Principles of digital data

communication
Block diagram of a digital
communication system
Advantages of digital communication
 Different parts
• Source
• Multiplexer
• Line coder
• Repeater
 Line Coder
• See the other ppt
Baseband and passband
communication
PSD of different line code
 Pulse shaping
• PSD of a digital signal is controlled by line
code.
• Influencing the PSD by shaping the pulse is
termed as pulse shaping.
 Intersymbol Interferences and
Effect
• Why does ISI arise?

• Spreading of a pulse beyond its allotted time

interval Tb will cause it to interfere with
neighboring pulses. This is known as
intersymbol interference-or ISI.
 How to eliminate ISI
• To eliminate ISI, Nyquist proposed three
different criteria for pulse shaping where the
pulses are allowed to overlap. Yet, they are
shaped to cause zero (or controlled)
interference with all the other pulses at the
decision-making instants, Thus, by limiting the
noninterference requirement only at the
decision-making instants, we eliminate the
need for the pulse to be totally
nonoverlapping.
 Nyquist's First Criterion for Zero ISI
• Proposed pulse shape

• Positive pulse—as an example

Expected Spectrum of Nyquist first
criterion pulse
• Pulse is termed sinc pulse.

• Is it practical?
 Drawback of first criterion pulse
• We will have to wait an infinite time to generate
it. Any attempt to truncate it would increase its
bandwidth beyond Rb/2 Hz.
• But even if this pulse were realizable, it would
have an undesirable feature: namely, it decays
too slowly at a rate 1/t. This causes some serious
practical problems. For instance, if the nominal
data rate of Rb bit/s required for this scheme
deviates a little, the pulse amplitudes will not
vanish at the other pulse centers.
 Solution
• The solution is to find a pulse p(t) that satisfies
Eq. (7.23) but decays faster than 1/t. Nyquist
has shown that such a pulse requires a
bandwidth kRb/2, with 1 < k < 2.
Derivation of zero ISI Nyquist pulse
• Lets take the spectrum of the pulse as follows
• Sampled desired pulse

• Taking FX on both sides of 7.25

• Graphically
• over the range 0 < f < Rb

• or
• Use of conjugate property
• The curve has an odd symmetry about the set
of axes intersecting at point alpha [the point
on curve at f = Rb/2]. Note that this
requires that
• r is called roll-off factor.
• One family that satisfies above spectrum
• r=1, PSD

• Corresponding time domain

• It should be remembered that it is the pulses
received at the detector input that should
have the form for zero ISI. In practice, because
the channel is not ideal (distortionless), the
transmitted pulses should be shaped so that
after passing through the channel with
transfer function Hc(f), they will be received
with the proper shape (such as raised-cosine
pulses) at the receiver. Hence, the transmitted
pulse pi (f) should satisfy
Class problems-
 Controlled ISI or Partial Response
Signaling
• Pulse
• What happens- (i) both pulses are positive (ii)
both are negative (iii) both are opposite

• Knowledge of the previous bit allows the

determination of the present bit.
• Advantages- error detection

• Between these full-valued samples of the same polarity,

there are an even number of zero-valued samples. If one of
the sample values is detected wrong, this rule is violated,
and the error is detected.
• The pulse p(t) goes to zero at t = —Tb and 2Tb, resulting
in the pulse width (of the primary lobe) 50% higher than
that of the first criterion pulse.
• This pulse broadening in the time domain leads to
reduction of its bandwidth. This is the second criterion
proposed by Nyquist.
• This scheme of controlled ISI is also known as correlative
or partial-response scheme.
• A pulse satisfying the second criterion in Eq. (7.37) is also
known as the duobinary pulse.
 Example of a Duobinary Pulse
• Pulse

• Decays with 1/t2

• Spectrum
• Problem-

• This pulse is not ideally realizable because p(t) is noncausal

and has infinite duration [because P(f) is band-limited].
However, it decays rapidly (as 1/t2), and therefore can be
closely approximated.

• It may come as a surprise that we are able to achieve the

theoretical rate using the duobinary pulse. In fact, it is an
illusion. The theoretical rate of transmission is 2 pieces of
independent information per second per hertz bandwidth.
We have achieved this rate for binary information. Here is the
catch!
 Modified Duobinary pulse

• See the text

Duobinary pulse generator
• Therefore, a simple detector of duobinary signaling is
to first detect all the bits corresponding to bk = ±2.
The remaining {bk) are zero-valued samples that
imply transition: that is, the current digit is 1 and the
previous digit is 0, or vice versa. This means the digit
detection must be is based on the previous digit.
• That is, error propagates.
• Soln-Differential coding
 Equalizers
• A pulse train is attenuated and distorted by the transmission
medium.

• The attenuation can be compensated by the preamplifier,

whereas the distortion is compensated by the equalizer.

• Channel distortion is in the form of dispersion, which is

caused by an attenuation of certain critical frequency
components of the data pulse train.

