John Waters Tue 31 Oct 2023 12:12PM
Around 20 years ago (more or less) there was a huge amount of activity on mesh networks, especially in extremely rural areas. As I recall, Locust Networks had a pioneering project across rural Yorkshire and they were a familiar sight a technology exhibitions. (Prompted by this fain memory, I ran a search for "Locust Networks" and found https://locustworld.com/ which, looking at the dates, is probably the same organization/company.)
Around that time, a few of us in rural Wales (mostly in Llanidloes) were working with Powys County Council to explore wireless solutions to extending some sort of connectivity to rural areas (especially the village schools). At this time there was no sign that ADSL would reach us any time soon and we needed a solution. Initially, we were exploring a tree topology using exceptionally high gain omnidirectional antennae developed by a colleague in Bath, so the mesh option appealed because of the redundant routing an more greatly improved load balancing. (I still have a quite a lot of the experimental kit we used, but high antenna gains would probably exceed legal limits these days.)
However, we were always acutely aware that, whichever solutions might turn out to be practical, bandwidth was always going to be a severe restriction - yet another commons in need of effective (and ethical) management. We didn't get as far a exploring that because the first sniffs of rapidly approaching ADSL drained momentum from the project, but the value of redundant networks has always remained in my mind
Although expectation have been raised enormously (through the bandwidth provided by ADSL/VDSL and 4G/5G) that doesn't take away the immense value in having an independent and highly-redundant network infrastructure to fall back on. (I like the following quotation I found on https://locustworld.com/: "I'd strongly encourage people today to ignore the IETF, and get focused on mobile, unlicensed wireless, highly reconfigurable and pervasive networking. Pursue overlays and co-existence, and create the next bigger "Internet" - the universal glue for networking things together. " -- David P. Reed")
Right now, however, I think something even more immediately urgent is an increase in the number of accessible LoRaWAN APs (especially in rural areas, for gathering data from wirelessly-networked environmental sensors) given the extremely patchy coverage at present (see https://lora-alliance.org/lorawan-coverage/ and https://www.thethingsnetwork.org/community for example).
Danyl Strype Mon 6 Nov 2023 5:21AM
@John Waters
Around 20 years ago (more or less) there was a huge amount of activity on mesh networks,
I remember a number of these types of projects being discussed by Indymedia folks around that time, and the Indymedia listserver hosting email lists for some of them. Interest in mesh networks seems to bubble up every decade or so, but it always seem to seep away again, with very little explanation of why it doesn't take off.
I'd strongly encourage people today to ignore the IETF
Well that's spectacularly bad advice 😆 Given that the IETF define core protocols like TCP/IP which are just as essential in mesh networks as they are in any other implementation of the net.
John Waters Mon 6 Nov 2023 10:12AM
@Danyl Strype The quotation I found on https://locustworld.com/ clearly does not refer to the overwhelming majority of IETF material, which would be impossible to discard given that without it we'd have nothing to work with. Given the context, I inferred that this refers to one specific piece of advice - one the source of which is unfortunately not specified. The important part from my perspective is "get focused on mobile, unlicensed wireless, highly reconfigurable and pervasive networking. Pursue overlays and co-existence, and create the next bigger 'Internet' - the universal glue for networking things together." All that means (from my perspective) is that above the immutable protocol layers there remains room for novelty.
The level of interest in such novel solutions depends upon the current drivers. Twenty years ago, the urgency was largely driven by economic considerations - an attempt to reduce the information poverty gap in the the shorter time. A more recent resurgence in interest may be a response to a general recognition of how vulnerable our communications infrastructure is to potential interference and a need for at least one independent, redundant layer to keep things running (even with severely reduced bandwidth).
Josef Davies-Coates Mon 6 Nov 2023 10:48AM
Yeah, I've always like the idea of mesh networks too. Guifi.net the best example.
See also https://mazizone.eu/
Sam Rossiter Wed 24 Jan 2024 6:32PM
I was involved in Bristol Wireless. These days I'm excited by https://meshtastic.org/ it's not how large your bandwidth is, it's how you use it.
Simon Grant · Tue 31 Oct 2023 10:51AM
I've always liked the idea of mesh networks. For one thing, to maintain communication if (or rather, when) the mainstream networks go down. Of course "they" won't want us to do this, as it will bite into incumbent extractive profits. All the more reason to do it, intelligently, cunningly, with the wisdom of serpents.