Modul #06

TE3223
SISTEM KOMUNIKASI 2

QAM & FSK

Modulasi,Demodulasi,Kinerja
Program Studi S1 Teknik Telekomunikasi
Departemen
p
Teknik Elektro - Sekolah Tinggi
gg Teknologi
g Telkom
Bandung 2008

QAM (Quadrature Amplitude Modulation)

Gabungan Modulasi ASK dan QPSK

Amplitude
p
Shift Keying
y g ((ASK)) modulation:

2 Ei
si (t ) =
cos(c t + )
T
si (t ) = ai 1 (t ) i = 1, K , M

1 (t ) =

O
On-off
ff k
keying
i
(M
(M=2):
2)
0

s2

s1

E1

1 (t )

2
cos(c t + )
T

ai = Ei

Modul 6 - Siskom 2 - QAM dan FSK

M-ary ASK

M-ary ASK sering disebut M-ary Pulse

Amplitude
p
modulation ((M-PAM))
si (t ) = ai

2
cos(c t )
T
4-ASK = 4-PAM:

si (t ) = ai 1 (t ) i = 1, K , M

00

s1

2
1 (t ) =
cos(c t )
T
2

s2

3 E g Eg 0

ai = (2i 1 M ) E g
Ei = s i

01

11

s3

Eg

10

s4

1 (t )

3 Eg

= E g (2i 1 M )

( M 2 1)
Es =
Eg
3
Modul 6 - Siskom 2 - QAM dan FSK

Error probability .

Coherent detection of M-PAM (M-ary

ASK )
Decision

variable:
00

4-PAM

1 (t )

r (t )

01

s1

3 Eg

z = r1
11

s3

s2

Eg

r1

Eg

10

s4

1 (t )

3 Eg

ML detector
(Compare with M-1 thresholds)

Modul 6 - Siskom 2 - QAM dan FSK

m
4

Error probability .

C h
Coherent
d
detection
i off M
M-PAM
PAM .

Error happens if the noise, n1 = r1 s m , exceeds in amplitude

one-half
one
half of the distance between adjacent symbols. For symbols
on the border, error can happen only in one direction. Hence:

(
P (s ) = Pr (n = r s

Pe (s m ) = Pr | n1 |=| r1 s m |> E g
e

1
PE ( M ) =
M

> Eg

Pe (s m ) =
m =1

for 1 < m < M ;

and Pe (s M ) = Pr n1 = r1 s M < E g

M 2
1
1
Pr | n1 |> E g + Pr n1 > E g + Pr n1 < E g
M
M
M

2( M 1)
2( M 1)
2( M 1) 2 E g
=
pn1 (n)dn =
Q
Pr n1 > E g =

E
N0
g
M
M
M

( M 2 1)
Es = (log 2 M ) Eb =
Eg
3

PE ( M ) =

2( M 1) 6 log 2 M Eb
Q
2
M
M
1 N 0

Gaussian pdf with

zero mean and variance N

Modul 6 - Siskom 2 - QAM dan FSK

/2

Modulation: representation
Any modulated signal can be represented as (include QPSK, QAM)
s(t) = A(t) cos [ct + (t)]
amplitude

phase or frequency

s(t) = A(t) cos (t) cos ct

in-phase

- A(t) sin (t) sin ct

quadrature

Linear versus nonlinear modulation impact on spectral efficiency

Linear: Amplitude or phase modulation
N
Non-linear:
li
f
frequency
modulation:
d l ti
spectral
t lb
broadening
d i
Constant envelope versus non-constant envelope
hardware implications
p
with impact
p
on p
power efficiency
y
(=> reliability: i.e. target BER at lower SNRs)
Modul 6 - Siskom 2 - QAM dan FSK

Teknik Modulasi M-ary Kuadratur

( diagram blok konseptual )
sI (t )

xI (t )
A cos(( 2f c t )

d I (t )

b(t )

yI (t )

y (t )
A sin( 2 f c t )

dQ ((t)

yQ (t)
sQ (t )

A cos( 2f c t )

y (t
(t )

xQ (t )

~
dI (t)
~
b (t
(t )

2
A sin( 2 f c t )

~
d Q (t )

Modul 6 - Siskom 2 - QAM dan FSK

Bentuk gelombang 16 QAM

1.5
1
b(t ) 0.5
0
-0.5
0
1.5
1
d I (t )0.5
0
-0.5
0
4
sI ( t ) 2
0
X I (t )
-2
-4
0
4
2
yI (t ) 0
-2
-4
4
0
5