• Theoretically, an equalizer should have a frequency

characteristic that is the inverse of that of the transmission
medium. This will restore the critical frequency components
and eliminate pulse dispersion. Unfortunately, this also
enhances the received channel noise by boosting its
components at these critical frequencies. This undesirable
phenomenon is known as noise amplification.
 Equalizers
• For digital signals, however, complete equalization is really not
necessary, because a detector only needs to make relatively
simple decisions—such as whether the pulse is positive or
negative (or whether the pulse is present or absent).

• Therefore, considerable pulse dispersion can be tolerated.

Pulse dispersion results in ISI and the consequent increase in
error detection.

• Noise increase resulting from the equalizer (which boosts the

high frequencies) also increases the detection error
probability.

• For this reason, design of an optimum equalizer involves an

inevitable compromise between reducing ISI and reducing the
channel noise.
 Zero-Forcing Equalizer
• The design goal is to force the equalizer output-pulse to have
zero ISI values at the sampling (decision-making) instants.

• In other words, the equalizer output pulses satisfy the Nyquist

or the controlled ISI criterion, The time delay T between
successive taps is chosen to be Tb, the interval between
pulses.
Zero-Forcing Equalizer
Zero-Forcing Equalizer
Zero-Forcing Equalizer
 Zero-Forcing Equalizer
• Disadvantages-
Minimum Mean Square Error (MMSE)
Method
• In practice an alternative approach is to minimize the mean square
difference between the equalizer output response po[k] and the
desired zero ISI response. This is known as the minimum mean
square error (MMSE) method for designing transversal filter
equalizers.

• The MMSE method does not try to force the pulse samples to zero
at 2N points. Instead, we minimize the squared errors averaged
over a set of output samples.

• This method involves more simultaneous equations. Thus we must

find the equalizer tap values to minimize the average (mean) square
error over a larger window [—K,K].
Problems on equalizer
 PAM: M-ARY BASEBAND SIGNALING FOR
HIGHER DATA RATE
• pulse amplitude modulation (PAM): the data
information is conveyed by the varying pulse
amplitude.

• Information carried by each level

 PAM
• Advantages- date rate increases
• Cost of PAM: As M increases, the transmitted
power also increases as M. This is because to
have the same noise inununity, the minimum
separation between pulse amplitudes should be
comparable to that of binary pulses. Therefore,
pulse amplitudes increase with M. (Fig. 7.28). It
can be shown that the transmitted power
increases as M2 (Prob. 7.7-5). Thus, to increase
the rate of communication by a factor of log2 M ,
the power required increases as M2.
 DIGITAL CARRIER SYSTEMS
• Basic Binary Carrier Modulations
• Amplitude Shift Keying (ASK)-
• What is required to be noted to draw signals
• Phase shift keying (PSK)-

• Frequency shift keying (FSK)-

PSD of Digital Carrier Modulation
 Connections between Analog and
Digital Carrier Modulations
• There is a natural and clear connection between ASK and AM
because the message information is directly reflected in the
varying amplitude of the modulated signals. Because of its
nonnegative amplitude, ASK is essentially an AM signal with
modulation index = 1.

• There is a similar connection between FSK and FM. FSK is

simply an FM signal with only limited number of
instantaneous frequencies.

• The connection between PSK and analog modulation is a bit

more subtle.
 Connections between Analog and
Digital Carrier Modulations
• PSK is a digital manifestation of the DSB-SC
amplitude modulation.
• DSB-SC amplitude modulation is more power
efficient than AM.
• Binary PSK is therefore more power efficient than
ASK.
• In terms of bandwidth utilization, we can see
from their connection to analog modulations that
ASK and PSK have identical bandwidth
occupation while FSK requires larger bandwidth.
 Demodulation
• ASK detection- Just like AM, ASK can be demodulated both
coherently (for synchronous detection) or noncoherently
(for envelope detection).

• The coherent detector requires more elaborate equipment

and has superior performance, especially when the signal
power (hence SNR) is low. For higher SNR, the envelope
detector performs almost as well as the coherent detector.

• Hence, coherent detection is not often used for ASK

because it will defeat its very purpose (the simplicity of
detection). If we can avail ourselves of a synchronous
detector, we might as well use PSK, which has better power
efficiency than ASK.
 Demodulation
• FSK Detection-
 Demodulation
• PSK detection-
 Demodulation
• Why PSK detection- not possible by envelope
detection?

• these signals cannot be demodulated via

envelope detection because the envelope
stays constant for both 1 and 0
 Differential PSK

• Although envelope detection cannot be used

for PSK detection, it is still possible to exploit
the finite number of modulation phase values
for noncoherent detection.

