訊號與系統 updated 4

Post on 12-Feb-2018

222 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/23/2019 Updated 4

    1/55

    1

    (Signals) (information) (data)

    (

    )

    (Systems)

    ( inputs) ( outputs)

    1-1

    1.1.1

    1.

    (continuous-time signals,

    1(a)) (discrete-time signals,

    1(b))

    1.

  • 7/23/2019 Updated 4

    2/55

    2. (a) (b) (c) (d)

    2.

    ()

    (analog signals) (

    ) (digital signals) ( 2)

    3.

    (deterministic signals)

    (stochastic or random signals)

    4.

    g(t) T0

    0( ) ( ),= + "g t g t T t (1)

    g(t) (periodic signals) (aperiodic signal)(1)

    T0g(t)

  • 7/23/2019 Updated 4

    3/55

    5.

    (energy) (power)

    g(t)

    2

    ( ) .

    -= gE g t dt (2)

    /2 2

    /2

    1lim ( ) .

    -=

    T

    gT T

    P g t dtT

    (3)

    (energy signals)

    (power signals)(2)(3)

    ()

  • 7/23/2019 Updated 4

    4/55

    3.

    1.1.2

    1. (Time Shifting)

    g(t)g(t T) T > 0 () T< 0

    ()

    2.

    (Time Scaling)

    ( 3)

  • 7/23/2019 Updated 4

    5/55

    3.

    (Time Reversal)

    g(t)g(t)

    4.

    g(t)g(t) = g(t)g(t) (odd function)g(t) =g(t)g(t)

    (even function)

    [ ] [ ]1 1

    ( ) : ( ) ( ) , ( ) : ( ) ( ) .2 2

    o eg t g t g t g t g t g t= - - = + -

    go(t)g(t)ge(t)g(t)

    1.1.3

    (a)

    (Sinusoidal Signal){ }0 0( 2 )0 0 0 0) cos(2 ) Re( ) cos( .j fx t A t A f t Ae p qw q p q ++ = + ==

    (b) (Rectangular Pulse)

    elsewhere

    121,( )

    0,( )x t

    tt

    = =-

  • 7/23/2019 Updated 4

    6/55

    1.1.4 (Phasor Signals and Spectra)

    A useful periodic signal in system analysis is the signal:

    0( )( ) ,j tx t Ae tw q+= - < <

    which is characterized by three parameters, amplitudeA, phase qin radians, and frequency

    w0in radians per second orf0= w0/2pHz. We refer to ( )x t as a rotating phasor to distin-

    guish it from the phasor jAe q , for which is 0j te w implicit. Using Eulers theorem, we may

    readily show that 0( ) ( )x t x t T = + , where 00 /2T p w= . Thus ( )x t is a periodic signal withperiod 02 /p w .

    The rotating phasor 0( )j tAe w q+ can be related to a real, sinusoidal signal 0 )cos(A tw q+ in two ways:

    1. 0 )0 ) Re( ( )) Re( ),( ) cos( j tt xA tx t Ae w qw q ++ = ==

    2. 0 0( ) ( )01 1 1 1

    ) ( ) (( ) cos( )2 2 2 2

    j t j tt x t x t Aex t A Aew q w qw q + - +*+ = + == +

    Frequency-Domain Representation of Signal: Line Spectra

  • 7/23/2019 Updated 4

    7/55

    1-2 (Signal, Singularity Functions)

    4.

    1.2.1 (Unit Impulse Function)

    ( )td P.A.M. Dirac (Dirac delta function)

    ( ) 0, 0

    ( ) 1

    t t

    t dt

    d

    d

    -

    =

    =

    (singularity

    function) (

    4)

    where0

    1

    ( ) lim ( ), ( ) 2 2

    t

    t t td d d

    = = P

    where

    2

    0

    1( ) lim ( ), ( ) sin

    tt t t

    t

    pd d d

    p

    = =

    (a) ( ) ( ) ( ) ( )t t T T t T f d f d - = - , f(t) t= T

    (b) ( ) ( ) ( ) ( ) ( )t t T dt T t T dt T f d f d f

    - -

    = =- -

    (sampling property)(sifting property)

  • 7/23/2019 Updated 4

    8/55

    (c)1

    ( ) ( ),at ta

    d d= a

    (d)d(t) = d(t).

