Atheros DFS radar detection

[Wright, Brett]

Hi Adrian,

> -----Original Message-----
> From: adrian.chadd at gmail.com [mailto:adrian.chadd at gmail.com] On Behalf
> Of Adrian Chadd
> 
> hi,
> 
> would you mind filing a couple of PRs, just so I remember to get this
> stuff done:
> 
Sure.

> I've whacked an AR5212 reference cardbus NIC into a laptop and added
> the AR5212 parameters to dfs_null.c. I get around 40 false detects a
> second when doing around 25mbit/sec TCP iperf test.

That seems to be about the rate of false detects I now get without
setting the parameters too aggressively... However, even with that rate
of false detects and iperfing continuously, I inevitably get a radar
pattern match and channel change (usually within about 20 minutes of
continuous iperf). With the more aggressive parameters it is much better
- but am suspicious that then real radar pulse matching will fail.
Perhaps false detects causing the radar pattern to be triggered and
channel change is just going to be inevitable to some degree?

> 
> Adrian