Hey Chris, List,
The dB value which wmwlmon fetches, is not the 100dBm which you see
when you do an ifconfig on your interface. It's the value which you
get - or at least I get it with my ath - when you use the ifconfig -M
option. For example "ifconfig -M ath0" :
ath0: flags=8863<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST> mtu
1500 lladdr 00:0e:9b:1b:0b:4a
groups: egress
media: IEEE802.11 autoselect (DS11 mode 11b)
status: active
ieee80211: nwid nazgul chan 6 bssid 00:0d:88:87:ad:a8 nwkey
<not displayed> nwid nazgul chan 6 bssid 00:0d:88:87:ad:a8 41dB
^^^^
54M privacy,short_preamble,short_slottime
inet6 fe80::20e:9bff:fe1b:b4a% ath0 prefixlen 64 scopeid 0x2
inet 192.168.1.5 netmask 0xffffff00 broadcast 192.168.1.255
So, there an access point scan is done (SIOCG80211ALLNODES)
returning the singal strength from the access points in dB. That's
exactly what you see in wmwlmon. I played around a bit with my ath,
and that value seems to be pretty exact. The 88dB which you see in
wmwlmon are therefore probably right. For 11Mbit you need at least
10dB. If you have more, it doesn't increase the bandwidth. At my
home I have usualy 44dB, or moving some more towards the access point
63 dB.
Now, I am no wireless guru, but I think you could calculate a dBm value
(SNR Signal to Noise Ratio) to dB, by substracting the minimumg dBm
which a client needs (-96dBm). For example:
-47dBm - (-96dBm) = 49dB
*But* I doubt that the results of 100dBm which you are getting is
right. All people I asked for the dBm displayed in their ifconfig
with an iwi or some other chipsets was 100dBm. So that value seems
not to be right. As you have seen in the ifconfig, my ath even don't
returns the dBm.
I hope that was a little explanation. Maybe reyk@ could give some more
details on that topic :)
> All the "LEDs" are dark on an unconfigured iwi. That's what I'd
> expect.
Ok, the LEDs where dark on your unconfigured iwi or is this a "wish"?
Regards,
Marcus
On Mon, 17 Oct 2005 14:37:11 -0600 Chris Kuethe
<[EMAIL PROTECTED]> wrote:
> On 17/10/05, Marcus Glocker <[EMAIL PROTECTED]> wrote:
> > Hi Chris, List,
> >
> > Thanks for pointing out the problem. Your patch is not exactly
> > right, as you are passing now 2 times the signal value to
> > draw_signal. But we need the speed (speed) and signal strength (r)
> > to calculate the bar. The problem is if we can't catch the speed
> > for a moment, the speed pointer contains NULL, and if NULL is
> > passed to atoi(3) it cores.
> >
> > I fixed it now in version 0.2. Attached the new port.
>
> OK, this time it doesn't core.
>
> All the "LEDs" are dark on an unconfigured iwi. That's what I'd
> expect.
>
> Ifconfig'ing iwi0 up, there seems to be a discrepancy between what
> ifconfig says the signal strength is "100dBm" and what wmwlmon says it
> is "88dB". I'm about 8ft away from the AP, and traffic is flowing
> across it.
>
> iwi0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> lladdr 00:12:f0:c4:56:97
> media: IEEE802.11 autoselect (DS1)
> status: active
> ieee80211: nwid linksys chan 6 bssid 00:06:25:61:1d:ea 100dBm
> inet6 fe80::212:f0ff:fec4:5697%iwi0 prefixlen 64 scopeid 0x2
> inet 10.0.0.66 netmask 0xffffff00 broadcast 10.0.0.255
>
> No errors in dmesg from the card locking up though. I'm definately
> looking forward to getting rid of the iwi in favor of an ath...
>
> --
> GDB has a 'break' feature; why doesn't it have 'fix' too?
--
Marcus Glocker, [EMAIL PROTECTED], http://www.nazgul.ch -----------------
