Signal Representation in Communication Questions and Answers

Answer: B

Because aliasing is irreversible. Once a component above half the sampling rate has folded down, it occupies exactly the same frequency as a legitimate component and no subsequent processing can tell them apart or separate them. The only remedy is to remove the offending energy before the fold happens, which is what the anti-alias filter does. Filters operate on discrete-time signals perfectly well, which is the whole of digital filtering, so the third option is simply false. Nothing damages the filter. And a delay is a benign and correctable effect, quite unlike an irreversible loss of information.

Answer: C

Time-reversal is exactly what makes the contributions add coherently. Convolving a pulse with its own time-reverse brings every frequency component into phase at one instant, so the signal builds to a peak there while the noise, being uncorrelated, adds in power only - which is why the signal-to-noise ratio at that instant is the best any linear filter can achieve. It cannot remove noise; it optimises the ratio, which is a different and achievable goal. Bandwidth is set by the transmitted pulse. And the output shape is the pulse's autocorrelation, which is generally rounded rather than rectangular.