Chapter 8 MS

EE4512 Analog and Digital Communications Chapter 8
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital to Analog
Digital to Analog
Conversion
Conversion
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Sampling and Quantization

Sampling and Quantization
Pages 390
Pages 390
-
-
391
391
Traditional
Traditional
analog transmission
analog transmission
(AM, FM and PM) are less
(AM, FM and PM) are less
complex than digital data transmission have been the basis
complex than digital data transmission have been the basis
of broadcasting and communication for 100 years.
of broadcasting and communication for 100 years.
Analog television signal Analog television spectrum
S&M Figure 8
S&M Figure 8
-
-
1a
1a
Digital data transmission

Digital data transmission
(PAM, ASK, PSK, FSK and QAM)
(PAM, ASK, PSK, FSK and QAM)
is more complex but (perhaps) offers higher performance
is more complex but (perhaps) offers higher performance
with control of accuracy and easier storage, simpler signal
with control of accuracy and easier storage, simpler signal
processing for noise reduction, error detection and
processing for noise reduction, error detection and
correction and encryption.
correction and encryption.
S&M Figure 8
S&M Figure 8
-
-
1b
1b
Digital data transmission requires analog

Digital data transmission requires analog
-
-
to
to
-
-
digital (ADC)
digital (ADC)
and digital
and digital
-
-
to
to
-
-
analog (DAC) converters. The ADC process
analog (DAC) converters. The ADC process
utilizes sampling and quantization of the continuous analog
utilizes sampling and quantization of the continuous analog
signal.
signal.
ADC
ADC
DAC
DAC
S&M Figure 8
S&M Figure 8
-
-
1b
1b
ADC sampling occurs at a uniform rate (the

ADC sampling occurs at a uniform rate (the
sampling rate
sampling rate
)
)
and has a continuous amplitude.
and has a continuous amplitude.
Continuous
amplitude
Uniform
sampling
rate
Analog signal
S&M Figure 8
S&M Figure 8
-
-
2a,b
2a,b
The continuous amplitude sample is then quantized to

The continuous amplitude sample is then quantized to
n
n
bits or resolution for the
bits or resolution for the
full scale
full scale
input or
input or
2
2
n n
levels
levels
.
.
Continuous
amplitude
Uniform
sampling
rate
Quantized
S&M Figure 8 S&M Figure 8- -2b,c 2b,c
Quantized
amplitude
Here n =
Here n =
4 and there are 2
4 and there are 2
4 4
= 16
= 16
levels for a full scale input
levels for a full scale input
of 2 V (
of 2 V (
1 V). The
1 V). The
step size
step size
= 2 V / 16 = 0.125 V and the
= 2 V / 16 = 0.125 V and the
quantized value
quantized value
is the
is the
midpoint
midpoint
of the
of the
voltage
voltage
range
range
.
.
S&M Table 8.1
S&M Table 8.1
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Sampling Baseband
Sampling Baseband
Analog Signals
Analog Signals
Pages 392
Pages 392
-
-
399
399
The analog signal

The analog signal
x(t
x(t
) which is continuously, uniformly
) which is continuously, uniformly
sampled is represented by:
sampled is represented by:
Multiplication in the temporal domain is convolution in the
Multiplication in the temporal domain is convolution in the
frequency domain and the frequency domain representation
frequency domain and the frequency domain representation
is:
is:
=
s
k =
x (t) x(t) (t k )
S
T
S&M Eq. 8.1
S&M Eq. 8.1

=
`
)

=
`
)
=
s
k =
s
k =
s
k =
X (f) X(f) (t k )
X (f) X(f) (f k )
X (f) X(f k )
S
S S
S S
T
f f
f f
F
S&M Eq. 8.2
S&M Eq. 8.2
Temporal and spectral representation of the continuous

