Recommended Beacon Interval for 2.4GHz vs 5GHz dual band networks with single SSID

[Jeremy Ward]

Jouni,

Thanks for offering your insight on this.

My understanding of Ruckus's thinking surrounding retarding 2.4GHz beacons
while accelerating 5GHz beacons is to gain the marginal chance that STAs
will associate with the 5GHz radio, specifically in high density Wi-Fi
networks; this difference, albeit marginal, has enough of an impact when
used in conjunction with controller-based STA steering methods, standards
based or otherwise.

In high density, retarding the beacons in 2.4GHz has the added benefit of
reducing the the utilization percentage of a channel, marginally increasing
effective data rates in light of the heavy frequency reuse needed for HD
networks -- directional or otherwise.

But back to hostAPd:

We are looking to use hostAPd with on appliance that will never have any
other APs present on the network other than one 2.4GHz and one 5GHz; also,
there will be no 802.11-based authentication (open system).

Given this fact, and our desire to "offload" as many STAs to the 5GHz AP as
possible, how do you suggest that we implement our requirements with
hostAPd?

My thinking was that we could utilize 802.11k action frames with clients
that support it (we are dealing with a LOT of iPhones), scripting a
"graceful" disassociation from the 2.4GHz AP, and invoking a "forced" roam
to the 5GHz AP. Please let me know if my logic is flawed.

My understanding, however, is that hostAPd does not currently support
802.11k; is this also correct?

Thanks for your time.

Jeremy D. Ward, CWNE
(954) 661-4965
On Jan 10, 2015 6:25 AM, "Jouni Malinen" <j at w1.fi> wrote:

> On Mon, Jan 05, 2015 at 03:31:40PM -0500, Jeremy Ward wrote:
> > Looking for a reccomendation as to what should be specified for a beacon
> > interval on each wireless interface when using two cards under hostAPd,
> one
> > operating at 2.4GHz the other at 5GHz, with the same SSID on each.
> >
> > The goal is to increase the likelihood that an STA will see the 5GHz
> beacon
> > before the 2.4GHz beacon, and thus associate with the 5GHz AP and not the
> > 2.4GHz AP.
>
> Please note that most station use active scanning on 2.4 GHz and
> changing the beacon interval won't have almost any effect on them
> finding the AP from the 2.4 GHz band. On 5 GHz, passive scanning is much
> more commonly used, but even then, I'd assume there to be only a very
> small difference when changing for AP discovery when changing the beacon
> interval between reasonable values. It needs to be understood that
> beacon interval changes will have significant drawbacks for other areas
> by either increasing latency on power save data delivery or increasing
> power consumption on stations.
>
> > A number of enterprise-class dual-band APs use exactly this tactic to
> > "entice" a dual-band STA to associate with the 5GHz radio, mostly because
> > out in the wild, 5GHz is less encumbered than 2.4GHz and has close to 10x
> > the amount of available spectrum (regulatory domain dependent, of course)
>
> I would not recommend doing this for the purpose of trying to steer STAs
> to 5 GHz. This is yet another doomed attempt at doing something with
> vendor specific hacks on AP side. Proper way of providing guidance for
> STAs is a mechanism that is actually designed and standardized for that
> exact purpose and is used in co-operation by both the APs and STAs. BSS
> transition management and neighbor reports are such examples.
>
> --
> Jouni Malinen                                            PGP id EFC895FA
> _______________________________________________
> HostAP mailing list
> HostAP at lists.shmoo.com
> http://lists.shmoo.com/mailman/listinfo/hostap
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.shmoo.com/pipermail/hostap/attachments/20150112/9094821f/attachment.htm>