Radio Astronomy Expertise Wanted

Regarding Anderson's paper on an anomalous acceleration of
Pioneer 10 I had the following quest which
and expert in Radio Astronomy could answer:
You are all blithely assuming that the low amplitude
oscillations of charge around some
average that are received when the receiving antenna is pointing
in the region of the
sky containing the spacecraft show a clear frequency over a
period of four hours or
so when the receiver antenna is so positioned and constantly
readjusted to point
to the spacecraft as the earth moves.
This may be a perfectly correct assumption on your part
and Anderson et al.; parts of the Anderson paper and others
describing the method of amplifying and detecting the expected
doppler shifted frequency are hard to decipher though
George Dishman has been helpful in translating some of the
FFT and PLL jargon.
If the reception was perfectly clear then one would see every
(1/2.11.....) times 10^-9 seconds a crossing at zero ( average
value) of the oscillating voltage.
But because of thermal noise and other radiation other
voltages
are added to this expected set of voltages and so to those at the
zero crossings every (1/2.11.....) times 10^-9 seconds
obscuring|
the zero and to other voltages making zero crossings appear which
are not part of the systematic pattern.
Using phase locked loops one could find a first true zero
and then one could take a second or billions of successive
voltage values
and add them to the next second or billion etc for 60 seconds and
the
random obscuring voltages would cancel out.
Apparently they have done so until recently for Pioneer 10 but
I
imagine there had been a steady degradation and I would like
to know from a radio astronomer what form this degradation
takes??
eg How many regularly spaced zeros do you have to miss before
you decide the data is too noisy to establish a specific
frequency?