Introduction To Digital Data Transmission

Chapter 4
Introduction to Digital Data

Transmission
Contents
4.1 Modulation . . . . . . . . . . . . . . . . . . . . . . . . 4-4
4.1.1 Modulation Classes . . . . . . . . . . . . . . . . 4-4
4.2 Linear Modulation . . . . . . . . . . . . . . . . . . . . 4-6
4.2.1 Binary Phase-Shift Keying (BPSK) . . . . . . . 4-6
4.2.2 Quadriphase-Shift Keying (QPSK) and Variations 4-10
4.3 Pulse Shaping . . . . . . . . . . . . . . . . . . . . . . 4-14
4.3.1 Raised Cosine Pulse . . . . . . . . . . . . . . . 4-15
4.3.2 Square-Root Raised Cosine . . . . . . . . . . . 4-18
4.4 Complex Baseband Representation . . . . . . . . . . 4-22
4.4.1 Complex Baseband Representation . . . . . . . 4-23
4.5 Signal Space Representation . . . . . . . . . . . . . . 4-24
4.5.1 BPSK . . . . . . . . . . . . . . . . . . . . . . . 4-25
4.5.2 QPSK . . . . . . . . . . . . . . . . . . . . . . . 4-26
4.5.3 Other Two-Dimensional Schemes . . . . . . . . 4-26
4.6 Nonlinear Modulation Techniques . . . . . . . . . . . 4-30
4.6.1 Binary Frequency-Shift Keying . . . . . . . . . 4-30
4-1
CHAPTER 4. INTRODUCTION TO DIGITAL DATA TRANSMISSION
4.6.2 Minimum Shift Keying . . . . . . . . . . . . . . 4-34
4.7 FDMA . . . . . . . . . . . . . . . . . . . . . . . . . . 4-42
4.7.1 Adjacent Channel Interference . . . . . . . . . . 4-43
4.7.2 Power Amplier Nonlinearity . . . . . . . . . . 4-45
4.8 Modulation Comparison . . . . . . . . . . . . . . . . 4-48
4.8.1 Linear Channel . . . . . . . . . . . . . . . . . . 4-48
4.8.2 Nonlinear Channel . . . . . . . . . . . . . . . . 4-49
4.9 Channel Estimation and Tracking . . . . . . . . . . . 4-50
4.9.1 Differential Detection . . . . . . . . . . . . . . 4-51
4.9.2 Pilot Transmission . . . . . . . . . . . . . . . . 4-53
4.10 Receiver Bit Error Probability Performance . . . . . 4-55
4.10.1 AWGN Channel . . . . . . . . . . . . . . . . . 4-56
4.10.2 Frequency Flat, Slow fading Channel . . . . . . 4-58
4.11 Theme Example: OFDM . . . . . . . . . . . . . . . . 4-60
4.11.1 Cyclic Prex . . . . . . . . . . . . . . . . . . . 4-63
4.12 Theme Example: Cordless Telephone . . . . . . . . . 4-64
4-2 ECE 5625 Communication Systems I
This chapter introduces a variety digital modulation schemes found
in modern wireless systems. The block diagram of a generic band-
pass wireless link is shown below.
Baseband
Processor
Baseband
Processor
Demodulator
Modulator
d(t )
d(t )
s(t )
s(t )
s
t
(t )
s
t
(t )
f
S
t
( f )
f
c
f
c
f
S( f )
0
f
f
c
f
c
N
0
/2
S
t
( f )
f
0
S( f )
C
h
a
n
n
e
l
Generic bandpass wireless system
By its very nature, sending digital information involves a mes-
sage signal that takes on only a nite number of values. At the
waveform level the encoded digital message signal can be a con-
tinuous function time, t . The data signal J(t ) may likely speech that
has been digitally encoded using a compression algorithm.
ECE 5625 Communication Systems I 4-3
4.1 Modulation
The process of modulation varies some aspect of the carrier wave
with respect to the modulating signal, e.g., J(t ). Demodulation per-
forms the opposite operation, so as to gain recover an estimate of the
information bearing signal s(t ).
Three practical benets of modulation include:
1. Shift the spectral content to an operating frequency band that
is easily transmitted and received
2. The modulation operation may make the signal less vulnerable
to noise and interference, e.g., frequency modulation
3. The scheme easily supports the use of a multiple-access tech-
nique
Generic Carrier Model
c(t ) =
c
cos(2
c
t 0)
4.1.1 Modulation Classes
Modulation schemes can be compared via a number of classica-
tions. Three such classications are briey explored before detailed
investigations of selected schemes begins.
Linear vs Nonlinear
Two basic classes of modulation are linear and nonlinear. If the
principle of superposition holds the modulation is linear.
4.1. MODULATION
The attributes of both linear and nonlinear modulation are ex-
plored in the remainder of this chapter.
Analog vs Digital
Modulation can also be classied along the lines of analog and dig-
ital. Here analog means the message signal, m(t ) is a continuous
function of time, so the modulated carrier, s
t
(t ), has attributes vary
continuously over some parameter range.
With digital modulation the message signal, J(t ), takes on dis-
crete values, e.g., 1 with the switching instants occurring at time
interval T
s
. The modulated carrier s
t
(t ) will still be a function of
continuous varibale t , but the values the signal takes on will have a
discrete nature, e.g., amplitude, frequency, or phase.
The focus in this chapter will be on digital modulation schemes.
Amplitude vs Angle
The particular attribute of the carrier signal c(t ) that is varied forms
another basis for modulation classication. Two broad classes are
amplitude and angle modulation. Angle modulation further breaks
down into phase and frequency modulation.
v Amplitude modulation simply requires the carrier amplitude
c
to vary linearly with respect to m(t ) or J(t )
v With angle modulation we consider the entire argument of cos( )
in c(t ) as the angle
[(t ) = 2
c
t 0(t )
of the carrier, and design a modulator so that it varies linearly
with m(t ) or J(t )
Specically with frequency modulation, the derivative J[(t ),Jt
is made to vary linearly with the message signal
Specically with phase modulation the carrier phase 0(t )
is made to vary linearly with the message
4.2 Linear Modulation
4.2.1 Binary Phase-Shift Keying (BPSK)
To create BPSK we rst construct a baseband data signal of the form
J(t ) =
k
b
k
(t kT )
where b
k
is a bipolar bit sequence of the form
b
k
=
_
1. binary symbol 1
1. binary symbol 0
v A fundamental pulse shape (t ) is the rectangle shape
(t ) =
_
1. 0 _ t _ T
0. otherwise
p(t )
1.0
0 T
t
4.2. LINEAR MODULATION
v BPSK constitutes a form a digital phase modulation in the car-
rier phase is switched between 0(t ) = 0 and radians depend-
ing upon the sign of b
k
v Since a phase of radians simply changes the sign of the car-
rier signal, we observe that for the case of BPSK
s
t
(t ) = J(t )c(t )
which is of the same form as double sideband suppressed car-
rier (DSB-SC) modulation
d(t )
s
t
(t )
t
t
1 2 3 4 5 6
1
0.5
0.5
1
1 2 3 4 5 6
1
0.5
0.5
1
T = T
b
= 1
BPSK using a rectangle pulse
v The power spectral density of BPSK can be shown to be of the
form
S
t
( ) =

