[Fwd: delay in adhoc mode]

[Jay Zhang]

Pavel,

Thanks a lot for the information.
Ideally, we would like to build a ad hoc system that all stations can 
receive and broadcast information. The security is not an issue here. It 
is very attractive if we can avoid authentication and association 
completely.
If I am right here, the firmware processes part of those frames ( We are 
using Prism2/3 based cards ), which is beyond our control.
One way I can think to achieve it is to put the card in Monitor mode 
first for listening and then switch back to ad hoc mode for 
broadcasting. But it looks awkward and I don't know if it will be fast 
enough as the information exchange should finish within several hundreds 
of milliseconds .

It will be great if you can give me some further advise or direct me to 
the right direction.

Thank you very much.

Jay.


Pavel Roskin wrote:

>Maybe you should avoid 802.11 authentication and association completely.
>One station could just monitor all packets and the other would send a
>special packet with the required information in it.  That packet would be
>selected and decrypted.
>
>N-way handshakes are bound to be slow because 802.11 is based on collision
>avoidance.  If you read the standard, you'll see that the stations have to
>wait a lot to avoid jamming other stations.
>
>Ideally, include extra data for forward error correction to reduce the
>risk that the frame will need to be retransmitted.
>
>Passive scanning is not going to help you - it's slow because the scanning
>station doesn't request the response, but waits for beacons passively.
>
>  
>