Interference hunting using a RTSA

Interference hunting is by nature tricky. The interference can be intermittent and getting information on when and where it occurs and what it affects is often not straightforward. This was the exactly the case in this example: a mobile network operator (MNO) had been contacted by a hospital saying that a 5G upgrade was affecting the hospital paging system.

Information was required from the hospital, the paging service provider and the MNO. What we were told didn’t quite stack up – plus whilst possible it’s probably unlikely that a 5G service will be causing interference on a 450 MHz paging system.
On site an initial investigation (OK we were sat in the hospital coffee shop) revealed the presence of wide band noise across the 450 MHz pager spectrum. What made it easy to spot was using a real time spectrum analyser (RTSA), a Narda SignalShark.

Fig 1: Wideband noise clearly visible in the spectrum (red line) & spectrogram (waterfall) display.

The noise was intermittent and invisible to the conventional swept spectrum analyser we had on site. This isn’t surprising given a swept analyser will be blind to transient signals that occur whilst the sweep is looking at a different part of the set frequency range. Different when using a RTSA which captures pretty much everything in the instruments’ real time bandwidth.
What we found was that the noise level increased (or rather the signal to noise ratio decreased) as we went up floors in the hospital. By the time we got to the rooftop and accessed the plant room it seemed we could have found the culprit, the noise seemed to coincide with switching of the lift motors.

One of the frustrations of interference hunting is that to get to a definitive conclusion you often need control over the given systems. Despite representatives of all organisations involved being on site we didn’t get to that, understandable given it’s not easy to shut off lifts or affect a safety critical paging service (or indeed network coverage) in a busy hospital.

We can’t say there was never any interference caused by telecoms services but we can say there was no indication of that whilst we were on site. We can say the only anomaly detected was the presence of wideband noise, completely across the 450 MHz paging spectrum, which was very easy to identify using the RTSA function of the Narda SignalShark.

Back to News