Temporal and spectral representation of the continuous
sampling process for a sum of three sinusoids.
sampling process for a sum of three sinusoids.
k =
(t k )
S
T
=
s
k =
x (t) x(t) (t k )
S
T
=
s
k =
X (f) X(f k )
S S
f f
k =
(f k )
S
f
S&M Figure 8
S&M Figure 8
-
-
3
3
2 V, 20
2 V, 20
initial
initial
phase,
phase,
500 Hz
500 Hz
sinusoid
sinusoid
sampled at
sampled at
5 k samples/sec
5 k samples/sec
S&M Figure 8
S&M Figure 8
-
-
4a,b
4a,b
Aliased samples
Aliased samples
can be
can be
reconstructed
reconstructed
for a 4500 Hz
for a 4500 Hz
and a 5500 Hz
and a 5500 Hz
sinusoid that
sinusoid that
appears to be
appears to be
a 500 Hz
a 500 Hz
sinusoid
sinusoid
S&M
S&M
Figure 8
Figure 8
-
-
4a,c,d
4a,c,d
The
The
aliasing
aliasing
of
of
the signal can
the signal can
be predicted by
be predicted by
the magnitude
the magnitude
spectrum of the
spectrum of the
original 500 Hz
original 500 Hz
sampled signal.
sampled signal.
If the 4500 Hz
If the 4500 Hz
and 5500 Hz
and 5500 Hz
signals are then
signals are then
sampled at
sampled at
5 k samples/sec
5 k samples/sec
aliasing at occurs at | 4500
aliasing at occurs at | 4500
5000 | and (5500

5000 | and (5500
5000) Hz
5000) Hz
S&M Figure 8
S&M Figure 8
-
-
4a,b
4a,b
The sum of
The sum of
three sinusoids
three sinusoids
does not have
does not have
any aliased
any aliased
frequencies
frequencies
since the
since the
sampling
sampling
frequency
frequency
f
f
S S
is greater than
is greater than
twice the
twice the
highest
highest
frequency
frequency
f
f
max max
f
f
S S
> 2
> 2
f
f
max max
S&M Figure 8
S&M Figure 8
-
-
4a,c
4a,c
S&M Figure 8
S&M Figure 8
-
-
5
5
The frequency
The frequency
2
2
f
f
max max
is called
is called
the
the
Nyquist
Nyquist
frequency
frequency
.
.
Harry Nyquist,
Harry Nyquist,
who contributed
who contributed
to the understanding of thermal noise
to the understanding of thermal noise
while at Bell Labs, is also remembered
while at Bell Labs, is also remembered
in electrotechnology for his analysis of
in electrotechnology for his analysis of
sampled data signals.
sampled data signals.
S&M Figure 8
S&M Figure 8
-
-
4a
4a
Harry Nyquist
Harry Nyquist
1889
1889
-
-
1976
1976
The analog signal is

The analog signal is
reconstructed
reconstructed
from the quantized
from the quantized
samples by a DAC and a low pass filter (LFP).
samples by a DAC and a low pass filter (LFP).
S&M Figure 8
S&M Figure 8
-
-
6
6
For
For
practical signals
practical signals
f
f
S S
> 2
> 2
f
f
max max
using a
using a
guard band
guard band
for
for
LPFs
LPFs
S&M Figure 8
S&M Figure 8
-
-
7
7
f
f
S S
= 2
= 2
f
f
max max
f
f
S S
> 2
> 2
f
f
max max
guard band
guard band
With
With
out
out
-
-
of
of
-
-
band noise
band noise
and sample signals, aliases of the
and sample signals, aliases of the
noise now appear
noise now appear
in
in
-
-
band
band
and should be filtered before the
and should be filtered before the
sampling process.
sampling process.
S&M Figure 8
S&M Figure 8
-
-
8
8
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Sampling Baseband
Sampling Baseband
Analog Signals
Analog Signals
Pages 149
Pages 149
-
-
182
182
The periodic baseband

The periodic baseband
signal consisting of
signal consisting of
three sinusoids is
three sinusoids is
impulse sampled
impulse sampled
,
,
sampled
sampled
-
-
and
and
-
-
held
held
,
,
processed by an
processed by an
8
8
-
-
bit ADC
bit ADC
-
-
DAC
DAC
and a
and a
quantizer
quantizer
in
in
Simulink
Simulink
.
.
MS Figure 4.1
MS Figure 4.1
The periodic baseband signal is the sum of a 1 V 500 Hz, a

The periodic baseband signal is the sum of a 1 V 500 Hz, a
0.5 V 1.5 kHz and a 0.2 V 2.5 kHz sinusoid.
0.5 V 1.5 kHz and a 0.2 V 2.5 kHz sinusoid.
MS Figure 4.2
MS Figure 4.2
The power spectral density (PSD) of the periodic baseband

The power spectral density (PSD) of the periodic baseband
signal has the expected peaks at 0.5, 1.5 and 2.5 kHz.
signal has the expected peaks at 0.5, 1.5 and 2.5 kHz.
MS Figure 4.3
MS Figure 4.3 Three sinusoids
Three sinusoids
The periodic baseband signal is overlaid with the continuous

