than speech'. People keep telling that no speech is not actually 39bps and what you're talking about is identical to just written text in this context.
https://github.com/mozilla/LPCNet https://github.com/drowe67/codec2
Conceivably one could reach 39 bps with near-realtime speech, tone, inflection and tempo recognition. Put the result into a zstd compressed SSML stream and perhaps even reach 16 bps. https://stackoverflow.com/questions/46108940/is-there-a-way-...
Does anyone know such a cryptographic system?
APRS is probably the closest thing to this in ham radio, but it typically requires multi-watt VHF radios.
In related news, see https://www.rrmediagroup.com/News/NewsDetails/NewsID/18899
We practice every week, so everybody who is an active ham knows the local frequency and they know that there will be updates every hour on the hour.
On top of that, the local WIN System repeater (Western Intertie Network, check their website) was up on backup power, meaning I could have used my walkie-talkie-type radio to get help from distant places like Ireland or Hawaii, if it would have been useful. To say nothing of the amateur radio satellites and HF operation.
One of our local nets during the emergency was a tech-talk net, with about 41 check-ins, all but one on emergency power, when the usual is about 7-12 people. The time was used to discuss lessons learned and one ham is an expert with generators and warned everybody: Change the oil every few days if you are running 24/7. A lot of people were surprised to hear that. And many of them only had one generator to burn through.
To give another example of the usefulness: Imagine, your power goes out, and a day passes without anything exciting happening, so you take a nap, forgetting that the reason the power is out is actually high fire risk.
In a few minutes you wake to a phone call. Your ham radio friend tells you he just heard on the radio a wildfire started a few miles north of your community. This is not on Facebook, it's not anything your neighbors know about.
So you have no idea what to do but fortunately they keep calling with updates from the radio. Eventually they call and give you the all clear and let you know your evacuation warning was lifted, so you don't have to take whatever you can and leave town with your family, spending money and using every last nerve you have just to figure out the next 24 hours.
This is just one specific example that happened here. Everybody else would have warning via text if they were lucky (had battery, tower connection, and proper system configuration), and final warning via a sheriff's vehicle siren if that didn't get through to them. In this case the person receiving calls never got a single text message.
This highlights a critical element of emergency notifications and response: you have to have a plan for how to deal with warnings and alerts. If you don't, then the alert is worthless.
It may even be worse than useless by throwing an otherwise orderly and predictable population into a panic.
How you respond to an emergency, in general, is to reduce risks and consequences. This may mean moving out of the area, but more generally, it means moving out of immediate harm's way. In the event of an earthquake, that can simply be to more out of range of falling debris, or into a solidly-constructed building. For a tsunami, gaining elevation on a secure base (terrain, a very solidly-constructed building). And, of course, responding to changing circumstances and conditions as appropriate.
For widespread disasters -- hurricane and wildfires -- exiting the region or moving to a location that's unlikely to be overwhelmed by the forces at play, is helpful.
For high wind, rain, cold, heat, etc., the challenges often play out over a longer period of time.
Disaster response as with security risks should revolve around threat models. What are the foreseeable threats, what are their specific mechanisms of action (e.g., fire, smoke, heat, wind, flood, ground movement), what's the effected range, and how can the hazards be reduced, mitigated, or countered?
Then there are the long-term survival needs: water, food, shelter, transport, injury and health treatment. Ultimately, rebuilding or relocation.
In one of the biggest disasters of all time, the Bianqao hydroelectric dam collapse following the intersection of a tropical storm and a cold-weather front, about 30,000 deaths were the immediate result of flooding. Another 150,000 or more resulted over the ensuing days and weeks from disease and starvation given both the disruption to ordinary life and the inability to move rescue, recovery, and relief personnel and supplies into the affected area.
None of which has an immediate relation to the ham vs. mesh-network radio debate, but calls into consideration that each are only a small part of an overall disaster response plan.
BTW my recent emergency experience is documented here:
http://www.friendlyskies.net/fmk/index.php?tpl=PSPS-Ham-Radi...
After reading your blog on the fires I think that may have been just the nudge I needed to get back into it.
VE7XML qsl KM6NHH(I messed up my cs in my earlier post!)
Ps just looked up your call sign, its interesting that in the US your licenses expire after 10 years, here in Canada its a lifetime license.
Yeah, but we don't have to re-test, and it can all be renewed online for free.
Personally, I am active with the Connecticut Amateur Radio Emergency Services(CT-ARES), which is a local organization part of the larger national ARES organization(http://www.arrl.org/ares). Both of these offer training and procedures for facilitating communication during emergencies. CT-ARES works with the state of CT, in particular the Red Cross and many local town governments.
