Re: [Wifiadmin-users] Noise and Silence values

[Christopher N. <ch...@ne...>]

On Fri, 2005-06-24 at 02:53 +0300, Thanos Panousis wrote:
>  From what I remember, I substracted 256 because our card readings were 
> positive integers, and since I figured the results should be in dBs, I 
> had to make them negative...So that's what I did. Its just as Rob said, 
> its right to do it in some cards, and it is wrong on others.

I downloaded the wireless tools source last night and poked through it.
There was a huge comment about confusions on how to read the wireless
signal values using some 8 bit integers.

Our boxen use Prism cards and report these values.

sunny1# cat 00\:06\:25\:ac\:83\:db | grep last_rx:
last_rx: silence=-91 dBm signal=-40 dBm rate=2 Mbps

> Anyway I think it provides some proportional taste of the signal levels 
> your clients are getting, either negative or positive. Any insights on a 
> better way are welcome.

I will write a quick fix to check if it is positive or not.  The
wireless tools code (v27 iwlib.c line 1301) checks the range, which we
don't have but it does say this in a comment.

   * Relative/percent values are always encoded unsigned, between 0 and 255.
   * Absolute/dBm values are always encoded negative, between -255 and 0.

So I will put a check in to see if the negation should take place.
Should a text unit text be applied so that the user knows if they are
looking at Relative or Absolute values?

Cheers,
Chris
-- 
Wireless Group,
McMaster University

finger.localdomain
11:13:37 up 4 days, 16 min, 1 user, load average: 0.07, 0.05, 0.05