Discrete-time signals and systems

See Oppenheim and Schafer, Second Edition pages 893, or First Edition
pages 879.
1 Discrete-time signals
A discrete-time signal is represented as a sequence of numbers:
. = {.n|]. o < n < o.
Here n is an integer, and .n| is the nth sample in the sequence.
Discrete-time signals are often obtained by sampling continuous-time signals.
In this case the nth sample of the sequence is equal to the value of the analogue
signal .
o
(t ) at time t = nT :
.n| = .
o
(nT ). o < n < o.
The sampling period is then equal to T , and the sampling frequency is

x
= 1,T .
x[1]
... ...
t
.
o
(1T )
For this reason, although .n| is strictly the nth number in the sequence, we
often refer to it as the nth sample. We also often refer to the sequence .n|
when we mean the entire sequence.
Discrete-time signals are often depicted graphically as follows:
1
x[3]
... ...
n
0 1 2 3 4
1 2 3
4
x[2]
x[4]
x[4]
x[0]
x[1]
x[3]
x[1]
x[2]
(This can be plotted using the MATLAB function stem.) The value .n| is
undened for noninteger values of n.
Sequences can be manipulated in several ways. The sum and product of two
sequences .n| and ,n| are dened as the sample-by-sample sum and product
respectively. Multiplication of .n| by a is dened as the multiplication of
each sample value by a.
A sequence ,n| is a delayed or shifted version of .n| if
,n| = .n n
0
|.
with n
0
an integer.
The unit sample sequence
1
n
0
is dened as
n| =
8
<
:
0 n = 0
1 n = 0.
This sequence is often referred to as a discrete-time impulse, or just impulse.
It plays the same role for discrete-time signals as the Dirac delta function does
for continuous-time signals. However, there are no mathematical
2
complications in its denition.
An important aspect of the impulse sequence is that an arbitrary sequence can
be represented as a sum of scaled, delayed impulses. For example, the
sequence
... ...
n
0 1 2 3 4
1 2 3
4
a
4
a
3
a
2
a
1
a
0
a
1
a
2
a
3
a
4
can be represented as
.n| = a
4
n 4| a
3
n 3| a
2
n 2| a
1
n 1| a
0
n|
a
1
n 1| a
2
n 2| a
3
n 3| a
4
n 4|.
In general, any sequence can be expressed as
.n| =
1
X
kD1
.k|n k|.
The unit step sequence
1
n
0
is dened as
un| =
8
<
:
1 n _ 0
0 n < 0.
3
The unit step is related to the impulse by
un| =
n
X
kD1
k|.
Alternatively, this can be expressed as
un| = n| n 1| n 2| =
1
X
kD0
n k|.
Conversely, the unit sample sequence can be expressed as the rst backward
difference of the unit step sequence
n| = un| un 1|.
Exponential sequences are important for analysing and representing
discrete-time systems. The general form is
.n| =
n
.
If and are real numbers then the sequence is real. If 0 < < 1 and is
positive, then the sequence values are positive and decrease with increasing n:
...
n
0
...
For 1 < < 0 the sequence alternates in sign, but decreases in magnitude.
For [[ > 1 the sequence grows in magnitude as n increases.
A sinusoidal sequence
...
n
0
...
4
has the form
.n| = cos(o
0
n ) for all n.
with and real constants. The exponential sequence
n
with complex
= [[e
}o
0
and = [[e
}
can be expressed as
.n| =
n
= [[e
}
[[
n
e
}o
0
n
= [[[[
n
e
}(o
0
nC)
= [[[[
n
cos(o
0
n ) [[[[
n
sin(o
0
n ).
so the real and imaginary parts are exponentially weighted sinusoids.
When [[ = 1 the sequence is called the complex exponential sequence:
.n| = [[e
}(o
0
nC)
= [[ cos(o
0
n ) [[ sin(o
0
n ).
The frequency of this complex sinusoid is o
0
, and is measured in radians per
sample. The phase of the signal is .
The index n is always an integer. This leads to some important differences
between the properties of discrete-time and continuous-time complex
exponentials:
v Consider the complex exponential with frequency (o
0
2):
.n| = e
}(o
0
C2t)n
= e
}o
0
n
e
} 2tn
= e
}o
0
n
.
Thus the sequence for the complex exponential with frequency o
0
is
exactly the same as that for the complex exponential with frequency
(o
0
2). More generally, complex exponential sequences with
frequencies (o
0
2r), where r is an integer, are indistinguishable from
one another. Similarly, for sinusoidal sequences
.n| = cos(o
0
2r)n | = cos(o
0
n ).
v In the continuous-time case, sinusoidal and complex exponential
sequences are always periodic. Discrete-time sequences are periodic (with
period N) if
.n| = .n N| for all n.
5
Thus the discrete-time sinusoid is only periodic if
cos(o
0
n ) = cos(o
0
n o
0
N ).
which requires that
o
0
N = 2k for k an integer.
The same condition is required for the complex exponential sequence
Ce
}o
0
n
to be periodic.
The two factors just described can be combined to reach the conclusion that
there are only N distinguishable frequencies for which the corresponding
sequences are periodic with period N. One such set is
o
k
=
2k
N
. k = 0. 1. . . . . N 1.
Additionally, for discrete-time sequences the interpretation of high and low
frequencies has to be modied: the discrete-time sinusoidal sequence
.n| = cos(o
0
n ) oscillates more rapidly as o
0
increases from 0 to ,
but the oscillations become slower as it increases further from to 2.
6
1
0
1
1
0
1
1
0
1
1
0
1
1
0
1
8 6 4 2 0 2 4 6 8
1
0
1
!
0
D
0
!
0
D
8
!
0
D
4
!
0
D

