Modulation and Coding

2 | Comm 2 Techniques
53

UNIT 2: DIGITAL COMMUNICATIONS

2.0. Intended Learning Outcomes

By the end of this unit you should be able to:

• Define and describe pulse code modulation

• Explain PCM coding efficiency
• Define and explain dynamic range
• Explain the difference between linear and nonlinear PCM codes
• Explain how to determine PCM line speed
• Describe delta modulation PCM
• Describe the composition of digital pulses
• Differentiate the different shift keying
• Identify the unique characteristics of the different shift keying
• Articulate on M-ary Encoding
• Explain the bandwidth considerations for ASK, FSK, PSK, QAM
• Generate waveforms to show the graphs of ASK, FSK, PSK, and QAM

2.1. Introduction

In case of analog modulation techniques described so far, sine wave is used as the
carrier signal. Sine wave values are defined for all the instants of time and hence
analog modulation is also termed as continuous wave (CW) modulation. Nothing
prevents us from replacing the sine wave with another wave as the carrier. The most
useful one that helped in advancing the communication field is the pulse train in place
of sine wave. On the similar lines of sine wave being characterized in terms of its
parameters amplitude, frequency and phase, the pulse train can also be characterized
in terms of its parameters, namely, amplitude, width and position of the pulse. CW
modulation is obtained by varying one of the parameters of the sine wave with the
instantaneous variations of the message. Similarly, pulse modulation can be obtained
by varying one of the parameters of the pulse train with respect to the message. Pulse
modulation is further classified as pulse analog and pulse digital, depending on
whether the parameter of the pulse is continuous or discrete in nature. Collectively all
are termed as pulse modulation techniques. This chapter deals with studying different
pulse modulation techniques.

In case of pulse train, the pulses by themselves occur at discrete instants of time.
However, the parameters of the pulse, namely, amplitude, width and position are
continuous in nature. If amplitude of the pulse is made proportional to the message,
then it is termed as pulse amplitude modulation (PAM). Alternatively, if the width of
the pulse is made proportional to the message, then it is termed as pulse width
modulation (PWM). The position of the pulse, i.e., its instant of occurrence compared
to its position in the reference pulse train is varied in proportion to the message in
case of pulse position modulation (PPM). Finally, the amplitude of the pulse can be
approximately represented by a discrete amplitude value which leads to the pulse code

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
54

modulation (PCM). Further variants of PCM include delta modulation (DM) and
differential PCM (DPCM). To summarize, in case of pulse analog modulation, time is
discrete, but the pulse parameters are analog, whereas, both time and pulse
parameters are discrete in case of pulse digital modulation.

The major difference between CW and pulse modulations need to be understood. CW

modulation translates message from baseband to the passband-range and helps in
transmitting it for a longer distance as described in the earlier chapters.
Alternatively, pulse modulation translates message from analog fom1 to the discrete
form. That is, continuously varying message information is now represented et
discrete instants of time. Both forms of the message will remain in the baseband itself!
This is an important fact and should be in view when we are studying the various
pulse modulation techniques. Thus, the word modulation from the context of
frequency translation is a misnomer in this case. ln case of pulse modulation, it refers
to the process of modifying the pulse parameters with respect to message and nothing
else. The natural question then will be why pulse modulation? The answer is even
though it does not help in frequency translation, it helps in other aspects of signal
processing, namely, digital representation of message signal. As will be discussed in
detail later, some of the pulse modulation techniques are fundamental to the digital
communication field.

2.2. Topic/Discussion (with Assessment/Activities)

2.2.1. Pulse Analog Modulation Techniques

The pulse analog modulation techniques are of three types namely, PAM, PWM and
PPM. This section describes each of them and also about the recovery of message from
them.

2.2.1.1. Pulse Amplitude Modulation

Pulse amplitude modulation is defined as the process of varying the amplitude of the
pulse in proportion to the instantaneous variations of message signal. Let the message
signal be given by
𝑣𝑚 = 𝑉𝑚 𝑠𝑖𝑛𝜔𝑚 𝑡
If x(t) is a periodic signal with period 𝑇0 , then it should satisfy the definition stated
as x(t) = x (t + 𝑇0 ). The pulse train is a periodic signal with some fundamental period
say 𝑇0 . Then the information present in each period of the pulse train is given by

𝑝 = 𝑉𝑝 0≤𝑡≥∆
=0 ∆≤ 𝑡 ≥ 𝑇0
where ∆ is the width of the pulse and the leading edge of the pulse is assumed to be
coinciding with the starting of the interval in each period.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
55

The pulse amplitude modulated wave in the time domain is obtained by multiplying
the message with the pulse train and is given by

𝑃𝑎 = 𝑝 × 𝑣𝑚
Substituting p in the above equation we get

𝑃𝑎 = 𝑉𝑝 𝑉𝑚 𝑠𝑖𝑛𝜔𝑚 𝑡 0≤𝑡≥∆
=0 ∆≤ 𝑡 ≥ 𝑇0
Figure 5.1 shows the message, pulse train and PAM signal. The amplitude of the PAM
signal follows the message signal contour and hence the name. It can be shown that
the spectrum of PAM signal is a sine function present at all frequencies (for
derivation, please refer to the topic of Fourier series in any of the signals and systems
textbook). Of course, its significant spectral amplitude values will be in the low
frequency range and tapers off as we move towards the high frequency range. The
message signal is a low frequency signal. Multiplication of the two for generating the
PAM signal results in the convolution of their spectra in the frequency domain. Thus
PAM signal still retains the message spectrum in the low frequency range after
modulation. This is the difference between amplitude modulation of sine wave and
pulse train. Therefore, PAM is not useful like AM for communication. Alternatively,
PAM is found to be useful in understanding the sampling process to be described next.

Sampling Process Sampling is a signal processing operation that helps in sensing the
continuous time signal values at discrete instants of time. The sampled sequence will
have amplitudes equal to signal values at the sampling instants and undefined at all

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
56

other times. This process can be conveniently performed using PAM described above.
The sampling process can be treated as an electronic switching action as shown in Fig.
5.2, The continuous time signal to be sampled is applied to the input terminal. The
pulse train is applied as the control signal of the switch. When the pulse occurs, the
switch is in ON condition, that is, acts as short circuit between input and output
terminals. The output value will therefore be equal to input. During the other intervals
of the pulse train, the switch is in OFF condition, that is, acts as open circuit. The
output is therefore undefined. The output of the switch will be essentially a PAM
signal. Any active device like diode, transistor or FET can be used as a switch.

In the context of sampling process, there are other aspects that need to be considered
with respect to the pulse train. The first and foremost is how often the signal needs to
be sampled or sensed, so that when needed an approximate version of the continuous
time signal can be reconstructed. This is based on the well-known sampling theorem
which states that the sampling frequency (f) i.e., number a/samples per second should be
greater than or equal to twice the maximum frequency component (𝐹𝑚 ) of the input signal.

