t •·~•

• • ·• DIGITAL MODULATION TECHNIQUES ◄

s.-,<(-.,~_,.~ • . •,.,..,.~~~,,..,,...,..,.,..>,.
...,_4;'_,,,_,._; ~., ,Jlllf/<,,·~ ~~Mv• .-,r-.,
31.7·_
1
• Integrator output= b (t)
2 0
JP: [2 k Tb + Tb - 2k Tb + Tb] ... (6.56)

/· Similarly we can prove that output of the lower integrator is given by b.(t) .JP. Tb. Thus, at th~
output of the two integrator~, we obtai~ the bit streams b 0 (t) and be(t). As explained for the BPSK
detection, we need to u~e _b~t synchronizers in QPSK receiver as well. It is used to establish the
beginning and end of the bit intervals of each bit stream. The bit synchronizer is-also used to operate
the sampling switch. At the end of each integration time for each integrator, the integrater output is
sampled. The samples are taken alternately from the two integrator outputs, at the end of each bit
time Tb and these samples are then held in the latch for the bit time Tb. Each individual integrator
output is thus sampled at intervals of 2 Tb. At the output of the latcp., we get the signal b(t) .
., .
&.13.4. Signal Space Repr~sentatlon of QPSK Signal .• • ,
. I

w~ know that a QPSK signal is m~thematically represented as under:·

VQPSK (t) = .J2 P8 cos [ ro0 t + (2m + 1) :] m =O, 1, 2, 3- ...(6.57}

By substituting different values of 'm', we can get the sinusoidal signals with different phase
shifts. Let us expand equation (6.57) using the standard trigonometric rule i.e.,
cos (A + B) = cos A cos B - sin A sin B
-Applying this to equation (6.57), we get
1
· VQPSK (t) .J2 P8 {[cos o>ct x cos (2m + 1) : ]-[sin ro0t sin (2m + 1) : ]} .. ,(6.58)

Rearrange equation (6.57), we get,

V QPSK (t) = {.J~~~~ ~~~ I (2m + 1) : ] x ~-:~cos ro.t}-

.... ..'

-{.JP8T.'sin [<2m + 1) :] x.J21T. _sin ro.t} ...(6.59)

In equation (6.59}, let us a_ssume that,

b 8 (t) = ✓2, ~s [<2m +1) :] an~b (t) =✓2 sin [<2II1 + 1) ¾] 0 .. (6.60)
/ .

We also assume that, ci>1(t) = U1 (t) = ✓2/Ts cos roct ...(6.61)

and • 411
2
(t) = u 2 (t) =.J2IT 8
' ro0 t ...(6.62)
Substituting equations (6.60), (6.61) and (i.62} in equation (6.59), _we get
. • I

- • 11 • 1
. VQPSK(t) = ..JP. T8 x-:Ji
b 0 (t) x u 1 (t) + .JP. T8 x ✓2, b 0 (t) x u 2(t) ...(6.63)

Where, u 1(t) and u 2(t) are the two orthonormal basis functions. In the generation of QPSK signal,
these orthonormal signals are used as carriers. In equation (6.63), T9 .is the sy~bol duration. Iri
QPSK, two successive bits will form one symbol. Hence, T9 = 2 Tb. Substituting in equa~ion (6.63),
we get
 ~;_J
MUNICATIONS
DIGITA COM
r2P ~x- =bo (t)> <u2 (t)
1 L
-----::::--1 b (t) )( U1 (t) +"' ~ s ✓2 E
VQPSK(t) == V ~ r2 PS Tb ><-Fn2 e
"'~
• l
_ E i e energy per b1•t • fE x b (t) •.. (6.64)
But , Ps Tb - 0
'E >< b (t) >< Uz
• ., (t) x u 1 (t) + ✓ O
fa QP SK sign al, as shown
~b
V (t) == ✓ ~b e rese ntat
•
1on o
Therefore, QPSK 1 t the signal space rep f eren ce axis.
Equation (6.64) can ?e u;ed t(ij:nd u2 (t) are use
in figure 6.35. The two s1gna s u 1
d as re •
-u2(t)

b0 (t) = b8 (t) = 1
bo(t} = 1, be(t) = - 1
sit) = 10
---------- s4 (t) = 11

J
U2(t) = 2/Ts sin roct
Fig. 6.35. Signal space representation of QPS
K signal
\

Imp orta nt Conclus~ons 1 .

