Click here to load reader
Upload
donhu
View
222
Download
3
Embed Size (px)
Citation preview
1
QPSKQAM
QPSK,.QAMMATLABQPSKQAMMonte CarloQPSKQAM
QPSK QAM
Title: QPSK, QAM modulation correspondence simulation
Major: Electronic Information of Science and Technology
Name: Signature:
Supervisor: Signature:
ABSTRACT
This article mainly discusses the modulation demodulation principle of the QPSK ,QAM. And it analyzes their modulation demodulation realization process. Then it carries on the communications system simulation with MATLAB to the QPSK, QAM modulation process, analyzes how each step of the modulation process specifically realizes. And obtains their error rate diagram of curves using Monte the Carlo simulation and analyzes. Finally carries on comparison to the QPSK, QAM two modulation ways, obtains their good and bad points respectively.
Key words QPSK, QAM, Scatter diagramError rate
Type of ThesisTheories Analysis Software Development
191837F.B.Mores,,;1837();1948(Shannon);2070,
AMFMFSKOOKPWMPPMPAM
//QAMQPSKBPSKQPSKQAM
11
11.1
11.2
11.2.1
21.2.2
21.3
42 MATLAB
42.1MATLAB
73
84 QPSK
84.1QPSK
84.1.1QPSK
94.1.2 QPSK
104.1.3 QPSK
124.2MATLABQPSK
124.2.1
124.2.2
145 QAM
145.1QAM
145.1.1 MQAM
145.1.2MQAM
155.2 QAM
155.2.1
155.2.2
165.3 MQAM
175.4 16QAM
185.5 16QAM
206 QPSKQAM
22
23
24
25
34
1
1.1
1.1.1
1.2
1.2.1
1
1/45351605kHZ187560m,10cm,900MHz
2
4kHz----PCM30
3
10kHz
1.2.2
---m(t),c(t),s(t),m(t)c(t)
1m(t)
1m(t)
2
2c(t)
1
2
3
1PAMASK
2FMFSKPDM
3PMPSKPPM
4
1ASK
2FMPMFSK
1.3
12SNR
2
log(1)
CWSNR
=+
Cbit/s,WHzSNRdB1.3.11.3.2
1.3.1
1.3.2SNR(dB)
h,bit/Symbol,bit/s/Hz
h
hSNR1.3.3
h
1.3.3
2 MATLAB
BasicFortranc, ,Mathwork1967Matrix LaboratoryMATLAB6.xwindowsMATLABMATLABMATLAB--BasicFortranCMATLAB. MATLAB
2.1MATLAB
1)randint
(1)OUT=randint 01
(2)OUT=randint(M) MM0101
(3)OUT=randint(M,N) MN0101
(4)OUT=randint(M,N,RANGE) MNRANGE
2) length
Length(M)M
3) mod
Mod(x,y)xy
4)qammod
Y = QAMMOD(X,M) QAMXMMQAMMX0M-1
5scatterplot
Scatterplot(x)xx
6) fft
Fft(x)xFFT
7abs
Abs(x)x
8)plot
(1)ploty
yplot[1 2 3]y
(2)plotx,y
Xy
(3)plotx,y,
9)stem
plot
10)axis
axis[XMIN XMAX YMIN YMAX ]
XY
11) Xlabel, ylabel
xyxlabel
xlabel(text,property1propertyValuel,property1,property Valye2,)
12)title
xlabel
13)text
Text(x,y,string)(x,y)
14)subplot
H=subplot(m,n,p)mnp
3
1
0
()
2
n
n
S
w
=
w
p
0
n
---W/Hz
(2)
0
n
/2
(
)
(
)
1
0
()
2
nn
n
RFS
twdt
-
==
t
0
30
4 QPSK
4.1QPSK
4.1.1QPSK
M
(
)
(
)
(
)
(
)
2
2(1)/
Re
2
()cos21
22
()cos1cos2()sin(1)sin2
c
jft
jmM
m
c
cc
stgtee
gtftm
M
gtmftgtmft
MM
p
p
p
p
pp
pp
-
=
=+-
=---
m=1,2,M
0
tT
g(t)
2(1)/
m
mM
qp
=-
m=1,2,MM
22
00
11
()()
22
TT
mg
stdtgtdt
ee
===
1
()
ft
2
()
ft
1122
()()()
mmm
stsftsft
=+
1
2
()()cos2
c
g
ftgtft
p
e
=
2
2
()()sin2
c
g
ftgtft
p
e
=-
1,2
mmm
sss
=
22
cos(1),sin(1)
22
gg
m
smm
MM
ee
pp
=--
m=1,2,M
M=44PSKQPSKQPSK0
/2,,3/2
ppp
/4
QPSK
p
-
/4
p
QPSK
2
()cos(1)
2
s
QPSKc
s
E
Stti
T
p
w
=+-
0
S
tT
i=1234
S
E
0
S
tT
c
w
s
T
QPSKQPSK
QPSK4.1.1
4.1.1 QPSK
4.1.2 QPSK
AWGN
(
)
()()
()()cos(2)()sin(2)
m
mccsc
rtutnt
utntftntft
pp
=+
=+-
()
c
nt
()
s
nt
1
()()cos(2)
Tc
tgtft
yp
=
2
()sin(2)
Tc
gtft
yp
=-
1
()
t
y
2
()
t
y
22
(cossin)
mscss
mm
rsnnn
MM
pp
xx
=+=++
c
n
s
n
1
()()
2
cTc
ngtntdt
-
=
1
()()
2
sTs
ngtntdt
-
=
()
c
nt
()
s
nt
()()0
cs
EnEn
==
()0
cs
Enn
=
c
n
s
n
22
0
()()
2
cs
N
EnEn
==
rM{
m
s
}
(,)
mm
Crsrs
=
m=0,1,M-1
r=(
c
r
,
s
r
)
arctan
s
r
c
r
r
q
=
{
m
s
}
r
q
AWGNPAM
2
0
2
b
PQ
N
x
=
b
x
1M
f
4
M
P
0
0
2
2sin
2
2sin
s
M
b
PQ
NM
k
Q
NM
x
p
x
p
2
log
kM
=
/
4.1.3 QPSK
QPSKPSK4.1.2
4.1.2 QPSK
+1V1-1V02IQ
sin
t
w
cos
t
w
QPSKQPSK4IQ4[0 0][0 1][1 0][1 1]290
2
m
m
M
p
j
=
m=0,1,,M-1,45-45QPSK
sin
t
w
cos
t
w
-QPSK4.1.3
4.1.3 QPSK
QPSKQPSK
,
()()cos()
()cos()
Sck
k
ck
stgtkTt
gtt
wj
wj
=-+
=+
g(t)
k
j
c
w
1
0
00
()cos()cos
11
()cos(2)cos()
22
s
ss
T
ckc
TT
kk
Ugtttdt
gttdtgtdt
wjw
wjj
=+
=++
t=
s
T
I
U
s
T
s
T
s
T
t=
s
T
0t=
s
T
1/2cos
k
j
g(t)I
cos
k
j
1
cos
k
U
j
Q
sin
k
j
sin
kk
U
j
10
4.2MATLABQPSK
4.2.1
1)QPSK
2)MATLAB
3)MATLABM-File
4)
(1)
(2)
(3)
(4)QPSK
(5)
(6)
5)
4.2.2
MATLAB
1) 01
2) IQ
3) BPSKsin(2*pi*Fc*t)cos(2*pi*Fc*t)
4) QPSK
5) Monte Carlo
Monte Carlo
r4201400.25,(0.25,0.5),(0.5,0.75),(0.75,1.0)
00,01,1110,
m
s
c
n
s
n
2
d
2
1
d
=
s
x
SNR
m
rsn
=+
r44.2.1SNR
0
/
b
N
x
/2
bs
xx
=
10000
4.2.1
5 QAM
M
APKMQAM
MQAM4QAM8QAM16QAM64QAM16QAM
5.1QAM
QAMcos2
t
f
c
p
)
2
sin(
t
f
c
p
5.1.1 MQAM
xtytMQAM
t
f
t
y
t
f
t
x
S
c
c
QAM
p
p
2
sin
)
(
2
cos
)
(
+
=
)
(
)
(
b
k
k
kT
t
g
x
t
x
-
=
)
(
)
(
b
k
k
kT
t
g
y
t
y
-
=
b
T
k
x
k
y
3
,
1
MQAM
()Re[()()]Re[]
c
cs
jwt
QAMiiTi
stajagtVe
=+=
i=1,2,3,..,M
0
s
tT
arctan(/)
ii
ab
q
=
,MQAM
5.1.2MQAM
MQAM
1122
()()()
iii
stsftsft
=+
i=1,2,3,..,M
0
s
tT
1
2
()()cos
Tc
g
ftgtwt
E
=
0
s
tT
2
2
()()sin
Tc
g
ftgtwt
E
=-
0
s
tT
1
1
0
()()
2
s
c
T
g
iii
E
sstftdta
==
i=1,2,3,..,M
22
0
()()
2
s
c
T
g
iii
E
sstftdta
==
i=1,2,3,..,M
MQAM
12
,
[,]
22
cs
gg
iiiii
EE
sssaa
==
i=1,2,3,..,M
5.2 QAM
5.2.1
.,MQAM.
