Erratic WiFi 2.4 GHz channel spikes, what gives?

2

2

Sorry guys, first a gripe about my neighbor's WiFi access point (it is related): they totally hog the center nine 2.4 GHz channels (3-11), centered right at 7! I know the outer regions of the signal don't make as much of a difference, and technically they're running channels 5 & 9. Anyway, their signal is clearly interfering with mine, which is necessarily centered at 3 or 11 to evade their interference. I guess it's somewhat a case of access point envy: they happen to have both a stronger signal and a higher data rate, while occupying twice the band width that I do.

Getting to the point, I've noticed that they tend to sit nice and pretty centered at 7, but they definitely auto-select their channel, and I've noticed that the auto-selection algorithm tends to shift towards the higher channels; hence I decided to pick channel 3, and I don't get so many intermittent lag spikes any more.

Anyway, the thing that weirded me out was the reason they have to auto-select sometimes: unexplained, powerful (talking order of 0dB here), giant spikes of 2.4 GHz activity in consistent regions of the spectrum. I don't think it's just noise, since my wireless monitoring software is registering a MAC address, a manufacturer, and usually a fairly coherent ascii name... and it seems to be a fairly well-confined signal. But these signals are fairly common, and they do some weird stuff to my signal.

So my question is what are these signals? Where are they coming from? Where are they going? Why are they so ridiculously strong? Why don't they ever last very long?

Here's an inSSIDer screenshot I took, for your perusal. I am labeled with "me", my greedy neighbor labeled with "neighbor", and the 2 quasar signals are labeled with "WTF?".

Pic related: it's me with my bi....g problem

machine yearning

Posted 2012-04-04T08:05:57.677

Reputation: 123

Those who see "spikes" - are you attached to a network while scanning? Because, the sniffer dongle will end up being quite close the antenna of the laptop you're using... :-D But that graph looks like they are at >0dBm: I suspect a software bug. – conny – 2014-12-04T16:50:41.007

1Interesting. What software are you using for this? What sort of area do you live in? Have you tried moving around (assuming you're using a laptop) to see how localised the interference is? Perhaps try a mobile phone app. – James P – 2012-04-04T11:36:07.667

@James That's inSSIDer from MetaGeek. – Spiff – 2012-04-04T17:26:39.820

@Francis What is your software reporting as the manufacturer of the equipment making the super strong networks on 6 and 11? Do you have a way to capture a beacon or probe response frame (with Wireshark or something) from those networks and post it somewhere (like cloudshark.com) so we can look at the Information Elements in the frames? That might provide more clues. This might be a high power outdoor point-to-point link. Tell us the name(s) of the two high-power networks too. – Spiff – 2012-04-04T17:36:49.497

@Spiff : Thanks for the responses, I am indeed using inSSIDer; I'll sniff out some frames at some point when I get a moment, and upload and update this post accordingly. – machine yearning – 2012-04-05T01:25:22.883

1I ran inSSIDer on my own site and can see the exact same spikes from one of my wifi-routers. It's a Linksys WRT610N v2 running DD-WRT. So either this is by design, or my router has the same fault... – Mattias Åslund – 2012-04-05T18:24:52.127

It has been some days, any update on the issue? – Daniel Andersson – 2012-04-10T10:45:41.163

1I wish OP hadn't disappeared. This was an interesting question. – Spiff – 2012-06-14T07:21:55.843

Answers

2

I see the same 0dB spikes when running inSSIDer 2.1. Because they don't seem to cause a disconnect to APs on the same channel as the spike, it seems possible to me that they are not really 0 dB "spikes", but really signal drops.

That is, if some nearby APs are occasionally blasting out signal at orders of magnitude higher amplitude, then I would expect severe issues with my WAN. Because I don't see any issues associated with these "spikes", I think it possible that they are a bug in the way the inSSIDer software plots signal. If signal below some arbitrary lower limit, for example -100 dB, is rounded off to nothing, that "nothing" might be encoded as "0". Unfortunately "0" dB is a very high signal not a very low signal. So when plotted we see your "quasar" effect.

Phillip

Posted 2012-04-04T08:05:57.677

Reputation: 36

Aha! This explanation fits so well, I'm almost certain it solves the mystery. – Spiff – 2012-06-24T18:40:59.030

1

Microwave ovens completely kill my connection. I don't know how they would be represented in your program, but I would try to do some controlled tests of that, to see the characteristics of microwave oven signals.

This corresponds well with your assertion that the signals don't last very long. It does not correspond well with that you say your analysis program seems to register it as a legitimate WLAN signal.

If nothing else works, and believe me, this should definitely be the last option (after e.g. asking strangers on the internet), you can try actually knocking on their door when it happens and see if they know if they did something special.

Daniel Andersson

Posted 2012-04-04T08:05:57.677

Reputation: 20 465

Asked: 2012-04-04T08:05:57.677

Viewed: 5 080 times

Active: 2016-02-04T01:35:38.997