5
y (t ) 0
-5

10

15

20

25

30

35

40

10

15

20

25

30

35

40

10

15

20

25

30

35

40

10

15

20

25

30

35

40

10

15

20

25

30

35

40

10

15

20

25

30

35

40

10

15

20

25

30

35

40

~s ( t ) 2
I
0
-2
5

10

15

20

25

30

35

40

10

15

20

25

30

35

40

10

15

20

25

30

35

40

y(t ) 0
-5
5
0

0
4
2
x~I (t ) 0
-2
-4
0

0
1.5
1
0.5
~
d I (t ) 0
-0.5
05
0
1.5
1
~
b (t ) 0.5
0
-0.5
0.5
0

Modul 6 - Siskom 2 - QAM dan FSK

M-PSK and M-QAM

Complex Vector Spaces: Constellations
M-PSK (Circular Constellations)
bn

M-QAM (Square Constellations)

4-PSK
16-PSK

bn
16-QAM
4-PSK

an

an

Tradeoffs
Higher-order modulations (M large) are more spectrally
efficient but less power efficient (i.e. BER higher).
M-QAM is more spectrally efficient than M-PSK but
also more sensitive to system nonlinearities.

Modul 6 - Siskom 2 - QAM dan FSK

Two dimensional mod. (M-QAM)

M ary Quadrature Amplitude Mod

M-ary
Mod. (M
(M-QAM)
QAM)
2 Ei
si (t ) =
cos(c t + i )
T
si (t ) = ai1 1 (t ) + ai 2 2 (t ) i = 1, K , M

1 (t ) =

2
2
cos(c t ) 2 (t ) =
sin (c t )
T
T

where
h ai1 and
d ai 2 are PAM symbols
b l andd E s =

2( M 1)
3

( M + 1, M 1) ( M + 3, M 1) L ( M 1, M 1)

L
(

M
+
1
,
M

3
)
(

M
+
3
,
M

3
)
(
M

1
,
M

3
)

(ai1 , ai 2 ) =

M
M
M
M

( M + 1, M + 1) ( M + 3, M + 1) L ( M 1, M + 1)

Modul 6 - Siskom 2 - QAM dan FSK

10

Two dimensional mod. (M-QAM)

16-QAM
0000

s1

1000

s5

2 (t )
0001
0011

s2

1001

s6

s3

0010

s4

1011

1010

s7

s8

-3
3

-1
1

s9

s10

s12

1100

s13
0100

-1 11
1101
1111

s14
0101

-3

s15
0111

1 (t )

1110

s16
0110

Modul 6 - Siskom 2 - QAM dan FSK

11

Two dimensional mod. (M-QAM)

Coherent detection of M-QAM

1 (t )

z1

ML detector
(Compare with M 1 thresholds)

r (t )

Parallel-to-serial
converter

2 (t )

z2

ML detector
(Compare with M 1 thresholds)

Modul 6 - Siskom 2 - QAM dan FSK

12

Error probability

Coherent detection
of M
M-QAM
QAM

2 (t )
0000

s1

1000
s
5

0001

s2

0010
s 3 0011
s4

s1001
s8 1010
s 7 1011
6

1 (t )

16-QAM

s9
1100

s13

1 (t )

r1

0100

ML detector

r (t )

Parallel-to-serial
converter

s11

1101

s14
0101

s12
1111

s15

1110

s16
0111

0110

(Compare with M 1 thresholds)

2 (t )

s10

r2

ML detector
(Compare with M 1 thresholds)

Modul 6 - Siskom 2 - QAM dan FSK

13

Error probability

Coherent detection of M-QAM

M-QAM can be viewed as the combination of two M PAM
modulations on I and Q branches,
branches respectively.
respectively
No error occurs if no error is detected on either I and Q branches.
Hence:
C
Considering
id i th
the symmetry
t off th
the signal
i
l space and
d orthogonality
th
lit
of I and Q branches:
PE ( M ) = 1 PC ( M ) = 1 Pr(no error detected on I and Q branches)

Pr(no error detected on I and Q branches) = Pr(no error on I).Pr(no error on Q)

= Pr(no error on I) 2 = 1 PE

1 3 log 2 M Eb

PE ( M ) = 41
Q
M M 1 N 0

( M ))

Average probability of
symbol error for M PAM

Modul 6 - Siskom 2 - QAM dan FSK

14

Probability of symbol error for M-PAM

Note!

