[wireless-regdb] [PATCH] Specify Australian DFS region as DFS-ETSI

[Seth Forshee]

On Tue, May 05, 2015 at 12:58:58PM +0200, Bjørn Mork wrote:
> Jouni Malinen <jouni at qca.qualcomm.com> writes:
> > On Tue, May 05, 2015 at 11:32:16AM +0300, Dmitry Ivanov wrote:
> >> Currently, DFS region for Australia is not specified. This prevents starting AP with AU country code on DFS-affected channel in 5 GHz band with latest cfg80211 and hostapd because there is an explicit check for such condition.
> >> 
> >> This patch specifies DFS region for Australia as DFS-ETSI. Alternatively, DFS-FCC can be used.
> >
> > I would have used DFS-FCC here unless you have a clear indication of
> > DFS-ETSI being closer to requirements. In addition, there should be a
> > Signed-off-by: tag in the commit log as described in the CONTRIBUTING
> > file.
> 
> From what I can gather from
> http://www.comlaw.gov.au/Details/F2014L01253
> the details are hidden in the standard "AS/NZS 4268:2012". The scope
> here:
>   http://shop.standards.co.nz/catalog/4268:2012%28AS|NZS%29/scope?
> shows the existence of an appendix "B" titled
> 
>  "DYNAMIC FREQUENCY SELECTION (DFS) AND TRANSMIT POWER CONTROL (TPC)
>   REQUIREMENTS FOR RADIO LOCAL AREA NETWORK (RLAN) TRANSMITTERS
>   OPERATING IN 5250 TO 5350 MHz OR 5470 TO 5725 MHz"
> 
> But I'm afraid you'll have to buy the standard to get the full details...
> 
> Given hysterical raisins, and the general tendency to reference ETSI
> standards and ERC decisions, I'd say that DFS-ETSI is more likely.  Why
> would they use FCC rules?
> 
> But it should probably be kept aligned with the NZ entry, which
> currently is DFS-FCC.  Don't know the source of that, though.  Could
> very well be wrong...  Ref the "EN 301 893 V1.7.1" alternative given
> here:  https://gazette.govt.nz/notice/id/2015-go694

That's the most official-looking (free) source of information I could
trun up as well.

Dmitry, you site AS/NZS 4268 as your source - do you have access to
this, and can you confirm that the requirements are closest to DFS-ETSI?
It would be great if someone else with access to the document could ack
this as well, since at first blush this appears to be in contradiction
to Peter's email.

I agree that it doesn't seem to make any sense for Australia and New
Zealand to be different here. So if Australia is specified to use
DFS-ETSI I think New Zealand should be changed to use it as well.

Thanks,
Seth