Discussion:
Progress on Ath9k HT support on Adhoc mode
(too old to reply)
Peizhao Hu
2010-12-03 05:30:30 UTC
Permalink
Hi

Could anyone please tell me the progress on Ath9k HT support for adhoc
mode?

On the latest OpenWRT trunk build, I can only get 22Mbps throughput with
54Mbit/s rate.

below are my setting for the AR9220 radio.

config 'wifi-device' 'radio0'
option 'type' 'mac80211'
option 'macaddr' '00:0c:42:64:bf:70'
option 'hwmode' '11na'
option 'htmode' 'HT40-'
list 'ht_capab' 'SHORT-GI-40'
list 'ht_capab' 'TX-STBC'
list 'ht_capab' 'RX-STBC1'
list 'ht_capab' 'DSSS_CCK-40'
option 'country' 'US'
option 'channel' '44'
option 'txpower' '1'

config 'wifi-iface'
option 'device' 'radio0'
option 'ssid' 'ath9kTest2'
option 'encryption' 'none'
option 'network' 'wan'
option 'mode' 'adhoc'
--
regards;

Peizhao
Mohammed Shafi
2010-12-03 06:51:01 UTC
Permalink
Post by Peizhao Hu
Hi
Could anyone please tell me the progress on Ath9k HT support for adhoc
mode?
adhoc mode is handled in mac80211 and that is yet to support HT.
Post by Peizhao Hu
On the latest OpenWRT trunk build, I can only get 22Mbps throughput with
54Mbit/s rate.
below are my setting for the AR9220 radio.
config 'wifi-device' 'radio0'
    option 'type' 'mac80211'
    option 'macaddr' '00:0c:42:64:bf:70'
    option 'hwmode' '11na'
    option 'htmode' 'HT40-'
    list 'ht_capab' 'SHORT-GI-40'
    list 'ht_capab' 'TX-STBC'
    list 'ht_capab' 'RX-STBC1'
    list 'ht_capab' 'DSSS_CCK-40'
    option 'country' 'US'
    option 'channel' '44'
    option 'txpower' '1'
config 'wifi-iface'
    option 'device' 'radio0'
    option 'ssid' 'ath9kTest2'
    option 'encryption' 'none'
    option 'network' 'wan'
    option 'mode' 'adhoc'
--
regards;
Peizhao
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Alexander Simon
2010-12-09 12:36:19 UTC
Permalink
Im working on it.

I took the initial patches vom Benoit Papillault. Right now, i have two cells
deploy and read HT beacons and have them entered into the sta_info structure.
Minstrel_ht also seems to drive at HT rates. I hope I can provide a first
patch tomorrow, as at the moment I also have problems with txpower.

Greetings, Alex
Luis R. Rodriguez
2010-12-09 18:55:46 UTC
Permalink
Post by Alexander Simon
Im working on it.
I took the initial patches vom Benoit Papillault. Right now, i have two cells
deploy and read HT beacons and have them entered into the sta_info structure.
Minstrel_ht also seems to drive at HT rates. I hope I can provide a first
patch tomorrow, as at the moment I also have problems with txpower.
This is great news, please feel free to edit:

http://wireless.kernel.org/en/users/Drivers/ath9k/todo
http://wireless.kernel.org/en/developers/todo-list

Luis
Robert Chan
2010-12-22 08:25:39 UTC
Permalink
Hi All,

May I ask whether the HT support (for Adhoc) is already in the trunk?

Thanks.

Regards,

Robert


--------------------------------------------------
From: "Alexander Simon" <***@web.de>
Sent: Thursday, December 09, 2010 8:36 PM
To: <ath9k-***@lists.ath9k.org>
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Alexander Simon
Im working on it.
I took the initial patches vom Benoit Papillault. Right now, i have two cells
deploy and read HT beacons and have them entered into the sta_info structure.
Minstrel_ht also seems to drive at HT rates. I hope I can provide a first
patch tomorrow, as at the moment I also have problems with txpower.
Greetings, Alex
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Robert Chan
2010-12-29 06:10:13 UTC
Permalink
Dear all,

After some research done on ath9k's support of HT on Adhoc mode, I found on
the mac80211 driver page (http://wireless.kernel.org/en/users/Drivers) that
ath9k seems to support IBSS *and/or* mesh *and/or* 802.11N:

Driver Manufacturer cfg80211 AP IBSS mesh monitor PHY
modes Buses
ath9k Atheros yes yes yes yes yes A/B/G/N PCI
/ PCI-E / AHB / PCMCIA

So should I assume that the details here are "too updated" or are the modes
above are only independently supported (but not able to work together such
as IBSS + N + mesh)?

Any clarification will be very much appreciated.

Regards,

Robert

--------------------------------------------------
From: "Robert Chan" <***@gmail.com>
Sent: Wednesday, December 22, 2010 4:25 PM
To: "Alexander Simon" <***@web.de>; <ath9k-***@lists.ath9k.org>
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
--------------------------------------------------
Sent: Thursday, December 09, 2010 8:36 PM
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Alexander Simon
Im working on it.
I took the initial patches vom Benoit Papillault. Right now, i have two cells
deploy and read HT beacons and have them entered into the sta_info structure.
Minstrel_ht also seems to drive at HT rates. I hope I can provide a first
patch tomorrow, as at the moment I also have problems with txpower.
Greetings, Alex
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Brian Prodoehl
2010-12-29 20:20:58 UTC
Permalink
Post by Robert Chan
Dear all,
After some research done on ath9k's support of HT on Adhoc mode, I found on
the mac80211 driver page (http://wireless.kernel.org/en/users/Drivers) that
Driver    Manufacturer    cfg80211    AP    IBSS    mesh    monitor    PHY
modes    Buses
ath9k     Atheros    yes     yes     yes     yes     yes     A/B/G/N     PCI
/ PCI-E / AHB /     PCMCIA
So should I assume that the details here are "too updated" or are the modes
above are only independently supported (but not able to work together such
as IBSS + N + mesh)?
Any clarification will be very much appreciated.
Regards,
Robert
--------------------------------------------------
Sent: Wednesday, December 22, 2010 4:25 PM
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
--------------------------------------------------
Sent: Thursday, December 09, 2010 8:36 PM
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Alexander Simon
Im working on it.
I took the initial patches vom Benoit Papillault. Right now, i have two cells
deploy and read HT beacons and have them entered into the sta_info structure.
Minstrel_ht also seems to drive at HT rates. I hope I can provide a first
patch tomorrow, as at the moment I also have problems with txpower.
Greetings, Alex
Mesh in that list is entirely different than IBSS. It is an
802.11s-compliant MAC-layer mesh. Of course, you're free to run mesh
routing protocols over an IBSS network, but that's not what they mean
by mesh in that list of supported features. Adding support for HT
rates under IBSS is ongoing, outside of anything you'll find in
compat-wireless. So, you're correct in that 802.11n and IBSS do not
work together yet. I believe the same is true for the 802.11s mesh.

