undefined | Better HN

0 pointsrapjr96y ago0 comments

Interesting. I wonder how well it would work in practice though. They assume uniform randomly distributed nodes, whereas people don't tend to distribute themselves randomly. They don't address practical concerns either, like how narrow the MIMO beamwidths need to be and how easy that is to achieve. Still a nice theoretical result.

0 comments

1 comments · 1 top-level

foobarian6y ago

This seems like it requires point-to-point communication; and with accurate MIMO configuration my intuition for the endpoints is that it's like they are communicating with laser-like narrow beams but in RF. But AFAIK multihop communication is still stuck at sqrt(n) scaling, i.e. not viable.

This is all pretty theoretical though, for a small number of nodes like up to 100 it should work pretty well like in the OP application.

j / k navigate · click thread line to collapse