An old lab I worked in had a malfunctioning RF power supply for a vacuum plasma generator.
None of the researchers were cognizant of the fact it was spewing noise outside its frequency allocation until some relevant government agencies quickly visited and had it shut down. We were impressed with the response time :)
Related, the FCC still maintains a network of monitoring stations around the US. Back in the day they were manned, now they're automated and remotely controlled.
(And yes, GPS-linked direction finding units are definitely a thing that can be purchased commercially. They're big in the amateur radio world for "fox hunts" - competitions to try and find hidden radio transmitters.)
Fun fact about modern radio direction-finding techniques: you don't need directional antennas, just an array of antennas you can observe phase differences across. It works shockingly well.
That's got some usual Nyquist limitations, right? For high enough freqs you'll be able to identify the plane, but the direction will potentially alias?
If you infringe on a mobile carrier’s spectrum, for which they have paid a handsome dollar, you can expect to find a man with a Yagi antenna and frequency analyzer in your backyard within a couple hours.
How far off from its allotted frequency does a device have to be in order to interfere? We are all aware of tuning an old radio where a station (e.g. broadcasting an FM signal) will come in slightly at 102.3 but much clearer at its real value at 102.4 so there is some range for each station.
Does the range change as frequencies increase (i.e. does the margin of error get smaller as you go up the spectrum)?
You should visit a WebSDR site like this one http://websdr.ewi.utwente.nl:8901 as it nicely visualizes the bandwidth actually taken by a single broadcast, as well as how tightly neighboring signals can be packed. To answer your question, a broadcast of a certain kind will take the same amount of bandwidth regardless of the frequencies, but it's typical for the higher bands to simply offer more bandwidth and as such offer more broadcasting spots.
What you're describing isn't anything related to what the original post describes. The "real value" an FM broadcast is not 102.3 but 102.3 +/- 75 kHz. It's expected to be able to observe the signal even with the dial 100 kHz off because the signal width is 150 kHz.
You don't even have to be off your frequency.. insufficient filtering of your harmonics can cause interference in other bands. Your interference is characterized as the amount of energy you are radiating above the "noise floor."
None of the researchers were cognizant of the fact it was spewing noise outside its frequency allocation until some relevant government agencies quickly visited and had it shut down. We were impressed with the response time :)