Professional Documents
Culture Documents
Power Spectrum
For a deterministic signal x(t), the spectrum is well defined: If X ( )
represents its Fourier transform, i.e., if
X ( ) x(t )e j t dt ,
(18-1)
x (t )dt
2
1
2
2
|
X
(
)
|
d E.
(18-2)
X (t )
Fig 18.1
Energy in ( , )
PILLAI
X T ( ) T X (t )e j t dt
so that
(18-3)
2
| X T ( ) |2
1 T
j t
X (t )e dt
(18-4)
T
2T
2T
represents the power distribution associated with that realization based
on ( T, T ). Notice that (18-4) represents a random variable for every
, and its ensemble average gives, the average power distribution
2
based on ( T, T ). Thus
PILLAI
| X T ( ) |2
1
PT ( ) E
2T
2T
j ( t1 t2 )
*
E
{
X
(
t
)
X
(
t
)}
e
dt1dt2
1
2
T T
1 T T
j ( t1 t2 )
R
(
t
,
t
)
e
dt1dt2
(18-5)
XX
1 2
T T
2T
represents the power distribution of X(t) based on ( T, T ). For wide
sense stationary (w.s.s) processes, it is possible to further simplify
(18-5). Thus if X(t) is assumed to be w.s.s, then RXX (t1 , t2 ) RXX (t1 t2 )
and (18-5) simplifies to
1
PT ( )
2T
j ( t1 t2 )
R
(
t
t
)
e
dt1dt2 .
T T XX 1 2
)
e
(2T | |)d
XX
2T
2T
2T
2T RXX ( )e j (1 2|T| )d 0
(18-6)
PILLAI
S XX ( ) lim PT ( ) RXX ( )e j d 0
T
(18-7)
to be the power spectral density of the w.s.s process X(t). Notice that
FT
RXX ( )
S XX (
) 0.
(18-8)
1
2
j
S
(
)
e
d
XX
(18-9)
2
S
(
)
d
R
(0)
E
{|
X
(
t
)
|
} P,
XX
XX
S XX ( )
Fig 18.2
The nonnegative-definiteness property of the autocorrelation function
in (14-8) translates into the nonnegative property for its Fourier
transform (power spectrum), since from (14-8) and (18-9)
n
ai a RXX (ti t j ) ai a
i 1 j 1
*
j
i 1 j 1
1
2
*
j
1
2
S XX ( )e
( ) i 1 ai e
n
XX
j ti
j ( ti t j )
d 0.
(18-11)
S XX ( ) 0.
(18-12)
5
PILLAI
S XX ( ) RXX ( )e j d
RXX ( ) cos d
2 0 RXX ( ) cos d S XX ( ) 0
(18-13)
PILLAI
g (t ) G ( )
f (t ) g (t ) F ( )G ( )
(18-15)
(18-16)
since
F { f (t ) g (t )} f (t ) g (t )e j t dt
PILLAI
F { f (t ) g (t )}=
f ( ) g (t )d e j t dt
= f ( )e
g (t )e j ( t ) d (t )
=F ( )G ( ).
(18-17)
h ( )e
*
h(t )e
j t
dt
(18-18)
H * ( ),
where
H ( ) h(t )e jt dt
(18-19)
(18-20)
PILLAI
(18-22)
Notice that the output spectrum captures the system transfer function
characteristics entirely, and for rational systems Eq (18-22) may be
used to determine the pole/zero locations of the underlying system.
9
PILLAI
| H ( )|2
qB
1
B / 2
(a) LPF
B/2
Fig. 18.4
(b)
10
PILLAI
t T
2T t T
X ( )d
(18-25)
1 / 2T
T
Y (t ) h(t ) X ( )d h(t ) X (t )
Fig 18.5
(18-26)
so that
Here
SYY ( ) S XX ( ) | H ( ) |2 .
T
H ( ) T
1
2T
e j t dt sinc(T )
(18-27)
(18-28) 11
PILLAI
so that
SYY ( ) S XX ( ) sinc 2 ( T ).
(18-29)
sinc 2 ( T )
S XX ( )
SYY ( )
Fig 18.6
Notice that the effect of the smoothing operation in (18-25) is to
suppress the high frequency components in the input (beyond / T ),
and the equivalent linear system acts as a low-pass filter (continuoustime moving average) with bandwidth 2 / T in this case.
