[Battlemesh] mesh protocol for lora

[Paul Gardner-Stephen]

Hello all,

Related somewhat to the above, I've just returned from a meeting with some
HF radio vendors, and implemented a simple Rhizome over HF proof-of-concept
(see http://servalpaul.blogspot.com.au/2016/08/bridging-rhizome-over-hf.html),
which we hope to advance with the cooperation of the vendors over coming
months and years. There seems to be sufficient interest from the
humanitarian sector for us to get traction on this. It helps that one NGO
emailed one of the vendors less than 12 hours after hearing of the
proof-of-concept asking to ensure that their HF radios would work with it.

In this context we don't want to reinvent anything unnecessarily.  In
particular, I expect that we will be talking again with the amateur radio
community to find out about what waveforms are good candidates for
improving on our current proof-of-concept.  This will likely require the
use of a vendor-neutral modem, or if we can't find one, then the creation
of one (probably a simple audio-band FPGA-based SDR using the microphone
and headphone jacks on the radios, if fidelity is sufficient), but again,
hopefully, using the existing work on HF waveforms.  Our understanding that
between 1200 - 9600 bps is possible under favourable conditions, which is
ample for Rhizome, and much better than the effective ~100bps we are
getting out of ALE 2G right now.

If anyone is interested in finding out more, or helping out with this
effort, that would be very welcome.

Paul.



On Sat, Sep 3, 2016 at 9:20 PM, Kristoff <kristoff at skypro.be> wrote:

> Hi Jonathan,
>
>
> Yes, that is a very good and interesting analysis of APRS.
>
> (The only thing I can add is that APRS on HF is not 1200 baud but 300
> baud. But as it is mainly used by hams on a boat "as a cheap way to notify
> the people at home where I am sailing today", it's being replaced by
> techniques that use some of the more specialised HF keyboard-to-keyboard
> modes.
>
>
> Anycase, about lora.
>
> As much as it is not my goal to recreate a lora IoT network, I do not want
> to "re-implement APRS" neither. APRS has a very solid installe-based and
> althou it is an old protocol and a lot can now be done better I do not
> think there are that much need to "replace something that does work".
> For me, this is more to learn about RF networking, mesh-protocols,
> applications and how to "mix it all together". APRS is for me a very
> interesting idea of how to do mesh-networking which is very different from
> -say- HSMM.
>
>
> As you explained yourself, one of the basic principles of APRS is it's
> limited hopcount. It's a "broadcast" network, but messages are limited to
> (usually) 3 hops. The idea is that APRS is mainly used for "located-related
> information", say the location of a voice-repeater. And "locality" is the
> key to this.
> It does not make sense to know there is a voice-repeater 100 km further
> down to the south, as there is no way you will be able to work that
> repeater.
> A hop-count of 3 on a usual "big stick" APRS network (high-power
> transmittors are high location) is roughly some 50 to 60 km, so that is OK.
>
> Now, the network I try to build is different. Instead of "a few big
> sticks", it's "a lot of low-power transmittors". It's a network of 10 to 20
> nodes spanning a small city. So the actual diameter (in hops) of a network
> will be much larger.
> So it may be necessairy to rethink some of the ideas of ARPS. (but that is
> what learning
>
>
> That is my though at this time. More to follow once we start implementing
> and testing things.
>
>
> This evening, I'll start with testing out the "RadioHead" protocol on top
> of these lora-modules. (radiohead is a library for embedded devices for RF
> modules).
> RH can run on top of "serial" radio-modules (like the lora-modules I have)
> ... and it also includes basic routing and a mesh mode!
> As far as I see, it's based on "on demand requesting routing-information
> when you want to reach somebody" which is probably best for a network with
> few users and low usage, like my current setup.
>
>
>
> Kristoff (ON1ARF)
>
>
> _______________________________________________
> Battlemesh mailing list
> Battlemesh at ml.ninux.org
> http://ml.ninux.org/mailman/listinfo/battlemesh
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ml.ninux.org/pipermail/battlemesh/attachments/20160904/90668063/attachment.htm>