Brian
Robert Chan
2010-12-29 06:10:34 UTC
Permalink
Dear all,

After some research done on ath9k's support of HT on Adhoc mode, I found on
the mac80211 driver page (http://wireless.kernel.org/en/users/Drivers) that
ath9k seems to support IBSS *and/or* mesh *and/or* 802.11N:

Driver Manufacturer cfg80211 AP IBSS mesh monitor PHY
modes Buses
ath9k Atheros yes yes yes yes yes A/B/G/N PCI
/ PCI-E / AHB / PCMCIA

So should I assume that the details here are "too updated" or are the modes
above are only independently supported (but not able to work together such
as IBSS + N + mesh)?

Any clarification will be very much appreciated.

Regards,

Robert

--------------------------------------------------
From: "Robert Chan" <***@gmail.com>
Sent: Wednesday, December 22, 2010 4:25 PM
To: "Alexander Simon" <***@web.de>; <ath9k-***@lists.ath9k.org>
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
--------------------------------------------------
Sent: Thursday, December 09, 2010 8:36 PM
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Alexander Simon
Im working on it.
I took the initial patches vom Benoit Papillault. Right now, i have two cells
deploy and read HT beacons and have them entered into the sta_info structure.
Minstrel_ht also seems to drive at HT rates. I hope I can provide a first
patch tomorrow, as at the moment I also have problems with txpower.
Greetings, Alex
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Mohammed Shafi
2010-12-29 06:32:04 UTC
Permalink
On Wed, Dec 29, 2010 at 11:40 AM, Robert Chan
Post by Robert Chan
Dear all,
After some research done on ath9k's support of HT on Adhoc mode, I found on
the mac80211 driver page (http://wireless.kernel.org/en/users/Drivers) that
Driver    Manufacturer    cfg80211    AP    IBSS    mesh    monitor    PHY
modes    Buses
ath9k     Atheros    yes     yes     yes     yes     yes     A/B/G/N     PCI
/ PCI-E / AHB /     PCMCIA
So should I assume that the details here are "too updated" or are the modes
above are only independently supported (but not able to work together such
as IBSS + N + mesh)?
I think 'mesh mode' also handled in mac80211 , if mesh mode is
supported with 11n then ibss(which is a sub-set ?) should have 11n
support.
If I am wrong some one please correct me.
Post by Robert Chan
Any clarification will be very much appreciated.
Regards,
Robert
--------------------------------------------------
Sent: Wednesday, December 22, 2010 4:25 PM
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
--------------------------------------------------
Sent: Thursday, December 09, 2010 8:36 PM
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Alexander Simon
Im working on it.
I took the initial patches vom Benoit Papillault. Right now, i have two cells
deploy and read HT beacons and have them entered into the sta_info structure.
Minstrel_ht also seems to drive at HT rates. I hope I can provide a first
patch tomorrow, as at the moment I also have problems with txpower.
Greetings, Alex
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Jouni Malinen
2010-12-29 06:52:29 UTC
Permalink
Post by Mohammed Shafi
On Wed, Dec 29, 2010 at 11:40 AM, Robert Chan
Post by Robert Chan
So should I assume that the details here are "too updated" or are the modes
above are only independently supported (but not able to work together such
as IBSS + N + mesh)?
The modes in that table should be considered independent and support for
802.11n in the table should really be read as at least one of the modes
supporting it, not all modes supporting it. So no, that does not mean
that either IBSS or mesh would support 802.11n (AP and station modes
do).
Post by Mohammed Shafi
I think 'mesh mode' also handled in mac80211 , if mesh mode is
supported with 11n then ibss(which is a sub-set ?) should have 11n
support.
Mesh and IBSS are independent modes and both will need to be
independently addressed for 802.11n support. In addition, both of them
have protocol limitations on which 802.11n HT features can be supported.
Finally, HT support for either mode is not yet complete in mac80211 and
as such, cannot be complete with ath9k either.

- Jouni
Mohammed Shafi
2010-12-29 07:00:12 UTC
Permalink
On Wed, Dec 29, 2010 at 12:22 PM, Jouni Malinen
Post by Jouni Malinen
Post by Mohammed Shafi
On Wed, Dec 29, 2010 at 11:40 AM, Robert Chan
Post by Robert Chan
So should I assume that the details here are "too updated" or are the modes
above are only independently supported (but not able to work together such
as IBSS + N + mesh)?
The modes in that table should be considered independent and support for
802.11n in the table should really be read as at least one of the modes
supporting it, not all modes supporting it. So no, that does not mean
that either IBSS or mesh would support 802.11n (AP and station modes
do).
Post by Mohammed Shafi
I think 'mesh mode' also handled in mac80211 , if mesh mode is
supported with 11n then ibss(which is a sub-set ?) should have 11n
support.
Mesh and IBSS are independent modes and both will need to be
independently addressed for 802.11n support. In addition, both of them
have protocol limitations on which 802.11n HT features can be supported.
Finally, HT support for either mode is not yet complete in mac80211 and
as such, cannot be complete with ath9k either.
Thanks a lot.
Post by Jouni Malinen
- Jouni
Robert Chan
2010-12-29 07:21:13 UTC
Permalink
Hi Jouni and Mohammed,

Thanks very much for the quick reply.

As I am implementing a project that requires a 802.11N mesh network at low
(or zero actually, for a poor area deployment) budget. May I ask what is
the progress in that end (mesh over 11N)? Or if I want to jump in, which
part of the code should I begin with (of course I really doubt I could do
much all by myself, but I will try to probably help out)?

Thanks very much again in advance.

Regards,

Robert

--------------------------------------------------
From: "Jouni Malinen" <***@atheros.com>
Sent: Wednesday, December 29, 2010 2:52 PM
To: "Mohammed Shafi" <***@gmail.com>
Cc: "Robert Chan" <***@gmail.com>;
<ath9k-***@lists.ath9k.org>
Subject: Re: [ath9k-devel] Progress on Ath9k HT support on Adhoc mode
Post by Jouni Malinen
Post by Mohammed Shafi
On Wed, Dec 29, 2010 at 11:40 AM, Robert Chan
Post by Robert Chan
So should I assume that the details here are "too updated" or are the modes
above are only independently supported (but not able to work together such
as IBSS + N + mesh)?
The modes in that table should be considered independent and support for
802.11n in the table should really be read as at least one of the modes
supporting it, not all modes supporting it. So no, that does not mean
that either IBSS or mesh would support 802.11n (AP and station modes
do).
Post by Mohammed Shafi
I think 'mesh mode' also handled in mac80211 , if mesh mode is
supported with 11n then ibss(which is a sub-set ?) should have 11n
support.
Mesh and IBSS are independent modes and both will need to be
independently addressed for 802.11n support. In addition, both of them
have protocol limitations on which 802.11n HT features can be supported.
Finally, HT support for either mode is not yet complete in mac80211 and
as such, cannot be complete with ath9k either.
- Jouni
Jouni Malinen
2010-12-29 08:30:09 UTC
Permalink
Post by Robert Chan
As I am implementing a project that requires a 802.11N mesh network at low
(or zero actually, for a poor area deployment) budget. May I ask what is
the progress in that end (mesh over 11N)? Or if I want to jump in, which
part of the code should I begin with (of course I really doubt I could do
much all by myself, but I will try to probably help out)?
I haven't followed this closely, so I don't have a good answer to that
question. Anyway, this is not really the best mailing list for that
either since the mesh code is almost completely in mac80211 and not in
the drivers like ath9k. linux-wireless mailing list would be better
place for this discussion:
http://wireless.kernel.org/en/developers/MailingLists


