6

DIGITAL MODULATION
TECHNIQUES

So far we have considered baseband digital signals and systems. In

baseband communication, the signals are transmitted without any frequency
shift. These signals however, are not suitable for transmission over a radio
link because they have a sizable power at low frequencies. Baseband
communication is only suitable for transmission over short distances using
twin-wire line or coaxial cables. For long distance digital transmission it is
necessary to generate bandpass signals suited to
the transmission medium.
This is achieved by varying the characteristics (amplitude or phase) of a
sinusoidal carrier in accordance with the digital baseband signal. In this
that
chapter we shall discuss various techniques of digital modulation
enables one to transmit binary data over a bandpass communication channel.
in
The basic concepts involved in digital modulation are same as those apply
modulation. The basic difference lies in the nature of the
the case of analog
modulating signal. We begin our discussion with binary modulation
techniques.
A digital signal can be used to modulate the amplitude, frequency or

phase ofa sinusoidal carrier wave producing three different forms of digital
modulation e.g., Amplitude-Shift-Keying (ASK), Frequency shift-Keying
illustrates the waveforms of
(FSK) and Phase-Shift-Keying (PSK). Fig.6.1 of NRZ
ASK, FSK and PSK for a waveform
modulating consisting
rectangular pulses. techniques of digital modulation
In addition to these basic
there are some modulation schemes that employ a combination of amplitude
and phase modulation. The details of various modulation and demodulation
schemes are discussed in the subsequent sections of this chapter.
 256 DIGITAL COMMUNICATION
DIGITAL MODULATION TECHNIQUES 257
system, binary symbol 1 is represented by transmitting a sinusoidal carrier
wave of fixed amplitude
A and fixed frequency f for the bit duration T

JUUVWVUVV seconds whereasbinary symbol 0 is represented by switching off the carrier

for T, seconds. This signal can be generated simply by turning the carrier of
a sinusoidal oscillator on and off for the prescribed periods indicated by the
(a) modulating pulse train. For this reason the scheme is also known as on-off
keying (OOK).
Let the sinusoidal carrier be represented by

e(t)=A, cos{27tf.t (6.1)

Then, the binary ASK signal can be represented by a wave st) given by

st) = A, cos2Tf,t). symbol 1

T, (6.2)
(b) = 0, symbol 0

A typical ASK waveform is illustrated in Fig.6.2 for a binary data

representedby 10110101
MA
UUUUUUUWUU (c)
MAAAA
(a)

UWVVV VWUUU VUUU

(d) -T (b)
Fig.6.1 Illustrating binary modulated waveforms. (a) Unmodulated carrier. (b) ASK.
c) FSK. (d) PSK.

6.1 AMPLITUDE-SHIFT KEYING (ASK)

The binary ASK system was one of the earliest forms of digital modulation
used in wireless telegraphy. This simplest form of digital modulation is no
longer used widely in digital communication. Nevertheless it serves as a (c)
useful model which helps in understanding certain concepts. In an ASK Unmodulated carrier.
Fig.6.2 Amplitude-shift keying
waveforms. (a)
ASK waveform.
(6) Unipolar bit sequence. (c)
 258 DIGITAL COMMUNICATION DIGITAL MODULATION TECHNIQUES 259
Generation of ASK signal Detection of binary ASK

ASK signal can be generated by applying the The demodulation of binary ASK wave
incoming binary data can be done with the
(represented in unipolar form) and the sinusoidal carrier to the two inputs of coherent detector shown in help of
Fig.6.5. It consists of a product modulator
followed by an integrator and a decision
a product modulator
(balanced modulator). The resulting output is the ASK making device. The incoming ASK
wave. This is illustrated in Fig.6.3. Modulation causes a shift of the
baseband signal is applied to one input of the product modulator. The other input of the
product modulator is supplied with a sinusoidal carrier
the help of a local oscillator. The generated with
output of the product modulator goes to
the integrator. The
integrator operates on the output of the multiplier for
Product successive bit intervals and
essentially performs a low-pass filtering action.
Binary wave in Modulator Binary ASK wave s(t) . The output of the
integrator goes to the input of a decision making device.
unipolar form b(t)