𝐹𝑠 ≥ 2𝐹𝑚
The minimum possible value of sampling frequency is termed as Nyquist rate. Thus
the sampling theorem will decide the periodicity associated with the pulse train. The
second important aspect is, the width of the pulse ∆ should not influence the
amplitude of the sampled value. Even though this point is not obvious in the time
domain, it can be understood by observing the frequency domain behavior of the
PAM process due to the convolution of sine function of pulse train with the input
signal spectrum. To minimize this effect, for all practical processing ∆→ 0. so that the
pulse train becomes on impulse train. The Fourier transform of an impulse train is
also an impulse train in the frequency domain. Therefore, convolution will not affect
the shape of the sampled signal. It only leads to periodicity of the spectrum!

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
57

Example 1:
A message signal made of multiple frequency; components has a maximum frequency
value of 4 kHz. Find out the minimum sampling frequency required according to the
sampling theorem.

Solution:
𝐹𝑚 = 4 𝑘𝐻𝑧
𝐹𝑠 ≥ 2 × 𝐹𝑚 = 2 × 4 𝑘𝐻𝑧 = 8 𝑘𝐻𝑧

Example 2:
A message signal has the following frequency components: a single tone sinewave
of 500 Hz and sound of frequency components with lowest value o/750 Hz and
highest value of 1800 Hz. What should be the minimum sampling frequency to sense
the information present in this signal according to the sampling theorem?
Solution:
𝐹𝑚 = 1800 𝐻𝑧
𝐹𝑠 ≥ 2 × 𝐹𝑚 = 2 × 1800 𝐻𝑧 = 3600 𝐻𝑧

2.2.1.2. Pulse Width Modulation

Pulse width modulation (PWM) is defined as the process of varying the width of the
pulse in proportion to the instantaneous variations of message.

Let ∆ be the width of the pulse in the unmodulated pulse train. ln PWM

∆ 𝛼 𝑣𝑚
Mathematically, the width of pulse in PWM signal is given by

∆𝑚 = ∆ (1 + 𝑣𝑚 )
When there is no message, i.e., 𝑣𝑚 = 0, then the width of the pulse will be equal to the
original width ∆ For positive values of message, the width will be proportionately
increases by (1 + 𝑣𝑚 ) factor. For negative values of message, the width decreases by
(1 − 𝑣𝑚 ) factor.

Figure 5.3 shows the generation of PWM signal. The amplitude of the pulse remains
constant in this case. Thus, PWM is more robust to noise compared to PAM. This is
the difference with respect to PAM signal. The mathematical treatment about the
frequency domain aspect of PWM is an· involved process. However, the resulting
PWM will still have the spectrum in the baseband region itself. The illustration given
in Fig. 5.3 is made only using trailing edge of the pulse. We can also perform the same
using either leading edge or both. Even though, the PWM signal also contains the
message information in the pulse train, it is seldom used as a sampling process to
discretize the continuous time signal as in PAM case due to its indirect way of storing
message information and also the randomness involved in the width modification.
Thus, PWM has limited use in signal processing and communication field.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
58

Alternatively, PWM finds use in power applications like direct current (de) motor
speed control as described next.

Speech Control of DC Motors using PWM. The speed of the de motor depends on the
average dc voltage applied across its terminals. Suppose if V volts is the voltage for
running the de motor at its full speed, then 0 volt is the voltage for the rest condition
of dc motor. Now, the speed of the dc motor can be varied from its rest to full speed
value by varying the dc voltage. This can be conveniently performed with the help of
PWM as illustrated in Fig. 5.4. The constant dc voltage source is applied across the
terminals of dc motor through a gating circuit controlled by the PWM signal. The
gating circuit will essentially convert the constant dc source into a variable dc source.
Suppose when there is no modulation, the width of the pulse will be the original value
∆ and let this run the dc motor at some speed. Now when the width increases, the
voltage value increases from its unmodulated case and hence the speed. It happens in
the opposite way for the decrease in width. Thus, PWM provides a convenient and
efficient approach for the speed control of dc motors.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
59

2.2.1.3. Pulse Position Modulation

Pulse position modulation (PPM) is defined as the process of varying the position of
the pulse with respect to the instantaneous variations of the message signal.

Let 𝑡𝑝 indicates the timing instant of the leading or trailing edge of the pulse in each
period of the pulse train. In PPM
𝑡𝑝 𝛼 𝑣𝑚
Mathematically, the position of the leading or trailing edge of the pulse (in each
period) in PPM signal is given by

𝑡𝑝 = 𝑓(𝑣𝑚 )
When there is no message, then the position of the leading or trailing edge of the
pulse will be equal to the original position and hence 𝑡𝑝 = 0. For positive values of
message, the position will be proportionately shifted right by 𝑡𝑝 = 𝑓(𝑣𝑚 ). For negative
values of message, the position will be proportionately shifted left by −𝑡𝑝 = −𝑓(𝑣𝑚 )
factor. One way of generating PPM is to generate PWM and post process the same to
get PPM.

Figure 5.5 shows the generation of PPM signal. As illustrated in the figure, if PWM is
generated by varying the width of the trailing edge, then this edge will be extracted
to get the position of the pulse in each period. Once the position is extracted, the
leading or trailing edge of the pulse is placed at this instant. The amplitude and width
of the pulse remain constant as in the original pulse train. Thus, PPM is equally robust
to noise like PWM. The mathematical treatment about the frequency domain aspect

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
60

of PWM is an involved process. However, the resulting PPM will also have the
spectrum in the baseband region itself. Alternatively, if PWM is generated by varying
the leading edge, then this edge needs to be extracted to generate PPM and any edge
can be used in case of modification of both edges. Even though, the PPM signal also
contains the message information in the pulse train; it is seldom used due to its
indirect way of storing message. information as in PWM and also the randomness
involved in the position modification. Thus, PPM is of theoretical interest only and
has limited use in signal processing and communication field.

2.2.1.4. Demodulation of Pulse Analog Modulated Signals

PAM, PWM and PPM stores the message is the baseband itself. They essentially
represent the message information at discrete instants of time. Further the message
signal is coded in one of the pulse parameters. We can recover the message that is,
reconstruct the approximate version of the continuous time signal from them when
needed. This is illustrated in Fig. 5.6. The process is straightforward in case of PAM.
The PAM signal can be passed through a low pass niter which retains essentially the
low frequency message signal and smoothing out the pulse train information.
Alternatively, demodulation of message from PWM and PPM appears to be difficult,
since visually the message information is not available as amplin1de variations.
However, the same is available in the other forms as width and position variations.
One simple way of thinking the possibility of demodulation process is to first convert
PWM and PPM to PAM and then perform low pass filtering.

2.2.2. Digital Communications

2.2.2.1. Pulse Digital Modulation Techniques

The most important pulse digital modulation techniques include PCM, DM and
DPCM. This section describes each of them and also recovering approximate analog
message signal from them.

2.2.2.2. Pulse Code Modulation

The fundamental and most important pulse digital modulation technique is the pulse
code modulation (PCM). This technique is the breakthrough for moving from analog

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
61

to digital communication. PCM technique is essentially the result of the thought

process to represent message signal in digital form rather than the original analog
form. The motivation is the merit of digital signal over analog signal for
communication, namely, noise robustness.

