A 60-square-meter apartment should not need enterprise networking gear. But after upgrading to a 300 Mbps plan and still measuring only 40-60 Mbps at my desk 8 meters from the router, I wanted to understand exactly why — not just accept it.
Building a Repeatable Test Protocol
Random speed tests are not useful for diagnosis because too many variables change between tests. I built a repeatable protocol: same device, same location, same time of day (6 PM), same test server, ten consecutive runs averaged. This took about 20 minutes per test session but produced data I could actually compare.
Baseline measurement with the laptop connected via Ethernet directly to the router: 287 Mbps. Same laptop connected via Wi-Fi from the desk: 53 Mbps. That 234 Mbps gap was the problem I was trying to explain.
Testing Each Variable Separately
First I checked Wi-Fi channel congestion using WiFi Analyzer, a free Android app. My router was set to automatic channel selection and had landed on channel 6 — the same channel as four neighboring networks, including one powerful enough to show -52 dBm signal strength on my device. Switching to channel 1 and then channel 11 improved throughput to 89 Mbps. Better, but not close to the wired speed.
Next I checked the frequency band. My router was broadcasting on 2.4 GHz, which has better range but significantly lower throughput than 5 GHz. The router had a 5 GHz radio enabled, but my laptop was connecting to 2.4 GHz automatically because both networks shared the same SSID. Separating them into distinct network names forced the laptop onto 5 GHz. Speed jumped to 198 Mbps.
Two configuration changes, zero cost, and speeds went from 53 Mbps to 198 Mbps. The hardware was never the problem.
The Last Gap
The remaining difference — 198 vs 287 Mbps — came from the physical environment. A concrete wall between the router and desk was attenuating the 5 GHz signal. I confirmed this with signal strength readings: -71 dBm at the desk, compared to -45 dBm in the same room as the router. At -71 dBm, 5 GHz connections reliably lose about 30-40% of their theoretical throughput.
Moving the router from behind a bookshelf to a higher, more central shelf position brought the signal up to -58 dBm and speeds to 241 Mbps. A $0 fix that required repositioning one device.
Total Cost of the Experiment
| Action Taken | Tool Used | Speed Gain | Cost |
|---|---|---|---|
| Changed Wi-Fi channel | WiFi Analyzer (free) | +36 Mbps | $0 |
| Separated 2.4/5 GHz SSIDs | Router admin panel | +109 Mbps | $0 |
| Repositioned router | Physical adjustment | +43 Mbps | $0 |
Total improvement: from 53 Mbps to 241 Mbps. Total cost: nothing. The key was measuring each variable in isolation rather than making all changes at once and not knowing which one helped. That discipline — changing one thing at a time — is the single most useful habit in network troubleshooting.