The exposed station problem can be depicted as follows:

A==B==C==D

When B is transmitting frames to A, and C hears the medium is busy so
it defers its transmisison to D.

The MACAW avoids the this problem in that B will send RTS to A waiting
for CTS from A, if C doesn't hear CTS, then C can decide that it can
send frame to D. Is that right?

If it is, Why not C simply get RTS and use the SA&RA(sender&Receiver)
address pair in RTS to determine whether it can send frames to D?

Can somboy explain this, thanks very much.