[Aironet] Re : Max number of nodes

[Jean T. <jt...@bo...>]

Matthew Wilson wrote :
> It's 64 as far as I know...
> 
> AP's can handle like 2048 or something like that.

	I'm really interested to know where those numbers are comming from.
	Facts :
	o 802.11 use IEEE 48 bits MAC addresses, so addressing is
usually not the issue.
	o The CSMA/CA contention has an exponential increase of the
contention window, which should handle any load, but at some point the
latency is no longer practical and you exhaust the MAC retry count.
	o High number of nodes increase the probability of collisions,
use RTS/CTS to decrease collision overhead.
	o High number of nodes increase the probability of having
hidden nodes, use RTS/CTS to decrease collision overhead.
	o The bridge software in the AP has a limited briding table,
so may have a max on the number of nodes. This is not part of the
802.11 standard, so different for each node.

	My belief :
	o Max number of *simultaneously active* nodes is around a
dozen per cell
	o Max number of nodes in ad-hoc net is unlimited
	o Max number of nodes in AP net is limited by AP bridging capability

	The only way to know is to try it...

	Jean