- Jouni
Alexander Simon
2011-01-19 13:42:17 UTC
Permalink
Hi there,

i just want to let you know i just released my patch for this on linux-
wireless.
http://article.gmane.org/gmane.linux.kernel.wireless.general/63191
I really apreciate any tests/comments/proposals on this!

Regards, Alex
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
Peizhao Hu
2011-02-04 03:04:20 UTC
Permalink
Hi Alex,

I am testing the patch. I got the following dmesg message.

I am using compat-wireless-2011-01-17.tar.bz2 and iw-0.9.21.tar.bz2 with
both patches. The linux kernel is 2.6.35 from linux-wireless.

[ 7549.786227] wlan0: No active IBSS STAs - trying to scan for other
IBSS networks with same SSID (merge)
[ 7549.925224] ath: cf559e20
[ 7549.928625] ath: cf559e48
[ 7549.944108] ath: cf559e04
[ 7550.073159] ath: cf559e20
[ 7550.077125] ath: cf559e48
[ 7550.092581] ath: cf559e04
[ 7550.277151] ath: cf559e20
[ 7550.281123] ath: cf559e48
[ 7550.296581] ath: cf559e04
[ 7550.369159] ath: cf559e20
[ 7550.373132] ath: cf559e48
[ 7550.388592] ath: cf559e04
[ 7550.461160] ath: cf559e20
[ 7550.465117] ath: cf559e48
[ 7550.480581] ath: cf559e04
[ 7550.553160] ath: cf559e20
[ 7550.557127] ath: cf559e48
[ 7550.572581] ath: cf559e04
[ 7550.701149] ath: cf559e20
[ 7550.705107] ath: cf559e48
[ 7550.720570] ath: cf559e04
[ 7550.793161] ath: cf559e20
[ 7550.797116] ath: cf559e48
[ 7550.812569] ath: cf559e04
[ 7550.885171] ath: cf559e20
[ 7550.889124] ath: cf559e48
[ 7550.904582] ath: cf559e04

regards;

Peizhao
Post by Alexander Simon
Hi there,
i just want to let you know i just released my patch for this on linux-
wireless.
http://article.gmane.org/gmane.linux.kernel.wireless.general/63191
I really apreciate any tests/comments/proposals on this!
Regards, Alex
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Mohammed Shafi
2011-02-04 05:27:18 UTC
Permalink
Post by Peizhao Hu
Hi Alex,
I am testing the patch. I got the following dmesg message.
I am using compat-wireless-2011-01-17.tar.bz2 and iw-0.9.21.tar.bz2 with
both patches. The linux kernel is 2.6.35 from linux-wireless.
[ 7549.786227] wlan0: No active IBSS STAs - trying to scan for other
IBSS networks with same SSID (merge)
[ 7549.925224] ath: cf559e20
[ 7549.928625] ath: cf559e48
[ 7549.944108] ath: cf559e04
[ 7550.073159] ath: cf559e20
[ 7550.077125] ath: cf559e48
[ 7550.092581] ath: cf559e04
[ 7550.277151] ath: cf559e20
[ 7550.281123] ath: cf559e48
[ 7550.296581] ath: cf559e04
[ 7550.369159] ath: cf559e20
[ 7550.373132] ath: cf559e48
[ 7550.388592] ath: cf559e04
[ 7550.461160] ath: cf559e20
[ 7550.465117] ath: cf559e48
[ 7550.480581] ath: cf559e04
[ 7550.553160] ath: cf559e20
[ 7550.557127] ath: cf559e48
[ 7550.572581] ath: cf559e04
[ 7550.701149] ath: cf559e20
[ 7550.705107] ath: cf559e48
[ 7550.720570] ath: cf559e04
[ 7550.793161] ath: cf559e20
[ 7550.797116] ath: cf559e48
[ 7550.812569] ath: cf559e04
[ 7550.885171] ath: cf559e20
[ 7550.889124] ath: cf559e48
[ 7550.904582] ath: cf559e04
Can you please try with this patch named [PATCH 1/1] compat-wireless:
Fix ath9k debug log issue.
Post by Peizhao Hu
regards;
Peizhao
Post by Alexander Simon
Hi there,
i just want to let you know i just released my patch for this on linux-
wireless.
http://article.gmane.org/gmane.linux.kernel.wireless.general/63191
I really apreciate any tests/comments/proposals on this!
Regards, Alex
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Peizhao Hu
2011-02-08 03:38:51 UTC
Permalink
Hi all,

I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load

Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).

Any one have idea how I can achieve 100Mbps+ throughput with my current
setup?


By the way, I tried to the "[PATCH 1/1] compat-wireless: Fix ath9k debug
log issue" patch. However, I am still getting these error logs.

[ 228.982367] cfg80211: Found new beacon on frequency: 5180 MHz (Ch 36)
on phy0
[ 229.699391] cfg80211: Found new beacon on frequency: 5745 MHz (Ch
149) on phy0
[ 229.853470] cfg80211: Found new beacon on frequency: 5805 MHz (Ch
161) on phy0
[ 231.971055] wlan0: Trigger new scan to find an IBSS to join
[ 232.108921] ath: ce24be20
[ 232.112313] ath: ce24be48
[ 232.127773] ath: ce24be04
[ 232.200898] ath: ce24be20
[ 232.204872] ath: ce24be48
[ 232.220331] ath: ce24be04
[ 232.292898] ath: ce24be20
[ 232.296855] ath: ce24be48
[ 232.312308] ath: ce24be04
[ 232.384899] ath: ce24be20
[ 232.388865] ath: ce24be48
[ 232.404319] ath: ce24be04
[ 232.532887] ath: ce24be20
[ 232.536848] ath: ce24be48
[ 232.552297] ath: ce24be04
[ 232.624899] ath: ce24be20
[ 232.628859] ath: ce24be48
[ 232.644320] ath: ce24be04
[ 232.716898] ath: ce24be20
[ 232.720870] ath: ce24be48
[ 232.736331] ath: ce24be04
[ 232.808909] ath: ce24be20
[ 232.812880] ath: ce24be48
[ 232.828342] ath: ce24be04
[ 232.900900] ath: ce24be20
[ 232.904865] ath: ce24be48
[ 232.920320] ath: ce24be04
[ 232.992898] ath: ce24be20
[ 232.996849] ath: ce24be48
[ 233.012308] ath: ce24be04
[ 233.084898] ath: ce24be20
[ 233.088861] ath: ce24be48
[ 233.104319] ath: ce24be04
[ 233.176898] ath: ce24be20
[ 233.180870] ath: ce24be48
[ 233.196332] ath: ce24be04
[ 234.932420] wlan0: Selected IBSS BSSID ae:d9:24:0e:d9:52 based on
configured SSID
[ 236.433296] wlan0: no IPv6 routers present





regards;

