Discrete-Time Fourier Series Concept
A signal can be represented as a linear combination of sinusoids.
Discrete-Time Fourier Methods
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
1
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
Discrete-Time Fourier Series Concept
2
Discrete-Time Fourier Series Concept
The relationship between complex and real sinusoids
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
3
Discrete-Time Fourier Series Concept
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
5
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
4
Discrete-Time Fourier Series Concept
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
6
1
Discrete-Time Fourier Series Concept
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
7
Discrete-Time Fourier Series Concept
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
∑
c x [ k ] e j 2 π kn/N
cx [ k ] =
k= N
1 N
∑
∑
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
10
A periodic discrete-time signal can be expressed as x[n] =
x [ n ] e− j 2 π kn/N
1 N
∑ X[ k ]e
k= N
j 2 π kn/N
X[ k ] =
n0 + N −1
∑ x[n]e
− j 2 π kn/N
n=n0
where X [ k ] is the DFT harmonic function, N is any period of x [ n ]
n=n0
and the notation,
means a summation over any range of
∑
means a summation over any range of
k= N
k= N
consecutive k’s exactly N in length. The main difference between the
consecutive k’s exactly N in length.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
Discrete-Time Fourier Series Concept
The discrete Fourier transform (DFT) is almost identical to the DTFS.
where c x [ k ] is the harmonic function, N is any period of x [ n ] and the notation,
8
The Discrete Fourier Transform
The discrete-time Fourier series (DTFS) is similar to the CTFS. A periodic discrete-time signal can be expressed as x[n] =
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
9
The Discrete-Time Fourier Series
n0 + N −1
Discrete-Time Fourier Series Concept
DTFS and the DFT is the location of the 1/N term. So X [ k ] = N c x [ k ].
11
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
12
2
The Discrete Fourier Transform
The Discrete Fourier Transform
Because the DTFS and DFT are so similar, and because the DFT is so widely used in digital signal processing (DSP), we will concentrate on the DFT realizing we can always form the DTFS from c x [ k ] = X [ k ] / N.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
13
DFT Example
)
x ⎡⎣ n ⎤⎦ = u ⎡⎣ n ⎤⎦ − u ⎡⎣ n − 3⎤⎦ ∗ δ 5 ⎡⎣ n ⎤⎦ using its fundamental period as the representation time. X ⎡⎣ k ⎤⎦ =
∑
n= N 4
x ⎡⎣ n ⎤⎦ e
2
(
= ∑e n=0 2
n=0
∑ X[ k ]e
j 2 π kn/N
so we can find x [ n ]
k= N
function from x [ n ]? We use the principle of orthogonality like
− j 2 π kn/5
we did with the CTFS except that now the orthogonality is in discrete time instead of continuous time.
1− e− j6π k /5 e− j3π k /5 e j3π k /5 − e− j3π k /5 = = × jπ k /5 − jπ k /5 1− e− j 2π k /5 e− jπ k /5 e −e
= ∑ e− j 2π kn/5 = e− j 2π k /5
1 N
from its harmonic function. But how do we find the harmonic
)
− j 2 π kn/5
We know that x [ n ] =
− j 2 π kn/ N
= ∑ u ⎡⎣ n ⎤⎦ − u ⎡⎣ n − 3⎤⎦ ∗ δ 5 ⎡⎣ n ⎤⎦ e n=0
14
The DFT Harmonic Function
Find the DFT harmonic function for
(
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
( (
) = 3e )
sin 3π k / 5 sin π k / 5
− j 2 π k /5
(
)
drcl k / 5,3
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
15
The DFT Harmonic Function
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
16
The DFT Harmonic Function
17
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
18
3
The DFT Harmonic Function
The DFT Harmonic Function
Below is a set of complex sinusoids for N = 8. They form a set of basis vectors. Notice that the k = 7 complex sinusoid rotates counterclockwise through 7 cycles but appears to rotate
The projection of a real vector x in the direction of another real vector y is xT y y yT y If p = 0, x and y are orthogonal. If the vectors are complexvalued
clockwise through one cycle. The k = 7 complex sinusoid is exactly the same as the k = −1 complex sinusoid. This must be true because the DFT is periodic with period N.