My activities include participating in regular 'nets' (http://www.ctares.org/networks.php) which are station tests which serve to prove that my radios work in various failure scenarios. Other members are even more active and practice their communication skills by offering supplemental communication for various events like marathons, fairs, etc.
It comes down to this; you don't know something works unless you regularly use it.
de K2CHA
PS: One should also test their backups by actually restoring from them...
https://othernet.is/products/dreamcatcher-v3-05
It has the 2.4 GHz LoRa radio, the SX1281. It also has a wideband mixer and synthesizer, which allows transmission across 85 - 6000 MHz. You should be a ham in order to play with the tx-side. The SX1281 is not compatible with sub-GHz LoRa.
To use the transmit outside of ISM bands, you definitely need to be a ham (and stay within ham bands).
We also act as a relay for ham packet radio (APRS). Hams can send messages that eventually get picked up by internet-connected gateways. We take APRS messages tagged with OTHER and broadcast then over the two satellite beams.
Happy to answer questions as I am capable.
It's cool to see this up on the front page, thanks for all the interest!
I'm a bit stuck at trying to manually change frequency of the LoRa radios. It seems the standard does not allow it but the HopeRF chip I use have methods for selecting frequency and bandwidth around that frequency.
If so, that would allow for a multi-frequency protocol over LoRa that could scale!
but most hams around are stuck in 2M analog, some even find configuring a subtone too complicated, so it is difficult to work on something no one will bother to use.
It would be neat to see what sort of traffic spike they get after being linked from a story that made it to the front page of HN.
This is the radio chip used in their hardware: https://www.semtech.com/products/wireless-rf/lora-transceive...
Very similar to the CC1200 I used in a radio I designed in 2013: https://www.ti.com/product/CC1200
Low data rate networks are super interesting. The long range let’s you do things you wouldn’t do with WiFi. But what i would really like is a high bandwidth long range mesh. The issue you run in to is who owns the spectrum. In the US it’s private communications companies that buy large chunks of spectrum. I’d like to see more valuable spectrum allotted for public use. This would pave the way for free or very low cost municipal cell phone service. Unfortunately there is big money interested in making sure that doesn’t happen. And with “business friendly” people in the FCC, the idea of free communications doesn’t have much traction.
This radio project looks cool. Ideally we’d have a generic 915mhz radio in our cell phones. I guess since this device also has WiFi, you could connect a phone pretty easily. Let’s beg the Librem folks to add a sub-Gigabertz radio (915/868/433MHz) to their next phone. :-D
Two parts can be replaced I guess with better alternatives though:
1. Risc-v chips for esp8266. Why? Because they are built on open source standard, are cheaply available, and are generally very powerful.
2. 802.11ah for LoRa. Why? Well, lora is highly optimized towards power optimization at the expense of data transfer rate. Data rates offered by LoRa might not be suitable for anything much demanding other that IoT. 802.11ah on other hand is also optimized for power, but still can operate up to at multiple 100mbps of data rates, with range in many KM's and supporting thousands of connections per installment.
This might help: https://www.electronicdesign.com/industrial-automation/what-...
802.11ah has only 1km range which is not enough for this purpose. LoRa is rated at more than 10km range at 292bit/s up to 50kbit/s. Also the chip used in this project costs only around $6.
One of the goals is to set up a pan tilt camera to remotely watch for wildfires, and we are seeking Ubiquiti gear for use on the project. We'd like to link it from inland to the coast which will take some doing but is probably possible. There's a really impressive mesh network project in Washington state, one I recall in Texas, etc...we are talking hilltops and mountaintops connecting whole communities.
Ham radio has served us well here during recent emergencies. After the recent 4-5 day power outage & wildfires, one of the county-coverage repeaters that didn't have backup power was set up with backup power.. Other than that even my non-ham friends were saying that listening to the ham radio frequencies was more helpful than listening to first responders via radio. Also, broadcast FM radio was really nice and there were hourly updates, calls with county officials, etc. Facebook was ridiculous because it has no idea that you are suddenly needing updates from your sheriff first, and everything else later. Cell data speeds were about ISDN level or maybe a tad faster. I accidentally made duplicate purchases on Amazon because I couldn't tell if my first purchase went through due to slow load times.
On ham frequencies we had people on hilltops directly watching and monitoring the fires, calling out what they were seeing long before the fires hit the news.
Anyway if you want to get into this, definitely get your ham license. Give yourself 2 weeks of study and dive in. Working with hams you'll find it much easier to get access to know-how, gear, people, etc. Even if you work independently somehow, the free education is worth it.
https://www.cisco.com/c/en/us/about/csr/stories/tactical-ope...
