Digital Signal Processing Reference
In-Depth Information
Linear Constant Coefficient Difference Equations
The difference equation is another fundamental relation between the input
and output of LTI systems. It describes the discrete-time process executed
by the system, such as a high-pass or a low-pass filtering. The difference
equation takes the following form:
()
+
(
)
+…
(
)
=
()
+
(
)
+…
bbxn M
(
)
ay n
ay n
−
1
a y n N
−
bx n
bx n
−
1
−
0
1
N
0
1
M
or compactly as follows:
N
M
∑
∑
ayn k
(
−=
)
bxn k
(
−
)
(2.3)
k
k
k
=
0
k
=
0
Important properties of the difference equations are:
•
The order of the difference equation is
N
, where
N
≥
M
.
•
The coefficients
a
i
, i =
are constant,
real numbers and define the properties of the system. For example,
one set of coefficients could generate a low-pass filter, while a dif-
ferent set of coefficients could generate a highpass filter. The latter
aspect shows the great simplicity of digital systems.
0, 1, 2,
… N
, and
b
i
, i =
0, 1, 2,
… M
2.1.1
Introduction to Z-Transforms and the System Function
H
(
z
)
In the previous section, we introduced two fundamental properties of LTI
systems, linear convolution and difference equation. The link between these
two fundamental aspects of LTI systems is provided by the Z-transform. The
Z-transform is a very important time-to-frequency domain transformation
of the basic discrete-time signal
x
(
n
). The Z-transform is defined by the
equation:
∞
∑
xnz
n
−
Xz
()
=
()
(2.4)
n
=−∞
where the variable
z
is complex and the function
X
(
z
) is defined in the
complex plane.
property (2) from Table 2.1, in Equation 2.2, we get the transformed equation:
Y
(
z
) =
X
(
z
)
H
(
z
)
(2.5)