p=
p=
xH y y yH y
where the x H is the complex-conjugate transpose of x. xT y and x H y are the dot product of x and y. M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
19
The DFT Harmonic Function
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
20
The DFT Harmonic Function
21
The DFT Harmonic Function
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
22
The DFT Harmonic Function
23
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
24
4
The DFT Harmonic Function
Convergence of the DFT
The most common definition of the DFT is N −1
X [ k ] = ∑ x [ n ] e− j 2 π kn/N ,
x[n] =
n=0
1 N
∑ X[ k ]e
j 2 π kn/N
• The DFT converges exactly with a finite number of terms. It does not have a Gibbs phenomenon in the same sense that the CTFS does
k= N
Here the beginning point for x [ n ] is taken as n0 = 0 . This is the form of the DFT that is implemented in practically all computer languages.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
25
The Discrete Fourier Transform
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
26
DFT Properties
X [ k ] is called the DFT harmonic function of x [ n ] and k is the
harmonic number just as we have seen in the CTFS. x [ n ] and X [ k ] form a DFT pair based on N points. D FT x [ n ] ←⎯⎯ → X[ k ] N
1 From x [ n ] = N
∑ X[ k ]e
j 2 π kn/N
we see that x [ n ] is formed
k= N
by a linear combination of functions of the form e j 2 π kn/N each of which has a period N. Therefore x [ n ] must also be periodic with period (but not necessarily fundamental period) N.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
27
DFT Properties
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
28
DFT Properties
It can be shown (and is in the text) that if x [ n ] is an even
function, X [ k ] is purely real and if x [ n ] is an odd function X [ k ] is purely imaginary.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
29
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
30
5
The Dirichlet Function
The functional form
DFT Pairs
D FT e j 2 π n/N ←⎯⎯ → mNδ mN [ k − m ] mN
D FT cos ( 2π qn / N ) ←⎯⎯ → ( mN / 2 ) (δ mN [ k − mq ] + δ mN [ k + mq ]) mN
sin (π Nt ) N sin (π t )
D FT sin ( 2π qn / N ) ←⎯⎯ → ( jmN / 2 ) (δ mN [ k + mq ] − δ mN [ k − mq ]) mN
appears often in discrete-time
D FT δ N [ n ] ←⎯⎯ → mδ mN [ k ] mN
signal analysis and is given the special name Dirichlet function. That is drcl ( t, N ) =
D FT 1←⎯⎯ → Nδ N [ k ] N
( u [ n − n ] − u [ n − n ]) ∗ δ [ n ]←⎯⎯→ e 0
sin (π Nt ) N sin (π t )
31
γ N = 2γ 1 2 2 4 3 8 4 16 5 32 6 64 7 128 8 256 9 512 10 1024
(Acquire the input data in an array x with N elements.)
. Initialize the DFT array to a column vector of zeros.
X = zeros(N,1) ; Compute the Xn’s in a nested, double for loop.
for k = 0 :N - 1 for n = 0 :N - 1 X(k + 1) = X(k + 1) + x(n + 1) * exp(-j * 2 * pi * n *k / N) ; end end
ADFT 2
M DFT 4
AFFT 2
M FFT 1
ADFT / AFFT 1
M DFT / M FFT 4
12 56 240
16 64 256
8 24 64
4 12 32
1.5 2.33 3.75
4 5.33 8
992 4032 16256 65280 261632
1024 4096 16384 65536 262144
160 384 896 2048 4608
80 192 448 1024 2304
6.2 10.5 18.1 31.9 56.8
12.8 21.3 36.6 64 113.8
1047552 1048576 10240
5120
102.3
204.8
33
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
34
Generalizing the DFT for Aperiodic Signals
Generalizing the DFT for Aperiodic Signals
Pulse Train
DFT of
Pulse Train
As the period of the rectangular wave increases, the period of the DFT increases
This periodic rectangular-wave signal is analogous to the
continuous-time periodic rectangular-wave signal used to
illustrate the transition from the CTFS to the CTFT.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
32
The Fast Fourier Transform
One could write a MATLAB program to implement the DFT.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
There is a function in MATLAB fft which accomplishes the same goal and is typically much faster. This table compares the speeds of the two methods. M stands for computer multiplies and A stands for computer additions.