    (e)

    2

    1

    ( ) ( )

    1 2( ) ( ) ( 1) ( ),

    n nnt

    t t t T dt T t T tf d f- = - <

  • 7/23/2019 Updated 4

    9/55

  • 7/23/2019 Updated 4

    10/55

  • 7/23/2019 Updated 4

    11/55

    Solution:

    2Write the signal)( ) sin( 2 cos(10 )6x t tt

    p p

    += as

    (a) The real part of a sum of rotating phasors.

    (b)A sum of rotating phasors plus their complex conjugates.

    (c) From your results in parts (a) and (b), sketch the single-sided and double-sided amplitude

    and phase spectra of x(t).

  • 7/23/2019 Updated 4

    12/55

    Solution:

    4For the signal g(t) shown in the following figure, sketch (a) g( t); (b) g(t+ 6); (c) g(3t);

    (d) g(6 t). Also find their energies.

    3For an energy signal g(t) with energy Eg, show that the energy of any one of the sig-

    nals g(t), g(t), and g(t T) is Eg. Show also that the energy of g(at) as well as g(at b) is Eg/a.

    This shows that time inversion and time shifting do not affect signal energy. On the other hand,

    time compression of a signal by a factor areduces the energy by the factor a. What is the effect

    on signal energy if the signal is (a) time-expanded by a factor a(a>1), and (b) multiplied by a

    constant a?

  • 7/23/2019 Updated 4

    13/55

    Solution:

    1-3

    1.3.1 Signals versus Vectors

    x(t)[a, b]

    N

    where31 2 , 2 , ..., ( 1) ,1

    , Nb a

    t a t a N ba t aN

    t -

    = += = + = + - = =-

    N x

    [ ]1 2( ) ( ) ( )Nx t x t x t=x

    N x Nx

    x

    l [ , ]im ( ), .N

    a bx t t

    = x

  • 7/23/2019 Updated 4

    14/55

    This relationship clearly shows that continuous-time signals are straightforward generali-zations of finite dimension vectors. Thus, basic definitions and operations in a vector spacecan be applied to continuous-time signals as well.

    1. (Complex Exponential Fourier Series)

    x(t) (t0, t0+T) 0 00

    22 f

    T

    pw p= = x(t)

    0

    0 0 0( ) ,jn

    ntx t X t te Ttw

    -

  • 7/23/2019 Updated 4

    15/55

    (Euler formula)

    0 0( )cos( ( )sin) )( Re( ) Im( ) ,nj X

    n n n nX x t n j x t n X j X X t t ew w= - = + =

    x(t)Re(Xn)Im(Xn)

    and ,n n n nX X X X - -= = -

    *n nX X- = x(t) 0( )cos( )nX x t n tw= x(t) Xn=

    0( )sin( )j x t n tw-

    (a)

    (Linearity)

    x(t) y(t) Xn YnAx(t)+By(t)AXn+BYn

    A, B(b)

    (Time Shifting)

    x(t) Xn x(t-a) 0jnaX e w-

    (c)

    (Frequency Shifting)

    x(t) Xn 0( ) j tkx t e w Xn-k k

    (d) (Conjugate)

    x(t) Xn x*(t)*

    nX-

    (e)

    (Time Reversal) x(t) Xn x(t) Xn

  • 7/23/2019 Updated 4

    16/55

    (f) Parsevals Theorem0 0

    0

    2 2

    0

    1( ) :

    t T

    n avt

    n

    x t dt X PT

    +

    =-

    = = Pav

    1.

  • 7/23/2019 Updated 4

    17/55

    2.

    (Line Spectra)

    x(t)(t0, t0+T0)

    0 0( )cos( ( )sin) )( Re( ) Im( ) ,nj X

    n n n nX x t n j x t n X j X X t t ew w= - = + =

    nj X

    nX e

    Xn (phasor)

    Xnw= nw0(n)

    (line spectra)X0 x(t) DCXn x(t)

    n (n-th harmonic)

    Solution:

    1-4 (Fourier Transform)

    x(t) x(t) ( )x t dt-

  • 7/23/2019 Updated 4

    18/55

    { } 21 ( ) ( ) ( ) .ftjX f x t X f e df p--

    = = F

    X(f)

    ( )) )( ) ( )cos(2 ( )sin(2 Re( ( )) Im( ( )) ( ) .j X fX f x t dt j x t dt X f j X f X f eft ftp p - -

    = - = + =

    x(t) Re(X(f)) Im(X(f)) 2 2 2( ) Re( ( )) Im( ( ))X f X f X f = + tan ( ) Im( ( )) /Re( ( ))X f X f X f = ( )X f- =

