Modln Techniques

Digital Modulation Formats:[1, 2, 3, 4]
Manjunatha. P
manjup.jnnce@gmail.com
Professor
Dept. of ECE
J.N.N. College of Engineering, Shimoga
January 1, 2016
Overview
Digital Modulation Formats

1
2
3
4
5
6
7
8
9
10
11
Coherent Binary Modulation Techniques

Coherent Quadrature Modulation Techniques
NonCoherent Binary Modulation Techniques
Comparison of Binary and Quaternary Modulation Techniques
M-ary Modulation Techniques
Power Spectra, Bandwidth Efficiency
M-ary Modulation formats viewed in the Light of the channel capacity
Theorem
Effect of Intersymbol Interference
Bit Versus Symbol Error Probabilities
Synchronisation
Applications
Manjunatha. P
(JNNCE)
January 1, 2016
2 / 118
Manjunatha. P
(JNNCE)
January 1, 2016
3 / 118
Trigonometric identities
Manjunatha. P
cos(x)cos(y )
1/2[cos(x y ) + cos(x + y )]
sin(x)sin(y )
1/2[cos(x y ) cos(x + y )]
sin(x)cos(y )
1/2[sin(x + y ) + sin(x y )]
cos(x)sin(y )
1/2[sin(x + y ) sin(x y )]
(JNNCE)
sin(x y )
sin x cos y cos x sin y
cos(x y )
cos x cos y sin x sin y
January 1, 2016
4 / 118
Modulation is the process by which an information signal is converted to a sinusoidal waveform by

varying the amplitude, frequency,or phase or a combination of them of an RF carrier.
The general form the the carrier wave is
s(t) = A(t) cos(t)
(1)
where A(t) is the time varying amplitude and (t) is the time varying angle.
s(t) = A(t) cos[t + (t)]
(2)
where is the angular frequency of the carrier and (t) is the phase. The frequency f in in hertz and
is in radians per second and are related by = 2f .
The carrier wave amplitude coefficient
The general form the the carrier wave is
s(t) = A cost
(3)
where A is peak value of the waveform. The peak value of the sinusoidal waveform equals
root-mean square(rms) value. Hence
q
s(t) = 2Arms cost = 2A2rms cost

p
2 times the
2A2rms represents the average power P normalized to 1. Therefore s(t)is
s(t) = 2P coswt
(4)
(5)
Replacing P watts by E joules/T seconds i.e., P = E /T

r
2E
s(t) =
cost
T
(6)
The energy of a received signal is the key parameter in determining the error performance of the
detection process, hence it is often more convenient to use the amplitude notation because it facilitates
solving the probability of error Pe as a function of signal energy.
Manjunatha. P
(JNNCE)
January 1, 2016
5 / 118
Noise in Communication Systems
Noise in Communication Systems
The term noise refers to unwanted electrical signals that are always present in electrical systems.
The noise arises from a variety of sources, both man made and natural.
The man made noise includes such sources as spark-plug ignition noise, switching transients, and other
radiating electromagnetic signals.
Natural noise includes such elements as the atmosphere, the sun and other galactic sources,thermal
noise or Johnson noise.
Noise can be eliminated through filtering, shielding, the choice of modulation and the selection of an
optimum receiver.
Thermal noise is caused by the thermal motion of electrons in all dissipative components like resistors,
wires and so on.
Thermal noise cannot be eliminated due to the same electrons are responsible for electrical conduction.
The thermal noise can be described by a zero mean Gaussian random process.
A Gaussian process n(t) is a random function whose value n at any arbitrary time t is statically
characterized by the Gaussian probability density function.
Manjunatha. P
(JNNCE)
January 1, 2016
6 / 118
Gaussian, Normal distribution

In probability theory, the normal (or Gaussian) distribution is a continuous probability distribution,
defined by the formula:
(x)2
1
2 2
f (x) = e
2
The parameter is the mean or expectation of the distribution (and also its median and mode). The
parameter is its standard deviation; its variance is therefore 2 .
A random variable with a Gaussian distribution is said to be normally distributed and is called a normal
deviate.
If = 0 and = 1, the distribution is called the standard normal distribution or the unit normal
distribution, and a random variable with that distribution is a standard normal deviate.
(x)2
1
2
f (x) = e
2
Many things closely follow a Normal Distribution:
Heights of people
Size of things produced by machines
Errors in measurements
Blood pressure
Marks on a test
Manjunatha. P
(JNNCE)
January 1, 2016
7 / 118
Figure 1:
Gaussian Distribution
Figure 2:
Manjunatha. P
(JNNCE)
Gaussian Distribution
January 1, 2016
8 / 118
White Noise
White Noise
In signal processing, white noise is a random signal with a flat (constant) power spectral density (PSD).
White noise draws its name from white light, which is commonly (but incorrectly) assumed to have a
flat spectral power density over the visible band.
The thermal noise is that its PSD is same for all frequencies from dc to 1012 Hz.
The thermal noise is assumed that its PSD Gn (f )
Gn (f ) = N0 /2
watts/hertz
(7)
where the factor 2 is included to indicate that Gn (f ) is two sided PSD.

The autocorrelation function of white noise is given by the inverse Fourier transform of the PSD
Rn ( ) = F {Gn (f )} =
N0
( )
2
(8)
The autocorrelation of white noise is a delta function weighted by the factor N0 /2 and occurring at
= 0 as shown in Figure 3.
Rn ( ) = 0 for 6= 0 that is any two different samples of white noise no matter how close together in
time they are uncorrelated.
Figure 3:
Manjunatha. P
(JNNCE)
Power spectral density and Autocorrelation of white noise.
January 1, 2016
9 / 118
Binary Phase Shift Keying (BPSK)
Manjunatha. P
(JNNCE)
January 1, 2016
10 / 118
The carrier is shifted in phase according to the input data stream, frequency and amplitude of the
carrier are constant.
Binary PSK (BPSK): two phases (0o 180o ) represent two binary digits:
Figure 4:
BPSK waveform
In binary PSK (BPSK) symbols 1 and 0 are represented by S1 (t) S2 (t) and are defined as
s
s1 (t) =
s
2Eb
2Eb
0
cos[2fc t + 0 ]=
cos[2fc t]
Tb
Tb
0 t Tb
s
2Eb
2Eb
0
cos[2fc t + 180 ] =
cos[2fc t]
0 t Tb
Tb
Tb
s
2Eb
0
0
cos[2fc t + (t)] i = 1, 2 (t) = 0 or 180
si (t) =
Tb
s2 (t) =
where Eb is the transmitted signal energy per bit.

Eb =
Manjunatha. P
(JNNCE)
A 2 Tb
2
A=
2Eb
Tb
January 1, 2016
11 / 118
There is a single basis function 1 (t) and is given by:

s
2
cos(2fc t)
1 (t) =
Tb
p
s1 (t) = Eb 1 (t)
p
s2 (t) = Eb 1 (t)
0 t Tb
Binary PSK
wave
Eb or Eb
1 ( t ) =
Figure 5:
(JNNCE)
0 t Tb
Product
Modulator
Binrary Wave
Manjunatha. P
0 t Tb
2
cos ( 2 f ct )
Tb
BPSK Tranmitter
January 1, 2016
12 / 118
BPSK system is having a signal space with one dimension and having two message points that have the
coordinate points as follows:
TB
Z
s11 =
TB
s21 =
TB
Z
s1 (t)1 (t)dt =
TB
Z
s2 (t)1 (t)dt =
2Eb
cos(2fc t)
Tb
cos x =
2Eb
cos(2fc t)
Tb
p
2
cos(2fc t)dt = Eb
Tb
p
2
cos 2fc t)dt = Eb
Tb
1
(1 + cos2x)
2
Decision
Boundary
Region Z1
Region Z2
Eb
Eb
1 ( t)
Message
Point 1
Message
Point 2
Figure 6: Signal Space diagram

Figure 7: Signal Space diagram
The message s1 (t) is located at s11 = + Eb s2 (t) is at s21 = Eb .

The signal space has two regions z1 and z2 corresponding to message 1 and 0.
The distance between message points is 2 Eb .

Manjunatha. P
(JNNCE)
January 1, 2016
13 / 118
Error Performance for Digital Modulation
Probability of Bit Error

An important measure of performance used for comparing digital modulation schemes is the probability
of error Pe .
The probability of the detector making an incorrect decision is termed the probability of symbol error Pe .
It is often convenient to specify system performance by the probability of bit error Pb even when
decisions are made on the basis of symbols for which M > 2.
Manjunatha. P
(JNNCE)
January 1, 2016
14 / 118
Probability of Error
Channel
Transmitted
Signal
si(t)
Received
Signal
x(t)=si(t)+w(t)
P(t )
x (t )
Sample at
every Tb Sample at
every Tb
t
Noise
w(t)
Figure 8: Receiver Model
x1 < 0 0
Binary
Data
dt
2
cos(2fc t )
Tb
1 (t ) =
x1 > 0 1
Vth=0
Figure 9: BPSK Receiver
The received signal x(t) in the presence of AWGN w(t)with the assumption that symbol 0 or s2 (t) is
transmitted
x(t) = s2 (t) + w (t)
0 t Tb
Z Tb
Z Tb
x1 =
x(t)1 (t)dt =
[s2 (t) + w (t)]1 (t)dt
0
0
Tb
Z
=
Tb
Z
s2 (t)1 (t)dt +
w (t)1 (t)dt
0
x1 = s21 + w1
p
s21 = Eb
p
x1 = Eb + w1
w1 is the sample value of random variable W1 having Gaussian distribution with mean zero and variance
N0 /2 The expected value of the random variable X1
Manjunatha. P
(JNNCE)
January 1, 2016
15 / 118
The expectation (mean) of the random variable X1

p
p
p
E [X1 ] = E [ Eb + w1 ] = Eb + 0 = Eb
The variance of the random variable X1

Var [X1 ] = Var [
p
N0
N0
E b + w1 ] = 0 +
=
2
2
Conditional pdf of random variable X1 given that symbol 0 is transmitted is given by

"
#
(x)2
1
(x1 + Eb )2
1
2 2
fX1 (x1 |0) =
exp
e
f (x) =
N0
N0
2 2
When symbol 0 is transmitted, an error will occur, if x1 > 0 in which case a decision is made in favor of symbol
1 Pe (0) = P(x1 > 0|symbol 0 is transmitted) Pe (0) can be computed by integrating conditional pdf fX1 (x1 |0)
Z
Pe (0) =
0
fX1 (x1 |0)dx1 =
N0 du
Manjunatha. P
N0
"
exp
0
(x1 + Eb )2
dx1
N0
(x1 + Eb )2
dx1
N0
N0 0
(x1 + Eb )
Let u =
N0
p
Lower limits when x = 0 u = Eb /N0 and higher limit is
Z
h
i
1
2
Pe (0) =
exp u du
Eb /N0
Pe (0) =
dx1 =
(JNNCE)
"
exp
January 1, 2016
16 / 118
1
Pe (0) =
2
erf (u) =
h
i
2
exp u du
Eb /N0
u
exp(x )dx
0
Z
2
2
erfc(u) = 1 erf (u) =
exp(x )dx
u
s
Eb
1
Pe (0) = erfc
2
N0
Similarly when symbol 1 is transmitted, if error occurs, then a decision is made in favor of symbol 0. The
probability of error when symbol 1 is transmitted is Pe (1) = P(x1 < 0|symbol 1 is transmitted. Pe (0) can be
computed by integrating conditional pdf fX1 (x1 |0)
s
1
Eb
Pe (1) = erfc
2
N0
Symbols 0 and 1 are equiprobable, that is Pe (0) = Pe (1) = 1/2 then the average probability of symbol error is
Pe =
1
1
Eb
[Pe (0) + Pe (1)] = erfc

