Networking

Leading research in computer science has led to advances in robust mesh networks. Field proven legacy systems such as AX.25 may also be suitable for this system. AX.25 is one method used to pass data through changing networks with a source and destination address. https://en.m.wikipedia.org/wiki/AX.25 For the first generation of field prototypes, it is OK for every node to have a preprogrammed identifier that looks for a destination identifier. APRS uses AX.25 with very inexpensive transmitters. As the mobile node (perhaps on a balloon floating around the world) travels, the network automatically routes packets. The source and destination identifiers are fixed. We are quite flexible with the network connectivity, as long as loss of a node doesn't bring the network down when other nodes are in range.

Max number of hops: let's start by assuming 2. Yes, that substantially constrains max data throughput. I don't want to start off too optimistic that we'll always be only one hop from the Internet. We will have to adapt our science data stream rate to this reality.

Minimum number of nodes

For ionospheric measurements, even two sites are useful as an advanced ionosonde or as a critical data link for remote areas (e.g. hospitals, clinics).