Summary of Signal Types: Pulses

1) The document summarizes different types of signal representations and probability distributions related to signals and noise. It provides information on the relationships between time and frequency domain representations of pulses and periodic signals. 2) It describes Gaussian, binomial, Poisson, and Laplacian probability distributions and their characteristics. Applications include modeling noise and describing random processes. 3) Thermal noise in transmission lines is analyzed by determining the electromagnetic energy density spectrum based on photon distributions in electromagnetic modes using Bose-Einstein statistics and Planck's law.

Uploaded by

varunmr

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

42 views

Summary of Signal Types: Pulses

Uploaded by

varunmr

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 22

Slide 1

Summary of Signal Types

G6
v(t) V(f) [ V ] [VHz
1
] [ V ] [VHz
1
]

R() S(f) = V(f)
2
[ V
2
s ] [VHz
1
]
2
[J ] [J Hz
1
]
Typical Sets of Units
Pulses:
1-ohm load
v(t) V
m
[v] [ V ] [V] [ V ] [ V ]

R()
m
= V
m

2
[ V
2
] [ V
2
] [ W ] [ W ]
Periodic:
1-ohm load
v(t) [?] [ V ] [?] [ V ] [ ? ]

() (f) [ V
2
] [ V
2
Hz
1
] [ W ] [ W/Hz ]
Random:
6.661 Fall 2001
Slide 2
Probability Distributions
G7
[ ]
2 2 2
2 / ) / v (
dv v ) v ( p
e
2
1
} v { P
2
=

=

p(v)

0
[ ]
o
o
2 2
=
Gaussian Noise:
v
Band-limited Gaussian white noise, e.g.
N
o
/2
(f)
( ) WHz N
1
o

[ ] WHz
1
f
B
(f)
Single-sided
equivalent
Power
spectral
density
2 N
o
f
6.661 Fall 2001
v E =

B N
B N v E
=
=
Slide 3
Probability Distributions
G8
{ } { } p 1 r Assume
{ } ( )
( ) n k
n
where p 1 p
k
n
k n k

=

[ ] ( )

=
= =
n
0 k
np
Binomial Distribution:
6.661 Fall 2001
Note: There are n positions possible for the first 1,
n 1 for the second 1, and a total of
n(n 1)(n k + 1)/k! ways to arrange those
k 1s among the n available positions.
0 p , p 1 p where 1, o 0 bits, n have we
k! ! k
! n
s 1' k p
k p k k E
Slide 4
Probability Distributions
G9
Poisson Distribution:
{ } { } p 1 r Assume
{ }

= = + =

=

0 xy ; y x
e
2
1
2
o
2 2 2 2 2
/
o
o
o
2 / 1
0
2 /
o

[ ] ( )

=
= =
n
0 k
np
N
~0
If n np i (

>> << =

e
k
Mean of k = = np
Variance of k =
6.661 Fall 2001
0 p , p 1 p where 1, o 0 bits, n have we
mean zero , variance Gaussian, are y x, where
x , y r if example, for , Arises
r p
r 2
Laplacian Distribution:
k p k k E
1, p 1: 1; var ance np 1-p) =

! k
p{k} then
Slide 5
Receiver-Noise Processes
Receivers are limited by noise, many types
1-D (TEM transmission line)
3-D (Multimode waveguide)
Equation of radiative transfer (1-D)
RF and optical limits; IR case
Cases:
Thermal noise:
2 6.661 Fall 2001
Slide 6
Thermal noise, 1-D (TEM) case
Z
o
, TK Z
o
(), c(ms
-1
)
Electromagnetic
energy
Heat
Heat
3 6.661 Fall 2001
lossless TEM line
Slide 7
Approach:
closed container
very slightly lossy
1) Find average energy density W(f)[J /m Hz]
2) Find average power P
+
[W/Hz] power flow
power P
+
P
-
D
z
4 6.661 Fall 2001
Slide 8
Find average energy density W(f)[J m
-1
Hz
-1
]
W(f) =
modes
Hz
photons
mode
energy
photon
hf [J oules] (h Plancks constant)
1
D
f is frequency (Hz)
h = 6.6252 10
-34
(J s)
5 6.661 Fall 2001
Slide 9
W(f) =
modes
Hz
photons
mode
energy
photon
1
D
Find modes/Hz:
Resonator modes
Therefore (v
p
= phase velocity)
modes/Hz
m =
2D

m
=
2Df
m
v
p
dm
df
2D
v
p
=
V(z)
z = 0
m = 3
m = 2
m = 1
z
D
6 6.661 Fall 2001
Find photons/mode
n
j
; (j
th
mode)

Photons obey Bose-Einstein statistics; therefore
any number can occupy each mode.
Total energy fixed; combinations favor more likely distributions
p
j
(n) = Q e
-nW
j
/kT
, Boltzmann distribution
where

n=0

p
j
(n) 1, W
j
hf
j
, Q = constant

n
j
= n p
j
(n)

n=0

p
j
(n)
7 6.661 Fall 2001
Slide 10
p{n photons in state j}
p
j
(n) = Q e
-nW
j
/kT
, Boltzmann distribution
where

n=0

p
j
(n) 1, W
j
hf
j
, Q = constant

n=0

Recall x
n
) if x < 1
-W
j
/kT
Therefore Q =
p
j
(n) = Q

n=0

e
-W
j
/kT
n
=
-W
j
/kT
Q

n=0

8 6.661 Fall 2001
Slide 11
= 1/(1 - x
1 - e
1 - e
p
j
(n) = Q e
-nW
j
/kT
, Boltzmann distribution
-W
j
/kT
Where Q =
p
j
(n) =
-W
j
/kT
e
-nW
j
/kT
-W
j
/kT
n
j
=

