hlexa@hotmail.com (Axel Hammerschmidt) hath wroth:

>Jeff Liebermann <jeffl@comix.santa-cruz.ca.us> wrote:
>> Yep. The problem is that interference is usually quite variable. It
>> comes and goes. There are very few sources of CONTINUOUS
>> interference. See list at:
>> http://wireless.wikia.com/wiki/Wi-Fi#Interference

>Nothing about spotty reception in the Wiki :-(

True. The Wiki/FAQ is built on user contributions. Contributions and
corrections are always welcome. There are plenty of topics that have
not been mentioned or are incomplete.

The issue I raised is about the characteristics of different types of
interference that could possibly cause the OP's "dead spot" or "spotty
reception". I suggested that this would imply continuous
interference. Since interference is *YOUR* theory, and allegedly
responsible for possibly causing the OP's problems, could I trouble
you to suggest what manner of device could possibly cause continuous
interference? Also, some remedial measures as I find it odd that you
would suggest tweaking timing parameters rather than identifying the
cause and either eliminating or avoiding it.

>Installing a reflector on the wireless router was your suggestion

Correct. "spotty reception" implies that there is some reception in
the bedroom. An increase in signal strength should improve the
situation. If the house is long and narrow, redirecting some of the
RF in the direction of the bedroom will be a big help. If the
wireless router is at one end of the house, and the bedroom at the
other, this would be ideal for an antenna reflector arrangement.

>> Reading a bit between the lines, it would appear that the entire
>> bedroom is a dead spot, which would imply that it's not a reflections
>> issue. If you've ever had to deal with a reflections issue, you
>> probably would have noticed that the effect varies enormously with
>> location and position. Move a few cubits one way or other and the
>> signal could easily be stable and usable. My guess(tm) is that the OP
>> would have tried a few possible positions before asking for help. It's
>> probably not reflections.
>
>Spotty reception in the bedroom, which an entire dead spot, and the OP's
>already moved the walls around...?

Sorry. I wasn't specific enough for you. I'm suggesting that the OP
move the wireless device (laptop or desktop) to various locations in
the bedroom to avoid any possible null and reflection problems. I
don't think that moving walls will be cost effective.

>The facts are, you don't know.

True. But it's so much fun to speculate or guess(tm).

Usually, the OP supplies additional detail at this point, but he's
apparently given up and considering power line networking. That will
work and is easier if he can find a common power line segment.

So, what does one do for dead spots? In order starting with my
favorite:
1. Bigger or better antennas.
2. Additional access point with CAT5 wire backhaul to main router.
3. Better technology (MIMO).
4. Powerline, phone line, or CATV bridging.
5. WDS bridge.
6. Repeater.
7. Bi-directional amplifier.
8. Dead Spot Remover:
http://www.doityourself.com/stry/blood
Out damn spot, out I say...

>But flow control is much easier, and should it fix the spotty reception
>in the bedroom, the PO will have saved himself the trouble.

As I previously mentioned... I advocate playing with every single last
lousy setting in the router just to see what happens. Who knows, one
might find something interesting, useful, or entertaining. One might
also learn something. In every router that I have owned, I make it a
point of understanding and testing all the settings. No big deal on a
commodity router, but a major challenge with a WRT54G and DD-WRT
firmware. It's all part of Learn By Destroying(tm). In the case of
the current "dead spot" problem, I predict that adding flow control
will not do anything useful.

>And the walls...

Use a shofar. ...and the walls came tumbling down.

>> Incidentally, the paper was from 1998 and deals exclusively with early
>> analog 802.11 chipsets. I think (not sure) that modern all digital
>> 802.11g chipsets don't use carrier detection for interference
>> detection but only use data validity detection or preamble S/N
>> measurement. If the data arrives corrupted, it's assume that
>> something else interfered with the transmission. RSSI detection is
>> just too slow and subject to blocking and overload. I gotta do some
>> digging and reading to be sure.

>Well, the author does note, that RTS/CTS is [...] less efficient with
>non-CSMA interferences, implying that it can be used to improve spotty
>reception, when caused by interference.

I don't see where you get that implication, but I can guess(tm).
Usually, interference reduction is measured using the signal to noise
ratio. That was fairly straight forward with analog receivers. The
signal strength versus the remaining noise level between packets.
However, with the introduction of all digital receivers, there was a
change to measuring the SNR using the error rate. 100% reception
success, with no corrupted or trashed packets, was deemed perfect SNR.
If corrupted packets were decoded, the SNR value was reduced. The
correlation is marginal with the previous analog methods but good
enough for site surveys and antenna aiming.

In this SNR case, you are correct. Increasing the probability of a
packet arriving without corruption, by reducing its size (i.e.
fragmentation), or preventing transmission when the channel is "noisy"
(RTS/CTS flow), would improve the SNR. The only problem is that both
will slow down the thruput considerably.

Incidentally, CTS/RTS is the common term for modem flow control.
RTS/CTS is the common term for wireless. I guess I'll have to change
over to using RTS/CTS. Sigh.

--
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