2
c
4T
_
[1(
c
)[
2
[1(
c
)[
2
_
where 1( ) = F
_
(t )
_
v For the rectangular pulse shape
1( ) = T
sin(2 T )
2 T
= T sinc( T )
where sinc(.) sin(.),(.)
f
c
f
c
f
S
t
( f )
2/T 2/T
T A
2
c
4
BPSK spectrum
v For a rectangular pulse shape the main lobe bandwidth, also
known as the RF bandwidth is T
RF
= 2,T = 21
b
, where
1
b
= 1,T is the bit rate
v Another useful bandwidth measure is the fractional contain-
ment bandwidth, T
, dened as
1
=
_
c
T
,2
c
T
,2
S
t
( ) J
_
o
0
S
t
( ) J
where 1
is a fraction of the total power in s

t
(t )
Example 4.1: BPSK with Rectangle Pulse Shape T
v The fractional containment bandwidth of rectangle pulse shaped

BPSK can be found from
1
=
_
T
,2
0
sinc
2
( T ) J
_
o
0
sinc
2
( T ) J
v We rst work with the denominator using Parsevals theorem
_
o
0
sinc
2
( T ) J =
1
2
_
o
o
sinc
2
( T ) J
=
1
2T
_
o
o
sinc
2
(.) J. =
1
2T
v Inserting in the numerator, and changing variables, we have
1
= 2
_
T
T,2
0
sinc
2
(.) J.
v At T
T = 2 the contained power is only 0.9028, while at

T
T = 5 the contained power has only increased to 0.9592

v The 99% containment bandwidth occurs when T
T = 20.57
v Clearly pulse shaping beyond the rectangular pulse shape is
needed
2 4 6 8 10
0.05
0.1
0.15
0.2
0.25
1
P
f
Normalized RF Containment Bandwidth B
f
T
90%
95%
Rect pulse BPSK fractional out-of-band power
4.2.2 Quadriphase-Shift Keying (QPSK) and Vari-
ations
To make more effective use of the available spectrum we may choose
to modulate both the sin and cos carriers (quadrature multiplexing).
v Consider a binary data stream J(t ) demultiplexed into two
equal bit rate streams, J
1
(t ) and J
2
(t ) respectively
J(t ) =
k
b
k
(t kT
b
)
J
1
(t ) =
k
b
2k
(t kT
s
)
J
2
(t ) =
k
b
2k1
(t kT
s
)
where T
s
= 2T
b
is the symbol duration, which is twice the bit
duration
Note that the symbol rate is 1
s
= 1,T
s
= 1
b
,2
With QPSK we transmit two bits per symbol
v The data streams are applied to orthogonal, but at the same
frequency, carriers
s
t
(t ) = s
1
(t ) s
2
(t )
s
t
(t ) =
c
_
J
1
(t ) cos(2
c
t ) J
2
(t ) sin(2
c
t )
_
1:2
Demux
d(t )
d
1
(t )
d
2
(t )
s
1t
(t )
s
2t
(t )
s
t
(t )
A
c
cos(2 f
c
t )
A
c
sin(2 f
c
t )
QPSK
QPSK modulator block diagram
Standard QPSK
With this form of QPSK it looks like we have two equal bit rate
BPSK modulators operating in parallel.
v The two carrier signals are aid to be in phase quadrature since
the sine lags the cosine by 90

v The signal J
1
(t ), modulating the cosine carrier, is known as
the in-phase signal or 1 component, while the signal J
2
(t ),
modulating the sine carrier, is known as the quadrature signal
v Assuming equal bit rates and pulse shaping on the 1 and Q
components, the spectrumof QPSKis identical to that of BPSK,
that is
S
t
( ) =

2
c
2T
s
_
[1(
c
)[
2
[1(
c
)[
2
_
Note the scale factor difference from BPSK because we have
two equal power carriers forming the QPSK signal.
v Rectangular pulse shaping the main lobe RF bandwidth is now
T
RF
= 21
s
= 1
b
, and bandwidth reduction of 2 compared
with BPSK
v Since each carrier phase is modulated between 0
and 180
,
and are in phase quadrature, composite carrier phase takes on
the four values of {45
. 135
. 225
. 315
]
v A specic property of standard QPSK is that the composite
carrier phase may undergo phase jumps of 0
, 90
, or 180
every 2T
b
= T
s
seconds
2 4 6 8 10
1
0.5
0.5
1
2 4 6 8 10
1
0.5
0.5
1
2 4 6 8 10
1
0.5
0.5
1
d
1
(t )
d
2
(t )
t
(symbols)
t
t
s
t
(t )
2
T
s
= 1
QPSK using a rectangle pulse
Offset QPSK (OQPSK)
The 180
phase jumps in QPSK can be a problem when the signal

is ltered and then amplied by a nonlinear power amplier.
v OQPSK, also known as staggered QPSK, is formed by delay-
ing the quadrature signal bit stream by T
s
,2 = T
b
, thus limit-
ing phase jumps to just 0
and 90
v The transmitted signal now takes the form

s
t
(t ) =
c
_
J
1
(t ) cos(2
c
t ) J
2
(t T
s
,2) sin(2
c
t )
_
v The waveforms change accordingly
2 4 6 8 10
1
0.5
0.5
1
d
1
(t ) t
(symbols)
t
t
s
t
(t )
2
2 4 6 8 10
1
0.5
0.5
1
2 4 6 8 10
1
0.5
0.5
1
d
2
(t T
s
/2)
T
s
= 1
OQPSK using a rectangle pulse
4.3 Pulse Shaping
For both BPSK and QPSK we have seen how the rectangular pulse
shape, while easy to implement, creates a wide spectral footprint in
the neighborhood of the carrier frequency,
c
. We now consider the
use of pulse shaping or a premodulation lter to better match the
transmitted signal spectrum to the available channel bandwidth. We
specically desire:
4.3. PULSE SHAPING
1. A more compact spectrum to allow more digitally modulated
carriers to occupy a frequency band allocation
2. A means to manage channel induced bandlimiting, e.g., due
to multipath, which introduces intersymbol interference (ISI),
which occurs when the energy of previous symbols interferes/overlaps
with the energy of the present symbol (pulse)
4.3.1 Raised Cosine Pulse
v A premodulation lter and/or pulse shaping satises, in part,
both of the above requirements
v A popular class of pulse shaping that achieves both band lim-
iting and ISI control is the raised cosine (RC) pulse
v The RC pulse has a spectrum with adjustable bandwidth in the
RF (two-sided) sense running from W = 1
b
to 2W = 21
b
v The pulse spectrum (Fourier transform) is dened by
1( ) =
_
_
1
2W
. 0 _ [ [ <
1
1
4W
_
1 cos
_