MQAM,MQAM
MQAM1.1-abc16QAM(M=4)
5.2.1 MQAM
5.2.2
MQAM,,:MQAM ,dminmin,
(1)MQAM
m QAM MQAM,,
(2)MQAMdmin
dminMQAM,MQAMMQAMdmin,MQAM
(3)MQAMmin
MQAM,MQAM,MQAMmin,
5.2.2
min
q
min
d
g
450
0
0.43
E
1.7
0
30
0
0.59
E
1.3
0
18
0
0.63
E
1.8
5.2.2
,E0,dmin,,,,,,min,-,
, MQAMMQAMMQAMMQAMMQAMMQAM
5.3 MQAM
MQAM5.3.1
5.3 .1 16QAM
{
n
a
}IQ
b
R
/2K/2
M
I(t)Q(t)
M
PAM0I(t)Q(t)ASKQAM
5.3.2 16QAM
5.4 16QAM
MAPKM
1616QAMQAM
16QAM
5.4.1 16QAM
MQAM
t
f
t
y
t
f
t
x
S
c
c
QAM
p
p
2
sin
)
(
2
cos
)
(
+
=
)
(
)
(
b
k
k
kT
t
g
x
t
x
-
=
)
(
)
(
b
k
k
kT
t
g
y
t
y
-
=
b
T
k
x
k
y
3
,
1
16QAM
k
k
y
x
,
3
,
1
)
(
b
kT
t
g
-
a a
5.5 16QAM
16QAM:
1
2
101416QAM4bit Str=[1 0 0 1 1 1 0 1 0 1 0 0 0 1 1 1 1 1 0 1]
2IQ1/2IQstr1=[1 0 1 0 0 0 0 1 1 0]str2=[0 1 1 1 1 0 1 1 1 1 ]
3BPSKA1Fc1Fs50
4BPSKQAM
Monte Carlo
16
1
b
2
b
3
b
4
b
4
[
]
,
mcms
AA
[
]
,
cs
nn
f
0
[
]
,
mccmss
rAnAn
=++
r5.5.1SNRN=10000
/4
bs
xx
=
5.5.1 16QAM
6 QPSKQAM
h
[bit/(sHz)],
h
=
,,,,,2 bit/(sHz),,.,
1
2
bit/(sHz),,QPSKQAM2 bit/(sHz)QPSK116QAM2
eb
P
BERQPSKQAM
1QPSKMQAMMQAMM4
2MQAMM4MPSKM=16
MPSK:M=4
2sin2
4
p
dAA
p
==
M=16
2sin0.39
16
p
dAA
p
==
MQAM:M=4
2
2
1
Q
A
dA
L
==
-
M=16
22
0.47
13
Q
A
dAA
L
===
-
M4M=16QAM16PSKMMPSK
3 QPSK
2
A
16QAM0.47A,,,QPSK16QAM.
4 MPSKM2
p
M
2
p
/MM
16QPSK8PSK
MATLABQPSKQAMQPSKQAM
QPSKQAM
QPSKQAMMATLABMonte CarloQPSKQAM
MATLABMATLABMATLAB
QPSKQAM
[1]..1994Vol.14(No.1)
[2].m-QAM.1997Vol.2(No.11)
[3].m-QAM.-
[4].16QAMHDTV.
[5].FSKPSKQAM.
[6].M-QAM.1996Vol.24(No.10)
[7] .QAM modulations with q-ary turbo codes.SCIENCE IN CHINA,1997,Vol.40(No.1)
[8].QAMHFC.2002
[1]...2001
[2].MATLAB..2001
[3].G.....2001
[4]...2000
5.3 Amplitude and Phase Modulation
In amplitude and phase modulation the information bit stream is encoded in the amplitude and/or phase of the transmitted signal. Specifically, over a time interval of
s
T
,
M
2
log
=
k
bits are encoded into the amplitude and/or phase of the transmitted signal
s
T
t
t
s
p/p
p0/p
p),/p
p(/p
. The transmitted signal over this period div class="embedded" id="_1207225318"/p)/p
p2/p
psin(/p
p)/p
p(/p
p)/p
p2/p
pcos(/p
p)/p
p(/p
p)/p
p(/p
pt/p
pf/p
pt/p
ps/p
pt/p
pf/p
pt/p
ps/p
pt/p
ps/p
pc/p
pQ/p
pc/p
pI/p
pp/p
pp/p
p-/p
p=/p
can be written in terms of its signal space representation as div class="embedded" id="_1207225542"/p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p2/p
p2/p
p1/p
p1/p
pt/p
ps/p
pt/p
ps/p
pt/p
ps/p
pi/p
pi/p
pj/p
pj/p
p+/p
p=/p
with basis functions div class="embedded" id="_1207228743"/p)/p
p2/p
pcos(/p
p)/p
p(/p
p)/p
p(/p
p0/p
p1/p
pj/p
pp/p
pj/p
p+/p
p=/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
and div class="embedded" id="_1207225850"/p)/p
p2/p
psin(/p
p)/p
p(/p
p)/p
p(/p
p0/p
p2/p
pj/p
pp/p
pj/p
p+/p
p-/p
p=/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
,where div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
is a shaping pulse. To send the div class="embedded" id="_1207226080"/pi/p
th message over the time interval [div class="embedded" id="_1207226139"/pT/p
pk/p
pkT/p
p)/p
p1/p
p(/p
p,/p
p+/p
], we setdiv class="embedded" id="_1207226217"/p)/p
p(/p
p)/p
p(/p
p1/p
pt/p
pg/p
ps/p
pt/p
ps/p
pi/p
pI/p
p=/p
anddiv class="embedded" id="_1207226315"/p)/p
p(/p
p)/p
p(/p
p2/p
pt/p
pg/p
ps/p
pt/p
ps/p
pi/p
pQ/p
p=/p
. These in-phase and quadrature signal components are baseband signals with spectral characteristics determined by the pulse shapediv class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
. In particular, their bandwidth div class="embedded" id="_1207226478"/pB/p
equals the bandwidth of div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
, and the transmitted signal div class="embedded" id="_1207226534"/p)/p
p(/p
pt/p
ps/p
is a passband signal with center frequency div class="embedded" id="_1207226578"/pc/p
pf/p
and passband bandwidthdiv class="embedded" id="_1207226607"/pB/p
p2/p
. In practice we take div class="embedded" id="_1207226636"/ps/p
pg/p
pT/p
pK/p
pB/p
p=/p
where div class="embedded" id="_1207226726"/pg/p
pK/p
depends on the pulse shape: for rectangular pulsesdiv class="embedded" id="_1207226762"/pg/p
pK/p
= .5 and for raised cosine pulses div class="embedded" id="_1207226821"/p1/p
p5/p
p./p
p/p
p/p
pg/p
pK/p
, as discussed in Section 5.5. Thus, for rectangular pulses the bandwidth ofdiv class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
isdiv class="embedded" id="_1207226986"/pS/p
pT/p
p5/p
p./p
and the bandwidth ofdiv class="embedded" id="_1207227067"/p)/p
p(/p
pt/p
ps/p
is div class="embedded" id="_1207227162"/pS/p
pT/p
p1/p
. Since the pulse shape div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
is fixed, the signal constellation for amplitude and phase modulation is defined based on the constellation point:div class="embedded" id="_1207227337"/pM/p
pi/p
pR/p
ps/p
ps/p
pi/p
pi/p
p,...,/p
p1/p
p,/p
p)/p
p,/p
p(/p
p2/p
p2/p
p1/p
p=/p
p/p
. The complex baseband representation of div class="embedded" id="_1207226534"/p)/p
p(/p
pt/p
ps/p
is /pp class=""div class="embedded" id="_1207227755"/p{/p
p}/p
p)/p
p2/p
p(/p
p0/p
p)/p
p(/p
p)/p
p(/p
pt/p
pf/p
pj/p
pj/p
pc/p
pe/p
pe/p
pt/p
px/p
pR/p
pt/p
ps/p
pp/p
pj/p
p=/p
(5.51)/pp class=""wherediv class="embedded" id="_1207227828"/p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p2/p
p1/p
pt/p
pg/p
pjs/p
ps/p
pt/p
pjs/p
pt/p
ps/p
pt/p
px/p
pi/p
pi/p
pQ/p
pI/p
p+/p
p=/p
p+/p
p=/p
.The constellation pointdiv class="embedded" id="_1207228036"/p)/p
p,/p
p(/p
p2/p
p1/p
pi/p
pi/p
pi/p
ps/p
ps/p
ps/p
p=/p
is called the symbol associated with the div class="embedded" id="_1207228134"/pM/p
p2/p
plog/p
bits and div class="embedded" id="_1207228189"/pS/p
pT/p
is called the symbol time. The bit rate for this modulation is div class="embedded" id="_1207228254"/pK/p
bits per symbol or div class="embedded" id="_1207228282"/ps/p
pT/p
pM/p
pR/p
p2/p
plog/p
p=/p