Binary
Carrier wave ASK wave Symbol 1,
if threshold
A cos(2tft) Decision exceeded
Fig. 6.3 Generation of binary ASK. device
Symbol 0,
otherwise
signal spectrum. The ASK signal which is basically the product of the
sequence and the carriers signal has a power spectral binary
that of the baseband on-off density (PSD) same as cos(21tf.t) Threshold
signal but shifted in the frequency domain by
+f. This is illustrated in Fig.6.4. Note that two impulses occur at fe. The
spectrum of the ASK signal reveals that it has an
infinite bandwidth. For
practical
purposes, the bandwidth is often defined as the bandwidth of an Fig.6.5 Coherent detection of binary ASK signals
ideal bandpass filter centered at
f whose output contains say 95% of the
total average power content of the ASK
signal. It can be shown that The decision making device compares the output of the
according to the above criterion the bandwidth of the ASK integrator with a
signal is
approximately 3/T, Hz. The bandwidth of the ASK signal can however, preset threshold. It makes a decision in favour of
symbol 1 if the threshold is
be exceeded and in favour of symbol 0 otherwise. The
coherent detection
reduced by using smoothed versions of the involves the use of linear operation. In the method discussed above it
pulse waveform b(t) instead of has
rectangular pulse waveforms. been assumed that the local carrier is in
perfect synchronisation with the
carriers used in the transmitter. This means that the
frequency and phase of
the localy generated carrier is same as those of the carriers
used in the
transmitter.
PSD
The following two forms of
synchronisation are needed for the
operation of coherent (or synchronous detector)

i) Phase synchronisation which ensures that carrier wave

generated locally in the receiver is locked in phase with
respect
to that employed in the transmitter

(i) Timing synchronisation which enable proper timing of the

T decision making operation in the receiver with
respect to
switching instants (switching between symbols 1 and 0) in the
Fig.6.4 Power spectral density of ASK signal. original binary data.
 DIGITAL MODULATION TECHNIQUES 261
DIGITAL COMMUNICATION
260
be demodulated non-coherently using envelope
Binary ASK signal can also the design consideration needed in
detector. This greatly simplifies
synchronous detection.
Non-coherent detection schemes do not require a
This Scheme involves some form of
AAAAAAAMAMAAL
phase-coherent local oscillator.
rectification and low pass filtering at the
receiver. The block diagram of a
non-coherent receiver for ASK signal is shown in Fig.6.6.
UVVUVUVUVUVUUUUU
(a)

Decision
Bandpass Rectifier Lowpass
Filter Filter
DeviceOutput
Envelope detector

AAAAAAAAAA-
Threshold

Fig.6.6 Non-coherent ASK detector.

(b)
6.2 PHASE-SHIFT KEING (PSK

In a PSK system, a sinusoidal carrier wave of fixed amplitude and fixed Fig.6.7 Phase-shift keying waveforms. (a) Unmodulated carier wave. (b) Phase-shift
keying waveform.
frequency f is used to represent both symbols 1 and 0, except that the
carrier phase of each symbol differs by a phase of 180°. Let the unmodulated
carrier be represented by