.
%
D FT N
D FT sinc ( n / w ) ∗ δ N [ n ] ←⎯⎯ → wrect ( wk / N ) ∗ δ N [ k ] N
The Fast Fourier Transform
%
N
D FT N
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
%
− jπ k ( n1 +n0 ) /N
( n1 − n0 ) drcl ( k / N, n1 − n0 ) e− jπ k /N 2 tri ( n / N w ) ∗ δ N [ n ] ←⎯⎯→ N w drcl ( k / N, N w ) , N w an integer 1
35
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
36
6
Generalizing the DFT for Aperiodic Signals
Generalizing the DFT for Aperiodic Signals
Normalized
DFT of
Pulse Train
The normalized DFT approaches this limit as the
period approaches infinity.
By plotting versus k / N 0 instead of k, the period of the normalized DFT stays at one.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
37
Definition of the Discrete-Time Fourier Transform (DTFT)
Forward x[n] = ∫ X( F ) e
F Form j 2 π Fn
1
1 2π
∫
2π
The function e− jΩ appears in the forward DTFT raised to the nth power. It is periodic in Ω with fundamental period 2π . n is an integer. Therefore
∞
∑ x[n]e
e- jΩn is periodic with fundamental period 2π / n and 2π is also a period
− j 2 π Fn
of e− jΩn . The forward DTFT is
n=−∞
Ω Form
Forward x[n] =
dF ←⎯→ X ( F ) =
( )
38
The Discrete-Time Fourier Transform
Inverse
F
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
∞
( ) ∑ x[n]e
X e jΩ =
Inverse
− jΩn
n=−∞
∞
( ) ∑ x[n]e
F X e jΩ e jΩn dΩ ←⎯ → X e jΩ =
a weighted summation of functions of the form e− jΩn , all of which repeat with
− jΩn
( )
every 2π change in Ω. Therefore X e jΩ is always periodic in Ω with period
n=−∞
2π . This also implies that X ( F ) is always periodic in F with period 1.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
39
DTFT Pairs
By generalizing the CTFT to include transform that have impulses we were able to find CTFT's of some important practical functions. The same is true of the DTFT. The DTFT of a constant
F δ [ n ] ←⎯ →1
e jΩ 1 e jΩ 1 F = , α 1 e − α 1 − α e− jΩ e − α 1 − α e− jΩ e jΩα sin ( Ω0 ) e jΩα sin ( Ω0 ) F F n n α sin ( Ω0 n ) u [ n ] ←⎯→ j 2Ω , α < 1 , − α sin ( Ω0 n ) u [ −n − 1] ←⎯→ j 2Ω , α >1 jΩ 2 jΩ e − 2α e cos ( Ω0 ) + α e − 2α e cos ( Ω0 ) + α 2 F α n cos ( Ω0 n ) u [ n ] ←⎯ →
jΩ
e jΩ ⎡⎣ e jΩ − α cos ( Ω0 ) ⎤⎦
e
j 2Ω
− 2α e cos ( Ω0 ) + α jΩ
2
F , α < 1 , − α n cos ( Ω0 n ) u [ −n − 1] ←⎯ →
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
40
The Generalized DTFT
We can begin a table of DTFT pairs directly from the definition. (There is a more extensive table in the text.)
F α n u [ n ] ←⎯ →
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
e jΩ ⎡⎣ e jΩ − α cos ( Ω0 ) ⎤⎦
e
j 2Ω
− 2α e cos ( Ω0 ) + α jΩ
2
, α >1
41
X(F ) =
∞
∑ Ae
n=−∞
− j 2 π Fn
∞
= A ∑ e− j 2 π Fn n=−∞
does not converge. The CTFT of a constant turned out to be an impulse. Since the DTFT must be periodic that cannot the the transform of a constant in discrete time. Instead the transform must be a periodic impulse.