!
0
D
7

4
!
0
D
1
5

8
The sequence corresponding to o
0
= 0 is indistinguishable from that with
o
0
= 2. In general, any frequencies in the vicinity of o
0
= 2k for integer
k are typically referred to as low frequencies, and those in the vicinity of
o
0
= ( 2k) are high frequencies.
7
2 Discrete-time systems
A discrete-time system is dened as a transformation or mapping operator that
maps an input signal .n| to an output signal ,n|. This can be denoted as
,n| = T {.n|].
y[n] x[n]
T {]
Example: Ideal delay
,n| = .n n
d
| :
y[n]=x[n2]
... ...
n
0 1 2 3 1 2 3
... ...
n
0 1 2 3 1 4 5
x[n]
This operation shifts input sequence later by n
d
samples.
8
Example: Moving average
,n| =
1
M
1
M
2
1

2
X
kD
1
.n k|
For M
1
= 1 and M
2
= 1, the input sequence
y[3]
x[n]
... ...
n
0 1 2 3 1 4 5
y[2]
yields an output with
.
.
.
,2| =
1
3
(.1| .2| .3|)
,3| =
1
3
(.2| .3| .4|)
.
.
.
In general, systems can be classied by placing constraints on the
transformation T {].
2.1 Memoryless systems
A system is memoryless if the output ,n| depends only on .n| at the same n.
For example, ,n| = (.n|)
2
is memoryless, but the ideal delay
9
,n| = .n n
d
| is not unless n
d
= 0.
2.2 Linear systems
A system is linear if the principle of superposition applies. Thus if ,
1
n| is the
response of the system to the input .
1
n|, and ,
2
n| the response to .
2
n|, then
linearity implies
v Additivity:
T {.
1
n| .
2
n|] = T {.
1
n|] T {.
2
n|] = ,
1
n| ,
2
n|
v Scaling:
T {a.
1
n|] = aT {.
1
n|] = a,
1
n|.
These properties combine to form the general principle of superposition
T {a.
1
n| b.
2
n|] = aT {.
1
n|] bT {.
2
n|] = a,
1
n| b,
2
n|.
In all cases a and b are arbitrary constants.
This property generalises to many inputs, so the response of a linear system to
.n| =
P
k
a
k
.
k
n| will be ,n| =
P
k
a
k
,
k
n|.
2.3 Time-invariant systems
A system is time invariant if a time shift or delay of the input sequence causes
a corresponding shift in the output sequence. That is, if ,n| is the response to
.n|, then ,n n
0
| is the response to .n n
0
|.
For example, the accumulator system
,n| =
n
X
kD1
.k|
10
is time invariant, but the compressor system
,n| = .Mn|
for M a positive integer (which selects every Mth sample from a sequence) is
not.
2.4 Causality
A system is causal if the output at n depends only on the input at n and earlier
inputs.
For example, the backward difference system
,n| = .n| .n 1|
is causal, but the forward difference system
,n| = .n 1| .n|
is not.
2.5 Stability
A system is stable if every bounded input sequence produces a bounded output
sequence:
v Bounded input: [.n|[ _ T
x
< o
v Bounded output: [,n|[ _ T
,
< o.
For example, the accumulator
,n| =
n