It may be noted that unlike ASK signal, the PSK transmission is

e.t)= A,cos(27f.t) (6.3) polar. The PSK signal may be viewed as the product of the binary data signal
bt) and the sinusoidal carrier where b{t) is a polar signal. Therefore, the
Then the binary PSK signal slt) can be written as power-spectral density of PSK is the same as that of the polar baseband
signal shifted to t f . The power spectral density (PSD) ofthe binary PSK is
shown in Fig.6.8. The shape of the psd of the binary PSK signal and ASK
st) =A, cos(21tf,1) symbol 1 signals are similar. The only difference is that the former does not have an
impulse at the carrier frequeney. The bandwidth requirement of the PSK
A, cos27f,t +t symbol 0
(6.4) signal is same as that of the ASK signal. The most significant difference
between the two is that ASK is a linear modulation scheme whereas the PSK
is a non-linear modulation scheme. The main advantage of the PSK signaling
The PSK waveform is illustrated in Fig.6.7. scheme is that it has a superior performance over the ASK in a noisy
environment for a given peak power
 DIGITAL COMMUNICATION 263
262 DIGTAL MODULATION TECHNIQUES
problem. The design of the coherent detection becomes complicated to tackle
the synchronisation problem. Moreover, PSK cannot be detected non-
PSD coherently like ASK. This is because of the fact that the envelope of a Sk
Wave is the same for both
used for the modulation
symbols 1 and0 and a single carrier frequency 1s
process. The need for synchronisation in binary PSK
can be avoided by suitably modifying the PSK
generation scheme.
Diferential phase shift keying (DPSK) is a modification of PSK that avoids
the need to provide synchronous carrier
required detecting PSK signals.
for

6.3 DIFFERENTIAL PHASE-SHIFT KEYING

f
- -, In order to eliminate the need for
phase synchronisation of coherent receiver
with PSK, a differential encoding system is used in this modified scheme
Fig.6.8 Power spectral density of binary PSK signal The digital information content of the
binary data is encoded in terms of
signal transitions. For example, the symbol 0 may be used to represent
in
Generation of binary PSK signal
transition a given binary sequence (with respect to the previous encoded
bit) and symbol 1 to indicate no transition. This new signaling technique that
combines differential encoding with phase-shift keying is called differentia
PSK signal can be generated by using the same scheme as used in the phase-shift keying (DPSK).
generation of ASK. The only difference is that the incoming binary data
should be in the polar form. This is illustrated in Fig.6.9. From this A schematic arrangement for generating DPSK signal is shown in
arrangement, we may conclude that a binaryPSK may also be viewed as a
double-side band suppressed camier (DSB-SC) modulated wave.
Fig.6.10. The data stream b{t) is applied to the input of the encoder. The
output of theencoder is applied to one input of the product modulator. To the
other input of this product modulator a sinusoidal carrier of fixed amplitude
and frequency is applied. The relationship between the binary sequence and

Product
its diferentially encoded version is illustrated in Table 6.1 for a assumed
datasequence 0 0 100100111. In this illustration it has
been assumed
Binary wave in polar Modulator Binary PSK wave s() that the encoding has been done in such a way that trunsition in the given
form bt)
binary sequence with respect to the previous encoded bit is represented by a
symbol 0 and no transition by symbol 1. Note that an extra bit (symbol 1)
has been arbitrarily added as an initial bit. This is necessary to determine the

Carier wave encoded sequence. The phase of the generated DPSK signal is showm in the
third row of the Table 6.1.
A, cos(2tf

Fig.6.9 Generation of binary PSK

(b(k) fd(k)) Product 0oxA, cos{27£,t)
Encoder
Detection of binary PSK Modulator
The coherent detection of binary PSK is similar to that of ASK. The (dud
consists of a product modulator supplied with a locally
detector
generated sinusoidal Delay
carrier, an integrator that operates on the multiplier output for successive bit
T
A, cos(27,0)
interval and a decision device that compares the
integrated output with a
preset threshold. The difference between coherent detection of PSK and ASK
lies in the choice of the threshold. Like the ASK coherent
detector, PSK
coherent detectors also suffers from synchronisation (phase and timing) Fig.6.10 Scheme for generating DPSK signals
 DIGITAL MODULATION TECHNIQUES 265
DIGITAL COMMUNICATION
264
carrier phase angle of the received DPSK signal and íts delayed version,
encoded sequences with phase. measured in the same bit interval. The phase angle of the DPSK signal and
Table 6.1 Differentially
1 its delayed version are shown in 3rd and 5th rows respectively. The
Binary data 0 0 10 0 1 0 0 pnase
ditterence between the two sequences for each bit interval is used to
{bk) 0 1 1 0 1 determine the sign of the phase comparator output. When = 0 , the
Differentially integrator output is positive whereas when o = T, the integrator output 1s
encoded data

d(k)} 0 0 T 00 T 0 negative. By comparing the integrator output with decision level of

a
volt, the decision device can reconstruct the binary sequence by assigning a
zero