PCM may be treated as an extension of PAM. In PAM the time parameter is

discretized, but the amplitude still remains continuous. That is, within the allowable
amplitude limits; the signal value can take on infinite values. However, all these
infinite values may not be distinct from the perception (auditory or visual)
point of view. For instance, in case of speech signal, all amplitude values may not
be important from the auditory perception point of view. Therefore, we may not lose
information by discretizing the amplitudes to some finite values. What is essentially
done is to round or approximate a group of nearby amplitude values and
represent them by a single discrete amplitude value. This process is termed as
quantization. The signal with discretized amplitude values is termed as quantized
signal. There will be error between the original analog signal and its quantized version
which is measured and represented in terms of quantization noise. What is preferable
is minimum quantization noise and hence more closely quantizing signal amplitudes.
This leads to a greater number of discrete levels. Hence it is a tradeoff.

The quantization can be carried out either by dividing the whole amplitude range
into uniform or nonuniform intervals. Accordingly, we have uniform and nonuniform
quantization. PCM is also named after the same as uniform or nonuniform PCM. The
nonuniform quantization and hence PCM are based on the observation of the
nonuniform distribution of signal values within the allowable limits. For instance,
in case of speech, most of the signal values are around the zero level and few will be
in the maximum range. Hence benefit can be achieved in terms of quantization noise
by using nonuniform quantization. However, nonuniform quantization is relatively
difficult to implement compared to uniform quantization.

Each of the discrete amplitude levels can be uniquely represented by a binary word.
To facilitate this, the total number of discrete levels are decided to be in powers of 2.
For instance, if the binary word is of 8-bit length, then we will have 256 discrete levels
possible. Thus, each analog value is sampled by PAM process, quantized and
represented by a binary word. Hence the name pulse code modulation where the
pulse modulation involves coding the sampled analog values. The PCM technique is
illustrated in Fig. 5.7. The sampler block essentially performs PAM process and the
only difference is the pulse width ∆→ 0. The input of sampler block will have signal
which is continuous both in time and amplitude. The output of sampler block will
have the signal which is discrete in time and continuous in amplitude. The output of
quantizer will have signal which is discrete both in time and amplitude. The output
of the encoder will have unique binary code for each discrete amplitude value. The
whole process of sampling, quantizing and encoding is also termed as analog to
digital conversion (ADC) operation. Thus, for any analog signal, the output of ADC is
nothing but PCM signal.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
62

2.2.2.3. Delta Modulation

Delta modulation (DM) is obtained by simplifying the quantization and encoding

process of PCM. To enable this, the signal is sampled at much higher than the required
Nyquist rate. This oversampling process will result in the sequence of samples which
are very close and hence high correlation among successive samples. Under this
condition, it may be safe to assume that any two successive samples are different by
an amplitude of 𝛿. That is, the current sample is either larger or smaller than the
previous value by 𝛿. If it is larger, then it is quantized as +𝛿 and as −𝛿 in smaller case.
Since it is decided a priori, only its sign is important. The sign information can be coded
using one-bit binary word, say; l represents + and 0 represent − . The quantization
and encoding blocks therefore become very simple. Thus, if we have the first signal
value and 1 bit quantization information we can reconstruct the complete quantized
signal.

The block diagram of delta modulator is given in Fig. 5.8 drawn by referring to the
block diagram of PCM given in Fig. 5.7. The sampler block remains 'same as in the
PCM, except that the sampling frequency is much higher than in PCM case (say 4
times or more). According to the principles of DM, the quantizer needs to discretize
the amplitude value by referring to the previous value and say whether it is larger or
smaller. Hence an accumulator is needed to store previous sample, a summer as a
comparing device and producing output into two discrete levels as +𝛿 and −𝛿 . The
encoder is trivial which directly maps the signs of 𝛿 into I or 0. The sequence of 1's
and 0's at the output of encoder constitutes the DM wave.

2.2.2.4. Differential Pulse Code Modulation

Differential pulse code modulation (DPCM) first estimates the predictable-part from
the signal and then codes the unpredictable or error signal in terms of unique binary
words as in PCM and hence the name. The motivation for the same is that most

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
63

message signals have high correlation. Therefore, it is possible to classify the

information present in them into predictable and the unpredictable parts. The main
merit in this approach is the significantly less variance among the samples in the
unpredictable version of the signal compared to the original. Roughly the variance
among the samples will be about half of that of the original signal. As a result, binary
words of smaller length are sufficient for coding unpredictable part Hence the saving
in the bandwidth requirement, measured as bit rate defined as number of kilobits per
second (kbps). For instance, if64 kbps is required for PCM, then DPCM requires about
48 kbps.

The block diagram of DPCM modulator is given in Fig. 5.9 drawn again by referring
to the PCM block diagram in Fig. 5. 7. The input analog signal is passed through the
predictor block whose function is to segregate the information into predictable and
unpredictable parts. The unpredictable part is passed through sampler, quantizer and
encoder blocks to get PCM corresponding to it. The predictable part is directly passed
through the encoder to get the codes. Both these are combined to get the DPCM wave
representing sequence of binary words corresponding to both the parts.

2.2.2.5. Demodulation of Pulse Digital Modulated Signals

The demodulation of PCM is straightforward. Figure 5.10 shows the block diagram
for the reconstruction of analog signal in case of PCM. For obtaining PCM from
analog signal, ADC was employed. Therefore, for obtaining analog signal from PCM,
the reverse of ADC namely, digital to analog conversion (DAC) is required. Thus, the
binary words are applied one at a time to a DAC circuit to obtain equivalent analog
value. How close the reconstructed analog value to the original depends on the
amount of approximation errors introduced due to ADC and DAC conversion. By
the proper choice of binary word length, it has been found that the errors are indeed
negligible from the perception point of view.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
64

The block diagram of demodulation in case of DM is given in Fig. 5. 11. The DM needs
to transmit the first sample and then the OM wave. By combining both, the analog
signal can be reconstructed from the DM wave in the following way: The second
sample is constructed from the first sample by adding to ±𝛿. The second sample is
then stored in the accumulator for future reference. The third sample is constructed
from the second sample using ±𝛿. The process continues till the last sample is
reconstructed.

The reconstruction of analog signal in case of DPCM is more involved and is

illustrated in Fig. 5.12. The approximate analog signal of the unpredictable part is
reconstructed by DAC as in PCM. This signal is used as input to a block constructed
using the predictable part and the approximate version of the original analog signal
is obtained at the output of this block.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
65

2.2.3. Digital Modulation Techniques and Bandwidth Considerations

The basic motivation for analog modulation is to develop techniques for shifting the
analog message signal from low to high frequency range so that it can be conveniently
transmitted over high frequency communication channels. This resulted in AM, FM
and PM techniques. The pulse modulation represents the message signal at discrete
instants of time. However, the resulting message will still be in the low-frequency
region. Thus, pulse modulation is essentially used for the digitization of analog
message (like PCM) and represent if possible, in compact manner (like DPCM). The
digitized message is nothing but sequence of 0's and 1's termed more commonly as
digital or binary message. Thus, using a suitable pulse modulation technique, we can
convert analog message into digital from. Alternatively, the message may be directly
generated in digital form like in the case of computer.

The requirement in the digital communication field is to transfer the digital message
from one place to the other. There are broadly two approaches, namely, baseband
transmission and passband transmission. Baseband digital transmission involves
transmission of digital message in the low frequency (baseband) range itself. Passband
transmission involves transmission of digital message in the high frequency
(passband) range. Since, original digital message is in baseband range, it is first
modulated to the high frequency range and then transmitted. The set of modulation
techniques for shifting the digital message from the baseband to passband are termed
as digital modulation techniques. The detailed study of these techniques is the aim of this
chapter.

