IEMS5701
Wireless
Communica7ons
Systems
Part
III:
Noise,
Digital
Modula7on
1
Outline
• Introduc7on
to
addi7ve
white
Gaussian
noise
• Digital
modula7on
and
detec7on
2
Introduc7on
to
Noise
• Noise
is
one
of
the
most
cri7cal
and
fundamental
concepts
affec7ng
communica7on
systems
• The
subject
of
communica7on
systems
is
largely
about
methods
to
overcome
the
distor7ng
or
bad
effects
of
noise
• Noise
is
a
Gaussian
random
process
• Power
spectral
density
is
N0/2
per
dimension
3
Gaussian
Random
Variable
• A
random
variable
X
is
Gaussian
(or
Normal)
distributed
if
its
density
func7on
is
( x− µ )2
1 −
f X (x) = e 2σ 2
2πσ 2
where
µ
is
the
mean
and
σ
2
is
the
variance.
We
can
also
write
X ~ N( µ, σ 2 )
• A
linear
func7on
of
Gaussian
variables
is
s7ll
Gaussian
4
Characteriza7on
of
Noise
in
Systems
• Signal
to
noise
ra7o
(SNR):
The
most
common
way
to
characterize
noise
in
communica7on
systems
5
Outline
• Introduc7on
to
addi7ve
white
Gaussian
noise
• Digital
modula7on
and
detec7on
6
Representa7on
of
Informa7on:
Analog
Signals
• Defini7on:
A
con7nuous
signal
for
which
the
7me
varying
feature
of
the
signal
is
a
representa7on
of
some
other
7me
varying
quan7ty,
i.e.,
message
• Example:
Sound
recoding
– Sound
(i.e.,
fluctua7ons
in
air
pressure)
introduces
fluctua7ons
in
the
voltage
or
the
current
of
an
electrical
waveform.
– The
voltage
of
current
is
the
“analog
signal”
of
the
sound.
• Feature:
Small
fluctua7ons
in
the
signal
are
meaningful
• Advantages:
– Can
poten7ally
carry
an
infinite
amount
of
signal
resolu7on
– Processing
may
be
achieved
more
simply,
for
signals
can
be
processed
directly
by
analog
components
• Disadvantages:
Vulnerable
to
noise
7
Representa7on
of
Informa7on:
Digital
Signals
Hi, Alice!
Hello, Bob! I am doing very well!
How are you doing!
001101011100100
…
Example:
ASCII
Code
100 0001 A 110 0001 a
100 0010 B 110 0010 b
100 0011 C 110 0011 c
100 0100 D 110 0100 d
100 0101 E 110 0101 e
100 0110 F 110 0110 f
100 0111 G 110 0111 g
100 1000 H 110 1000 h
100 1001 I 110 1001 i
100 1010 J 110 1010 j
100 1011 K 110 1011 k
Example:
Morse
Code
Titanic 1905
Representa7on
of
Informa7on:
Digital
Signal
• Defini7on:
a
physical
signal
that
is
a
representa7on
of
a
sequence
of
discrete
values,
e.g.,
a
bit
stream
or
a
digi7zed
analog
signal
• Analog
to
digital
converter:
0100100…
How to transmit the 0, 1 sequence in the air?