Peizhao
Post by Mohammed Shafi
Post by Peizhao Hu
Hi Alex,
I am testing the patch. I got the following dmesg message.
I am using compat-wireless-2011-01-17.tar.bz2 and iw-0.9.21.tar.bz2 with
both patches. The linux kernel is 2.6.35 from linux-wireless.
[ 7549.786227] wlan0: No active IBSS STAs - trying to scan for other
IBSS networks with same SSID (merge)
[ 7549.925224] ath: cf559e20
[ 7549.928625] ath: cf559e48
[ 7549.944108] ath: cf559e04
[ 7550.073159] ath: cf559e20
[ 7550.077125] ath: cf559e48
[ 7550.092581] ath: cf559e04
[ 7550.277151] ath: cf559e20
[ 7550.281123] ath: cf559e48
[ 7550.296581] ath: cf559e04
[ 7550.369159] ath: cf559e20
[ 7550.373132] ath: cf559e48
[ 7550.388592] ath: cf559e04
[ 7550.461160] ath: cf559e20
[ 7550.465117] ath: cf559e48
[ 7550.480581] ath: cf559e04
[ 7550.553160] ath: cf559e20
[ 7550.557127] ath: cf559e48
[ 7550.572581] ath: cf559e04
[ 7550.701149] ath: cf559e20
[ 7550.705107] ath: cf559e48
[ 7550.720570] ath: cf559e04
[ 7550.793161] ath: cf559e20
[ 7550.797116] ath: cf559e48
[ 7550.812569] ath: cf559e04
[ 7550.885171] ath: cf559e20
[ 7550.889124] ath: cf559e48
[ 7550.904582] ath: cf559e04
Fix ath9k debug log issue.
Post by Peizhao Hu
regards;
Peizhao
Post by Alexander Simon
Hi there,
i just want to let you know i just released my patch for this on linux-
wireless.
http://article.gmane.org/gmane.linux.kernel.wireless.general/63191
I really apreciate any tests/comments/proposals on this!
Regards, Alex
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Mohammed Shafi
2011-02-10 14:07:51 UTC
Permalink
Post by Peizhao Hu
Hi all,
I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch suggested by
Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
Any one have idea how I can achieve 100Mbps+ throughput with my current
setup?
By the way, I tried to the "[PATCH 1/1] compat-wireless: Fix ath9k debug log
issue" patch. However, I am still getting these error logs.
[  228.982367] cfg80211: Found new beacon on frequency: 5180 MHz (Ch 36) on
phy0
[  229.699391] cfg80211: Found new beacon on frequency: 5745 MHz (Ch 149) on
phy0
[  229.853470] cfg80211: Found new beacon on frequency: 5805 MHz (Ch 161) on
phy0
[  231.971055] wlan0: Trigger new scan to find an IBSS to join
[  232.108921] ath: ce24be20
[  232.112313] ath: ce24be48
[  232.127773] ath: ce24be04
[  232.200898] ath: ce24be20
[  232.204872] ath: ce24be48
[  232.220331] ath: ce24be04
[  232.292898] ath: ce24be20
[  232.296855] ath: ce24be48
[  232.312308] ath: ce24be04
[  232.384899] ath: ce24be20
[  232.388865] ath: ce24be48
[  232.404319] ath: ce24be04
[  232.532887] ath: ce24be20
[  232.536848] ath: ce24be48
[  232.552297] ath: ce24be04
[  232.624899] ath: ce24be20
[  232.628859] ath: ce24be48
[  232.644320] ath: ce24be04
[  232.716898] ath: ce24be20
[  232.720870] ath: ce24be48
[  232.736331] ath: ce24be04
[  232.808909] ath: ce24be20
[  232.812880] ath: ce24be48
[  232.828342] ath: ce24be04
[  232.900900] ath: ce24be20
[  232.904865] ath: ce24be48
[  232.920320] ath: ce24be04
[  232.992898] ath: ce24be20
[  232.996849] ath: ce24be48
[  233.012308] ath: ce24be04
[  233.084898] ath: ce24be20
[  233.088861] ath: ce24be48
[  233.104319] ath: ce24be04
[  233.176898] ath: ce24be20
[  233.180870] ath: ce24be48
[  233.196332] ath: ce24be04
[  234.932420] wlan0: Selected IBSS BSSID ae:d9:24:0e:d9:52 based on
configured SSID
[  236.433296] wlan0: no IPv6 routers present
have no idea why those errors are coming for you, but the patch does
fixed the issue.
thanks,
shafi
Post by Peizhao Hu
regards;
Peizhao
Post by Mohammed Shafi
Post by Peizhao Hu
Hi Alex,
I am testing the patch. I got the following dmesg message.
I am using compat-wireless-2011-01-17.tar.bz2 and iw-0.9.21.tar.bz2 with
both patches. The linux kernel is 2.6.35 from linux-wireless.
[ 7549.786227] wlan0: No active IBSS STAs - trying to scan for other
IBSS networks with same SSID (merge)
[ 7549.925224] ath: cf559e20
[ 7549.928625] ath: cf559e48
[ 7549.944108] ath: cf559e04
[ 7550.073159] ath: cf559e20
[ 7550.077125] ath: cf559e48
[ 7550.092581] ath: cf559e04
[ 7550.277151] ath: cf559e20
[ 7550.281123] ath: cf559e48
[ 7550.296581] ath: cf559e04
[ 7550.369159] ath: cf559e20
[ 7550.373132] ath: cf559e48
[ 7550.388592] ath: cf559e04
[ 7550.461160] ath: cf559e20
[ 7550.465117] ath: cf559e48
[ 7550.480581] ath: cf559e04
[ 7550.553160] ath: cf559e20
[ 7550.557127] ath: cf559e48
[ 7550.572581] ath: cf559e04
[ 7550.701149] ath: cf559e20
[ 7550.705107] ath: cf559e48
[ 7550.720570] ath: cf559e04
[ 7550.793161] ath: cf559e20
[ 7550.797116] ath: cf559e48
[ 7550.812569] ath: cf559e04
[ 7550.885171] ath: cf559e20
[ 7550.889124] ath: cf559e48
[ 7550.904582] ath: cf559e04
Fix ath9k debug log issue.
Post by Peizhao Hu
regards;
Peizhao
Post by Alexander Simon
Hi there,
i just want to let you know i just released my patch for this on linux-
wireless.
http://article.gmane.org/gmane.linux.kernel.wireless.general/63191
I really apreciate any tests/comments/proposals on this!
Regards, Alex
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
bcoll
2011-01-28 08:58:05 UTC
Permalink
Dear Peizhao,