bits per second./pp class="" There are three main types of amplitude/phase modulation:/pp class="" Pulse Amplitude Modulation (MPAM):information encoded in amplitude only./pp class="" Phase Shift Keying (MPSK): information encoded in phase only./pp class="" Quadrature Amplitude Modulation (MQAM): information encoded in both amplitude and phase./pp class=""The number of bits per symboldiv class="embedded" id="_1207228390"/pM/p
pK/p
p2/p
plog/p
p=/p
, signal constellationdiv class="embedded" id="_1207227337"/pM/p
pi/p
pR/p
ps/p
ps/p
pi/p
pi/p
p,...,/p
p1/p
p,/p
p)/p
p,/p
p(/p
p2/p
p2/p
p1/p
p=/p
p/p
, and choice of pulse shape div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
determines the digital modulation design. The pulse shape div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
is designed to improve spectral efficiency and combat ISI, as discussed in Section 5.5 below./pp class=""Amplitude and phase modulation over a given symbol period can be generated using the modulator structure shown in Figure 5.10. Note that the basis functions in this figure have an arbitrary phasediv class="embedded" id="_1207228604"/p0/p
pj/p
associated with the transmit oscillator. Demodulation over each symbol period is performed using the demodulation structure of Figure 5.11, which is equivalent to the structure of Figure 5.7 for div class="embedded" id="_1207228794"/p)/p
p2/p
pcos(/p
p)/p
p(/p
p)/p
p(/p
p0/p
p1/p
pj/p
pp/p
pj/p
p+/p
p=/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
and div class="embedded" id="_1207225850"/p)/p
p2/p
psin(/p
p)/p
p(/p
p)/p
p(/p
p0/p
p2/p
pj/p
pp/p
pj/p
p+/p
p-/p
p=/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
. Typically the receiver includes some additional circuitry for carrier phase recovery that matches the carrier phase div class="embedded" id="_1207302856"/pj/p
at the receiver to the carrier phase div class="embedded" id="_1207228604"/p0/p
pj/p
at the transmitter, which is called coherent detection. If div class="embedded" id="_1207302892"/p0/p
p0/p
p/p
pD/p
p=/p
p-/p
pj/p
pj/p
pj/p
then the in-phase branch will have an unwanted term associated with the quadrature branch and vice versa, i.e. div class="embedded" id="_1207302984"/p1/p
p2/p
p1/p
p1/p
p)/p
psin(/p
p)/p
pcos(/p
pn/p
ps/p
ps/p
pr/p
pi/p
pi/p
p+/p
pD/p
p+/p
pD/p
p=/p
pj/p
pj/p
and div class="embedded" id="_1207303131"/p2/p
p2/p
p1/p
p2/p
p)/p
pcos(/p
p)/p
psin(/p
pn/p
ps/p
ps/p
pr/p
pi/p
pi/p
p+/p
pD/p
p+/p
pD/p
p=/p
pj/p
pj/p
which can result in significant performance degradation. The receiver structure also assumes that sampling function everydiv class="embedded" id="_1207303230"/ps/p
pT/p
seconds is synchronized to the start of the symbol period, which is called synchronization or timing recovery. Receiver synchronization and carrier phase recovery are complex receiver operations that can be highly challenging in wireless environments. These operations are discussed in more detail in Section 5.6. We will assume perfect carrier recovery in our discussion of MPAM, MPSK and MQAM, and therefore set div class="embedded" id="_1207303278"/
EMBED Equation.3 p0/p
p0/p
p=/p
p=/p
pj/p
pj/p
for their analysis./pp class=""5.3.1 Pulse Amplitude Modulation (MPAM)/pp class=""We will start by looking at the form of linear modulation, one-dimensional MPAM, which has no quadrature component (div class="embedded" id="_1207303388"/p0/p
p2/p
p=/p
pi/p
ps/p
) For MPAM all of the information is encoded into the signal amplitudediv class="embedded" id="_1207303424"/pi/p
pA/p
. The transmitted signal over one symbol time is given by/pp class=""div class="embedded" id="_1207303642"/p{/p
p}/p
pc/p
ps/p
pc/p
pi/p
pt/p
pf/p
pj/p
pi/p
pi/p
pf/p
pT/p
pt/p
pt/p
pf/p
pt/p
pg/p
pA/p
pe/p
pt/p
pg/p
pA/p
pR/p
pt/p
ps/p
pc/p
p1/p
p0/p
p),/p
p2/p
pcos(/p
p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p2/p
p>>
=
=
p
p
5.52
where
M
i
d
M
i
A
i
,...,
2
,
1
,
)
1
2
(
=
-
-
=
defines the signal constellation, parameterized by the distance d which is typically a function of the signal energy, and
)
(
t
g
is the pulse shape satisfying (5.12) and (5.13).
N-dimensional space, which must be mapped to a sequence of constellations in 2-dimensional space in order to be generated by the modulator shown in Figure 5.10. The general conclusion in [18] is that for uncoded modulation, the increased complexity of spherical constellations is not worth their energy gains, since coding can provide much better performance at less complexity cost. However, if a complex channel code is already being used and little further improvement can be obtained by a more complex code, constellation shaping may obtain around 1 dB of additional gain. An in-depth discussion of constellation shaping, as well as constellations that allow a noninteger number of bits per symbol, can be found in [18].
The minimum distance between constellation points is
d
A
A
d
j
i
j
i
2
|
|
min
,
min
=
-
=
.The amplitude of the transmitted signal takes on M different values, which implies that each pulse conveys
K
M
=
2
log
bits per symbol time
s
T
.
Over each symbol period the MPAM signal associated with the
i
th constellation has energy
2
0
2
2
2
0
2
)
2
(
cos
)
(
)
(
i
T
c
i
T
i
s
A
dt
t
f
t
g
A
dt
t
s
E
s
s
i
=
=
=
p
(5.53)
Since the pulse shape must satisfy(5.12). Note that the energy is not the same for each signal
)
(
t
s
i
, i=1,,M. Assuming equally likely symbol, the average energy is
=
=
M
i
i
s
A
M
E
1
2
1
(5.54)
The constellation mapping is usually done by Gray encoding, where the message associated with signal amplitudes that are adjacent to each other differ by one bit value, as illustrated in Figure 5.12. With this encoding method, if noise causes the demodulation process to mistake one symbol for an adjacent one (with most likely type of error), this results in only a single bit error in the sequence of K bits. Gray codes can be designed for MPSK and square MQAM constellations, but not rectangular MQAM.
