On Sun, 17 Jun 2001, Fernando Ronci wrote:

> Hi,
>
> Is there a simple and easy way to measure the Signal Strength from a
> customer´s building to our bridge/AP without having to do a complete client
> PCI card, cable, accessories and 6 dBi patch wall mount antenna install?

Yes, there is a simple way. Either buy or rent something of a spectrum
analyser. Think I'm joking? Read on..

If you, at your customers site, wish to make a quick, and swift assesment
of their signal conditions, a simple PCMCIA card & antenna are probably
the more missleading and less informative way to do.

The criteria I think you should be concerned with are not going to be
easily shown to you by linking a card to some antenna and wafting it in
the air. Those criteria are:

-Ambient noise floor in the 2400 to 2480 Mhz band

-Impulse noise floor in the 2400 to 2480 Mhz band

and..

-How hot your target signal is, and how much stronger than the noise floor
you can measure it to be.

Now, to gain reliable information about these criteria, you could get a
(sorry, this won't be pleasent to some) FreeBSD laptop, some PCMCIA card
that that generates or keeps a "signal cache" and leave it at that. You
can walk arround (or remain still ;) and do a:

wicontrol -i wi0 -C

Which will generate:

[1/10]: 00:02:a5:0e:57:01, 172.16.4.45, sig: -87, noise: -96, qual: 9
[2/10]: 00:50:8b:48:89:28, 172.16.4.16, sig: -88, noise: -91, qual: 3
[3/10]: 00:90:27:43:27:2f, 172.16.1.50, sig: -88, noise: -90, qual: 2
[4/10]: 00:08:c7:29:a5:77, 172.16.4.120, sig: -89, noise: -94, qual: 5
.......

What this table shows is somewhat inaccurate; each MAC listed is _NOT_
another wireless client, necessarily. It's merely the signal conditions
durring the reception of a frame FROM the listed MAC address. This
reception is always from the Access Point in BSS/ESS opperation. So, in
summary, you can get a good idea of your signal to noise ratio from this
data, but keep in mind that each MAC you see will only repesent the signal
quality data at the time of the last frame recieved from that station via
the AP!

As a general rule, when doing a site survey, it is ill-adviseable to use
an antenna that exhibits gain. Why is that? Well, we want to know the REAL
signal strength at the VERY POINT you're measuring it. We don't want to
skew the results by using an antenna which gives us 6, or 12, or 24 dB of
gain. Something that exhibits unity gain would be appropriate. Hence, a
PCMCIA laptop card is ideal for the cheap route.

Getting back to a Spec-An. The use of this device would be great for
you; you'd get far more usefull data about the RF happenings and goings-on
in your target band. With it, you would need to setup some sort of
continous packet stream to watch RF from. Something like a:

 ping -f <some.target.ip.address>

would create a nice flurry of activity, although you will want to do this
when other hosts are not trying to use the medium very heavily ;)

Anyway, with traffic a' flying, you would then be off to measure signal
from your radio headend at a clients site. A small 1/4 wave antenna with
near unity gain would be great to start with. Setup the Spec-An to show
you a window of 2400-2480 MHz, and then simply take a few minutes to study
the resulting visual information. See, spec-an's are neet-o like
windows. What you should see are several 'hills' of RF energy above a
random noise floor. The differenve in height of the 'hills' above the
noise floor will be the data you use to understand what sort of signal to
noise ratio you can expect in that area.

If your client device is rated to opperate at 11 mbits with a recieved
signal strength of -80 dB, and your signal is at -71 dB with the noise
floor at -96 dB, you'd probably be in good running for near full 11mbit
transmission. However, if you saw a noise floor of -85 dB, with the same
signal strength, be more concerned. You may then want to try a directional
antenna that exhibits gain. See if you can determine if using that sort of
antenna will help negate the sources of the noise floor, that is, if the
direction you're going to point it is not the general direction of the
noise source!

At any rate, I would sugest either method as an o.k. way to learn more
about the signal characteristics in the desired recieving area. Just
remember, it is important to make measurements with things that will not
effect the measurements; this means unity gain, wide-band antennas, very
nice, low-loss cabling, and accurate test gear. Hell, I have no 'real'
idea that my Lucent or Aironet card is even dB-Calibrated! I'm assuime it
is only because of the data format it yeilds and it sort-of correlates to
other measurements I've taken at the same time.. But anyway, good luck!

-Xam

> We have just started to provide wireless internet access to bussineses in
> our area (100% with Cisco products) and are looking for a practical and
> quick way to measure the SS, something like a signal strength meter. This
> would save us a lot of time, especially when the customer is far enough,
> near the maximum allowed range. It´s a waste of time to do a complete
> install and then say to the customer, "We´re sorry, no signal here".
>
> The embedded antennas on PCMCIA cards have a much shorter range than patch
> ones, so going with a notebook doesn´t help.
>
> Thank you,
>
> Fernando Ronci
> E-mail: fernandowaycom.com.ar
>
> _______________________________________________
> Aironet mailing list - Aironetcsl.cse.ucsc.edu
> http://csl.cse.ucsc.edu/mailman/listinfo/aironet
>

_______________________________________________
Aironet mailing list - Aironetcsl.cse.ucsc.edu
http://csl.cse.ucsc.edu/mailman/listinfo/aironet

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]