n=0

n=0

n p
j
(n) =
e
-W
j
/kT
n
n
Therefore
9 6.661 Fall 2001
Slide 12
1 - e
1 - e
1 - e
Recall

n=0

n x
n
= x
d
dx

n=0

x
n
= x
d
dx
)
-1
=
2
x
So n
j
=
-W
j
/kT
e
-W
j
/kT -W
j
/kT
2

=

=

= =
0 n
n
kT W kT W
0 n
j
j
j j
e 1 ) n ( n
[ ]
j j
kT W
j
hf e 1 n
j
=

=
10 6.661 Fall 2001
Slide 13
(1 - x
1 - x
1 - e 1 - e
e n p n
W de photons/mo 1
W(f) =
modes
Hz
photons
mode
energy
photon
1
D
J m
-1
Hz
-1
W(f) =
2D
v
p e
W
j
/kT
1
(hf)
1
D
=
2hf
v
p
e
hf/kT
J m
-1
Hz
-1
W(f) = W
+
+ W
-
= 2W
+
(powers and energies superimpose
if waves are orthogonal)
W
+
= forward-moving energy density
11 6.661 Fall 2001
Slide 14
Solution - Average Energy Density
- 1
- 1
W(f) =
2D
v
p e
W
j
/kT
1
(hf)
1
D
=
2hf
v
p
e
hf/kT
J m
-1
Hz
-1
W(f) = W
+
+ W
-
= 2W
+
If the TEM line is non-dispersive, then v
p
= v
g
and
P
+
(f) WH
z
-1
=
e
hf/kT
hf
P
+
WH
z
-1
= v
g
W
+
= v
g
W/2
12 6.661 Fall 2001
Slide 15
Solution - Thermal power in TEM line:
- 1
- 1
- 1
Recall e
x
= 1 + x + x
2
/2! + 1 + x for x << 1
P kTB watts in uniform bandwidth B(Hz)
Rayleigh-J eans region (radio vs. optical)
B 0
f(Hz)
kT
f
o
P
+
(f)
hf
o
kT, so f
o
= kT/h
Plancks constant: h 10
-34
[J sec]
Boltzmanns constant: 10
-23
[ ] kT
1 e
hf
WH ) f ( P
1
z
<<

=

+
Rayleigh-J eans limit
13 6.661 Fall 2001
Slide 16
20 T(K) GHz
6.6
k = 1.38 [J /K]
hf for kT
kT hf
Problem: Find thermal radiation intensity I(watts/Hz m
2
ster)
Approach:
slightly lossy, filled with photons
b
a
d
z
x
y
1) Find energy density spectrum W(f)[J /m
3
Hz]
2) Relate W(f) to I(f)
e.g. antenna
Intercepting
surface
W(f)
I(f)
14
6.661 Fall 2001
Slide 17
Assume closed container, very
First:
waveguide
modes
Find energy density spectrum W(f)[J m
-3
Hz
-1
]
W(f) =
modes
Hz
photons
mode
energy
photon
1
vol.
TE
m,n
1 e -1
hf/kT
hf
E
z
0 H
z
0
m
y
2
= a,
n
x
2
= b

x
b
x
m=4
n=3
a
y

y
a
b
d
z
15
6.661 Fall 2001
Slide 18
m,n
, TM
Begin finding modes/Hz
2 2 2
cp cn cm
f

= Claim:
Recall wave eqn:
2
+
2
E = 0

2
x
2

2
y
2

2
z
2
+ +
wave
direction
x

y
y

o
k
k
2
x
k
2
y
k
2
z
+ + =
k
2
o
=
2

o
=
2

o
2
E = E
o
e
x y z
z
[k
x
= 2/
x
]
16
6.661 Fall 2001
Slide 19
p , n , m
d 2 b 2 a 2
- jk x - jk y - jk
Uniform plane wave
k
2
x
k
2
y
k
2
z
+ + =
k
2
o
=
2

o

o
=
2

o
2
E = E
o
e
x y z
z
[k
x
= 2 /
x
]
m

y
2
= a (m/2a)
2
+ (n/2b)
2
+ (p/2d)
2
= (f/c)
2
= QED
Next, use this relation to find modes/Hz
1/
2
X
+
1/
2
y
+
1/
2
z
=
1/
2
o
= (f/c)
2
Therefore
Note:
wave
direction
x

y
y

x

o
k
( ) 0 E
2 2
=
17
6.661 Fall 2001
Slide 20
- jk x - jk y - jk
+
f
m,n,p
=
cm
2a
2
cn
2b
2
+
cp
2d
2
+
#modes in df shell =
= vol. of shell 2/vol. cell
4f
2
8
c
2a

c
2b

c
2c
=
8f
2
c
3
V
ol
df
modes
Hz
df
=
abd
df
1 2 3 0
0
1
1
2
p(c/2d)
m(c/2a)
n(c/2b)
Find modes/Hz:
18
6.661 Fall 2001
Slide 21
TE + TM
df 2
W(f) =
modes
Hz
photons
mode
energy
photon
1
vol.
Find energy density spectrum W(f):
W(f) =
8f
2
c
3
V
e
hf/kT
1 8
c
3
hf
3
e -1
hf/kT
[J m
-3
Hz
-1
]
19
6.661 Fall 2001
Slide 22
- 1
hf 1/V =