2Wj
_
[ [
W(1 j)
_
__
.
1
_ [ [ < 2W
1
0. otherwise
where
1
sets the edge of the at portion of the spectrum and
is related to the roll-off factor j and W via
0 _ j = 1

1
W
_ 1
v The parameter j (elsewhere denoted ) controls the excess
bandwidth relative to the minimum value of W = 1
b
(one-
sided W/2) when j = 0
1 0.5 0.5 1
0.2
0.4
0.6
0.8
1
Normalized Frequency f /W
P( f )
= 0, 1/2, 1
RC spectrum, 1( ), for various j values
v The RC pulse gets its name from the 1 cos() term in the
spectrum denition
v The corresponding RC pulse, (t ), can be obtained by inverse
Fourier transforming 1( )
(t ) = F
1
{1( )] =
_
cos(2jWt )
1 16j
2
W
2
t
2
_
sinc(2Wt )
v The zero ISI property of the RC pulse is that although (0) =
1, (nT ) = 0 for . . . . 2. 1. 1. 2. . . .
4.3. PULSE SHAPING
3 2 1 1 2 3
0.2
0.2
0.4
0.6
0.8
1
p(t )
Normalized Time t /T
= 0, 1/2, 1
RC pulse, (t ), for various j values
Example 4.2: RC Waveform for a Bit Pattern
Consider the waveformproduced by the bit sequence {0. 0. 1. 1. 0. 1]
or in bipolar form {1. 1. 1. 1. 1. 1].
t
Individual pulses
forming seq.
..,-1,-1,1,1,-1,1,..
Note zero ISI
4 2 2 4
1
0.5
0.5
1
Waveform created with bit pattern -1,-1,1,1,-1,1
t
Composite of
pulses
forming seq.
..,-1,-1,1,1,-1,1,..
4 2 2 4
1
0.5
0.5
1
Composite waveform created with bit pattern -1,-1,1,1,-1,1
4.3.2 Square-Root Raised Cosine
The zero ISI response holds for the RC pulse, but optimal ltering in
an additive noise environment, requires that ltering/pulse shaping
be distributed between the transmitter and receiver.
v The root raised-cosine or square-root raised-cosine (SRC) l-
ter satises this requirement
SRC
Tx
Channel SRC
Rx
WGN
SRC (also denoted RRC) used as a Tx/Rx lter pair
4.3. PULSE SHAPING
v The spectrum at the receiver becomes the 1
2
( ), which for
the case of the SRC is again the RC
v The SRC pulse spectrum is dened as
1
SRC
( ) =
_
_
1
_
2W
. 0 _ [ [ <
1
1
_
2W
cos
_

4Wj
_
[ [
W(1 j)
_
_
.
1
_ [ [ < 2W
1
0. otherwise
1 0.5 0.5 1
0.2
0.4
0.6
0.8
1
Normalized Frequency f /W
P( f )
= 0, 1/2, 1
SRC spectrum, 1( ), for various j values
v The corresponding RC pulse, (t ), can be obtained by inverse
Fourier transforming 1( )
(t ) = F
1
{1( )]
=
_
2W
1 64j
2
W
2
t
2
_
sin(2W(1 j)t )
2Wt
4j
cos
_
2W(1 j)t
_
_
v The SRC pulse does not have a zero ISI property as with the
RC, but it does have an orthogonality condition
_
o
o
(t )(t nT ) Jt = 0 for n = 1. 2. . . .
3 2 1 1 2 3
0.2
0.2
0.4
0.6
0.8
1
1.2
p(t )
Normalized Time t /T
= 0, 1/2, 1
SRC pulse, (t ), for various j values
4.3. PULSE SHAPING
Example 4.3: SRC Waveform for a Bit Pattern
Consider the waveformproduced by the bit sequence {0. 0. 1. 1. 0. 1]
or in bipolar form {1. 1. 1. 1. 1. 1].
4 2 2 4
1
0.5
0.5
1
Individual pulses
forming seq.
..,-1,-1,1,1,-1,1,..
Note not zero ISI
t
Waveform created with bit pattern -1,-1,1,1,-1,1
t
Composite of
pulses
forming seq.
..,-1,-1,1,1,-1,1,..
4 2 2 4
1.5
1
0.5
0.5
1
Composite waveform created with bit pattern -1,-1,1,1,-1,1
v Notice that the zero ISI condition is not met until the above
waveform is passed through an SRC lter in the receiver
v Also note that the equivalent transmit waveform (gure imme-
diately above) requires a greater dynamic range than the corre-
sponding RC waveform
4.4 Complex Baseband Representation
As we continue to study various digital modulation schemes, we are
motivated to consider the complex envelope representation. To es-
tablish the equivalence of the complex envelope form we start with
the so-called band-pass signal canonical form:
s(t ) = s
1
(t ) cos(2
c
t ) s
Q
(t ) sin(2
c
t )
v Here the in-phase modulating signal is denoted s
1
(t ) and the
quadrature modulating signal is denoted s
Q
(t )
v The complex envelope is the complex signal
s(t ) = s
1
(t ) s
Q
(t )
v The relationship between s(t ) and s(t ) is obtained by noting
that
s(t ) = Re
s(t )e
2
c
t
_
where exp( 2
c
t ) = cos(2
c
t ) sin(2
c
t )
v When dealing with the complex envelope the carrier frequency
is effectively suppressed
v The complex envelope (complex baseband) form is also lends
itself to simplied computer simulation and actual hardware
implementations in ASIC/FPGA/general purpose DSP
4.4. COMPLEX BASEBAND REPRESENTATION
v The complex envelope representation suggests the following
IQ modulator/demodulator structures:
Synth.
Osc.
90
IQ Modulator
s
I
(t )
s
Q
(t )
s(t )
cos(2 f
c
t )
Synth.
Osc.
90
LPF or
SRC
LPF or
SRC
s(t )
s
I
(t )
s
Q
(t )
Recovered
IQ Signals
IQ Demodulator
2 sin(2 f
c
t )
2 cos(2 f
c
t )
sin(2 f
c
t )
IQ Mod/Demod
4.4.1 Complex Baseband Representation
We can carry the complex envelope idea further by modeling band-
pass ltering in terms of a complex baseband impulse response.
v The impulse response of a bandpass lter can be written as
h(t ) = Re
_
h(t )e
2
c
t
_
where
h(t ) = h
1
(t ) h
Q
(t )
v Consider the bandpass signal .(t ) ltered by bandpass lter
h(t ) to produce ,(t )
,(t ) = .(t ) + h(t ) =
_
o
o
.(z)h(t z) Jz
v Using complex envelopes and the complex baseband impulse
response, we can equivalently write
,(t ) =
1
2
.(t ) +

