Meshtastic is an encrypted communications platform for the Lora RF protocol

https://meshtastic.org/docs/about

by buescher • 4 years ago

249 88 4 years ago

meshtastic.org

olah_12• 4 years ago

Another similar project: https://reticulum.network/

RF_Savage2• 4 years ago

Meshtastic seems to be the most widespread one so far. And with the best hardware support.

Reticulum looks interesting, but what does it bring to the table compared to previous efforts?

ryukafalz3• 4 years ago

Having looked at things like this before (including Meshtastic) the thing that stuck out to me about Reticulum is that it’s carrier-agnostic. LoRa is cool, but being able to extend the network over arbitrary channels sounds very appealing.

olah_12• 4 years ago

answer from the Reticulum creator:

“My knowlegde about Meshtastic is a bit limited, so I might get some details wrong, but from what I understand, Meshtastic is more or less an application. It is focused on carrying out a single task well, in a way that is easy to understand and deploy, but that also means that it will not really do anything else than that task, and stop working well if you push it close to, or outside, its intended operational domain.

Reticulum itself is a general purpose communications stack. It will handle any sort of digital communication for you, from sending a single data packet to sending messages or streaming a video call, to serving a library of all human knowledge. But it will not do so itself. Someone will have to write the applications that utilise Reticulum to do those tasks. This also means that Reticulum, in itself, will not allow anyone to have a chat with someone else. You need an application using Reticulum for that, and a Reticulum network.

Right now, only a handful of programs and applications exist that use Reticulum. One of those systems is LXMF, a distributed messaging protocol that enables direct (and offline) messaging between users and machines. There is currently several LXMF clients available that people can use on mobile devices and computers. This is similar to Meshtastic, in that they both serve the same end-goal, of sending messages between users.

I don't know how open or interoperable the Meshtastic protocol is, but I do know that LXMF is very open and easy to integrate in other system, and messages passed with LXMF can be transported over any number of mediums, such as LoRa, WiFi, packet radio or over the Inernet. And you can mix and match the mediums as you please. This is possible since the messaging system is built on top of Reticulum (a general purpose networking stack), and it then gains all the capabilities of it (such as efficient multihop routing, global addressing, strong encryption, good privacy and so on).

I hope this clarifies it a bit. If not please ask away and I will try to answer as best I can!“

https://github.com/markqvist/Reticulum/discussions/77#discus...

nonrandomstring• 4 years ago

Over the years I've had students doing "people net" style meshing, BATMAN, opportunistic routing, stuff for warzones or emergency coms for disaster areas.

We learned a great place to test this is festivals.

Lots of endpoint mobility. New nodes coming in and out of the network. Terrible 4/4G cell coverage, so few alternatives. Dead batteries. Shadow zones. Fairly chilled out delivery time constraints. Everything you need to tweak your protocols.

I hope someone into playing with this will set up a larger scale experiment at Glastonbury, Burning Man or another big music festival.

livueta4• 4 years ago

Burning Man is one of the the primary usecases of one of the core devs, so yeah, it should definitely get some good exercise in that kind of context. Another acquaintance is a PAX admin and looking into it for similar reasons.

bigiain4• 4 years ago

I've not been following along too closely for the last couple of years (not a lot of need for or opportunity to use it in Covid lockdown times), but the original lead dev's most memorable (for me) main use case was for paragliding. Being able to see where your friends are (and who's in good lift) as a group of paragliders would be very useful. Having a reliable long-is range way of letting your friends know where you landed out would obviously be super useful as well.

c7h• 4 years ago

We have tested it successfully with 4 nodes last week at the Fusion Festival in Germany - one of those where you have to battle constant cellular network outages - and were surprised to randomly see fellow meshtastic users extending the mesh network. It was one of those "open source technology is amazing" moments :D.

aaaaaaaaata• 4 years ago

This comment will go down in history.

Super neat stuff.

wrycoder2• 4 years ago

Paging Cory Doctorow

buescher6• 4 years ago

I bet! I can imagine festivals being a good disaster simulation.