12
PILLAI
rk ( kT ).
rk e jT 0,
(18-30)
(18-31)
(18-32)
13
PILLAI
)
e
d
(18-33)
B XX
2B
and
1 B
2
(18-34)
r0 E{| X ( nT ) | }
S ( )d
B XX
2B
represents the total power of the discrete-time process X(nT). The
input-output relations for discrete-time system h(nT) in (14-65)-(14-67)
translate into
S XY ( ) S XX ( ) H * (e j )
(18-35)
SYY ( ) S XX ( ) | H (e j ) |2
(18-36)
and
where
j
H (e )
h( nT ) e j nT
(18-37)
14
PILLAI
Matched Filter
Let r(t) represent a deterministic signal s(t) corrupted by noise. Thus
r (t ) s(t ) w(t ),
0 t t0
(18-38)
h(t)
t t0
y(t)
y (t ) ys (t ) n(t )
(18-39)
where
y s (t ) s(t ) h(t ),
(18-40)
2
Output
signal
power
at
t
t
|
y
(
t
)
|
0
s 0
( SNR )0
| y s (t0 ) |
1
2
nn
( )d
1
2
1
2
S ( ) H ( )e
WW
j t0
( ) | H ( ) | d
2
(18-41)
j t0
S ( ) H ( )e d
( SNR )0
(18-42)
2
2 N 0 | H ( ) | d
and a direct application of Cauchy-Schwarz inequality
in (18-42) gives
16
PILLAI
1
2 N 0
( SNR )0
| S ( ) |
d 0
s(t ) 2 dt
N0
Es
N0
(18-43)
h (t )
(a)
T / 2
h (t )
t0
(b) t0=T/2
(c) t0=T
Fig 18.8
Fig 18-8 shows the optimum h(t) for two different values of t0. In Fig
18.8 (b), the receiver is noncausal, whereas in Fig 18-8 (c) the
17
PILLAI
receiver represents a causal waveform.
(18-46)
t0
s 2 (t )dt
(18-47)
f (t )
T
q(t)
s (t )
Fig 18.9
T
s (t ) f (t ) q(t ) 0 f ( )q (t ) d ,
(18-48)
where q(t) represents the target impulse response. One interesting
question in this context is to determine the optimum transmit 18
PILLAI
signal f (t) with normalized energy that maximizes the receiver output
SNR at t = t0 in Fig 18.7. Notice that for a given s(t), Eq (18-45)
represents the optimum receiver, and (18-43) gives the corresponding
maximum (SNR)0. To maximize (SNR)0 in (18-43), we may substitute
(18-48) into (18-43). This gives
T
( SNR )0 0 |{0 q(t 1 ) f ( 1 )d 1 } |2 dt
1
N0
0 0
*
q
(
t
)
q
10 4 4 4412 4(t4 42 4)3dt f ( 2 )d 2 f ( 1 )d 1
( 1 , 2 )
1
N0
0 { 0 ( 1 , 2 ) f ( 2 )d 2 } f ( 1 )d 1 max / N 0
(18-49)
where ( 1 , 2 ) is given by
( 1 , 2 ) 0 q (t 1 )q* (t 2 ) dt
(18-50)
0 ( 1 , 2 ) f ( 2 )d 2 max f (1 ),
0 1 T .
(18-51)
19
PILLAI
and
T
f 2 (t )dt 1.
(18-52)
f (t )
h(t )
T
(a)
t0
(b)
Fig 18.10
(c)
( 1 , 2 ) 0 q(t 1 ) q* (t 2 )dt.
and the optimum transmit signal is given by (18-51). Notice
that in the causal case, information beyond t = t0 is not used.
(18-53)
20
PILLAI
r (t ) s (t ) w(t )
colored noise
Fig 18.11
sg (t ) n(t )
white noise
Notice that the signal part of the whitened output sg(t) equals
sg (t ) s (t ) g (t )
(18-54)
where g(t) represents the whitening filter, and the output noise n(t) is
white with unit spectral density. This interesting idea due to 21
PILLAI
1
| G ( ) |
.