Some of the imp orta nt cone us1ons a bout the \ign al space rep rese nta tion may be described as
under:
• · • •
(i) The distance of a sign al poin t from the or1g b
1n is give n y'
[Distance (0 - s4)] 2 = ✓(P. Tb)2 + ✓<P. Tb)2
or '
(O- s4) 2·= T8 Tb+ PsT b=2 Eb= Es
Hence, (0 - s4) = .{E: ... (6.65)

As shown in figure 6.35, all the sign al poi nts .

s 1(t) to s 4 (t) are at a d1sta 'E from the
nce v us
origin.
. _·· e
(ii) The ability to determine a bit wit hou t erro
r is mea sur ed by the dist anc e in sig~a~ s~a
between the poin ts corr espo ndin g to diff eren ~
t val ues of the bit. Fro m figu re 6.35, it 18
tha t the distance betw een the poin ts whi ch ce
diff er in a sing le bit is giv en by,
... (6.66)
whe re Eb= Ene rgy con tain ed in a bit ttan
smi tted for tim e T .
(iii) The distance 'd' for QPSK (shown by equ I(
atio n (6.66)) is exa ~tly sam e as tha t for BP~
system. Thus, the noise.immunity of the QP SK !(
sys tem is exa ctly sam e as tha t of the BP
systeqi. 1 •
 IQUES ◄
DIGITAL MODULATION TECHN
6.13.5. Spec trum of QPSK
Earlier, we have obtain ed the pow
. • • •
bit extend1ng over a period Tb as underer
: spectr al densi•t Y (PSD) of an NRZ bipolar signal with each

S-(f) = Tb [sin (m Tb)]2

PS

(7tf Tb)
This is the PSD of signa l b(t). In QPSK th· · .•
i.e., be (t) and b 0 (t) respectively. Each symb ol:: ~?;al b~t) 18 divide
d into even and odd bit stream s
Hence, their power spectr al densit ies are given by, se bit streams has
period of T 8 = 2 Tb seconds.

Se (f) = p T [sin (m Ts)]2

s s 7tf T ... (6.67)
s

and S (f) = P T [sin (7tf Ts)]2

o s s 7tf T ... (6.68)
s

.,_...-;. ..__.... .;::11i -t-e---- -=:*:_ _--k_ _..: :~~ ~~c _~~ -- t
-1/2Ts -1/Ts O 1/Ts 1/2Ts
=-f@ =f@
(a) Power spectral density of a baseband QPSK signal

S(f)

. . , a . c . . . _ ~ ~ - - - - L - - ~ ~ - ~ ~ + -.........
~-~ ~-_ _J- ---~ ...£ -~~ ~-- ~t
- fC f -
C
i2 fC f
c+2
fb
I•
BW = fb .. ,
(b) Power spectral density of a QPSK signal

Fig. 6.36.
As b 0 (t) and be (t) are statis tically indep enden t, the baseb and power /

spectr al densit y of a QPSK

signa l is given by,

... (6.69)

This equat ion gives us the baseb and power spectr al densit y of a QPSK
signal , which is plotted
in figure 6.36(a). When we use this signa l to modu late the sinuso idal
carrie r of frequency fc, the PSD
given by equat ion (6.69) will be shifte d to ± fc· The graph s of PSD
for QPSK have been shown in
figure 6.36(b). It may be obser ved that these are very simila r to the
PSD plots of BPSK.
 ► DIGITAL COMMUNICATIONS