M = 2k

The same average symbol

energy
e
e gy for
o different
d e e t sizes
s es of
o
signal space

PE

Eb / N 0 dB
Modul 6 - Siskom 2 - QAM dan FSK

15

Probability of symbol error for M-QAM

Note!

M = 2k

The same average symbol

energy
e
e gy for
o different
d e e t sizes
s es of
o
signal space

PE

Eb / N 0 dB
Modul 6 - Siskom 2 - QAM dan FSK

16

Signal Space of several modulation

Modul 6 - Siskom 2 - QAM dan FSK

17

Binary FSK

s1 (t ) =
s2 (t ) =

2 Eb
cos(2f 0 t ) 0 t Tb
Tb
2 Eb
cos(2f1 t ) 0 t Tb
Tb

f0 and f1 are chosen such that

Tb

cos(2f t )cos(2f t )dt = 0 orthogonal

0

Modul 6 - Siskom 2 - QAM dan FSK

18

Coherent Detection
by Integrate
teg ate a
and
d Dump
u p / Matched
atc ed Filter
te Receiver
ece e

Coherent detection utilizes carrier phase information and requires inphase replica of the carrier at the receiver (explicitly or implicitly)
It is easy to show that these two techniques have the same
performance:

y (t )

h(t ) = s ( t )

v(T )

v(t ) = s ( t ) y (t )
= 0 s (t ) y ( ) d

y (t )

v(T )

ss(t t ))

v(t ) = 0 s (t ) y ( )d

Modul 6 - Siskom 2 - QAM dan FSK

19

Coherent Detection
y0(Tb)

2
cos
Tb

x(t)

(2 f 0 t )
y1(Tb)

2
cos (2 f 1 t )
Tb
When s1(t) is transmitted:

y 0 (T b ) E b + n

y (T ) =
n
1 b

When s2(t) is transmitted:

y 0 (T b )
y (T ) =
1 b

E b + n
n

20

Lets compare Error probability:

BPSK vs BFSK

BPSK and BFSK with coherent detection:

2 (t )

0
BPSK

s1 s 2 / 2

PB = Q
N /2
0

1 ((tt )

s1

s2

Eb

Eb 1 ((tt )

BFSK

s1

s1 s 2 = 2 Eb

s 21

s1 s 2 = 2 Eb

2 Eb
PB = Q
N0

Eb

2 (t )

Eb

Eb

PB = Q

N0
Modul 6 - Siskom 2 - QAM dan FSK

21

Non-coherent Detection

Base on filtering signal energy on allocated spectra and using

envelope detectors
Has performance degradation of about 1-3 dB when compared
to coherent detection (depending on Eb/N0)
Examples:

2-ASK

2-FSK

Modul 6 - Siskom 2 - QAM dan FSK

22

Error probability BFSK

Non-coherent detection of BFSK

Decision variable:
Diff
Difference
off envelopes
l
z = z1 z2

2 / T cos(( 1 t )

r11

( )2

z1 = r11 + r12
2

2 / T sin( 1 t )

r ((t )

2 / T cos( 2 t )

r12

r21

( )2
( )2

Decision rule:

if z (T ) > 0, m = 1
if z (T ) < 0, m = 0

2 / T sin( 2 t )

z 2 = r21 + r22
2

r22

( )2

Modul 6 - Siskom 2 - QAM dan FSK

23

Error probability BFSK contd

Non-coherent detection of BFSK

1
1
Pr(( z1 > z 2 | s 2 ) + Pr(( z 2 > z1 | s1 )
2
2
= Pr( z1 > z 2 | s 2 ) = E [Pr( z1 > z 2 | s 2 , z 2 )]

PB =

= Pr( z1 > z 2 | s 2 , z 2 ) p ( z 2 | s 2 )dz 2 =

PB =

E
1
exp b
2
2N0

p ( z | s )dz p ( z | s )dz
1
2
1
2
2
2
z2

Rayleigh
y g p
pdf

Rician p
pdf

Similarly, non-coherent detection of DBPSK

Eb
1

PB = exp
2
N0
Modul 6 - Siskom 2 - QAM dan FSK

24

Demodulation & Detection M-FSK

M-ary Frequency Shift Keying (M-FSK)