I actually had an interesting experience relating to this. A small handful of remote disaster coordination groups have self-organized over Zello (a PTT walkie talkie app for phones that uses data.) Their primary purpose seems to be to organize broadcasts of relevant information to groups of people who would "listen in" during dangerous conditions. The groups volunteers consist primarily of people that have internet connections and can broadcast from a safe location during the storm. These are similar to this idea [0].
I participated in one of these groups during a major hurricane in the US a year ago. My experience was mixed. It was interesting and fun to monitor social media, news sites, and NWS for alerts, hazards, and announcements. On the other hand, I was struck with how disorganized the groups were.
It was like there was a "vigilante" group of disaster responders who were giving guidance and advice that was of questionable quality. I encountered more than one member of the group that actively used the channels to spread fear and uncertainty, (like claiming that the hurricane would cause a risk of "meltdown" at a nearby nuclear power plant.) After I tried to bring up the fact that this was potentially harmful, I was verbally abused by the group leader and banned from the channel.
While these "bad eggs" are not the only part of these groups, other more well-meaning members might still not be well trained to handle very real crises.
I'm not sure the best way to handle this sort of problem, but I felt like sharing my story might provide value to anyone else who might consider getting involved with this sort of thing.
[0]: https://fortune.com/2016/08/20/cajun-navy-tech-louisiana-flo...
For even longer range it would be nice to have a certain percentage of nodes that can also communicate with HAM radio amateur satellites.
1. What is your baseline calculation for Wh per day produced by the PV panel?
2. More precise specifications on the PV panel?
3. What is the calculated load in Wh per day? How much of this is dependent on data traffic level? (eg. Does it go into a sleep mode?)
4. Have you calculated this for sunlight levels in mid winter, off grid, at latitudes 48 and above? The Wh produced in a day, week or month by that size panel somewhere like North Vancouver, BC, will be really low from November to February due to short day lengths and winter weather.
> disaster.radio nodes are capable of functioning in most moderately sunny locations
I live in Northern Europe (54N) and wanted to use an old phone as a time lapse camera in a remote location without power or internet (hence the phone rather than something like a Raspberry Pi). The battery when fully charged would last ~3 days, so I just needed a solar panel big enough to keep it topped up.
I figured a 20W solar panel would be sufficient, as most phones use way less than that when charging. In my initial testing in direct sunlight this panel was barely able to keep the battery % the same, let alone charge it. I'm not sure whether I just needed or a bigger panel, or it was just the 20W solar panel I got from eBay not really being 20W. If anyone has any experience with this I'd still like to do the project :-)
The 20W rating is at what is called standard test conditions, basically equivalent to direct sunlight in the Arizona desert.
You'd probably want at least a nominal 180W panel and a basic 25 Euro PWM PV charge controller.
Is polarisation of the reflected light an issue?
We are working on a new board design with an integrated solar panel and two LoRa radios, but I am less involved in that part of the project, so I can't answer all your questions on the power system. Perhaps this link will help, though it has not been updated for the ESP32, https://github.com/sudomesh/disaster-radio/wiki/Power
Currently, it does not have a sleep mode, I recall there being some concerns with waking up from sleep mode back when I was dev'ing with the ESP8266, not sure if that has changed on ESP32, have not re-approached the idea it since switching to ESP32.
We did have some do some tests with an off-the-shelf 15W solar panel and a super-cap, it seemed to work well. Check out their report, https://sudoroom.org/pipermail/disasterradio/2019-December/0...
I'm actually getting a ham radio for christmas (don't tell my wife I know) and while that will fill almost every NEED for communication, it certainly leaves room for improvements. Text chat is very convenient, maps are very convenient, so I think this project is a great start.
Kickstarter the hardware for like $100 and I bet it would do really well. I'd buy one right now. Or let me donate!
Looks like there's an "app store". Curious how that works. I assume that would be a major bandwidth/tragedy of the commons issue, but perhaps there's some way to have a higher QoS channel for the critical apps.
there is also this: https://www.nycmesh.net/ , somewhat related and can be used as a back up information/communication source
Disaster.radio uses LoRa which is very low bandwidth and typically omnidirectional antennas. It only uses WiFi to connect users’ “terminal” devices to the node within a short range.
disaster.radio is not a replacement to a existing internet infrastructure like NYC Mesh. Setting up a high bandwidth mesh network is a lot of work. Each node needs a non-trivial amount of power (enough to make solar hard) and nodes generally need line of sight which requires good mounting locations and planning.