Meshtastic appears to use a simple flooding algorithm which is appropriate to what I understand to be the application - a small group of outdoorspeople keeping tabs on each other with short messages. One of my takeaways from a similar project I worked on in a past life was that flooding works well for use cases like that and just about anything more sophisticated is a pretty serious research project. BATMAN etc looks like fun to experiment with.

bigiain4• 4 years ago

what I understand to be the application - a small group of outdoorspeople keeping tabs on each other with short messages.

One really neat thing (in my opinion) about Meshtastic, it that it has support for GPS location in it's message scheme, so GPS equipped devices (or devices paired to phones with GPS on) can see where other mesh members are relative to them. This is really cool in the festival-type use case. Being able to easily see where your crew is whenever you want is great.

wgx• 4 years ago

Over my decades of visiting Glastonbury I can report that cellphone coverage has gone from “it works for some calls but no data” to “perfect 3G data everywhere on site”. The cell phone infrastructure companies ship in lots of portable masts.

erichocean2• 4 years ago

You can do the same thing with Bluetooth LE, which everyone already has.

Low-bandwidth, but doesn't require any "human in the loop" to establish the mesh.

tagami• 4 years ago

100m vs up to 8km...

noman-land2• 4 years ago

LoRa can get you tens of kilometers line of sight, and sometimes over a hundred. What kind of reasonable distances can one get with BLE?

eternityforest• 4 years ago

I wonder how hard it would ve to extend the Meshtastic code to send data via BLE phone to phone as well as LoRa.

It's usually done on an ESP32, so fixed repeaters can already handle it, and it would let you try it out without actually having the hardware.

retrac• 4 years ago

Bluetooth LE is aiming at something slightly different. When people say several kilometres with LoRa, they do mean in actual real-world applications.

Two LoRa transceivers with dinky antennas indoors will do > 1 km in a suburban environment. With a well-sited outdoor antenna for one of the transceivers that will increase to 5+ km. Two well-sited outdoor antennas can do 20+ km if they have line-of-sight.

There's nothing else quite like it at the moment. Cellular networks are close, but higher bandwidth, power, and of course, licensed spectrum. One could cover an entire large city with a LoRa network with a dozen well-placed nodes. Its most common application to date is along those lines, with utility meters and such.

dan0008923• 4 years ago

I have this same need and am preparing an evaluation of Meshtastic in the field this month.

I'm part of a volunteer EMS division within a paid fire department and we staff foot teams and medical carts at large events at our 90,888-capacity stadium (football, concerts, etc; well over 100k including tailgaters at our biggest game of the year) and music festivals with 10-40k attendees on the adjacent golf course.

While we have fancy Motorola APX 8000XE, our on-site dispatch wholly lacks visibility into unit locations and the abysmal cell service precludes software solutions leveraging mobile phones.

jonah2• 4 years ago

Doesn't Motorola have location tracking systems which integrate or add-on to their radio systems? I know we used to have this on some of our VHF radios.

You could also look at using APRS for location tracking...

Edit: Also - you should be able to set up your phones with priority network access on FirstNet, right?

whatthewonk2• 4 years ago

What about LoRa and Xray Satellite Comm? Couldnt you set up like noded repeaters as well for a ground back up / stronger signal along side as well? I have a bunch different arduino coding to do it, but finding a group of people to actually try to implement it wider scale deems to be harder than one would think. lmfao

na85• 4 years ago

I'd love to see a protocol designed for that sort of thing that hops from phone to phone without explicit intervention from the user.

No setting up a wifi hotspot and having people connect to you. No need to be connected to wifi at all. No fucking with Bluetooth.

I'm not even sure if there's a way to do it but that's what I really want. Carrying a separate LoRa device is just one more thing I'll forget to pack or charge.

foolfoolz• 4 years ago

they test this stuff regularly at festival and sporting events. i met a guy walking around outside lands with a suitcase full of phones. he was working for verizon to ensure the mobile base stations were appropriately powered

angst_ridden• 4 years ago

Around October 2020, there was a group of people putting up battery powered LoRa repeater nodes in the hills around Los Angeles. I had a few LilyGo TTGO units, one with DisasterRadio and one with Meshtastic installed.

I could get occasional packets through to one of the nodes that was about 6 miles away (line-of-sight).