The decision regions
i
Z
, i=1,,M associated with the pulse amplitude
d
M
i
A
i
)
1
2
(
-
-
=
for M=4 and M=8 are shown in Figure 5.13. Mathematically, for any M, these decision regions are defines by
=
-
-
+
-
=
+
-
=
M
i
d
A
M
i
d
A
d
A
i
d
A
Z
i
i
i
i
i
)
,
[
1
2
)
,
[
1
)
,
(
Form (5.52) we see that MPAM has only a single basis function
)
2
cos(
)
(
)
(
1
t
f
t
g
t
c
p
f
=
. Thus, the coherent demodulator of Figure 5.11 for MPAM reduces to the demodulator shown in Figure 5.14, where the multithreshold device maps x to a decision region
i
Z
and outputs the corresponding bit sequence
)
,
,
{
1
K
i
b
b
m
m
K
=
=
.
5.3.2 Phase Shift Keying (MPSK)
For MPSK all of the information is encoded in the phase of the transmitted signal. Thus, the transmitted signal over one symbol time is given by
{
}
t
f
M
i
t
Ag
t
f
M
i
t
Ag
M
i
t
f
t
Ag
T
t
e
e
t
Ag
R
t
s
c
c
c
s
t
f
M
i
j
i
c
p
p
p
p
p
p
p
p
2
sin
)
1
(
2
sin
)
(
2
cos
)
1
(
2
cos
)
(
)
1
(
2
2
cos
)
(
0
,
)
(
)
(
2
/
)
1
(
2
-
-
-
=
-
+
=
=
-
(5.55)
Thus, the constellation points or symbol
)
,
(
2
1
i
i
s
s
are given by
-
=
M
i
A
s
i
)
1
(
2
cos
1
p
and
-
=
M
i
A
s
i
)
1
(
2
sin
2
p
for i=1,,M. The pulse shape
)
(
t
g
satisfied (5.12) and (5.13), and
K
i
M
i
M
i
2
,
,
1
,
)
1
(
2
=
=
-
=
K
p
q
are the different phases in the signal constellation points that convey the information bits. The minimum distance between constellation points is
)
sin(
2
min
M
A
d
p
=
, where A is typically a function of the signal energy. 2PSK is often referred to as binary PSK or BPSK, which 4PSK is often called quadrature phase shift keying (QPSK), and is the same as MQAM with M=4 which is defined below.
All possible transmitted signals
)
(
t
s
i
have equal energy
2
0
2
)
(
A
dt
t
s
E
s
i
T
i
s
=
=
(5.56)
Note that for
s
s
T
t
T
t
g
=
0
,
2
)
(
, i.e. a rectangular pulse, this signal has constant envelope, unlike the other amplitude modulation techniques MPAM and MQAM. However, rectangular pulse are spectrally-inefficient, and more efficient pulse shapes make MPSK nonconstant envelope. As for MPAM, constellation mapping is usually done by Gray encoding, where the messages associated with signal phases that are adjacent to each other differ by one bit value, as illustrated in Figure 5.15. With this encoding method, mistaking a symbol for an adjacent one causes only a single bit error.
The decision regions
i
Z
i=1,,M, associated with MPSK for M=8 are shown in Figure 5.16. If we represent
2
R
re
r
j
=
q
in polar coordinates then decision for any M are defined by
{
}
M
i
M
i
re
Z
j
i
)
5
.
(
2
)
5
.
(
2
:
+
-
=
p
q
p
q
(5.57)
Form (5.55) we see that MPSK has both in-phase and quadrature components, and coherent demodulator is as shown in Figure 5.11. For the special case of BPSK, the decision regions as given in Example 5.2 simplify to
)
0
:
(
1
>
=
r
r
Z
and
)
0
:
(
1
=
r
r
Z
. Moreover BPSK has only a single basis function
)
2
cos(
)
(
)
(
1
t
f
t
g
t
c
p
f
=
and , since there is only a single bit transmitted per symbol time
s
T
, the bit time
s
b
T
T
=
. Thus, the coherent demodulator of Figure 5.11for BPSK reduces to the demodulator shown in Figure 5.17, where the threshold device maps x to the positive or negative half of the real line, and outputs the corresponding bit value. We have assumed in this figure that the message corresponding to a bit value if 1,
1
1
=
m
, is mapped to constellation point
A
s
=
1
and the message corresponding to a bit value of 0,
0
2
=
m
, is mapped to the constellation point
A
s
-
=
2
.
5.3.3 Quadrature Amplitude Modulation (MQAM)
For MQAM, the information bits are encoded in both the amplitude and phase of the transmitted signal. Thus, whereas both MPAM and MPSK have one degree of freedom in which to encode the information bits (amplitude or phase), MQAM has two degrees of freedom. As a result, MQAM is more spectrally-efficient than MPAM and MPSK, in that it can encode the most number of bits per symbol for a given average energy.
The transmitted signal is given by
{
}
s
c
i
i
c
i
i
t
f
j
j
i
i
T
t
t
f
t
g
A
t
f
t
g
A
e
t
g
e
A
R
t
s
c
i
-
=
=
0
),
2
sin(
)
(
)
sin(
)
2
cos(
)
(
)
cos(
)
(
)
(
2
p
q
p
q
p
q
(5.58)
Where the pulse shape
)
(
t
g
satisfies (5.12) and (5.13). The energy in
)
(
t
s
i
is
2
0
2
)
(
i
T
i
s
A
t
s
E
s
i
=
=
(5.59)
The same as for MPAM. The distance between any pair of symbols in the signal constellation is
2
2
2
2
1
1
)
(
)
(
j
i
j
i
j
i
ij
s
s
s
s
s
s
d
-
+
-
=
-
=
(5.60)
For square signal constellations, where
1
i
s
and
2
i
s
take values on
l
L
i
d
L
i
2
,
,
2
,
1
,
)
1
2
(
=
=
-
-
K
, the minimum distance between signal points reduces
d
d
2
min
=
, the same as for MPAM. In fact, MQAM with square constellation of size
2
L
is equivalent to MPAM modulation with constellations of size L on each of the in-phase and quadrature signal components. Common square constellation are 4QAM and 16QAM, which are shown in Figure 5.18 below. These square constellation have
2
2
2
L
M
l
=
=
constellation points, which are used to send 2l bit/symbol, or l bits per dimension. It can be shown that the average power of a square signal constellation with l bits per dimension,
l
S
, is proportional to
3
4
l
,and it follows that the average power for one more bit per dimension
l
l
S
S
4
1
+
. Thus, for square constellation it takes approximately 6 dB more power to send an additional 1 bit/dimension or 2bits/symbol while maintaining the same minimum distance between constellation points.
Good constellation mapping can be hard to find for QAM signals, especially for irregular constellation shapes. In particular, it is hard to find a Gray code mapping where all adjacent symbol differ by a single bit. The decision regions
i
Z
, i=1,,M, associated with MQAM for M=16 are shown in Figure 5.19. From (5.58)we see that MQAM has both in-phase and quadrature components, and thus the coherent demodulator is as shown in Figure 5.11.
5.3.4 Differential Modulation
The information in MPSK and MQAM signals is carried in the signal phase. Thus, these modulation techniques require coherent demodulation, i.e. the phase of the transmitted signal carrier 0 must be matched to the phase of the receiver carrier . Techniques for phase recovery typically require more complexity and cost in the receiver and they are also susceptible to phase drift of the carrier. Moreover, obtaining a coherent phase reference in a rapidly fading channel can be difficult. Issues associated with carrier phase recovery are dicussed in more detail in Section 5.6. Due to the difficulties as well as the cost and complexity associated with carrier phase recovery, differential modulation techniques, which do not require a coherent phase reference at the receiver, are generally preferred to coherent modulation for wireless applications.