h(t ) =
1
2
_
o
o
.(z)
h(t z) Jz
v Once ,(t ) is obtained we can return to ,(t ) via
,(t ) = Re
,(t )e
2
c
t
_
4.5 Signal Space Representation
As modulation schemes become more complex, a signal space rep-
resentation becomes convenient for performance analysis purposes.
Two dimensional signal constellations have been studied extensively.
v Traditionally the coordinate system is established via energy
normalized version of the in-phase and quadrature signals, de-
noted
1
(t ) and
2
(t )
1
(t ) =
_
2
T
cos(2
c
t ). 0 _ t _ T
2
(t ) =
_
2
T
sin(2
c
t ). 0 _ t _ T
4.5. SIGNAL SPACE REPRESENTATION
v Two immediate properties of this construction approach are:
1. Orthogonality
_
T
0
1
(t )
2
(t ) Jt = 0
2. Unit length/unit energy
_
T
0
[
1
(t )[
2
Jt =
_
T
0
[
2
(t )[
2
Jt = 1
v The functions (basis functions)
1
(t ) and
2
(t ) constitute an
orthonormal set
v With these two functions a variety of digital modulation schemes
can be represented
4.5.1 BPSK
BPSK requires just a single dimension, say
1
(t ) to describe it signal
constellation. Notice that in signal space the length squared has units
of signal energy per bit, 1
b
(or symbol 1
s
).
1
(t )
E
b

E
b
0
'0' '1'
BPSK 1-D signal space
4.5.2 QPSK
QPSK and it variations require two dimensions. Note 1
s
= 21
b
.
E
b

E
b
2E
b
2E
b
2E
b
2E
b
1
(t )
2
(t )
2
E
b
1
(t )
2
(t )
'00' '11'
'01'
'10'
E
b
E
b
QPSK 2-D signal space: two variations
4.5.3 Other Two-Dimensional Schemes
A very large number of digital modulation schemes can be reduced
to a two-dimensional signal space. Two schemes found in wireless
systems are M-ary PSK (MPSK) and quadrature-amplitude modu-
lation (QAM).
MPSK
With BPSKand QPSKdened, the logical extension is MPSKwhich
encodes the transmission bits into one of M phase values.
v The values that M takes on are 2, 4, 8, 16, 32, . . .
v The number of bits per symbol is log
2
(M), so considering the
null-to-null bandwidth, MPSK has bandwidth efciency com-
pared to BPSK of
T
eff
=
1
b
2
log
2
(M) bits/s/Hz
where 1
b
is the serial bit rate in bits/s
v The signal space of MPSK is 2-dimensional with adjacent sig-
nal points moving closer together as M increases for a xed
1
b
v Notice that the bits are encoded using Gray coding which in-
sures that two adjacent symbols differ by no more than one
bit
2
E
b
2
(t )
1
(t )
M = 8
2
M
'000'
'001'
'011'
'111'
'110'
'010'
'110'
'100'
decision
region
MPSK with M = 8
v Note that the point in signal space where a receive signal must
lie in order to make a symbol decision (decision regions), are
pie shaped in the case of MPSK
QAM
If we include amplitude modulation along with phase modulation
we maintain a 2-D signal space, but now allow a much denser array
of signal points. The new modulation scheme is known as quadra-
ture amplitude modulation (QAM). At complex baseband the gen-
eral form of QAM is
s(t ) =
o
k=o
a
k
(t kT
s
)
o
k=o
b
k
(t kT
s
)
where a
k
and b
k
are the in-phase and quadrature amplitude values of
the QAM symbols taking on values of 1. 3. 5. . . .
2
(t )
1
(t )
M = 16
1/2 3/2
1/2
3/2
'1101' '1100'
'0011'
decision
region
decision
region

decision
region
QAM with M = 16 (16QAM)
v With QAM symbol decisions are made with different size de-
cision regions
v The fact that information is carrier in the amplitude and the
phase also means that QAM is more sensitive to nonlinearities
4.6 Nonlinear Modulation Techniques
A large class of modulation schemes employ nonlinear signal pro-
cessing in the modulation process. For these schemes it is convenient
to represent the transmitted signal in polar form, that is
s(t ) =
s
1
(t ) s
Q
(t ))e
2
c
t
_
= a(t ) cos2
c
t 0(t )|
where
a(t ) =
_
s
2
1
(t ) s
2
Q
(t )
0(t ) = tan
1
_
s
Q
(t )
s
1
(t )
_
4.6.1 Binary Frequency-Shift Keying
To create a digital FM scheme we need only make the instantaneous
frequency of the carrier a function of the transmission bit stream.
With binary frequency-shift keying the carrier switches between two
frequencies accoding to the transmitted bit type 0 or 1, e.g.,
s(t ) =
_
_
_
_
21
b
T
cos(2
1
t ). 0 _ t _ T
b
. 1s transmission
_
21
b
T
cos(2
2
t ). 0 _ t _ T
b
. 0s transmission
where
1
and
2
are the BFSK tone frequencies and may be mod-
eled as a random variable.
v The choice of
1
and
2
, or better yet [
2

1
[ is a design
parameter
4.6. NONLINEAR MODULATION TECHNIQUES
v In Sundes FSK
1
we choose
i
=
n
c
i
T
. i = 1. 2
with n
c
a xed integer and assume that = 0
v As each successive bit is transmitted phase continuity is main-
tained, thus this variation of FSK is known as continuous-
phase FSK (CFSK)
v When
i
is chosen as described above, it turns out that
_
T
0
_
21
b
T
cos(2
i
t )
_
21
b
T
cos(2
t ) Jt = 0. i =
v The corresponding signal space orthonormal basis functions
are
i
(t ) =
_
_
_
_
2
T
cos(2
i
t ). 0 _ t _ T
0. otherwise
v M-ary FSM (MFSK) can be created by using more orthogo-
nal frequencies, hence the dimensionality of the signal space
increases proportionately
1
(t )
E
b
0
'0'
'1'
E
b
2
1
Haykin and Moher p. 132.
Orthogonal BFSK signal space
BFSK Power Spectrum
The exact power spectrum of BFSK is difcult to obtain, but an ap-
proximate spectrum can be obtained using a quasi-static analysis.
This analysis assumes the power spectrum can be formed as a super-
position of the power spectrum due to on-off keying of the individual
carriers
1
and
2
.
v Assuming a rectangular pulse shape for (t ), which is also
the easiest to implement in simple FSK hardware, we have the
spectrum for binary on-off keying of a carrier at
i
as given by
S
i
( ) =
1
2

2
c
4T
_
[1(
i
)[
2
[1(
i
)[
2
_
where the rst 1/2 factor is the duty cycle associated with equally
likely 1s and 0s and as before 1( ) = T sinc( T )
v Since we have two carrier frequencies in BFSK, superposition
yields the approximate BFSK spectrum
S
t
( ) .