6.13.6. Bandwidth of QPSK tem .

hal f of the ban dw idt h of BP SK sys
The ban dw idth of QP SK sys tem is one
2 fb _ £
There£ore, BW = - 2 - b
reduction
wn in figu re 6.3 6 (b). Thu s, the adv antagEl of mu ltil eve l mo dul atio n is
. ~s has bee n sho •
1n req uir ed ban dw idth . Fin d th
ing inp ut bin ary seq uen ce b(k ) = {l, - 1, 1, - 1, - 1, - 1, 1, l}.
EXAMPLE 6.2 For the foll ow for QP SK .
e
ske tch the tra nsm itte d wa vef orm
tra nsm itte d pha se seq uen ce and \I
hav e bee n sho wn in figu re 6.37 .
Solution: The req uire d wav efor ms
-1 1 1 II

-1 1 -1 -1
b(k) 1
4 5 6- 7 8
Bit nubmer 1 2 3

Even bit sequence t

be(t) 0

Odd bit sequence

b0 (t) t
0

Phase shift
t t t t t t .~ t
goo goo goo oo
Fig. 6.37
6 13 • .
• .7. Error Probability of QPSK S
The expression for erro r .. ystem
. .
probab1hty of a QPSK s~s tem is
. ' given by
1 '
Pe== 2 erfc -JF;; !No ,,-
 •
DIGITAL MODULATION TECHNIQUES ◄

will
Compare this with the expression for error probability of the coherent BPSK system. You
in
find that the erro~ prob_ability of QPSK is same as that of BPSK. So, the performance of QPSK
presence of noise 1s equivalent to that of a BPSK system.

6.13.8. Advantages of QPSK

(i) Very good noise immunity.
dth
(ii) Baud rate is half the bit rate therefore more effective utilization of the available bandwi
of the transmi ssion channel. ·
(iii) Low error probability.
Due to these advantages, the QPSK is used for very high bit rate data transmission.
6.13.9. Drawback
The generation and detection of QPSK is quite complex.
6.13.10. QPSK is Better than PSK
The QPSK is better than PSK because of the following reasons:
(i) Due to multilevel modulation used in QPSK, it is possible to increase the bit rate to double
the bit rate of PSK without increasing the bandwidth.
(ii) The noise immun ity of QPSK is same as that of PSK system. .
(iii) Available channe l bandwi dth is utilized in a better way by the QPSK system than PSK
system.
EXAMPLE 6.3. What are the phase states of the carrier when the bit stream,
10011 01100
is applied to a QPSK modula tor.
Solution: (i) First, we divide the input data stream into groups of two bits i.e., dibits.

101 011 101 111 oo I

(ii) Also, the phase shift to each dibit are as under:

Modulator input 10 01 10 11 00

Phase state 180° 90° 180° 270° oo

,

6.13.11. Performance Comparison of BPSK and QPSK

S;No. Param eter/ch aracter istics BPSK QPSK '

1. Variable characteristics of Phase Phase

the carrier
2. Type of modulation Two level (binary) ' FourJev el
3. Type of representation A binary bit is represe nted A group of two binary
by one phase state bits is represe nted by
' I one phase state
4. Bit rate/bau d ra@ Bit rate =baud rate Bit rate =2 baud rate
5. Detection method Coherent Coherent
6. Complexity Complex Very complex
7. Applications Suitable f0r applications Suitable for applications
that need high bit rate needing very high bit
rates
 J
► DIGITAL COMMUNICATIONS
QPS K (4 PSK) sign al:
t he
• d t stre am' d raw
EXAMPLE 6.4. For the following a a y has ~ot
11010001 . • 8 It ay be noted that the carr ier frequenc
3 m
hOwn 1n .figurhe 6. d• acco bOI
c • rding to the sym •
Solution: The required waveiorm is shift is being c ange
e ph as es
been changed at all. Only t h 00 1
1 0
1 1 0 S2
Bit stream S1
S2 goo
S4
symbol goo oo
Phase shift 270°

Wave
Fig. 6.38.