Differential modulation falls in the more general class of modulation with memory, where the symbol transmitted over time
]
)
1
(
,
[
s
s
T
k
kT
+
depends on the bits associated with the current message to be transmitted and the bits transmitted over prior symbol times. The basic principle of differential modulation is to use the previous symbol as a phase reference for the current symbol, thus avoiding the need for a coherent phase reference at the receiver. Specifically, the information bits are encoded as the differential phase between the current symbol and the previous symbol. For example, in differential BPSK, referred to as DPSK, if the symbol over time
]
,
)
1
[(
s
s
kT
T
k
-
has phase
p
q
q
q
,
0
,
)
1
(
=
=
-
i
j
i
e
k
, then to encode a 0 bit over
]
)
1
(
,
[
s
s
T
k
kT
+
, the symbol would have phase
i
j
e
k
q
q
=
)
(
and to encode a 1 bit the symbol would have phase
p
q
q
+
=
i
j
e
k
)
(
. In other words, a 0 bit is encoded by no change in phase, whereas a 1 bit is encoded as a phase change of . Similarly, in 4PSK modulation with differential encoding, the symbol phase over symbol interval
]
)
1
(
,
[
s
s
T
k
kT
+
depends on the current information bits over this time interval and the symbol phase over the previous symbol interval. The phase transitions for DQPSK modulation are summarized in Table 5.1. Specifically, suppose the symbol over time
]
,
)
1
[(
s
s
kT
T
k
-
has phase
i
j
e
k
q
q
=
-
)
1
(
. Then, over symbol time
]
)
1
(
,
[
s
s
T
k
kT
+
, if the information bits are 00, the corresponding symbol would have phase
i
j
e
k
q
q
=
)
(
, i.e. to encode the bits 00, the symbol from symbol interval
]
,
)
1
[(
s
s
kT
T
k
-
is repeated over the next interval
]
)
1
(
,
[
s
s
T
k
kT
+
. If the two information bits to be sent at time interval
]
)
1
(
,
[
s
s
T
k
kT
+
are 01, then the corresponding symbol has phase
)
2
/
(
)
(
p
q
q
+
=
i
j
e
k
. For information bits 10 the symbol phase is
)
2
/
(
)
(
p
q
q
-
=
i
j
e
k
, and for information bits 11 the symbol phase is
)
(
)
(
p
q
q
+
=
k
j
e
n
. We see that the symbol phase over symbol interval
]
)
1
(
,
[
s
s
T
k
kT
+
depends on the current information bits over this time interval and the symbol phase
i
q
over the previous symbol interval. Note that this mapping of bit sequences to phase transitions ensures that the most likely detection error, that of mistaking a received symbol for one of its nearest neighbors, results in a single bit error. For example, if the bit sequence 00 is encoded in the kth symbol then the kth symbol has the same phase as the (k1)th symbol. Assume this phase is
i
q
. The most likely detection error of the kth symbol is to decode it as one of its nearest neighbor symbols, which have phase
2
/
p
q
i
. But decoding the received symbol with phase
2
/
p
q
i
would result in a decoded information sequence of either 01 or 10, i.e. it would differ by a single bit from the original sequence 00. More generally, we can use Gray encoding for the phase transitions in differential MPSK for any M, so that a message of all 0 bits results in no phase change, a message with a single 1 bit and the rest 0 bits results in the minimum phase change of
M
/
2
p
, a message with two 1 bits and the rest 0 bits results in a phase change of
M
/
4
p
, and so forth. Differential encoding is most common for MPSK signals, since the differential mapping is relatively simple. Differential encoding can also be done for MQAM with a more complex differential mapping. Differential encoding of MPSK is denoted by D-MPSK, and for BPSK and QPSK this becomes DPSK and D-QPSK, respectively.
assuming the transmitted symbol at the (k 1)th symbol time was s(k 1) =
e
j
A
p
. Solution: The first bit, a 1, results in a phase transition of , so s(k) =
A
. The next bit, a 0, results in no transition, so s(k + 1) =
A
. The next bit, a 1, results in another transition of , so s(k + 1) =
e
j
A
p
, and so on. The fullsymbol sequence corresponding to 101110 is
A
,
A
,
e
j
A
p
,
A
,
e
j
A
p
,
e
j
A
p
.
The demodulator for differential modulation is shown in Figure 5.20. Assume the transmitted constellation at time k is s(k) =
e
k
j
A
0
)
(
f
q
+
The received vector associated with the sampler outputs is
)
(
)
(
)
(
)
(
0
)
(
2
1
k
n
A
k
jr
k
r
k
z
e
k
j
+
=
+
=
+
f
q
, (5.61)
where n(k) is complex white Gaussian noise. The received vector at the previous time sample k 1 is thus
)
1
(
)
1
(
)
1
(
)
1
(
0
)
1
(
2
1
-
+
=
-
+
-
=
-
+
-
k
n
A
k
jr
k
r
k
z
e
k
j
f
q
. (5.62)
The phase difference between z(k) and z(k 1) dictates which symbol was transmitted. Consider
)
1
(
)
(
)
(
)
1
(
)
1
(
)
(
0
0
)
1
(
)
1
(
))
1
(
)
(
(
2
-
+
+
-
+
+
=
-
*
+
-
-
*
+
-
-
-
*
k
n
k
n
k
n
A
k
n
A
A
k
z
k
z
e
e
e
k
j
k
j
k
k
j
f
q
f
q
q
q
. (5.63)
In the absence of noise (n(k) = n(k 1) = 0) only the first term in (5.63) is nonzero, and this term yields the desired phase difference. The phase comparator in Figure 5.20 extracts this phase difference and outputs the corresponding symbol.
Differential modulation is less sensitive to a random drift in the carrier phase. However, if the channel has a nonzero Doppler frequency, the signal phase can decorrelate between symbol times, making the previous symbol a very noisy phase reference. This decorrelation gives rise to an irreducible error floor for differential modulation over wireless channels with Doppler, as we shall discuss in Chapter 6.
5.3.5 Constellation Shaping
Rectangular and hexagonal constellations have a better power efficiency than the square or circular constellations associated with MQAM and MPSK, respectively. These irregular constellations can save up to 1.3 dB of power at the expense of increased complexity in the constellation map [18]. The optimal constellation shape is a sphere in N-dimensional space, which must be mapped to a sequence of constellations in 2-dimensional space in order to be generated by the modulator shown in Figure 5.10. The general conclusion in [18] is that for uncoded modulation, the increased complexity of spherical constellations is not worth their energy gains, since coding can provide much better performance at less complexity cost. However, if a complex channel code is already being used and little further improvement can be obtained by a more complex code, constellation shaping may obtain around 1 dB of additional gain. An in-depth discussion of constellation shaping, as well as constellations that allow a noninteger number of bits per symbol, can be found in [18].
5.3.6 Quadrature Offset
A linearly modulated signal with symbol
)
,
(
2
1
i
i
i
s
s
s
=
will lie in one of the four quadrants of the signal space. At each symbol time kTs the transition to a new symbol value in a different quadrant can cause a phase transition of up to 180 degrees, which may cause the signal amplitude to transition through the zero point: these abrupt phase transitions and large amplitude variations can be distorted by nonlinear amplifiers and filters. These abrupt transitions are avoided by offsetting the quadrature branch pulse g(t) by half a symbol period, as shown in Figure 5.21. This offset makes the signal less sensitive to distortion during symbol transitions.
Phase modulation with phase offset is usually abbreviated as O-MPSK, where the O indicates the offset. For example, QPSK modulation with quadrature offset is referred to as O-QPSK. O-QPSK has the same spectral properties as QPSK for linear amplification, but has higher spectral efficiency under nonlinear amplification, since the maximum phase transition of the signal is 90 degrees, corresponding to the maximum phase transition in either the in-phase or quadrature branch, but not both simultaneously. Another technique to mitigate the amplitude fluctuations of a 180 degree phase shift used in the IS-54 standard for digital cellular is /4-QPSK [13]. This technique allows for a maximum phase transition of 135 degrees, versus 90 degrees for offset QPSK and 180 degrees for QPSK. Thus, /4-QPSK does not have as good spectral properties as O-QPSK under nonlinear amplification. However, /4-QPSK can be differentially encoded, eliminating the need for a coherent phase reference, which is a significant advantage. Using differential encoding with /4-QPSK is called /4-DQPSK. The /4-DQPSK modulation works as follows: the information bits are first differentially encoded as in DQPSK, which yields one of the four QPSK constellation points. Then, every other symbol transmission is shifted in phase by /4. This periodic phase shift has a similar effect as the time offset in OQPSK: it reduces the amplitude fluctuations at symbol transitions, which makes the signal more robust against noise and fading.