The digital modulation techniques are based on the conventional analog modulation
techniques. Since the digital message will have only two levels, 0 and l, the modulation
process needs to store this information in the high frequency range. This can be done
using AM, FM and PM techniques. Accordingly, we have amplitude shift keying (ASK),
frequency shift keying (FSK) and phase shift keying (PSK) as basic digital modulation
techniques. ASK deals with shifting the amplitude of the carrier signal between two
distinct values. FSK deals with shifting the frequency of the carrier signal between two
distinct values. Similarly, PSK deals with shifting the phase of the carrier signal
between two distinct values.

Apart from these basic digital modulation techniques, their variants are also available
termed as M-ary digital modulation techniques. These include M-ary ASK, M-ary FSK
and M-ary PSK. The hybrid schemes involving more than one parameter variation like
amplitude-phase shift keying (APK) are also present under M-ary digital modulation
techniques. The main merit of M-ary techniques is the increased transmission rate for
the given channel bandwidth. From the perspective of M-ary, the basic digital
modulation techniques are also termed as binary digital modulation techniques.
Accordingly, we have binary ASK (BASK), binary FSK (BFSK) and binary PSK
(BPSK).

Depending on the nature of demodulation scheme, the digital modulation

techniques are classified as coherent and non-coherent detection techniques. In case
of coherent detection, the carrier in the receiver is in synchronism with that of the

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
66

transmitter and no such constraint in non-coherent detection. The digital modulation

techniques may be further grouped as binary or M-ary signaling schemes. In binary
signaling scheme, the parameters of the carrier are varied between only two levels
whereas they are varied between M levels in case of M-ary signaling.

Thus, there are a number of digital modulation techniques for passband digital
message transmission. The choice of a particular technique is based on the two
important resources of communication namely, transmitted power and channel
bandwidth. The ideal requirement is the one which uses minimum transmitted power
and channel bandwidth. But this will be conflicting requirements, i.e., to conserve
bandwidth we need to spend more power and hence trade off needs to be achieved.

2.2.3.1. Amplitude Shift Keying (ASK)

ASK is a digital modulation technique defined as the process of shifting the amplitude
of the carrier signal between two levels, depending on whether 1 or 0 is to be
transmitted.

Let the message be binary sequence of 1's and 0's. It can be represented as a function
of time as follows:
𝑣𝑚 = 𝑉𝑚 when symbol is 1
=0 when symbol is 0

Let the carrier be defined as

𝑣𝑐 = 𝑉𝑐 𝑐𝑜𝑠𝜔𝑐 𝑡

The corresponding ASK signal is given by the product of 𝑣𝑚 and 𝑣𝑐 as

𝑣𝐴𝑆𝐾 = 𝑉𝑚 𝑉𝑐 𝑐𝑜𝑠𝜔𝑐 𝑡 when symbol is 1

=0 when symbol is 0

Figure 6.1 shows the time domain representation of the generation of ASK signal. The
digital message i.e., binary sequence can be represented as a message signal as shown
in Fig. 6.1 a. The carrier signal of frequency 𝑓𝑐 = 𝜔𝑐 /2𝜋 is generated continuously from
an oscillator circuit as shown in Fig. 6. lb. When the oscillator output is multiplied by
the message signal, it results in a signal as shown in Fig. 6.1 c termed as ASK signal.

When the binary symbol is 1, the ASK signal will have information equal to the carrier
multiplied by message amplitude and when the binary symbol is 0, it will be zero.
Thus the output shifts between two amplitude levels, namely, 𝑉𝑚 𝑉𝑐 and 0. Hence the
name amplitude shift keying. Based on this discussion a block diagram for the
generation of ASK signal can be written as given in Fig. 6.2. ASK modulator is
essentially an analog multiplier that takes baseband message 𝑉𝑚 and passband carrier
𝑉𝑐 and multiplies the two resulting in the product signal termed as ASK.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
67

The next question is whether such a process results in the shift of spectrum of
baseband message to the passband? The answer is from the amplitude modulation
process discussed in the earlier chapter. This can be illustrated pictorially as follows:
Without worrying about the mathematical intricacies, let the spectrum of 𝑉𝑚 be as
shown in Fig. 6.3a. It will be essentially a sine function in the frequency domain and
has information concentrated mainly in the low frequency range. The sinusoidal
carrier 𝑉𝑐 will have impulses 𝑓𝑐 and −𝑓𝑐 shown in Fig. 6.3b. The product of the two in
the time domain results convolution in the frequency domain giving rise to the
spectrum of ASK signal as shown in Fig. 6.3c. Thus, the ASK signal will have the
message shifted to the passband range.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
68

Demodulation of ASK Signal The demodulation is also tem1ed as detection. There are
two ways in which the message can be demodulated, namely, coherent and non-
coherent detection. Due to the requirement of carrier in the receiver which is in
synchronism with that of the transmitter, the coherent detection circuit is more
complex compared to non-coherent detector. However, the coherent detector
provides better performance under noisy condition.

ln coherent detection, a copy of carrier used for modulation is assumed to be available

at the receiver. The incoming ASK signal is multiplied with the carrier signal. The
output of the multiplier will be a low frequency component representing amplitude
scaled version of baseband message and ASK signal at twice the carrier frequency.
The baseband message is retrieved by passing this signal through a low pass filter.
Figure 6.4 shows the block diagram of a coherent ASK detector.

Let the synchronous carrier at the receiver be given by

𝑣′𝑐 = 𝑉′𝑐 𝑐𝑜𝑠𝜔𝑐 𝑡
The output of the multiplier is given by
𝑉 𝑉 𝑉′𝑐
𝑆1 = 𝑣𝐴𝑆𝐾 𝑉′𝑐 = 𝑚 2𝑐 (1 + cos 2𝜔𝑐 𝑡) when symbol is 1
=0 when is symbol is 0

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
69

The output of the low pass filter is given by

𝑆2 = 𝑉𝑚 (𝑉𝑐 𝑉′𝑐 ) when is symbol is 1

=0 when symbol is 0

Thus, the filter output is

𝑆2 𝛼 𝑣𝑚
Hence, the recovery of baseband message is carried out.

In non-coherent detection, there is no reference carrier made available at the receiver.

Hence, we have to follow other approach. In case of ASK. simple envelope detector
will suffice. The incoming ASK signal is passed through an envelope detector which
tracks the envelope of the ASK signal which is nothing but the baseband message.

Figure 6.5 shows the block diagram of non-coherent ASK detector. The output of the
diode will be a unipolar signal containing the envelope information. The high
frequency variations are further removed by passing it through a low pass filter.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
70

The output of the low pass filter may be further refined by passing it through a
comparator which compares the output of the envelope detector to a preset
threshold and sets all values greater than or equal to the threshold to high level and
rest to the low level. The waveforms at various stages of the non-coherent ASK
detector are shown in Fig. 6.6.