2 Es
2 Es
si (t ) =
cos(i t ) =
cos(c t + (i 1)t )
T
T

1
=
f =
2 2T
3 (t )

si (t ) = aij j (t ) i = 1, K , M

s3

j =1

2
i (t ) =
cos(i t )
T
Es = Ei = s i

Es i = j
aij =
i j
0

Es

s2

2 (t )

Es

s1
Es
1 (t )
Modul 6 - Siskom 2 - QAM dan FSK

25

Demodulation & Detection M-FSK

Coherent detection of M-FSK

1 ((tt )

z1

r (t )
M (t )

zM

z1
M

zM

ML detector:

=z

Modul 6 - Siskom 2 - QAM dan FSK

Choose
the largest element
in the observed vector

26

Error probability

Coherent detection of M-FSK

The dimensionality of signal space is M

M. An upper
bound for average symbol error probability can be
obtained by using union bound. Hence
Es

PE ( M ) (M 1)Q

N0

or, equivalently

PE ( M ) (M 1)Q

(log 2 M )Eb
N0

Modul 6 - Siskom 2 - QAM dan FSK

27

Bit error probability versus symbol

error probability
Number of bits per symbol k = log 2 M
For orthogonal M
M-ary
ary signaling (M
(M-FSK)
FSK)
PB
2 k 1
M /2
= k
=
PE 2 1 M 1
PB 1
lim
=
k P
2
E

For M-PSK, M-PAM and M-QAM

PE
PB
for PE << 1
k
Modul 6 - Siskom 2 - QAM dan FSK

28

Probability of symbol error for M-FSK

Note!

M = 2k

The same average symbol

energy
e
e gy for
o different
d e e t sizes
s es of
o
signal space

PE

Eb / N 0 dB
Modul 6 - Siskom 2 - QAM dan FSK

29

Power Spectral FSK

The FSK wave can be written as

s (t ) =

2 Eb
t
cos 2f c t
Tb
Tb

s (t ) =

t
2 Eb
cos cos (2f c t ) m
Tb
Tb

0 t Tb
t
2 Eb
sin sin (2f c t )
Tb
Tb

The in-phase
Th
i h
componentt is
i completely
l t l iindependent
d
d t off th
the
input binary wave. The power spectral density of this
component consists of two delta functions at f=+- 1/2Tb
Th quadrature
The
d t
componentt iis related
l t d tto the
th input
i
t binary
bi
wave and is given by
2Eb

0 t Tb
g (t ) = T b
0
otherwise

8 E b T b cos 2 ( T b f )
g ( f )=
2
2 4 T b2 f 2 1

Modul 6 - Siskom 2 - QAM dan FSK

30

Power Spectral FSK

E
PSD = b
2Tb

1
1 8 E bTb cos 2 (Tb f )
+
+ f +
f
2Tb
2Tb 2 4Tb2 f 2 1 2

Modul 6 - Siskom 2 - QAM dan FSK

31

Power Spectral M-FSK

M - ary FSK for h = 0.5 (freq deviation) BW log 2 M

Modul 6 - Siskom 2 - QAM dan FSK

0 .5
Tb

32

Bandwidth vs. Power Efficiency

(contoh M
Mary
ary PSK,
PSK FSK dan M-ary QAM pada BER = 10-66 )
M-ary PSK

16

32

Rb / BW null

05
0,5

15
1,5

25
2,5

Eb / o

10,5

10,5

14

18,5

23,4

M-ary FSK

16

32

Rb / BW null

0,4

0,57

0,55

0,42

0,29

Eb / o

13 5
13,5

10 8
10,8

93
9,3

82
8,2

75
7,5

M-ary QAM

16

64

256

1024

Rb / BW nullll

Eb / o

10,5

15

18,5

24

28

Modul 6 - Siskom 2 - QAM dan FSK

33

GMSK, varian of Frequency Shift

Keying (FSK)
Continuous Phase FSK (CPFSK)
digital data encoded in the frequency shift
typically implemented with frequency
modulator to maintain continuous phase
s(t) = A cos [ct + 2 kf d() d]
t

nonlinear modulation but constant-envelope

Minimum Shift Keying

y g ((MSK))
minimum bandwidth, sidelobes large
can be implemented using I-Q receiver

Gaussian Minimum Shift Keying (GMSK)

reduces sidelobes of MSK using a premodulation filter
Data CDPD,
CDPD and HIPERLAN
used by RAM Mobile Data,
Modul 6 - Siskom 2 - QAM dan FSK

34

Selamat Idul Fitri 1428 H

MOHON MAAF LAHIR DAN BATIN