5.4 Frequency Modulation
Frequency modulation encodes information bits into the frequency of the transmitted signal. Specifically, each symbol time K = log2M bits are encoded into the frequency of the transmitted signal s(t), 0 t < Ts, resulting in a transmitted signal si(t) = Acos(2fit + i), where i is the index of the ith message corresponding to the log2M bits and i is the phase associated with the ith carrier. The signal space representation is si(t) = _j sijj(t) where sij = A(i j) and j(t) = cos(2fjt + j), so the basis functions correspond to carriers at different frequencies and only one such basis function is transmitted in each symbol period. The orthogonality of the basis functions requires a minimum separation between different carrier frequencies of f = minij |fjfi| = .5/Ts. Since frequency modulation encodes information in the signal frequency, the transmitted signal s(t) has a constant envelope A. Because the signal is constant envelope, nonlinear amplifiers can be used with high power efficiency, and the modulated signal is less sensitive to amplitude distortion introduced by the channel or the hardware. The price exacted for this robustness is a lower spectral efficiency: because the modulation technique is nonlinear, it tends to have a higher bandwidth occupancy than the amplitude and phase modulation techniques.
53
/
s
T
M
2
log
=
k
s
T
t
t
s
p/p
p0/p
p),/p
p(/p
/div class="embedded" id="_1207224938"/ps/p
pT/p
div class="embedded" id="_1207225318"/p)/p
p2/p
psin(/p
p)/p
p(/p
p)/p
p2/p
pcos(/p
p)/p
p(/p
p)/p
p(/p
pt/p
pf/p
pt/p
ps/p
pt/p
pf/p
pt/p
ps/p
pt/p
ps/p
pc/p
pQ/p
pc/p
pI/p
pp/p
pp/p
p-/p
p=/p
div class="embedded" id="_1207225542"/p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p2/p
p2/p
p1/p
p1/p
pt/p
ps/p
pt/p
ps/p
pt/p
ps/p
pi/p
pi/p
pj/p
pj/p
p+/p
p=/p
div class="embedded" id="_1207228743"/p)/p
p2/p
pcos(/p
p)/p
p(/p
p)/p
p(/p
p0/p
p1/p
pj/p
pp/p
pj/p
p+/p
p=/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
div class="embedded" id="_1207225850"/p)/p
p2/p
psin(/p
p)/p
p(/p
p)/p
p(/p
p0/p
p2/p
pj/p
pp/p
pj/p
p+/p
p-/p
p=/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
[div class="embedded" id="_1207226139"/pT/p
pk/p
pkT/p
p)/p
p1/p
p(/p
p,/p
p+/p
]div class="embedded" id="_1207226080"/pi/p
div class="embedded" id="_1207226217"/p)/p
p(/p
p)/p
p(/p
p1/p
pt/p
pg/p
ps/p
pt/p
ps/p
pi/p
pI/p
p=/p
div class="embedded" id="_1207226315"/p)/p
p(/p
p)/p
p(/p
p2/p
pt/p
pg/p
ps/p
pt/p
ps/p
pi/p
pQ/p
p=/p
in- phasediv class="embedded" id="_1207226478"/pB/p
div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
div class="embedded" id="_1207226578"/pc/p
pf/p
div class="embedded" id="_1207226607"/pB/p
p2/p
div class="embedded" id="_1207226636"/ps/p
pg/p
pT/p
pK/p
pB/p
p=/p
div class="embedded" id="_1207226726"/pg/p
pK/p
div class="embedded" id="_1207226762"/pg/p
pK/p
=.5,div class="embedded" id="_1207226821"/p1/p
p5/p
p./p
p/p
p/p
pg/p
pK/p
5.5div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
div class="embedded" id="_1207226986"/pS/p
pT/p
p5/p
p./p
div class="embedded" id="_1207227067"/p)/p
p(/p
pt/p
ps/p
div class="embedded" id="_1207227162"/pS/p
pT/p
p1/p
,div class="embedded" id="_1207227337"/pM/p
pi/p
pR/p
ps/p
ps/p
pi/p
pi/p
p,...,/p
p1/p
p,/p
p)/p
p,/p
p(/p
p2/p
p2/p
p1/p
p=/p
p/p
div class="embedded" id="_1207226534"/p)/p
p(/p
pt/p
ps/p
/pp class="" div class="embedded" id="_1207227755"/p{/p
p}/p
p)/p
p2/p
p(/p
p0/p
p)/p
p(/p
p)/p
p(/p
pt/p
pf/p
pj/p
pj/p
pc/p
pe/p
pe/p
pt/p
px/p
pR/p
pt/p
ps/p
pp/p
pj/p
p=/p
(5.51)/pp class=""div class="embedded" id="_1207227828"/p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p)/p
p(/p
p2/p
p1/p
pt/p
pg/p
pjs/p
ps/p
pt/p
pjs/p
pt/p
ps/p
pt/p
px/p
pi/p
pi/p
pQ/p
pI/p
p+/p
p=/p
p+/p
p=/p
div class="embedded" id="_1207228036"/p)/p
p,/p
p(/p
p2/p
p1/p
pi/p
pi/p
pi/p
ps/p
ps/p
ps/p
p=/p
div class="embedded" id="_1207228134"/pM/p
p2/p
plog/p
div class="embedded" id="_1207228189"/pS/p
pT/p
div class="embedded" id="_1207228254"/pK/p
div class="embedded" id="_1207228282"/ps/p
pT/p
pM/p
pR/p
p2/p
plog/p
p=/p
/pp class=""//pp class="" MPAM/pp class="" MPSK/pp class="" MQAM/pp class=""div class="embedded" id="_1207228390"/pM/p
pK/p
p2/p
plog/p
p=/p
div class="embedded" id="_1207227337"/pM/p
pi/p
pR/p
ps/p
ps/p
pi/p
pi/p
p,...,/p
p1/p
p,/p
p)/p
p,/p
p(/p
p2/p
p2/p
p1/p
p=/p
p/p
div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
ISI5.5/pp class=""5.10div class="embedded" id="_1207228604"/p0/p
pj/p
5.11,5.7,5.7div class="embedded" id="_1207228794"/p)/p
p2/p
pcos(/p
p)/p
p(/p
p)/p
p(/p
p0/p
p1/p
pj/p
pp/p
pj/p
p+/p
p=/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
div class="embedded" id="_1207225850"/p)/p
p2/p
psin(/p
p)/p
p(/p
p)/p
p(/p
p0/p
p2/p
pj/p
pp/p
pj/p
p+/p
p-/p
p=/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
div class="embedded" id="_1207302856"/pj/p
div class="embedded" id="_1207228604"/p0/p
pj/p
,div class="embedded" id="_1207302892"/p0/p
p0/p
p/p
pD/p
p=/p
p-/p
pj/p
pj/p
pj/p
in- phasediv class="embedded" id="_1207302984"/p1/p
p2/p
p1/p
p1/p
p)/p
psin(/p
p)/p
pcos(/p
pn/p
ps/p
ps/p
pr/p
pi/p
pi/p
p+/p
pD/p
p+/p
pD/p
p=/p
pj/p
pj/p
div class="embedded" id="_1207303131"/p2/p
p2/p
p1/p
p2/p
p)/p
pcos(/p
p)/p
psin(/p
pn/p
ps/p
ps/p
pr/p
pi/p
pi/p
p+/p
pD/p
p+/p
pD/p
p=/p
pj/p
pj/p
Tssynchronizationtiming recoveryoperation5.6MPAM, MPSKMQAMdiv class="embedded" id="_1207303321"/p0/p