--- Modulation
11
Modula7on
• Defini7on:
Process
of
encoding
informa7on
in
a
manner
suitable
for
transmission
• Example:
BPSK
signals
(baseband
representa7on)
12
Basic
Digital
Signal
Modula7on
Methods
s(t) = A cos(ω c t + ϕ ) t ∈[ 0,T ]
13
Amplitude
Shi^
Keying:
On-‐Off
Keying
(OOK)
s0 (t) = A0 cos(ω c t) t ∈[ 0,T ]
s1 (t) = A1 cos(ω c t) t ∈[ 0,T ]
14
Frequency
Shi^
Keying
(FSK)
s0 (t) = A cos(ω 0 t) t ∈[ 0,T ]
s1 (t) = A cos(ω 1t) t ∈[ 0,T ]
15
Phase-‐Shi^
Keying
(PSK)
Binary PSK (BPSK)
s0 (t) = −A cos(ω c t) t ∈[ 0,T ]
s1 (t) = A cos(ω c t) t ∈[ 0,T ]
Baseband representation
s0 (t) = −A t ∈[ 0,T ]
s1 (t) = A t ∈[ 0,T ]
16
Quadrature
PSK
(QPSK)
Transmitted signal
s00 (t) = −A cos(ω c t) + A sin(ω c t) t ∈[ 0,T ]
s01 (t) = −A cos(ω c t) − A sin(ω c t) t ∈[ 0,T ]
s10 (t) = A cos(ω c t) + A sin(ω c t) t ∈[ 0,T ]
s11 (t) = A cos(ω c t) − A sin(ω c t) t ∈[ 0,T ]
Baseband representation
s00 (t) = −A − Ai t ∈[ 0,T ]
s01 (t) = −A + Ai t ∈[ 0,T ]
s10 (t) = A − Ai t ∈[ 0,T ]
s11 (t) = A + Ai t ∈[ 0,T ]
17
M-‐ary
PSK
(MPSK)
• Let
phase
take
on
M=2m
levels
• One
symbol
carries
m
bits
8PSK
18
Pulse
Amplitude
Modula7on
(PAM)
• Let
amplitude
take
on
M=2m
levels
• One
symbol
carries
m
bits
19
Quadrature
AM
• Equal
to
2
PAM
channels
16 QAM 4 QAM (QPSK) 8 QAM
20
Op7mal
Receiver
for
BPSK
y(t) = s(t) + n(t) Y
y '(t) ŝ(t) = 1
LPF ∫ (i) dt ≥0 ?
N
T
ŝ(t) = 0
cos ω c t
21
Performance
Analysis
N 0T
η(t) ~ N(0, σ 2 ) σ 2 = in one dimension
y '(t) s1 = AT + ∫ n(t) dt 2
T
N 0T
N(AT , σ 2 ) σ2 = in one dimension
2
0
1 ⎛ (x − AT )2 ⎞
P(E 1) = Pr ( y '(t) < 0 ) = ∫ exp ⎜ − ⎟ dx
−∞ π N 0T ⎝ N 0T ⎠
− AT
1 ⎛ t2 ⎞ x − AT
= ∫ −∞
N 0T 2
2π
exp ⎜ − ⎟ dt
⎝ 2⎠
where t =
N 0T 2
~ N(0,1)
∞ 1 ⎛ t2 ⎞ ∞
⎛ −x 2 ⎞
= ∫ AT
N 0T 2 2π
exp ⎜ − ⎟ dt
⎝ 2⎠
Q(y) = ∫
1
2π
exp ⎜
⎝ 2 ⎟⎠
dx Q function
y
⎛ AT ⎞ ⎛ 2A 2T ⎞ ⎛ 2Eb ⎞
= Q⎜ ⎟ = Q ⎜ ⎟ == Q ⎜ N ⎟
⎝ N 0T 2 ⎠ ⎝ N0 ⎠ ⎝ 0 ⎠
P(E) = P(E 1)P(1) + P(E 0)P(0) = P(E 1)
22
Performance
Analysis
23
QPSK
Detector
Y
ŝI = 1
LPF ∫ (i) dt ≥0 ?
N
T
ŝI = 0
cos ω c t Y
ŝQ = 1
LPF ∫ (i) dt ≥0 ?
N
T
ŝQ = 0
sin ω c t
24
Performance
Analysis
(
Pr ( correct detection s11 ) = Pr ηI ≥ −AT & ηQ ≥ −AT )
(
= Pr (ηI ≥ −AT ) Pr ηQ ≥ −AT )
2
⎛ 2A 2T ⎞
= Q⎜− ⎟
⎝ N 0 ⎠
2
⎛ ⎛ 2A 2T ⎞⎞
= ⎜1 − Q ⎜ ⎟ ⎟⎟
⎜⎝ ⎝ N0 ⎠⎠
⎛ 2A 2T ⎞
≈ 1 − 2Q ⎜ ⎟
⎝ N0 ⎠
⎛ 2A 2T ⎞ ⎛ 2A 2T ⎞
Pr ( correct detection ) ≈ 1 − 2Q ⎜ ⎟ Pr ( symbol error ) ≈ 2Q ⎜ ⎟
⎝ N0 ⎠ ⎝ N0 ⎠
⎛ 2A 2T ⎞
Pr ( bit error ) ≈ Q ⎜ ⎟
⎝ N0 ⎠
25
Tradeoff
Between
Data
Rate
and
BER
26
