[PATCH 1/6] bss coloring: add support for handling collision events and triggering CCA

Ryder Lee ryder.lee at mediatek.com
Wed Nov 3 23:41:41 PDT 2021


> > (!eloop_is_timeout_registered(hostapd_switch_color_timeout_handler,
> > hapd, NULL))
> > +		eloop_register_timeout(DOT11BSS_COLOR_COLLISION_AP_PERI
> > OD, 0,
> > +				       hostapd_switch_color_timeout_han
> > dler, hapd, NULL);
> > +}
> 
> What is this behavior based on? DOT11BSS_COLOR_COLLISION_AP_PERIOD
> definition refers to IEEE Std 802.11ax-2021, 26.17.3.5.1 which notes
> that the BSS Color Disabled subfield is set to 1 if the BSS color
> collision persistent for at least dot11BSSColorCollisionAPPeriod.
> However, this timeout seems to be used for changing the BSS color,
> not
> only for disabling BSS color.

This is a series of continuous actions I think. It depends on how we
interpret the 26.17.3.5 and what's the best time to change the color,
but I think most of implementations are similar.

1. BSS color collision persistent for at least
dot11BSSColorCollisionAPPeriod.
2. BSS Color Disabled subfield is set to 1.
3. Switch color out.

> > diff --git a/src/ap/hostapd.h b/src/ap/hostapd.h
> > +	int cca_in_progress;
> 
> bool
> 
Will do, It's better to use bool indeed. I thought of this but changed
my mind after looking at csa_in_progress, so keeping int for the sake
of consistency.

> 
> > diff --git a/src/common/ieee802_11_defs.h
> > b/src/common/ieee802_11_defs.h
> > +/* IEEE802.11/D6.0 - 26.17.3.5.1
> > + * the minimum default timeout between color collision and color
> > change is defined as 50s
> > + */
> > +#define DOT11BSS_COLOR_COLLISION_AP_PERIOD	50
> 
> That comment is a bit confusing.. IEEE Std 802.11ax-2021, 26.17.3.5.1
> seems to be talking about disabling BSS color after
> dot11BSSColorCollisionAPPeriod and so does the definition of this MIB
> variable while this comment talks about time between the color
> collision
> and color change which sounds quite different.

I will revise this comment to reflect the IEEE Std 802.11ax-2021,
26.17.3.5.1.

Thank you for reviewing I will fix other parts.

Ryder




More information about the Hostap mailing list