Phase of symbol 0 for negative output and a symbol1 for positive output. 1he
DPSK 1 0 1 0 1 binary data is shown in the last
Shifted reconstructed row of the table. It is thus seen
differentially that in the absence of noise thebe viewed
receiver can reconstruct the transmitted
encoded data binary data exactly. DPSK may as a non-coherent version of PSK.

dPhase of 0 T 0 0 T 0 0 T 0 0
Note that the reconstruction is invariant with the choice of the initial bit in
the encoded data. This is illustrated in the example 6.1 given below.
shifted Example 6.1 A binary data stream 0010010011 is to be transmitted
DPSK
Phase using DPSK. Show that the reconstruction of the DPSK signal by the
technique discussed in the previous section is independent of the choice of
comparison the extra bit.
output
Detected 0 1 0 0 1 0 011 Solution. In the previous section we have seen that DPSK signal can be
binary detected accurately (in the absence of channel noise) without having a local
sequence oscillator for generation of synchronous carrier. The initial bit in the
arbitrary starting reference bit differentially encoded data was assumed to be 1. In this example, we use the
initial bit to be symbol 0 and verify that the reconstruction is invariant with
the choice of the initial bit. The results obtained for this case are given in
For detection of the differentially encoded PSK (DPSK) we my use Table 6.2 below. It can be easily verified that the extra chosen bit 0 changes
the receiver urrangement shown in Fig6.11. The received DPSK signal is
applied to one input of the multiplier. To the other input of the multiplier
the phase of the DPSK
invariant.
sequence but the detected sequence remains

Table 6.2 Differentially encoded sequences with phase

Symbol 1,
Binary data {b(k)} 0 100 1 0. 0 1
f0
DPSK wave
Decision Differentially 0 1 0 0 1 001 0 0 0
device
Symbol 0,
encoded data {d(k)}_
Phase of DPSK TtT 0 TT 0T
otherwise

Delay Shifted 01 0 0 1 0 0 1 0 0
T Threshold of
zero volt
differentially
encoded data {dl
Phase of T

Fig.6.11 Receiver for the detection of DPSK signals.

shifted DPSK
Phase comparison
Output
a delayed version of the received DPSK signal by the time interval T, is
Detected binary b() 001 0 0 1 0 0
applied. The delayed version of the received DPSK signal (in the absence of starting reference bit
channel noise) is shown in the 4th row of the table. The output of the
diference is proportion to cos0), where 6 is the difference between the
 DIGITAL COMMUNICATION DIGITAL MODULATION TECHNIQUES 267
266
6.4 FREQUENCY-SHIFT KEYING (FSK)

In an FSK system, two sinusoidal carrier waves of the same amplitude A PSD
but different frequencies f and f are used to represent binary symbols 1

M M
and 0 respectively. The binary FSK wave s) may be represented as

st)=A,cos2mf,) symbol 1
(6.5)
= A_ cos2) symbol 0 M
fa fa
The binary FSK waveform is illustrated in Fig.6.12 for the arbitrary

sequence10110101. Itcan be easily verified that binary FSK Fig.6.13 Power spectral density of binary FSK signal.
waveform is essentially a superposition of two binary ASK waveforms, one
with a frequency f and the other with frequency f. Hence the power
density spectrum of FSK is the sum of two ASK spectra at frequencies Generation of FSK signal
f
and f2. No discrete components appear in the spectrum of FSK sigmal. It
PSK
can be shown that by properly choosing fo and f the discrete The signal can be generated by applying the incoming binary data
represented in polar form) to a frequency modulator as shown in Fig.6.14.
components can be totally eliminated. It may be noted that the bandwidth of To the other input a sinusoidal carrier wave of constant amplitude A, and
FSK is higher than that of ASK or PSK. The spectrum of FSK signal frequency f is applied. As the modulating voltage changes from one level
is shown in Fig.6.13. FSK signaling scheme find a wide range of applications
in to another (both non-zero being polar), the frequency modulator output
low-speed digital data transmission. The main advantage of FSK lies in its
easy hardware implementation. changes its frequency in the corresponding fashion.