The periodic baseband signal is overlaid with the continuous
amplitude sample
amplitude sample
-
-
and
and
-
-
hold signal with
hold signal with
f
f
S S
= 8 kHz.
= 8 kHz.
MS Figure 4.4
MS Figure 4.4
0.125 msec
0.125 msec
The analog signal

The analog signal
x(t
x(t
) here is sampled and held rather than
) here is sampled and held rather than
impulse sampled:
impulse sampled:
The power spectral density (PSD
The power spectral density (PSD
s s- -h h
) of the sample and hold
) of the sample and hold
operation is:
operation is:
= =
=
s-h
n
y (t) x(n ) h(t n ) h(t) 1 0 t
h(t) 0
S S S
T T T
otherwise
MS Eq. 4.3
MS Eq. 4.3
( )
( )
=
=
2 2 2 2
s-h S
k =
2 2
s-h
k =
PSD | X(f k ) | T sinc 2 f
PSD | X(f k ) | sinc 2 f

S S S
S S
f f T
f T
MS Eq. 4.4
MS Eq. 4.4
However, if the analog signal

However, if the analog signal
x(t
x(t
) is impulse sampled:
) is impulse sampled:
Then the power spectral density (PSD) does not have a
Then the power spectral density (PSD) does not have a
sinc
sinc
2 2
term:
term:
The PSD
The PSD
s s- -h h
does have the sinc
does have the sinc
2 2
term:
term:
MS Eq. 4.1
MS Eq. 4.1
( )
=
=
2 2
k =
2 2
s-h
k =
PSD | X(f k ) |
S S
S S
f f
f T
MS Eq. 4.4
MS Eq. 4.4
=
n
x(n ) x(t) (t n )
S S
T T
MS Eq. 4.2
MS Eq. 4.2
The PSD of the impulse sampled sum of three sinusoid

The PSD of the impulse sampled sum of three sinusoid
signal with
signal with
f
f
S S
= 8 kHz is:
= 8 kHz is:
MS Figure 4.5
MS Figure 4.5 8 kHz
8 kHz
16 kHz
16 kHz
No sinc
No sinc
2 2
term
term
2 2
k =
PSD | X(f k ) |
S S
f f
MS Eq. 4.2
MS Eq. 4.2
The PSD of the continuous amplitude sample and hold

The PSD of the continuous amplitude sample and hold
sum of three sinusoid signal with
sum of three sinusoid signal with
f
f
S S
= 8 kHz is:
= 8 kHz is:
MS Figure 4.6
MS Figure 4.6
8 kHz
8 kHz
16 kHz
16 kHz
sinc
sinc
2 2
term
term
( )
( )
=
=
2 2 2 2
s-h
k =
2 2
s-h
k =

S S S S
S S
f f T T
f T
MS Eq. 4.4
MS Eq. 4.4
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Sampling Bandpass
Sampling Bandpass
Analog Signals
Analog Signals
Pages 399
Pages 399
-
-
400
400
A bandpass signal does not need to be sampled at 2

A bandpass signal does not need to be sampled at 2
f
f
2 2
.
.
Nyquist
Nyquist
s bandpass sampling theory states that the

s bandpass sampling theory states that the
sampling rate
sampling rate
f
f
S S
> 2(
> 2(
f
f
2 2

f
f
1 1
) which is substantially less than
) which is substantially less than
2
2
f
f
2 2
S&M Figure 8
S&M Figure 8
-
-
9
9
f f
1 1
f f
2 2
f f
S S
= 20 ksamples/sec = 20 ksamples/sec
f f
S S
= 7 ksamples/sec = 7 ksamples/sec
LPF 10 kHz LPF 10 kHz
BPF 8 BPF 8- -10 kHz 10 kHz
8 10 kHz 8 10 kHz
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Sampling Bandpass
Sampling Bandpass
Analog Signals
Analog Signals
Pages 180
Pages 180
-
-
181
181
The
The
Simulink
Simulink
simulation uses the DSB AM modulation
simulation uses the DSB AM modulation
block and the sum of three sinusoids source.
block and the sum of three sinusoids source.
MS Figure 4
MS Figure 4
-
-
32
32
MS Figure 4
MS Figure 4
-
-
33
33
The
The
Simulink
Simulink
simulation initially uses a sampling rate of
simulation initially uses a sampling rate of
5 MHz and results in 4 194 304 = 2
5 MHz and results in 4 194 304 = 2
22 22
sampling points. The
sampling points. The
PSD shows the DSB
PSD shows the DSB
-
-
LC AM signal with the LSB and USB.
LC AM signal with the LSB and USB.
MS Figure 4
MS Figure 4
-
-
33
33
LSB USB
LSB USB
f
f
C C
Scaled PSD
Scaled PSD
f
f
max max
= 50 kHz
= 50 kHz
The bandwidth of the bandpass signal is

