I suppose isolation is the reciprocal
of coupling but I have not seen it mentioned as
an antenna parameter. Coupling has 2 meanings: there
is coupling between adjacent elements in an array
which is either desired because it enhances either
sidelobe levels or gain, and unwanted coupling that
degrades performance or complicates the design process.
So far as isolation is concerned, there is a parameter
called cross-polar isolation, which is the degree
to which two antennas with nominally opposite polarisation
still cross-couple due to imperfection. Thus if
an antenna is nominally vertically polarised but
with a small say -15dB of horizontal vector, then
the isolation to an antenna that is hrizontally
polarised will be -15dB.
Diversity takes several
forms and is normally encountered when a path between
antennas has a time-varying refractive index and
a possibility of two or more paths: one direct and
the others by reflection. Spacing two receive antennas
maximises the probability that the paths to one
of them will result in a vector sum that is a maximum,
whilst at the other antenna the received signal
has a vector sum that is a minimum, ie has faded
due to out of phase vector addition. The antennas
can be spaced apart vertically or horizontally (space
diversity), or the polarisation can be opposite
(polarisation diversity), or two different frequencies
can be used (frequency diversity). The ITU-R publish
recommendations how these can be implemented.
At bottom, life is all
Sucking in and blowing out.