Frequency
AAAAAANAAAAAA Binary wave in polar Modulator Binary FSK wave s(t)

JUVVVVV JU
form b(t)

(a) Carrier wave

A, cos(2nf

Fig.6.14 Generation of binary FSK waveforms

Detection of FSK signal

FSK signal can be demodulated using synchronous or coherent detector. A
b) scheme for demodulation of binary FSK wave using coherent detection
technique is shown in Fig
Fig.6.12 Frequency shift keying waveforms, (a) Unmodulated carrier. (b) FSK
signal
 DIGITAL COMMUNICATION DIGITAL MODULATION TECHNIQUES 269
268

H
BPF
Envelope Sampler
,fol Detector
0
Symbol 1,
if 4>
Binary cos(27fe) Comparator FSK
Symbol 0.
FSK
otherwise
Comparator
wave

BPF Envelope
fa Detector Sampler
cos(27f.0)

Fig.6.15 Coherent detectoion of FSK binary wave. Fig.6.16 Non-coherent detection of FSK binary signals.

The detector consists of two correlators that are individually tuned to two A decision is made in favour of symbol 1 if the envelope detector
different carrier frequencies to represent symbols 1 and 0. A correlator output derived from the filter tuned to frequency fe is larger than that
consists ofa multiplier followed by an integrator. The received binary FSK
derived from the second filter. Otherwise a decision is made in favour of
signal is applied to the multipliers of both the correlators. To the other input
symbol 0.
of the multipliers, camiers with frequency and applied
f f are as
shown
in Fig.6.15. The multiplied output of each multiplier is subsequently passed 6.5 QUADRIPHASE-SHIFT KEYING (QPSK
through integrators generating output and 42 in the two paths. The output
of the two integrators are then fed to the decision making device. The The digital modulation techniques discussed thus far are spectrally
decision making device is essentially a comparator which compares the inefticient in the sense that the available channel bandwidth is not fully used.
output (in the upper path) and output &, (in the lower path). If the output The spectral efficiency can be improved upon substantially by using
qudriphase-shifi keying (QPSK) and minimum-shift keying (MSK). In this
produced in theupperpath (associated with frequency f) is greater than
section we discuss the QPSK technique. QPSK is an extension of binary
the output produced in the lower path (associated with frequeney f), the PSK. In binary data transmission, we transmit only one of two possible
detector makes a decision in favour of symbol 1. If the output , is less than signals during each bit interval T. On the other hand, in an M-ary data
, then the decision making device decides in favour of symbol 0 (say). transmission it is possible to send any one of M possible signals, during each
signaling interval T
This type of digital communication reccivers are
receivers. As before, the detector based coherent
also called correlation
on detection requires phase
and timing synchronisation. In most of the cases, the number of possible signal is

Non-coherent detection of FSK M 2" (6.6)

Binary FSK waves may be demodulated non-coherently using envelope where n is an integer.
detector. The received FSK signal is applied to a bank of
two bandpass
filters, one tuned to frequency fa and the other tuñed to f Each filter is The signaling interval is
followed by envelope detector. The resulting outputs of the two envelope
an
detectors sampled and then compared with each other. The arrangement
are T nl 6.7)
for non-coherent detection of FSK signal is shown in Fig.6.16.
270 DIGITAL COMMUNICATION
DIGITAL MODULATION TECHNIQUES 271
It can be easily appreciated that binary data transmission system is a
special
case of an M-ary data transmission system with M=2. Each of the M oCt)=-S7T dibit 00
Signals in an M-ary signal is called a symbol.