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
42
7
The Generalized DTFT
The Generalized DTFT
Now consider the function X ( F ) = Aδ1 ( F − F0 ) , − 1 / 2 < F0 < 1 / 2. Its inverse DTFT is
To shorten the process, find the inverse DTFT of a periodic
x [ n ] = ∫ Aδ1 ( F − F0 ) e j 2 π Fn dF = A
impulse of the form Aδ1 ( F ) . Using the formula x [ n ] = ∫ Aδ1 ( F ) e j 2 π Fn dF = A 1
1/2
∫ δ (F)e
1
j 2 π Fn
1/2
∫ δ (F − F )e
j 2 π Fn
0
−1/2
dF = Ae j 2 π F0 n
(
)
Now change x [ n ] to x [ n ] = A cos ( 2π F0 n ) = ( A / 2 ) e j 2 π F0 n + e− j 2 π F0 n .
dF = A
−1/2
Then
proving that the DTFT of a constant A is Aδ1 ( F ) or, in radian-
F A cos ( 2π F0 n ) ←⎯ → ( A / 2 ) ⎡⎣δ1 ( F − F0 ) + δ1 ( F + F0 ) ⎤⎦
F frequency form A ←⎯ → 2π Aδ 2 π ( Ω ) .
or F A cos ( Ω0 n ) ←⎯ → π A ⎡⎣δ 2 π ( Ω − Ω0 ) + δ 2 π ( Ω + Ω0 ) ⎤⎦
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
43
Forward DTFT Example
∞
∑(
n = −∞
)
u [ n − n0 ] − u [ n − n1 ] e− j 2 π Fn =
F u [ n − n0 ] − u [ n − n1 ] ←⎯ → e− j 2 π Fn0
n1 −1
∑ e− j 2π Fn
u [ n − n0 ] − u [ n − n1 ] ←⎯→
n1 − n0 −1
∑
e
− j 2 π F ( m + n0 )
=e
− j 2 π Fn0
m=0
n1 − n0 −1
∑
e
e− jπ F ( n1 − n0 ) e jπ F ( n1 − n0 ) − e− jπ F ( n1 − n0 ) e − jπ F e jπ F − e − jπ F By the definition of the sine function in terms of complex exponentials F u [ n − n0 ] − u [ n − n1 ] ←⎯ → e− j 2 π Fn0
− j 2 π Fm
m=0
Summing this geometric series
F u [ n − n0 ] − u [ n − n1 ] ←⎯ →
1 − e− j 2 π F ( n1 − n0 ) u [ n − n0 ] − u [ n − n1 ] ←⎯→ e− j 2 π Fn0 1 − e− j 2 π F F
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
45
u [ n − n0 ] − u [ n − n1 ] ←⎯→
(
sin π F ( n1 − n0 ) sin (π F )
)
)
46
More DTFT Pairs
We can now extend the table of DTFT pairs. F δ [ n ] ←⎯ →1
1 + (1 / 2 )δ1 ( F ) 1 − e− j 2 π F F sinc ( n / w ) ←⎯ → w rect ( wF ) ∗ δ1 ( F ) F u [ n ] ←⎯ →
, n0 + n1 = 1
, ,
F tri ( n / w ) ←⎯ → w drcl2 ( F, w )
F tri ( n / w ) ←⎯ → w drcl2 ( Ω / 2π , w ) F 1←⎯ → 2πδ 2 π ( Ω )
,
F δ N0 [ n ] ←⎯ → (1 / N 0 )δ1/ N0 ( F )
F δ N0 [ n ] ←⎯ → ( 2π / N 0 )δ 2 π / N0 ( Ω )
,
F cos ( 2π F0 n ) ←⎯ → (1 / 2 ) ⎡⎣δ1 ( F − F0 ) + δ1 ( F + F0 ) ⎤⎦
sin (π wf ) πf The DTFT is a periodically-repeated sinc function.
1 F u [ n ] ←⎯ → + πδ1 ( Ω ) 1 − e− jΩ F sinc ( n / w ) ←⎯ → w rect ( wΩ / 2π ) ∗ δ 2 π ( Ω )
,
F 1←⎯ → δ1 ( F )
Compare this to the CTFT of a rectangular pulse of width w centered at t = 0.
F cos ( Ω0 n ) ←⎯ → π ⎡⎣δ 2 π ( Ω − Ω0 ) + δ 2 π ( Ω + Ω0 ) ⎤⎦
,
F sin ( 2π F0 n ) ←⎯ → ( j / 2 ) ⎡⎣δ1 ( F + F0 ) − δ1 ( F − F0 ) ⎤⎦ ,
F rect ( t / w ) ←⎯ → w sinc ( wf ) =
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
(
e− jπ F ( n0 + n1 ) sin π F ( n1 − n0 ) e − jπ F sin (π F )
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
Forward DTFT Example
Consider the special case of n0 + n1 = 1 ⇒ n0 = 1 − n1 (making the function a discrete-time rectangular pulse of width 2n0 +1 centered at n = 0. F
1 − e− j 2 π F ( n1 − n0 ) 1 − e− j 2 π F
Factor out e− jπ F ( n1 − n0 ) from the numerator and e− jπ F from the denominator
n = n0
Let m = n − n0 . Then F
44
Forward DTFT Example
Find the forward DTFT of x [ n ] = u [ n − n0 ] − u [ n − n1 ]. F u [ n − n0 ] − u [ n − n1 ] ←⎯ →
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
F u [ n − n0 ] − u [ n − n1 ] ←⎯ → F u [ n − n0 ] − u [ n − n1 ] ←⎯ →
47
F sin ( Ω0 n ) ←⎯ → jπ ⎡⎣δ 2 π ( Ω + Ω0 ) − δ 2 π ( Ω − Ω0 ) ⎤⎦
e j 2π F e− jπ F ( n0 + n1 ) e− j 2 π n0 F − e− j 2 π n1 F = ( n1 − n0 ) drcl ( F, n1 − n0 ) e −1 e − jπ F
(
j 2π F
)
e jΩ e− jΩ( n0 + n1 ) /2 e− jn0 Ω − e− jn1Ω = ( n1 − n0 ) drcl (Ω / 2π , n1 − n0 ) e jΩ − 1 e− jΩ /2
(
)
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
48
8
DTFT Properties
DTFT Properties
n
F x [ n − n0 ] ←⎯ → e− j 2 π Fn0 X ( F ) ,
(
F e j 2 π F0 n x [ n ] ←⎯ → X ( F − F0 ) ,
F e jΩ0 n x [ n ] ←⎯ → X e j (Ω−Ω0 )
(
F x* [ n ] ←⎯ → X* ( −F ) ,
(
)
F x [ n ] − x [ n − 1] ←⎯ → 1 − e− j 2 π F X ( F ) ,
∞
∑e
) x [ n ] − x [ n − 1] ←⎯→ (1 − e ) X ( e )
∑ x[n]
n=−∞
(Note:
49
DTFT Properties
F
j0
− jΩ
2π
− jΩ
F
,
jΩ
jΩ
F
∑e
= δ1 ( F ) ,
jΩ see note
jΩ
∞
n=−∞ ∞
jΩ
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
j 2 π Fn
X e jΩ
∑ x [ m ]←⎯→ 1 − e
m=−∞
F
see note
)
( ) + π X ( e )δ ( Ω ) x [ −n ] ←⎯→ X ( e ) x [ n ] ∗ y [ n ] ←⎯→ X ( e ) Y ( e ) x [ n ] y [ n ] ←⎯→ (1 / 2π ) X ( e ) Y ( e ) n
1 X ( 0 )δ1 ( F ) , 2
F x [ n ] y [ n ] ←⎯ → X ( F ) Y( F )
(
− jΩ
+
F x [ n ] ∗ y [ n ] ←⎯ → X ( F ) Y( F ) ,
F x* [ n ] ←⎯ → X* e− jΩ F
− j 2π F
F x [ −n ] ←⎯ → X ( −F ) ,
)
⎧x [ n / m ] , n / m an integer ⎫ F F jmΩ If y [ n ] = ⎨ ⎬ then y [ n ] ←⎯→ X ( mF ) or y [ n ] ←⎯→ X e , otherwise ⎩0 ⎭
F
m=−∞
( )
F x [ n − n0 ] ←⎯ → e− jΩn0 X e jΩ
X(F )
∑ x [ m ]←⎯→ 1 − e
F F α x [ n ] + β y [ n ] ←⎯ → α X ( F ) + β Y ( F ) , α x [ n ] + β y [ n ] ←⎯ → α X ( e jΩ ) + β Y ( e jΩ )
jΩn
= 2πδ 2 π ( Ω )