• Indeed, PSK signals may be demodulated

noncoherently by means of an ingenious
method known as differential PSK.
 Differential PSK
• The principle of differential detection is for the receiver to
detect the relative phase change between successive
modulated phases. Since the phase value in PSK is finite
(equaling to 0 and pi in binary PSK), the transmitter can
encode the information data into the phase difference. For
example, a phase difference of zero represents 0 whereas a
phase difference of pi signifies 1.
• This technique is known as differential encoding (before
modulation). In one differential code, a 0 is encoded by the
same pulse used to encode the previous data bit (no
transition), and a 1 is encoded by the negative of the pulse
used to encode the previous data bit (transition).
 Differential PSK

• Addition is modulo-2
 Differential PSK
• The modulated signal consists of pulses
 Differential PSK
• Demodulation-
 Differential PSK
• In demodulation of DPSK, we avoid generation of a local carrier by
observing that the received modulated signal itself is a carrier (±A
ens wet) with a possible sign ambiguity.

• For demodulation, in place of the carrier, we use the received signal

delayed by Tb (one bit interval).

• If the received pulse is identical to the previous pulse, the product

y(t) = A2 cos2 wct = (A2/2)(1 cos 2wct), and the low-pass filter
output z (t) = A2/2. We immediately detect the present bit as 0. If
the received pulse and the previous pulse are of opposite polarity,
y(t) = —A2 oos2wct and z(t) = —A2/2, and the present bit is detected
as 11.

• Table 7.3 illustrates a specific example of the encoding and

decoding.
Differential PSK
 M-ARY DIGITAL CARRIER
MODULATION
• M-ary ASK:
 M-ary ASK
• This is still an AM signal that uses M different
amplitudes and a modulation index of 1.

• Its bandwidth remains the same as that of the

binary ASK, while its power is increased
proportionally with M2.

• Its demodulation would again be achieved via

envelope detection or coherent detection.
M-ary FSK
 Minimum frequency separation

• First term zero-why?

• This choice of minimum frequency separation is
known as the minimum shift FSK. Since it forms
an orthogonal set of symbols, it is often known as
orthogonal signaling. Other way to find-GS
orthogonalization procedure
 Bandwidth of M-ary FSK

• From, Carsons rule

 Bandwidth of M-ary FSK
• In fact, it can be in general shown that the bandwidth of an
orthogonal M-ary scheme is M times that of the binary scheme.
•
• Therefore, in an M-ary orthogonal scheme, the rate of
communication increases by a factor of log2 M at the cost of M -fold
transmission bandwidth increase.

• For a comparable noise ,immunity, the transmitted power is

practically independent of M in the orthogonal scheme.

• Therefore, unlike M-ary ASK, M-ary FSK does not require more
transmission power. However, its bandwidth requirement increases
almost linearly with M (compared with binary FSK or M-ary ASK).
 M-ary PSK,PAM,QAM
• PSK signalling:

• M-ary PSK in general

 QPSK
• The special PSK signaling with M = 4 is an extremely
popular and powerful digital modulation format. It in fact
is a summation of two binary PSK signals, one using the
(in-phase) carrier of cos wct while the other uses the
(quadrature) carrier of sin wct of the same frequency. For
this reason, it is also known as quadrature PSK (QPSK).
We can transmit and receive both of these signals on the
same channel, thus doubling the transmission rate.
 M-ary QAM
• Signalling:
 QAM
• M1(t) and m2(t)-sqrt(M) ary data/pulse
sequence
16 point QAM
 M-ary PAM
• Note that if we disable the data stream that modulates sin wct
in QAM, then all the signaling points can be reduced to a
single dimension. Upon setting m2(t) = 0, QAM becomes

• This degenerates into the pulse amplitude modulation or

PAM.

• Comparison of the signal expression of p (t) with the analog

DSB-SC signal makes it clear that PAM is the digital version of
the DSB-SC signal.
• Just as analog QAM is formed by the superposition of
two DSB-SC amplitude modulations in phase
quadrature, digital QAM consists of two PAM signals,
each having sqrt(M) signaling levels.

• Similarly, like the relationship between analog DSB-SC

and QAM, PAM requires the same amount of
bandwidth as QAM does. However, PAM is much less
efficient because it would need M modulation signaling
levels in one dimension, whereas QAM requires only
sqrt(M) signaling levels in each of the two orthogonal
QAM dimensions.
 Trading Power and Bandwidth
• The nature of the exchange between the transmission
bandwidth and the transmitted power (or SNR) depends
on the choice of M-ary scheme. For example, in
orthogonal signaling, the transmitted power is practically
independent of M but the transmission bandwidth
increases with M. Contrast this to the PAM case, where
the transmitted power increases roughly with M2 while
the bandwidth remains constant. Thus, M-ary signaling
allows us great flexibility in trading signal power (or SNR)
for transmission bandwidth. The choice of the
appropriate system will depend upon the particular
circumstances. For instance, it will he appropriate to use
QAM signaling if the bandwidth is at a premium (as in
telephone lines) and to use orthogonal signaling when
power is at a premium (as in space communication).