2
2
N0
s
1
Eb
Pe = erfc
2
N0
Manjunatha. P
(JNNCE)
January 1, 2016
17 / 118
Error Function Table
Complementary Error Function Table

x
erfc(x)
1.000000
0.5
0.479500
0.157299
1.5
0.01
0.988717
0.51
0.470756
1.01
0.153190
1.51
0.02
0.52
0.462101
erfc(x)
1.02
0.149162
1.52
erfc(x)
0.033895
x
2
0.032723 2.01
0.031587 2.02
erfc(x)
0.004678
2.5
0.000407
0.00002209
0.004475 2.51
0.000386
3.01
0.00002074
0.004281 2.52
erfc(x)
0.000365
erfc(x)
3.02
0.00001947
0.03
0.966159
0.53
0.453536
1.03
0.145216
1.53
0.030484 2.03
0.004094 2.53
0.000346
3.03
0.00001827
0.04
0.954889
0.54
0.445061
1.04
0.141350
1.54
0.029414 2.04
0.003914 2.54
0.000328
3.04
0.00001714
0.05
0.943628
0.55
0.436677
1.05
0.137564
1.55
0.028377 2.05
0.003742 2.55
0.000311
3.05
0.00001608
0.06
0.932378
0.56
0.428384
1.06
0.133856
1.56
0.027372 2.06
0.003577 2.56
0.000294
3.06
0.00001508
0.07
0.921142
0.57
0.420184
1.07
0.130227
1.57
0.026397 2.07
0.003418 2.57
0.000278
3.07
0.00001414
0.08
0.909922
0.58
0.412077
1.08
0.126674
1.58
0.025453 2.08
0.003266 2.58
0.000264
3.08
0.09
0.898719
0.59
0.404064
1.09
0.123197
1.59
0.024538 2.09
0.003120 2.59
0.000249
3.09
0.00001243
0.1
0.887537
0.6
0.396144
1.1
0.119795
1.6
0.023652
0.002979
2.6
0.000236
3.1
0.00001165
0.11
0.876377
0.61
0.388319
1.11
0.116467
1.61
0.002845 2.61
0.000223
3.11
0.00001092
0.12
0.865242
0.000211
3.12
0.00001023
0.854133
0.63
0.372954
1.13
0.110029
1.63
0.021157 2.13
0.002593 2.63
0.000200
3.13
0.00000958
0.843053
0.64
0.365414
1.14
0.106918
1.64
0.020378 2.14
0.002475 2.64
0.000189
3.14
0.00000897
0.65
0.380589
0.357971
1.12
1.15
0.113212
0.103876
1.62
0.022793 2.11
1.65
0.021962 2.12
0.019624 2.15
0.002716 2.62
0.00001326
0.13
0.832004
0.62
2.1
0.14
0.15
0.002361 2.65
0.000178
3.15
0.00000840
0.16
0.820988
0.66
0.350623
1.16
0.100904
1.66
0.018895 2.16
0.002253 2.66
0.000169
3.16
0.00000786
0.17
0.810008
0.67
0.343372
1.17
0.098000
1.67
0.018190 2.17
0.002149 2.67
0.000159
3.17
0.00000736
0.18
0.799064
0.68
0.336218
1.18
0.095163
1.68
0.017507 2.18
0.002049 2.68
0.000151
3.18
0.00000689
0.19
0.788160
0.69
0.329160
1.19
0.092392
1.69
0.016847 2.19
0.001954 2.69
0.000142
3.19
0.00000644
0.2
0.777297
0.7
0.322199
1.2
0.089686
1.7
0.016210
0.001863
0.000134
3.2
0.00000603
0.21
0.766478
0.001776 2.71
0.000127
0.755704
0.72
0.308567
1.22
0.084466
1.72
0.014997 2.22
0.001692 2.72
0.000120
3.22
0.00000527
0.744977
0.73
0.301896
1.23
0.081950
1.73
0.014422 2.23
0.001612 2.73
0.000113
3.23
0.00000493
0.000107
3.24
0.00000460
0.288845
1.25
0.077100
1.75
0.013328 2.25
0.001463 2.75
0.000101
3.25
0.00000430
0.282463
1.26
0.074764
1.76
0.012810 2.26
0.001393 2.76
0.000095
3.26
0.00000402
1.77
0.012309 2.27
0.001536 2.74
0.00000564
0.75
0.072486
0.013865 2.24
3.21
0.76
1.27
1.74
0.015593 2.21
0.723674
0.276179
0.079495
1.71
0.713100
0.77
1.24
0.087045
0.25
0.702582
0.295322
1.21
0.26
0.27
0.74
0.315335
2.7
0.22
0.734300
0.71
2.2
0.23
0.24
0.001326 2.77
0.000090
3.27
0.00000376
0.28
0.692120
0.78
0.269990
1.28
0.070266
1.78
0.011826 2.28
0.001262 2.78
0.000084
3.28
0.00000351
0.29
0.681717
0.79
0.263897
1.29
0.068101
1.79
0.011359 2.29
0.001201 2.79
0.000080
3.29
0.00000328
0.3
0.671373
0.8
0.257899
1.3
0.065992
1.8
0.010909
2.3
0.001143
2.8
0.000075
3.3
0.00000306
0.31
0.661092
0.81
0.251997
1.31
0.063937
1.81
0.010475 2.31
0.001088 2.81
0.000071
3.31
0.00000285
0.32
0.650874
0.82
0.246189
1.32
0.061935
1.82
0.010057 2.32
0.001034 2.82
0.000067
3.32
0.00000266
0.33
0.640721
0.83
0.240476
1.33
0.059985
1.83
0.009653 2.33
0.000984 2.83
0.000063
3.33
0.00000249
0.34
0.630635
0.84
0.234857
1.34
0.058086
1.84
0.009264 2.34
0.000935 2.84
0.000059
3.34
0.00000232
0.35
0.620618
0.85
0.229332
1.35
0.056238
1.85
0.008889 2.35
0.000889 2.85
0.000056
3.35
0.00000216
0.36
0.610670
0.86
0.223900
1.36
0.054439
1.86
0.008528 2.36
0.000845 2.86
0.000052
3.36
0.00000202
0.37
0.600794
0.87
0.218560
1.37
0.052688
1.87
0.008179 2.37
0.000803 2.87
0.000049
3.37
0.00000188
0.38
0.590991
0.88
0.213313
1.38
0.050984
1.88
0.007844 2.38
0.000763 2.88
0.000046
3.38
0.00000175
0.39
0.581261
0.89
0.208157
1.39
0.049327
1.89
0.007521 2.39
0.000725 2.89
0.000044
3.39
0.00000163
0.4
0.571608
0.9
0.203092
1.4
0.047715
1.9
0.007210
0.000689
0.000041
3.4
0.00000152
0.41
Manjunatha. P
0.977435
erfc(x)
0.91
0.198117
1.41
0.046148
1.91
0.006910 2.41
2.9
0.000039
3.41
0.00000142
0.42
0.552532
0.92
0.193232
1.42
0.044624
1.92
0.006622 2.42
0.000621 2.92
0.000036
3.42
0.00000132
0.43
0.543113
0.93
0.188437
1.43
0.043143
1.93
0.006344 2.43
0.000589 2.93
0.000034
3.43
0.00000123
0.533775
0.94
0.183729
0.000032
3.44
0.00000115
(JNNCE)
0.44
0.562031
2.4
0.000654 2.91
Digital
Modulation
Formats:[1,
2, 3, 4] 2.94
1.44
0.041703
1.94 0.006077
2.44 0.000559
January 1, 2016
18 / 118

2
erf(x) =
x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0
3.1
3.2
0
0.00000
0.11246
0.22270
0.32863
0.42839
0.52050
0.60386
0.67780
0.74210
0.79691
0.84270
0.88021
0.91031
0.93401
0.95229
0.96611
0.97635
0.98379
0.98909
0.99279
0.99532
0.99702
0.99814
0.99886
0.99931
0.99959
0.99976
0.99987
0.99992
0.99996
0.99998
0.99999
0.99999
1
0.01128
0.12362
0.23352
0.33891
0.43797
0.52924
0.61168
0.68467
0.74800
0.80188
0.84681
0.88353
0.91296
0.93606
0.95385
0.96728
0.97721
0.98441
0.98952
0.99309
0.99552
0.99715
0.99822
0.99891
0.99935
0.99961
0.99978
0.99987
0.99993
0.99996
0.99998
0.99999
0.99999
2
0.02256
0.13476
0.24430
0.34913
0.44747
0.53790
0.61941
0.69143
0.75381
0.80677
0.85084
0.88679
0.91553
0.93807
0.95538
0.96841
0.97804
0.98500
0.98994
0.99338
0.99572
0.99728
0.99831
0.99897
0.99938
0.99963
0.99979
0.99988
0.99993
0.99996
0.99998
0.99999
0.99999
3
0.03384
0.14587
0.25502
0.35928
0.45689
0.54646
0.62705
0.69810
0.75952
0.81156
0.85478
0.88997
0.91805
0.94002
0.95686
0.96952
0.97884
0.98558
0.99035
0.99366
0.99591
0.99741
0.99839
0.99902
0.99941
0.99965
0.99980
0.99989
0.99994
0.99997
0.99998
0.99999
1.00000
et dt
2
Hundredths digit of x
4
5
0.04511 0.05637
0.15695 0.16800
0.26570 0.27633
0.36936 0.37938
0.46623 0.47548
0.55494 0.56332
0.63459 0.64203
0.70468 0.71116
0.76514 0.77067
0.81627 0.82089
0.85865 0.86244
0.89308 0.89612
0.92051 0.92290
0.94191 0.94376
0.95830 0.95970
0.97059 0.97162
0.97962 0.98038
0.98613 0.98667
0.99074 0.99111
0.99392 0.99418
0.99609 0.99626
0.99753 0.99764
0.99846 0.99854
0.99906 0.99911
0.99944 0.99947
0.99967 0.99969
0.99981 0.99982
0.99989 0.99990
0.99994 0.99994
0.99997 0.99997
0.99998 0.99998
0.99999 0.99999
1.00000 1.00000
6
0.06762
0.17901
0.28690
0.38933
0.48466
0.57162
0.64938
0.71754
0.77610
0.82542
0.86614
0.89910
0.92524
0.94556
0.96105
0.97263
0.98110
0.98719
0.99147
0.99443
0.99642
0.99775
0.99861
0.99915
0.99950
0.99971
0.99983
0.99991
0.99995
0.99997
0.99998
0.99999
1.00000
7
0.07886
0.18999
0.29742
0.39921
0.49375
0.57982
0.65663
0.72382
0.78144
0.82987
0.86977
0.90200
0.92751
0.94731
0.96237
0.97360
0.98181
0.98769
0.99182
0.99466
0.99658
0.99785
0.99867
0.99920
0.99952
0.99972
0.99984
0.99991
0.99995
0.99997
0.99999
0.99999
1.00000
8
0.09008
0.20094
0.30788
0.40901
0.50275
0.58792
0.66378
0.73001
0.78669
0.83423
0.87333
0.90484
0.92973
0.94902
0.96365
0.97455
0.98249
0.98817
0.99216
0.99489
0.99673
0.99795
0.99874
0.99924
0.99955
0.99974
0.99985
0.99992
0.99995
0.99997
0.99999
0.99999
1.00000
ECE 3610 Engineering Probability & Statistics
Manjunatha. P
(JNNCE)
9
0.10128
0.21184
0.31828
0.41874
0.51167
0.59594
0.67084
0.73610
0.79184
0.83851
0.87680
0.90761
0.93190
0.95067
0.96490
0.97546
0.98315
0.98864
0.99248
0.99511
0.99688
0.99805
0.99880
0.99928
0.99957
0.99975
0.99986
0.99992
0.99996
0.99998
0.99999
0.99999
1.00000
3-1
January 1, 2016
19 / 118
Frequency Shift Keying (FSK)
Manjunatha. P
(JNNCE)
January 1, 2016
20 / 118
The carrier frequency is shifted in according to the input data stream, phase and amplitude of the
carrier are constant.
s
si (t) =
2Eb
cos[2fi t]
Tb
0 t Tb
where i=1,2 and Eb is the transmitted signal energy per bit, and the transmitted frequency fi = nTc +i for
b
some fixed integer nc and i=1,2.
In binary FSK (BFSK) two different carrier frequencies (f1 f2 ) are used and symbols 1 and 0 are
represented by S1 (t) S2 (t) and are defined as
s
s
2Eb
2Eb
s1 (t) =
cos[2f1 t]
0 t Tb
s2 (t) =
cos[2f2 t]
0 t Tb
Tb
Tb
There is set of orthonormal basis function 1 (t) and 2 (t) and is given by:
s
2
i (t) =
cos(2fi t)
0 t Tb
Tb
s1 (t) =
Manjunatha. P
(JNNCE)
Eb 1 (t)
0 t Tb
s2 (t) =
Eb 2 (t)
i = 1, 2
0 t Tb
January 1, 2016
21 / 118
BFSK system is having a signal space with two dimension and having two message points represented by
coordinate points as follows
s
Z T
Z T s
p
B
B
2Eb
2
s11 =
s1 (t)1 (t)dt =
cos(2f1 t)
cos(2f1 t)dt = Eb
T
T
b
b
0
0
and
TB
Z
s12 =
TB
Z
s1 (t)2 (t)dt =
2Eb
cos(2f1 t)
Tb

Eb
S1 =
0
2
cos(2f2 t)dt = 0
Tb
2 ( t )
Similarly the coefficients of s2 (t) are s21 and s22
Z
s21 =
TB
s2 (t)1 (t)dt = 0
and
TB
Z
s22 =
Region Z2
Message
Point S2
( 0, 0 )
s2 (t)2 (t)dt =
Eb
The dist. betwn., 2 nodes=
Eb
Decision
Boundary
Eb , 0
Message
Point S1
S2 =
( 0,
1 ( t)
Region Z1
Eb
Eb +
Eb = 2Eb
Figure 10:
Signal Space diagram for
BPSK system
Manjunatha. P
(JNNCE)
January 1, 2016
22 / 118
m(t )
Tb
+
Binary wave
1 ( t ) =
m(t )
2
cos ( 2 f1t )
Tb
Binary FSK
wave
Received
Signal
x(t )
2
cos ( 2 f1t )
Tb
Decision
Device
m(t )
Choose 1 if l f 0
Choose 0 if l p 0
Tb
Figure 11:
x1 (t )
+
1 ( t ) =
Inverter
2 ( t ) =
dt
2 ( t ) =
2
cos ( 2 f 2t )
Tb
x2 (t )
2
cos ( 2 f 2t )
Tb
Figure 12:
BFSK Transmitter model
dt
BFSK Receiver model
The observation vector has two elements x1 and x2 (message points)which are defined by
TB
Z
x1 =
x(t)1 (t)dt
0
TB
Z
x2 =
x(t)2 (t)dt
where x(t) is the received signal and when symbol 1 is transmitted

x(t) = s1 (t) + w (t)
and when symbol 0 is transmitted
x(t) = s2 (t) + w (t)
where w (t) is the sample function of white Gaussian noise process of zero mean and variance (PSD) is N0 /2
Manjunatha. P
(JNNCE)
January 1, 2016
23 / 118
Define a new Gaussian random variable L whose sample value l is

l = x1 x2
The mean value of the random variable L depends on which binary symbol was transmitted. When symbol 1
was transmitted
the Gaussian random variables X1 and X2 and whose sample values x1 and x2 have mean
equal to Eb and zero. The conditional mean value of the random variable L given that symbol 1 was
transmitted is given by
p
E [L|1] = E [X1 |1] E [X2 |1] = Eb
Similarly the conditional mean value of the random variable L given that symbol 0 was transmitted is given by
p
E [L|0] = E [X1 |0] E [X2 |0] = Eb =
Random variables are statistically independent, each with a variance equal to N0 /2

