> The mostly unpredictable and extremely low duty cycle modes of non-line of sight VHF/UHF propagation like tropospheric ducting are not a realistic option for communications networks. Alternately the more reliable tropospheric scatter is a brute force solution requiring high output power of a kilowatt or more with big horn antennas.

A friend and myself are currently prototyping a solution where we transmit data using the near vertical incidence skywave (NVIS). Possibly the only option, when you want to avoid infrastructure at all costs. Of course you have the disadvantage of huge HF-antennas. To make this setup usable at all, we are trying with 20m long copper cables close to the ground. If this does not work we will try magnetic loop antennas.

In case you have something to share even if on documentation level, I'd invite you to join https://github.com/radio3-network so more people can help.

Over there we are building an operating system for ESP32 which then controls the LoRa module. Some boards are already coming with LoRa built-in and is possible to talk with chinese manufacturers to customize them.

This might be interesting for you: https://www.kk5jy.net/LoG/

HF is not legally useful for fast data networks.

Not only will the antennas have to be huge (or really inefficient like electrically small mag loops which throw away ~20dB of signal versus a resonant size antenna) but the legal limits on channel bandwidth and baud rate kick in.

You cannot legally do high rate networks on HF NVIS.

It's true, this will certainly not be a high speed network as 2.7 kHz is the max allowed bandwidth (compared 500 kHz for LoRa). But it should be fast enough for transmitting text messages.

Antenna size is problematic, but NVIS does not require the antenna to be high up in the air. Also the polarisation must not be vertical, so throwing a simple wire dipole on the ground, might actually do the job. But we will see, it's an experiment.

Antennas in cellular networks are higher on poles to extand coverage, yes, but the key for coverage is that they are cellular. They add more antennas/base stations to cover huge areas and the whole complexity of these networks is in managing this.

You've got it backwards. The key to cellular networks is that they pay money for height above terrain and thus line of sight for the cells' antennas. Otherwise it'd be infeasible. Trying to build a cellular network with the antennas at ground level is worse than a fools errand. Just look at the number of basestations required for deployal of mm-wave networks in urban cores.

Macrocells have 'macro' coverage because they are indeed better located and more powerful (and lower frequency that mm-wave).

But this tends to get smaller and smaller because of higher throughput and higher capacity demands. 5G/mm-wave/densification means more, smaller cells.

In any case, the key to ubiquitous coverage is the cellular aspect: You don't build a taller mast with a more powerful transmitter. You add more cells.