SWW ( )
2
(18-55)
From there, any spectral density that satisfies the finite power constraint
XX
( )d
(18-56)
S XX ( ) | H ( j ) |2 H ( s ) H ( s) |s j
(18-57)
(18-58)
where H(s) together with its inverse function 1/H(s) represent two
filters that are both analytic in Re s > 0. Thus H(s) and its inverse 1/ H(s)
can be chosen to be stable and causal in (18-58). Such a filter is known
as the Wiener factor, and since it has all its poles and zeros in the left
half plane, it represents a minimum phase factor. In the rational case,
if X(t) represents a real process, then S XX ( ) is even and hence (18-58)
reads
23
PILLAI
2
0 S XX ( 2 ) S%
(
s
) |s j H ( s) H ( s) |s j .
XX
Example 18.3: Consider the spectrum
(18-59)
( 2 1)( 2 2) 2
S XX ( )
( 4 1)
which translates into
1 j
s
2
(1 s )(2 s )
2
%
S XX ( s )
.
4
1 s
2
2 2
( s 1)( s 2 j )( s 2 j )
s 1 j s 1 j
2
d
d s 2 j
s 1 j
2
d
s 1
d
s1
d
1 j
d s 1 j
s
2
2 s 2j
Fig 18.12
( s 1)( s 2 2)
2
s 2s 1
24
PILLAI
ln H ( ) ln A( ) j ( ) 0 b(t )e j t dt.
Thus
ln A( ) 0 b(t ) cos t dt
t
( ) 0 b(t )sin t dt
and since cos t and sin t are Hilbert transform pairs, it follows that
the phase function ( ) in (18-60) is given by the Hilbert 25
PILLAI
transform of ln A( ). Thus
( ) H {ln A( )}.
(18-61)
Eq. (18-60) may be used to generate the unknown phase function of
a minimum phase factor from its magnitude.
For discrete-time processes, the factorization conditions take the form
(see (9-203)-(9-205), Text)
XX
( )d <
(18-62)
( ) d > .
(18-63)
and
ln S
XX
In that case
S XX ( ) | H (e j ) |2
where the discrete-time system
H ( z ) h(k ) z k
k 0
26
PILLAI
is analytic together with its inverse in |z| >1. This unique minimum
phase function represents the Wiener factor in the discrete-case.
Matched Filter in Colored Noise:
Returning back to the matched filter problem in colored noise, the
design can be completed as shown in Fig 18.13.
r ( t ) s ( t ) w( t )
G ( j ) L1 ( j )
Whitening Filter
s g (t ) n (t )
t t0
h0(t)=sg(t0 t)
Matched Filter
Fig 18.13
(Here G ( j ) represents the whitening filter associated with the noise
spectral density SWW ( ) as in (18-55)-(18-58). Notice that G(s) is the
inverse of the Wiener factor L(s) corresponding to the spectrum SWW ( ).
i.e.,
L( s ) L( s ) |s j | L( j ) |2 SWW ( ).
The whitened output sg(t) + n(t) in Fig 18.13 is similar
(18-64)
27
PILLAI
H ( )
(a)
t t0
r (t )
L-1(s)
(b)
H ( )
L(s)
1 4 44 2 4 4 43
t t0
H 0 ( )
Fig 18.14
Notice that Fig 18.14 (a) and (b) are equivalent, and Fig 18.14 (b) is
equivalent to Fig 18.13. Hence (see Fig 18.14 (b))
H 0 ( ) L( j ) H ( )
or
28
PILLAI
H ( ) L1 ( j ) H 0 ( ) L1 ( ) S g* ( )e j t0
L1 ( ){L1 ( ) S ( )}* e j t0
(18-65)
turns out to be the overall matched filter for the original problem.
Once again, transmit signal design can be carried out in this case also.
(18-66)
(18-67)
where
FM
PM.
(t )
c t m( )d
0
cm(t )
(18-68)
29
PILLAI
Here m(t) represents the message signal and a random phase jitter
in the received signal. In the case of FM, (t ) (t ) c m(t ) so that
the instantaneous frequency is proportional to the message signal. We
will assume that both the message process m(t) and the noise process
n(t) are w.s.s with power spectra S mm ( ) and S nn ( ) respectively.
We wish to determine whether the AM and FM signals are w.s.s,
and if so their respective power spectral densities.
