The distinction is between analogue and digital pulse modulation, and it decides everything.

PAM, PWM and PPM are all analogue pulse schemes: the information lives in a continuously variable property of the pulse.

Because a corrupted amplitude, width or position is a corrupted sample value, noise in an analogue pulse scheme translates directly into error at the output, and it accumulates at every repeater along the route.

PCM breaks that chain. The receiver has to answer only a yes/no question: was a pulse present? Provided the noise stays below half the decision threshold, the answer is correct and the regenerated pulse is perfect — the noise picked up on that hop is discarded rather than passed on. A chain of regenerative repeaters therefore has essentially the noise performance of a single hop, which is exactly why long-haul and undersea links are digital. So (R) states precisely the mechanism behind (A), and the code is 1.

What PCM pays for the immunity. Sampling at \(f_{s}\ge2f_{m}\) and coding each sample into n bits needs a bandwidth of roughly

\(B_{PCM}\approx n f_{s}\)

far more than the message bandwidth. It also introduces quantisation noise, which for a uniform quantiser of step \(\Delta\) has power \(\Delta^{2}/12\) and gives

\(\left(\dfrac{S}{N}\right)_{dB}\approx6n+1.76\)

— the familiar 6 dB per bit rule. Quantisation noise is unavoidable, but it is fixed at the encoder and, unlike channel noise, does not grow along the route.

The threshold effect is the other side of the coin: PCM performance is excellent above a certain carrier-to-noise ratio and degrades abruptly below it, whereas analogue schemes fade gracefully.

Hence, both (A) and (R) are true and (R) is the correct explanation of (A).