2.2.3.2. Frequency Shift Keying (FSK)

FSK is a digital modulation technique defined as the process of shifting the frequency
of the carrier signal between two levels, depending on whether 1 or 0 is to be
transmitted.

Let the two carriers be defined as

𝑣𝑐1 = 𝑉𝑐 𝑐𝑜𝑠𝜔𝑐1 𝑡
𝑣𝑐2 = 𝑉𝑐 𝑐𝑜𝑠𝜔𝑐2 𝑡
The corresponding FSK signal is defined as
𝑣𝐴𝑆𝐾 = 𝑉𝑚 𝑉𝑐 𝑐𝑜𝑠𝜔𝑐1 𝑡 when symbol is 1
= 𝑉𝑚 𝑉𝑐 𝑐𝑜𝑠𝜔𝑐2 𝑡 when symbol is 0

Figure 6.7 shows the time domain representation of the generation of FSK signal. The
digital message, i.e., binary sequence can be represented as a message signal as shown

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
71

in Fig. 6.7a. Two carrier signals of frequencies 𝜔𝑐1 and 𝜔𝑐2 as shown in Figs. 6. 7 b
and c. When binary symbol is 1, the FSK signal will have the carrier signal with
frequency 𝜔𝑐1. Alternatively, the FSK signal will have the carrier signal with
frequency 𝜔𝑐2 when the binary symbol is 0. This can be achieved by using a suitable
combinational logic circuit which selects one of the two carrier signals based on the
input signal value applied at its control input. For instance, a 2 × 1 multiplexer can
be used for this purpose. Thus, the output of the multiplexer shifts between the two
distinct frequency values, namely, 𝜔𝑐1 and 𝜔𝑐2. Hence, the name frequency shift
keying. Based on this discussion a block diagram for the generation of FSK signal can
be written as given in Fig. 6.8. FSK modulator is essentially a 2 × 1 multiplexer that
takes baseband message 𝑣𝑚 at the control input and two carriers 𝑣𝑐1 and 𝑣𝑐2 at its
input, and produces the FSK signal at its output.

The next question is whether such a process results in the shift of spectrum of
baseband message to the passband? The answer is yes. To appreciate this, we can treat
the FSK modulation process conceptually as two ASK processes, one using carrier
signal with frequency 𝜔𝑐1, and other using 𝜔𝑐2. This is shown in Fig. 6.9. Thus the first
ASK modulator shifts the baseband message to passband centered around 𝜔𝑐1, and
the second ASK modulator shifts the baseband message to passband centered around

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
72

𝜔𝑐2. This can be illustrated pictorially as follows: Let the spectrum of 𝑣𝑚 be as

shown in Fig. 6.10a. The output of the first ASK modulator is shown in Fig. 6.10b and
that of second in Fig. 6.10c. The spectrum of FSK modulator may be viewed as given
in Fig. 6.10d. Thus the FSK signal will have the message shifted to the passband
range.

Demodulation of FSK Signal. In this case also, the message can be demodulated either
by coherent or non-coherent detection. Both demodulation processes can be
understood easily by considering the ASK view of FSK as illustrated in Fig. 6.9.

The block diagram for the coherent detection of FSK is drawn as given in Fig. 6.11.
The incoming FSK signal is multiplied by the carrier signal with frequency 𝜔𝑐1 in the
upper channel and carrier signal with frequency 𝜔𝑐2 in the lower channel. The output
of the multiplier in the upper channel will be low frequency message and ASK signal
at twice 𝜔𝑐1 during the intervals when the FSK is due to the carrier of frequency cod
and will be ASK signals at (𝜔𝑐1 ± 𝜔𝑐2 ) during intervals when the FSK is due to the
carrier of frequency 𝜔𝑐2. Thus the output of the low pass filter in the upper channel
will contain baseband message during intervals belonging to the carrier frequency
𝜔𝑐1 and zero during the intervals belonging to 𝜔𝑐2· Exactly opposite happens in the
lower channel. The outputs of the two channels are further passed onto a comparator.
The output of the comparator will be high where upper channel output is greater than
the lower channel and low when lower channel output is greater than the upper
channel. In this way the baseband message is retrieved from the FSK signal

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
73

Let the synchronous carriers at the receiver be given by

𝑣′𝑐1 = 𝑉′𝑐 cos 𝜔𝑐1 𝑡
𝑣′𝑐2 = 𝑉′𝑐 cos 𝜔𝑐2 𝑡

The output of the multiplier in the upper channel during the interval having frequency
𝜔𝑐1 is given by
𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑆1𝑢 = 𝑣𝐹𝑆𝐾 𝑣′𝑐1 = (1 + cos 2𝜔𝑐1 𝑡)
2
The output of the multiplier in the upper channel during the interval having
frequency 𝜔𝑐2 is given by

𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑆1𝑢 = (cos(𝜔𝑐1 − 𝜔𝑐2 )𝑡 + cos(𝜔𝑐1 + 𝜔𝑐2 )𝑡)
2
The output of the low pass filter in the upper channel during the interval having
frequency 𝜔𝑐1 is given by
𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑆2𝑢 =
2

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
74

The output of the low pass filter in the upper channel during the interval having
frequency 𝜔𝑐2 is given by
𝑆2𝑢 = 0
Thus, the filter output in the upper channel is
𝑆2𝑢 𝛼 𝑣𝑚
during the interval having frequency 𝜔𝑐1 and

𝑆2𝑢 𝛼 0
during the interval having frequency 𝜔𝑐2.

The output of the multiplier in the lower channel during the interval having frequency
𝜔𝑐1 is given by

𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑆1𝑙 = (cos(𝜔𝑐1 − 𝜔𝑐2 )𝑡 + cos(𝜔𝑐1 + 𝜔𝑐2 )𝑡)
2
The output of the multiplier in the lower channel during the interval having frequency
𝜔𝑐2 is given by
𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑆1𝑙 = 𝑣𝐹𝑆𝐾 𝑣′𝑐1 = (1 + cos 2𝜔𝑐1 𝑡)
2
The output of the low pass filter in the lower channel during the interval having
frequency 𝜔𝑐1 is given by
𝑆2𝑙 = 0
The output of the low pass tilter in the lower channel during the interval having
frequency 𝜔𝑐2 is given by

𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑆2𝑙 =
2
Thus, the filler output in the lower channel is
𝑆2𝑙 𝛼 0
during the interval having frequency 𝜔𝑐1 and

𝑆2𝑙 𝛼 𝑣𝑚

during the interval having frequency 𝜔𝑐2

Therefore, the output of the comparator is given by

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
75

𝑆3 𝛼 𝑣𝑚
Hence. the recovery of baseband message is carried out.

In case of non-coherent detection, envelope detectors can be used as shown in the

arrangement given in Fig. 6.12. The incoming FSK signal is passed through a filter
tuned to 𝜔𝑐1 and then an envelope detector in the upper channel. Similarly, the
same FSK signal is passed through a filter tuned to 𝜔𝑐2 and then an envelope detector
in the lower channel. Thus the distinction between the upper and lower channels is
due to the two filters. During the interval represented by the carrier signal with
frequency 𝜔𝑐1, the output the upper channel will be high whereas that of the lower
channel is low. Exactly opposite happens during the interval represented by the
carrier signal with frequency 𝜔𝑐2. The outputs of the upper and lower channels
envelope detectors are applied to a comparator which produces the output
proportional to the message. The waveforms at various stages of the non-coherent
FSK detector are shown in Fig. 6.13.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
76

2.2.3.3. Phase Shift Keying

PSK is a digital modulation technique defined as the process of shifting the phase of
the carrier signal between two levels, depending on whether 1 or 0 is to be
transmitted.