p0/p
p=/p
p=/p
pj/p
pj/p
/pp class=""5.3.1 (MPAM)/pp class=""div class="embedded" id="_1207303388"/p0/p
p2/p
p=/p
pi/p
ps/p
MPAMMPAMdiv class="embedded" id="_1207303424"/pi/p
pA/p
one symbol timeThe transmitted signaldiv class="embedded" id="_1207303230"/ps/p
pT/p
div class="embedded" id="_1207226012"/p)/p
p(/p
pt/p
pg/p
(5.12)(5.13)div class="embedded" id="_1207202038"/pd/p
pA/p
pA/p
pd/p
pj/p
pi/p
pj/p
pi/p
p2/p
p|/p
p|/p
pmin/p
p,/p
pmin/p
p=/p
p-/p
p=/p
Mdiv class="embedded" id="_1207202292"/ps/p
pT/p
div class="embedded" id="_1207202252"/pK/p
pM/p
p=/p
p2/p
plog/p
div class="embedded" id="_1207202373"/pi/p
MPAM div class="embedded" id="_1207202411"/p2/p
p0/p
p2/p
p2/p
p2/p
p0/p
p2/p
p)/p
p2/p
p(/p
pcos/p
p)/p
p(/p
p)/p
p(/p
pi/p
pT/p
pc/p
pi/p
pT/p
pi/p
ps/p
pA/p
pdt/p
pt/p
pf/p
pt/p
pg/p
pA/p
pdt/p
pt/p
ps/p
pE/p
ps/p
ps/p
pi/p
p=/p
p=/p
p=/p
p/p
p/p
pp/p
(5.53)/pp class=""(5.12)div class="embedded" id="_1207202658"/p)/p
p(/p
pt/p
ps/p
pi/p
, i=1,,M/pp class="" div class="embedded" id="_1207202822"/p/p
p=/p
p=/p
pM/p
pi/p
pi/p
ps/p
pA/p
pM/p
pE/p
p1/p
p2/p
p1/p
(5.54)/pp class="",5.12,KMPSKMQAMconstellationsMQAM/pp class=""M=4M=8div class="embedded" id="_1207203409"/pd/p
pM/p
pi/p
pA/p
pi/p
p)/p
p1/p
p2/p
p(/p
p-/p
p-/p
p=/p
div class="embedded" id="_1207203337"/pi/p
pZ/p
i=1,,M5.13/pp class=""div class="embedded" id="_1207203870"/p/p
p/p
p/p
p/p
p/p
p=/p
p/p
p-/p
p-/p
p/p
p/p
p+/p
p-/p
p=/p
p+/p
p-/p
p=/p
pM/p
pi/p
pd/p
pA/p
pM/p
pi/p
pd/p
pA/p
pd/p
pA/p
pi/p
pd/p
pA/p
pZ/p
pi/p
pi/p
pi/p
pi/p
pi/p
p)/p
p,/p
p[/p
p1/p
p2/p
p)/p
p,/p
p[/p
p1/p
p)/p
p,/p
p(/p
/pp class=""(5.52)MPAMdiv class="embedded" id="_1207204048"/p)/p
p2/p
pcos(/p
p)/p
p(/p
p)/p
p(/p
p1/p
pt/p
pf/p
pt/p
pg/p
pt/p
pc/p
pp/p
pf/p
p=/p
5.11MPAM5.145.14xdiv class="embedded" id="_1207203337"/pi/p
pZ/p
div class="embedded" id="_1207204306"/p)/p
p,/p
p,/p
p{/p
p/p
p1/p
pK/p
pi/p
pb/p
pb/p
pm/p
pm/p
pK/p
p=/p
p=/p
/pp class="" 5.3.2 (MPSK)/pp class=""MPSK,/pp class=""div class="embedded" id="_1207205024"/p{/p
p}/p
pt/p
pf/p
pM/p
pi/p
pt/p
pAg/p
pt/p
pf/p
pM/p
pi/p
pt/p
pAg/p
pM/p
pi/p
pt/p
pf/p
pt/p
pAg/p
pT/p
pt/p
pe/p
pe/p
pt/p
pAg/p
pR/p
pt/p
ps/p
pc/p
pc/p
pc/p
ps/p
pt/p
pf/p
pM/p
pi/p
pj/p
pi/p
pc/p
pp/p
pp/p
pp/p
pp/p
pp/p
pp/p
pp/p
pp/p
p2/p
psin/p
p)/p
p1/p
p(/p
p2/p
psin/p
p)/p
p(/p
p2/p
pcos/p
p)/p
p1/p
p(/p
p2/p
pcos/p
p)/p
p(/p
p)/p
p1/p
p(/p
p2/p
p2/p
pcos/p
p)/p
p(/p
p0/p
p,/p
p)/p
p(/p
p)/p
p(/p
p2/p
p//p
p)/p
p1/p
p(/p
p2/p
p/p
p/p
p/p
p/p
p/p
p/p
p-/p
p-/p
p/p
p/p
p/p
p/p
p/p
p/p
p-/p
p=/p
p/p
p/p
p/p
p/p
p/p
p/p
p-/p
p+/p
p=/p
p/p
p/p
p=/p
p-/p
/pp class="" (5.55)/pp class=""div class="embedded" id="_1207205144"/p)/p
p,/p
p(/p
p2/p
p1/p
pi/p
pi/p
ps/p
ps/p
div class="embedded" id="_1207205187"/p/p
p/p
p/p
p/p
p/p
p/p
p-/p
p=/p
pM/p
pi/p
pA/p
ps/p
pi/p
p)/p
p1/p
p(/p
p2/p
pcos/p
p1/p
pp/p
i=1,,Mdiv class="embedded" id="_1207205250"/p/p
p/p
p/p
p/p
p/p
p/p
p-/p
p=/p
pM/p
pi/p
pA/p
ps/p
pi/p
p)/p
p1/p
p(/p
p2/p
psin/p
p2/p
pp/p
(5.12)(5.13)div class="embedded" id="_1207205353"/pK/p
pi/p
pM/p
pi/p
pM/p
pi/p
p2/p
p,/p
p,/p
p1/p
p,/p
p)/p
p1/p
p(/p
p2/p
p=/p
p=/p
p-/p
p=/p
pK/p
pp/p
pq/p
div class="embedded" id="_1207205576"/p)/p
psin(/p
p2/p
pmin/p
pM/p
pA/p
pd/p
pp/p
p=/p
A2PSKPSKBPSK4PSK4PSKMQAMM= 4/pp class=""/pp class=""div class="embedded" id="_1207205981"/p2/p
p0/p
p2/p
p)/p
p(/p
pA/p
pdt/p
pt/p
ps/p
pE/p
ps/p
pi/p
pT/p
pi/p
ps/p
p=/p
p=/p
p/p
(5.56)/pp class=""div class="embedded" id="_1207206110"/ps/p
ps/p
pT/p
pt/p
pT/p
pt/p
pg/p
p/p
p/p
p=/p
p0/p
p,/p
p2/p
p)/p
p(/p
MPAMMQAMMPSKMPAM5.15/pp class=""M= 8MPSKdiv class="embedded" id="_1207203337"/pi/p
pZ/p
i=1,,M5.16div class="embedded" id="_1207206816"/p2/p
pR/p
pre/p
pr/p
pj/p
p/p
p=/p
pq/p
M/pp class=""div class="embedded" id="_1207207000"/p{/p
p}/p
pM/p
pi/p
pM/p
pi/p
pre/p
pZ/p
pj/p
pi/p
p)/p
p5/p
p./p
p(/p
p2/p
p)/p
p5/p
p./p
p(/p
p2/p
p:/p
p+/p
p/p
p/p
p-/p
p=/p
pp/p
pq/p
pp/p
pq/p
(5.57)/pp class=""(5.55)MPSK in-phase5.11BPSK5.25.2div class="embedded" id="_1207207398"/p)/p
p0/p
p:/p
p(/p
p1/p
p>
=
r
r
Z
)
0
:
(
1
=
r
r
Z
BPSK
)
2
cos(
)
(
)
(
1
t
f
t
g
t
c
p
f
=
s
T
s
b
T
T
=
5.11BPSKreduces to5.17thresholdx1
1
1
=
m
A
s
=
1
0,
0
2
=
m
A
s
-
=
2
5.3.3 (MQAM)
MQAMMPAMMPSKMQAM2MQAMMPAMMPSKMQAM
{
}
s
c
i
i
c
i
i
t
f
j
j
i
i
T
t
t
f
t
g
A
t
f
t
g
A
e
t
g
e
A
R
t
s
c
i
-
=
=
0
),
2
sin(
)
(
)
sin(
)
2
cos(
)
(
)
cos(
)
(
)
(
2
p
q
p
q
p
q
(5.58)
(5.12)(5.13)
)
(
t
s
i
2
0
2
)
(
i
T
i
s
A
t
s
E
s
i
=
=
(5.59)
MPAM
2
2
2
2
1
1
)
(
)
(
j
i
j
i
j
i
ij
s
s
s
s
s
s
d
-
+
-
=
-
=
(5.60)
1
i
s
2
i
s
l
L
i
d
L
i
2
,
,
2
,
1
,
)
1
2
(
=
=
-
-
K
d
d
2
min
=
MPAM MQAMMPAML 4QAM16QAM5.18
2
2
2
L
M
l
=
=
2l /l /.6db1/2/
QAMM= 16MQAM
i
Z
i=1,,M,5.19(5.58)MQAMin- phase5.11
5.3.4
MPSKMQAMdemodulation
0
j
j
5.6
]
)
1
(
,
[
s
s
T
k
kT
+
BPSKDPSK
]
,
)
1
[(
s
s
kT
T
k
-
p
q
q
q
,
0
,
)
1
(
=
=
-
i
j
i
e
k
]
)
1
(
,
[
s
s
T
k
kT
+
0
p
q
q
+
=
i
j
e
k
)
(
1014PSK
]
)
1
(
,
[
s
s
T
k
kT
+
DQPSK5.1
]
,
)
1
[(
s
s
kT
T
k
-
i
j
e
k
q
q
=
-
)
1
(
0000
]
,
)
1
[(
s
s
kT
T
k
-
]
)
1
(
,
[
s
s
T
k
kT
+
01
)
2
/
(
)
(
p
q
q
+
=
i
j
e
k
10
)
2
/
(
)
(
p
q
q
-
=
i
j
e
k
11
)
(
)
(
p
q
q
+
=
k
j
e
n
]
)
1
(
,
[
s
s
T
k
kT
+
i
q
00kk(k1)
i
q
k
2
/
p
q
i
2
/
p
q
i
011000,MPSKM01010MPSKMQAMMPSKD- MPSKBPSKQPSKDPSKD- QPSK
5.20ks(k) =
e
k
j
A
0
)
(
f
q
+
)
(
)
(
)
(
)
(
0
)
(
2
1
k
n
A
k
