Merits of Digital Communication:
1. Digital signals are very easy to receive. The receiver has to just detect
whether the pulse is low or high.
2. AM - FM signals become corrupted over much short distances as
compared to digital signals. In digital signals, the original signal can be
reproduced accurately.
3. The signals lose power as they travel, which is called attenuation. When
AM and FM signals are amplified, the noise also get amplified. But the
digital signals can be cleaned up to restore the quality and amplified by
the regenerators.
4. The noise may change the shape of the pulses but not the pattern of the
pulses.
5. AM and FM signals can be received by any one by suitable receiver. But
digital signals can be coded so that only the person, who is intended for,
can receive them.
6. AM and FM transmitters are ‘real time systems’. I.e. they can be received
only at the time of transmission. But digital signals can be stored at the
receiving end.
7. The digital signals can be stored, or used to produce a display on a
computer monitor or converted back into analog signal to drive a loud
speaker.
Pulse Modulation
Analog Pulse Modulation Digital Pulse Modulation
Pulse Amplitude (PAM) Pulse Code (PCM), DPCM
Pulse Width (PWM) Delta (DM), ADM
Pulse Position (PPM)
Pulse Code Modulation
PCM is a method of converting an analog
signal into a digital signal. (A/D conversion)
The amplitude of Analog signal can take any
value over a continuous range i.e. it can take
on an infinite values.
Digital signal amplitude can take on finite
values.
Analog signal can be converted into digital by
sampling and quantizing.
Pulse Code Modulation (PCM):
* Analog signal is converted into digital signal by using a digital
code.
* Analog to digital converter employs two techniques:
1. Sampling: The process of generating pulses of zero width
and of amplitude equal to the instantaneous amplitude of the
analog signal. The no. of pulses per second is called
“sampling rate”.
2. Quantization: The process of dividing the maximum value
of the analog signal into a fixed no. of levels in order to
convert the PAM into a Binary Code.
The levels obtained are called “quanization levels”.
* A digital signal is described by its ‘bit rate’ whereas analog
signal is described by its ‘frequency range’.
* Bit rate = sampling rate x no. of bits / sample
V Sampling,
o Quantization and
l Coding
t
a
g
e
Time
7 111
L 6 110 B
e 5 101 i C
v 4 100 n o
e 3 011 a d
l 2 010 e
r
s 1 001 s
0 000 y
Time
V
o 010101110111110101010
l
t
a
g
e
Time
Cont.
The amplitude of analog signal m(t) lie in the
range (-mp, mp) and is partitioned into L
sub-intervals each of magnitude 2mp/L
Binary pulse codes
Quantizing
Digital signals come from variety of sources
e.g. computer
Some sources are analog but are converted
into digital form by variety of techniques such
as PCM and DM
For quantizing , we limit the amplitude of m(t)
to a range(-mp, mp) as shown in the previous
slides
This amplitude is uniformly divided into L
subintervals and each interval is ,
Cont.
A sample value is approximated by the mid
point of the interval
The quantized samples are coded and
transmitted as binary pulses
At the receiver some pulses will be detected
incorrectly
There are two types of errors
Quantization error
Pulse detection error
Cont.
In almost all practical schemes, the pulse
detection error is very small compared to the
quantization error and can be ignored
Now we analyze the quantization error
Cont.
Cont.
Cont.
The integral of the cross product terms is
zero and we obtain,
Because the sampling rate is 2B, hence the
total number of samples over the averaging
interval is 2BT
This is called the mean of the quantization
error
Cont.
The quantized levels are separated by 2mp/L
Since sample value is approximated by the
midpoint of the subinterval in which the
sample falls
The maximum quantization error is
The mean square quantizing error is
Cont.
Cont.
Non-uniform quantization
SNR is an indication of the quality of the received
signal
Ideally we would like to have constant SNR
Unfortunately, the SNR is directly proportional to the
signal power, which varies from talker to talker
The signal power can also vary because of the
connecting circuits
SNR vary even for the same talker, when the person
speaks softly
Smaller amplitudes pre-dominate in speech and larger
amplitude much less frequent.
This means the SNR will be low most of the time
Cont.
The root of this difficulty is that the
quantization steps are of uniform value
The quantization noise is directly
proportional to the square of the step size.
The problem can be solved by using smaller
steps for smaller amplitudes as shown in fig.
on the next slide
Cont.
Cont.
The same result can be obtained by first
compressing a signal and then using uniform
quantization
The input-output characteristics of
compressor are shown in fig.
Cont.
The horizontal axis is normalized input signal
and the vertical axis is the output signal y.
The compressor maps the input signal into
larger increments
Hence the interval delta(m) contains large
number of steps when m is small
The quantization noise is small for smaller
input signal
Thus loud talker and stronger signals are
penalized with higher noise steps in order to
compensate the soft talker and weak signals
Compression Laws
There are two laws regarding compressions
(1)
This law is used in North America and Japan
(2) A-Law
This law is used in Europe and the rest of the
word
Cont.
The compressed samples are restored to
their original values at receiver by using an
expander
The compressor and expander together are
called compandor.
Compression of a signal increases its
bandwidth but in PCM, we are not
compressing the signal but its samples the
number of samples does not change,
therefore bandwidth does not rise
When meu-law compandor is used then
Transmission BW and output SNR
For binary PCM, we assign distinct group of n
binary digits to each of the L quantization
levels
Each quantized level is encoded into n-bits
Minimum channel BW is
This is the theoretical minimum transmission
bandwidth required to transmit the PCM
signal
Example 6.2
A signal m(t) band-limited to 3kHz is sampled
at a rate 33.33% higher than Nyquist rate, a
maximum acceptable error in the sample
amplitude is 0.5% of the peak amplitude. The
quantized samples are binary coded. Find the
minimum channel BW required to transmit
the coded signal. If 24 such channels are
time-division multiplexed, determine the
minimum transmission BW required to
transmit the multiplexed signal
Solution
Exponential Increase of output SNR
SNR in decibel scale
Cont.
Prove that the output signal to quantization noise
ratio in decibel can be expressed as 1.8 - 20log 10L.
or
1.8 +6R.
Example 6.3
Comments on Logarithmic Units
Very small and very large values are
expressed in logarithmic units
T1 carrier system
A schematic of T1-system is shown in fig.
Cont.
Cont.
Differential Pulse Code Modulation
Taylor's series
Cont.
Analysis of DPCM
Cont.
Delta Modulation
Cont.
Cont.
Delta Modulator
Delta Demodulator
Delta Modulator output
Working of DM