Var [L] = Var [X1 ] + Var [X2 ] = N0 /2 + N0 /2 = N0
Suppose when symbol 0 was transmitted, then the conditional value of the conditional probability density
function of the random variable L equals
"
#
(x)2
1
(l + Eb )2
1
2 2
f (x) =
e
fL (l|0) =
exp
2N0
2N0
2 2
Since the condition x1 > x2 or l > 0 then the receiver is making decision in favor of symbol 1 the conditional
probability of error, given that symbol 0 was transmitted is given by,
Z
Pe (0) = Pe (l > 0| symbol 0 was sent) =
fL (l|0)dl
0
Pe (0) =
Manjunatha. P
(JNNCE)
1
2N0
Z
0
(l + Eb )2
exp
dl
2N0
"
January 1, 2016
24 / 118
Pe (0) =
1
2N0
"
exp
0
Let
(l + Eb )2
dl
2N0
Eb )
=z
2N0
p
dl = 2N0 dz
Z
h
i
1
2
r
Pe (0) =
exp z dz
E
b
(l +
2N0
Eb
1
Pe (0) = erfc
2
2N0
Similarly conditional probability of error Pe (1), when symbol 1 was transmitted has the same value of Pe (0)
i.e.,
s
1
Eb
Pe (1) = erfc
2
2N0
The average probability of symbol error Pe for coherent binary FSK is
s
1
1
1
Eb
Pe = Pe (0) + Pe (1) = erfc

2
2
2
2N0
In order to maintain the same average errorrate as in coherent BPSK system, the BFSK system should have
twice the bit energy to noisedensity ratio, Eb /2N0 . In coherent BPSK
system the distance between the two
message points is equal to 2 Eb whereas in coherent BPSK system it is 2Eb . Larger the distance between
the message points smaller is the average probability of error Pe .
Manjunatha. P
(JNNCE)
January 1, 2016
25 / 118
Quadriphase Shift Keying (QPSK)
Manjunatha. P
(JNNCE)
January 1, 2016
26 / 118
The important goal in design of a digital communication system is to provide low probability of error and
efficient utilization of channel bandwidth.
In QPSK, the phase of the carrier takes on one of four equally spaced values, such as /4, 3/4 5/4 7/4
as given below
r
si (t) =

2E
cos 2fc t + (2i 1)

T
4
0tT
where i=1,2,3,4, E is the transmitted signal energy per symbol, T is symbol duration, and the carrier frequency
fc equals nc /T for some fixed integer nc .
r
si (t) =
r

2E
2E
cos (2i 1)
cos 2fc t
sin (2i 1)
sin 2fc t
T
4
T
4
b1 (t )
There are two orthonormal basis functions

contained in the si (t)
+
Binary wave
b(t )
r
1 (t) =
r
2 (t) =
Serial to
parallel
converter
0tT
2
sin(2fc t)
T
0tT
QPSK
wave
b2 (t )
2 ( t ) =
Figure 13:
(JNNCE)
2
cos ( 2 f ct )
T
2
cos(2fc t)
T
Manjunatha. P
1 ( t ) =
2
sin ( 2 f c t )
T
QPSK Transmitter
January 1, 2016
27 / 118
Constellation Diagram
A constellation diagram helps us to define the

amplitude and phase of a signal when we are
using two carriers, one in quadrature of the
other.
The X-axis represents the in-phase carrier and
the Y-axis represents quadrature phase carrier.
Figure 14:
Figure 15:
Manjunatha. P
(JNNCE)
January 1, 2016
28 / 118
2 ( t )
Decision
Boundary
There are four message points

Si =

E cos (2i 1) 4
E sin (2i 1) 4
Region Z3
i = 1, 2, 3, 4
Region Z4
Message
Point 3
01
E /2
Message
Point 4
11
Decision
Boundary
Input dibit
0tT
Phase of
QPSK signal
10
00
01
11
/4
3/4
5/4
7/4
Coordinates of
message points
si1
si2
p
p
+ pE /2
pE /2
pE /2
p E /2
p E /2
+pE /2
+ E /2
+ E /2
E /2
Message
Point 2
00
Region Z2
Figure 16:
Manjunatha. P
(JNNCE)
1 ( t)
E/ 2
E/2 Message
Point 1
10
Region Z1
QPSK-signalspace
January 1, 2016
29 / 118
QPSK
The received signal is defined by

In phase channel
x(t) = si (t) + w (t)
0tT
Tb
i = 1, 2, 3, 4
The observation vector x of a coherent QPSK receiver
has 2 elements x1 and x2 and are defined by
Received
Signal
1 ( t ) =
x(t )
x(t)1 (t)dt =
0

E cos (2i 1)
+ w1
4
2 ( t ) =
Decision
Device
2
sin ( 2 f c t )
T
Output
Binary
wave
Multiplexer
Tb
x1
2
cos ( 2 f ct )
T
x1 =
dt
x2
dt
Decision
Device
Quadrature phase channel

Figure 17: QPSK Receiver model
+ w2
x(t)2 (t)dt = E sin (2i 1)
4
0
Suppose signal s4 (t) was transmitted, then the receiver will make correct decision if the observation vector x
lies inside the region Z4 . The probability of a correct decision Pc , when signal s4 (t) is transmitted is given by
Z
x2 =
Pc (s4 (t)) = P(x1 > 0 and x2 > 0)

The probability of P(x1 > 0 and x2 > 0)
fx1 (x1 |s4 (t)) transmitted) =
fx2 (x2 |s4 (t)) transmitted) =
Manjunatha. P
(JNNCE)
1
N0
1
N0
"
exp
"
exp
(x1
E /2)2
E /2)2
N0
(x2
p
N0
January 1, 2016
30 / 118
QPSK
Pc =
0
1
N0
exp

2
2
p
p
Z
x1 E /2
x2 E /2
1
exp
dx1 .
dx2
N0
N0
N0
0

2
p
p
x1 E /2
x2 E /2
=
=z
N0
N0
#2
"
Z

1
2
exp z dz
Pc =
E /2N0

From the definition of complementary error function

1
E /2N0

E
2
exp z dz = 1 erfc
2N0
2
s
1
E
Pc = 1 erfc
2
2N0
s
= 1 erfc
Manjunatha. P
(JNNCE)
E
1
E
2
+ erfc
2N0
4
2N0
January 1, 2016
31 / 118
QPSK
The average probability of symbol error for coherent QPSK

s
Pe = 1 Pc = erfc
1
E
E
erfc 2
2N0
4
2N0
By ignoring the second term probability of symbol error of QPSK

s
Pe = erfc
2N0
Signal points s1 ,s2 ,s3 and s4 are symmetrically located in the two dimensional signal space diagram, then
Pe (s4 (t)) = Pe (s3 (t)) = Pe (s2 (t)) = Pe (s1 (t))
(9)
The probability of occurrence of four messages is an equiprobable, then

Pe =
1
[Pe (s4 (t)) + Pe (s3 (t)) + Pe (s2 (t)) + Pe (s1 (t))]
4
(10)
In QPSK there are two bits per symbol E = 2Eb

s
Pe = erfc
Manjunatha. P
(JNNCE)
Eb
N0
January 1, 2016
32 / 118
QPSK
Figure 18:
Manjunatha. P
(JNNCE)
QPSK Waveform
January 1, 2016
33 / 118
QPSK
b1 (t )
+
Binary wave
b(t )
Serial to
parallel
converter
1 ( t ) =
2
cos ( 2 f ct )
T
QPSK
wave
+
b2 (t )
2 ( t ) =
Manjunatha. P
(JNNCE)
2
sin ( 2 f c t )
T
Figure 19:
QPSK Transmitter model
Figure 20:
January 1, 2016
34 / 118
QPSK
Figure 21:
Manjunatha. P
(JNNCE)
QPSK input bits
January 1, 2016
35 / 118
QPSK
Figure 22:
Manjunatha. P
(JNNCE)
January 1, 2016
36 / 118
QPSK
In phase channel
Tb
Received
Signal
1 ( t ) =
x(t )
(JNNCE)
Decision
Device
2
sin ( 2 f c t )
T
Output
Binary
wave
Multiplexer
Tb
Manjunatha. P
x1
2
cos ( 2 f ct )
T
2 ( t ) =
dt
x2
dt
Decision
Device
Quadrature phase channel
Figure 23:
QPSK Receiver model
Figure 24:
QPSK Receiver model
January 1, 2016
37 / 118
Minimum Shift Keying (MSK)
Manjunatha. P
(JNNCE)
January 1, 2016
38 / 118
The performance of the FSK receiver is improved by the proper utilization of phase. Continuous-phase
frequency-shift-keying (CPFSK) signal is defined as follows:
q
2Eb
cos[2f1 t + (0)]
q Tb
s(t) =
2Eb
T cos[2f2 t + (0)]
for symbol 1
for symbol 2
Eb is the transmitted signal energy per bit and Tb is the bit duration, phase (0) is the value of the phase at
time t = 0, depends on the past history of the modulation process. Frequencies f1 and f2 represent the symbol
1 and 0 respectively.
The conventional form of an angle modulated wave as follows:
s
s(t) =
2Eb
cos[2fc t + (t)]
Tb
The nominal frequency fc is chosen as the arithmetic mean of the two frequencies f1 and f2
fc =
1
(f1 + f2 )
2
The phase (t) increases or decreases linearly with time during each bit period of Tb seconds
(t) = (0)
h
t
Tb
0 t Tb
where the plus sign to send symbol 1 and the minus sign to send symbol 0. The parameter h is defined by
h = Tb (f1 f2 )
where h refers to deviation ratio, when t = Tb
Manjunatha. P
(JNNCE)
January 1, 2016
39 / 118
where h refers to deviation ratio, when t = Tb

h for symbol 1
h for symbol 0
4 h
(t ) (0), radians

(Tb ) (0) =
To send symbol 1, phase is increased by h radians whereas to

send symbol 0, phase is reduced by h radians. The phase is an
odd or even multiple of h radians at odd or even multiples of the
bit duration Tb respectively. Since all phase shifts are modulo-2
, the case of h = 1/2 is of special case in which phase can take
on only two values /2 at odd multiples of Tb and only the two
values 0 and at even multiples of Tb . This graph is called a phase
trellis.
s
2Eb
s(t) =
cos[2fc t + (t)]
Tb
s
s
2Eb
2Eb
cos[(t)] cos(2fc t)
sin[(t)] sin(2fc t)
=
Tb
Tb
p
Consider the in-phase component 2Eb /Tb cos[(t)] with deviation h = 1/2 then
Manjunatha. P
(JNNCE)
2Tb
4Tb
6Tb
8Tb
3 h
4 h
Figure 25:
0 t Tb
where the plus sign corresponds to symbol 1 and the minus sign
corresponds to symbol 0. A similar results holds for (t) in the
interval Tb t 0. The phase (0) is 0 or depending on the
past history of the modulation process.
2 h
(t ) (0), radians
(t) = (0)
t
2Tb
3 h
2 h
/2
/ 2
2Tb
Figure 26:
Phase tree
4Tb
6Tb
8Tb
Phase trellis for 1101000
January 1, 2016
40 / 118
(t) = (0)
t
0 t Tb
2Tb
s
s
s

2Eb
2Eb
2Eb
cos[(t)] =
cos[(0)] cos
t =
cos
t
sI (t) =
Tb
Tb
2Tb
Tb
2Tb
where plus sign corresponds to (0) = 0 and the minus corresponds to (0) = . Similarly in the interval
0 t 2Tb the quadrature component sQ (t) consists of half sine pulse and depends on (Tb ) and is given
by
s
s
s

2Eb
2Eb
2Eb
sQ (t) =
sin[(t)] =
sin[(Tb )] sin
t =
sin
t
Tb
Tb
2Tb
Tb
2Tb
where plus sign corresponds to (Tb ) = /2 and the minus corresponds to (Tb ) = /2. Consider the
equation h = Tb (f1 f2 ) with h = 1/2, then the frequency deviation equals the half the bit rate. This is the
minimum frequency spacing in FSK signals to make f1 and f2 orthogonal to each other. For this reason
CPFSK signal with deviation ratio of one-half is referred to as minimum shift keying (MSK).The phase states
of (0) and (Tb ) can each assume one of two possible values, and has any four possible combinations and are
as follows:
1 The phase (0) = 0 and (Tb ) = /2 to transmit symbol 1.
2 The phase (0) = and (Tb ) = /2 to transmit symbol 0.
3 The phase (0) = and (Tb ) = /2 to transmit symbol 1.
4 The phase (0) = 0 and (Tb ) = /2 to transmit symbol 0.
s
s
2Eb
2Eb
cos[(t)] cos(2fc t)
sin[(t)] sin(2fc t)
s(t) =
Tb
Tb
There are two orthonormal basis functions in si (t)
s