jr
k
r
k
z
e
k
j
+
=
+
=
+
f
q
(5.61)
)
1
(
)
1
(
)
1
(
)
1
(
0
)
1
(
2
1
-
+
=
-
+
-
=
-
+
-
k
n
A
k
jr
k
r
k
z
e
k
j
f
q
(5.62)
z(k)z(k 1)
)
1
(
)
(
)
(
)
1
(
)
1
(
)
(
0
0
)
1
(
)
1
(
))
1
(
)
(
(
2
-
+
+
-
+
+
=
-
*
+
-
-
*
+
-
-
-
*
k
n
k
n
k
n
A
k
n
A
A
k
z
k
z
e
e
e
k
j
k
j
k
k
j
f
q
f
q
q
q
(5.63)
(5.63)5.20
DopplerDoppler6
5.3.5
MQAMMPSK1.3dBN5.102[18]1dB[18]
PAGE
2
_1207291969.unknown
_1211175010.unknown
_1211176224.unknown
_1211179025.unknown
_1211180565.unknown
_1211355624.unknown
_1211711303.unknown
_1211711609.unknown
_1211711638.unknown
_1211711660.unknown
_1211711679.unknown
_1211711624.unknown
_1211711370.unknown
_1211356164.unknown
_1211709678.unknown
_1211711284.unknown
_1211356241.unknown
_1211356310.unknown
_1211355830.unknown
_1211355915.unknown
_1211355809.unknown
_1211188690.unknown
_1211189203.unknown
_1211189692.unknown
_1211188791.unknown
_1211180647.unknown
_1211180894.unknown
_1211180600.unknown
_1211180434.unknown
_1211180449.unknown
_1211180514.unknown
_1211180442.unknown
_1211179209.unknown
_1211180425.unknown
_1211179186.unknown
_1211176691.unknown
_1211178808.unknown
_1211178896.unknown
_1211178934.unknown
_1211178848.unknown
_1211177608.unknown
_1211178799.unknown
_1211176840.unknown
_1211176490.unknown
_1211176594.unknown
_1211176648.unknown
_1211176501.unknown
_1211176297.unknown
_1211176321.unknown
_1211176249.unknown
_1211175627.unknown
_1211175800.unknown
_1211176011.unknown
_1211176217.unknown
_1211175864.unknown
_1211175702.unknown
_1211175711.unknown
_1211175675.unknown
_1211175337.unknown
_1211175521.unknown
_1211175548.unknown
_1211175449.unknown
_1211175203.unknown
_1211175224.unknown
_1211175095.unknown
_1211018834.unknown
_1211019865.unknown
_1211045957.unknown
_1211174682.unknown
_1211174929.unknown
_1211174995.unknown
_1211174874.unknown
_1211046193.unknown
_1211174502.unknown
_1211174664.unknown
_1211046321.unknown
_1211046352.unknown
_1211046229.unknown
_1211046014.unknown
_1211046127.unknown
_1211045983.unknown
_1211045524.unknown
_1211045872.unknown
_1211045890.unknown
_1211045622.unknown
_1211045433.unknown
_1211045487.unknown
_1211026786.unknown
_1211026480.unknown
_1211026510.unknown
_1211019910.unknown
_1211019548.unknown
_1211019757.unknown
_1211019812.unknown
_1211019627.unknown
_1211019234.unknown
_1211019492.unknown
_1211018873.unknown
_1207303278.unknown
_1211017513.unknown
_1211018165.unknown
_1211018680.unknown
_1211018746.unknown
_1211018598.unknown
_1211017916.unknown
_1211018144.unknown
_1211017598.unknown
_1210922459.unknown
_1210922933.unknown
_1211017102.unknown
_1210922718.unknown
_1210922053.unknown
_1210922219.unknown
_1207303642.unknown
_1210878361.unknown
_1207304018.unknown
_1207303388.unknown
_1207303424.unknown
_1207303321.unknown
_1207293132.unknown
_1207293154.unknown
_1207302856.unknown
_1207302892.unknown
_1207303131.unknown
_1207303230.unknown
_1207302984.unknown
_1207293156.unknown
_1207293155.unknown
_1207292268.unknown
_1207292761.unknown
_1207292871.unknown
_1207292946.unknown
_1207292987.unknown
_1207292910.unknown
_1207292817.unknown
_1207292376.unknown
_1207292460.unknown
_1207292349.unknown
_1207292244.unknown
_1207292267.unknown
_1207292073.unknown
_1207292025.unknown
_1207226578.unknown
_1207227162.unknown
_1207228036.unknown
_1207228282.unknown
_1207228604.unknown
_1207291301.unknown
_1207291693.unknown
_1207291719.unknown
_1207291471.unknown
_1207228794.unknown
_1207228390.unknown
_1207228189.unknown
_1207228254.unknown
_1207228134.unknown
_1207227755.unknown
_1207227828.unknown
_1207227337.unknown
_1207226762.unknown
_1207226986.unknown
_1207227067.unknown
_1207226821.unknown
_1207226636.unknown
_1207226726.unknown
_1207226607.unknown
_1207205342.unknown
_1207208026.unknown
_1207210264.unknown
_1207226139.unknown
_1207226315.unknown
_1207226478.unknown
_1207226534.unknown
_1207226217.unknown
_1207210748.unknown
_1207225156.unknown
_1207225542.unknown
_1207226012.unknown
_1207226080.unknown
_1207225850.unknown
_1207225318.unknown
_1207210784.unknown
_1207225056.unknown
_1207210862.unknown
_1207224938.unknown
_1207210506.unknown
_1207208883.unknown
_1207209305.unknown
_1207209606.unknown
_1207209666.unknown
_1207210006.unknown
_1207209632.unknown
_1207209408.unknown
_1207209258.unknown
_1207208126.unknown
_1207208188.unknown
_1207208081.unknown
_1207205981.unknown
_1207207398.unknown
_1207207641.unknown
_1207207684.unknown
_1207207444.unknown
_1207206816.unknown
_1207207000.unknown
_1207206110.unknown
_1207205576.unknown
_1207205926.unknown
_1207205353.unknown
_1178112355.unknown
_1207203409.unknown
_1207204306.unknown
_1207205024.unknown
_1207205187.unknown
_1207205250.unknown
_1207205144.unknown
_1207203870.unknown
_1207204048.unknown
_1207202292.unknown
_1207202411.unknown
_1207202822.unknown
_1207203337.unknown
_1207202658.unknown
_1207202373.unknown
_1207202038.unknown
_1207202252.unknown
_1179665260.unknown
_1179665182.unknown
_1179665195.unknown
_1146837637.unknown
_1178112319.unknown
_1178110081.unknown
_1178110212.unknown
_1178110260.unknown
_1178110391.unknown
_1178110243.unknown
_1178110153.unknown
_1146839823.unknown
_1178109996.unknown
_1146837645.unknown
_1146836080.unknown
_1146836857.unknown
_1146837195.unknown
_1146837248.unknown
_1146836965.unknown
_1146836603.unknown
_1022311942.unknown
_1146835743.unknown
_1022311702.unknown