Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 1 Dec 2011 01:26:43 -0500
From:      "Viet-Ha Pham" <hapvbk@yahoo.co.uk>
To:        "Adrian Chadd" <adrian@freebsd.org>
Cc:        freebsd-net@freebsd.org, Bernhard Schmidt <bschmidt@freebsd.org>
Subject:   Re: Wifi throughput test issue (FreeBSD 8.2)
Message-ID:  <CB7EB42661084A168C508C6A2AC07BF2@PC>
In-Reply-To: <CAJ-Vmo=B3H3_Wefhq4fKQQ7LiYF27oQ06osm4QzuWKT2a3wDxA@mail.gmail.com>
References:  <69C75F6ECC9646C79374299EDF0E4304@PC><CAAgh0_Ym0Np4Py-aaxH=bEJcSJPAKcx2CFECMtgSD=%2B%2BVhUvLQ@mail.gmail.com><5939430EC6AF43B79CB9BE99940E9B91@PC> <CAJ-Vmo=B3H3_Wefhq4fKQQ7LiYF27oQ06osm4QzuWKT2a3wDxA@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Hi Adrian,

Thanks for your reply. I will try the sysctl today.

Let me clarify what we are doing and our objectives.

We want to investigate the relationship between SNR and throughput at each 
transmission rate. So we fix the rate, vary the input noise and measure the 
throughput at that noise level for that rate. We repeat this procedure for 
all transmission rates.

The issue that I described happened even with no-noise scenario.

For the samplerate source code, we have read it since we are working also on 
rate adaptation algorithm.

I'll update the dmesg result.

Thanks.


HAPV


-----Original Message----- 
From: Adrian Chadd
Sent: Wednesday, November 30, 2011 11:06 AM
To: Viet-Ha Pham
Cc: freebsd-net@freebsd.org ; Bernhard Schmidt
Subject: Re: Wifi throughput test issue (FreeBSD 8.2)

Hi,

Hm. Selecting a static rate would mean that you'd end up populating
the rate control statistics with more "wins" for that particular TX
rate, so when you go back to "any", it will choose 54mbit more often.

I suggest looking at the code in
sys/dev/ath/ath_rate/sample/sample.[ch] and see what's going on.  You
can investigate the current statistics by "sysctl
dev.ath.X.sample_stats=1" and then check dmesg.

It sounds like your noise is causing some tx failures to occur at
higher rates, which makes sample back off very quickly. It may not
then sample the higher rates often enough if it sees failures early
on.


Adrian
_______________________________________________
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to "freebsd-net-unsubscribe@freebsd.org" 




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CB7EB42661084A168C508C6A2AC07BF2>