2
1 (t) =
cos
t cos(2fc t)
Tb
2Tb
Manjunatha. P
(JNNCE)
s
2 (t) =
2
sin
Tb
t
2Tb

sin(2fc t)
January 1, 2016
41 / 118

2
MSK signal is expressed as
Decision
Boundary
s(t) = s1 (t)1 (t) + s2 (t)2 (t) 0 t Tb
Region Z3
Region Z4
[ (0) = 0, (Tb ) = / 2]
[ (0) = , (Tb ) = / 2]
In phase component of s(t) is
Message Point 4
Message Point 3
Eb
Decision
Boundary
Tb
Z
s1 (t) =
s(t)1 (t) =
Eb cos[(0)]
Eb
Tb t Tb
Tb
Message Point 2
Eb
Eb
[ (0) = , (Tb ) = / 2]
Message Point 1
[ (0) = 0, (Tb ) = / 2]
Quadrature component of s(t) is

Region Z2
2Tb
Z
s2 (t) =
s(t)2 (t) =
Region Z1
Eb sin[(Tb )] 0 t 2Tb
The signal constellation diagram for an MSK signal is of two dimensional with four message points.
The
ofthe message
points
areas follows:
coordinates
(+ Eb , - Eb ), (- Eb , - Eb ), (- Eb , + Eb ),(+ Eb , + Eb ).
The possible values of (0) and (Tb ) are as shown in Table
Table 1:
Signal Space characterization of MSK
Input bit
0 t T
1
0
1
0
Manjunatha. P
(JNNCE)
Phase states
radians
(Tb )
+/2
+/2
-/2
-/2
(0)
0
Coordinates of
message points
s1
s2
p
p
+ pEb
pEb
pEb
p Eb
p Eb
+pEb
+ Eb
+ Eb
January 1, 2016
42 / 118
The received signal is defined by: x(t) = s(t) + w (t)

where s(t) is the Tx signal, and w (t) is sample value
of white Gaussian noise process of zero mean and power
spectral density of N0 /2. In order to determine whether
symbol 1 or 0 was transmitted, determine (0) and
(Tb ) To determine the phase (0) is as follows
In phase channel
Tb
Z
x1 (t) =
x(t)1 (t)dt = s1 (t)+w (t) Tb t Tb
Tb
Tb
Received
Signal
1 ( t ) =
x(t )
Tb
2Tb
2Tb
Z
x2 (t) =
x1
Decision
Device
x2
dt
Phase estimate ( 0 )
Logic ckt for Output Binary

wave
interleaving
phase
decisions

2
cos
cos ( 2 fc t )
Tb
2Tb
From the signal space diagram it is observed that if x1 >

0 the receiver decides (0) = 0, otherwise when x1 < 0
the receiver decides (0) = . Similarly the detection
of (Tb ) is as follows
dt
Decision
Device
Quadrature phase
channel
Phase estimate ( Tb )

2
2 ( t ) =
sin
sin ( 2 f ct )
Tb
2Tb
Figure 27:
MSK Receiver
x(t)2 (t)dt = s2 (t)+w (t) 0 t 2Tb
If x2 > 0 the received signal decides (Tb ) = /2, otherwise when x2 < 0 the receiver decides
(Tb ) = /2. The MSK and QPSK signals have similar signal space-diagram. It follows that the same
average probability of symbol error for coherent MSK
s
E
1
E
2
Pe = 1 Pc = erfc
erfc
2N0
4
2N0
By ignoring the second term probability of symbol error of MSK
s
Eb
Pe = erfc
N0
Manjunatha. P
(JNNCE)
January 1, 2016
43 / 118
MSK Transmitter and Receiver
cos A cos B = [cos(A B) + cos(A + B)]

1 (t )
BPF
(f1)
+
cos ( 2 f c t )
m1 ( t )
MSK
wave
s (t )
t
cos
2Tb
BPF
(f2)
2 (t )
m2 ( t )
Figure 28:
MSK Transmitter
In phase channel
Tb
Tb
Received
Signal
1 ( t ) =
x(t )
dt
x1
Decision
Device
Logic ckt for Output Binary

wave
interleaving
phase
decisions

2
cos
cos ( 2 fc t )
Tb
2Tb
2Tb
x2
dt
Decision
Device
Quadrature phase
channel
2 ( t ) =
Manjunatha. P
(JNNCE)
Phase estimate ( 0 )
Phase estimate ( Tb )

2
sin
sin ( 2 f ct )
Tb
2Tb
January 1, 2016
44 / 118
Figure 30:
Manjunatha. P
(JNNCE)
MSK Transmitter and Receiver
MSK waveform
January 1, 2016
45 / 118
Non Coherent Modulation Techniques
Manjunatha. P
(JNNCE)
January 1, 2016
46 / 118
Non Coherent Binary Frequency Shift Keying (FSK)
Non Coherent Binary Frequency Shift Keying

(FSK)
Manjunatha. P
(JNNCE)
January 1, 2016
47 / 118
In binary FSK the transmitted is defined by:

( q
Si (t) =
Eb
2Tb
Non Coherent Binary Frequency Shift Keying (FSK)
cos(2fi t)
0 t Tb
elsewhere
where the carrier frequency fi may be f1 or f2 .

Non coherent FSK receiverq
consists of a pair of matched filters followed by envelope detectors.
q The filter in the
2
2
upper path is matched to
T cos(2f1 t) and in the the lower path of the is matched to
T cos(2f2 t).
b
The upper and lower envelope outputs l1 and l2 are sampled at t = Tb and their values are compared. If
l1 > l2 the receiver decides in favor of symbol 1 and if l1 < l2 the receiver decides in favor of symbol 0.
Sample at t=Tb
Filter matched to
2
cos(2 f1t )
Tb
Envelope
detctor
l1
Received
Signal
If l1 is >l2
then chose 1
Compariso
n device
x(t )
Filter matched to
2
cos(2 f 2 t )
Tb
Figure 31:
l2
Envelope
detctor
If l1 is <l2
then chose 0
Sample at t=Tb
Noncoherent FSK receiver
The noncoherent binary FSK is special case of noncoherent orthogonal modulation with T = Tb and E = Eb .
The average probability of error for non coherent FSK is given by

1
Eb
Pe = exp
2
2N0
Manjunatha. P
(JNNCE)
January 1, 2016
48 / 118
Differential Phase Shift Keying
Differential Phase Shift Keying (DPSK)
Manjunatha. P
(JNNCE)
January 1, 2016
49 / 118
Differential Phase Shift Keying (DPSK) is noncoherent version of the binary PSK system. DPSK eliminates
need for a coherent reference signal at the receiver. It consists of differential encoder and phase shift keying.
To transmit the symbol 0 the phase is advanced the current signal waveform by 180 and to send symbol 1 the
phase of the current signal waveform is unchanged. The differentially encoded sequence dk is generated by
using the logic equation and is given by:
dk = bk dk1
Input
binary
sequence
bk
Binrary Wave
Amplitude
Level
Shifter
Product
Modulator
DPSK
Signal
Eb or Eb
Table 2:
dk-1
dk
Delay Tb
A
0
0
1
1
2
1 ( t ) =
cos ( 2 fct )
Tb
Figure 32:
Ex-OR & Ex-NOR Truth
Table
AB
0
1
1
0
B
0
1
0
1
AB
1
0
0
1
DPSK Transmitter
Table 3:
bk
dk1
dk
Transmitted Phase
Illustrating the generation of DPSK signal
1(Initial assumption)
0
1
1
1
0
0
1
0
0
0
1
0
1
1
1
0
0
1
0
0
0
1
0
1
1
1
0
1
1
1
0
Note: means 1800 phase shift with respect to unmodulated carrier. Phase shift is 1800 when the inputs to
Ex-Nor gate are dissimilar.
Manjunatha. P
(JNNCE)
January 1, 2016
50 / 118
dk = bk dk1
bk
dk1
dk
Transmitted Phase
1 (Initial assumption)
0
1
1
0
1
0
1
0
0
1
1
0
1
1
0
1
0
1
0
0
1
1
0
0
1
1
0
0
1
1
0
Note: means 1800 phase shift with respect to unmodulated carrier. Phase shift is 1800 when the inputs to
Ex-Or gate are similar.
Manjunatha. P
(JNNCE)
January 1, 2016
51 / 118
The receiver is equipped with a storage capability, so that it measure the relative phase difference between the
waveforms received during two successive bit intervals.
Correlator
Product
modulator
x (t )
p(t )
BPF
Sample at
every Tb
Tb
dt
l > 0 1
l <00
Decision
device
Delay Tb
Binary
Data
Vth=0
Figure 33: DPSK Receiver

Let S1 (t) denote the transmitted DPSK signal for 0 t 2Tb for the case when binary symbol 1 at the
transmitter input for the second part of this interval namely Tb t 2Tb . The transmission of 1 leaves the
carrier phase unchanged, and so S1 (t) is as follows:
q
Eb
cos(2fc t)
0 t Tb
q 2Tb
(11)
S1 (t) =
Eb
Tb t 2Tb
2T cos(2fc t)
b
Let S2 (t) denote the transmitted DPSK signal for 0 t 2Tb for the case when binary symbol 0 at the
transmitter input for the second part of this interval namely Tb t 2Tb . The transmission of 0 advances
the carrier phase by 1800 , and so S2 (t) is as follows:
q
Eb
cos(2fc t)
0 t Tb
q 2Tb
S2 (t) =
(12)
Eb
Tb t 2Tb
2T cos(2fc t + )
b
Manjunatha. P
(JNNCE)
January 1, 2016
52 / 118
Let the x(t) is the received signal. The BPF passes only the spectrum of the DBPSK signal and rejects the
noise components. In the absence of the noise, the received signal is same as the transmitted signal
s
Eb
cos(2fc t + i (t)) Tb t 2Tb
2Tb
q
q
Eb
E
where i=1,2 and i (t) = 00 or 1800
2Tb cos(2fc t)
2T cos(2fc t) or
b
IF the bits bk = dk1 (both 0 or 1) then the input to the product modulator are inphase and hence the
product modulator output is:
Eb
Eb 1
2
cos [2fc t] =
[1 + cos 2(2fc t)]
2Tb
2Tb 2
p(t) =
at t = Tb the integrator output is:
Z
=
Tb
Eb
[1 + cos 2(2fc t]dt = Eb /4
4Tb
IF the bits bk = 0 and dk1 = 1 or bk = 1 and dk1 = 0 then the input to the product modulator are out of
phase by radians hence the product modulator output is:
p(t) =
Eb
Eb 1
2
cos [2fc t] =
[1 + cos 2(2fc t)]
2Tb
2Tb 2
at t = Tb the integrator output is:

Tb
Z
=
0
Manjunatha. P
(JNNCE)
Eb
[1 + cos 2(2fc t+)]dt = Eb /4
4Tb
January 1, 2016
53 / 118
Comparison of Binary and quaternary Modulation

Techniques
Manjunatha. P
(JNNCE)
January 1, 2016
54 / 118
Error Function
In mathematics, the error function (also called the Gauss error function) is a special function (non-elementary)
of sigmoid shape which occurs in probability, statistics and partial differential equations. It is defined as
Z u
2
2
erf (u) =
exp(x )dx
0
The complementary error function, denoted erfc, is defined as
2
erfc(u) = 1 erf (u) =
exp(x )dx
u
Q Function
In statistics, the Q-function is the tail probability of the standard normal distribution. It is defined as
1
Q(u) =
2
exp(x /2)dx
u
The Q-function can be expressed in terms of the error function, or the complementary error function, as
Q(u) =
Manjunatha. P
(JNNCE)
1
1
1
erfc(u/ 2) = erf (u/ 2)
2
2
2
January 1, 2016
55 / 118
Table 4:
Comparison of Modulation Schemes with Error Probability
Modulation
Detection Method
BPSK
Coherent
FSK
Coherent
QPSK
Coherent
MSK
Coherent
Error Probability Pe
p

1
Eb /N0
2 erfc
p

1
Eb /2N0
2 erfc
p
p

erfc
Eb /N0 14 erfc 2
Eb /2N0
p
p

erfc
Eb /N0 14 erfc 2
Eb /2N0
DPSK
FSK
Non Coherent
Non Coherent
1
2 exp
1
2 exp
(Eb /N0 )
(Eb /2N0 )
The error rates for all the systems decease monotonically with increasing values of Eb /N0 .
Coherent PSK produces a smaller error rate than any of other systems.
Coherent PSK and DPSK require an Eb /N0 that is 3 dB less than the corresponding values for
conventional coherent FSK and non coherent FSK respectively to realize the same error rate.
At high values of Eb /N0 DPSK and noncoherent FSK perform almost as well as coherent PSK and
conventional coherent FSK, respectively, for the same bit rate and signal energy per bit
p
p
In QPSK two orthogonal carriers 2/T cos(2fc t) and 2/T sin(2fc t) are used, where the carrier
frequency fc is an integral multiple of the symbol rate 1/T with the result that two independent bit
streams can transmitted and subsequently detected in the receiver. At high values of Eb /N0 coherently
detected binary PSK and QPSK have about the same error rate performance for the same value of
Eb /N0 .
p
p
In MSK two orthogonal carriers 2/Tb cos(2fc t) and 2/Tb sin(2fc t) are modulated by the two
antipodal symbol shaping pulses cos(t/Tb ) and sin(t/Tb ) respectively over 2Tb intervals.
Correspondingly, the receiver uses a coherent phase decoding process over two successive bit intervals to
recover the original bit stream. MSK has the same error rate performance as QPSK.
Manjunatha. P
(JNNCE)
January 1, 2016
56 / 118
Performance comparision for different modulation schemes
10
10
10
BPSK
FSK
QPSK
DPSK
Noncoherent FSK
10
10
Figure 34:
Manjunatha. P
(JNNCE)
5
Eb/No, dB
10
15