X
kD1
.n|
11
is an example of an unbounded system, since its response to the unit step un|
is
,n| =
n
X
kD1
un| =
8
<
:
0 n < 0
n 1 n _ 0.
which has no nite upper bound.
3 Linear time-invariant systems
If the linearity property is combined with the representation of a general
sequence as a linear combination of delayed impulses, then it follows that a
linear time-invariant (LTI) system can be completely characterised by its
impulse response.
Suppose h
k
n| is the response of a linear system to the impulse n k| at
n = k. Since
,n| = T
(
1
X
kD1
.k|n k|
)
.
the principle of superposition means that
,n| =
1
X
kD1
.k|T {n k|] =
1
X
kD1
.k|h
k
n|.
If the system is additionally time invariant, then the response to n k| is
hn k|. The previous equation then becomes
,n| =
1
X
kD1
.k|hn k|.
This expression is called the convolution sum. Therefore, a LTI system has
the property that given hn|, we can nd ,n| for any input .n|. Alternatively,
,n| is the convolution of .n| with hn|, denoted as follows:
,n| = .n| + hn|.
12
The previous derivation suggests the interpretation that the input sample at
n = k, represented by .k|n k|, is transformed by the system into an
output sequence .k|hn k|. For each k, these sequences are superimposed
to yield the overall output sequence:
n 0 0 1 n
n 0 0 n
1
n 0 0 1 n
1
n 0
.n| hn|
.1|n 1|
.1|n 1|
.1|hn 1|
.1|hn 1|
,n| = .1|hn 1| .1|hn 1|
A slightly different interpretation, however, leads to a convenient
computational form: the nth value of the output, namely ,n|, is obtained by
multiplying the input sequence (expressed as a function of k) by the sequence
with values hn k|, and then summing all the values of the products
.k|hn k|. The key to this method is in understanding how to form the
sequence hn k| for all values of n of interest.
To this end, note that hn k| = h(k n)|. The sequence hk| is seen to
be equivalent to the sequence hk| reected around the origin:
13
Shift
0
0
0
k
k
k
2
2
n n+2
5
5
n5
Reflect
h[nk]
h[k]
h[k]
The sequence hn k| is then obtained by shifting the origin of the sequence
to k = n.
To implement discrete-time convolution, the sequences .k| and hn k| are
multiplied together for o < k < o, and the products summed to obtain the
value of the output sample ,n|. To obtain another output sample, the
procedure is repeated with the origin shifted to the new sample position.
Example: analytical evaluation of the convolution sum
Consider the output of a system with impulse response
hn| =
8
<
:
1 0 _ n _ N 1
0 otherwise
to the input .n| = a
n
un|. To nd the output at n, we must form the sum over
all k of the product .k|hn k|.
14
10 5 0 5 10
0
0.5
1
k
x
[
n
]
10 5 n(N1) 0 n 5 10
0
0.5
1
k
h
[
n