Let the two carriers be defined as

𝑣𝑐1 = 𝑉𝑐 cos 𝜔𝑐 𝑡
𝑣𝑐2 = −𝑉𝑐 cos 𝜔𝑐 𝑡
The corresponding PSK signal is defined as
𝑣𝑃𝑆𝐾 = 𝑉𝑚 𝑉𝑐 cos 𝜔𝑐 𝑡 when symbol is 1
𝑣𝑃𝑆𝐾 = −𝑉𝑚 𝑉𝑐 cos 𝜔𝑐 𝑡 when symbol is 0

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
77

Figure 6.14 shows the time domain representation of the generation of PSK signal. The
digital message, i.e., binary sequence can be represented as a message signal as shown
in Fig. 6.14a. Two carrier signals of opposite phases generated from an oscillator and
an inverter (180° phase shifter) are as shown in Figs. 6.14b and c. When the binary
symbol is 1, the PSK signal will have the original carrier signal. Alternatively, the PSK
signal will have the 180° phase shifted carrier signal when the binary symbol is 0. This
can be achieved by using a suitable combinational logic circuit like 2 X l multiplexer
as described in the case of FSK. Thus, the output of the multiplexer shifts between the
two distinct phase values; namely, 0° and 180°. Hence the name phase shift keying.
Based on this discussion a block diagram for the generation of PSK signal can be
written as given in Fig. 6.15.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
78

We can also treat the PSK modulation process conceptually ns two ASK processes,
one using carrier signal, with 0°phase shift and other using l80° phase shift. This is
shown in Fig. 6. 16. Thus the first ASK modulator shift is the baseband message to
passband centered around 𝜔𝑐 , but with phase shift of 0° and the second ASK
modulator also shifts the baseband message to passband centered around 𝜔𝑐 , but with
phase shift of 180°. This can be illustrated pictorially as follows: Let the spectrum 𝑣𝑚
be as shown in Fig. 6. l 7a. Since the difference he1ween the two carrier signals are in
terms of phase values. the magnitude spectrum of the output of both the ASK
modulators will be same as shown in Fig. 6. 17b. Thus the two ASK signals are
indistinguishable in their magnitude spectra. Their distinction lies only in their phase
spectra which are not shown. The magnitude spectnm1 of PSK modulator will also be
same as in Fig. 6.17b. However, we can appreciate the fact that the PSK signal will
have the message shifted to the passband range.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
79

Demodulation of PSK Signal The demodulation of PSK can also be understood easily
by considering the ASK view of PSK. However, the message can only be demodulated
by coherent detection. This can be appreciated from the non-coherent detection of FSK
signal which was made possible due to the frequency selective operation of the filters
present in the upper and lower channels. In PSK, the two ASK signals are separated
in phase values, not in frequency.

The block diagram for the coherent detection of PSK may be drawn as given in Fig.
6.18. The incoming PSK signal is multiplied with the carrier signal with phase shift
0° in the upper channel and carrier signal with phase shift 180° in the lower channel.
The output of the multiplier in the upper channel will be low frequency message and
ASK signal at twice 𝜔𝑐 , during the intervals when the PSK is due to the carrier with
phase shift 0°. It will be 180° phase shifted versions during intervals when the PSK is
due to the carrier of phase shift I 80°. Thus, the output of the low pass filter in the
upper channel will contain baseband message during intervals belonging to 0° phase
shift and its 180-phase shifted version during the intervals belonging to the phase shift
of 180°. Exactly opposite happens in the lower channel. The outputs of the two
channels arc further passed onto a comparator. The output of the comparator will be
high when upper channel output is greater than the lower channel and low when
lower channel output is greater than the upper channel. In this way the passband
message is retrieved from the PSK signal

Let the synchronous carriers at the receiver be given by

𝑣′𝑐1 = 𝑉′𝑐 𝑐𝑜𝑠 𝜔𝑐 𝑡
𝑣′𝑐2 = −𝑉′𝑐 𝑐𝑜𝑠 𝜔𝑐 𝑡

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
80

The output of the multiplier is the upper channel during the interval having 0° phase
shift is given by
𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑠1𝑢 = 𝑣𝑃𝑆𝐾 𝑣′𝑐1 = (1 + 𝑐𝑜𝑠 2𝜔𝑐 𝑡)
2
The output of the multiplier in the upper channel during the interval having 180°
phase shift is given by
𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑠1𝑢 = − (1 + 𝑐𝑜𝑠 2𝜔𝑐 𝑡)
2
The output of the low pass filter in the upper channel during the interval having 0°
phase shift is given by
𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑠2𝑢 =
2
The output of the low pass filter in the upper channel during the interval having
180° phase shift is given by
𝑉𝑚 𝑉𝑐 𝑉 ′ 𝑐
𝑠2𝑢 = −
2
Thus, the filter output in the upper channel is
𝑠2𝑢 𝛼 𝑣𝑚
during the interval having 0° phase shift and
𝑠2𝑢 𝛼 − 𝑣𝑚
during the interval having 180° phase shift.

The exact opposite phenomenon happens in the lower channel. As a result, the filter
output in the lower channel is
𝑠2𝑙 𝛼 𝑣𝑚
during the interval having 0° phase shift and
𝑠2𝑙 𝛼 − 𝑣𝑚
during the interval having 180° phase shift.

Therefore, the output of the comparator is given by

𝑠3 𝛼 𝑣𝑚
Hence the recovery of baseband message is carried out.

2.2.3.4. M-ary Digital Modulation Techniques

In the previous section, we described the basic digital modulation techniques which
involve transmitting information in two levels. Hence, they may also be termed as
binary digital modulation techniques. Accordingly, we can rename them as binary ASK
(BASK), binary FSK (BFSK) and binary PSK (BPSK). We can extend the same

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
81

principles to transmit information in more than two levels, in general, M levels. These
modulation techniques are termed as M-ary digital modulation techniques. As will be
apparent from later description, the main merit of M-ary techniques is increased
transmission rate on the same channel bandwidth. The signals with M different levels
may be generated by changing the amplitude, frequency or phase of a carrier in M
discrete steps as opposed to two levels in binary modulation scheme. Accordingly, we
have M-ary ASK, M-ary FSK and M-ary PSK digital modulation techniques. Another
way of generating M-ary signals is to combine different methods of binary digital
modulation schemes. For instance, M-ary amplitude-phase shift keying (APK) is
obtained by combing ASK and PSK. A special fom1 of this hybrid modulation that
exploits the merits of quadrature amplitude modulation (QAM) and M-ary scheme is
M-ary QAM technique. Among all the M-ary digital modulation techniques the
mostly used ones include M-ary PSK, M-a.ry FSK and M-ary QAM which are
described in the rest of the section.