January 1, 2016
57 / 118
M-Ary Modulation Techniques
Manjunatha. P
(JNNCE)
January 1, 2016
58 / 118
In M-ary signaling scheme, M possible signals s1 (t), s2 (t)..sM (t) are sent during each symbol interval T.
M = 2n possible states where n is an integer.
The symbol duration T = nTb where Tb is the bit duration.
These M signals are generated by changing the amplitude, frequency or phase of a carrier in M discrete
steps.
Hybrid form: M-ary QAM
Manjunatha. P
(JNNCE)
January 1, 2016
59 / 118
M-Ary PSK
M-Ary PSK
Manjunatha. P
(JNNCE)
January 1, 2016
60 / 118
M-Ary PSK
In M-ary PSK the phase of the carrier takes one of M possible values.
i = 2i/M where i = 0,1,M - 1
r
si (t) =
2E
cos
T
2i
M
2fc t +
where E is the signal energy/symbol and fc = nc /T

r
si (t) =
2E
cos
T
2i
M

cos 2fc t
2E
sin
T
2i
M

sin 2fc t
Two basis functions 1 (t) and 2 (t) are defined as:

r
1 (t) =
2
cos 2fc t
T
0tT
2
sin 2fc t
T
0tT
r
2 (t) =
Figure 35:
Signal Space
Constellation for octaphase shift
keying
Message points, and the associated signal vectors defined by:
sMPSK (t) =
Manjunatha. P

E cos
(JNNCE)
2i
M

E sin
2i
M

i = 0, 1, 2, 3, 4
January 1, 2016
61 / 118
The optimum M-ary PSK receiver consists of pair of correlators with reference signals in phase quadrature.
The two correlator outputs xI and xQ are fed to phase discriminator to compute the phase estimate.

xQ
1
= tan
xI
The phase discriminator selects the set {si (t), i = 0, ..M 1} to a particular signal whose phase is closest to
In the presence of noise, the input to the phase discriminator is:
the estimate .

2i
2i
+ wI
xQ = E sin
+ wQ
xI = E cos
M
M
where wI and wQ are samples of two independent Gaussian random variables WI and WQ whose mean is zero
and common variance 2 = N0 /2.
The probability of error is approximately given by
r
2
Pe = 1
E sin( )
N0
M
x I (t )
2
exp(z )dz
s
Pe = erfc
E
sin
N0
Received
Signal
x (t )
cos(2 f ct )
Phase
Discriminator
(JNNCE)
Parallel-toReconstructed
serial
binary data
converter
For large values of E /N0 and M 4 the probability of symbol error is approximately given by
s

2E
Pe = erfc
sin
M 4
N0
2M
Manjunatha. P
dt
sin(2 f c t )
Figure 36:
dt
xQ (t )
Receiver for M-ary PSK
January 1, 2016
62 / 118
M-Ary Quadrature Amplitude Modulation
Manjunatha. P
(JNNCE)
January 1, 2016
63 / 118
In M-Ary quadrature amplitude modulation (QAM)

the carrier amplitude as well as phase modulation is
done.
In this modulation the modulated signal envelope is
not constant.
By using this higher spectral efficiency can be
achieved.
The general form of an M-ary QAM signal is :
r
si (t) =
2E0
ai cos(2fc t) +
T
2E0
bi sin(2fc t)
T
where 0 t T i = 1, 2, ..., M, E0 is the energy of

the signal with the lowest amplitude, ai and bi are the
pair of independent integers.
The basis function for the si (t):
r
1 (t) =
2
cos 2fc t
T
2 (t) =
The coordinates of the i th message point are ai
a i , bi =
E and bi
2
sin 2fc t
T
Figure 37:

a i , bi =
Constellation for L=4, M=16.
(3, 3)
(3, 1)
(3, 1
(3, 3)
(1, 3)
(1, 1)
(1, 1)
(1, 3)
(1, 3)
(1, 1)
(1, 1)
(1, 3)
(3, 3)
(3, 1)
(3, 1)
(3, 3)
E are:
(L + 1, L 1)
(L + 1, L 3)
(L + 3, L 1)
(L + 3, L 3)
...
...
(L 1, L 1)
(L 1, L 3)
.
.
.
(L + 1, L + 1)
(L + 3, L + 1)
...
(L 1, L + 1)
where L =
M
Manjunatha. P
(JNNCE)
January 1, 2016
64 / 118
The probability of symbol error for M-ary QAM is given by

s

1
E0
Pe = 2 1
erfc
N0
M
M 4
2-to-L
converter
Binary
Data
Serial-to parallel converter accepts a binary sequence at

a bit rate Rb = 1/Tb and produces two parallel binary
sequences whose bit rates are Rb = 1/2.
cos(2 f ct )
Serial-toParallel
converter
M-ary Qam
wave
2-to-L
converter
sin(2 f c t )
Figure 38:
Received
Signal
x (t )
cos(2 f ct )
sin(2 f c t )
Figure 39:
(JNNCE)
Decision
Circuit
(L-1)
Thresholds
Manjunatha. P
dt
dt
M-ary-QAM Transmitter.
Parallel-toserial
converter
Demodulated
binary data
Decision
Circuit
(L-1)
Thresholds
M-ary-QAM Receiver.
January 1, 2016
65 / 118
M-ary FSK
M-ary FSK
Manjunatha. P
(JNNCE)
January 1, 2016
66 / 118
M-ary FSK
M-ary FSK
Frequencies are chosen in a special way so that they are easily separated at the demodulator
(orthogonality principle).
M-ary FSK transmitted signals are:
r
si (t) =

2E
cos
(nc + i)t 0 t Ts i = 0, 1, ..., M
T
Ts
fc = nc /2T for some integer nc

The M transmitted signals are of equal energy and equal duration.
The signal frequencies are separated by 1/2T Hz, making the signals orthogonal to one another.
Manjunatha. P
(JNNCE)
January 1, 2016
67 / 118
Power Spectra
Power Spectra
Power Spectra
Manjunatha. P
(JNNCE)
January 1, 2016
68 / 118
Introduction
Baseband Data Transmission
Discrete Pulse Modulation

Pulse Amplitude Modulation (PAM)
Pulse Duration Modulation (PDM)
Pulse Position Modulation (PPM)
Pulse Amplitude Modulation (PAM)
PAM Systems are preferred as they are most efficient in terms of power and bandwidth utilization.
Digital data, a sequence of binary digits is coded into electrical pulses or waveforms for the purpose of
transmission over the channel.This process is called line coding or transmission coding.
Line codes are also called as signaling formats.
Manjunatha. P
(JNNCE)
January 1, 2016
69 / 118
Introduction
Binary Data Formats
Each line code is described by

1
2
A symbol mapping function Ak

A pulse shape v (t)
Symbol mapping functions Ak .

1
2
3
Unipolar
Polar
Bipolar
A pulse shape v (t)

1
2
3
NRZ (Nonreturn-to-zero)
RZ (Return to Zero)
Manchester (split phase)
Manjunatha. P
(JNNCE)
January 1, 2016
70 / 118
Introduction
Binary Data Format
Figure 40: Binary Data Format
Manjunatha. P
(JNNCE)
January 1, 2016
71 / 118
Introduction
Power Spectral Density

Power Spectral Density (PSD) of a random process is a function describing the distribution of power as
a function of frequency[?].
This process is important when considering filtering in communication systems while evaluating the
signal and noise at the filter output.
Manjunatha. P
(JNNCE)
January 1, 2016
72 / 118
Introduction
PSD of Discrete PAM Signals

PSD can be evaluated by using
1
Deterministic approach - the waveform for a line code that results from a
particular data sequence.
Stochastic approach - line codes with a random data sequence.
Stochastic approach
The line code is viewed as a random process X(t) defined by
X (t) =
Ak v (t kT )
k=
where v(t) is the symbol pulse shape (rect(t/T))

T is the duration of one symbol
Ak is the discrete random variable.
Manjunatha. P
(JNNCE)
January 1, 2016
73 / 118
Introduction
PSD of Discrete PAM Signals

In general the PSD of a digital signal is described by the equation
Sx (f ) =
X
1
2
j2fnT
|v (f )|
RA (n)e
T
n=
where v(f) is the Fourier Transformer of the pulse shaping function v(t)
RA (n) is the autocorrelation function of the data.
The autocorrelation function RA (n) is,
RA (n) = E [Ak Akn ]
I
P
=
[Ak Akn ]i Pi
i=1
where Ak and Akn are the voltage levels of the data pulses at the k th and
k nth symbol positions,
Pi is the probability of the i th product Ak Akn .
I is the number of possible products Ak Akn .
Manjunatha. P
(JNNCE)
January 1, 2016
74 / 118
Polar NRZ Signaling
Figure 41: Polar NRZ signal format

Figure 42: Polar NRZ PSD
The Amplitude is represented by Ak and takes the value
Ak =
+a
symbol 1
symbol 0
The value Ak is assumed to be equally likely to occur and the data are independent.
P[Ak = +a] = P[Ak = a] =
Manjunatha. P
(JNNCE)
1
2
January 1, 2016
75 / 118
Polar NRZ Signaling
The basic pulse shaping function is rectangular,which is defined by

rect
t
Tb

=
1,
|t| <
Tb
2
|t| >
Tb
2
0,
The fourier Transformer pair is

v (t) = rect(
V (f ) =
t
) v (f )
Tb
v (t)e j2ft dt
Tb
R2
e j2ft dt = Tb
T
b
2
sin(fTb )
fTb
= Tb sin c (fTb )
Manjunatha. P
(JNNCE)
January 1, 2016
76 / 118
Polar NRZ Signaling
For n 6= 0, there are four combinations (I=4) for

Ak .Akn
axa
-axa
ax-a
-ax-a
1
2
for n = 0
3
2
P
RA (0) =
(Ak Ak )i Pi
i=1

= a2 21 + (a)2 12 = a2
For this random data,the probability is

RA (n) = E [Ak Akn ]
4
P
=
[Ak Akn ]i Pi = a2
1
4
i=1
1
4
+ (a)(a)
1
4
+ (a)(a)
1
4
+ (a)2
=0
Summarizing we get

RA (n) =
a2
0
n=0
n 6= 0
Power Spectral Density is:

Sx (f ) =
X
1
2
j2fnT
|v (f )|
RA (n)e
T
n=
PSD of the polar NRZ is

2
Sx (f ) = a Tb sin c (fTb )
Manjunatha. P
(JNNCE)
January 1, 2016
77 / 118
1
4
Power Spectra
Introduction
Power Spectra
The description given by s(t) for a pass-band signal contains the definitions of ASK, PSK and FSK
signals,depending on the way in which the in-phase component sI (t) and the Quadrature component
sQ (t) are defined.
The analysis of these kind of signals may be simplified by using complex notation.The signal s(t) may
be expressed as
s(t) = sI (t) cos(2fc t) sQ (t) sin(2fc t)
= Re[
s (t) exp(j2fc t)]
Also we have
s (t) = SI (t) + jSQ (t)

And
exp(j2fc t) = cos(2fc t) + j sin(2fc t)
is called the complex envelop of the band-pass signal s(t).
The signal s(t)
The signals SI (t) ,SQ (t) and
s (t) are all low pass signals.They are uniquely defined in terms of the
band-pass signal s(t) and the carrier frequency fc provided that the half-bandwidth of s(t) is less than
the carrier frequency fc .
Manjunatha. P
(JNNCE)
January 1, 2016
78 / 118
Power Spectra
Introduction
Let SB (f ) denote the power spectral density of the complex envelope

s (t), where SB (f ) is referred to as
the baseband power spectral density.
The power spectral density SS (f ), of the original band-pass signal s(t) is a frequency shifted version of
SB (f ), except the scaling factor shown as
SS (f ) =
1
[SB (f fc ) + SB (f + fc )]
4
Since
s (t) is a low-pass signal,the calculation of SB (f ) would be simpler than the calculation of
SS (f ).Therefore it is sufficient to evaluate the baseband power spectral density SB (f ).
Manjunatha. P
(JNNCE)
January 1, 2016
79 / 118
Power Spectra
Binary PSK and FSK Signals
Binary PSK
The Binary PSK wave has an In-phase component only.
Depending on whether we have a symbol 1 or 0 at the modulator input during the signaling interval
0 t Tb we find that this in-phase component equals +g(t) or -g(t) respectively, where g(t) is the
symbol shaping function defined by,
q
2Eb
0 t Tb
Tb
g (t) =
0
elsewhere
The baseband power spectral density of a binary PSK wave equals
SB (f ) =
2Eb sin2 (Tb f )

(Tb f )2
= 2Eb sin c 2 (Tb f )
Manjunatha. P
(JNNCE)
January 1, 2016
80 / 118
Power Spectra
Figure 43: Power Spectra of binary PSK
Figure 44: Power Spectra of binary PSK