My conclusion was that for a city, a much higher transceiver density would be needed if you wanted viable communications. It's not outside the realm of possibility. The units themselves are less than $20/each in bulk, and could be powered with a $5 solar panel/battery rig. Placement of the units would be key.

I saw some Hong Kong activists online post designs for "throwable" battery-powered units. The idea there was to toss out dozens of them during events where non-internet mesh communications would be needed. Seems like an interesting use case, although jamming and the end points (e.g., burner phones with WiFi->LoRa or Bluetooth->LoRa) are still the weak points in a scheme like that.

delabay• 4 years ago

Check out helium network. Crazy LoRaWAN density in LA.

ajsnigrutin• 4 years ago

I tried it some time ago, and the only way to communicate was to join a channel with someone you already knew (so basically it needed two(+) people to synchronize first in person or over some other channel, and then chat via meshtastic).

Is there something that supports a "public" chatroom? Something that would allow you to set up a node on a window in a large city, become a point in a mesh and be able to join a chatroom with all the other people (that you don't know yet) and chat there?

I don't personally know anyone else who'd use this over some "normal" chat platform, but live in a building high enough to be able to set a possibly usable meshpoint to connect with other enthousiasts and chat about random stuff there.

livueta4• 4 years ago

The firmware has a 'default' channel where the psk is well-known: https://meshtastic.org/docs/settings/channel

Selecting a default or any of the simple values from the following table will use publicly known encryption keys. They're shipped with Meshtastic source code and thus, anyone can listen to messages encrypted by them. They're great for testing and public channels.

In $west_coast_city I've gotten a fair amount of random pings on default settings.

Other projects (using other base technologies) like https://www.arednmesh.org/ are more focused on joining an already-existing network than making your own.

goodpoint2• 4 years ago

It's a pity that LoRa is all proprietary and patented.

An open alternative would provide more opportunities for experimenting.

EDIT: there are alternatives like HaLow and DASH7 on paper but nothing you can buy (for 4 euro like LoRa)

wmf• 4 years ago

Maybe HaLow is more open.

edrxty2• 4 years ago

M17 is possibly what you're looking for? Not specifically mesh but could be used similarly or extended. It's a a protocol that provides data and digital voice transmissions and works (with some modification) on a few existing commercial grade handheld radios running OpenRTX open source firmware. The only thing we need is a radio that has a Bluetooth module.

marcodiego• 4 years ago

I sometimes think about buying one of those LoRa arduino shields for some experiments. What exactly is proprietary in these?

goodpoint2• 4 years ago

The chipset, the modulation, the protocol itself, it's all proprietary and patented.

pabs3• 4 years ago

Range is lower for those though I thought.

ChrisMarshallNY3• 4 years ago

This project has been going on for a while.

I heard about it (here), a couple of years ago. I think the main reason that I didn't get involved, was because there's quite a bit of "some assembly required" with the project.

I don't really have a problem with that, but I wasn't really up to setting up a tech bench, all over, again.

I was trying to work with a proprietary system, and they were quite uncooperative. Once I have the wraps on the project I'm doing now, I may see what I can do.

livueta4• 4 years ago

You're correct that if you buy the most popular board for it (LilyGO T-Beam) from the source, you need to do some light soldering to get the screen on, and need to figure out a case on your own. That said, there's now an ecosystem of people on Etsy and other places who'll sell you a pre-soldered board in a nice printed case, e.g. everything under https://www.etsy.com/shop/QuantumShadow3D. Maybe worth considering if you just want to kick the tires without literally getting your hands dirty.

ChrisMarshallNY3• 4 years ago

Thanks!

My dance card is a bit full, now, but this is the kind of stuff that makes my heart warm. I've been messing with hardware forever (my current project is all software).

Is that the main guy behind Meshtastic? I'd probably just get a couple of the prebuilt radios, but not until it was time for me to start working on the project.

thcipriani• 4 years ago

I bought a couple of these from aliexpress fully assembled[0] (along with some massive terrifying batteries[1]). They're cool little devices. You can send messages via the app on your phone or by plugging them into your computer via USB. There's evidently also some wifi connectivity that I haven't experimented with at all.

Message shows up in the chat room on the phone and on the other devices linked to the room. Planning to play with them for camping Soon™.

