How to interpret wifi statistics?

On my MR42, it seems one radio (the meraki scanning radio) shows a variable noise level. I take this to imply it is the only one that does potentially return useful noise measurements.

If that is the case, how do I interpret the graph below and what is the significance of the signal and noise converging at around 21:30?
Am I right in thinking, at that point, the signal becomes indistinguishable from the noise?
BTW I have been trying very low tx powers for that radio.

Any insight appreciated.

I see noise around -86 dBm (not between -20 and -30, your chart also says that the average is -84.9). Follow the red line.

Given your interpretation is slightly flawed - do you still have questions?

I don't follow.

At around 9:30pm, the graph as I'm reading it, shows noise the same as signal (approx -20dBm). Am I reading it right and if so, does that mean signal is 'lost' in the noise?

If you only mean at 9:30pm, yes - that's a fair assessment (I assume you'd have better knowledge, since it seems 9:30pm particularly interested you - ignoring e.g., the preceeding time).

Out of curiousicity, how does this relate to OpenWrt?

Is there detail or context regarding the chart that might be helpful in our understanding?

Tx?

OK.

Thanks. OK First was to ensure I understood the graph.

Here's why I came to enabling and trying to understand the owrt collectd data:
I have roku streaming stick, that when it's streaming, appears to interfere with IOT devices. . Monitoring these IOT devices shows that when the roku is streaming, they have slower/erratic pings and briefly unreachable.

Ther roku is now the only client connected to wlan2-ap0. Always on and active on the network, and for 24hr period above was streaming content between 18:30 and 21:00, with no apparent issues.
I have attempted to isolate the roku on its own radio, different channel and, since it is very close to the physical access point, used the meraki scanning radio and set the maximum transmit power to very low levels.
I see a marked improvement with IOT devices through the day but still some interruption to them once the roku is streaming. The roku uses a second wifi connection to coomunicate with the remote control, but I have yet to catch this on my phone based wifi scanner. I believe it is on the same channel as the roku's main wifi connection. So I don't understand how either is interrupting the IOT devices, now they are separated.

Perhaps check the channel bandwidth (when on the same band) ?

radio0 is the 5GHz radio.
Both 2.4Ghz radios are set to HT20.
radio 1 (iot devices) is channel 1, radio2 (roku only) is channel 6.

But perhaps radio2 is operating in some wide band mode despite the setting above? It's the meraki scanning radio.

It would seem that the OpenWRT device is configured adequately (2.4 GHz, channels well separated). It would be prudent to look at the logs at the time of interference. I would however guess you'd need to look for external sources of interference.

Can it be related to the roku Wifi Direct settings (remote) ?

Thanks. There are no settings as such for the roku voice remote. I will need to dig around to see what channel it uses. I'm guessing it will be the same as the roku to local network since I doubt the roku stick has two radios.

Usually [the first] 20 MHz of the channel you've set for the AP.

You may have hit on something! The door to the lounge tends to be closed only whilst the tv (and therefore roku) are in use. This door is in direct line of sight between the AP and majority of the IoT devices (which reside in the stair cupboard across the hall).

Update: The state of the door I refer to above had no bearing on the signal.

Keep in mind that you are dealing with radio waves so interference could be radio waves from sources other than your WIFI devices. 2.4gHz WIFI is in the ISM radio band (https://en.wikipedia.org/wiki/ISM_radio_band) so interference could be coming from any of a wide range of sources with legitimate legal use that interferes with your desired use. In your home, that could be cordless phones, baby monitors, intercoms, microwave ovens etc. It could also be sourced outside from your house/apartment from a neighbor's equipment, municipal, police, utility company equipment at the street or a nearby facility and lastly industrial use could potentially be further away than a few hundred feet depending regulations in your jurisdiction.

2 Likes

True - from a previous discussion that also mentions this:

There's a lot of factors to consider regarding WiFi - especially since it doesn't have legal priority usage of the band (e.g., your microwave have higher priority than WiFi).

:fork_and_knife_with_plate:

1 Like

The symptom I'm investigating (protracted ping of IoT devices) is closely correlated timewise with use of the roku tv stick.


But I take the point, the cause may still involve other factors.
Is DFS a consideration with 2.4GHz?