Quadriphase-shift keying (QPSK) is an example of M-ary data dibit 10

transmission with M=4. In QPSK, one of four possible signals is (6.10)
transmited during each signaling
dibit. For
interval, withbe each signal uniquely related dibit 11
to a example
the four dibits may 00, 01, 10,
1l in natural
coded form 3Tt
00, 10, 11, 0l in Gray encoded form. In QPSK system, we
or
4 dibit 01
may represent the four possible dibits in Gray encoded form by
a sinusoidal carrier with one of four possible values such as
transmitting
The phase variation of the carrier of QPSK for different dibits is shown in
Fig.6.18.
40-A co 2mi dibit 00

=
A cos 27nft- dibit 10 2T
(6.8) dibit 0 0
=A, cos2m,t+ dibit 11 3Tt

-
A, co 2m,t+ dibit 01

where 0t<T, T being the symbol duration. The QPSK waveform for an
dibit 1 0 2T
arbitrary binary sequence 11 0 01001 is shown in Fig.6.17.

ULASAUUAU 11 00 10 01 o(t)

dibit 11
Fig.6.17 QPSK waveform for the binary sequence 11001001.

t
The QPSK signal represented by (6.8) may be viewed as a
equ
special form of phase modulation. This can be easily appreciated by
expressing st) in the following form dt)
37t
dibit 01 4
st) A, cos =

27f,t+o()) (6.9)
where o(t) is the instantaneous phase value that assumes constant 21b
values
for each dibit of the incoming data. It can be
easily verified that
Fig.6.18 The coding of carrier phase of QPSK with dibits (Gray-coded form).
 272 DIGITAL COMMUNICATION
DIGITAL MODULATION TECHNIQUES 273
QPSK signal represented by cquation (6.9) can be further expressed
as
The the desired QPSK signal. It is apparent that the signaling interval T in a
QPSK system is twice as long as the bit duration T, of the input binary data
sd) = A, coso()cos(27f t)-A, sin(o()sin(27f, (6.11) stream. Therefore, for a given bit rate 1/T,,a QPSK system requires half the
transmission bandwidth of corresponding binary PSK system. In other
In this form, the QPSK wave s{t) may be considered to be composed of an words, for a given transmission bandwidth, a QPSK system carries twice as
in-phase component equal to A. cos o() and a quadrature component many bits of information as the corresponding binary PSK system.

equal to A, sin(o(t). The QPSK signal can be detected by using a pair of corelators
(multiplier followed by integrator) in parallel. The arrangement is shown in
On the basis of the above representation it is possible to set up a Fig.6.20.
block diagram for generation of QPSK signal. A typical arrangement for
QPSK signal generation is shown in Fig.6.19. It consists of a serial to In-phase channel
parallel converter, a pair of product modulators, two carrier sources in phase
quadrature and a summer. The serial to parallel converter represents each
successive pair of bits (dibits) of the incoming binary data stream as two
separate bits with one bit applied to the in-phase channel of the transmitter
Decision
Device
and the other bit applied to the quadrature channel. The output of the serial to 0

parallel converter in the in-phase channel fed to one input of the product cos(27tf.0)
input. Oscillator
modulator. The carrier signal A cos 27tf.t) is applied to the other
Threshold
The output of the serial to parallel converter in the quadrature channel is fed QPSK
to another product modulator input in the channel, other input of which is fed -90 Parallel-to- Binary
Signa Serial
with a quadrature carrier derived from the Phase-shifter
signal original
carrier through Converter | Wave
a 90 phase shifter. The output of the two multipliers are then added to get b(t)
Threshold
In-phase channel sin(2tf)
x Ja
Decision
Device
A cos(2f)
Oscillator Quadrature channel

Binary QPSK
Serial-to 90
Signal
Parallel Phase-shifter Mg.6.20 Block diagram of QPSK coherent receiver.
Wave Converter
b)
The correlator in the upper path computes the cosine of the carrier
A, Sin(27tf) phase whereas the correlator in the lower path computes the sine of the
carrier phase. The sign of the two correlator outputs are compared with the
help of a pair of decision devices and a unique resolution of the four

