[ath9k-devel] varying RSSI values with ath9k_htc drivers for TL-WN721N USB wifi adapter.

[Harshal Vora <harshal@amideeptech.com>]

Hi,

I did some testing with the latest version of the ath9k_htc drivers in 
3.9.0-rc8-wl which already has the patch
https://patchwork.kernel.org/patch/2177921/
....and the results are much better.

Can you point out the AR9285 RX descriptor contents relating to antenna 
configuration?

Regards,


On Friday 26 April 2013 08:08 PM, Adrian Chadd wrote:
> There's some driver work that Felix and others did or are talking
> about, which tidy up the RSSI measurement.
>
> There's also the antenna diversity behaviour; depending upon which
> antenna is used, there's a few dB difference (same as the AR9285.)
>
> I'm sorry, I can't work on it any more than that. I can point out what
> the AR9285 RX descriptor contents are relating to antenna
> configuration; that way someone can add that to ath9k and possibly
> link antenna selection / configuration to the RSSI differences. But I
> can't do the legwork, sorry.
>
>
>
>
> Adrian
>
> On 26 April 2013 06:56, Trevor Cullen <tcc@innovelec.co.uk> wrote:
>> This issue totally destroyed deployment of DNUA-93F (AR9271) USB Wifi in a
>> similar application. Now use Realtek with very stable results.
>>
>> But there is history.
>>
>> See https://patchwork.kernel.org/patch/2177921/ and attached.
>>
>> Trevor
>> -----Original Message-----
>> From: ath9k-devel-bounces at lists.ath9k.org
>> [mailto:ath9k-devel-bounces at lists.ath9k.org] On Behalf Of Harshal Vora
>> Sent: 26 April 2013 11:43
>> To: Adrian Chadd
>> Cc: ath9k-devel at lists.ath9k.org
>> Subject: Re: [ath9k-devel] varying RSSI values with ath9k_htc drivers for
>> TL-WN721N USB wifi adapter.
>>
>> On Friday 26 April 2013 01:39 AM, Adrian Chadd wrote:
>>> Have you actually verified that the output power of those devices is
>>> actually constant?
>>>
>>>
>>>
>>> Adrian
>>>
>>> On 25 April 2013 07:08, Harshal Vora <harshal@amideeptech.com> wrote:
>>>> Hi,
>>>>
>>>> I are trying to achieve approximate location detection (room level
>>>> granularity) based on RSSI.
>>>> I am using TP-LINK TL-WN721N USB wifi adapter which require ath9k_htc
>>>> drivers.
>>>> Operating system: Ubuntu 12:04 LTS.
>>>>
>>>> The RSSI values recorded vary a lot, even when the transmitter
>>>> (iPhone 5 and sony xPeria and other phones) and the receiver are left
>>>> untouched at the same location.
>>>>
>>>> There is a difference of almost 15 to 20 dbm in the readings.
>>>> Sometimes it is in the range of -40 to -50dbm .. sometimes -50 to
>>>> -60dbm
>>>>
>>>> Also sometimes the RSSI values are always between -75 to -85 dbm no
>>>> matter how short or long the distance is between the two devices.
>>>>
>>>> I am using airmon-ng to set the device in monitor mode and tshark to
>> sniff.
>>>> I have tried with the drivers that come with ubuntu 12.04, with
>>>> compat-wireless 3.6.8-1 as well as compat-drivers 3.8.3-2
>>>>
>>>> The todo list on
>>>> http://wireless.kernel.org/en/users/Drivers/ath9k_htc#TODO
>>>> mentions
>>>> Low RSSI issue for UB91/94.
>>>>
>>>> Should this be a concern?
>>>>
>>>>
>>>> Regards,
>>>>
>>>> _______________________________________________
>>>> ath9k-devel mailing list
>>>> ath9k-devel at lists.ath9k.org
>>>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>>
>> Hi,
>>
>> I have a MAC laptop which has inbuilt broadcom BCM 4331 chipset running on
>> b43 drivers.
>>
>> I put the iPhone in charging and it is always unlocked and on the wifi
>> settings page which displays available networks.
>>   From what I understand, iPhone goes in power saver mode only when locked.
>> The readings on bcm4331 are much stable as compared to the wn721n usb
>> adapter.
>> On broadcom, the dbm values differ mostly by 3-4 dbm and sometimes I see a
>> spike with a difference of 7-8dbm.
>> On the other hand the dbm values using ath9k_htc drivers differ by 10-15dbm.
>>
>> I have also tested with a locked iPhone to get similar results.
>>
>> In another setup, I connected one wn721n chipset to hackberry and another to
>> my laptop.
>> Again I took two sets of readings with transmitter as wn721n on hackberry
>> and receiver as bcm4331 in the first run and receiver as wn721n on my mac in
>> the second run.
>> Again similar observaions as the experiment with iPhone as transmitter.
>>
>> How can I measure the output transmission power? I deployed the ath9k_htc
>> drivers in debug mode with value 0xffffffff. But cannot get any relevant
>> information from debugfs.
>> When I do ifconfig, it says power management off.
>> When I run sudo iwconfig "dev" txpower "value", it returns an error.
>>
>> I have taken a lot of different sets of readings for ath9k_htc, and
>> relatively the readings are correct. i.e. the dbm values will change as the
>> distance between the transmitter and receiver changes. But sometimes the dbm
>> values remain in a particular range say -75dbm to -85dbm no matter how near
>> or far the transmitter is from the receiver.
>>
>> Regards,
>> _______________________________________________
>> ath9k-devel mailing list
>> ath9k-devel at lists.ath9k.org
>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel