Scribd
Upload
56 views46 pages

信號與系統Chap. 3 Class Notes

1) Chapter 3 introduces Fourier series representation for periodic signals as an alternative to impulse representation from Chapter 2. 2) Periodic signals can be represented as a linear combination of complex exponentials with harmonically related frequencies. 3) The Fourier series representation provides a way to analyze periodic signals into their frequency components and to synthesize signals from a given set of frequency coefficients.

Uploaded by

wangli
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
56 views46 pages

信號與系統Chap. 3 Class Notes

1) Chapter 3 introduces Fourier series representation for periodic signals as an alternative to impulse representation from Chapter 2. 2) Periodic signals can be represented as a linear combination of complex exponentials with harmonically related frequencies. 3) The Fourier series representation provides a way to analyze periodic signals into their frequency components and to synthesize signals from a given set of frequency coefficients.

Uploaded by

wangli
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 46

Chapter 3 Fourier Series

Representation of Periodic Signals

Chapter 2 is based on representing signals as


linear combinations of shifted impulses.

In this and the following two chapters, we

explore an alternative representation for signals


and LTI systems.

3.2 THE RESPONSE OF LTI SYSTEMS TO


COMPLEX EXPONENTIALS

Represent signals as linear combinations of basic


signals that possess the following two properties:
z The set of basic signals can be used to
construct a broad and useful class of signals.
z The response of an LTI system to each signal
should be simple enough in structure.
With s and z in complex

For x(t) = est

For x[n] = zn

Let

3.3 FOURIER SERIES REPRESENTATION OF


CONTINUOUS-TIME PERIODIC SIGNALS
3.3.1 Linear Combination of Harmonically
Related Complex Exponentials

Harmonically related complex exponential

Linear combination

is also periodic with period T --- F.S. representation

Exp. 3-2 --- x(t) with fundamental freq. 2, Fig.


3.4

--- a real periodic signal

For real x(t)

let k = -k

x(t) = x*(t)

The alternative form of the F.S. for real x(t)

Sub. ak* for a-k

If ak in polar form

If ak in rectangular form

For real periodic function (3.25) = (3.31) = (3.32)

3.3.2 Determination of the Fourier Series


Representation of a Continuous-time Periodic
Signal

Determine ak

(3.38) --- synthesis eq.


(3.39) --- analysis eq.
{ak}: F.S. coeff. or spectral coeff. of x(t)
ao: dc or constant component of x(t)
--- average value of x(t) over one
period

Exp. 3-4 ---

Fig. 3.5

Exp. 3-5 --- Fig. 3.6,


period T

Fig. 3.7 --- fixed T1, various T

with

3.4 CONVERGENCE OF THE FOURIER SERIES

Every continuous periodic signal has a Fourier


series representation.
This is also true for many discontinuous signals.
The Dirichlet conditions --- guarantees that x(t)
equals its Fourier series representation, except
at isolated values of t for which x(t) is
discontinuous.
At these values, the infinite series converges to
the average of the values on either side of the
discontinuity.

Condition 1: Over any period, x(t) must be


absolutely integrable.

Fig. 3.8(a) ---

Condition 2: In any finite interval of time, x(t) is of


bounded variation.
z

Fig. 3.8 (b) ---

Condition 3 --- In any finite interval of time,

there are only a finite number of discontinuities.


z

Fig. 3.8 (c) ---

Gibbs phenomenon --- Fig. 3.9


1. xN(t) has the average value of the discontinuity
2. for any other value of t = t1,

3. constant 9% overshot

3.5 PROPERTIES OF CONTINUOUS-TIME


FOURIER SERIES

3.5.1 Linearity

3.5.2 Time Shifting


y(t) = x(t to)
let = t to

i.e.

When a periodic signal is shifted in time, the


magnitudes of its Fourier series coefficients
remain unaltered, i.e., |ak| = |bk|.

3.5.3 Time Reversal


y ( t) = x ( - t)

let k = -m

Time reversal applied to a continuous-time signal


results in a time reversal of the corresponding
sequence of F.S. coefficients.
If x(t) is even
x(-t) = x(t)
a-k = ak
ak is even
ak is odd
x(t) is odd
x(-t) = -x(t) a-k = -ak

3.5.4 Time Scaling

If x(t) is periodic with period T, then x(t), where


is a positive real number, is periodic with
period T/.

The Fourier coefficients have not changed, the


Fourier series representation has changed.
Because the fundamental frequency has
changed.

3.5.5 Multiplication

x(t) and y(t) are both periodic with period T.

discrete-time convolution

3.5.6 Conjugation and Conjugate Symmetry

Take * on both side and let k = -m

x(t) real
x(t) = x*(t)
a-k = ak*
x(t) real and even
a-k = ak*
ak = ak*
ak = a-k
ak = a-k
ak real and even
x(t) real and odd
ak pure imaginary and odd

3.5.7 Parsevals Relation for Continuous-Time


Periodic Signals

average power

|ak|2 : average power in the kth harmonic


component of x(t)
The total average power in a periodic signal
equals the sum of the average powers in all of
its harmonic components.
Tab. 3.1 --- summary of properties

Exp. 3.6 --- from Exp. 3-5, T = 4, T1 = 1


Fig. 3.10 --x(t)
ak
x(t 1)
-1/2
g(t)

Exp. 3.9 --- x(t) = ?


1. x(t) is a real signal.
2. x(t) is periodic with period T = 4, and it has
F.S. coefficients ak.
3. ak = 0 for k > 1.
4. The signal with Fourier coefficients bk = e-jk/2a-k
is odd.
5. 14 x(t ) 2 dt = 1/2
4

from (3)
ak = a-1, ao, a1
from (2), T = 4
from (1)
a1 = a-1*

from (4),
x(-t)
a-k
signal shift by 1 to the right (i.e., replace by t -1)
akeik/2
x(-(t 1)) = x(-t + 1)
must be odd
x(t) real

x(-t + 1) is real

bo = 0, b-1 = - b1

bk = e-jk/2a-k

From (5),
average power for x(t) = that of x(-t +1)

|b1| = , and b1 has to be purely


b 1 = -b- 1
imaginary b1 = j/2 or j/2
bo = 0

ao = 0

a1 = e-j/2b1 = -jb-1 = jb1


=
-1/2
or
1/2
x(t) = -cos (t/2) or

cos(t/2)

3.6 FOURIER SERIES REPRESENTATION OF


DISCRETE-TIME PERIODIC SIGNALS

A finite series, no issues of convergence

3.6.1 Linear Combinations of Harmonically


Related Complex Exponentials

Harmonically related but only N distinct signals


in the set, i.e.,

3.6.2 Determination of the Fourier Series


Representation of a Periodic Signal

the sum over one period of the values of a


periodic complex exponential is zero, unless that
complex exponential is a constant.

= 0, unless k r = 0 or an
integer multiple of N

choose k and r over same range, k = r and

(3.94) --- synthesis eq.


(3.95) --- analysis eq.

Take k = 0,, N 1

Take k = 1,, N

fo[n] = fN[n]

ak repeat periodically with period N


ak is a sequence defined for all values of k, but
where only N successive elements in the
sequence will be used in the Fourier series
representation.

ao = aN

Exp. 3.10 --- x[n] = sinon

Fig. 3.13 --- N = 5

Exp. 3.11 ---

Fig. 3.15 --- N = 10

Exp. 3.12 --- Fig. 3.16, x[n] = 1, for N1 < n < N1


let m = n + N1

Geometric series

Fig. 3.17 --- ak for 2N1 + 1 = 5, N1 = 2, N = 10,


20, 40
Convergence ? --- Fig. 3.18

N = 9, 2N1 + 1 = 5
for M = 4
x[n] = x[n]
There are no convergence issues and no Gibbs
phenomenon

3.7 PROPERTIES OF DISCRETE-TIME FOURIER


SERIES

Tab. 3.2

3.7.1 Multiplication
both periodic with period N

periodic convolution

3.7.2 First Difference

--- Time shifting and linearity


3.7.3 Parseval's Relation for Discrete-Time
Periodic Signals

the average power in a periodic signal equals


the sum of the average powers in all of its
harmonic components.

3.8 FOURIER SERIES AND LTI SYSTEMS

Cont t
x(t) = est

y(t) = H(s)est

focus on Re{s} = 0, i.e., s = j


Disc t
x[n] = zn
y[n] = H(z)zn

focus on |z| = 1, i.e., z = ej

Frequency
response

Take cont t as an example

F.S. coefficient for output

The effect of the LTI system is to modify


individually each of the Fourier coefficients of the
input through multiplication by the value of the
frequency response at the corresponding
frequency.

Exp. 3.16 --- input Fig. 3.4 (Exp. 3.2) to an LTI


sys.

o = 2

bk = akH(jko) = akH(jk2)

1/(1+jko)

In disc - t

3.9 FILTERING
3.9.1 Frequency-Shaping Filters

Fig. 3.22 --- dB (decibel) = 20 log|H(j)|


Fig. 3.23 --- y(t) = dx(t)/dt
x(t) = ejt

y(t) = jejt

used to enhance rapid variations or transitions of


signals.
Fig. 3.24 --- enhance edge in picture processing
x(t1, t2) --- brightness of the image

Fig. 3.25 --- disc-t filters

h[n] = ([n] + [n 1])

e.g. stock market average

3.9.2 Frequency-Selective Filters

Select some bands and reject others


Lowpass filter
Highpass filter
Bandpass filter
Cutoff frequencies End of Chap. 2
Passband, stopband
Fig. 3.26 --- ideal lowpass
Fig. 3.27 --- ideal highpass
ideal bandpass

End of Chap. 3

You might also like