k
]
Since the sequences are non-overlapping for all negative n, the output must be
zero:
,n| = 0. n < 0.
For 0 _ n _ N 1 the product terms in the sum are .k|hn k| = a
k
, so it
follows that
,n| =
n
X
kD0
a
k
. 0 _ n _ N 1.
Finally, for n > N 1 the product terms are .k|hn k| = a
k
as before, but
the lower limit on the sum is now n N 1. Therefore
,n| =
n
X
kDn1C1
a
k
. n > N 1.
15
4 Properties of LTI systems
All LTI systems are described by the convolution sum
,n| =
1
X
kD1
.k|hn k|.
Some properties of LTI systems can therefore be found by considering the
properties of the convolution operation:
v Commutative: .n| + hn| = hn| + .n|
v Distributive over addition:
.n| + (h
1
n| h
2
n|) = .n| + h
1
n| .n| + h
2
n|.
v Cascade connection:
x[n] y[n]
x[n] y[n]
h
1
n|
h
1
n| h
2
n|
h
2
n|
,n| = hn| + .n| = h
1
n| + h
2
n| + .n| = h
2
n| + h
1
n| + .n|.
v Parallel connection:
x[n] y[n]
h
1
n|
h
2
n|
,n| = (h
1
n| h
2
n|) + .n| = h
]
n| + .n|.
Additional important properties are:
v A LTI system is stable if and only if S =
P
1
kD1
[hk|[ < o. The ideal
16
delay system hn| = n n
d
| is stable since S = 1 < o; the moving
average system
hn| =
1
M
1
M
2
1

2
X
kD
1
n k|
=
8
<
:
1

1
C
2
C1
M
1
_ n _ M
2
0 otherwise.
the forward difference system hn| = n 1| n|, and the backward
difference system hn| = n| n 1| are stable since S is the sum of a
nite number of nite samples, and is therefore less than o; the
accumulator system
hn| =
n
X
kD1
k|
=
8
<
:
1 n _ 0
0 n < 0
= un|
is unstable since S =
P
1
nD0
un| = o.
v A LTI system is causal if and only if hn| = 0 for n < 0. The ideal delay
system is causal if n
d
_ 0; the moving average system is causal if
M
1
_ 0 and M
2
_ 0; the accumulator and backward difference systems
are causal; the forward difference system is noncausal.
Systems with only a nite number of nonzero values in hn| are called Finite
duration impulse response (FIR) systems. FIR systems are stable if each
impulse response value is nite. The ideal delay, the moving average, and the
forward and backward difference described above fall into this class. Innite
impulse response (IIR) systems, such as the accumulator system, are more
difcult to analyse. For example, the accumulator system is unstable, but the
17
IIR system
hn| = a
n
un|. [a[ < 1
is stable since
S =
1
X
nD0
[a
n
[ _
1
X
nD0
[a[
n
=
1
1 [a[
< o
(it is the sum of an innite geometric series).
Consider the system
delay
Forward
difference
Onesample
which has
hn| = (n 1| n|) + n 1|
= n 1| + n 1| n 1| + n|
= n| n 1|.
This is the impulse response of a backward difference system:
delay
Forward
difference
Backward
difference
Onesample
Here a non-causal system has been converted to a causal one by cascading with
a delay. In general, any non-causal FIR system can be made causal by
cascading with a sufciently long delay.
Consider the system consisting of an accumulator followed by a backward
difference:
18
Accumulator
difference
Backward
The impulse response of this system is
hn| = un| + (n| n 1|) = un| un 1| = n|.
The output is therefore equal to the input because .n| + n| = .n|. Thus the
backward difference exactly compensates for (or inverts) the effect of the
accumulator the backward difference system is the inverse system for the
accumulator, and vice versa. We dene this inverse relationship for all LTI
systems:
hn| + h
i
n| = n|.
5 Linear constant coefcient difference equations
Some LTI systems can be represented in terms of linear constant coefcient
difference (LCCD) equations
1
X
kD0
a
k
,n k| =