The bandwidth of the bandpass signal is
f
f
2 2

f
f
1 1
= 22.5
= 22.5
17.5 = 5 kHz and the

17.5 = 5 kHz and the
Simulink
Simulink
sampling rate is set to
sampling rate is set to
50 kHz
50 kHz
and results in only 32 768 = 2
and results in only 32 768 = 2
15 15
sampling points.
sampling points.
MS Figure 4
MS Figure 4
-
-
34
34
LSB USB
LSB USB
f
f
C C
Scaled PSD
Scaled PSD
f
f
max max
= 50 kHz
= 50 kHz
Aliased frequency range > 25 kHz
Aliased frequency range > 25 kHz
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Quantizing Process:
Quantizing Process:
Uniform Quantization
Uniform Quantization
Pages 400
Pages 400
-
-
404
404
The
The
quantizing process
quantizing process
divides the
divides the
range
range
(
(
full scale) into

full scale) into
2
2
n n
(n = 4 here) regions which are assigned an n
(n = 4 here) regions which are assigned an n
-
-
bit binary
bit binary
code.
code.
S&M Figure 8
S&M Figure 8
-
-
10
10
The
The
error
error
associated with the quantizing process is
associated with the quantizing process is
assumed to have a uniform probability density function.
assumed to have a uniform probability density function.
The maximum error for uniform quantization is:
The maximum error for uniform quantization is:
The quantizer range is
The quantizer range is
V
V
max max
and the uniform quantizer
and the uniform quantizer
voltage step size is:
voltage step size is:
The mean square quantizing E
The mean square quantizing E
q q
is the normalized
is the normalized
quantizing noise power:
quantizing noise power:
S&M Figure 8
S&M Figure 8
-
-
11
11
| |
= =
|
\ .
max max
n n
2 V V
q 0.5
2 2
= =
max max
n n-1
2 V V
2 2
( )
( )
/ 2
/ 2
= = = =
2 2 2
2
max max
q
2
2n
n
V V 1
E q dq
12
3 2
3 2
MS Eq. 4.7
MS Eq. 4.7
MS Eq. 4.6
MS Eq. 4.6
The
The
signal to quantizing noise power
signal to quantizing noise power
(SNR
(SNR
q q
) is:
) is:
P
P
S S
is the normalized power of the signal that is quantized.
is the normalized power of the signal that is quantized.
For the ADC here
For the ADC here
= 10 mV and n = 8. The sum of three

= 10 mV and n = 8. The sum of three
sinusoids as the input signal has a peak amplitude of 1.1 V
sinusoids as the input signal has a peak amplitude of 1.1 V
and the quantizing noise has a peak amplitude of 10 mV.
and the quantizing noise has a peak amplitude of 10 mV.
MS Figure 4.7
MS Figure 4.7
MS Eq. 4.8
MS Eq. 4.8
( )
= =
2n
S S
q
2 2
max
12 P P
SNR 3 2
V
10 mV
10 mV
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Quantizing Process:
Quantizing Process:
Nonuniform Quantization
Nonuniform Quantization
Pages 400
Pages 400
-
-
404
404
Nonuniform quantization divides the dynamic range of an

Nonuniform quantization divides the dynamic range of an
analog signal into nonuniform quantization regions. Lower
analog signal into nonuniform quantization regions. Lower
magnitudes have smaller quantization regions than high
magnitudes have smaller quantization regions than high
magnitudes.
magnitudes.
The quantization of speech benefits from nonuniform
The quantization of speech benefits from nonuniform
quantization since the perception of hearing is
quantization since the perception of hearing is
logarithmical
logarithmical
rather than linear.
rather than linear.
Uniform
Uniform
quantization
quantization
(top) results in
(top) results in
a large amount
a large amount
of error for
of error for
small sample
small sample
amplitude.
amplitude.
Non
Non
-
-
uniform
uniform
quantization
quantization
(bottom)
(bottom)
reduces the
reduces the
error for small
error for small
sample
sample
amplitudes.
amplitudes.
S&M Figure 8
S&M Figure 8
-
-
13
13
Uniform quantization Uniform quantization
Nonuniform quantization Nonuniform quantization
Uniform quantization is simpler to implement so a

