Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We're making an open-source $30 GPS/mesh radio, would like advice

meshtastic.org

Post reply on HN
Print view View original

81–90 of 176 posts

Re: We're making an open-source $30 GPS/mesh radio, would like advice

#81

Post by punkgeek »

th0ma5 wrote: post #80
punkgeek wrote: post #78

Earlier quoted context omitted.

Thanks for the feedback. I think this project is a little bit different, we are focusing for now on making a nice end-user friendly consumer product using hardware that is already shipping from various aliexpress vendors. our tag line is "An opensource hiking, pilot, skiing, Signal-App-extending GPS mesh communicator for $30" We are currently using RadioHead mesh library, but we might shift to a common implementation…

Keep the hacking ability. Your large target demographic may not use or need it, but tinkerers will hype your product for free, at least in theory. See the RTL_SDR as an example, although it is a low cost DVB tuner.

oh yes - definitely want people to hack and add features (or use it as auto configuring pipes for other projects etc...). I'm just trying to say "at least for 1.0" we are trying to stay focused on this one primary thing. ;-)

Re: We're making an open-source $30 GPS/mesh radio, would like advice

#82

Post by masukomi »

The idea of these radio mesh network things is great. The reality is terrible because radio waves need line of sight unless you've got a repeater. This is completely impractical when hiking, or in a city filled with buildings. They end up with crazy short range between devices and you need an implausibly long string of them to get from a person off the grid to a mesh point on the grid. The only exception to this being someone high on a hill with no bumpy bits of other hills between them and whoever has the other mesh point that happens to be on-grid.

Gotenna couldn't make this work with all their $$ because of the simple physics of radio waves. Yes they sell devices that technically do what they claim, they just fall flat on their faces when they encounter hills, or buildings... or anything that isn't flat.

I think if you want to make something that actually works, you should drop the "consumer" and make having a ham radio license a requirement. Then use the GPS to find the closest public repeater. Their locations are pretty well mapped at this point.

Use the most appropriate of the many ham radio digital packet systems for whatever wavelength you're broadcasting on and choose the appropriate part of the spectrum that is designated for digital packet use.

Also be a good radio citizen and try other frequencies if the current one is in use.

I understand that the ham license is pretty much a deal breaker for adoption BUT it's what's required to use the decent frequencies and access to repeaters. As this seems to be an open source kind of project and not a "hey lets sell bajillions" kinda thing, that seems like a plausible option.

Re: We're making an open-source $30 GPS/mesh radio, would like advice

#83

Post by bigiain »

punkgeek wrote: post #47
bigiain wrote: post #40

Earlier quoted context omitted.

Also which of the ISM bands are valid in which country. (We sold off the bottom half the 915 band here to telcos, so only the upper half of it is available in .au)

ooh. good to know. here's what we are doing now for the regions we know. Someone is sending in a PR soon for the extra 440ish band in EU soon. https://github.com/meshtastic/Meshtastic-esp32/blob/master/s...

If you'd like some reference:

https://www.acma.gov.au/step-2-show-your-product-complies

Says:

(2) The 900 MHz ISM band within Australia is 915–928 MHz. 900 MHz ISM devices working outside of this range cannot be operated in Australia (that is, ISM devices that use the frequencies 902–915 MHz are not authorised for use within Australia).

(3) Using the frequency range 902–915 MHz would interfere with Australian mobile telephone networks. Severe criminal penalties, including fines of up to $255,000, exist under the Radiocommunications Act 1992 for interfering with radiocommunications services.

Re: We're making an open-source $30 GPS/mesh radio, would like advice

#85

Post by rapjr9 »

keithwinstein wrote: post #65
rapjr9 wrote: post #50

Note that there are limits to how large a mesh network can scale, it's likely to start running into problems with several hundred users. You can scale larger by limiting data to short text messages, but you'll probably never be able to handle more than 10,000 users, unless you can find a way to create high bandwidth backbone links between clusters of users. There has been a lot of research on this in the wireless sen…

That's true in practice with current radios, but it was a beautiful discovery in 2006 that it's not true in theory! In a world where radios are perfect and nodes cooperate to do decentralized beamforming, the capacity-per-node of a random mesh network scales as the number of nodes increases. See Ozgur, Leveque, and Tse, "Hierarchical Cooperation Achieves Optimal Capacity Scaling in Ad Hoc Networks," IEEE Trans. Info.…

Interesting. I wonder how well it would work in practice though. They assume uniform randomly distributed nodes, whereas people don't tend to distribute themselves randomly. They don't address practical concerns either, like how narrow the MIMO beamwidths need to be and how easy that is to achieve. Still a nice theoretical result.

Re: We're making an open-source $30 GPS/mesh radio, would like advice

#87

Post by donquichotte »

gardners wrote: post #70

Howdy, as the founder of the Serval Mesh adhoc networking protocols and hardware, what I would look for in a new radio is: 1. True packet radio mode, that doesn't hide the radioness of the link etc, but lets the programmer use it's strengths. 2. UART link, ideally using 3DR/RFD900 compatible connector (then we can include it directly in our existing systems, and so can others). 3. Field-flashable, but with robust pro…

Good advice. I also recommend to make sure the toolchain is usable and documentation is readily available.

I worked on a project that had an Silicon Labs EFR32, and while the hardware may have been great, their proprietary, closed-source "Radio Abstraction Interface Layer" was hard to use and badly documented.

The Eclipse based IDE that you HAD to use to configure the radio front-end (the EFR32 is a system-on-chip with an RF front-end and an ARM Cortex M4F core) is a GUI, and porting or comparing radio configs between different chip generations, and even different versions of the IDE, is a nightmare.

The Java IDE crashes regularly, debugging does not really work etc., so it really is worth taking some time to evaluate the whole package around an RF IC.

Re: We're making an open-source $30 GPS/mesh radio, would like advice

#89

Post by ChrisMarshallNY »

I’m interested, but we’ll see how it goes. This is a non-trivial endeavor.

I’ve been trying to write an open-source Apple driver for goTenna units, and have been unable to get them to help me out by providing their BLE spec.

That’s a shame, because I write really nice SDKs. Even if my driver was open and free, it would provide significant brand reinforcement. Win-win.

I believe that this tech is important, as it can provide service in disasters. I’m also big on open-source.

But I cut my teeth on RF stuff, and know that it’s a real black art. Software can get complicated, but is child’s play, compared to RF tech.

You have your work cut out for you.

Re: We're making an open-source $30 GPS/mesh radio, would like advice

#90

Post by alexisread »

punkgeek wrote: post #41
alexisread wrote: post #18

Not an expert by any means, but I'd suggest looking at the Dash7 protocol/ firmware as a reference for the mesh, security and positioning. It's been around a while and seems robust. Open source, royalty free, low power, high range (several km even in urban areas), medium bandwidth, runs on several lora boards. https://dash7-alliance.org/

ooh. That is great advice! We used the RadioHead mesh solution but it is super non optimal for us and we plan to switch. I'll definitely do some reading on DASH7.

Firmware available at: https://github.com/jpnorair/OpenTag
Post reply on HN