X
nD0
b
n
.n m|.
Example: difference equation representation of the accumulator
Take for example the accumulator
x[n]
difference
Backward
Accumulator
x[n] y[n]
Here ,n| ,n 1| = .n|, which can be written in the desired form with
N = 1, a
0
= 1, a
1
= 1, M = 0, and b
0
= 1. Rewriting as
,n| = ,n 1| .n|
19
we obtain the recursion representation
y[n]
Onesample
delay
x[n]
where at n we add the current input .n| to the previously accumulated sum
,n 1|.
Example: difference equation representation of moving average
Consider now the moving average system with M
1
= 0:
hn| =
1
M
2
1
(un| un M
2
1|).
The output of the system is
,n| =
1
M
2
1

2
X
kD0
.n k|.
which is a LCCDE with N = 0, a
0
= 1, and M = M
2
, b
k
= 1,(M
2
1).
Using the sifting property of n|,
hn| =
1
M
2
1
(n| n M
2
1|) + un|
so
x[n]
sample delay

+
y[n]
Attenuator
Accumulator
1,(M
2
1)
.
1
n|
(M
2
1)
20
Here .
1
n| = 1,(M
2
1)(.n| .n M
2
1|) and for the accumulator
,n| ,n 1| = .
1
n|. Therefore
,n| ,n 1| =
1
M
2
1
(.n| .n M
2
1|).
which is again a (different) LCCD equation with N = 1, a
0
= 1, a
1
= 1,
b
0
= b

2
C1
= 1,(M
2
1).
As for constant coefcient differential equations in the continuous case,
without additional information or constraints a LCCDE does not provide a
unique solution for the output given an input. Specically, suppose we have
the particular output ,
]
n| for the input .
]
n|. The same equation then has the
solution
,n| = ,
]
n| ,
h
n|.
where ,
h
n| is any solution with .n| = 0. That is, ,
h
n| is an homogeneous
solution to the homogeneous equation
1
X
kD0
a
k
,
h
n k| = 0.
It can be shown that there are N nonzero solutions to this equation, so a set of
N auxiliary conditions are required for a unique specication of ,n| for a
given .n|.
If a system is LTI and causal, then the initial conditions are initial rest
conditions, and a unique solution can be obtained.
6 Frequency-domain representation of
discrete-time signals and systems
The Fourier transform considered here is strictly speaking the discrete-time
Fourier transform (DTFT), although Oppenheim and Schafer call it just the
21
Fourier transform. Its properties are recapped here (with examples) to show
nomenclature.
Complex exponentials
.n| = e
}on
. o < n < o
are eigenfunctions of LTI systems:
,n| =
1
X
kD1
hk|e
}o(nk)
= e
}on