Uniform quantization is simpler to implement so a
compressor
compressor
(a non
(a non
-
-
linear transfer function) is used before
linear transfer function) is used before
the quantizer.
the quantizer.
The
The
-
-
Law
Law
compressor
compressor
is used in
is used in
telephony
telephony
with MS Eq. 4.9
with MS Eq. 4.9
= 255. At the receiver

= 255. At the receiver
an expander has the
an expander has the
inverse non
inverse non
-
-
linear
linear
transfer function and
transfer function and
results in
results in
companding
companding
(
(
COMpressing
COMpressing
and
and
exPANDING
exPANDING
).
).
( )
| |
|
|
\ .

in
max
in
out max
max
V
ln 1+
V
V
V = V 0 1
ln 1+ V
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Companding
Companding
Pages 157
Pages 157
-
-
159
159
The
The
-
-
Law compander concept can be simulated in
Law compander concept can be simulated in
Simulink
Simulink
with the
with the
-
-
Law Compressor and
Law Compressor and
-
-
Law Expander
Law Expander
blocks. The A
blocks. The A
-
-
Law Compressor and A
Law Compressor and A
-
-
Law Expander
Law Expander
blocks are included for comparison.
blocks are included for comparison.
MS Figure 4.13
MS Figure 4.13
The
The
-
-
Law compressor voltage transfer function is
Law compressor voltage transfer function is
sigmoidal
sigmoidal
(S
(S
-
-
shaped).
shaped).
MS Figure 4.14
MS Figure 4.14
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Pulse Code Modulation

Pulse Code Modulation
Pages 171
Pages 171
-
-
175
175
The
The
pulse code modulator (PCM) transmitter utilizes a
pulse code modulator (PCM) transmitter utilizes a
Simulink
Simulink
-
-
Law compressor block, an 8
Law compressor block, an 8
-
-
bit ADC subsystem,
bit ADC subsystem,
an 8
an 8
-
-
bit DAC subsystem and a
bit DAC subsystem and a
-
-
Law expander block.
Law expander block.
MS Figure 4.21
MS Figure 4.21
The
The
Simulink
Simulink
8
8
-
-
bit ADC subsystem has a sample
bit ADC subsystem has a sample
-
-
and
and
-
-
hold
hold
block controlled by a
block controlled by a
sampling
sampling
pulse generator, an 8
pulse generator, an 8
-
-
bit
bit
encoder block, an integer
encoder block, an integer
-
-
to
to
-
-
bit converter block which
bit converter block which
provides an 8
provides an 8
-
-
bit
bit
vector
vector
to a demultiplexer block and a
to a demultiplexer block and a
multiport switch.
multiport switch.
An 8
An 8
-
-
level
level
staircase
staircase
subsystem
subsystem
sequences the multiport switch to select 1 of the 8 inputs for
sequences the multiport switch to select 1 of the 8 inputs for
bit serial output.
bit serial output.
MS Figure 4.22
MS Figure 4.22
The 8
The 8
-
-
level
level
staircase Simulink
staircase Simulink
subsystem sequences the
subsystem sequences the
multiport switch with a 3
multiport switch with a 3
-
-
bit counter and a 3
bit counter and a 3
-
-
bit DAC for the
bit DAC for the
output.
output.
MS Figure 4.22
MS Figure 4.22
3
3
-
-
bit DAC
bit DAC
3
3
-
-
bit counter
bit counter
The 8
The 8
-
-
bit DAC
bit DAC
Simulink
Simulink
subsystem for the PCM system
uses a 8
uses a 8
-
-
bit
bit
shift register
shift register
and an 8
and an 8
-
-
bit DAC.
bit DAC.
MS Figure 4.24
MS Figure 4.24
8
8
-
-
bit DAC
bit DAC
8
8
-
-
bit shift
bit shift
register
register
The 8
The 8
-
-
bit DAC
bit DAC
Simulink
Simulink
uses a 8
uses a 8
-
-
bit
bit
shift register
shift register
and an 8
and an 8
-
-
bit DAC.
bit DAC.
MS Figure 4.24
MS Figure 4.24
8
8
-
-
bit DAC
bit DAC
8
8
-
-
bit shift
bit shift
register
register
Analog input signal to the PCM system

MS Figure 4.25
MS Figure 4.25
MS Figure 4.21
MS Figure 4.21
8
8
-
-
bit DAC output after 8
bit DAC output after 8
-
-
bit ADC and
bit ADC and
-
-
Law Compressor
Law Compressor
MS Figure 4.25
MS Figure 4.25
MS Figure 4.21
MS Figure 4.21
-
-
Law Expander block
Law Expander block
output of the PCM system
MS Figure 4.25
MS Figure 4.25
MS Figure 4.21
MS Figure 4.21
LPF
LPF
MS Figure 4.25
MS Figure 4.25
MS Figure 4.21
MS Figure 4.21

