Wi-Fi Troubleshooting Explained: Reading Signal, Noise and Channels

Most wireless complaints arrive as "the Wi-Fi is slow", which describes a symptom and not a cause. Analysis tools turn that into a measurement, and the exam expects you to know what to measure and what the numbers mean.

The measurements

RSSI — received signal strength, in dBm. Always negative, and closer to zero is stronger. Around -30 dBm is excellent, -67 dBm is the design target for voice and video, -70 dBm is adequate for data, and below -80 dBm is unusable.

Noise floor — background RF energy, also in dBm. Typically around -95 dBm in a quiet environment. A higher noise floor means less usable signal margin.

SNR — signal-to-noise ratio, in dB, the difference between the two. This matters more than raw signal: aim for 25 dB or better for reliable high data rates, and below 20 dB throughput degrades noticeably.

The reason SNR is the better metric: a client at -65 dBm with a -95 dBm noise floor has 30 dB of margin and performs well. The same -65 dBm signal in an environment with a -75 dBm noise floor has only 10 dB and performs badly. The signal bar looks identical in both cases, which is why users report "full bars and it does not work".

Channel problems

Co-channel interference. Two access points on the same channel within range of each other. They cooperate politely — each waits for the other — so nothing breaks, but they share airtime and capacity halves. Symptom: performance degrades as more clients connect, with good signal throughout.

Adjacent channel interference. Overlapping but different channels. Worse than co-channel, because the radios cannot decode each other well enough to defer, so they transmit over one another and corrupt frames. Symptom: retransmissions and errors rather than simple slowness.

In 2.4 GHz only channels 1, 6 and 11 do not overlap. Any other assignment guarantees adjacent channel interference, and it is one of the most common findings in an analysis.

In 5 GHz there are many non-overlapping channels, so the usual issue is channel width — wide channels give higher peak throughput and fewer non-overlapping options, which in a dense deployment produces worse aggregate performance than narrower channels.

Non-Wi-Fi interference

The category a Wi-Fi analyser cannot see, and a point the exam makes. Wi-Fi analysis tools decode Wi-Fi frames; they are blind to energy that is not Wi-Fi.

Microwave ovens, cordless phones, video senders, wireless cameras, Bluetooth devices and some industrial equipment all radiate in 2.4 GHz. They raise the noise floor and corrupt frames without appearing as a network in any scan.

The diagnostic signature is a high noise floor or poor performance with no Wi-Fi explanation. The tool for confirming it is a spectrum analyser, which shows raw RF energy rather than decoded Wi-Fi. If a question describes interference invisible to a Wi-Fi scanner, the answer is a spectrum analyser.

The classic case: a break room microwave degrading 2.4 GHz performance for a few minutes at a time, several times a day.

Sticky clients and roaming

A device connects to an access point and holds on as the user walks away, refusing to roam to a nearer one until the connection is nearly unusable.

It happens because the roaming decision belongs to the client, not the infrastructure, and client implementations vary. Excessive transmit power makes it worse — a distant AP that is still audible at -80 dBm gives the client no reason to switch.

The counterintuitive fix is reducing transmit power so cells are smaller and clients are pushed to roam. Band steering and minimum data rate settings also help by refusing to serve clients at very low rates.

A troubleshooting order

Establish whether the problem is one client, one location or everywhere — that alone narrows it enormously. Check signal and SNR where the problem occurs, not at the desk of whoever reported it. Look at channel assignment and neighbouring networks. Check client count and airtime utilisation, since a strong signal on a saturated AP is still slow. Check for retransmissions, which indicate interference rather than coverage. And confirm the problem is wireless at all — a slow application over a perfect wireless link is not a Wi-Fi problem, and testing to a local target rather than the internet settles that quickly.

What to remember

RSSI is negative with -67 dBm the voice target; SNR matters more than raw signal, with 25 dB a good target. Co-channel interference shares airtime; adjacent channel interference corrupts frames and is worse. 2.4 GHz has only channels 1, 6 and 11. Wi-Fi analysers cannot see non-Wi-Fi interference — that needs a spectrum analyser. Reducing transmit power often fixes sticky clients.

Related reading: Wireless site surveys and heat maps · RSSI · SSID, BSSID and ESSID