DTFT, DFT, ZT, and FFT 469
R.5.28 Observe that DTFT closely resembles the analog FT presented in Chapter 4 of this
book.
The syntaxes (notation) used to indicate the direct and inverse FT pairs are as
follows:
F(ejW) = F[f(n)] (denotes the discrete time Fourier transform [DTFT])
f(n) = F–1[ejW] (denotes the inverse discrete Fourier transform [IDFT])
f(n)↔F(ejW) (denotes the transform equivalent time–frequency pair)
R.5.29 The DTFT F(ejW) is a periodic function of the real variable W, with a period equal
to 2π. This is the reason why the integral of the inverse transform is evaluated over
the range −π ≤ W ≤ +π. Also observe that
FejW f nejnW
n
() ()
∑
and its expansion is given by
Fe fe fe fe fne
fe f
jW jW j W jnW
jW
( ) (0) (1) (2) ( )
(1) (2
^02
))()efnej2W jnW
consisting of a sum of exponentials of the form e−jnW, for n = 0, ±1, ±2, ..., ±∞.
Observe also that the function e−j^2 πn repeats in frequency of every 2π rad.
R.5.30 Recall that the FT of an analog signal is a function of w, with radian/second as unit,
whereas W is used to denote discrete frequencies, with radian as unit.
Some authors differentiate the analog case from the discrete case by using the
variable Ω; in this book, W (uppercase) is used to denote discrete frequencies.
R.5.31 Since F[ejW] is a complex function of the real variable W, then F[eJW] can be repre-
sented by two parts, real and imaginary in rectangular form, or with a magnitude
and phase in polar form (similar to the continuous case) as follows:
F[eJW] = real {F[ejW]} + jimag {F[ejW]} (rectangular)
F[eJW] = |F[ejW]|e−jθ(n) (polar)
R.5.32 Let |F[eJW]| represent the magnitude, whereas θ(W) the phase of DTFT of f(n), then
the plot |F[eJW]|v e r s u s W is referred to as the magnitude spectrum plot of f(n),
whereas θ(W) versus W is referred to as the phase spectrum plot of f(n).
R.5.33 Recall that the relation between the polar and rectangular forms is
real {F[ejW]} = F[ejW]cos[θ(W)]
imag {F[ejW]} = F[ejW]sin[θ(W)]
where
F e( )jW real F e^22 [( )]jW imag F e[( )]jW