n=−∞ 2
= ∫ X ( F ) dF , 2
1
∞
∑ x[n]
2
n=−∞
= (1 / 2π ) ∫
2π
( )
X e jΩ
x ( t ) y ( t ) = ∫ x (τ ) y ( t − τ ) dτ where T is a period of both x and y T
2
dΩ
)
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
50
DTFT Properties
Find and graph the inverse DTFT of X ( F ) = ⎡⎣ rect ( 50 ( F − 1 / 4 )) + rect ( 50 ( F + 1 / 4 )) ⎤⎦ ∗ δ1 ( F )
Frequency shifting the other direction
F e− jπ n/2 (1 / 50 ) sinc ( n / 50 ) ←⎯ → rect ( 50 ( F + 1 / 4 )) ∗ δ1 ( F )
Start with F sinc ( n / w ) ←⎯ → wrect ( wF ) ∗ δ1 ( F )
Combining the last two results and using cos ( x ) =
In this case w = 50.
F → (1 / 25 ) sinc ( n / 50 ) cos (π n / 2 ) ←⎯ ⎡⎣ rect ( 50 ( F − 1 / 4 )) + rect ( 50 ( F + 1 / 4 )) ⎤⎦ ∗ δ1 ( F )
F → rect ( 50F ) ∗ δ1 ( F ) (1 / 50 ) sinc ( n / 50 ) ←⎯
Then, using the frequency-shifting property F e j 2 π F0 n x [ n ] ←⎯ → X ( F − F0 )
e
jπ n/2
e jx + e− jx 2
F → rect ( 50 ( F − 1 / 4 )) ∗ δ1 ( F ) (1 / 50 ) sinc ( n / 50 ) ←⎯
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
51
DTFT Properties
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
52
DTFT Properties
Time scaling in discrete time is quite different from time scaling in continuous-time. Let z [ n ] = x [ an ]. If a is not an
Time scaling does not work for time compression because of
integer some values of z [ n ] are undefined and a DTFT cannot
decimation. But it does work for a special type of time expansion. ⎧x [ n / m ] , n / m an integer Let z [ n ] = ⎨ . Then Z ( F ) = X ( mF ) . , otherwise ⎩0 So the time-scaling property of the DTFT is
be found for it. If a is an integer greater than one, some values of x [ n ] will not appear in z [ n ]
⎧x [ n / m ] , n / m an integer z[n] = ⎨ , , otherwise ⎩0
because of decimation and there cannot be a unique relationship between their DTFT’s M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
53
F z [ n ] ←⎯ → X ( mF )
(
F or z [ n ] ←⎯ → X e jmΩ
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
)
54
9
DTFT Properties
DTFT Properties
Find the signal energy of x [ n ] = (1 / 5 ) sinc ( n /100 ) . The
In the time domain, the response of a system is the convolution
straightforward way of finding signal energy is directly from
of the excitation with the impulse response
the definition E x =
y[n] = x[n] ∗ h[n] In the frequency domain the response of a system is the product of the excitation and the frequency response of the system
( )
∞
∑ x[n]
2
.
n=−∞
Ex =
∞
∑ (1 / 5 ) sinc ( n /100 )
n=−∞
( ) ( )
2
∞
= (1 / 25 ) ∑ sinc 2 ( n /100 ) n=−∞
In this case we run into difficulty because we don't know how to sum this series.
