Noise measurements of metrobuses

Metrobuses: Monstrous in appearance, monstrous in its dominance of the traffic infrastructure and news of Trondheim since its introduction in August last year. Turns out to be pretty monstrous also in the frequency domain.

LA1BFA had noticed that the squelch on his 2m radio would sometimes open when he drove by a metrobus. LA3WUA took responsibility and went to Høgskoleparken in the darkness in order to point his antenna towards buses and do some noise measurements. (The author of this blog post was also forced outside in order to provide photographs.)

The experimental setup was that we first measured the driver side using vertical and horizontal polarization, and then the same for the passenger side at approximately the same distance by walking back a bit.

• 

• 

• 

There was definitely something around 146 MHz whenever a metrobus was passing.

• 

• 

• 

• 

• 

• 

The span is 80 MHz since we didn’t know what to look for in advance, so it doesn’t really give a good impression of the nature of the noise. We therefore limited the frequency range and did another measurement at vertical polarization:

• 

• 

The metrobus noise seems to span 3 MHz with a center frequency at around 146 MHz, covering the amateur radio bands.

The next step was to see whether we could sense the metrobus using the equipment at ARK. We directed our 2m array towards the road, and waited patiently for interesting signals.

• 

• 

Once in a while, we’d see comb noise splattered across the spectrum, as illustrated above. We could run to the window whenever we got this, look outside, and ascertain that, yes, there was a metrobus passing.

We then reduced the frequency range and recorded IQ data while carefully noting down metrobus passing times. (Thanks to LB5LI for long-legged assistance with metrobus spotting.)

• 

• 

The characteristic comb noise is present in the waterfall diagram every time a metrobus was passing (at time 44 seconds, 456 seconds, 516 seconds and 599 seconds). Some of the signals are a bit weak, so we also plotted individual spectrograms below for clarity:

The comb noise spikes are clear even for the weaker signals. For fun, it is possible to create a metrobus detector using simple classification techniques:

Is atb.no down, and you need to catch a bus, but the weather is too stormy outside? No problem, just point a 2m antenna towards the road, wait for the metrobus’ gargantuan presence across the 2m amateur radio band, and run very fast.

We’ll get back to more metrobuses in later posts. This represents yet another noise source we have to cope with, and it will be interesting to see whether the noise is within the allowable limits.