[0]: < https://www.aliexpress.com/item/2255800992363816.html?spm=a2...> [1]: <https://www.18650batterystore.com/products/sanyo-ncr18650ga>

nope96• 4 years ago

What kind of bandwidth and latency would you get using this? Say, in the suburbs of a major city, two people 5 miles apart... how many bytes/second?

I'd love to see something like old school BBS systems take off again, an off the grid/off the internet network for hobbyists.

According to https://meshtastic.discourse.group/t/data-bandwidth-and-late... "But for the default very long range config it takes about 3 seconds to send 60ish bytes." Hmm, a bit slower than a 300 baud modem.

AdamH12113• 4 years ago

I'm not sure about Meshtastic, but in LoRaWAN, with the largest spreading factor (= maximum range) the maximum packet size is 11 bytes. And you get maybe one packet per second at most. Meshtastic is its own protocol and uses repeaters (and wider channels?), so I would expect them to do better, but not several orders of magnitude better.

I worked on LoRaWAN systems a couple years ago, and from what I found the biggest determinant of performance was what frequency band you're in. US915 has a 400ms dwell time limit for single transmissions, while EU868 has a 1% duty cycle limit. LoRa was designed for sending small amounts of data infrequently -- that's the "low-power" in LPWAN. Where I worked we were pushing it to the limits to get a couple hundred bytes per second at close range. LoRa does have some nice properties and I'm glad to see people using it outside of LoRaWAN, which is somewhat bulky for point-to-point communication.

rcoder2• 4 years ago

Meshtastic can carry a bit more data per packet -- 200-ish bytes, IIRC -- but the same duty cycle/dwell time constraints apply.

The routing model also makes it hard to add more than a few dozen nodes to a mesh. For small groups over wide distances that's absolutely fine, but it isn't a great option if you want to connect large numbers of _people_, unless said people are clustered around a few devices sharing WiFi or BLE connection time. (Meshtastic also doesn't really support this use case b/c of a "one device == one user/identity/mailbox" model, but that's an application-level choice, not something imposed by the underlying network.)

airbreather2• 4 years ago

If you are licensed there are 5W repeaters for 70cm under $100 - https://www.ebay.com.au/itm/194144864187.

Also, you can buy 1 watt Lora modules made by Ebyte, Sparkfun sell them.

wrycoder2• 4 years ago

In the US, the 70 cm ham band is 420 - 450 MHz. The transceiver you referenced is listed for 410 - 441 MHz. It could be used by a licensed ham to transmit data, if he/she determines how to modify or control it to avoid transmitting on the 410 - 420 MHz frequencies. It would also be necessary to send station identification[0] per Section 97.119(a) of the rules, which requires an amateur station to transmit its “assigned call sign on its transmitting channel at the end of each communication, and at least every 10 minutes during a communication.”

It would also be necessary to assure that any harmonics of 433 MHZ were within regulated limits.

EDIT: ETCI has specified 433 - 434 MHz as a LoRa band. In the US, the ARRL band plan[1] specifies those frequencies as auxiliary/repeater links, so they should be usable for LoRa, assuming no local frequency coordination disputes arise. The radio in question should be configured by the ham for that frequency.

EDIT2: In the US, and in general worldwide, hams can only communicate with other hams. So, 433 MHz LoRa would be limited to a mesh exclusively between hams, and with the necessary identification added. Unless you have specific reasons for using higher power, are a ham, and know what you're doing, it would be wise to use standard commercial LoRa radios, and stick to the 902 - 928 MHz radios in the US. (Which is shared with the 33 cm ham band - but the LoRa radios are low power.)

[0] http://www.arrl.org/news/fcc-proposes-18-000-fine-in-louisia...

[1] http://www.arrl.org/band-plan

jokowueu• 4 years ago

I've joined their discord a while back

But few time a year I go online and just try to find a simple Lora device with good range a qwerty keyboard and a simple OS

It still doesn't exist . Maybe in a few years

squarefoot• 4 years ago

The Pinephone has a LoRa capable add-on case, but the software still is in its infancy: https://pine64.com/product/pinephone-pinephone-pro-pindio-lo...

