What is the reasonable sampling rate for the wireless signal level?
 

If I want to compare the signal level for 2 wireless networks, what is the reasonable sampling rate during comparison?

What I mean here is, I get the signal level from iwconfig command, I want to average them over a period to minimize the fluctuation of the signal level. However what is the reasonable period? 5 seconds? 30 seconds? 1 minute? What is the reasonable rate?

Thanks in advance.

Generally if I'm trying to find the average or mean of anything, the bigger the sample, the more accurate the result is going to be. So it seems to me that sampling for as long as you feel is practical would be the thing to do. If I had the time for it, I'd sample over an hour or two. Also, depending on the physical location of the equipment, rates may vary by time of day, or even with the weather. So I'd try to sample many times, and for quite a while each time. Personally I don't feel that a few seconds' sampling will deliver meaningful results. I am ready to stand corrected, though! :)

If your machine and access points are stationary, I would not expect the signal levels to fluctuate much. Sampling every minute for a few hours would probably be adequate. It would help to know what you actually want to figure out.

If the signal path is partially outdoors, weather conditions (rain, snow) can have a big impact on signal strength at these frequencies. Also, if there are trees in the path, this can cause major fluctuations.

My router's signal varies from roughly 55% to 78%. It's sitting near an outside brick wall and double-glazed window. I usually access the router from my living room about twenty-five feet away with two more interior walls in between. I wonder if radio-frequency interference might cause the lowering of signal level at least part of the time. Weather (and we have the extremes in KC) seems to have no effect. But I live in a densely populated area close to tall buildings, heavy traffic. (Off-topic, but worth mentioning: my digital broadcast TV reception fails at almost exactly the same time every evening, then is OK again after a half-hour or so.)

But I think interference from other radio-frequency devices won't lower the signal level but increase the noise level. Lowering of signal level is mainly caused by blocking of building or object. (Am I wrong?)

Quite interesting, do you think what is the reason causing this happen? Will it be insufficient signal level? Low SNR? Insufficient bandwidth? Or others reasons? What will be the main factor decides the quality of IPTV?

Christyyim, I think you are right about RF interference increasing noise level, and not in itself lowering signal level. And definitely placement of a router near a wall, especially an exterior one, or near a window, does reduce signal level. Funny though the router never moves, I hardly ever move from the couch (laptop at hand), I have no cat, no dog... And yet the signal varies like that.

I read a recommendation to place a router as high as possible and preferably in the center of a room. Hmm. I'd say not too many people have those options, especially the second one!