As in the subject - is there a way to somehow force it to e.g. redo DFS check every N min or so and come back to correct channel (if it's available).
I mean on hostapd/hostapd_cli/configuration level.
I can do simple cron/timer check and just call wifi command to restart, but wanted something more fine grained.
Also I tried doing it with chan_switch (via hostapd_cli), but no matter the combination of options, the only result of that attempt is FAIL (with no particularly useful info in the logs either: IEEE 802.11 CHAN_SWITCH VHT CONFIG 0x3).
The above was tested/trierd on ZyXEL NBG6817 with recent snapshots (but on standard kernel firmware - not candelatech's one).
It's ugly, as without an independent radio to pre-scan the channels, you probably have to bring the wireless down, at least as I understand the spec. From what I understand, you can't start transmitting on a channel until the DFS check is complete, which I believe is on the order of 60 seconds. Even if you "knew" that the channel was clear using a different radio or scan, I believe that the radio would still need its minute to check.
I'm also wondering about the utility of such an action. If the radio has determined that a channel is unusable because of DFS, why would this change in a "stable" way in the future?
I'm also wondering about the utility of such an action. If the radio has determined that a channel is unusable because of DFS, why would this change in a "stable" way in the future?
cause majority of time the channel is ok to use, this happens from once - twice per day, to once per week; thus:
there are multiple accesspoints here configured on disjoint channels, so it's nice to ask them to try to return to their prefered frequencies - otherwise one of the most practical features of 5ghz becomes a total moot point
Are you intentionally trying to jam a military radar?
Well this is near middle of a big city, with 2 big shop centers nearby, in old building within concrete walls, with 10-20 dBm power. Even if I tried, I couldn't jam anything. What's even the point of the DFS in context of lower power indoor use ?
I watched some time ago a presentation about DFS, where guy was talking about kilowatt directional antennas mounted on roofs, but that's not my use case.
The splatter that even a consumer device can cause is immense. Remember that the radar receiver is looking for reflections off rain drops or even clouds, or objects hundreds of miles/km away. 10-20 dBm is a huge signal level compared to those returns and its wideband nature basically obscures everything else, as it can't be "processed out" by the radar receiver.
And even if it could, it won't be, since the radar use was there first, takes precedence, and often involves long-term capital investment -- tens or hundreds of thousands of US-equivalent moneys per station, often with lifetimes of a decade or more.