1
X
kD1
hk|e
}ok
!
.
Dening
H(e
}o
) =
1
X
kD1
hk|e
}ok
we have that ,n| = H(e
}o
)e
}on
= H(e
}o
).n|. Therefore, e
}on
is an
eigenfunction of the system, and H(e
}o
) is the associated eigenvalue.
The quantity H(e
}o
) is called the frequency response of the system, and
H(e
}o
) = H
T
(e
}o
) H
J
(e
}o
) = [H(e
}o
)[e
} ^1(e
j!
)
.
Example: frequency response of ideal delay:
Consider the input .n| = e
}on
to the ideal delay system ,n| = .n n
d
|:
the output is
,n| = e
}o(nn
d
)
= e
}on
d
e
}on
.
The frequency response is therefore
H(e
}o
) = e
}on
d
.
Alternatively, for the ideal delay hn| = n n
d
|,
H(e
}o
) =
1
X
nD1
n n
d
|e
}on
= e
}on
d
.
The real and imaginary parts of the frequency response are
22
H
T
(e
}o
) = cos(on
d
) and H
J
(e
}o
) = sin(on
d
), or alternatively
[H(e
}o
)[ = 1
^H(e
}o
) = on
d
.
The frequency response of a LTI system is essentially the same for continuous
and discrete time systems. However, an important distinction is that in the
discrete case it is always periodic in frequency with a period 2:
H(e
}(oC2t)
) =
1
X
nD1
hn|e
}(oC2t)n
=
1
X
nD1
hn|e
}on
e
} 2tn
=
1
X
nD1
hn|e
}on
= H(e
}o
).
This last result holds since e
} 2tn
= 1 for integer n.
The reason for this periodicity is related to the observation that the sequence

e
}on

. o < n < o
has exactly the same values as the sequence
n
e
}(oC2t)n
o
. o < n < o.
A system will therefore respond in exactly the same way to both sequences.
Example: ideal frequency selective lters
The frequency response of an ideal lowpass lter is as follows:
23
0
1
Only required part

2 2
o
o
c
o
c
H
lp
(e
}o
)
Due to the periodicity in the response, it is only necessary to consider one
frequency cycle, usually chosen to be the range to . Other examples of
ideal lters are:
0
0
0
1
1
1
Highpass
Bandstop
Bandpass

o
o
o
o
o
o
o
o
b
o
b
o
c
o
o
o
o
o
b
o
b
o
c
H
hp
(e
}o
)
H
bs
(e
}o
)
H
bp
(e
}o
)
In these cases it is implied that the frequency response repeats with period 2
outside of the plotted interval.
Example: frequency response of the moving-average system
24
The frequency response of the moving average system
hn| =
8
<
:
1

1
C
2
C1
M
1
_ n _ M
2
0 otherwise
is given by
H(e
}o
) =
1
M
1
M
2
1
e
}o(
2
C
1
C1){2
e
}o(
2
C
1
C1){2
1 e
}o
e

j!.M
2
M
1
C1/
2
=
1
M
1
M
2
1
e
}o(
2
C
1
C1){2
e
}o(
2
C
1
C1){2
e
}o{2
e
}o{2
e

j!.M
2
M
1
/
2
=
1
M
1
M
2
1
sino(M
1
M
2
1),2|
sin(o,2)
e

j!.M
2
M
1
/
2
.
For M
1
= 0 and M
2
= 4,
0
0
1
0
0

2
2
2
2
2t
5

2t
5
o
o
[
H
(
e
}
o
)
[
^
H
(
e
}
o
)
This system attenuates high frequencies (at around o = ), and therefore has
the behaviour of a lowpass lter.
25
7 Fourier transforms of discrete sequences
The discrete time Fourier transform (DTFT) of the sequence .n| is
X(e
}o
) =
1
X
nD1
.n|e
}on
.
This is also called the forward transform or analysis equation. The inverse
Fourier transform, or synthesis formula, is given by the Fourier integral
.n| =
1
2
Z
t
t
X(e
}o
)e
}on
Jo.
The Fourier transform is generally a complex-valued function of o:
X(e
}o
) = X
T
(e
}o
) X
J
(e
}o
) = [X(e
}o
)[e
} ^A(e
j!
)
.
The quantities [X(e
}o
)[ and ^X(e
}o
) are referred to as the magnitude and
phase of the Fourier transform. The Fourier transform is often referred to as
the Fourier spectrum.
Since the frequency response of a LTI system is given by
H(e
}o
) =
1
X
kD1
hk|e
}ok
.
it is clear that the frequency response is equivalent to the Fourier transform of
the impulse response, and the impulse response is
hn| =
1
2
Z
t
t
H(e
}o
)e
}on
Jo.
A sufcient condition for the existence of the Fourier transform of a sequence
.n| is that it be absolutely summable:
P
1
nD1
[.n|[ < o. In other words,
the Fourier transform exists if the sum
P
1
nD1
[.n|[ converges. The Fourier
transform may however exist for sequences where this is not true a rigorous
mathematical treatment can be found in the theory of generalised functions.
26
8 Symmetry properties of the Fourier transform
Any sequence .n| can be expressed as
.n| = .
e
n| .
o
n|.
where .
e
n| is conjugate symmetric (.
e
n| = .

