Re: More APs with single ­SSID

[Milan Krčmář]

Thank you all for your comments. I will make some tests with
Deauth/Disassoc frames and let you know whether it helps. It will take
some time.

I will also try to sort out the most "offending" STA implementations
and solve them by means of software upgrade and/or additional
application (I've seen such a scanning app for Android, at least).

Milan

On Sun, Feb 3, 2013 at 12:19 PM, Jouni Malinen <j at w1.fi> wrote:
> On Sun, Jan 27, 2013 at 01:35:39PM +0100, Milan Krčmář wrote:
>> Yes, I have forgotten to emphasize that it is primarily STA's problem
>> and that I would like to discuss how to mitigate it from AP's point of
>> view, because you usually have not enough control over STAs: we
>> operate AP's that should work with usual STA implementations. We
>> observed Android phones behave slightly better than brand (HP etc.)
>> notebooks (mostly Intel radios)...
>
> Generally, this should really be solved at the STA, but as far as the
> question of whether the AP side can somehow help with the issue you are
> seeing is concerned, there is not really a good mechanism available with
> already deployed devices. IEEE 802.11v added BSS Transition Management
> functionality which could potentially help with AP requesting the
> station to roam. However, that would require station changes and the
> station could as well be made to roam more easily on its own in such
> case..
>
> If the network side knows that the station would see another AP with
> better signal strength (which is difficult unless the STA is reporting
> such information), a Deauthentication or Disassociation frame could be
> sent to force the STA to do something. However, this comes at a price of
> disconnection (even if only a brief one) and some STAs could end up
> trying to return to the old AP anyway..
>
> --
> Jouni Malinen                                            PGP id EFC895FA
> _______________________________________________
> HostAP mailing list
> HostAP at lists.shmoo.com
> http://lists.shmoo.com/mailman/listinfo/hostap