MS Figure 4.25
MS Figure 4.25
LPF output of the PCM system

LPF output of the PCM system
startup
startup
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Differential Pulse Code

Modulation
Modulation
Pages 407
Pages 407
-
-
411
411
Sampled speech data are

Sampled speech data are
highly correlated
highly correlated
and
and
differential
differential
pulse code modulation
pulse code modulation
(DPCM) exploits this to lower the
(DPCM) exploits this to lower the
overall data
overall data
rate.
rate.
DPCM
DPCM
uses a
uses a
predictor
predictor
to subtract
to subtract
a predicted
a predicted
value from
value from
the input.
the input.
The
The
error
error
difference
difference
is sent.
is sent.
S&M Figure 8
S&M Figure 8
-
-
15
15
The predictor is a recursive equation, for example:

The predictor is a recursive equation, for example:
S
S
(n
(n
) = 0.75 s(n
) = 0.75 s(n
1) + 0.2 s(n
1) + 0.2 s(n
2) +0.05 s(n
2) +0.05 s(n
3)
3)
where
where
S
S
(n
(n
)
)
is the
is the
predicted
predicted
value of the
value of the
n
n
th
th
sample
sample
and
and
s(n
s(n
-
-
i)
i)
is the
is the
n
n
-
-
i
i
th
th
sample.
sample.
The error
The error
signal is
signal is
s(n
s(n
)
)
S
S
(n
(n
)
)
S&M Figure 8
S&M Figure 8
-
-
15
15
A typical continuous analog signal is sampled and results in

A typical continuous analog signal is sampled and results in
a discrete signal
a discrete signal
s(n
s(n
), The discrete predicted signal
), The discrete predicted signal
S
S
(n
(n
) is
) is
recursively computed. The discrete error signal is
recursively computed. The discrete error signal is
transmitted and has less quantizing bits than the actual
transmitted and has less quantizing bits than the actual
discrete signal.
discrete signal.
S&M Figure 8
S&M Figure 8
-
-
16b
16b
A DPCM example of actual discrete values, predicted

A DPCM example of actual discrete values, predicted
values and the error terms:
values and the error terms:
S&M Table 8
S&M Table 8
-
-
3
3
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion

Modulation
Modulation
Pages 175
Pages 175
-
-
180
180
A 4
A 4
-
-
bit first order
bit first order
differential pulse code modulator
differential pulse code modulator
(DPCM)
(DPCM)
can be simulated in
can be simulated in
Simulink.
Simulink.
MS Figure 4.26
MS Figure 4.26
The first order linear predictor MetaSystem determines the

The first order linear predictor MetaSystem determines the
error signal:
error signal:
e(n
e(n
) = s(n+1)
) = s(n+1)
2
2
s(n
s(n
) + s(n
) + s(n
-
-
1)
1)
MS Figure 4.27
MS Figure 4.27
error signal
error signal
input
input
ADC conversion command
ADC conversion command
The
The
Simulink
Simulink
4
4
-
-
bit ADC subsystem of the DPCM system is
bit ADC subsystem of the DPCM system is
similar to the 8
similar to the 8
-
-
bit ADC of the PCM system and illustrates
bit ADC of the PCM system and illustrates
design reuse.
design reuse.
MS Figure 4.28
MS Figure 4.28
The
The
Simulink
Simulink
4
4
-
-
bit DAC subsystem of the DPCM system is
bit DAC subsystem of the DPCM system is
also similar to the 8
also similar to the 8
-
-
bit DAC of the PCM system and
bit DAC of the PCM system and
again illustrates
again illustrates
design reuse.
design reuse.
MS Figure 4.29
MS Figure 4.29
4
4
-
-
bit DAC
bit DAC
4
4
-
-
bit shift
bit shift
register
register
The first order linear predictor