Manjunatha. P
(JNNCE)
January 1, 2016
81 / 118
Power Spectra
Binary FSK
The Binary FSK wave, the two transmitted frequencies f1 and f2 differ by an amount equal to the bit
rate 1/Tb and their arithmetic mean equals the nominal carrier frequency fc
with the assumption that phase continuity is always maintained, Binary FSK signal is given by
s
2Eb
t
s(t) =
cos(2fc t
)
0 t Tb
Tb
Tb
With the trigonometric identity,we get
q
q
2Eb
2Eb
t
t
s(t) =
Tb cos( Tb ) cos(2fc t)
Tb sin( Tb ) sin(2fc t)
q
q
2Eb
2E
t
t
b
=
T cos( T ) cos(2fc t)
T sin( T ) sin(2fc t)
b
The plus sign corresponds for symbol 0 transmission, and the minus sign for symbol 1.
The Binary FSK wave has an In-phase component and quadrature component.
Manjunatha. P
(JNNCE)
January 1, 2016
82 / 118
Power Spectra
Binary FSK
The In-phase component is completely independent of the input binary wave. It is equal to
s
2Eb
t
cos(
)
for all values of time t
Tb
Tb
The Power Spectral density of this component therefore consists of two delta functions, weighted by the
factor Eb /2Tb and occurring at f = 1/2Tb .
The Quadrature component is directly related to the input binary wave.
During the signaling interval 0 t Tb it equals +g(t) when symbol 1, and -g(t) when we have
symbol 0 where g(t) is the symbol shaping function defined by,
( q
2Eb
t
0 t Tb
Tb sin( Tb )
g (t) =
0
elsewhere
The energy spectral density of this symbol shaping function is
g (f ) =
8Eb Tb cos2 (Tb f )

2 (4Tb2 f 2 1)2
The Power spectral density of Quadrature component is

g (f )
Tb
The In-phase component and quadrature component of Binary FSK wave are independent of each other.
The base-band power spectral density of the binary FSK wave equals the sum of the power spectral
densities of these components and is given by,

Eb
1
1
8Eb cos2 (Tb f )
SB (f ) =
f
+ f +
+ 2
2Tb
2Tb
2Tb
(4Tb2 f 2 1)2
Manjunatha. P
(JNNCE)
January 1, 2016
83 / 118
Power Spectra
Figure 45: Power Spectra of binary PSK and FSK
Manjunatha. P
(JNNCE)
January 1, 2016
84 / 118
Power Spectra
QPSK and MSK Signals
QPSK
The binary wave at the input of the modulator is random with symbol 1 and 0 being equally likely and
the symbols transmitted during the adjacent time slots are statistically independent.
The QPSK has In-phase component and quadrature component.
Depending on dibit sent during the signaling interval 0 t Tb the in-phase component equals +g(t)
or -g(t) and similarly for quadrature component, where g(t) is the symbol shaping function defined by,
( q
g (t) =
E
T
0 t Tb
elsewhere
The In-phase component and quadrature component have a common Power spectral density given by
2
E sin c (Tf )
The In-phase and quadrature component are statistically independent.
The base-band power spectral density of the QPSK signal is the sum of the individual power spectral
densities of the In-phase and quadrature components and hence
SB (f ) = 2E sin c 2 (Tf )
= 4Eb sin c 2 (2Tb f )
Manjunatha. P
(JNNCE)
January 1, 2016
85 / 118
Power Spectra
MSK
The binary wave at the input of the modulator is random with symbol 1 and 0 being equally likely and
the symbols transmitted during the adjacent time slots are statistically independent.
Depending on the value of phase state (0) in-phase component equals +g(t) or -g(t), where g(t) is the
symbol shaping function defined by,
( q
g (t) =
0
2Eb
Tb
t
cos( 2T
)
Tb t Tb
elsewhere
g (f ) =
32Eb Tb
2
"
cos(2Tb f )
(16Tb2 f 2 1)
#2
The Power spectral density of In-phase component is

g (f )
2Tb
Manjunatha. P
(JNNCE)
January 1, 2016
86 / 118
Power Spectra
MSK
Depending on the value of phase state (Tb ) quadrature component equals +g(t) or -g(t), where g(t) is
defined by,
( q
2Eb
t
0 t Tb
Tb sin( 2Tb )
g (t) =
0
elsewhere
g (f ) =
32Eb Tb
2
"
cos(2Tb f )
(16Tb2 f 2 1)
#2
The In-phase and quadrature component have same power spectral density.
The In-phase and quadrature component are statistically independent.Hence The base-band power
spectral density of the MSK signal is given by
h (f ) i
g
SB (f ) = 2 2T
b

2
32E
cos(2Tb f )
= 2b
2 2
Manjunatha. P
(JNNCE)
(16T f 1)
b
January 1, 2016
87 / 118
Power Spectra
Figure 46: Power Spectra of QPSK and MSK
Manjunatha. P
(JNNCE)
January 1, 2016
88 / 118
Power Spectra
M-ary Signals
M-ary PSK
Binary PSK and QPSK are the special cases of M-ary PSK signals. The symbol duration is defined by
T = Tb log2 M
With the same procedure as that of QPSK the base-band power spectral density of M-ary PSK signal is
given by,
SB (f ) = 2E sin c 2 (Tf )
= 2Eb log2 M sin c 2 (Tb f log2 M)
Figure 47: Power Spectra of M-ary PSK signals for M=2,4,8
Manjunatha. P
(JNNCE)
January 1, 2016
89 / 118
Power Spectra
M-ary Signals
M-ary FSK
A particular case of interest occurs when the frequencies assigned to the multilevels make the frequency
spacing uniform and the frequency deviation ratio k=0.5.
M signal frequencies are separated by 1/2T where T is the symbol duration.
For k=0.5, the base-band power spectral density of M-ary FSK signals is defined by

M
M
M
1 X sin i 2
1 XX
sin i 2 sin j 2
SB (f ) = 4Eb
+ 2
cos(i + j )
2M i=1
i
M i=1 j=1
i
j
where
i = (fTb
i
)
4
i = 2i (M + 1)
i = 1, 2, ..., M
Figure 48: Power Spectra of M-ary FSK signals for M=2,4,8

Manjunatha. P
(JNNCE)
January 1, 2016
90 / 118
Bandwidth Efficiency
Manjunatha. P
(JNNCE)
January 1, 2016
91 / 118
The channel bandwidth and transmitted power are two primary communication resources which provides
spectral efficiency.
Spectral (bandwidth) efficiency is defined as the ratio of data rate to the channel bandwidth.
It is measured in units of bits per second per hertz.
The Bandwidth Efficiency is expressed as =
Rb
B
bits/s/Hz
where Rb is the data rate and B is the channel bandwidth.

Bandwidth Efficiency of M-ary PSK Siginals
The power spectra of M-ary PSK signals possess a main lobe bounded by well defined spectral nulls.
Most of the signal power of an M-ary PSK signal contains in the main lobe.
The channel bandwidth required to pass M-ary PSK signals is given by B =
2
T
where T is the symbol duration and T is related to the bit duration Tb as Rb = 1/Tb .
The channel bandwidth is defined in terms of the bit rate Rb as
B =
2Rb
log2 M
The Bandwidth Efficiency of M-ary PSK Signals is expressed as

=
M
bps/Hz
Manjunatha. P
(JNNCE)
2
.5
4
1
log2 M
Rb
=
B
2
8
1.5
16
2
32
2.5
64
3
January 1, 2016
92 / 118
Bandwidth Efficiency of M-ary FSK Siginals

M-ary FSK signal consists of an orthogonal set of M frequency-shifted signals.
The adjacent signals are separated from each other by frequency difference 1/2T to maintain
orthogonality.
The channel bandwidth required to pass M-ary FSK signals is given by
B =
M
2T
where T is the symbol duration and T is related to the bit duration Tb as Rb = 1/Tb and M is the
number of symbols.
The most of the signal power will be contained in the bandwidth B defined in the previous expression.
The channel bandwidth is defined in terms of the bit rate Rb as
B =
Rb M
2log2 M
The Bandwidth Efficiency of M-ary FSK Signals is expressed as

=
Manjunatha. P
(JNNCE)
Rb
2log2 M
=
B
M
January 1, 2016
93 / 118
Manjunatha. P
(JNNCE)
January 1, 2016
94 / 118
Synchronization
Synchronization
Synchronization
Manjunatha. P
(JNNCE)
January 1, 2016
95 / 118
Synchronization
Synchronization
Synchronization
In coherent reception the receiver must be synchronous to the transmitter [1, 3, 4]

Two sequences of events (Tx and Rx) are synchronous relative to each other when the events in one
sequence and the corresponding events in the other occur simultaneously.
The process of making a situation synchronous, and maintaining it in this condition is called
synchronization.
To demodulate the received signal the following two basic modes of synchronization are necessary:
1
Carrier Synchronization: Knowledge of both frequency and phase of the carrier

is necessary. The estimation of frequency and phase is called carrier recovery
or carrier synchronization.
Symbol Synchronization: The receiver has to know the starting and finishing
times of the individual symbols. The estimation of these times is called clock
recovery or symbol synchronization.
Manjunatha. P
(JNNCE)
January 1, 2016
96 / 118
Synchronization
Carrier Synchronization
Carrier Synchronization:
Coast loop method is used to recover the carrier frequency.
The loop consists of two paths, one in-phase and the other referred to as quadrature that are coupled
together via common voltage controlled oscillator (VCO) to form a negative feedback.
VCO is a circuit module that oscillates at a controlled frequency .
The Oscillating Frequency is controlled using Voltage VControl .
When synchronization attains the demodulated output appears at the output of in-phase path, and the
output is zero at quadrature path under ideal condition.
Demodulated
binary wave
Low pass
filter
Received
BPSK
Signal
Voltage
controlled
oscillator
Phase
discrimanator
-900
phase
shifter
Figure 50:
Low pass
filter
Figure 49:
Manjunatha. P
(JNNCE)
Voltage controlled oscillator.
Coast loop.
January 1, 2016
97 / 118
Synchronization
Carrier Synchronization
PLL:
PLL is an Electronic Module (Circuit) that locks the phase of the output to the input.
Figure 51:
PLL.
Figure 52:
Locked Vs. Unlocked Phase.
PLL is a feedback system that detects the phase error and then adjusts the phase of the output.
The Phase Detector (PD), detects between the output and the input through feedback system
Voltage Control Oscillator (VCO) adjusts the phase difference
Figure 53:
Manjunatha. P
(JNNCE)
Implementation of PD.
January 1, 2016
98 / 118
Synchronization
Symbol Synchronization
Simple PLL:
Structure.
Phase Detector ( XOR ) that detects the phase error

Low Pass Filter ( to smooth )
Voltage Control Oscillator (VCO)
Basic Idea
If VI and Vout are out of phase (unlocked), then the PD module detects the
error and the LPF smoothes the error signal.
The control signal slows down or speeds up the VCO module; hence, the phase
is corrected (locked)
Figure 54:
Manjunatha. P
(JNNCE)
Simple PLL.
January 1, 2016
99 / 118
Synchronization
Symbol Synchronization:
Symbol synchronization is achieved by transmitting clock along with the data-bearing signal in the
multiplexed form.
At the receiver the clock is extracted by filtering of the modulated waveform.
This approach minimizes the time required for carrier/clock recovery but the fraction of the transmitted
power is allocated to transmit the clock.
In another approach the clock is extracted by processing demodulated baseband waveforms, thereby avoiding
any wastage of transmitted power.
Consider a rectangular pulse defined by

g (t) =
a
0
g (t )
0tT
Otherwise
The output of the filter attains its peak value at time

t = T and that it is symmetric about this point.
Figure 55:
Rectangular pulse g(t).
From the figure it shows that the proper time to

sample the matched filter output is at t = T .
Suppose the matched filter output is sampled early at
t = T 0 T or late at t = T + 0 T .
The absolute values of the two samples will be equal
and smaller than the peak value at t = T .
Matched
filter
output
a 2T
The error signal, the difference between the absolute

values of the two samples is zero, and the proper
sampling time is the midpoint between t = T 0 T
and t = T + 0 T .
This special condition is viewed as an equilibrium
point, in that if it is deviated from it the error
becomes nonzero.
Manjunatha. P
(JNNCE)
Figure 56:
T - 0T
T + 0T
2T
Output of the filter matched
to g(t).
January 1, 2016
100 / 118
Synchronization
The early-late gate symbol synchronizer consists of correlators and these correlators will integrate over a
full symbol interval T , with one starting 0 T early relative to the transition time estimate and the
other starting 0 T late.
An error signal e(kT ), is generated by taking the difference between the actual transition times t(kT )
and their local estimate t (kT ), the error signal is zero when no transition occurs at t(kT ), otherwise it
is linearly proportional to irrespective of the polarity of .
The error signal is low pass filtered and then applied to a voltage controlled oscillator that controls
(symbol waveform generator) the charging and discharging instants of the correlators.
The instantaneous frequency of the local clock is advanced or retarded in an iterative manner until the
equilibrium point is reached, and symbol synchronization is thereby established.
Absolute
value
calculator
dt
Advance by
0T
Received
Signal
Symbol
waveform
generator
Voltage
Contolled
oscillator
Low pass
Filter
Retard by
0T
Symbol Timing
dt
Absolute
value
calculator
Figure 57: Early-late gate type of symbol synchronizer.

