Adam Frey Nyquist Sampling Theorem
Sampling Theorem: Let x(t) be a band-limited signal with X(j w) = 0 for |W| >Wm.
Then x(t) is uniquely determined by its samples x(nT) = 0,(+,-)[1,2,3]. . ., if
Ws > 2 * Wm,
where
Ws = (2* pi ) / T
Then if
X D [n] = X(nTs) are a collection of samples, then x(t) can be uniquely recovered from its samples if
Ts < (1/2) (2 pi)/ Wm
For example
if X(w) = u(w +2) - u(w-2), What is the largest Ts you can use to obtain xr(t)from x(t)?
well, Wm = 2, ( X(w) = 0 for |W | > Wm)
and Ws > 2 Wm
and Ts = 2(pi) / Wm
Ts = 2(pi) / (2 * 2) Ts = pi / 2 So the greatest Ts that can be used with out aliasing would be pi/ 2 .