We, the creators of disaster.radio, actually also run a small wifi-based mesh network https://peoplesopen.net/ and the idea for disaster.radio came out of frustration with difficulty of mounting wifi nodes (finding interested people in good locations, then negotiating with landlords for permission to mount on rooftops and running ethernet cable for PoE into building, then finding another location within line of sight and repeating the process). We thought: What if we could make a "mesh throwie" where installation was as easy as throwing it on a roof (and maybe strapping it to something).
LoRa can run in any band you want, but it's almost universally deployed in ISM bands as well, 315/433 or 868/915 MHz. I don't think I've ever seen a LoRa chipset that would push enough power to exceed part-15 regs.
HOWEVER, to be strictly legal, a part-15 device actually needs to be tested and certified thereas. Otherwise the parts can only be sold as a kit and stuff, which is how pretty much all lora stuff is sold right now. And you'd be hard pressed to piss anyone off enough to care.
Yeah. Disaster Radio does not look like a "kit" at all. Not for "an average user" at least. I'd be cautious building and operating hardware that can potentially get me in trouble with authorities.
Still great general idea though, to build a solar-powered kit that's easy to setup and operate, a kit that makes sure there is a "plug-and-play" solution in case of a disaster (natural or not), kit that also delegates legal responsibility to the "creator"/manufacturer.
Look at the unmitigated chaos on CB for a sense of how much the FCC feels like enforcing anything on the unlicensed bands. Probably 95% of CBers are running above-legal power, many of them by _several orders of magnitude_, and behold, the field in which the FCC grows its fucks, it is barren.
Wifi is 2.4/5ghz.
"LoRa uses license-free sub-gigahertz radio frequency bands like 433 MHz, 868 MHz (Europe) and 915 MHz (Australia and North America)." -Wikipedia
A good direction to develop would be to talk to FEMA and its global counterparts. There is literature on what is needed in disaster. Instead of an insular tech-world development of such a platform having disaster response practitioner insights might be useful.
And agreed, once we have stable hardware and firmware, talking to local disaster response practitioners would be a great way do test deployments and get feedback on the software design and usability. I'm curious about the disaster-related literature you referenced, would love to read it.
$40 for a pair https://www.aliexpress.com/item/4000396836096.html
Using HF would be better in almost all ways except size though.
[1] https://arstechnica.com/gadgets/2019/10/lostik-usb-lora-radi...
I implemented a pair of these launchers designed for my 900 MHz "wifi" setup using a ubiquiti 900 MHz tranceiver running broadband hamnet firmware. It seemed to carry the signal to the antenna as well as a similar length of RG-6 coax. http://erewhon.superkuh.com/swtl_launcher_complete.jpg It was a little fragile and the connection to the SMA connector broke on one. Some day I need to resolder it and put one end up in a tree. Glenn's 2.5d metallized paper designs look much easier to build and more robust.
Both of these could be turned over to an automated system though I suppose.
If there are too many options maybe put a hardware matrix?
We're also about to finalize some new custom boards with ESP32 and two LoRa modules (meshing on a single channel kinda sucks) and will make an announcement on our site as soon as that's ready to try out.
In the case of an EMP it depends on how close you are and how much damage it does. If it is close enough to fry the electronic circuits of the disaster radio, then it's toast. Then again if you're that close everything that could connect to the disaster radio would also be toast. You'd be OK if you kept the setup in a highly shielded bag and only deployed it after the disaster (EMP) struck... however you'd still have the problem that your only device to connect would also have to be stored in the shielded bag.... and it's not much use of a mesh network if everyone else got fried too (because they didn't shield bag storage).
As for Solar storm. I presume you're talking about something like the Carrington Event ( https://en.wikipedia.org/wiki/Solar_storm_of_1859 ). So again, it depends on the scale. Based on the research I've seen, the equipment may keep working - but the radio propagation is a different story. GNSS systems will likely be toast, so that's not going to work at all. Radio propagation, depending on a lot of factors, will likely be affected. If you're in a heavily populated area with lots of notes - it may not make a different at all. If you're out in the fringe areas and only have a few nodes barely within reach - then your mesh might get cut off. However in theory with varying "jamming" by the storm, and retries built into the system, your messages could eventually get through.
If you want disrupt an electrical grid, there's a lot of easier -- and less likely to have an obvious calling card -- than setting off a nuke.
And if you're thinking about some other type of EMP weapon, forget it. They're not practical. While a theoretical weapon, they're not a realistic one. It's fantasy.
https://www.vice.com/en_us/article/kwxq4v/we-asked-a-militar...
https://nationalinterest.org/blog/buzz/world-wont-end-danger...