Y e jΩ = X e jΩ H e jΩ
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
55
DTFT Properties
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
56
Transform Method Comparisons
We can use Parseval's theorem to find the signal energy from the DTFT of the signal. F → 20rect (100F ) ∗ δ1 ( F ) (1 / 5 ) sinc ( n /100 ) ←⎯
Parseval's theorem is ∞
∑ x[n]
n=−∞
2
= ∫ X ( F ) dF 2
1
For this case E x = ∫ 20rect (100F ) ∗ δ1 ( F ) dF = 2
1
1/2
∫
20rect (100F ) dF 2
−1/2
1/200
E x = 400
∫
dF = 4
−1/200 M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
57
Transform Method Comparisons
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
58
Transform Method Comparisons
59
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
60
10
Transform Method Comparisons
Transform Method Comparisons
Using the equivalence property of the impulse and fact that both
The DFT can often be used to find the DTFT of a signal. The
e j 2 π F and δ1 ( F ) have a fundamental period of one,
DTFT is defined by X ( F ) =
δ ( F − 1 /12 ) − jπ /6 δ1 ( F + 1 /12 ) ⎤ ⎡ Y ( F ) = (1 / 2 ) ⎢ e jπ /6 1 jπ /6 +e e − 0.9 e− jπ /6 − 0.9 ⎥⎦ ⎣ Finding a common denominator and simplifying, Y ( F ) = (1 / 2 )
− j 2 π Fn
and the DFT
N −1
is defined by X [ k ] = ∑ x [ n ] e− j 2 π kn/N . If the signal x [ n ] is causal n=0
δ1 ( F − 1 /12 ) (1 − 0.9e jπ /6 ) + δ1 ( F + 1 /12 ) (1 − 0.9e− jπ /6 )
and time limited, the summation in the DTFT is over a finite range of n values beginning with 0 and we can set the value of N by letting N − 1 be the last value of n needed to cover that finite
1.81 − 1.8 cos (π / 6 )
Y ( F ) = 0.4391 ⎡⎣δ1 ( F − 1 /12 ) + δ1 ( F + 1 /12 ) ⎤⎦
N −1
range. Then X ( F ) = ∑ x [ n ] e− j 2 π Fn . Now let F → k / N yielding
+ j0.8957 ⎡⎣δ1 ( F + 1 /12 ) − δ1 ( F − 1 /12 ) ⎤⎦
n=0
y [ n ] = 0.8782 cos ( 2π n /12 ) + 1.7914 sin ( 2π n /12 )
N −1
X ( k / N ) = ∑ x [ n ] e− j 2 π kn/N = X [ k ]
y [ n ] = 1.995 cos ( 2π n /12 − 1.115 )
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
∞
∑ x[n]e
n=−∞
n=0
61
Transform Method Comparisons
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
62
Transform Method Comparisons
We can also use the DFT to approximate the inverse DTFT. The inverse DTFT is defined by x [ n ] = ∫ X ( F ) e j 2 π Fn dF
The result N −1
X( k / N ) = ∑ x[n]e
1
− j 2 π kn/N
= X[ k ]
and the inverse DFT is defined by x [ n ] =
n=0
is the DTFT of x [ n ] at a discrete set of frequencies F = k / N or
We can approximate the inverse DTFT by
Ω = 2π k / N. If that resolution in frequency is not sufficient, N
N −1 ( k+1) /N
can be made larger by augmenting the previous set of x [ n ] values
x[n] ≅ ∑
with zeros. That reduces the space between frequency points thereby increasing the resolution. This technique is called zero
x[n] ≅ ∑ X( k / N )
k=0
∫
k /N
N −1
padding.
63
k=0
k /N
∫
e j 2 π Fn dF
1 N −1 ∑ X ( k / N ) e j 2π kn/N N k=0
M. J. Roberts - All Rights Reserved. Edited by Dr. Robert Akl
64
Transform Method Comparisons
For n