2
c
8T
_
[1(
1
)[
2
[1(
1
)[
2
[1(
2
)[
2
[1(
2
)[
2
_
v This approximation is best when [
2

1
[ is large compared
with the bit rate 1
b
v At complex baseband the power spectrum is of the form
S
T
( ) .

2
c
4T
_
[1( ^ )[
2
[1( ^ )[
2
_
where ^ = [
2

1
[,2
4 2 0 2 4
100
80
60
40
20
Normalized Frequency f /f
P
S
D

(
d
B
)
R
b
= 0.05f
Baseband BFSK power spectrum in dB for 1
b
= 0.05^
4 2 0 2 4
60
50
40
30
20
Normalized Frequency f /f
P
S
D

(
d
B
)
R
b
= 0.5f
Baseband BFSK power spectrum in dB for 1
b
= 0.5^
4.6.2 Minimum Shift Keying
v Recall that with CPFSK the carrier phase is continuous from
bit-to-bit.
v Rather than using two frequencies
1
and
2
we may write s(t )
in terms of
c
and phase 0(t )
s(t ) =
_
21
b
T
cos
_
2
c
t 0(t )
_
v With CPFSK the phase linearly ramps up or down during each
bit time as
0(t ) = 0(0)
h
T
. 0 _ t _ T
where h is the CPFSK modulation index or deviation ratio and
0(0) is the accumulated excess carrier phase up to time t = 0
v A phase trellis can be constructed by noting the possible phase
trajectories of 0(t ) 0(0) for t _ 0
t
T 2T 3T 4T 5T 6T 7T 8T
0
h
2h
3h
4h
4h
3h
2h
h
CPFSK phase trellis for arbitrary h
v For the special case of h = 1,2 we have what is know as
minimum-shift keying (MSK)
v The term minimum is used because when h = 1,2 the fre-
quency deviation between the FSK tones is given by
1

2
=
_
h
2T
_
h
2T
_
=
h
T
=
1
2T
=
1
b
2
v We see that setting the frequency difference between two FSK
tones to 1
b
,2 (h = 1,2) is the minimum tone that allows
the two tones to remain orthogonal and create the minimum
spectral bandwidth
v The signal space basis functions for MSK are typically ex-
pressed as
1
(t ) =
_
2
T
cos
_

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

2T
t
_
sin(2
c
t )
then it can be shown that
s(t ) = s
1
1
(t ) s
2
2
(t )
where s
1
and s
2
represent parallel data bits modulating I-Q car-
riers
v This expansion not only shows us that MSK can be generated
using a standard I-Q modulator using half-sine pulse shaping
v It is also clear that the IQ phase trajectories follow the circum-
ference of a circle traversing 90
each serial bit period T

! "! #! $! %! &!
!"
!
"
! "! #! $! %! &!
!"
!
"
s
I
(
t
)
s
Q
(
t
)
Serial Bit Normalized Time
MSK complex baseband IQ waveforms
MSK Power Spectrum
The complex baseband power spectral density of MSK can be shown
to be
S
T
( ) =
321
b
2
_
cos(2T )
16T
2
2
1
_
2
where T
b
is the serial bit duration as opposed to the parallel symbol
duration which is 2T
b
.
2 1 0 1 2 3
60
40
20
0
P
S
D

(
d
B
)
Normalized Frequency f /R
b
3
4
R
b
MSK baseband power spectrum
MSK Properties
v Constant envelope
v Relatively narrow bandwidth
Wide main lobe than QPSK
Faster sidelobe rolloff rate than QPSK
v Coherent receiver performance equivalent to QPSK (more on
this later)
Gaussian MSK
The spectrum of MSK is more compact that rectangular pulse shape
BPSK/QPSK, but is still not suitable for multiple access communi-
cations.
v Another feature of MSK is that it can be generated by di-
rect frequency modulation using a voltage controlled oscillator
(VCO)
v In particular it is possible to apply preltering to a rectangular
pulse shaped binary message stream
v With Gaussian MSK (GMSK) this shaping lter has a Gaus-
sian impulse response and a Gaussian frequency response
h(t ) =
_
2
ln 2
W exp
_
2
2
ln 2
W
2
t
2
_
H( ) = exp
_
ln 2
2
_
W
_
2
_
where W is the lter 3 dB bandwidth
Gaussian
LPF,
BW = W
FM Mod
(VCO)
s(t )
f
c
+1
1
Binary Bit Stream
T
Equivalent
Mod Index
of h = 1/2
GMSK modulator
v When this premodulation shaping lter is driven a rectangu-
lar pulse of duration T , the effective frequency pulse shaping
driving the FM modulator is of the form
g(t ) =
1
2
_
erfc
_
_
2
ln 2
W T
_
t
T
1
2
_
_
erfc
_
_
2
ln 2
W T
_
t
T
1
2
_
__
where erfc(.) is the complementary error function dened as
erfc(.) =
2
_
_
o
.
exp(:
2
) J:
v The design parameter controlling the spectral bandwidth of
GMSK is the time-bandwidth product W T (more commonly
TT )
v Note that W T o(no ltering) results in MSK
v There is no closed-form expression for the power spectral den-
sity of GMSK, so we have to resort to simulation
! "! #! $! %! &!
!"
!
"
! "! #! $! %! &!
!"
!
"
s
I
(
t
)
s
Q
(
t
)
Serial Bit Normalized Time
GMSK complex baseband IQ waveforms m
!! !" !# $ # " !
!%$
!&$
!'$
!"$
$
Bit Normalized Frequency
MSK
P
o
w
e
r