K
6.13.12. Difference between OQPSK and QPS -
QPSK
OQPSK
S.No. The re is no offset betw een b0 (t) and be (t)
1. There is an offset of Tb seconds
be~ween b0 (t) and he (t)
b0 (t) and be (t) can cha nge sim ulta neo usly
2. b0 (t) and be (t) will never change
simultaneously
Maximum pha se change,.,is 180° or 1t.
3. Maximum phase change in the
outp ut signal is 90° or (7t/2) pt
•Amplitude vari atio ns at the inst ants of abru
Amplitude variations at the inst ant
4. pha se cha nge s are of larg er amp litu desis
of abru pt phase changes are of
smaller amplitudes as compared compared to OQPSK
toQ PSK

6.14 M-ARY MODULATION TECHNIQUES

... 9M(t),
of the M possible sign als suc h as s 1(t), s 2 (t),
In ~n M-ary s~gnal~ng ~cheme, we can s~nd one given
~ds. The num ber of pos sibl e sign als i.e., Mis
dun ng each s1gna.hng interval of duration T seco
by, •
M =2N (6. 70)
where, N is an integer
Adv anta ge
nel
schemes whe n we wan t to conserve the chan
Th_e M-ary schemes are preferred to the bina ry
bandwidth.
Drawbacks
cost of following draw bac ks:
~e bandwi~th conservation is achieved at the
(1) Increase in transmitted power.
(ii) Increase in error probability.
Types ofM -ary syst ems
ems in details.
H(:)re, we shall discuss the follo~1ri~1~i~-ary syst •
1 M-ary PSK
(ii) M-ary QAM
(iii) M-ary FSK
 DIGITAL MODULATION TECHNIQUES ◄

6.15 M-ARY PHASE SHIFT KEYING (PSK)

In QPSK, we have grouped together two bits to form messages. Depending on the two bit message
(00, 01, _10 or 11), a sinuso_idal signal of duration equal to 2 Tb and having a parti~ular p~ase shift is
trans~1tted. The QPSK signals differ by 90° or 7t/2 radians in phase with respect to each other.
The same principle can be further extended to obtain the M-ary PSK system. The M-ary PSK
signals are obtained as_ under :
I

. (i) GroupN bits together to form N-bit symbols.

(ii) These symbols will extend over a period of NTb where Tb is the duration of one bit.
(iii) Due to the grouping of N bits per sy~bol, we can have 2N =M possible symbols.

1t!M

, (t) = u2(t) ~ •{!_ sin roct

2
j T5 • _

Fig. 6.39. Geometric representation of M-ary _PSK signals

(iv) These M symbols are represented by sinusoidal signals of duration Ts =~~b which differ
from one another by a phase 21t/M radians. Thus, the M-ary PSK waveform can be
mathematically represented as, •
VM-ary PS~ = ✓2 P 8 cos (met+ q>~) ... (6. 71)
=
where m 0, 1, ... M - 1
(v) In equation (6. 71), q,m represents
..
the symbol phase angle, which is given by

q,
m
=(2m+l)~
M .
._.. (6. 72)
(vi) The M-ary PSK waveforms corresponding to equation (6. 72) are represented by dots in
-. •-:: --::- figura.6_._3~~ This is a ·signal space diagram in which the co-ordinate axis are formed by the
orthonormal waveforms u 1(t) and ult) given by the following equations:
• =
_u 1(t) :.... ✓(2/T8 ) cos met and u 2(t) ✓(2/T8 ) sin met
These are same orthoi-iormal _signals ~sed for the signal spac~ representation of a QPSK
signal. 1
(vii) The distance of ~ach M-ary P~¥,.waveform from the ~~igin is_ given_ by
--- • .JE: =✓Perr. J' , ,

/
 ► DIGITAL COMMUNICATIONS

(d) 18
6.15.1 Distance between Signal Points also know n as the Eucl idea n dista nce. This dista nce •
. The distance between signal points (d) is
munication syte m. The erro r prob abili ty goes
impo rtan t in deciding the erro r probability of a com
d. In orde r to obta in tile expr essio n for the distanon
decreasing with increase in the Euclidean distance
ce
d, let us consider figure 6.40. the angle 0 is equa l to (2n;/M) beca use 2 7t corr espo nds to the complete
(i) A_s shown in figure 6.40,
circle and M is the num ber of messages. gle 0-8 -X. Usin g the defi nitio n of sin 0, we get
(ii) Now, let us consider tM righ t angle trian 1

d/2 ... (6. 73)