Recently has been post this patch that I think It could be interesting for you:
http://comments.gmane.org/gmane.linux.kernel.wireless.general/63191

Regards,
bcoll
le thanh son
2011-02-08 16:00:38 UTC
Permalink
Dear All

I have added the patch of HT support in adhoc mode by Alex to compat-wireless-2011-01-24 and other patch on iw tool.
My radio card is AR9220 on AR71xx processor board.
The set up is as following:

iw phy phy0 interface add wlan01 type ibss
ifconfig wlan01 up
iw wlan01 ibss join test2 5745 HT40+

There are some issues as follows:
1. It could not get higher MCS. Normally just MCS0
Station 00:xx:xx:xx:xx:ee (on wlan01)
inactive time: 300 ms
rx bytes: 63190
rx packets: 1173
tx bytes: 11724
tx packets: 100
tx retries: 0
tx failed: 0
signal: -51 dBm
signal avg: -52 dBm
tx bitrate: 6.5 MBit/s MCS 0
2. I tested throughput by iperf, tcp, it shows very low throughput, just about 1.3 Mbps. It means the lowest bit rate is in operation, not higher bit rate (such as MCS 15)

Please not that if I change to AP-STA mode, then immediately the higher MCS schemes activated (normally MCS15), so the signal is quite good.

3. Although I just set up 1 point to point link and expected only ONE adhoc station shown in the output of iw dev wlan0 station dump. But in fact it shows more than one stations ?!

***@OpenWrt:/# iw dev wlan01 station dump
Station 00:xx:xx:xx:xx:22 (on wlan01)
inactive time: 42740 ms
rx bytes: 229
rx packets: 2
tx bytes: 0
tx packets: 0
tx retries: 0
tx failed: 0
signal: -51 dBm
signal avg: -50 dBm
tx bitrate: 6.0 MBit/s
Station 00:xx:xx:xx:xx:10 (on wlan01)
inactive time: 24060 ms
rx bytes: 470
rx packets: 4
tx bytes: 0
tx packets: 0
tx retries: 0
tx failed: 0
signal: -50 dBm
signal avg: -51 dBm
tx bitrate: 6.0 MBit/s
Station 00:xx:xx:xx:xx:f5 (on wlan01)
inactive time: 0 ms
rx bytes: 7254
rx packets: 1499
tx bytes: 3412532
tx packets: 2411
tx retries: 0
tx failed: 0
signal: -52 dBm
signal avg: -52 dBm
tx bitrate: 6.0 MBit/s
Station 00:xx:xx:xx:xx:ee (on wlan01)
inactive time: 300 ms
rx bytes: 63190
rx packets: 1173
tx bytes: 11724
tx packets: 100
tx retries: 0
tx failed: 0
signal: -51 dBm
signal avg: -52 dBm
tx bitrate: 6.5 MBit/s MCS 0
***@OpenWrt:/# iw dev wlan01 link
Joined IBSS 02:xx:xx:xx:xx:45 (on wlan01)
SSID: test2
freq: 5745

The question is whether I set up correctly ? How should I set up ?

Thanks in advance

Le
Post by Peizhao Hu
Hi all,
I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
Any one have idea how I can achieve 100Mbps+ throughput with my current
setup?
By the way, I tried to the "[PATCH 1/1] compat-wireless: Fix ath9k debug
log issue" patch. However, I am still getting these error logs.
[ 228.982367] cfg80211: Found new beacon on frequency: 5180 MHz (Ch 36)
on phy0
[ 229.699391] cfg80211: Found new beacon on frequency: 5745 MHz (Ch
149) on phy0
[ 229.853470] cfg80211: Found new beacon on frequency: 5805 MHz (Ch
161) on phy0
[ 231.971055] wlan0: Trigger new scan to find an IBSS to join
[ 232.108921] ath: ce24be20
[ 232.112313] ath: ce24be48
[ 232.127773] ath: ce24be04
[ 232.200898] ath: ce24be20
[ 232.204872] ath: ce24be48
[ 232.220331] ath: ce24be04
[ 232.292898] ath: ce24be20
[ 232.296855] ath: ce24be48
[ 232.312308] ath: ce24be04
[ 232.384899] ath: ce24be20
[ 234.932420] wlan0: Selected IBSS BSSID ae:d9:24:0e:d9:52 based on
configured SSID
[ 236.433296] wlan0: no IPv6 routers present





regards;

Peizhao
Post by Peizhao Hu
Post by Peizhao Hu
Hi Alex,
I am testing the patch. I got the following dmesg message.
I am using compat-wireless-2011-01-17.tar.bz2 and iw-0.9.21.tar.bz2 with
both patches. The linux kernel is 2.6.35 from linux-wireless.
[ 7549.786227] wlan0: No active IBSS STAs - trying to scan for other
IBSS networks with same SSID (merge)
[ 7549.925224] ath: cf559e20
[ 7549.928625] ath: cf559e48
[ 7549.944108] ath: cf559e04
[ 7550.073159] ath: cf559e20
[ 7550.077125] ath: cf559e48
[ 7550.092581] ath: cf559e04
[ 7550.277151] ath: cf559e20
[ 7550.281123] ath: cf559e48
[ 7550.296581] ath: cf559e04
[ 7550.369159] ath: cf559e20
[ 7550.373132] ath: cf559e48
[ 7550.388592] ath: cf559e04
[ 7550.461160] ath: cf559e20
[ 7550.465117] ath: cf559e48
[ 7550.480581] ath: cf559e04
[ 7550.553160] ath: cf559e20
[ 7550.557127] ath: cf559e48
[ 7550.572581] ath: cf559e04
[ 7550.701149] ath: cf559e20
[ 7550.705107] ath: cf559e48
[ 7550.720570] ath: cf559e04
[ 7550.793161] ath: cf559e20
[ 7550.797116] ath: cf559e48
[ 7550.812569] ath: cf559e04
[ 7550.885171] ath: cf559e20
[ 7550.889124] ath: cf559e48
[ 7550.904582] ath: cf559e04
Fix ath9k debug log issue.
Post by Peizhao Hu
regards;
Peizhao
Post by Alexander Simon
Hi there,
i just want to let you know i just released my patch for this on linux-
wireless.
http://article.gmane.org/gmane.linux.kernel.wireless.general/63191
I really apreciate any tests/comments/proposals on this!
Regards, Alex
Post by Robert Chan
Hi All,
May I ask whether the HT support (for Adhoc) is already in the trunk?
Thanks.
Regards,
Robert
Baldomero Coll
2011-02-10 12:04:15 UTC
Permalink
Hi all,

I'm also testing the patch Alex and Benoit have released for the community.
After configuring my IBSS network, the station dump command shows this [1].
As you can see, although the IBSS network has been created in HT mode, one
link doesn't seem to support MCS rates.
I would like to know if it is possible to force the MCS scheme. I don't know
why the "NL80211_TXRATE_MCS" parameter is not defined in the bitrates.c
(iw), or what I have to do to enable it.