The first order linear predictor
Simulink
Simulink
subsystem
subsystem
reconstructs an estimate of the signal
reconstructs an estimate of the signal
s
s
e e
(n
(n
) from the error
) from the error
signal
signal
e(n
e(n
) received and past estimates:
) received and past estimates:
s
s
e e
(n+1) = e(n+1) + 2
(n+1) = e(n+1) + 2
s
s
e e
(n
(n
)
)
s
s
e e
(n
(n
1)
1)
MS Figure 4.30
MS Figure 4.30
reconstructed
reconstructed
signal
signal
input
input
Analog in
Analog in
put signal of the DPCM system
put signal of the DPCM system
MS Figure 4.31
MS Figure 4.31
MS Figure 4.31
MS Figure 4.31
Output of the 4
Output of the 4
-
-
bit first order
bit first order
DPCM system
DPCM system
MS Figure 4.31
MS Figure 4.31
MS Figure 4.31
MS Figure 4.31
Analog input signal to the DPCM system

Analog input signal to the DPCM system
MS Figure 4.31
MS Figure 4.31
Output of the 4
Output of the 4
-
-
bit first order DPCM system
startup
startup
Output of the 8
Output of the 8
-
-
bit PCM system
bit PCM system
MS Figure 4.31
MS Figure 4.31
Output of the 4
Output of the 4
-
-
startup
startup
startup
startup
MS Figure 4.25
MS Figure 4.25
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Delta Modulation
Delta Modulation
Pages 411
Pages 411
-
-
415
415
Delta modulation
Delta modulation
is an extreme example of DPCM
is an extreme example of DPCM
using
using
1
1
-
-
bit data representing
bit data representing
:
:
S
S
(n
(n
) =
) =
S
S
(n
(n
1) +
1) +
b
b
i i
= 1 if
= 1 if
S
S
(n
(n
1)
1)
s(n
s(n
1)
1)
S
S
(n
(n
) =
) =
S
S
(n
(n
1)
1)
b
b
i i
= 0 if
= 0 if
S
S
(n
(n
1) >
1) >
s(n
s(n
1)
1)
S&M Eq. 8.10
S&M Eq. 8.10
S&M Figure 8
S&M Figure 8
-
-
18
18
DM transmitter DM transmitter
DM receiver DM receiver
The reconstructed signal increments

The reconstructed signal increments
on each
on each
transmitted bit.
transmitted bit.
b
b
i i
= 1
= 1
S
S
(n
(n
) =
) =
S
S
(n
(n
1) +
1) +
b
b
i i
= 0
= 0
S
S
(n
(n
) =
) =
S
S
(n
(n
1)
1)
S&M Figure 8
S&M Figure 8
-
-
19
19
4 1s 4 0s 4 1s 4 0s
Chapter 8
Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion
Delta Modulation
Delta Modulation
Pages 72
Pages 72
-
-
75
75
Delta modulation (DM) can be simulated in

Delta modulation (DM) can be simulated in
Simulink
Simulink
. The
. The
DM receiver utilizes a sample and hold token as an
DM receiver utilizes a sample and hold token as an
accumulator and the step size
accumulator and the step size
= 20 mV.
= 20 mV.
MS Figure 2.61
MS Figure 2.61
DM transmitter
DM transmitter
DM receiver
DM receiver
f = 2 Hz
f = 2 Hz
A = 1 V
A = 1 V
f
f
S S
= 2 kHz
= 2 kHz
T
T
S S
= 0.5 msec
= 0.5 msec
DM can be subject to
DM can be subject to
slope overload
slope overload
which occurs when:
which occurs when:
/
/
T
T
S S
< max | d
< max | d
m(t
m(t
) /
) /
dt
dt
| SVU Eq. 2.61 modified
| SVU Eq. 2.61 modified
Here the sinusoid has A = 1 but f = 10 Hz and:
Here the sinusoid has A = 1 but f = 10 Hz and:
/
/
T
T
S S
= 20 mV / 0.5 msec = 40 < max | d
= 20 mV / 0.5 msec = 40 < max | d
m(t
m(t
) /
) /
dt
dt
| = 80
| = 80
and slope overload occurs.

and slope overload occurs.
MS Figure 2.63
MS Figure 2.63
sinusoidal
sinusoidal
input signal
input signal
Granular noise
Granular noise
occurs
occurs
in
in
DM because if the input
DM because if the input
m(t
m(t
) is
) is
constant the received signal
constant the received signal
oscillates
oscillates
by
by
because there
because there
is no
is no
0
0
possible.
possible.
Clocking
Clocking
occurs at the DM symbol interval
occurs at the DM symbol interval
T
T
S S
= 0.5 msec.
= 0.5 msec.
MS Figure 2.64
MS Figure 2.64
= 20 mV
= 20 mV
=
=
20 mV
20 mV
T
T
S S
= 0.5 msec
= 0.5 msec
The tradeoff between slope overload and granular noise is

