[ath9k-devel] Interpreting "link quality" of cards driven by ath9k to estimate usable bandwidth?

[ath9k-devel] Interpreting "link quality" of cards driven by ath9k to estimate usable bandwidth?

[Steffen Dettmer]

Hi,

man iwconfig tells, "Link quality... depends totally on the driver and hardware". I would like to know how to interpret the value for WPEA-127N driven by ath9k - ideally to have a base value for estimation of "useable bandwidth" (i.e. how much data could be sent and received per time).

Is this feasible?

Is parsing the output of "iwconfig" in intervals a good approach? Which polling interval would be suited?
What would be a better way / correct way?

I tried "wpa_cli -p $path signal_poll", but it returns only "FAIL" (however, commands work). Do I something wrong?

It would be good to estimate the order of magnitude (if there are 2 or 20 MBits) and having a value to compare several links. Surely this is a very complex topic (Google found heaps of papers about), but I'm looking for a starting point for a rough "guess", but hopefully still a bit reasonable.

Any pointers appreciated!

Regards
Steffen

Example iwconfig:
          Bit Rate=48 Mb/s   Tx-Power=17 dBm   
          Retry  long limit:7   RTS thr:off   Fragment thr:off
          Encryption key:off
          Power Management:on
          Link Quality=40/70  Signal level=-70 dBm  
          Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:0  Invalid misc:17   Missed beacon:0

[ath9k-devel] Interpreting "link quality" of cards driven by ath9k to estimate usable bandwidth?

[Holger Schurig]

Hi, I wouldn't use "iwconfig" and link quality.

"iwconfig" is outdated and virtually unmaintained since years. It uses the
deprecated WEXT kernel API, which is just emulation by mac80211-based
drivers like ath9k for such antique tools. Linux distributions should
really get rid of that stuff.

The basic idea of link quality was nice: do not only account for signal
strength, but also for channel noise, packet retries and probably
transmission rates. Shape that into a value between 0 and 100 that you can
present in a GUI to an end-user. Nice. Unfortunately, it was never normed,
and every driver (if they implemented it at all) did it differently There's
no way to compare that values.

So this concept has been dropped, as far as I know.



As for "wpa_cli signal_poll", it get output like this:

Selected interface 'wlan0'
RSSI=-37
LINKSPEED=1
NOISE=-93
FREQUENCY=2462

If you get a fail, then maybe your distribution-provided wpa_supplicant is
outdated and/or using the old WEXT interface as well. Find a way to use
wpa_supplicant with -Dnl80211, and retry.

Something that should work, even if you can't get wpa_supplicant with
-dnl80211 to work, is to use "iw wlan0 link":

SSID: CENSORED
freq: 2462
RX: 1353926 bytes (12159 packets)
TX: 795 bytes (17 packets)
signal: -36 dBm
tx bitrate: 1.0 MBit/s

bss flags: short-preamble short-slot-time
dtim period: 0
beacon int: 100



2013/4/3 Steffen Dettmer <steffen.dettmer@nomadrail.com>

> Hi,
>
> man iwconfig tells, "Link quality... depends totally on the driver and
> hardware". I would like to know how to interpret the value for WPEA-127N
> driven by ath9k - ideally to have a base value for estimation of "useable
> bandwidth" (i.e. how much data could be sent and received per time).
>
> Is this feasible?
>
> Is parsing the output of "iwconfig" in intervals a good approach? Which
> polling interval would be suited?
> What would be a better way / correct way?
>
> I tried "wpa_cli -p $path signal_poll", but it returns only "FAIL"
> (however, commands work). Do I something wrong?
>
> It would be good to estimate the order of magnitude (if there are 2 or 20
> MBits) and having a value to compare several links. Surely this is a very
> complex topic (Google found heaps of papers about), but I'm looking for a
> starting point for a rough "guess", but hopefully still a bit reasonable.
>
> Any pointers appreciated!
>
> Regards
> Steffen
>
> Example iwconfig:
>           Bit Rate=48 Mb/s   Tx-Power=17 dBm
>           Retry  long limit:7   RTS thr:off   Fragment thr:off
>           Encryption key:off
>           Power Management:on
>           Link Quality=40/70  Signal level=-70 dBm
>           Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
>           Tx excessive retries:0  Invalid misc:17   Missed beacon:0
>
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20130404/fd78cf1d/attachment.html 

[ath9k-devel] Interpreting "link quality" of cards driven by ath9k to estimate usable bandwidth?

[Flavio Leonel]

good that we got to this point.

1 year've been looking for information on how to create a threshold 
signal or a range of signal threshold for ath9k but iwconfig does not 
really work anymore ..

I've been studying the iw command but found no way to disconnect the 
client when it has a signal of -82 eg reading more about the driver if 
he does not take into account the signal level but the data rate that is 
connected to the radio more than the RSSI properly.

if you have anything you can say about the quality of MCS so we can find 
a way please help ..


thanks ..