I also wonder what rate control mechanism is been used in HT mode (minstrel
HT?), and whether such mechanism switches to non HT rates when the channel
quality is bad.

I have been using wireshark to capture data transmissions between 2 PCs
configured in IBSS HT mode. What I see in the radiotap header is that no HT
information is available (protocol, data rate). Do you know how the HT
information can be added to the ratiotap header?

Thanks for your help
Baldo


[1]
----------------------------------------------------------------------------------------------------------------------------------------
sudo iw dev wlan0 station dump
Station xx:xx:xx:xx:xx:xx (on wlan0)
inactive time: 8392 ms
rx bytes: 2476336
rx packets: 2149
tx bytes: 433552
tx packets: 1228
tx retries: 564
tx failed: 0
signal: -27 dBm
signal avg: -28 dBm
tx bitrate: 270.0 MBit/s MCS 14 40Mhz short GI
Station xx:xx:xx:xx:xx:xx (on wlan0)
inactive time: 200 ms
rx bytes: 27577
rx packets: 434
tx bytes: 803
tx packets: 8
tx retries: 2
tx failed: 0
signal: -26 dBm
signal avg: -26 dBm
tx bitrate: 6.5 MBit/s MCS 0
Station xx:xx:xx:xx:xx:xx (on wlan0)
inactive time: 96 ms
rx bytes: 692528
rx packets: 18938
tx bytes: 2979324
tx packets: 2001
tx retries: 3340
tx failed: 0
signal: -33 dBm
signal avg: -33 dBm
tx bitrate: 54.0 MBit/s
----------------------------------------------------------------------------------------------------------------------------------------
Mohammed Shafi
2011-02-10 14:11:30 UTC
Permalink
Post by Peizhao Hu
Hi all,
I'm also testing the patch Alex and Benoit have released for the community.
After configuring my IBSS network, the station dump command shows this [1].
As you can see, although the IBSS network has been created in HT mode, one
link doesn't seem to support MCS rates.
I would like to know if it is possible to force the MCS scheme. I don't know
why the "NL80211_TXRATE_MCS" parameter is not defined in the bitrates.c
(iw), or what I have to do to enable it.
I don't think we can force a particular MCS rate in the station side,
we have rate control algorithms to determine the rate.
Post by Peizhao Hu
I also wonder what rate control mechanism is been used in HT mode (minstrel
HT?), and whether such mechanism switches to non HT rates when the channel
quality is bad.
I have been using wireshark to capture data transmissions between 2 PCs
configured in IBSS HT mode. What I see in the radiotap header is that no HT
information is available (protocol, data rate). Do you know how the HT
information can be added to the ratiotap header?
If you are using wireshark, you need to take the latest bleeding edge
that has MCS rate support and also you need latest wireless-testing
that contains MCS rate update to radiotap.
Post by Peizhao Hu
Thanks for your help
Baldo
[1]
----------------------------------------------------------------------------------------------------------------------------------------
sudo iw dev wlan0 station dump
Station xx:xx:xx:xx:xx:xx (on wlan0)
    inactive time:    8392 ms
    rx bytes:    2476336
    rx packets:    2149
    tx bytes:    433552
    tx packets:    1228
    tx retries:    564
    tx failed:    0
    signal:      -27 dBm
    signal avg:    -28 dBm
    tx bitrate:    270.0 MBit/s MCS 14 40Mhz short GI
Station xx:xx:xx:xx:xx:xx (on wlan0)
    inactive time:    200 ms
    rx bytes:    27577
    rx packets:    434
    tx bytes:    803
    tx packets:    8
    tx retries:    2
    tx failed:    0
    signal:      -26 dBm
    signal avg:    -26 dBm
    tx bitrate:    6.5 MBit/s MCS 0
Station xx:xx:xx:xx:xx:xx (on wlan0)
    inactive time:    96 ms
    rx bytes:    692528
    rx packets:    18938
    tx bytes:    2979324
    tx packets:    2001
    tx retries:    3340
    tx failed:    0
    signal:      -33 dBm
    signal avg:    -33 dBm
    tx bitrate:    54.0 MBit/s
----------------------------------------------------------------------------------------------------------------------------------------
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Baldomero Coll
2011-02-10 16:45:05 UTC
Permalink
Post by Baldomero Coll
Post by Peizhao Hu
Hi all,
I'm also testing the patch Alex and Benoit have released for the
community.
Post by Peizhao Hu
After configuring my IBSS network, the station dump command shows this
[1].
Post by Peizhao Hu
As you can see, although the IBSS network has been created in HT mode,
one
Post by Peizhao Hu
link doesn't seem to support MCS rates.
I would like to know if it is possible to force the MCS scheme. I don't
know
Post by Peizhao Hu
why the "NL80211_TXRATE_MCS" parameter is not defined in the bitrates.c
(iw), or what I have to do to enable it.
I don't think we can force a particular MCS rate in the station side,
we have rate control algorithms to determine the rate.
Do you know where i can find the rate control mechanism that is been used
when i create the HT IBSS network?
I would like to know if such mechanism is able to switch to non-HT rates if
the channel quality is poor.
Post by Baldomero Coll
Post by Peizhao Hu
I also wonder what rate control mechanism is been used in HT mode
(minstrel
Post by Peizhao Hu
HT?), and whether such mechanism switches to non HT rates when the
channel
Post by Peizhao Hu
quality is bad.
I have been using wireshark to capture data transmissions between 2 PCs
configured in IBSS HT mode. What I see in the radiotap header is that no
HT
Post by Peizhao Hu
information is available (protocol, data rate). Do you know how the HT
information can be added to the ratiotap header?
If you are using wireshark, you need to take the latest bleeding edge
that has MCS rate support and also you need latest wireless-testing
that contains MCS rate update to radiotap.
I don't completely understand what do you mean. Can you tell me which
version of wireshark i should use? and where i could find the latest
radiotap version?
Post by Baldomero Coll
Post by Peizhao Hu
Thanks for your help
Baldo
[1]
----------------------------------------------------------------------------------------------------------------------------------------
Post by Peizhao Hu
sudo iw dev wlan0 station dump
Station xx:xx:xx:xx:xx:xx (on wlan0)
inactive time: 8392 ms
rx bytes: 2476336
rx packets: 2149
tx bytes: 433552
tx packets: 1228
tx retries: 564
tx failed: 0
signal: -27 dBm
signal avg: -28 dBm
tx bitrate: 270.0 MBit/s MCS 14 40Mhz short GI
Station xx:xx:xx:xx:xx:xx (on wlan0)
inactive time: 200 ms
rx bytes: 27577
rx packets: 434
tx bytes: 803
tx packets: 8
tx retries: 2
tx failed: 0
signal: -26 dBm
signal avg: -26 dBm
tx bitrate: 6.5 MBit/s MCS 0
Station xx:xx:xx:xx:xx:xx (on wlan0)
inactive time: 96 ms
rx bytes: 692528
rx packets: 18938
tx bytes: 2979324
tx packets: 2001
tx retries: 3340
tx failed: 0
signal: -33 dBm
signal avg: -33 dBm
tx bitrate: 54.0 MBit/s
----------------------------------------------------------------------------------------------------------------------------------------
Post by Peizhao Hu
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Mohammed Shafi
2011-02-11 10:24:32 UTC
Permalink
Post by Baldomero Coll
Post by Mohammed Shafi
Post by Peizhao Hu
Hi all,
I'm also testing the patch Alex and Benoit have released for the community.
After configuring my IBSS network, the station dump command shows this [1].
As you can see, although the IBSS network has been created in HT mode, one
link doesn't seem to support MCS rates.
I would like to know if it is possible to force the MCS scheme. I don't know
why the "NL80211_TXRATE_MCS" parameter is not defined in the bitrates.c
(iw), or what I have to do to enable it.
I don't think we can force a particular MCS rate in the station side,
we have rate control algorithms to determine the rate.
Do you know where i can find the rate control mechanism that is been used
when i create the HT IBSS network?
I would like to know if such mechanism is able to switch to non-HT rates if
the channel quality is poor.
ministrel -HT ?
I think it will(just like ath9k rate control).
Post by Baldomero Coll
Post by Mohammed Shafi
Post by Peizhao Hu
I also wonder what rate control mechanism is been used in HT mode (minstrel
HT?), and whether such mechanism switches to non HT rates when the channel
quality is bad.
Yes it should.
Post by Baldomero Coll
Post by Mohammed Shafi
Post by Peizhao Hu
I have been using wireshark to capture data transmissions between 2 PCs
configured in IBSS HT mode. What I see in the radiotap header is that no HT
information is available (protocol, data rate). Do you know how the HT
information can be added to the ratiotap header?
If you are using wireshark, you need to take the latest bleeding edge
that has MCS rate support and also you need latest wireless-testing
that contains MCS rate update to radiotap.
I don't completely understand what do you mean. Can you tell me which
version of wireshark i should use? and where i could find the latest
radiotap version?
latest development branch (not even 1.5.0 version).
http://www.wireshark.org/develop.html
Post by Baldomero Coll
Post by Mohammed Shafi
Post by Peizhao Hu
Thanks for your help
Baldo
[1]
----------------------------------------------------------------------------------------------------------------------------------------
sudo iw dev wlan0 station dump
Station xx:xx:xx:xx:xx:xx (on wlan0)
    inactive time:    8392 ms
    rx bytes:    2476336
    rx packets:    2149
    tx bytes:    433552
    tx packets:    1228
    tx retries:    564
    tx failed:    0
    signal:      -27 dBm
    signal avg:    -28 dBm
    tx bitrate:    270.0 MBit/s MCS 14 40Mhz short GI
Station xx:xx:xx:xx:xx:xx (on wlan0)
    inactive time:    200 ms
    rx bytes:    27577
    rx packets:    434
    tx bytes:    803
    tx packets:    8
    tx retries:    2
    tx failed:    0
    signal:      -26 dBm
    signal avg:    -26 dBm
    tx bitrate:    6.5 MBit/s MCS 0
Station xx:xx:xx:xx:xx:xx (on wlan0)
    inactive time:    96 ms
    rx bytes:    692528
    rx packets:    18938
    tx bytes:    2979324
    tx packets:    2001
    tx retries:    3340
    tx failed:    0
    signal:      -33 dBm
    signal avg:    -33 dBm
    tx bitrate:    54.0 MBit/s
----------------------------------------------------------------------------------------------------------------------------------------
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Baldomero Coll
2011-02-15 09:05:09 UTC
Permalink
Hello,

