26 Sep
2021
26 Sep
'21
11:21 a.m.
Greg Hall <greghall(a)brushbox.info> writes:
Lots of interesting thoughts and data points. Will comment only on one:
5) All of the systems I've played with allow WiFi
connex, but it
really is intended that each user has a LORA device, and then the BT
personal area network makes sense. It's the LORA device that has the
identity on the mesh, not the connected user(s).
I think this is going to be a big deal in community deployments. One
model is a bunch of people who all have disaster.radio (or whatever)
nodes that want to communicate with each other, and the other is also
extending some limited comms to the public (at a community center). So
architecturally keeping the concept of "user" that can access a node,
maybe a different one later, and "node", that speaks the over-lora
routing protocol, separate seems useful but also very difficult.
Not really in response to your message but in general:
When I talked about scalability, I didn't mean the ability to procure
and set up N nodes. I meant: once you have N nodes, does the protocol
work, or does control traffic overwhelm capacity, and do loss rates and
hop counts lead to messages not getting through. Going from 3 to 10 to
30 to 100 is usually a big deal in ad hoc radio systems.
We haven't talked about mobility. Generally, if nodes are known not to
move routing is easier because your update rates can be lower, because
you can believe your topology information for a longer period of time.