WLAN Design and Site Survey
Survey types
Predictive surveys simulate coverage from floor plans and attenuation values; passive surveys listen only and show the real RF environment; active surveys associate to the network and measure true throughput, roaming and application behavior. An AP-on-a-stick survey temporarily places one AP in many candidate locations to validate a predictive model. Always finish with a validation survey after deployment - a heat map is a model, not a measurement.
Design targets
Design coverage-first in sparse environments and capacity-first in dense ones. Common cell-edge targets are −67 dBm with 25 dB SNR for voice, and −70 dBm with 20 dB SNR for data; 15–20% cell overlap keeps roaming smooth. Plan 2.4 GHz with channels 1, 6 and 11 only, and 5 GHz with 20 MHz channels spaced to avoid adjacent-channel interference.
Attenuation and traps
Typical losses: drywall 3 dB, brick 6 dB, concrete 10–15 dB, metal 20 dB or more, and low-E glass can be worse than concrete. Frequent traps: calling a passive survey "active," forgetting the uplink (client radios transmit weaker than APs), trusting a survey laptop NIC as representative of phones and scanners, and designing 2.4 GHz with 40 MHz channels. Also remember DFS radar events and 6 GHz preferred scanning channels change the usable channel plan.
Worked example
For a 30,000 ft² open office with 250 laptops and 100 VoWiFi handsets, start from capacity: if each AP can comfortably carry 30 active clients at the required airtime, that is about 12 APs from client count alone; then overlay the −67 dBm voice cell-edge requirement to confirm no coverage holes and adjust placement rather than simply adding power.