Quadrature channel transmitted phase angles are made. The parallel-to-serial converter
interleaves the decision made by the in-phase and quadrature channels of the
receiver and then reconstructs the binary data stream. In the absence of noise
thisdetectedbinary stream is identical to the original one transmitted by the
Pig.6.19 Block diagram of QPSK transmitter QPSK transmitter
 DIGITAL MODULATION TECHNIQUES 275
274 DIGITAL COMMUNICATION

On the basis of the above discussion we may view QPSK scheme as

a combination of two binary schemes that operate in parallel and employ two
cos27f,t). By applying this output to a 90° phase shifter, the second

carriers inphase quadrature. QPSK is therefore, a

quadrature-carier
reference signal sin(27f,t) is obtained. The divide-by-four circuit
multiplexing scheme that gives bandwidth conservation to binary PSK however, produces four phase ambiguities in the interval (0, 2r) which my
be taken care of by differentíally encoding the incoming binary wave at the
For the coherent receiver shown in Fig. 6.20 to work successfully, it
transmitter input and decoding the same at the receiver output.
is necessary that the locally generated carriers are in perfect synchronisation
with that used on the transmitter side. In order to provide a pair of coherent
sinusoidal signals for detection an arrangement shown in Fig.6.21 is often
66 MINIMUM SHIFT KEYING (MSK)
used. The synchronisation loop used to generate a pair of reference signal
tracks the randomly varying phase of the received QPSK wave irrespective
It has been observed that in a binary FSK, the phase continuity is maintained
at the transition points as the incoming binary data stream switches back and
of which of the four dibits phase-modulates the carrier. The synchronisation
loop that produces coherent reference signals that are independent of the forth between symbols 1 and 0. This type of modulated wave is referred to as
modulation is a fourth power loop. The a continuous-phase frequency-shift keying (CPFSK) wave. Minimum-shift
received
signal is first
passed keying (MSK) is a special form of binary CPFSK with the change in the
through a bandpass filter to minimise the effects of noise and the raised to
the fourth power. The fourth harmonic of the carrier produced by this carrier frequency from symbol 0 to symbol 1 or symbol 1 to symbol 0 is
process is then tracked by a phase locked loop. The resulting sinusoidal equal to one half the rate of the incoming data. If öf denotes the change
output is next applied to a frequency divide-by-four circuit that yields in carrier frequency due to transition from one symbol to the other and T
denotes the bit duration, the MSK wave will satisfy the condition

Bandpass Narrow 1
Filter band
Filter
Lowpass
Filter 6f21 (6.12)

Let f and fa represent the transition frequency of symbol 1 and 0

respectively. The frequency in tems of of
fo can be represented as

Voltage follows
Controlled
Oscillator
Frequency
ataala
2
Divider 8f
4

90 where f= tf (6.13a)
Phase 2
Shifter
and f =f-foa (6.13b)

() Similarly we may express fa as

Coherent Reference Signals

Fig.6.21 Synchronization circuit for QPSK detection.

-f,-28f (6.14)
 DIOITAL MODULATION TECHNIQUES 217
276 DIGITAL COMMUNICATION The MSK may be viewed as another example of
wave
The frequency f is the unmodulated carrier frequency and is cqual to the multiplexing, Expressing the MSK wave sdt) as the following quadrature
arithmetic mean of the two transmitted modulated wave we get frequency
frequencies f and f
coresponding to the symbols 1 and0.
)= A, cou2mf,)+40)
The deviation infrequency from the unmodulated carrier

8f A, cosl600Jeo211,1)- A, sin[ø)Jain( 2nf, (6.19)