Can you please tell me how do you select one o two antennas?

I'm using a similar setting than you:
Linux kernel: 2.6.32-28-generic-pae.
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch suggested by
Alex.
Radio card: Ubiquiti SR71x

Thanks in advance,
Baldomero
Post by Peizhao Hu
Hi all,
I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
Sagar Bijwe
2011-02-15 09:10:28 UTC
Permalink
I would also like to know that...
-Sagar
Post by Baldomero Coll
Hello,
Can you please tell me how do you select one o two antennas?
Linux kernel: 2.6.32-28-generic-pae.
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch suggested
by Alex.
Radio card: Ubiquiti SR71x
Thanks in advance,
Baldomero
Post by Peizhao Hu
Hi all,
I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Peizhao Hu
2011-02-16 00:10:40 UTC
Permalink
Hi,

I meant the number of actual antenna was used. I am not sure whether it
has the same antenna setting in /sys/kernel/debug/ieee80211 like ath5k.

regards;

Peizhao
Post by Sagar Bijwe
I would also like to know that...
-Sagar
Hello,
Can you please tell me how do you select one o two antennas?
Linux kernel: 2.6.32-28-generic-pae.
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiquiti SR71x
Thanks in advance,
Baldomero
Hi all,
I would like to confirm my findings. My test platform
configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Mohammed Shafi
2011-02-16 12:16:23 UTC
Permalink
Post by Baldomero Coll
Hello,
Can you please tell me how do you select one o two antennas?
I don't know why you should do that. I guess changing the tx/rx
chainmask will do after it was read from eeprom.
Post by Baldomero Coll
Linux kernel: 2.6.32-28-generic-pae.
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch suggested by
Alex.
Radio card: Ubiquiti SR71x
Thanks in advance,
Baldomero
Post by Peizhao Hu
Hi all,
I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Baldomero Coll
2011-02-16 14:47:05 UTC
Permalink
I'm not sure, but I've read somewhere that by default the two antennas are
used.
It is true that I'm not interested in selecting the number of antennas, what
I really want is that the MIMO capability is exploited if I'm using 802.11n
HT IBSS operation mode.
Can someone confirm that by default the two antennas (spatial diversity) are
being used when we create the HT IBSS network?
Post by Mohammed Shafi
Post by Baldomero Coll
Hello,
Can you please tell me how do you select one o two antennas?
I don't know why you should do that. I guess changing the tx/rx
chainmask will do after it was read from eeprom.
Post by Baldomero Coll
Linux kernel: 2.6.32-28-generic-pae.
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch suggested
by
Post by Baldomero Coll
Alex.
Radio card: Ubiquiti SR71x
Thanks in advance,
Baldomero
Post by Peizhao Hu
Hi all,
I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Brian Prodoehl
2011-02-16 16:46:15 UTC
Permalink
Post by Baldomero Coll
I'm not sure, but I've read somewhere that by default the two antennas are
used.
It is true that I'm not interested in selecting the number of antennas, what
I really want is that the MIMO capability is exploited if I'm using 802.11n
HT IBSS operation mode.
Can someone confirm that by default the two antennas (spatial diversity) are
being used when we create the HT IBSS network?
Post by Mohammed Shafi
Post by Baldomero Coll
Hello,
Can you please tell me how do you select one o two antennas?
I don't know why you should do that. I guess changing the tx/rx
chainmask will do after it was read from eeprom.
Post by Baldomero Coll
Linux kernel: 2.6.32-28-generic-pae.
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch suggested by
Alex.
Radio card: Ubiquiti SR71x
Thanks in advance,
Baldomero
Post by Peizhao Hu
Hi all,
I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
If you can operate beyond MCS7 (MCS8-15), then you are using two
spatial streams over two antennas. I have seen this with HT IBSS and
ath9k.