. Distance S1 -X
sin (7t/M) = D'1stance O- s1 = ~v'E
.cis
... (6. 74)

Therefore,
d = 2JE.
sin (1t/ M)

0 = 21t/M
,
d = Euclidean distance

''
''
''
''
'' u1(t) or <j,1(t)
'
''
''
''
''
'

SM-1

•
Fig. 6.40. Signal space diagram t0
calculate dista nce d
Equation (6 74) · • W • n
. valu e of M t 0 get t h e Euclidea
· is a gene ral
. expr e ssion . e can subs titut e any
distance d.

B d •
6.15.2 Power Spectral Density (PSD) and an width of M-ary PSK
(i) In the prev ious secti on h
ht · , we ave
o ained the expression for power
spec
PSKtral• dens ity of the base b and 2P5 NTb
Q signal as under;

SB (QPSK) (f) = 2 PsTs [sin (1tf Ts)J2

..) 7tf T
(n However, QPSK is • s
M-ary PSK w·ith Ma special case of
can use the ab 4. Hence, -we
write the e o~e iXpr essio n to -~£___~- ~/ T~ -c oJ --~ ~~ d: _,
b- b xpress10 n for PSD of th 1; _ 1rr,
ase and M-ary PSK e
The 1
only modificati~n requ~rysdt e~.
e 1s, that Fig. 6.41. PSD of bases:
and M-ary PSK signal
•I

C
 N TECH NIQUES ~ ~
DIGITAL MODULATIO

pression.
e Ts = NTb in th e above ex
we ha ve to su bs tit ut 2 (6 75)
NTb>] ••• •
S (f) = 2 PsNTib [s in7tf(m
NTb
Hence, 6.40.
B(M-ary PSK)
ba se ba nd M -a ry PS K as shown in figure
D of th e
), we ca n plot th e PS
Using equation (6. 75 ~h ow n in figure 6.41, we ha
ve
e pl ot of PS D
ry PS K : From th
B an dw id th of M-a
BW = J_ __(_J_Ts_] = ~
Ts
Ts

B ut T5 = NTb
2
Therefore, BW = NTb
1
or, Tb =fb
... (6. 76)
BW = 2fb
Hence, N
ove expression
nd w jd th of a BP SK system is 2 fb. The ab
& lm po rt an t Po in
t: We know th at th e ba m es sa ge, th e ba nd w id th re
duces.
r of bi ts pe r
ea se in nu m be
I . tells tis th at w ith in cr
ansmitter
6.15.3. M-ary PSK Tr K sy st em ha s been shown in fig
ure 6.42.
an M -a ry PS
The block di ag ra m of bit symbol N
Analog voltage

