Point to Multi-Point Communication

[jmdejoanelli]

Hi all,

So I've been wondering how point to multipoint communication works. I have had a bit of a google around, and it seems most topologies use a system of time division multiplexing which makes perfect sense. But how does a new device connecting to the network know which time slice to use?
I'm guessing some kind of handshaking occurs at connection, where the client introduces itself to the host and the host allocates a frequency channel or time slice to that client. But what happens if two clients happen to try to connect to the host at the same time? Wont the handshaking be corrupted?

Any help clearing this up for me would be appreciated.

Thanks,
Josef
 

[PA0PBZ]

These kind of problems are mostly solved by a random wait time when a collision occurs, so that both clients will not try again at the same time.

[Rerouter]

or you could use similar to the can bus, where following a signal to say its initialised on the line, it spits out its serial number and reads back while doing it, if any high bits are read back as low, it waits till the other devices with a lower serial number is allocated and tries again (though with can it has a set ID, but serial numbers can be a convenient non repeating id)