Em 04-04-2013 05:25, Holger Schurig escreveu:
> Hi, I wouldn't use "iwconfig" and link quality.
>
> "iwconfig" is outdated and virtually unmaintained since years. It uses 
> the deprecated WEXT kernel API, which is just emulation by 
> mac80211-based drivers like ath9k for such antique tools. Linux 
> distributions should really get rid of that stuff.
>
> The basic idea of link quality was nice: do not only account for 
> signal strength, but also for channel noise, packet retries and 
> probably transmission rates. Shape that into a value between 0 and 100 
> that you can present in a GUI to an end-user. Nice. Unfortunately, it 
> was never normed, and every driver (if they implemented it at all) did 
> it differently There's no way to compare that values.
>
> So this concept has been dropped, as far as I know.
>
>
>
> As for "wpa_cli signal_poll", it get output like this:
>
> Selected interface 'wlan0'
> RSSI=-37
> LINKSPEED=1
> NOISE=-93
> FREQUENCY=2462
>
> If you get a fail, then maybe your distribution-provided 
> wpa_supplicant is outdated and/or using the old WEXT interface as 
> well. Find a way to use wpa_supplicant with -Dnl80211, and retry.
>
> Something that should work, even if you can't get wpa_supplicant with 
> -dnl80211 to work, is to use "iw wlan0 link":
>
> SSID: CENSORED
> freq: 2462
> RX: 1353926 bytes (12159 packets)
> TX: 795 bytes (17 packets)
> signal: -36 dBm
> tx bitrate: 1.0 MBit/s
>
> bss flags:short-preamble short-slot-time
> dtim period:0
> beacon int:100
>
>
>
> 2013/4/3 Steffen Dettmer <steffen.dettmer@nomadrail.com 
> <mailto:steffen.dettmer@nomadrail.com>>
>
>     Hi,
>
>     man iwconfig tells, "Link quality... depends totally on the driver
>     and hardware". I would like to know how to interpret the value for
>     WPEA-127N driven by ath9k - ideally to have a base value for
>     estimation of "useable bandwidth" (i.e. how much data could be
>     sent and received per time).
>
>     Is this feasible?
>
>     Is parsing the output of "iwconfig" in intervals a good approach?
>     Which polling interval would be suited?
>     What would be a better way / correct way?
>
>     I tried "wpa_cli -p $path signal_poll", but it returns only "FAIL"
>     (however, commands work). Do I something wrong?
>
>     It would be good to estimate the order of magnitude (if there are
>     2 or 20 MBits) and having a value to compare several links. Surely
>     this is a very complex topic (Google found heaps of papers about),
>     but I'm looking for a starting point for a rough "guess", but
>     hopefully still a bit reasonable.
>
>     Any pointers appreciated!
>
>     Regards
>     Steffen
>
>     Example iwconfig:
>               Bit Rate=48 Mb/s   Tx-Power=17 dBm
>               Retry  long limit:7   RTS thr:off   Fragment thr:off
>               Encryption key:off
>               Power Management:on
>               Link Quality=40/70  Signal level=-70 dBm
>               Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
>               Tx excessive retries:0  Invalid misc:17   Missed beacon:0
>
>     _______________________________________________
>     ath9k-devel mailing list
>     ath9k-devel at lists.ath9k.org <mailto:ath9k-devel@lists.ath9k.org>
>     https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>
>
>
>
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20130404/aaab7cfe/attachment.htm 

[ath9k-devel] Interpreting "link quality" of cards driven by ath9k to estimate usable bandwidth?

[Steffen Dettmer]

Hi,

thanks a lot for your helpful mail. I made several tests and good
progress, thanks to you :)

* Holger Schurig [mailto:holgerschurig at gmail.com] wrote:
> Hi, I wouldn't use "iwconfig" and link quality.
>
> "iwconfig" is outdated and virtually unmaintained since years.

Ohh, good to know... it is mentioned by a lot of howtos, so I
completely missed that it could be deprecated.

> So this concept has been dropped, as far as I know.

ok, I see.

> If you get a fail, then maybe your distribution-provided
> wpa_supplicant is outdated and/or using the old WEXT interface
> as well. Find a way to use wpa_supplicant with -Dnl80211, and
> retry.

Indeed the WEXT interface was used (I guess it is default). I
tested with -Dnl80211 and found it working, exactly as you told,
thank you!

However, if we start wpa_supplicant this way, we drop support for
all chips not (yet) supported by nl80211 (which probably is the
reason why it is not default), right? Are this only a few
outdated devices?

Documentation (of iw) just briefly tells "It supports all new
drivers that have been added to the kernel recently." but does
not tell a date.

> Something that should work, even if you can't get
> wpa_supplicant with -dnl80211 to work, is to use "iw wlan0 link":

Yes, this works for me both with and without -Dnl80211, very well!
I tested a bit monitoring this ten times per second while
changing anntennas and so on and it looks very good and promising.

It also shows the BSSID, so provides all information we need.

So I think forgetting about wpa_cli but instead using iw actually
is the right way to go here?

Thanks so much for pointing me to the right direction,
and have a nice weekend,

Steffen