S
p
e
c
t
r
u
m

(
d
B
)
WT = 0.35 WT = 0.25
GMSK baseband power spectrum
4.7 FDMA
Frequency division multiple access (FDMA) was briey described
in Chapter 1 as means to allow multiple users to communicate si-
multaneously. The frequency division duplex (FDD) scheme used
with the PCS bands is shown below.
... ...
Uplink Bandwidth Downlink Bandwidth
1 2 N 1 2 N
B
T
B
T
Guard band
(e.g., PCS blk A is
1850-1865 MHz)
(e.g., PCS blk A is
1930-1945 MHz)
FDD/FDMA channel allocations as used in the PCS band
v Guard bands are provided as buffer zones, so the power radi-
ated outside the assigned band stays between 6080 dB below
the in-band signal
v Common terminology is to refer to the base station to mobile
link as the forward-link or downlink; while the mobile to base
station link is referred to as the reverse-link or uplink
v When FDD is employed uplink and down link communica-
tions can occur simultaneously, meaning that the radio hard-
ware needs a diplexer to separate the two signal paths
4.7. FDMA
Diplexer
Modulator
Demodulator
Tx
Rx
Special purpose
microwave filter
Data
Data
Diplexer used in an FDD enabled radio
v One downside of FDMA is that the base station must have a
dedicated transmitter and receiver for each carrier frequency
in the downlink and uplink bandwidths, respectively
v Since the bandwidth assigned to each channel is narrow, the
fading encountered tends to be at
v Doppler spreading is still an issue however
4.7.1 Adjacent Channel Interference
One serious threat to FDMA is adjacent channel interference, that
is the performance degradation resulting from spectral energy from
adjacent channels leaking into the desired channel.
v Earlier we talked about bandwidth efciency of PSK modula-
tion, in bits/s/Hz
v A higher spectral efciency means that more channels can be
packed into bandwidth T
T
v Adjacent channel intereference (ACI) is particularly a problem
for non-bandlimited signals, such as rectangular pulse shaped
PSK
0 2 4 6 8
40
30
20
10
0
1 0 1 2 3 4
40
30
20
10
0
C
h
0
C
h
1
C
h
2
C
h
0
C
h
1
C
h
2
( f f
c
)/T
Rect
pulse
( f f
c
)/T
RC
pulse
S
p
e
c
t
r
u
m
(
d
B
)
S
p
e
c
t
r
u
m
(
d
B
)
ACI comparison for Rect and RC signals
4.7. FDMA
4.7.2 Power Amplier Nonlinearity
ACI as just discussed, limits how close signals can be spaced. RF
power amplier nonlinearity also factors into ACI as spectral re-
growth and intermodulation may occur.
v In portable electronics the efciency of the RF power amplier
factors into battery lifetime
v Ampliers that are more power efcient also tend to be more
nonlinear
v There are means to linearize nonlinear ampliers, and this is
an active area of research and development topic
v Apower amplier will have both a non-constant gain and phase
versus the input power level, that is also a function of fre-
quency
27.5
25.0
25.5
26.0
26.5
27.0
2 0 -2 -4 -6 -8 -10 -12 -14
Input Level (dBm)
A
m
p
l
i
e
r

G
a
i
n

(
d
B
)
1643 MHz
1626 MHz
Amplitude characteristics of a nonlinear amplier
2 0 -2 -4 -6 -8 -10 -12 -14
Input Level (dBm)
40
-80
-60
-40
0
20
A
m
p
l
i
e
r

P
h
a
s
e

(
d
e
g
)1643 MHz
1626 MHz
Phase characteristics of a nonlinear amplier
v In the above sample plots we see that when the input kept be-
low about -8 dBm, both the amplitude and the phase remain
constant
v For larger input levels we have both AM-to-AM and AM-to-PM
distortion taking place
v By operating the amplier below a particular input level, dis-
tortion can be minimized
v One measure of amplier saturation is when the amplier gain
is reduced by 1 dB, the so-called 1dB compression point
v If we dene the input and output 1 dB compression points at
V
in, sat
and V
out, sat
respectively, we can dene the amplier in-
4.7. FDMA
put back-off and output back-off respectively as
Input back-off = 10 log
10
_
V
in, rms
V
in, sat
_
2
Output back-off = 10 log
10
_
V
out, rms
V
out, sat
_
2
v As we get closer to the 1dB compression point more distortion
is introduced
10 log
10
[V
2
out, sat
]
10 log
10
[V
2
out, rms
]
1
0
l
o
g
1
0
[
V
2
i
n
,
r
m
s
]
1
0
l
o
g
1
0
[
V
2
i
n
,
s
a
t
]
Input Power (dBm)
O
u
t
p
u
t

P
o
w
e
r

(
d
B
m
)
B
a
c
k
o
f
f
Backoff
Idealized amplier AM-to-AM characteristic
4.8 Modulation Comparison
Modulation schemes can be compared on the basis of spectral ef-
ciency and power efciency. Spectral efciency as we know is mea-
sures in bits/s/Hz. The channel context is important however.
v Three relevant factors include:
1. Pulse shaping (rect, RC, SRC, Gaussian, etc.)
2. Other ltering (Image rejection and amplier spurious re-
sponses)
3. Presence of nonlinearities (power ampliers operating close
to saturation)
4.8.1 Linear Channel
For a linear channel we are just interested in the ideal transmitted
spectra, that is with only linear ltering at most.
v We have considered both linear and nonlinear modulation schemes
v Rectangular pulse shaping is known to be inefcient due to the
slow sidelobe roll-off rate
v MSK, with the half-sine pulse shape is better, but still has a
wide bandwidth
v GMSK is much better than MSK for spectral efciency, but
power efciency is not as good
v QPSK with RC or SRC shaping is very efcient, and is good
choice for many applications
4.8. MODULATION COMPARISON
0.5 1 1.5 2 2.5 3 3.5 4
50
40
30
20
10
0
Normalized Frequency - ( f f
c
)/T
S
p
e
c
t
r
u
m
(
d
B
)
MSK
QPSK- rect RC-QPSK
= 0.5
Comparison spectra: QPSK, MSK, RC-QPSK with j = 0.5
4.8.2 Nonlinear Channel
Anonlinear transmit power amplier can effectively destroy the spec-
tral bandlimiting achieved through linear ltering. Spectral side-
lobes tend to regrow depending upon how close the amplier is to
saturation.
v Constant envelope modulation schemes such as QPSK with
rectangular pulse shaping, MSK, and GMSK, are in theory un-
affected by a nonlinear amplier
v The sidelobe level can be kept 40 to 50 dB below the main lobe
level
v When RC and SRC pulse shaping is employed envelope varia-
tions are introduced in QPSK, OQPSK, and ,4-DQPSK
v With a nonlinear amplier near saturation, the sidelobe level
can come up to with 30 dB of the main lobe, followed by a
more gradual spectral rolloff
2 1.75 1.5 1.25 1.0 0.75 0.5 0.25 0.0
0
-50
-40
-30
-20
-10
Normalized Frequency ( f f
c
)/T
S
p
e
c
t
r
u
m
(
d
B
)
Linear
Nonlinear
Approximate OQPSK spectrum with 1 dB backoff
4.9 Channel Estimation and Tracking
v PSK based modulation requires a phase reference at the re-
ceiver to properly recover the message bits
v When slow fading is present recall that the channel introduces
phase variations that are small relative to the modulation in-
duced phase shifts
v There several receiver design options when it comes to phase
tracking
1. Attempt to coherently track the channel induced phase
variations using a carrier phase recovery algorithm
4.9. CHANNEL ESTIMATION AND TRACKING
2. Implement differential detection
3. Implement pilot symbol assisted modulation (PSAM)
v Coherent detection may not always be possible, or may be too
complex to implement in a low-cost receiver design
4.9.1 Differential Detection
v Differential detection relies on the fact that the carrier phase
changes little from one symbol to the next, so that the previous
symbol can be used to demodulate the present symbol
v Consider a transmitted signal of the form
s(t ) = Re
J(t )e
2
c
t
_
where is the carrier amplitude,
c
is the carrier frequency,
and J(t ) is the data modulation
v At the receiver the carrier frequency will not be known exactly,
so in complex baseband form we receive
.(t ) =
/
J(t )e
(2^ t )
n(t )
where
/
is the received signal amplitude, is the residual
frequency error (instabilities and Doppler), and n(t ) is com-
plex white Gaussian channel noise
v We further assume that
J(t ) =
k
b
k
(t kT )
where b
k
are data symbols (possibly complex) and (t ) is the
transmitted pulse shape
v The received signal is rst matched ltered to obtain improved
SNR
,
k
=
_
(n1)T
nT
+
(t kT ) .(t ) Jt
v Assuming a pulse shape having unit energy, the matched lter
output is approximately
,
k
.
/
b
k
e
(k2^ T )
n
k
v We assume that the data stream was differentially decoded at
the transmitter as
b
k
= a
k
b
k1
where a
k
is the original input data stream
v The differential detector performs a delay-and-multiply opera-
tion as follows:
a
k
= ,
k
,
+
k1
=
_
/
b
k
e
(k2^ T )
n
k
_
/
b
k1
e
((k1)2^ T )
n
k1
_
+
= b
k
b
+
k1
e
2^ T
j
k
~ a
k
j
k
where in the last line we have assumed that exp( 2^ T ) .
1
Matched
Filter
Delay T
x(t )
y
k
a
k
Received
Signal
kT
y
k1
( )

