Digital Signal Processing Reference
In-Depth Information
Fig. 3.9. Output
y
(
t
) computed
in Example 3.8.
0.4
0.3
0.2
0.1
t
0
−1
0
1
2
3
4
5
6
7
8
9
−2
Solution
Functions
x
(
τ
),
h
(
τ
),
h
(
−τ
), and
h
(
t
− τ
) are plotted in Figs. 3.10(a)-(d).
Depending on the value of
t
, the convolution integral takes five different forms.
We consider these five cases below.
Case 1
(
t
< −
3). As seen in Fig. 3.10(e), the non-zero parts of
h
(
t
− τ
) and
x
(
τ
) do not overlap. Therefore, the output signal
y
(
t
)
=
0.
Case 2
(
−
3
≤
t
≤
0). As seen in Fig. 3.10(f), the non-zero part of
h
(
t
− τ
)
partially overlaps with
x
(
τ
) within the region
τ =
[
−
2
,
t
+
1]. The product
x
(
τ
)
h
(
t
− τ
) becomes a rectangular function in the region with an amplitude
of 1
.
5
2
=
3. Therefore, the output for
−
3
≤
t
≤
0isgivenby
t
+
1
y
(
t
)
=
3d
τ
=
3(
t
+
3)
.
−
2
Case 3
(0
≤
t
≤
2). As seen in Fig. 3.10(g), the non-zero part of
h
(
t
− τ
)
overlaps completely with
x
(
τ
). The overlapping region is given by
τ =
[
t
−
2
,
t
+
1]. The product
x
(
τ
)
h
(
t
− τ
) is a rectangular function with an ampli-
tude of 3 in the region
τ
=
[
t
−
2
,
t
+
1]. The output for 0
≤
t
≤
2 is given by
t
+
1
y
(
t
)
=
3d
τ
=
9
.
t
−
2
Case 4
(2
≤
t
≤
5). The non-zero part of
h
(
t
− τ
) overlaps partially with
x
(
τ
)
within the region
τ
=
[
t
−
2
,
3]. Therefore, the output for 2
≤
t
≤
5isgiven
by
3
y
(
t
)
=
3d
τ
=
3(5
−
t
)
.
t
−
2
Case 5
(
t
≥
0). We see from Fig. 3.10(i) that the non-zero parts of
h
(
t
− τ
)
and
x
(
τ
) do not overlap. Therefore, the output
y
(
t
)
=
0.
Search WWH ::
Custom Search