2.2.3.4.1. M-ary PSK

In BPSK, the phase of the carrier can take on only two values and most convenient
being 0° and 180°. As opposed to this, M-ary PSK can take on M different phase shift
values within 2𝜋 range, given by∅𝑖 = 2𝜋𝑖 /𝑀, where, i = 0, 1, ..., M·- 1. Accordingly,
we have M carrier signals for modulation. For instance, when 𝑀 = 4, we have ∅𝑖 =
0, 𝜋/2, 𝜋, 3𝜋/2. Such a scheme is termed as quaternary PSK, since the phase values are
separated by 𝜋/2. Alternatively, in BPSK, if the phase shifts are separated by 𝜋/2,
then it is termed as quadrature PSK (QPSK).

The M different carrier signals can be defined as

2𝜋𝑖
𝑣𝑐𝑖 = 𝑉𝑐 𝑐𝑜𝑠 (𝜔𝑐 𝑡 + 𝑀 ) 𝑖 = 0,1, … . , 𝑀 − 1

For ease of illustration we discuss by considering quaternary PSK. In a symbol interval

we can transmit 2 different messages, namely, 𝑣𝑚1 and 𝑣𝑚2 using the carriers
𝑣𝑐1 , 𝑣𝑐2 , 𝑣𝑐3 and 𝑣𝑐4 separated by 𝜋/4. This is because M levels can be used to transmit
binary words of length n, where 𝑀 = 2𝑛 . For 𝑀 = 4 we have binary words of 2-bit
length and hence two independent binary sequences can be transmitted. For instance,
00 can be transmitted using a carrier with phase shift ∅ = 0°, 01 with ∅ = 90°, 10 with
∅ = 180° and 11 with ∅ = 270°. Figure 6.19 shows the two different messages, four
different carriers and corresponding quaternary PSK signal. Based on this a block
diagram can be drawn for the generation for a quaternary PSK using 4 x I multiplexer
as shown in Fig. 6.20. The two input message sequences are applied to the control
inputs. When 00 is to be transmitted 𝑣𝑐1 is selected, 𝑣𝑐2 for 01, 𝑣𝑐3 for 10 and 𝑣𝑐4 for 11.
Hence the generation of quaternary PSK.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
82

Demodulation of M-ary PSK Signal. For the demodulation. only coherent detection is
possible. In coherent detection. incoming quaternary PSK signal is multiplied with
four carrier signals 𝑣′𝑐1 , 𝑣′𝑐2 , 𝑣′𝑐3 and 𝑣′𝑐4 which are in synchronism with those at the
transmitter. In given symbol interval. the multiplier whose carrier phase matches

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
83

with that of the PSK signal will produce maximum output compared to other
multipliers. Accordingly, the corresponding binary word of two bits is decoded. For
instance, if the multiplier with 𝑣′𝑐1 produces maximum output, then 00 is decoded.
The two bit sequences can be separated to get the two message 𝑣𝑚1 and 𝑣𝑚2 . Figure
6.21 shows the block diagram for the demodulation of qutarne1y PSK. The
purpose of maximum finder is to find the channel that provides maximum output.
Accordingly, the binary word decoder will produce the corresponding binary word.

2.2.3.4.2. M-ary FSK

M-ary FSK is same as M-ary PSK, except that the carriers are separated in frequency
than phase. In BFSK, the frequency of the carrier can take only two values say, 𝜔𝑐1
and 𝜔𝑐2. As opposed to this, M-ary FSK can take on M different frequency values.
given by 𝜔 . where. 𝑖 = 0. 1, . . . . . , 𝑀 − 𝐼. Accordingly, we have M carrier signals for
modulation. For instance, when M = 4, we have 𝜔𝑐𝑖 = 𝜔𝑐1 , 𝜔𝑐2 , 𝜔𝑐3 and 𝜔𝑐4. Such a
scheme is termed as quaternary FSK.

The M different carrier signals can be defined as

𝑣𝑐𝑖 = 𝑉𝑐 𝑐𝑜𝑠 (𝜔𝑐𝑖 𝑡) 𝑖 = 0,1, … . , 𝑀 − 1

For ease of illustration we discuss by considering quaternary FSK. In a symbol interval

we can transmit 2 different messages, namely. 𝑣𝑚1 and 𝑣𝑚2 , using the carriers
𝑣𝑐1 , 𝑣𝑐2 , 𝑣𝑐3 and 𝑣𝑐4 For instance, 00 can be transmitted using a carrier with frequency
𝜔𝑐1, 01 with 𝜔𝑐2, 10 with 𝜔𝑐3 and 11 with 𝜔𝑐4. Figure 6.22 shows the two different
messages, four different carriers and corresponding quaternary FSK signal.
Therefore, the block diagram for the generation of quaternary FSK will remain same
as that of quadraphase PSK shown in Fig. 6.20. The only difference is that the different

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
84

carriers are separated in frequency than phase. The two input message sequences are
applied to the control inputs. When 00 is to be transmitted 𝑣𝑐1 is selected, 01 is to be
transmitted 𝑣𝑐2 is selected, 𝑣𝑐3 for 10 and 𝑣𝑐4 for 11. Hence the generation of
quaternary FSK.

Demodulation M-ary FSK Signal FSK can be demodulated by either coherent or non-
coherent detection. ln coherent detection incoming quaternary FSK signal is applied
to four analog multipliers having carrier signals 𝑣𝑐1 , 𝑣𝑐2 , 𝑣𝑐3 and 𝑣𝑐4 which are
separated in frequency. ln a given symbol interval. the analog multiplier whose carrier
frequency matches with that of the FSK signal will produce maximum output.
Accordingly, the corresponding binary word of two bits is decoded. For instance, if
the analog multiplier with 𝑣′𝑐1 produces maximum output, then 00 is decoded. The
two-bit sequences can be separated to get the two messages 𝑣𝑚1 and 𝑣𝑚2 .The block
diagram for the coherent detection of quaternary FSK is same as that of quaternary
PSK shown in Fig. 6.21, except that the carrier signals are now separated in frequency.

The block diagram of non-coherent detection of quaternary FSK is shown. in Fig.

6.23. In non-coherent detection, incoming quaternary FSK signal is applied to four

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
85

correlators or matched filters which are by design matched to the four carrier signals
𝑣′𝑐1 , 𝑣′𝑐2 , 𝑣′𝑐3 and 𝑣′𝑐4 . Thus, it avoids the requirement of reference carriers in the
receiver which is their main merit. The output of matched filter gives information
about the similarity of input wave with the matched filter design value. In a given
symbol interval, the matched filter which matches best with that of the FSK signal will
produce maximum output compared to other filters. The output of the matched filters
are passed through the envelope detectors. The output of the envelope detectors are
compared and the one with maximum output is taken as the channel and its
corresponding binary word is decoded. For instance, if the matched filter designed for
𝑣′𝑐1 produces maximum output, then 00 is decoded.

2.2.3.4.3. M-ary QAM

Quadrature amplin1de modulation (QAM) is a variant of AM to conserve bandwidth.

The two message signals 𝑣𝑚1 and 𝑣𝑚2 can be transmitted on the same bandwidth
using two carriers having same frequency, but separated by a phase shift of 𝜋/2. That
is, the two carrier signals are in phase quadrature and each of these carriers are
amplitude modulated and hence the name quadrature amplitude modulation (QAM).

