TL:DR, $50 Nodes with 1km+ range. Look at the map.
So today I was playing around with the idea of mesh network backbones. All the decentralization guys talk about a distributed and democratic internet. But how reasonable is it?
Not very. At least not with current hardware.
I used the specs of the flutter wireless board when figuring all this out. It’s great for mesh networking because
- It has good encryption
- It’s cheap ($20 per board).
- It’s a microcontroller, which means adding in solar charge controller and battery monitoring is easy
- It has a whopping 1km+ range, at 900Mhz (just look at the comparison picture).
- It has 1.2Mbps worth of bandwidth. Which isn’t nearly as good as wifi, but isn’t bad for that kind of range.
- The power requirements are low enough that it’s feasible to run them off solar.
Overall, it’s probably the best board I’ve ever seen for doing meshes that cover a large area. The bandwidths not great, but it’s a lot better then the very long range hobby stuff. It’s relatively cheap. It gets good range.
So I mapped it out. This shows a relatively simple connection between halifax and a nearby town.
Presuming we manage to build flutter nodes for 50$ (which is probably conservative, but might be doable). It would take around $3000 to cover the approximately 60 nodes between halifax and windsor. And a single point of failure would take them all out.
Not a good price for a 1.2Mbps network. But pretty good for a scientific sensor network, or as a way of getting drone telemetry.
The prices are similar if you wanted to just blanket halifax in nodes. At the point it might be usable as an indie cellphone network, although more then a few voice calls at once would saturate it. It would be great for email and text though. Maybe some light web browsing.
But the amount of money involved would be pretty nasty for a community backed project. A lot of money for something that isn’t even close to competitive with existing infrastructure.