Non-Wi-Fi Interference in 5GHz

 Up until just recently, I had never seen non-Wi-Fi interference in the 5GHz band. After some consultation with other folks who could lend their expertise, I realized that I was looking at sustained interference from an outside source. I still haven't found out what it is but that will be for the engineers on site to figure out. Hopefully, this will help other engineers identify interference and save time second guessing yourselves.

At a site recently, I noticed that every access point I looked into had at least 25% channel utilization and as much as 65%!  The hospital does have seven active SSID's, but that isn't enough management overhead to warrant those high utilization rates. Also, they had very low co-channel interference.

In this screen grab the channel utilization for this access point is 43%.

This screen grab shows very low management overhead...

In this screen grab, Ekahau shows very little co-channel interference.

If there is no co-channel interference to speak of, and the management overhead is not very high, it has to be non-Wi-Fi interference... right?

This is a screen grab of the Ekahau Spectrum Analyzer sitting in the nurse station of this building. Notice all the channels with some high utilization, particularly channel 48 and channel 161. There is 802.11 traffic on 40 MHz channel 44-48 as evidenced by the OFDM traffic (Bart!), but the utilization is high on only channel 48. The utilization on 161 is very odd as well.

In this screen grab the 802.11 traffic has stopped, but the interference is still there and still strong. I did not see any cordless phones I didn't know about and I still suspect it will turn out to be cameras or something continuously on and transmitting. The assessment was concluded by informing the client of the situation and having them identify and attempt to mitigate it. I am curious myself what it turns out to be. If a solution is discovered I will update this blog with the result.