In Chapter 2, we looked at a version of Parseval's theorem for the Fourier series. Here, we will look at a similar version of this theorem for the Fourier transform. Recall that the energy of a signal is given by
e
x
=
∫
-
∞
∞
x
(
t
)
2
d
t
e
x
=
∫
-
∞
∞
x
(
t
)
2
d
t
(1)
If the energy is finite then x(t)x(t) is an energy signal, as described in Chapter 1. Suppose x(t)x(t) is an energy signal, then the autocorrelation function is defined as
r
x
(
t
)
=
x
(
t
)
*
x
(
-
t
)
r
x
(
t
)
=
x
(
t
)
*
x
(
-
t
)
(2)
It can be shown that rx(t)rx(t) is an even function of tt and that rx(0)=exrx(0)=ex(see Exercises). The Fourier transform of rx(t)rx(t) is given by X(jΩ)X(jΩ)*=X(jΩ)2X(jΩ)X(jΩ)*=X(jΩ)2. If follows that
e
x
=
1
2
π
∫
-
∞
∞
X
(
j
Ω
)
2
e
j
Ω
t
d
Ω
t
=
0
=
1
2
π
∫
-
∞
∞
X
(
j
Ω
)
2
d
Ω
e
x
=
1
2
π
∫
-
∞
∞
X
(
j
Ω
)
2
e
j
Ω
t
d
Ω
t
=
0
=
1
2
π
∫
-
∞
∞
X
(
j
Ω
)
2
d
Ω
(3)
Which is Parseval's theorem for the Fourier transform.