A little while ago I was called into a major incident where the problem statement was the wireless is broken. Now having known this site a warehouse, which had recently under gone a wireless redesign, I though how could the wireless be that bad that they had to pushback the implementation of a new ware house management system (WMS), especially as the current system had been working for weeks beforehand on the new design.
So I was asked to head to the site and investigate. Couple of points to note; the old WMS worked in a way where it only really needed to communicate with its servers at the start and end of the orders, where as the new system communicated each time an item was picked. In other words one used a cache mythology and the other was streaming data, which meant it was more noticeable if there were WIFI or network issues.
So first thing after getting to site and verifying the exact issue and areas within the site the issue was present (or more present) I broke out the Sidekick and started a survey just to confirm the coverage was correct – having not implemented this site myself.
So I got back from the survey and started reviewing the results and thats when I noticed the narrow band interference within the UNII 3 channels
Now I had to begin the process of determining if this was the cause of the issues and what was causing it. So initially from all the surveying it was determined that it only existed in channels 157, 161 and 165 so we decided to remove them from the channel plan, and determine if the issues went away.
Well feedback from this change was good, but there were still issues on site, so I went back and did another survey and guess what the interference was now in channel 149 so was time to find out what exactly was causing the interference.
I spend hours trying to find the source only to every time I got close it disappeared, I by accident walked into the maintenance department of the warehouse and there was a massive pile of old lights and bunch of new ones. I grabbed one of the new ones and noticed the sensor on it, and remembered from someone in Twitter saying they had had non WI-FI interference on 5GHz caused by light sensors, thought lets dig out the vendors data sheet for these lights.
Well it was confirmed they have an operational frequency of 5.8GHz +/- 75MHz so that basically meant for our channel plan the whole of UNII-3. So we had discussions on do we just remove UNII-3 or do we remove the interference, in this case IT won the argument as these lights had been installed as a temporary measure until the permanent solution was ready.
Video below is what happens when the interference is turned off.
The model of the motion sensor was the Pierlite motion sensor for Pielux Eco Highbay
Without having access to a spectrum analyser in this case the SideKick I would still be looking for the issue on site, so it pays to have the right tools for the job