0
Sinewave
1 Digital gPnerator M-ary PSK output
Serial to VA phase
', b(t) to 2
analog controlled

''
Digital parallel converter by VA
input converter

r
I

N--1

K transmitter
Fig. 6.42. M- ar y PS
ts of
ki ng O pe ra tio n l co nv er te r. Th is block ca n st or e th e N bi
W or pa ra lle one af te r th e
applied to a se ria l to lly, in tim e sequence
The bit st re am b(t) is en pr es en te d se ria en, all
bo l. Th es e N bi ts pe r symbol ha ve be th e se ria l to pa ra lle l ~onverter block. Th
a sy m by sage
e N bi ts pe r sy m bo l ar e fir st assembled of th e co nv er te r. Thus, th e N-bit mes
other. Th ou tp ut lines tp ut of th e
te d at once on th e N lle l converter. The ou
these bits ar e pr es en of th e se ria l to pa ra
is th e time
ar s in th e pa ra lle l form at th e ou tp ut du ra tio n of N T b of a symbol which r
ap pe ra seconds, th e converte
ra lle l co nv er te r re m ai ns unchanged fo ts. A fte r ev er y N Tb
se ria l to pa w group of N bi Depending
d fo r th e co nv erter to assemble a ne r is th en ap pl ie d to a DIA converter.
re qu ire er te 2N = M
is up da te d. Th e N bi t ou tp ut of th e conv ut VA . Th e N bi t di gital in pu t can have
ou tp ut
pu t, it produces an an
alog ou tp mber of distinct
on th e N bi t digital in DI A co nv er te r ou tp ut VA will have M nu al signal
e
m be r of po ssib le co mbinations. Hence, th an al og vo lta ge is applied to a sinusoid hich
nu
de pe nd in g on th e symbols. Finally th is so id al ou tp ut vo lta ge, th e ph as e <Pm of w
values, itu de sinu ge t a fixed
uces a co ns ta nt am pl of ehe tra ns m itt er , we
generator, which prod VA· Th us , at th e ou tp ut th e N bit
te rm in ed by th e D IA converter ou tp ut ch ha s a on e to on e correspondence to d.
is de
si nu so id al w av ef or m, th e ph as e of whi NT b. Th us , th e M -ary PSK is generate
am pl itu de mbol time T6 =
m bo ls. Th e w ill ch ange only once pe r sy
sy
 6.15.4
The block diagram of M-ary PSK receiver has been shown 1n figure 6.43. This 1s same as the
non-offset QPSK receiver_.
s(t) = Pe cos coct s(t) sin coct
-p0 sin coct
M-ary ~ - - - - ~ - - - ,
PSK
Raise input
to
' th
M power
O Parallel
Analog to .,_1-----t to Bit sequence
Bandpass digital serial b(t)
Carrier
filter Lr_ec_o_n_ve_r_s_io_ni-:-:--:-1converter
recovery
circuit Mfc N-1
cos Mcoct
Frequency TSPO
divider
+M

s(t) cos coct

Fig. 6.43. M-ary PSK receiver
Workin g Operat ion .
The M-ary receiver operates on the principl e of synchro nous demodu lation .which we had
discusse d for BPSK and QPSK. The carrier recovery system will require a device which can raise
y
the received signal to Wh power. This signal is then applied to a bandpas s filter, the center frequenc
that is
of which has been selected to be Mfc. At the filter output, we get a sinusoid at frequen cy Mfc
Mth harmoni c of the carrier frequenc y f . This frequenc y is then divided by M to obtain the carrier
d
at frequency fc. The two carriers produce d at the filter output are cos coct and sin coct. These recovere
carriers are then applied to two multipli ers (balance d modulat ors). The other input of each multipli er
block is connecte d to the received M-ary PSK signal. The outputs of the balance d modulat ors are
applied to the integrat ors.Sinc e, the M-ary PSK system is a non-offs et or non-sta ggered type of
system, the integrat ors will extend their integrat ion over the same time interval . Tp.ese integrat ors
will work alongwi th a bit synchro nizer, which has not been shown in figure 6.43. The outputs of the
integrat ors are proporti onal to T 8 pe and T8 p 0 respecti vely and they (outputs ) change at the symbol
rate. These outputs are then applied to an· A to D convert er which yields the N bit transmi tted
g
signal. The signal is converte d into b(t) by using a parallel to serial convert er. Now, the operatin
systems with N =4 bits and M =2 =1~ are common . The bandwi dth of such a system will be given
4
by, I •

BW = 2 fb = 2 fb = fb ... (6. 77)

N. 4 2
Advant ages of M-ary PSK
(i) Bandwi dth reduces with increase in the number of bits per symbol (N).
(ii) A~ the informa tion is transmi tted through signal phase and not through· signal amplitud e,
this system operates ·
very well 1·nthe s1·tuat·ions w h ere the signal ·
. . amplitu de varies due to
the character1st_1cs of the transmis sion medium .
Disadv antages of M-ary PSK:
• N
(i) The probabil ity of error increase s with increase in th.e numb er Of b its b I the
distanc ed decrease s with increase in N.. per sym o, as
(ii) The transmi tter. and receiver of M-ary PSK is very comp1ex.