The Pocket Popcorn Computer might be closer to what you are looking for, if and when it will be ready for purchase. https://pocket.popcorncomputer.com/

noman-land2• 4 years ago

You still can't get this on one package but the closest I've come so far is using a LoRa capable feather board [0] and a keyboard feathering [1]. This gets you much of the way there.

[0] https://learn.adafruit.com/adafruit-feather/lora-radio-feath...

[1] https://www.solder.party/docs/keyboard-featherwing/

itintheory• 4 years ago

There's a LoRa model of the Popcorn Pocket P.C. which will have a physical keyboard and run Linux, however they're a couple years behind manufacturing schedule so it's a little doubtful if it'll ever actually be for sale.

foobarbecue• 4 years ago

What's a typical range between nodes for systems like this? Just looking for order of magnitude -- are we talking 1km? 10km?

edrxty2• 4 years ago

I work with similar systems a lot, given the modulation and power output, 1-10km would be a reasonable range band, 10km being line of sight under real world conditions, 1km being light urban propagation. It could theoretically be much worse in a dense urban environment with a high noise floor but I think you can safely put it around or slightly better than the performance of those blister pack FRS walkie talkies.

For comparison, my APRS mobilinkd modem attached to a 5 Watt handheld 144mhz radio will routinely do an order of magnitude better and the same modem on my 50w mobile car radio will approach two orders (double the base range plus additional antenna efficiencies).

metadaemon2• 4 years ago

Is there any concern with encrypted packets sent over radio and the FCC?

sitzkrieg• 4 years ago

this is ISM, so no

metadaemon2• 4 years ago

Yeah I guess if you're on 915MHz in the US you're good. 868, not so much.

buescher6• 4 years ago

Not if you're operating within part 15 limits. But if you're operating your Lora radio as an amateur radio operator, then yes.

jcbcn• 4 years ago

I've been thinking about this idea recently and finally I've stumbled across a real world implementation!

londons_explore• 4 years ago

Remember that if anyone manages to get any kind of mesh network working that regular users can use, they will get immediate and very hard pushback from mobile network providers.

People will stop spending $1000/year to have a cell connection if they can browse the web and message friends using your mesh network for free.

z3c0• 4 years ago

That's only if they make it past major ISPs, who have all decided that acting as a mesh bridge is against their terms of service.

derbOac• 4 years ago

Yet another reason for competition in the form of municipal ISPs.

buescher6• 4 years ago

My prediction is that no one will ever get the "join us now and share the bandwidth" model of mesh networking to work for any reasonable definition of "work", for both technical and social/political/human-being reasons. No industry conspiracy will be necessary.

chrononaut• 4 years ago

Doesn't this count? https://www.nycmesh.net/

supertrope• 4 years ago

Mesh networks will always perform poorly. Each wireless hop cuts throughout in half. People don’t want their device’s battery charge and timeslots used to relay other people’s traffic. To simulate the slowness, use Tor full time.

Saris• 4 years ago

The bandwidth issue could be helped by having multiple radios like LoRaWAN does right? I think they use up to 8 radios on 8 different channels to have simultaneous data transfer with multiple clients.

nullc• 4 years ago

Does meshtastic still have the problem that if there are too many nearby devices they will "capture" messages and use up all their hop count before they make it far?

ComputerCat• 4 years ago

Not going to lie, this is pretty neat!

sweis• 4 years ago

Is this using unauthenticated AES-CTR mode?

https://github.com/meshtastic/Meshtastic-device/blob/0447808...

https://github.com/meshtastic/Meshtastic-device/blob/285413c...

walterbell• 4 years ago

Are duty cycles voluntary with Meshtastic, as with LoRaWAN? https://www.thethingsnetwork.org/forum/t/lorawan-duty-cycle-...

iseanstevens• 4 years ago

This is a great project. Still work in progress but so much better than expensive commercial closed source products that are similar

brooksbp• 4 years ago

Why isn't the LoRa PHY spec open source?

ENadyr• 4 years ago

I was hacking on this back in 2019 but never finished it, glad this open source project came about and looking forward to trying it out on my next hike! Found a great case for the modules on here too https://www.thingiverse.com/thing:5200457 !

Meshtastic is an encrypted communications platform for the Lora RF protocol

Related Stories