e
n|) and .
o
n| is conjugate
antisymmetric (.
o
n| = .

o
n|). These two components of the sequence
can be obtained as:
.
e
n| =
1
2
(.n| .

n|) = .

e
n|
.
o
n| =
1
2
(.n| .

n|) = .

o
n|.
If a real sequence is conjugate symmetric, then it is an even sequence, and if
conjugate antisymmetric, then it is odd.
Similarly, the Fourier transform X(e
}o
) can be decomposed into a sum of
conjugate symmetric and antisymmetric parts:
X(e
}o
) = X
e
(e
}o
) X
o
(e
}o
).
where
X
e
(e
}o
) =
1
2
X(e
}o
) X

(e
}o
)|
X
o
(e
}o
) =
1
2
X(e
}o
) X

(e
}o
)|.
With these denitions, and letting
X(e
}o
) = X
T
(e
}o
) X
J
(e
}o
).
the symmetry properties of the Fourier transform can be summarised as
follows:
27
Sequence .n| Transform X(e
}o
)
.

n| X

(e
}o
)
.

n| X

(e
}o
)
Re{.n|] X
e
(e
}o
)
Im{.n|] X
o
(e
}o
)
.
e
n| X
T
(e
}o
)
.
o
n| X
J
(e
}o
)
Most of these properties can be proved by substituting into the expression for
the Fourier transform. Additionally, for real .n| the following also hold:
Real sequence .n| Transform X(e
}o
)
.n| X(e
}o
) = X

(e
}o
)
.n| X
T
(e
}o
) = X
T
(e
}o
)
.n| X
J
(e
}o
) = X
J
(e
}o
)
.n| [X(e
}o
)[ = [X(e
}o
)[
.n| ^X(e
}o
) = ^X(e
}o
)
.
e
n| X
T
(e
}o
)
.
o
n| X
J
(e
}o
)
9 Fourier transform theorems
Let X(e
}o
) be the Fourier transform of .n|. The following theorems then
apply:
28
Sequences .n|, ,n| Transforms X(e
}o
), Y(e
}o
) Property
a.n| b,n| aX(e
}o
) bY(e
}o
) Linearity
.n n
d
| e
}on
d
X(e
}o
) Time shift
e
}o
0
n
.n| X(e
}(oo
0
)
) Frequency shift
.n| X(e
}o
) Time reversal
n.n|
dA(e
j!
)
do
Frequency diff.
.n| + ,n| X(e
}o
)Y(e
}o
) Convolution
.n|,n|
1
2t
R
t
t
X(e
}0
)Y(e
}(o0)
)J0 Modulation
Some useful Fourier transform pairs are:
Sequence Fourier transform
n| 1
n n
0
| e
}on
0
1 (o < n < o)
P
1
kD1
2(o 2k)
a
n
un| ([a[ < 1)
1
1oe
j!
un|
1
1e
j!

P
1
kD1
(o 2k)
(n 1)a
n
un| ([a[ < 1)
1
(1oe
j!
)
2
sin(o
c
n)
tn
X(e
}o
) =
8
<
:
1 [o[ < o
c
0 o
c
< [o[ _
.n| =
8
<
:
1 0 _ n _ M
0 otherwise
sino(C1){2j
sin(o{2)
e
}o{2
e
}o
0
n
P
1
kD1
2(o o
0
2k)
29