-Brian
Baldomero Coll
2011-02-17 09:05:13 UTC
Permalink
Post by Jouni Malinen
Post by Baldomero Coll
I'm not sure, but I've read somewhere that by default the two antennas
are
Post by Baldomero Coll
used.
It is true that I'm not interested in selecting the number of antennas,
what
Post by Baldomero Coll
I really want is that the MIMO capability is exploited if I'm using
802.11n
Post by Baldomero Coll
HT IBSS operation mode.
Can someone confirm that by default the two antennas (spatial diversity)
are
Post by Baldomero Coll
being used when we create the HT IBSS network?
Post by Mohammed Shafi
Post by Baldomero Coll
Hello,
Can you please tell me how do you select one o two antennas?
I don't know why you should do that. I guess changing the tx/rx
chainmask will do after it was read from eeprom.
Post by Baldomero Coll
Linux kernel: 2.6.32-28-generic-pae.
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch suggested by
Alex.
Radio card: Ubiquiti SR71x
Thanks in advance,
Baldomero
Post by Peizhao Hu
Hi all,
I would like to confirm my findings. My test platform configurations are
follow.
Board: pcengine alix3d2
Linux kernel: 2.6.35 from linux-wireless git
Driver: compat-wireless-2011-01-17 and iw-0.9.21 with the patch
suggested by Alex.
Radio card: Ubiqiti SR71a on channel 36 with HT40+
Measurement tool and settings: Iperf, UDP, 100Mb offered load
Recored throughput: 50-54Mbps (one antenna); 78-80Mbps (two or three
antennas).
If you can operate beyond MCS7 (MCS8-15), then you are using two
spatial streams over two antennas. I have seen this with HT IBSS and
ath9k.
-Brian
I get the following results, so based on your comments I suppose the two
antennas are being used:

It is a scp transmission between two laptops joined at the same HT IBSS
network.
(The first parameter shows the file transmitted percentage, the second is
the amount of MB transmitted and the third one the estimated time of arrival
(or expected time to finish the transmission).

5% 51MB 10.9MB/s 01:22 ETA signal: -33 dBm tx bitrate: 300.0 MBit/s
MCS 15 40Mhz short GI
10% 100MB 12.3MB/s 01:09 ETA signal: -33 dBm tx bitrate: 300.0
MBit/s MCS 15 40Mhz short GI

This results belong to one of the laptops, the other one I don't know why is
not using HT rates (although both of them are similar).
Are you having HT rates bidirectionally? Sometimes it happens to me that one
of the HT laptops shows legacy 2.4 or 5 rates, and there is no way of
upgrading its data rates. I'm going to dig into this right now.

Baldo
Adrian Chadd
2011-02-16 17:03:14 UTC
Permalink
Hi,
Baldomero Coll
2011-02-17 09:27:55 UTC
Permalink
Thank you for your illustrative answer adrian.
I've sent an email today to the mailing list showing my performance in HT
IBSS mode.

What happens now is that my ad-hoc link don't support HT in both directions.
One of the stations seems to stay in legacy IBSS mode, although sometimes it
upgrades to HT mode.

I suppose it is due to the link adaptation mechanism (Minstrel HT?), but I
would also investigate if the station that stay in legacy IBSS mode is the
one who creates the HT network, the one who joins the HT network, or if it
don't really depend on that.
Hi,
Adrian Chadd
2011-02-17 16:11:43 UTC
Permalink
I suggest watching minstral "walk its walk" as it tries different rates and
see what the pass/fail/retry/times are for each of the rates.

Take a look for rc_stat in the /sys filesystem, somewhere under ieee80211 or
something similar; I don't have a Linux box handy to check.


Adrian
Post by Baldomero Coll
Thank you for your illustrative answer adrian.
I've sent an email today to the mailing list showing my performance in HT
IBSS mode.
What happens now is that my ad-hoc link don't support HT in both
directions. One of the stations seems to stay in legacy IBSS mode, although
sometimes it upgrades to HT mode.
I suppose it is due to the link adaptation mechanism (Minstrel HT?), but I
would also investigate if the station that stay in legacy IBSS mode is the
one who creates the HT network, the one who joins the HT network, or if it
don't really depend on that.
Hi,
Baldomero Coll
2011-02-18 18:49:07 UTC
Permalink
Has anyone of you been able to capture the frames generated withing the HT
IBSS network?

I'm using Wireshark wireshark 1.2.7 and kismet 2011-01, and there is no way
of captruring the data frames, although the beacons are well captured.
Xianghua Xiao
2011-03-07 21:49:02 UTC
Permalink
Will there be a new version of IBSS/HT patch in the near future?
Thanks,
Post by Baldomero Coll
Has anyone of you been able to capture the frames generated withing the HT
IBSS network?
I'm using Wireshark wireshark 1.2.7 and kismet 2011-01, and there is no way
of captruring the data frames, although the beacons are well captured.
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Mohammed Shafi
2011-03-08 05:25:03 UTC
Permalink
Post by Baldomero Coll
Has anyone of you been able to capture the frames generated withing the HT
IBSS network?
I'm using Wireshark wireshark 1.2.7 and kismet 2011-01, and there is no way
of captruring the data frames, although the beacons are well captured.
make use of this command for the sniffer, based on your ad-hoc
networks channel and mode
iw dev <devname> set channel <channel> [HT20|HT40+|HT40-]
Post by Baldomero Coll
_______________________________________________
ath9k-devel mailing list
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Baldomero Coll
2011-03-10 08:41:58 UTC
Permalink
Post by Xianghua Xiao
Will there be a new version of IBSS/HT patch in the near future?
Thanks,
Can you tell us something in advance about the new features thal will
appear in the IBSS/HT patch that you are announcing?

Thanks, baldo

Continue reading on narkive:
Loading...