Manjunatha. P
(JNNCE)
January 1, 2016
101 / 118
Applications
Applications
Applications:
Voice Grade Modems
1 Existing Telephone network developed for voice communication operates in the range 300-3400Hz.
2 FSK with two frequencies 1300 and 2100 with 1200 bps (Simplicity and economical).
3 DPSK with 1800 Hz operating at 4000 bps
4
M-ary QAM: 16-QAM, 2400 bauds (9600 bps)
Digital Radio.
Digital Communications by Satellite.
Manjunatha. P
(JNNCE)
January 1, 2016
102 / 118
Problems
Problems
Problems
Manjunatha. P
(JNNCE)
January 1, 2016
103 / 118
Problems
Problems
Hari Bhat
8.4 An FSK system transmits binary data at a rate of 106 bits per second. Assuming channel AWGN with zero
mean and power spectral density of N0 /2 = 2 1020 W/Hz. Determine the probability of error. Assume
coherent detection and amplitude of received sinusoidal signal for both symbol 1 and 0 to be 1.2 microvolt.
Solution:
s
1 2
2Eb
A=
Eb = A Tb
Tb
2
Signal energy per bit
Eb =
1 2
1
6 2
6
18
A Tb = (1.2 10 ) 10
= 0.72 10
Joulels
2
2
Given N0 /2 = 2 1020 W/Hz

N0 = 4 1020 W/Hz
The average probability of error for coherent PSK is
Pe =

1
Eb
0.72 1018
= 0.5 erfc
= 0.5 erfc
erfc
0.09 100 = 0.5 erfc(3)
2
2N0
8 1020
From complementary error function table

erfc(3) = 0.00002
6
Pe = 0.5 0.00002 = 0.00001 = 1 10
Manjunatha. P
(JNNCE)
January 1, 2016
104 / 118
Problems
Problems
8.8 A binary FSK system transmits binary data at a rate of 2 MBPS. Assuming channel AWGN with zero
mean and power spectral density of N0 /2 = 1 1020 W/Hz. The amplitude of the received signal in the
absence of noise is 1 microvolt. Determine the average probability of error for coherent detection of FSK.
Solution:
s
2Eb
1 2
A=
Eb = A Tb
Tb
2
Signal energy per bit
Eb =
1 2
1
6 2
6
18
A Tb = (1 10 ) 0.5 10
= 0.25 10
Joulels
2
2
Given N0 /2 = 1 1020 W/Hz

N0 = 2 1020 W/Hz
Pe =
1
Eb
0.25 1018
= 0.5 erfc
= 0.5 erfc( 6.25) = 0.5 erfc(2.5)
erfc
2
2N0
4 1020
From error function table

erfc(2.5) = 0.000407
Pe = 0.5 0.000407 = 0.0002035
Manjunatha. P
(JNNCE)
January 1, 2016
105 / 118
Problems
Problems
8.10 A binary data is transmitted over an AWGN channel using binary PSK at a rate of 1MBPS. It is desired
to have average probability of error Pe 104 . Noise power spectral density is N0 /2 = 1 1012 W/Hz.
Determine the average carrier power required at the receiver input, if the detector is of coherent type.
Solution:
Let P be the power required at the receiver then Eb = PTb where Tb is the bit duration.
Since N0 /2 = 1 1012 W/Hz
N0 = 2 1012 W/Hz
Tb =
1
1
6
=
= 10
bit rate
106

Eb
N0

=
P Tb
P 106
6
= 0.5 P 10
=
N0
2 1012

s
1
Eb
Pe = erfc
2
N0
4
Pe = 0.5 erfc(0.5 P 10 ) 10
4
erfc(0.5 P 10 ) 2 10
From complementary error function table
erfc(2.63) = 0.0002
6
0.5 P 10 2.63
Manjunatha. P
(JNNCE)
P 1.315 10
W
January 1, 2016
106 / 118
Problems
Problems
Proakis 5.18 Suppose that binary PSK is used for transmitting information over an AWGN with a power
spectral density of N0 /2 = 1010 W/Hz. The transmitted signal energy is Eb = 1/2A2 T where T is the bit
interval and A is the signal amplitude. Determine the signal amplitude required to achieve an error probability
of 106 when the data rate is a) 10 kbits/s, b) 100 kbits/s, and c)1 Mbits/s
Solution:
q
2Eb
A=
Eb =
T
b
1 2
2 A Tb
and Pb =
1
2 erfc
q
s
Pb = Pb =
Pb =
1
erfc
2
Eb
N0
For binary phase modulation, the error probability is
Eb
0.5 A2 Tb
1
= erfc
N0
2
N0
1
0.5 A2 Tb
0.5 A2 Tb
= 106 erfc
= 2 106
erfc
2
N0
N0
6
We
p know that erfc(x) = 2 10 2 and from erfc table the value of x=3.36
0.5A2 Tb /N0 = 3.36 0.5A Tb /N0 = 11.2896
A2 Tb = 22.579N0 A2 = 22.5792 p
1010 (1/Tb ) = 4.515 109 (1/Tb )
A2 = 4.515 109 (1/Tb ) A = 4.515 109 (1/Tb )
If the data rate is 10 Kbps, then
p
3
A = 4.515 109 10 103 = 6.7193 10
If the data rate is 100 Kbps, then

A=
4.515 109 100 103 = 21.2 10
If the data rate is 1 Mbps, then

A=
Manjunatha. P
(JNNCE)
4.515 109 1 106 = 67.19 10

January 1, 2016
107 / 118
Problems
Problems
5.18 Suppose that binary PSK is used for transmitting information over an AWGN with a power spectral
density of N0 /2 = 1010 W/Hz. The transmitted signal energy is Eb = 1/2A2 T where T is the bit interval
and A is the signal amplitude. Determine the signal amplitude required to achieve an error probability of 106
when the data rate is
a) 10 kbits/s
b) 100 kbits/s
c) 1 Mbits/s
For binary phase modulation, the error probability is
s
2Eb
A2 T
= Pb = Q
Pb = Q
N0
N0
With Pb = 106 from the table it is found that 4.74
Since in Q(x) x > 3
!
1
(x)2
6
6
10
=
= 4.05 10
exp
2
4.47 2
s
A2 T
2
10
= 4.74 A T = 44.9352 10
N0
If the data rate is 10 Kbps, then the bit interval is T = 104 and therefore, the signal amplitude is
p
3
A = 44.9352 1010 104 = 6.7034 10
p
2
A = 44.9352 1010 105 = 2.12 10
p
2
A = 44.9352 1010 106 = 6.7034 10
Manjunatha. P
(JNNCE)
January 1, 2016
108 / 118
Problems
Problems
Problems:Sklar
4.2 A continuously operating coherent BPSK system makes errors at the average rate of 100 errors per day.
The data rate is 1000 bits/s. The single-sided noise power spectral density is N0 = 1010 W /Hz.
If the system is ergodic, what is the average bit error probability?
If the value of received average signal power is adjusted to be 106 W will, this received power be
adequate to maintain the error probability found in part (a)?
Solution:
The total bit detected in one day= 1000 86400 == 8.64 107
100
6
= 1.16 10
8.64 107
s
!
r
2Eb
2ST
=Q
Pb = Q
N0
1010
Pb =
where S = 106 W and T =
1
1000
s
Pb = Q
2 106
= Q( 20) = Q(4.47)
1000 1010
Since in Q(x) x > 3

Pb =
1
exp
4.47 2
(4.47)2
2
!
6
= 4.05 10
No
Manjunatha. P
(JNNCE)
January 1, 2016
109 / 118
Problems
Problems
Find the expected number of bit errors made in one day by the following continuously operating coherent
BPSK receiver. The data rate is 5000 bps. The input digital waveforms are s1 (t) = Acos0 t and
s2 (t) = Acos0 t where A= 1 mV and the the single-sided noise power spectral density is
N0 = 1011 W /Hz. Assume that signal power and energy per bit are normalized to a 1 resistive load.
Solution:
s
A=
2Eb
Tb
Eb =
A 2 Tb
2
s
1
1
Eb
= erfc
Pe = erfc
2
N0
2
Tb =
1
Rb
A 2 Tb
2N0
1
(1 103 )2
1
= erfc(3.1622)
Pe = erfc
2
2 1011 5000
2
Pe =
1
6
0.00000786 = 3.93 10
2
Average no of errors in one day = 5000bits 24 60 60 3.93 106 ' 1698 bits in error
Manjunatha. P
(JNNCE)
January 1, 2016
110 / 118
Problems
Problems
Find the expected number of bit errors made in one day by the following continuously operating coherent
BPSK receiver. The data rate is 5000 bps. The input digital waveforms are s1 (t) = Acos0 t and
s2 (t) = Acos0 t where A= 1 mV and the the single-sided noise power spectral density is
N0 = 1011 W /Hz. Assume that signal power and energy per bit are normalized to a 1 resistive load.
Solution:
s
2Eb
A2 T
=Q
Pb = Q
N0
N0
where
Q(x) '
s
Pb = Q
Pb =
1
exp
x 2
x 2
2
!
x >3

106
=Q
20 = Q(4.47)
5000 1011
1
exp
4.47 2
(4.47)2
2
!
6
= 4.05 10
Average no of errors in one day = 5000bits 86400 4.05 106 ' 1750 bits in error
Manjunatha. P
(JNNCE)
January 1, 2016
111 / 118
Problems
Problems
Find the error probability for a BPSK system with a bit rate of 1 MBPS. The received waveform
s1 (t) = Acos0 t and s2 (t) = Acos0 t are coherently detected with a matched filter. The value of A is 10
mV. Assume that the single-sided noise power spectral density is n0 = 1011 W /Hz and that signal power and
energy per bit are normalized to a 1 load.
Solution:
r
2Eb
1
2
6
A=
= 10 V
T =
= 10 S
T
R
s
s
A2
2Eb
11
Eb =
T = 5 10
J
= 3.16
2
N0
s
2Eb
= Q(3.16)
Pb = Q
N0
From Q table Pb = 8 104
Manjunatha. P
(JNNCE)
January 1, 2016
112 / 118
Problems
Problems
Problem 7.34
Suppose that binary PSK is used for transmitting information over an AWGN with power-spectral density of
No/2 = 1010 W/Hz. The transmitted signal energy is Eb = A2 T /2, where T is the bit interval and A is the
signal amplitude. Determine the signal amplitude required to achieve an error probability of 106 , if the data
rate is (a) 10 kbps, (b) 100 kbps, (c) 1 Mbps.
Solution
s
2Eb
A2 T
=Q
P2 = Q
(13)
N0
N0
With P2 = 106 we find from tables that
s
A2 T
2
3
= 4.74 A T = 44.9352 10
N0
If the data rate is 10 Kbps, then the bit interval is T = 104 and therefore, the signal amplitude is
A=
44.9352 1010 104 = 6.7034 10
(14)
Similarly we find that when the rate is 105 bps and 106 bps, the required amplitude of the signal is
A = 2.12 102 and A = 6.703 102 A respectively.
Manjunatha. P
(JNNCE)
January 1, 2016
113 / 118
Problems
Problems
Problem 7.47
Consider a digital communication system that transmits information via QAM over a voice-band telephone
channel at a rate 2400 symbols/second. The additive noise is assumed to be white and Gaussian.
1
Determine the [hi No required to achieve an error probability of 105 at 4800 bps.
Repeat (1) for a bit rate of 9600 bps.
Repeat (1) for a bit rate of 19,200 bps.
What conclusions do you reach from these results.
1) The number of bits per symbol is

k =
4800
4800
=
=2
R
2400
Thus, a 4-QAM constellation is used for transmission. The probability of error for an M-ary QAM system with
M = 2k, is
#!2

"s
1
3kEb
PM = 1 1 2 1
Q
(M 1)N0
M
With PM = 105 and k = 2 we obtain
s
Q
Manjunatha. P
(JNNCE)
2Eb
E
= 5 106 = b = 9.7682
N0
N0
January 1, 2016
114 / 118
Problems
Problems
2 If the bit rate of transmission is 9600 bps, then

k =
4800
9600
=
=4
R
2400
In this case a 16-QAM constellation is used and the probability of error is
s
2

1
3 4 Eb
PM = 1 1 2 1
Q
4
15 N0
Thus,
s
Q
1
E
3Eb
= 105 = b = 25.3688
15N0
3
N0
3 If the bit rate of transmission is 19200 bps, then

k =
4800
19200
=
=8
R
2400
In this case a 256-QAM constellation is used and the probability of error is
s
2