    ( )X f ( ) ( )X f X f - = - X(f) = X*(f) () x(t)X(f) =

    Re(X(f)) x(t) X(f) =jIm(X(f)) ( )X f

    ( )X f

    x(t)x(t)

    2 2*: ( ) ( ) ( ) ,j ftE x t dt x t X f e df dtp

    - - -

    = =

    *

    2 2*

    *

    ( ) ( ) ( ) ( )

    ( ) ( )

    ft fj j tE X f x t e dt df X f x t e dt df

    X f X f df

    p p

    - - - -

    -

    = =

    =

    22

    ( ) ( )E x t dt X f df

    - -= =

    Rayleighs energy theorem Parseval 2

    ( )X f jouls/hertz (energy density)

    (energy spectral density)2

    ( ) : ( )G f X f =

    G(f)

    ()

    1. (Superposition Theorem)

    1 1 2 2 1 1 2 2( ) ( ) ( ) ( )a x t a x t a X f a X f + + ()Proof.

    2. (Time-Delay Theorem)02

    0( ) ( ) ftjx t t X f e p--

    Proof.

  • 7/23/2019 Updated 4

    19/55

    3. (Scale-Change Theorem)

    1( )

    fx at X

    a a

    Proof.

    4. (Duality Theorem)

    ( ) ( )X t x f -

    x(t) X(f)X(t) ( X t X(t))

    x(f)

    Proof.

    5. (Frequency Transition Theorem)02

    0( ) ( )f tjx t e X f f p -

    Proof.

    6. (Modulation Theorem)

    0 0 0( )1 1

    ) ( ) (co 2 )(2 2

    s f t X f t f fx f Xp - + +

    7.

    (Differentiation Theorem)( )

    ( 2 ) ( )n

    n

    nd x

    df f

    tj X

    tp

    8. (Integration Theorem)

    Proof.

    1( ( 21

    ) ) ( ) (0) ( )

    2

    t

    d f X f x f Xjl l p d -

    -

    +

  • 7/23/2019 Updated 4

    20/55

    9.

    (Convolution Theorem)

    Proof.

    10. (Multiplication Theorem)

    Proof.

    Solution:

    1 2 1 1 12 2 2( ) ( ) : ( ) ( ) ) ( )( ( ) ( )x tx t x t x x t X X d d fx fl l l l l l

    -

    -* = =- -

    1 2 1 2 1 2( ) ( ) ( ) ( )) ( ) (x t x t X f X f f X X dl l l

    -

    * = -

    4. Obtain the following Fourier transform pairs:

    (1)

    ( )A t Ad (2) 020( )

    j ftt t AA e pd -- (3)

    ( )A A fd

    (4) 0 02 ( )f tjAe A f fp d -

    Ziemer

  • 7/23/2019 Updated 4

    21/55

  • 7/23/2019 Updated 4

    22/55

    Solution:

    ys(t)

    2 1( ( )) , ,sjns n sm n

    fs

    s

    tt mTy t Y e f T

    pd

    =- =-

    = = =-

    2)1 .( ss

    n sT

    jn f t

    s

    Y fT

    t e dtpd -= =

    2( ) ,sjns sn

    f ty t f e p

    =-

    =

    4(4)

    2( .() )1 sjs s s sn

    nf t

    n

    f nY f f e f f p d

    =- =-

    = = - F

    ( ) ( )s ssm n

    ft mT t nf d d=-

    -

    =

    - - (5)

    (5)

    2( ) ( ) ftj

    m ms st mT t mT e dt

    pd d --=

    -

    =-

    =

    - -

    F

    2 22( ) ( ) sft j ft mj j T

    m

    f

    m s m st mT e t mT e edtdtp p pd d- -

    - -

    -

    =- = - - =

    - - = =

    (5)mm

    2 ( )sj sm

    m f

    n

    Tse f t nf

    p d=

    - -=

    = -

    5. As another example of obtaining Fourier transform of signals involving impulses,

    let us consider the signal

    ( ) ( .)s sm

    t my Tt d=

    -

    = - It is a periodic waveform referred to as the ideal sampling waveform and consists of adoubly infinite sequence of impulses spaced by Tss. Find the Fourier transform of ys(t).

    (Ziemer)

  • 7/23/2019 Updated 4

    23/55

    4

    x(t)p(t) ( /2 /2s sT t T-