Let the two carrier signals be given by

𝑣𝑐1 = 𝑉𝑐 𝑐𝑜𝑠 𝜔𝑐 𝑡
and
𝑣𝑐2 = 𝑉𝑐 𝑠𝑖𝑛 𝜔𝑐 𝑡
The corresponding QAM signal is defined as
𝑣𝑄𝐴𝑀 = 𝑉𝑚1 𝑉𝑐 𝑐𝑜𝑠 𝜔𝑐 𝑡 + 𝑉𝑚2 𝑉𝑐 𝑠𝑖𝑛 𝜔𝑐 𝑡

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
86

ln the above equation, the first term is termed as in-phase component and the second
term is termed as quadrature component.

The message signals can be recovered at the receiver by coherent detection. The
incoming QAM is simultaneously applied to in-phase and quadrature channels. The
output of the analog multiplier in the in-phase channel is given by

𝑣𝑚1 𝑉𝑐 𝑉 ′ 𝑐 𝑣𝑚1 𝑉𝑐 𝑉 ′ 𝑐
𝑠1 = 𝑣𝑄𝐴𝑀 𝑉′𝑐 𝑐𝑜𝑠 𝜔𝑐 𝑡 = + 𝑐𝑜𝑠 2𝜔𝑐 𝑡 + 𝑣𝑚2 𝑉𝑐 𝑉 ′ 𝑐 𝑠𝑖𝑛 𝜔𝑐 𝑡 𝑐𝑜𝑠 𝜔𝑐 𝑡
2 2

The first term is the scaled version of the message 𝑣𝑚1 which can be retrieved by
passing through a low pass filter.
The output of the analog multiplier in the quadrature channel is given by

𝑣𝑚2 𝑉𝑐 𝑉 ′ 𝑐 𝑣𝑚2 𝑉𝑐 𝑉 ′ 𝑐
𝑠2 = 𝑣𝑄𝐴𝑀 𝑉′𝑐 𝑠𝑖𝑛 𝜔𝑐 𝑡 = + 𝑠𝑖𝑛 2𝜔𝑐 𝑡 + 𝑣𝑚1 𝑉𝑐 𝑉 ′ 𝑐 𝑠𝑖𝑛 𝜔𝑐 𝑡 𝑐𝑜𝑠 𝜔𝑐 𝑡
2 2

The first term is the scaled version of the message 𝑣𝑚2 which can be retrieved by
passing through a low pass filter.

In this way we can transmit two independent message signals on the same bandwidth
with the help of two carriers which are in phase quadrature. The conventional QAM
is used for analog communication, but it applies equally to digital message signal also.

The transmission rate of the M-ary PSK can be further increased by combining the
QAM concept with it resulting in the hybrid M-ary amplitude-phase shift keying
(APK) termed as M-ary QAM. In case of M-ary PSK, the M carrier signals separated
in phase are used to transmit binary words of length n bits, where 𝑀 = 2𝑛 . This
transmission rate can be further increased by replacing these carriers with in-phase
and quadrature components and amplitude modulating each component by a suitable
in-phase and quadrature value.

The generation of the in-phase and quadrature values can be illustrated with the help
of Fig. 6.24, termed more commonly as signal constellation diagram. ∅1 represents the
in-phase component and ∅2 represents the quadrature component. Any point in the
constellation diagram can be identified by a unique binary word obtained by
dividing the whole region into smaller square blocks as shown and giving unique
binary code for each square. For instance, the point in the second square block
around the origin of the first quadrant is uniquely identified by 10 for ∅1 and 10 for
∅2 and accordingly it represents the binary word 1010. This can be uniquely
transmitted by using the in-phase and quadrature components a1 and b1,
respectively, whose values are as indicated in the figure.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
87

Accordingly, the transmitted signal can be written in generic form as

𝑣𝑖 (𝑡) = 𝑎1 cos 𝜔𝑐 𝑡 + 𝑏1 𝑠𝑖𝑛𝜔𝑐 𝑡 where 𝑖 = 1,2, … , 16

As defined in the equation, 𝑣𝑖 (𝑡) can take M distinct shapes. Each pulse can be used
to transmit distinct binary word and accordingly for 𝑀 = 16, we have 16 words, each
of length 4 bits. Thus, in each symbol interval. The bit rate has doubled compared to
M-ary PSK.

Demodulation of M-ary QAM Signal The message can be recovered by coherent

demodulation based on QAM demodulator as shown in Fig. 6.25. The incoming M-
ary QAM is applied to the in-phase and quadrature phase channels. The output of in-
phase channel will be proportional to the in-phase value 𝑎1 , which can be identified
by comparing the same with multilevel threshold. In case of M-ary QAM, there will
be 𝐿 = √𝑀 thresholds possible, one for each value of 𝑎1 . Based on this comparison, it
is possible to identify the most likely a, value and corresponding binary subword.
Same thing is true with respect to quadrature phase channel also. By combining the
two outputs, the binary word can be recovered.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
88

The digital modulation techniques are meant for translating the digital message from
baseband to passband. As described in this chapter it is indeed possible to do the same
with help of techniques that arc based on analog modulation techniques. Binary ASK
stores digital message infom1ation in two amplitude levels. Binary FSK stores the
same in two frequency levels and binary PSK in two phase levels. The transmission
rate possible is one bit per symbol interval. Alternatively, in M-ary digital modulation
techniques the transmission rate can be increased significantly. In case of M-ary
schemes, the transmission rate will be n bits per symbol interval where M = 2n. Except
for PSK and QAM, all other digital modulation schemes can employ both coherent
and non-coherent approaches for detecting the message. PSK and QAM schemes can
use only coherent detection scheme.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
89

Assessment of Learning

Quiz.

A. Define the following terms:

1. Pulse Code Modulation 7. Delta Modulation

2. PCM coding efficiency 8. Digital Pulses
3. Dynamic Range 9. M-ary Encoding
4. Linear PCM code 10. Quadrature Amplitude
5. Nonlinear PCM code Modulation
6. PCM line speed

B. Identify the unique characteristics of ASK, FSK, PSK, QAM

C. Explain the bandwidth considerations for ASK, FSK, PSK, QAM
D. Differentiate bit rate and baud rate. Explain when is bit rate and baud rate applicable.

Assignment.

Using Matlab, show the graph of ASK, PSK, and FSK using the following digital input values:

1. 0111010001
2. 0110110011
3. 1011000110

Sample output:

Printscreen your answers and save in a PDF format. Indicate the date and time of your
answers in your respective computers.

C. M. D. Hamo-ay
 Modulation and Coding
2 | Comm 2 Techniques
90

2.3. References

• Tomasi, W. (2014). Advanced Electronic Communications Systems, 6th Edition.

Upper Sadle River. New Jersey. Pearson Education Limited.
• Kennedy, G., Davis, B., Prasanna, SRM (2015). Kennedy’s Electronic
Communications Systems, 5th Edition. P-24, Green Park Extension, New Delhi,
India. McGraw Hill Education Private Limited.

2.4. Acknowledgement

The images, tables, figures and information contained in this module were taken from
the references cited above. Some parts of the instructional materials were taken from
the articles published by the respective references. The writers and their valuable work
are highly acknowledged.

C. M. D. Hamo-ay