MikroTik

Hi,

I'd like your help with understanding the following WiFi issue concerning my RB4011iGS+5HacQ2HnD with a Atheros AR9300 2.4 GHz chip.

How it started

The WiFi connection between a laptop and the router was very bad over 2.4 GHz. Pings of around 1000ms even with the laptop 1m away from the router. Solved by using a different channel.

Investigations

I started a spectral scan and noticed a very strong signal in part of the spectrum.
Signal persists even after moving the router to a different room
Tilting the antennas left reveils a second strong signal
Tilting the antennas right both signals almost disappear
Snooper shows the following
Question

Could this be a very strong signal from the surroundings or is the router broken?

Notes: I've repeated the measurements on multiple occasions and the results were the same, day or night. Personal appliances where plugged out during measurements (except router and laptop connected by wire to the router).

What device is making those spectral scans? So what happens if you turn off the device during the scan? Also, isn't this just the router working at the set frequency?

The router RB4011iGS+5HacQ2HnD is making the spectral scans. From what I can tell, it does not emit while doing that.

It seems like an external signal. When you compared spectral scan results in different rooms the signal was there, but a bit weaker, but that also reflects on almost the whole of the spectrum. Adjusting the antenna position did make that signal bit weaker, along with the second signal, but that also goes for most of the other signals.
Likely there is some device that creates interference, we would recommend using a different channel, as you already did, to avoid those frequencies.

Could this be a very strong signal from the surroundings or is the router broken?

Could be some WiFi device, running non-standard protocol instead of 802.11. It's similar to running wifi link with NV2 ... the last time I was checking, non-MT devices did not "detect" such activity on channel, but I guess a spectral scan would show a nice 20MHz-wide "interference".

An unexpected update: the mysterious signal disappears when I disconnect the SFP module from the router.