4.9. CHANNEL ESTIMATION AND TRACKING
Differential detection based receiver
v So long as the information is differentially encoded, this phase
difference demodulation will work
v There is a performance penalty over fully coherent demodula-
tion, since we form the product ,
k
,
+
k1
to make symbol deci-
sions
4.9.2 Pilot Transmission
PSAM is more complex that differential detection, but not only can
the carrier phase be tracked, the chanel state can also be estimated.
For this to function known pilot symbols are inserted in the transmit
data stream at regular intervals.
v Assuming the channel exhibits a fading pattern in complex
baseband of the form (t ), we can write the received signal
as
.(t ) = (t )J(t ) n(t )
v Assuming Nyquist pulse shaping samples of the matched out-
put are of the form
,
k
=
k
b
k
n
k
We assume that the fading is constant over the pulse length
v Suppose the pilot symbols are know at the receiver at times
k = 1i , where 1 is the pilot symbol spacing and i is an index
variable
v By correlating the received pilot symbols against the known
transmitted pilot symbols we can recover approximate values
of the unknown fading channel
k
1i
= b
+
1i
,
1i
=
1i
[b
1i
[
2
b
+
1i
n
1i
=
1i
n
1i
v Here we have assumed that the data is BPSK, i.e., [b
1i
[
2
= 1
v To obtain a better estimate of
1i
a linear minimum mean-
squared error (LMMSE) estimate is formed using multiple
values of h
1(i m)
in the neighborhood of 1i , i.e.,

1i
=
1
m=1
a
m
h
1(i m)
v The Haykin text explains in detail how the LMMSE coefcient
set {a
m
] is obtained
v A practical implementation is shown below
Matched
Filter
kT
Demux Delay T
Smoothing
Filter
1
()
x(t )
b
Ki
h
Ki
Pilot
Signal

K i
1

K i
y
k
b
k
PSAM receiver for tracking channel variations
4.10. RECEIVER BIT ERROR PROBABILITY PERFORMANCE
v With system fading as well as residual frequency and phase
errors can be tracked
v An obvious downside is overhead needed to transmit the pilot
symbols
4.10 Receiver Bit Error Probability Per-
formance
In this section we consider the bit error rate (BER) or bit error prob-
ability (BEP) of some of the modulation schemes discussed thus far.
A simple additive white Gaussian noise (AWGN) channel is consid-
ered rst followed by a frequency at, slow fading channel.
Bit error probability formulas
Signaling Scheme AWGN Slow Rayleigh
(a) Coh. BPSK, QPSK, MSK
1
2
erfc
__
1
b
N
0
_
1
2
_
1
_
;
0
1;
0
_
(b) Coh. BFSK
1
2
erfc
__
1
b
2N
0
_
1
2
_
1
_
;
0
2;
0
_
(c) Binary DPSK
1
2
exp
_
1
b
N
0
_
1
2(1;
0
)
(d) NC BFSK
1
2
exp
_
1
b
2N
0
_
1
2;
0
where 1
b
is the energy transmitted per bit, N
0
is the one-sided WGN
noise power spectral density, and ;
0
is the mean received value of
1
b
,N
0
under Rayleigh fading.
4.10.1 AWGN Channel
v The above AWGN expressions are exact for the indicated mod-
ulation schemes under coherent detection, differentially coher-
ent, and noncoherent detection
Coherent detection means a locally generated carrier is
obtained via synchronization of both residual phase and
frequency; this increases receiver complexity and may not
always be practical in severe fading environments
Differentially coherent detection means that the previous
symbol is used to demodulate the present, without the
need for a true locally generated carrier
Noncoherent means that a scheme such as frequency dis-
criminator is employed, again a separate locally generated
carrier is not obtained
v Exact closed-form BEP solutions are not available for GMSK,
MPSK with M > 4, QAM, and MFSK. Bound can be calcu-
lated and simulations can of course be performed
5 0 5 10 15
1. 10
6
0.00001
0.0001
0.001
0.01
0.1
1
DPSK
Coherent BPSK
Coherent QPSK
Coherent MSK
Noncoherent BFSK
Coherent BFSK
SNR (E
b
/N
0
dB)
B
i
t
E
r
r
o
r
P
r
o
b
a
b
i
l
i
t
y
(
B
E
P
)
BEP for an AWGN channel
v Note that in all cases the error probability decreases monoton-
ically with increasing 1
b
,N
0
; waterfall curves
v Coherent BPSK/QPSk/MSK has the best performance
v Coherent BFSK is 3 dB inferior to BPSK
v DPSK is less than 3 dB inferior to BPSK, with the value ap-
proaching 1 dB for high 1
b
,N
0
v Noncoherent BFSK is is within 1 dB coherent BFSK at high
1
b
,N
0
GMSK
GMSK is more practical than MSK, yet its performance cannot be
calculated exactly. A simple approximation for the BEP of GMSK
is
1
e
=
1
2
erfc
_
_
_
1
b
2N
0
_
_
where is a constant depending upon the time bandwidth product
W T .
v Note that when = 2 we have coherent MSK
v For W T = 0.3 the degradation is 0.46 dB which corresponds
to ,2 = 0.9
4.10.2 Frequency Flat, Slow fading Channel
When fading is present we characterize the BEP performence in
terms of the mean value of receiced 1
b
,N
0
given by
;
0
=
1
b
N
0
1
2
|
where ; is Rayleigh random variable characterizing the envelope
fading statistics.
0 5 10 15 20
0.001
0.005
0.01
0.05
0.1
0.5
1
Coherent BPSK
Coherent QPSK
Coherent MSK
Noncoherent BFSK
Coherent BFSK
SNR (E
b
/N
0
dB)
B
i
t
E
r
r
o
r
P
r
o
b
a
b
i
l
i
t
y
(
B
E
P
)
DPSK
Coherent AWGN BPSK
BEP for a Rayleigh at fading channel
v The BEP expression in the above table are radically different
from the AWGN case
v The exponential-law expressions become algebraic making the
waterfall much more gradual compared with the AWGN case
v A signicantly larger mean SNR is required to obtain the same
BEP, in fact up to 20 dB the BEP has not dropped below 10
3
v The Rayleigh fading channel poses a serious challenge; more
will be said about this is later chapters
4.11 Theme Example: OFDM
A modulation scheme that is becoming popular for xed wireless
data services is orthogonal frequency division multiplexing (OFDM).
In particular the WLANstandard IEEE 802.11a uses an OFDMscheme
to achieve a serial bit rate of up to 54 Mbps. The RF transmission
bandwidth of 802.11a is constrained to be less than 20 MHz. One
subrate of the standard sends user information at 36 MB/s, but for-
ward error correction and coding increases this up to 48 Mb/s. This
section will discuss some of the details of this rate standard.
v To pack 48 Mb/s in a 20 MHz wide RF channel requires better
than 2 bits/s/Hz of bandwidth efciency
v In this case the M-ary scheme 16-QAM is employed since it
offers 4 bits/symbol
v Normally with 16-QAM, each transmitted complex baseband
symbol is of the form
s(t ) = b
k
(t kT
J
). (k 1)T
J
_ t < kT
J
where here (t ) is a the pulse shape and T
J
is the symbol pe-
riod
v In 802.11a rather than sending the information over a single
carrier, 48 orthogonal carriers are used to send {b
k
] demulti-
plexed into 48 parallel streams
v The symbol duration of each of these streams is T = 48T
J
and
a pulse shape denoted, g(t ), applied to each stream is 48 times
4.11. THEME EXAMPLE: OFDM
linger than (t )
s
i
(t ) = b
k.i
g(t kT )e
2
i
t
.
(k 1)T _ t < kT
i = 1. 2. . . . . 48
v The complete complex baseband transmitted signal is of the
form
s(t ) =
48
i =1
s
i
(t )
v The orthogonal carriers are established through the use of the
inverse fast Fourier transform (IFFT) (formally the inverse dis-
crete Fourier transform (IDFT))
v There is a corresponding discrete Fourier transform(DFT) (the
fast version is the FFT), so both the forward and reverse oper-
ations can be performed
v For the FFT/IFFT to be fast and efcient, the size or number
of points used in the computation is often a power of two
v Since 48 is not a power of two, a 64-point IFFT is used in
practice
v The additional subcarriers are used for synchronization and
tracking operations; unused subcarriers can be set to zero
v The transceiver block diagram for the 48 Mb/s system is shown
below
Forward
Error
Correct.
Encoder
16-
QAM
Mod.
64
point
IFFT
S
e
r
i
a
l