1
3 8 Eb
Q
PM = 1 1 2 1
16
255 N0
Thus,
s
Q
Manjunatha. P
(JNNCE)
Eb
1
E
= 105 = b = 659.8922
N0
3
N0
January 1, 2016
115 / 118
Problems
Problems
Complementary Error Function Table

x
erfc(x)
erfc(x)
0.5
0.479500
0.157299
1.5
0.51
0.470756
1.01
0.153190
1.51
0.02
0.977435
0.966159
0.52
0.53
0.462101
0.453536
1.02
1.03
0.149162
0.145216
1.52
1.53
erfc(x)
0.033895
x
2
0.032723 2.01
0.031587 2.02
0.030484 2.03
erfc(x)
0.004678
2.5
0.000407
0.00002209
0.004475 2.51
0.000386
3.01
0.00002074
0.004281 2.52
0.004094 2.53
erfc(x)
0.000365
0.000346
erfc(x)
3.02
3.03
0.00001947
0.00001827
0.04
0.954889
0.54
0.445061
1.04
0.141350
1.54
0.029414 2.04
0.003914 2.54
0.000328
3.04
0.00001714
0.05
0.943628
0.55
0.436677
1.05
0.137564
1.55
0.028377 2.05
0.003742 2.55
0.000311
3.05
0.00001608
0.06
0.07
0.932378
0.921142
0.56
0.133856
1.56
0.025453 2.08
0.003266 2.58
0.000264
3.08
0.00001326
1.59
0.024538 2.09
0.003120 2.59
0.000249
3.09
0.00001243
0.023652
2.1
0.002979
2.6
0.000236
3.07
0.00001508
1.58
0.123197
1.6
0.000278
3.06
0.126674
0.119795
0.003418 2.57
0.000294
1.08
1.09
1.1
0.026397 2.07
0.003577 2.56
0.130227
0.412077
0.404064
0.396144
1.57
0.027372 2.06
1.07
0.58
0.59
0.6
0.420184
1.06
0.909922
0.898719
0.887537
0.57
0.428384
0.08
0.09
0.1
3.1
0.00001414
0.00001165
0.11
0.876377
0.61
0.388319
1.11
0.116467
1.61
0.022793 2.11
0.002845 2.61
0.000223
3.11
0.00001092
0.12
0.865242
0.62
0.380589
1.12
0.113212
1.62
0.021962 2.12
0.002716 2.62
0.000211
3.12
0.00001023
0.13
0.14
0.854133
0.843053
0.63
1.63
0.021157 2.13
0.019624 2.15
0.002361 2.65
0.000178
3.15
0.00000840
1.66
0.018895 2.16
0.002253 2.66
0.000169
3.16
0.00000786
0.018190 2.17
0.002149 2.67
0.000159
3.14
0.00000958
1.65
1.67
0.000189
3.13
0.103876
0.100904
0.098000
0.002475 2.64
0.000200
1.15
1.16
1.17
0.020378 2.14
0.002593 2.63
1.64
0.357971
0.350623
0.343372
1.14
0.110029
0.106918
0.65
0.66
0.67
0.365414
1.13
0.832004
0.820988
0.810008
0.64
0.372954
0.15
0.16
0.17
3.17
0.00000897
0.00000736
0.18
0.799064
0.68
0.336218
1.18
0.095163
1.68
0.017507 2.18
0.002049 2.68
0.000151
3.18
0.00000689
0.19
0.788160
0.69
0.329160
1.19
0.092392
1.69
0.016847 2.19
0.001954 2.69
0.000142
3.19
0.00000644
0.2
0.21
0.777297
0.016210
2.2
0.001863
2.7
0.001692 2.72
0.000120
3.22
0.00000527
0.014422 2.23
0.001612 2.73
0.000113
3.23
0.00000493
1.24
1.25
0.079495
0.077100
1.74
1.75
0.013865 2.24
0.013328 2.25
0.001536 2.74
0.001463 2.75
0.000107
0.000101
3.21
0.00000603
0.014997 2.22
0.295322
0.000127
3.2
1.72
1.73
0.288845
0.001776 2.71
0.000134
0.015593 2.21
0.084466
0.081950
0.74
0.087045
1.7
1.71
1.22
1.23
0.75
1.21
0.089686
0.308567
0.301896
0.734300
0.315335
1.2
0.72
0.73
0.723674
0.71
0.322199
0.755704
0.744977
0.24
0.766478
0.7
0.22
0.23
0.25
3.24
3.25
0.00000564
0.00000460
0.00000430
0.26
0.713100
0.76
0.282463
1.26
0.074764
1.76
0.012810 2.26
0.001393 2.76
0.000095
3.26
0.00000402
0.27
0.702582
0.77
0.276179
1.27
0.072486
1.77
0.012309 2.27
0.001326 2.77
0.000090
3.27
0.00000376
0.28
0.692120
0.78
0.269990
1.28
0.070266
1.78
0.011826 2.28
0.001262 2.78
0.000084
3.28
0.29
0.681717
0.79
0.263897
1.29
0.068101
1.79
0.011359 2.29
0.001201 2.79
0.000080
3.29
0.00000328
0.3
0.671373
0.8
0.257899
1.3
0.065992
1.8
0.010909
0.001143
0.000075
3.3
0.00000306
0.31
0.32
0.661092
0.650874
0.81
0.82
0.251997
0.246189
1.31
1.32
0.063937
0.061935
1.81
1.82
2.3
0.010475 2.31
0.010057 2.32
2.8
0.001088 2.81
0.001034 2.82
0.000071
0.000067
3.31
3.32
0.00000351
0.00000285
0.00000266
0.33
0.640721
0.83
0.240476
1.33
0.059985
1.83
0.009653 2.33
0.000984 2.83
0.000063
3.33
0.00000249
0.34
0.630635
0.84
0.234857
1.34
0.058086
1.84
0.009264 2.34
0.000935 2.84
0.000059
3.34
0.00000232
0.35
0.620618
0.85
0.229332
1.35
0.056238
1.85
0.008889 2.35
0.000889 2.85
0.000056
3.35
0.00000216
0.36
0.610670
0.86
0.223900
1.36
0.054439
1.86
0.008528 2.36
0.000845 2.86
0.000052
3.36
0.00000202
0.37
0.600794
0.87
0.218560
1.37
0.052688
1.87
0.008179 2.37
0.000803 2.87
0.000049
3.37
0.00000188
0.38
0.39
0.590991
0.581261
0.88
0.89
0.213313
0.208157
1.38
1.39
0.050984
0.049327
1.88
1.89
0.4
0.571608
0.9
0.203092
1.4
0.047715
1.9
0.41
0.562031
0.91
0.198117
1.41
0.046148
1.91
0.42
0.43
0.552532
0.543113
0.92
0.193232
1.42
0.044624
1.92
0.006622 2.42
0.000689
0.000046
0.000044
3.38
3.39
0.00000175
0.00000163
2.9
0.000041
3.4
0.00000152
0.000039
3.41
0.00000142
0.000621 2.92
0.041703
1.94
0.006077 2.44
0.000559 2.94
0.000032
3.44
0.00000115
0.040305
1.95
0.005821 2.45
0.000531 2.95
0.000030
3.45
0.00000107
0.005574 2.46
0.000503 2.96
0.000028
3.43
0.00000132
1.44
1.45
1.96
0.000034
3.42
0.006344 2.43
0.183729
0.179109
0.038946
0.000589 2.93
0.000036
1.93
0.94
0.95
1.46
0.043143
0.000763 2.88
0.000725 2.89
0.000654 2.91
0.533775
0.174576
1.43
2.4
0.524518
0.96
0.188437
0.007210
0.44
0.515345
0.93
0.007844 2.38
0.007521 2.39
0.006910 2.41
0.45
0.46
(JNNCE)
erfc(x)
1.000000
0.988717
0.03
Manjunatha. P
0
0.01
3.46
0.00000123
0.00000099
0.47
0.506255
0.97
0.170130
1.47
0.037627
1.97
0.005336 2.47
0.000477 2.97
0.000027
3.47
0.00000092
0.48
0.497250
0.98
0.165769
1.48
0.036346
1.98
0.005108 2.48
0.000453 2.98
0.000025
3.48
0.00000086
0.49
0.488332
0.99
0.161492
1.49
0.035102
1.99
0.004889 2.49
0.000429 2.99
0.000024
3.49
0.00000080
January 1, 2016
116 / 118
Problems
Problems

2
erf(x) =
x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0
3.1
3.2
0
0.00000
0.11246
0.22270
0.32863
0.42839
0.52050
0.60386
0.67780
0.74210
0.79691
0.84270
0.88021
0.91031
0.93401
0.95229
0.96611
0.97635
0.98379
0.98909
0.99279
0.99532
0.99702
0.99814
0.99886
0.99931
0.99959
0.99976
0.99987
0.99992
0.99996
0.99998
0.99999
0.99999
1
0.01128
0.12362
0.23352
0.33891
0.43797
0.52924
0.61168
0.68467
0.74800
0.80188
0.84681
0.88353
0.91296
0.93606
0.95385
0.96728
0.97721
0.98441
0.98952
0.99309
0.99552
0.99715
0.99822
0.99891
0.99935
0.99961
0.99978
0.99987
0.99993
0.99996
0.99998
0.99999
0.99999
2
0.02256
0.13476
0.24430
0.34913
0.44747
0.53790
0.61941
0.69143
0.75381
0.80677
0.85084
0.88679
0.91553
0.93807
0.95538
0.96841
0.97804
0.98500
0.98994
0.99338
0.99572
0.99728
0.99831
0.99897
0.99938
0.99963
0.99979
0.99988
0.99993
0.99996
0.99998
0.99999
0.99999
3
0.03384
0.14587
0.25502
0.35928
0.45689
0.54646
0.62705
0.69810
0.75952
0.81156
0.85478
0.88997
0.91805
0.94002
0.95686
0.96952
0.97884
0.98558
0.99035
0.99366
0.99591
0.99741
0.99839
0.99902
0.99941
0.99965
0.99980
0.99989
0.99994
0.99997
0.99998
0.99999
1.00000
et dt
2
Hundredths digit of x
4
5
0.04511 0.05637
0.15695 0.16800
0.26570 0.27633
0.36936 0.37938
0.46623 0.47548
0.55494 0.56332
0.63459 0.64203
0.70468 0.71116
0.76514 0.77067
0.81627 0.82089
0.85865 0.86244
0.89308 0.89612
0.92051 0.92290
0.94191 0.94376
0.95830 0.95970
0.97059 0.97162
0.97962 0.98038
0.98613 0.98667
0.99074 0.99111
0.99392 0.99418
0.99609 0.99626
0.99753 0.99764
0.99846 0.99854
0.99906 0.99911
0.99944 0.99947
0.99967 0.99969
0.99981 0.99982
0.99989 0.99990
0.99994 0.99994
0.99997 0.99997
0.99998 0.99998
0.99999 0.99999
1.00000 1.00000
6
0.06762
0.17901
0.28690
0.38933
0.48466
0.57162
0.64938
0.71754
0.77610
0.82542
0.86614
0.89910
0.92524
0.94556
0.96105
0.97263
0.98110
0.98719
0.99147
0.99443
0.99642
0.99775
0.99861
0.99915
0.99950
0.99971
0.99983
0.99991
0.99995
0.99997
0.99998
0.99999
1.00000
7
0.07886
0.18999
0.29742
0.39921
0.49375
0.57982
0.65663
0.72382
0.78144
0.82987
0.86977
0.90200
0.92751
0.94731
0.96237
0.97360
0.98181
0.98769
0.99182
0.99466
0.99658
0.99785
0.99867
0.99920
0.99952
0.99972
0.99984
0.99991
0.99995
0.99997
0.99999
0.99999
1.00000
8
0.09008
0.20094
0.30788
0.40901
0.50275
0.58792
0.66378
0.73001
0.78669
0.83423
0.87333
0.90484
0.92973
0.94902
0.96365
0.97455
0.98249
0.98817
0.99216
0.99489
0.99673
0.99795
0.99874
0.99924
0.99955
0.99974
0.99985
0.99992
0.99995
0.99997
0.99999
0.99999
1.00000
ECE 3610 Engineering Probability & Statistics
Figure 58:
Manjunatha. P
(JNNCE)
9
0.10128
0.21184
0.31828
0.41874
0.51167
0.59594
0.67084
0.73610
0.79184
0.83851
0.87680
0.90761
0.93190
0.95067
0.96490
0.97546
0.98315
0.98864
0.99248
0.99511
0.99688
0.99805
0.99880
0.99928
0.99957
0.99975
0.99986
0.99992
0.99996
0.99998
0.99999
0.99999
1.00000
3-1
January 1, 2016
117 / 118
References
S. Haykin, Digital communications, 2nd ed.
Wiley, 1988.
B. Sklar and P. K. Ray, Digital Communications: Fundamentals and Applications, 2nd ed.
Education, 2001.
J. G. Proakis, Digital communications, 4th ed.
Prentice Hall, 2001.
J. G. Proakis and M. Salehi, Communication Systems Engineering, 2nd ed.
Prentice Hall, 2002.
K. N. H. Bhat and D. G. Rao, Digital Communications: A Simppified Approach, 2nd ed.

Technical Publishers, 2005.
Manjunatha. P
(JNNCE)
Pearson
Sanguine
January 1, 2016
118 / 118

Modln Techniques

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Modln Techniques

Загружено:

Авторское право:

Доступные форматы

Digital Modulation Formats:[1, 2, 3, 4]

Digital Modulation Formats

Coherent Binary Modulation Techniques

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Coherent Binary Modulation Techniques

Coherent Binary Modulation Techniques

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Coherent Binary Modulation Techniques

sin x cos y cos x sin y

cos x cos y sin x sin y

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Coherent Binary Modulation Techniques

Modulation is the process by which an information signal is converted to a sinusoidal waveform by

s(t) = 2Arms cost = 2A2rms cost

2A2rms represents the average power P normalized to 1. Therefore s(t)is

Replacing P watts by E joules/T seconds i.e., P = E /T

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Noise in Communication Systems

Noise in Communication Systems

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Gaussian, Normal distribution

Gaussian, Normal distribution

Many things closely follow a Normal Distribution:

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Gaussian, Normal distribution

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

where the factor 2 is included to indicate that Gn (f ) is two sided PSD.

Power spectral density and Autocorrelation of white noise.

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Binary Phase Shift Keying (BPSK)

Binary Phase Shift Keying (BPSK)

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Binary Phase Shift Keying (BPSK)

where Eb is the transmitted signal energy per bit.

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Binary Phase Shift Keying (BPSK)

There is a single basis function 1 (t) and is given by:

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Binary Phase Shift Keying (BPSK)

Figure 6: Signal Space diagram

The message s1 (t) is located at s11 = + Eb s2 (t) is at s21 = Eb .

The distance between message points is 2 Eb .

Digital Modulation Formats:[1, 2, 3, 4]

Coherent Binary Modulation Techniques

Binary Phase Shift Keying (BPSK)

Error Performance for Digital Modulation

Probability of Bit Error

Digital Modulation Formats:[1, 2, 3, 4]

Binary Phase Shift Keying (BPSK)

Figure 8: Receiver Model

Figure 9: BPSK Receiver

Digital Modulation Formats:[1, 2, 3, 4]

Binary Phase Shift Keying (BPSK)