By performing a wireless site survey we visualise the state of a complete wireless network. We pay attention to signal strength, channel usage, channel width and the maximum attainable theoretical speed. A performance test to see if the wireless network (still) meets the requirements is certainly part of this exercise.

By performing a spectrum analysis, we examine the electromagnetic spectrum and then solve the problems. WiFi exists on the first two layers of the OSI model. Wi-Fi is very ‘polite’: does it hear another WiFi transmission on the same channel or is there other electromagnetic energy present at a strength of -85dBm or stronger? The other WiFi-devices remain silent.

Sometimes there are problems in a network that are located in the higher OSI layers. By means of packet capture we see all data streams from and to devices on a specific channel. For example, we can discover whether the driver of a WiFi adapter can be the problem or why the 4-way hand-shake is not going well and logging in to the network is not successful.

A WiFi signal is divided into three: ‘want’, ‘do not want’ and ‘do not care’. We want a signal of -67dBm (want), but we do not want an overlap of the same channel with a signal of -85dBm or better (do not want). A signal of less than -85dBm does not bother us (do not care). This is the basis of a good channel distribution.