t
o
P
a
r
a
l
l
e
l
P
a
r
a
l
l
e
l
t
o

S
e
r
i
a
l
D/A
A/D
C
h
a
n
n
e
l
16-
QAM
Demod.
P
a
r
a
l
l
e
l
t
o

S
e
r
i
a
l
64
point
IFFT
S
e
r
i
a
l

t
o
P
a
r
a
l
l
e
l
Forward
Error
Correct.
Decoder
36
Mbps
48
Mbps
12
Msps
48
sub-
carriers
extra
sub car.
64
250 kHz
wide
sub car.
16 MHz
wide
data
stream
Tx
Rx
I
n
p
u
t
S
t
r
e
a
m
E
s
t
i
m
a
t
e

o
f
I
n
p
u
t

S
t
r
e
a
m
IEEE 802.11a OFDM transceiver block diagram
v Subcarrier orthogonality is established by choosing
i
=
1
T
i = 1. 2. . . . . 48
v Each subcarrier is effectively sampled at M samples per sym-
bol of length T , thus the discrete-time version of s(t ) becomes
s(m) =
M1
i =0
b
n
e
2mn,M
. m = 0. 1. . . . . M 1
v Digital signal processing, particularly VLSI implementation
makes all of this possible
4.11. THEME EXAMPLE: OFDM
4.11.1 Cyclic Prex
v A special feature of OFDM is its ability to overcome ISI by the
use of a cyclic prex
v The cyclic prex extends the duration of each OFDM symbol
by a guard time corresponding to the maximum expected mul-
tipath delay
v By including the cyclic prex the overall transmission band-
width grows since a fraction of each symbol now contains a
guard interval
v The receiver must remove the guard interval before the FFT
processing
D
e
m
u
x
M
u
x
b
k
C
h
a
n
n
e
l
b
k,l s
l
(t )
exp( j 2 f
l
t )
s(t )
exp(j 2 f
l
t )
LPF

b
k
Subcarrier Mod Subcarrier Demod
Simplied view of the OFDM modulation and demodulation
process
4.12 Theme Example: Cordless Tele-
phone
In Europe a cordless telephone standard known as CT-2 exists. 40
channels are dened over a 4 MHz bandwidth from 864.15 to 868.15
MHz. The multiple access scheme is FDMA using time-division
duplexing (TDD) as opposed to FDD which was discussed earlier.
Each channel occupies 100 kHz of bandwidth.
v With TDD the same frequency/channel is used for both trans-
mit and receive
v In the case of CT-2 a form of GMSK, known as GFSK (Gaus-
sian FSK), is used to provide a bandwidth efcient modulation
v The particular form of GFSK is M-ary so multiple bits are sent
per symbol
v Speech is digitized using adaptive pulse code modulation (AD-
PCM) to a rate of 32 kbps
v Considering the TDD nature the overall bit rate is 72 kbps,
which includes overhead for guard intervals between transmit
and receive frames
66 Bits
5.5
Bits
66 Bits
6.5
Bits
2 ms
Base to
portable
(forward)
Portable
to base
(reverse)
CT-2 TDD frame structure
4.12. THEME EXAMPLE: CORDLESS TELEPHONE
v The GFSK modulation scheme simplies the transceiver de-
sign, and in particular allows for noncoherent reception
v Power control, a means to minimize interference between users,
is implemented in CT-2 via a simple two level scheme
A power level of 5 mW is used as a default, unless the
received signal power exceeds a particular level
When the signal is above threshold return path signaling
tels the transmitter to reduce its power by about 15 dB
.

Introduction To Digital Data Transmission

Diunggah oleh

Informasi Dokumen

Deskripsi Asli:

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Introduction To Digital Data Transmission

Diunggah oleh

Hak Cipta:

Format Tersedia

Chapter 4

Introduction to Digital Data

is a fraction of the total power in s

v The fractional containment bandwidth of rectangle pulse shaped

T = 2 the contained power is only 0.9028, while at

T = 5 the contained power has only increased to 0.9592

ECE 5625 Communication Systems I 4-11

phase jumps in QPSK can be a problem when the signal

v The transmitted signal now takes the form

each serial bit period T

4-52 ECE 5625 Communication Systems I

Anda mungkin juga menyukai