Survey the spectrum you are sitting in
Task
Scan the wireless networks around you, record channel, band, signal strength and width for each, and work out where the congestion is. Everything in objective 2.3 — channels, bands, RSSI, overlap — becomes concrete in one scan.
Steps
- Run the scan and capture the output to a file so you can work on it:
sudo iw dev wlan0 scan > /tmp/scan.txt. - Extract, for each network, the SSID, channel, signal in dBm, and channel width. Group them by band — channels 1 to 14 are 2.4 GHz, 36 and above are 5 GHz, and anything above 1 in the 6 GHz list is Wi-Fi 6E.
- Count how many networks sit on each 2.4 GHz channel. Then check how many are on 1, 6 or 11 versus something else. Any network not on 1, 6 or 11 overlaps two of them.
- Rank the networks by signal. Note where yours falls, and convert the dBm figures to the practical bands: better than −65 dBm is good enough for voice, −70 is adequate for data, below −80 is unusable.
- Decide which 2.4 GHz channel you would move to and justify it from your own counts, not from the usual advice.
Verify
sudo iw dev wlan0 scan | grep -E "SSID|freq|signal|DS Parameter" | head -40
sudo iw dev wlan0 scan | grep -c "^BSS"
python3 -c "
for ch in (1,6,11):
print('2.4 GHz channel', ch, 'centre', 2407 + ch*5, 'MHz')
print('non-overlapping 2.4 GHz channels: 1, 6, 11 only -- 22 MHz wide, 5 MHz apart')
print('good -65 dBm | adequate -70 dBm | unusable below -80 dBm')
"
The scan must list multiple BSSIDs with signal values in dBm. Your channel count per 2.4 GHz channel is the finding — and if the busiest channel is not 1, 6 or 11, you have found a network making the problem worse for everyone.
Notes
The 2.4 GHz arithmetic is worth understanding rather than memorising. Channels are 5 MHz apart and each carrier is about 22 MHz wide, so a channel overlaps its four neighbours on each side. Only 1, 6 and 11 are far enough apart to avoid each other — and channel 3 overlaps both 1 and 6, which is why "I picked an empty channel" so often makes things worse.
5 GHz has many non-overlapping channels, which is most of why it performs better in dense areas. The trade is range: higher frequency attenuates faster, so a 5 GHz cell is smaller than a 2.4 GHz one at the same power.
Channel width is the other lever. 40, 80 and 160 MHz channels are faster and consume more of the band, so in a crowded environment a narrower channel often delivers more real throughput than a wider one.