I haven't worked with demodulation for a while, so I can't explain the exact concept of demodulating it. But you need more then just one sample to demodulate. You need to get the frequencies and you can't detect frequency from a single sample.
You should consider using an array and a shift register, to append the samples to a package with specific length (the longer, the more precise your measurement will be).
Then you process a package of signal only every e.g. 512th iteration.
I hope this is not supposed to run with a live incoming radiosignal? I would guess that labview and the pc are both too slow for that.