Solution: AM signal: In this case from (18-66), if we assume
~ U (0, 2 ), then
1
RXX ( ) Rmm ( ) cos 0 Rnn ( )
(18-69)
2
so that (see Fig 18.15)
S XX ( 0 ) S XX ( 0 )
(18-70)
S XX ( )
Snn ( ).
2
S XX ( )
Smm ( )
(a)
Fig 18.15
(b)
S mm ( 0 )
30
PILLAI
E{cos[ 0 (t / 2) (t / 2)]
2
cos[2 0 t (t / 2) (t / 2) 2 ]}
A2
[ E{cos( (t / 2) (t / 2))}cos 0
2
E{sin( (t / 2) (t / 2))}sin 0 ]
since
E{cos(2 0t (t / 2) (t / 2) 2 )}
(18-71)
0
E{cos(2 0t (t / 2) (t / 2))}E{cos 2 }
0
E{sin(2 0t (t / 2) (t / 2))}E{sin 2 } 0.
31
PILLAI
a (t , ) E{cos( (t / 2) (t / 2))}
(18-72)
(18-73)
and
b( t , )
E{sin( (t / 2) (t / 2))}
(18-74)
d
R ( )
2
R ( ) c Rmm ( )
(18-75)
2
d
for the FM case. In that case the random variable
32
PILLAI
where
Y (t / 2) (t / 2) ~ N (0, Y2 )
(18-76)
Y2 2( R (0) R ( )).
(18-77)
jY
} e
2 Y2 / 2
( R (0) R ( )) 2
(18-78)
(18-79)
( R (0) R ( ))
(18-80)
and
b(t , ) 0
(18-81)
so that the FM autocorrelation function in (18-72) simplifies into
33
PILLAI
A2 ( R (0) R ( ))
RXX ( )
e
cos 0 .
(18-82)
2
Notice that for stationary Gaussian message input m(t) (or (t ) ), the
nonlinear output X(t) is indeed strict sense stationary with
autocorrelation function as in (18-82).
Narrowband FM: If R (0) = 1, then (18-82) may be approximated
as ( e x ; 1 x, | x | 1)
A2
RXX ( ) {(1 R (0)) R ( )}cos 0
2
(18-83)
PILLAI
2
1
c
(0) 2 L R (0) Rmm (0) 2 L (18-84)
R ( ) R (0) R
2
2
and substituting this into (18-82) we get
A2 c22 Rmm (0) 2 L
RXX ( )
e
cos 0
(18-85)
2
so that the power spectrum of FM in this case is given by
(18-86)
S XX ( ) 12 S ( 0 ) S ( 0 )
where
2
A
2 / 2 c 2 Rmm (0)
~
S ( ) e
.
(18-87)
2
Notice that S XX ( ) always occupies infinite bandwidth irrespective
of the actual message bandwidth (Fig 18.16)and this capacity to spread
the message signal across the entire spectral band helps to reduce the
noise effect in any band.
S XX ( )
Fig 18.16
35
PILLAI
rk e jk 0.
(18-88)
Notice that from (14-64), the given sequence satisfies Tn > 0, and at
every step of the extension, this nonnegativity condition must be
satisfied. Thus we must have
Tn k 0,
Let rn 1 x. Then
k 1, 2,L .
(18-89)
36
PILLAI
x
rn
Tn 1 =
Tn
rn 1
M
r1
x* , rn* , L r1* r0
Fig 18.17
(18-90)
or
2
| rn 1 n |
n
,
n 1
(18-91)
where
n a T Tn11 b , a [r1 , r2 ,L , rn ]T , b [ rn , rn 1 ,L , r1 ]T .
(18-92) 37
PILLAI
Eq. (18-91) represents the interior of a circle with center n and radius
n / n1 as in Fig 18.17, and geometrically it represents the admissible
set of values for rn+1. Repeating this procedure for rn 2 , rn 3 ,L , it
follows that the class of extensions that satisfy (18-85) are infinite.
It is possible to parameterically represent the class of all
admissible spectra. Known as the trigonometric moment problem,
extensive literature is available on this topic.
[See section 12.4 Youlas Parameterization, pages 562-574, Text for
a reasonably complete description and further insight into this topic.].
38
PILLAI