Scribd
Upload
51 views4 pages

Lecture07 Notes

This document summarizes key concepts from a lecture on baseband digital transmission and noise/errors: - A baseband binary receiver receives a signal plus noise at time k (yk(t)) and uses a comparator to regenerate the signal (xe(t)) based on a decision threshold V. - Additive noise can cause the comparator to output the incorrect value (an error), modeled by random variables Y (received signal) and N (noise). - The optimal threshold Vopt minimizes the average error probability Pe by making the conditional error probabilities Pe0 and Pe1 equal. - For binary pulse amplitude modulation (PAM) and zero-mean Gaussian noise, the error probability is

Uploaded by

NeeDog20
Copyright
© Attribution Non-Commercial (BY-NC)
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
51 views4 pages

Lecture07 Notes

This document summarizes key concepts from a lecture on baseband digital transmission and noise/errors: - A baseband binary receiver receives a signal plus noise at time k (yk(t)) and uses a comparator to regenerate the signal (xe(t)) based on a decision threshold V. - Additive noise can cause the comparator to output the incorrect value (an error), modeled by random variables Y (received signal) and N (noise). - The optimal threshold Vopt minimizes the average error probability Pe by making the conditional error probabilities Pe0 and Pe1 equal. - For binary pulse amplitude modulation (PAM) and zero-mean Gaussian noise, the error probability is

Uploaded by

NeeDog20
Copyright
© Attribution Non-Commercial (BY-NC)
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

EE413 Random Signals and Noise

Lecture 7
Chapter 11: Baseband Digital Transmission
Assignment:
HW 3A: 11.2-2, 11.2-4, 11.2-9
11.2: Noise and errors
Baseband binary receiver
Binary Receiver
At some optimum time t
k
yt ) -yt
k
)
yt
k
)=a
k
+nt
k
)
With the x
e
t )=

A y t
k
)>V
0 y t
k
)V
where x
e
t ) is the regenerator output.
Regeneration of a unipolar signal
(a) signal plus noise (b) S/H output (c) comparator output
Noise, Errors and Conditional Probability
As the previous slide indicates, additive noise
affects the value of yt
k
)causing the
to output the incorrect value
error
Let random variables be

Y yt
k
)
N nt
k
)
Let H
0
hypothesis that a
k
=0 and Y =N
Conditional PDF p
Y
yH
0
)=p
N
y)
Let H
1
hypothesis that a
k
=A and Y =A+N
Conditional PDF p
Y
yH
1
)=p
N
yA)
Decision Rule
Choose hypothesis H
0
a
k
=0) if Y V
Choose hypothesis H
1
a
k
=A) if Y >V
Conditional PDFs with decision threshold and error prob.
1 of 4
Error Probabilities
P
e0
=Pe0)PY >VH
0
)=

p
Y
yH
0
) dy
P
e1
=Pe1)PY VH
1
)=

p
Y
yH
1
)dy
with P
0
=P H
0
) and P
1
=P H
1
)
Average probability is
P
e
=P
e0
P
0
+P
e1
P
1
Threshold Regions For Conditional pdfs
Optimum Threshold
Optimum V
opt
To determine V
opt

dP
e
dV
=0
Using Leibniz's rule for differentiating integrals
P
Y
V
opt
H
0
) P
0
=P
Y
V
opt
H
1
) P
1
If we assume 1s and 0s equally likely
P
0
=P
1
=1/ 2
P
Y
V
opt
H
0
)=P
Y
V
opt
H
1
)
Probability of Errors
Assuming the noise is zero-mean
with variance c
2
p
N
n)=
1
.2nc
2
e
n
2
/2c
2
P
e0
=

p
N
y) dy=Q

V
c
)
P
e1
=

p
N
yA) dy=Q

AV
c
)
If V =V
opt
and P
e0
=P
e1
P
e
=Q

A
2c
)
Signal Types
This result applies for unipolar PAM with levels 0
and A with V
opt
=1/ 2
P
e
=Q

A
2c
)
The result is the same for PAM when
a
k
=!
A
2
V
opt
=0
P
e
=Q

A
2c
)
Note: The error probability is the same since the
space between levels and the threshold =A/ 2
Unipolar vs Polar Signals
Unipolar and polar signals have the same error
probability in this case, but the polar signal
uses the signal power of the
unipolar signal
Unipolar S
R
=A
2
/ 2, Polar S
R
=A
2
/ 4
A=

.
2 S
R
unipolar
.
4 S
R
polar
2 of 4
Signal-To-Noise Ratio

A
2c
)
2
=
A
2
4 N
R
=

1
2
S / N )
R
unipolar
S / N )
R
polar
The bit rate has an effect. To pass pulses of duration
T
b
=1/ r
b
the filter's bandwidth must satisfy B
N
r
b
/2
N
R
=N
0
B
N
N
0
r
b
/2
signalingdecreases S / N )
R
increase in P
e
Regenerative Repeaters
Recall from Chaps 3 and 9 that long haul transmission requires
repeaters.
For digital signals, as an alternative to amplifying the signal, we
can use _______________________________.

Detect signal
convert to message digits
re-transmit message digits
Regenerative Repeater Features
Conventional analog repeaters will amplify the signal,
and introduce noise and __________________
With regenerative repeaters, the repeater output is a
new signal without the cumulative effects of noise and
distortion building up at each hop
Regenerative repeaters => each repeater is a complete
receiver and transmitter.
Regenerative output can contain errors, but noise and
distortion are eliminated.
Error Performance of Regenerative Repeaters
o=Q
|
.

S
N
)
1

=error probability at each repeater


Probability of i errors in m successive conversions
m repeaters P
I
i )=

m
i
)
o
i
1o)
mi
Even errors cancel out ( signal)
for i odd
P
e
=

m
1
)
o1o)
m1
+

m
3
)
o
3
1o)
m3
+.mo
P
e
mQ
|
.

S
N
)
1

Error Performance of Regenerative Repeaters


With analog repeaters
P
e
=Q
|
.
1
m

S
N
)
1

Power Saving by m Regenerative Repeaters


3 of 4
In Class Exercises
Problem 11.2-1
Problem 11.2-8
4 of 4
P
e
=10
5

You might also like