The tradeoff between slope overload and granular noise is
that a large value of
that a large value of
(to avoid slope overload) would

(to avoid slope overload) would
increase granular noise. A decrease in
increase granular noise. A decrease in
T
T
s s
(again to avoid
(again to avoid
slope overload) would increase the data rate
slope overload) would increase the data rate
r
r
S S
.
.
The step size

The step size
= 20 mV and
= 20 mV and
T
T
S S
= 0.5 msec (
= 0.5 msec (
r
r
S S
= 2 kb/sec)
= 2 kb/sec)
here.
here.
For a 10 Hz sinusoidal input signal to a DM:

For a 10 Hz sinusoidal input signal to a DM:
m(t
m(t
) = sin (2
) = sin (2
10t)
10t)
max | d
max | d
m(t
m(t
) /
) /
dt
dt
|
|
= 20
= 20
If step size
If step size
= 20 mV and
= 20 mV and
T
T
S S
= 0.5 msec then
= 0.5 msec then
/ T
/ T
S S
= 40 < max | d
= 40 < max | d
m(t
m(t
) /
) /
dt
dt
| = 20
| = 20
so
so
slope overload is
slope overload is
predicted to occur
predicted to occur
.
.
For the 10 Hz sinusoidal input signal to a DM

For the 10 Hz sinusoidal input signal to a DM
m(t
m(t
) = sin (2
) = sin (2
10t)
10t)
max | d
max | d
m(t
m(t
) /
) /
dt
dt
|
|
= 20
= 20
and the step size remains

and the step size remains
= 20 mV but
= 20 mV but
T
T
S S
= 0.25 m
= 0.25 m
sec
sec
then
then
/ T
/ T
S S
= 80 > max | d
= 80 > max | d
m(t
m(t
) /
) /
dt
dt
| and slope overload is
| and slope overload is
mitigated but
mitigated but
r
r
S S
= 4 kb/sec.
= 4 kb/sec.
In comparison, an 8
In comparison, an 8
-
-
bit PCM system sampling a 10 Hz
bit PCM system sampling a 10 Hz
sinusoid at a reasonable sampling rate of 500 Hz (50
sinusoid at a reasonable sampling rate of 500 Hz (50
sampling points/period) has
sampling points/period) has
r
r
b b
= 8(500) = 4 kb/sec or
= 8(500) = 4 kb/sec or
r
r
b b
= r
= r
S S
but PCM is more complicated than DM.
but PCM is more complicated than DM.
End of Chapter 8
End of Chapter 8
Analog
Analog
-
-
to
to
-
-
Digital and
Digital and
Digital
Digital
-
-
to
to
-
-
Analog
Analog
Conversion
Conversion

Chapter 8 MS

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 8 MS

Uploaded by

Copyright:

Available Formats

EE4512 Analog and Digital Communications Chapter 8

Sampling and Quantization

Digital data transmission

Digital data transmission requires analog

ADC sampling occurs at a uniform rate (the

The continuous amplitude sample is then quantized to

The analog signal

Temporal and spectral representation of the continuous

5000 | and (5500

The analog signal is

The periodic baseband

The periodic baseband signal is the sum of a 1 V 500 Hz, a

The power spectral density (PSD) of the periodic baseband

The periodic baseband signal is overlaid with the continuous

The analog signal

However, if the analog signal

The PSD of the impulse sampled sum of three sinusoid

The PSD of the continuous amplitude sample and hold

A bandpass signal does not need to be sampled at 2

s bandpass sampling theory states that the

The bandwidth of the bandpass signal is

17.5 = 5 kHz and the

full scale) into

= 10 mV and n = 8. The sum of three

Nonuniform quantization divides the dynamic range of an

Uniform quantization is simpler to implement so a

= 255. At the receiver

Pulse Code Modulation

Analog input signal to the PCM system

Analog input signal to the PCM system

LPF output of the PCM system

Differential Pulse Code

Sampled speech data are

The predictor is a recursive equation, for example:

A typical continuous analog signal is sampled and results in

A DPCM example of actual discrete values, predicted

Differential Pulse Code

The first order linear predictor MetaSystem determines the

The first order linear predictor

Analog input signal to the DPCM system

The reconstructed signal increments

Delta modulation (DM) can be simulated in

and slope overload occurs.

The tradeoff between slope overload and granular noise is

(to avoid slope overload) would

The step size

For a 10 Hz sinusoidal input signal to a DM:

For the 10 Hz sinusoidal input signal to a DM

and the step size remains

You might also like