(6.15a)
Thus the MSK wave s()
fa-f,-28f
has a in-phase component equal to A, coso(t)
(6.15b)
and a quadrature component equal to A, sin|6()|. Like QPSK, there are
The MSK signal is represented in the generalised form as four dibits that may be considered in the
00,10,11 and 0 1(in Gray code).
case also. These dibits may be

d0-A, cos2nf, t + o(0) (6.16) Letus consider the case of transmission of the dibit 0 0. The phase
of the MSK wave decreases (representing the first symbol 0 ) upto the end of
where o(t) = tnôft. the bit period T, which is again followed by a decrease due to the second

Using equation (6.12) we find that the transmission of symbol 1(i.e.,

symbol 0 in the dibit. At the end of another T,, the total phase decreases to
frequency fa) changes the phase of the MSK signal s(t) by an amount -T The phase history of the MSK wave follows the path shown in
Fig.6.22 (a) for the transmission ofthe dibit 0 0. Likewise, we find the other
given by paths followed by the phase of the MSK during transmission of other dibits
10,11 and 01 as shown in Fig.6.22 (b),(¢) and (d) respectively. In the
o(t)= Töf illustration (Fig.6.22) the initial phase has been assumed to be o(0) =0. It
(symbol 1) (6,17) may be noted that at the end of the bit period t= T,, the phase MSK
wave equal to + Tt/2 or -t/2 and that at the end t= 2T, it is equal to 0
ofthe
At time t T, that is at the end of the termination of the or Tt radians. The pair of values of sin o(T,)and coso(27,)
=
interval
representing transmission of symbol 1, the phase of the MSK increases by an corresponding to each of the four possible dibits are shown in Table 6.5. It
amount t/2 radians. can be easily verified that each dibit in MSK is uniquely defined by a set of

On the other hand, the transmission of symbol 0 e,, frequency f2)

values of sin o(T.).co o(2T,)
changes the phase of the MSK wave s) by an amount MSK may be viewed as a quadrature multiplexed frequency-
modulated wave. The basic difference between QPSK and MSK is that
in
o() = -nôft constant distinct value for the entire
QPSK the phase shift o() assumes a
in MSK the
Ttt (symbol 0) duration of a symbol depending on the dibit transmitted whereas
=- (6.18) phase shift o(t) varies with time along a distinct straight path depending on
the dibit being transmited.
At time t= Th, that is at the termination of the interval representing
transmission of symbol 0, the phase of the MSK decreases by an amount
Tt/2 radians.
 DIGITAL COMMUNICATION

278
DIGITAL MODULATION TECHNIQUES 219
T 27, Table 6.5 Pair of sine and cosine values
of the four possible dibits

Dibit
dibit 00 (Graycoded)
00
sin o(T.) codo(27)
10 +1 +1
11 +1 1
01

(a)
MSK wave may be
generated by using a frequency-modulator that
satisfies equation (6.12). MSK signal can be detected
by using coherent
detection technique consisting of a pair of correlators with
built-in-memory
and decision circuits to make choice over successive
dibit 10
pairs of bit intervals.
Generation and detection of MSK signals involve further mathematical
treatment that is beyond the scope of the book.

2T

(b)

dibit 11

T 2T

(C)

T 2T,
dibit 01
t

(d)
Fig.6.22 Carrier phase coding for MSK for dibits in Gray-coded form
 DIGITAL COMMUNICATION
280

Problems

6.1 Sketch the binary PSK waveform for the bit sequence

1010110110

6.2 Sketch the binary ASK waveform for the bit sequence

0110101001

6.3 Sketch binary FSK waveform for the sequence 1 0 01101011. Assume

that
fe2fe T, being thebit duration

6.4 Sketch the QPSK waveform for the sequence 1101010 010, assuming
the carrier frequency to be equal to the bit rate.

6.5 Suggest a suitable synchronising circuit arrangement for extracting the carmier at
the receiver end for coherent detection of binary PSK.

6.6 Sketch the waveform of the in-phase and quadrature components of a QPSK
signals for the binary sequence 1 101011 010.

6.7 For the sequence given in problem 6.6, sketch the QPSK waveform

6.8 Sketch the waveform of the MSK signal for the input binary sequence
1100100010