btysgtmajor@gmail.com hath wroth:
> I'm decent with technology but my wireless knowledge doesn't go much
>further than basic networking and I can extrapolate just a little bit
>more.
I little knowledge is dangerous. More is often fatal.
>Regardless, at my work, I've setup a wireless AP (it actually
>sits a few feet away from my desk, more or less) for the building
>(it's about the size of a mechanic's bay or two). I selected its
>location because it would seem to cover the most distance and looks to
>be in an area where other wireless APs shouldn't affect things too
>much.
Is this a theoretical question? If not, kindly disclose the access
point maker and model, the topography in whatever ft, meters, metres,
cubits, or whatever, and some clue as to the environment (walls,
sources of interference, expected coverage, etc).
> I do know there are three Cisco APs that I can detect through my
>laptop (I know they're spread over channels 1, 6 and 11, which doesn't
>bode well for me), as well as a few other commerical/residential ones
>that seem to come and go.
Those are just the one's that you can see. If you happen to be using
a sniffer that detects access points and clients, such as Kismet
running on Linux, you'll probably find far more systems. Also, just
because a given AP is on some channel today, doesn't mean they will
stay there. There's "automatic channel selection" which allegedly
selects the least polluted channel, and moves the entire WLAN over to
a new channel. In addition, other WLAN systems are not the only
source of interference. See the FAQ at:
<http://wireless.wikia.com/wiki/Wi-Fi#Interference>
I think municipal wireless is currently the most popular method of
generating interference.
>Now, even when I'm right by my AP, I can
>sometimes connect, but there are other times I just can't.
Stop, look, and think. There are two possible causes, interference
and something else. Let's take interference out of the picture. Drag
your unspecified model access point and a portable wireless client
(also known as a laptop) to a dungeon. Lacking an RF shielded
anechoic test chamber, a dungeon should suffice. No need for an
internet connection in the dungeon, but it would be interesting. See
how things work in the dungeon, where presumably no stray RF signals
and interference from other WLAN's can enter. Use the built in web
server in the AP for simple testing.
If you want to do a more better test, use a benchmarking program such
as IPerf:
<http://dast.nlanr.net/Projects/Iperf/>
Setup a 2nd computer as a server, in the dungeon, and run:
iperf -s
On your wireless laptop, run:
iperf -c ip_address_of_server
You should see some benchmark results. Also try it without wireless,
using a direct CAT5 connection to the server. You should get:
Type-o-connect Thruput
100baseTX-FDX CAT5 80-90 Mbits/sec
10baseTX-FDX CAT5 8- 9 Mbits/sec
802.11g 54Mbits/sec 22-26 Mbits/sec
802.11b 11Mbits/sec 3- 6 Mbits/sec
If performance is good in the dungeon, and the symptoms do NOT follow
the hardware, then you probably have an interference problem.
However, if you have exactly the same problems below ground, as you do
above ground, then it's likely your unspecified model AP, client,
operating system, or client manager are in dire need of a firmware or
driver upgrade.
>It says
>the signal is at full bars, but it just won't let me connect sometimes
>(it'll cycle through the channels but be unable to lock onto
>anything). It's the same thing, no matter which channel I use. And
>yes, it's an 802.11g AP.
I see that you went through considerable effort to use the term "it"
instead of specifying what is actually on your screen. Is this some
type of secret project?
>Switching to 802.11a isn't really a viable
>option, either.
Viable not? If you happen to be in the middle of the interference
capital of the world, that may be your only option. A 2nd 802.11a
access point and dual mode MiniPCI clients aren't that expensive.
>This would lead me to believe there's some form of wireless
>interference, in spite of me being where my APs signal strength/
>connection quality should be at its best.
No, you're 86.3% wrong. Interference at the access point end will
show up in the signal to noise ratio figure. The signal will be
strong, but the noise level will climb sufficiently to cause problems.
Also note that if the interference is caused by a WLAN system, the
system usually has to be moving traffic in order to see the
interference. That means that there will probably be long periods
during the evening where there is no interference present. It really
depends on the source of interference. The worst case I've found is
in a high rise glass office building, with a fabulous view of perhaps
100 assorted wireless systems (half of which are SSID=linksys or
default). With such a large number of systems in view, just the
broadcasts and management traffic conspired to generate sufficient
interference to make the system useless. As they say in real estate:
location, location, location.
In addition, your AP search tool only finds other AP's. It will not
show clients. I found one office where I could just barely see the
access point through the window, but where the interference from all
the employees sitting by the large windows was awful. This mess only
showed up using Kismet.
>I've search the internet
>for suggestions but honestly can't really find many definitive
>answers.
That's because there are no definitive answers. The best we can do is
an FAQ or two. See:
<http://wireless.wikia.com/wiki/Wi-Fi>
<http://wireless.wikia.com/wiki/Fast_Fixes_to_Wi-Fi_Problems>
<http://wireless.wikia.com/wiki/Wi-Fi_How_To>
>What can I do? Would installing a stronger antenna help?
I don't think so. Most antennas are aluminium and fiberglass. Making
one out of steel would be stronger, but not necessarily have more gain
or directionality. Anyway, if you can't get a reliable standing next
to the access point, antennas are not going to make any difference.
>Any and all suggestions would be most welcome. I apologize if this
>question seems "newb"-ish, but I'm really in a bind.
Hint:
1. What are you trying to accomplish?
2. What do you have to work with? (Hardware, software, versions,
topography, environment, numbers)
3. Where are you stuck? What have you done and what happened?
You missed badly on #2 and part of #3.
Suggestions:
1. Disclose the pertinent hardware, firmware, and software numbers. I
don't like working in the dark.
2. Verify that you are using a known working client laptop for
testing. Drag it to a coffee shop with wireless to be sure.
3. Update the firmware and client drivers. This is the most likely
culprit.
4. Reset the AP back to defaults. If it's a turbo-enhanced
conglomeration of acronyms, kindly turn off all this crap and go back
to plain old 802.11g at 54Mbits/sec.
5. If the channel selection in the AP is set to auto, change it to
either channel 1, 6, or 11.
6. The best way to troubleshoot is by substitution. If you can
borrow a different model access point and try it in place of your
existing unspecified model access point, and it also acts flaky, then
interference (or a screwed up client computer) is certainly a
possibility.
7. There's also a possibility that you're testing too many things at
once. Start at the DSL or cable modem and plug in a wired computer.
Test for performance. If that works, add in your router and try
again. If that works, add in your wireless access point and connect
via wireless. Etc. At some point, you're going to see the problem.
Whatever you added last, is probably the culprit. It would be rather
embarrassing if this were a bad DSL or cable modem, broken CAT5, or
ISP problem.
I have lots of other suggestions, but I think this will get you going
in the general direction.
--
Jeff Liebermann
jeffl@comix.santa-cruz.ca.us
150 Felker St #D
http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Wireless Interference from Other APs? 
Linear Mode
