There are many kinds and sources of noise, and the method you would use to
characterize a transistor or an amplifiier's generated noise, is different than
the methods you'd use to chase down system level noises like the interferer
you mention.
A spectrum analyzer will let you find continuous fixed-frequency "noise". But
inj a digital system it may show you series of spurs that really attend a clock
or data edge. A real time 'scope looking for impulses is also useful. The spec-an
may also provide a trigger outout that can be used to trigger the 'scope at a
burst event that might be too fast to see on the display.
Some types of transistor noise (SOI RTN) are almost impossible to see on a
traditional noise measurement system because they are sub-Hz bilevel (not
gaussian, and most 1/f noise rigs start about 10Hz). I've used a 'scope to
characterize this peculiar type of noise after having my modeling "experts
first tell me there was none, then that it couldn't be measured. Beware of
experts who say "can't".