Nhập môn

Kỹ thuật Truyền thông

TS. Đỗ Công Thuần IT4593
Khoa Kỹ thuật máy tính, Trường CNTT&TT
Đại học Bách khoa Hà Nội
Email: [email protected]
Thông tin chung
• Tên học phần: Nhập môn kỹ thuật truyền thông
• Mã học phần: IT4593
• Khối lượng: 2 TC (2-1-0-4)
• Lý thuyết và bài tập: 16.5 buổi (3 tiết/buổi)
• Đánh giá (QT+CK):
30% QT (kiểm tra + bài tập/project + chuyên cần-quiz )
70% CK (trắc nghiệm + tự luận)
• Tài liệu tham khảo:
‒ Slide bài giảng
‒ Textbook: Communication Systems Engineering, 2nd
Edition, by John G. Proakis Masoud Salehi
‒ Internet
 Phần 2: Các kỹ thuật điều chế số
(Digital Modulations)
Bài 11: Không gian tín hiệu 4-PSK và m-PSK
Quadrature modulation
• Consider a 2-D constellation, suppose that basis signals =
cosine and sine

• Each constellation symbol corresponds to a vector with two

real components
Quadrature modulation DURATION T

Binary information sequence vT [n] H k DURATION T

sT [n] M  R 2

Symbol sequence

[n] R  [n] R

DURATION T

Transmitted signal

s(t) = [n]b1 (t − nT) +  [n]b2 (t − nT) = a(t) + b(t)

n n
Quadrature modulation
Spectrum of a(t):

when p(t) = ideal low pass filter

− f0 f0

1 R 1R
=R =R
T T
Quadrature modulation
Spectrum of b(t):

when p(t) = ideal low pass filter

− f0 f0

1 R 1R
=R =R
T T
Quadrature modulation

s(t) = a(t) + b(t)

It can be proved that

Gs ( f ) = Ga ( f ) + Gb ( f )
Quadrature modulation
s(t) = a(t) + b(t)
Gs = Ga + Gb
Ga = x  P( f − f 0 ) 2 + P( f + f 0 ) 2  xR
 
Gb = y  P( f − f 0 ) 2 + P( f + f 0 ) 2  yR
 
Gs = z  P( f − f 0 ) 2 + P( f + f 0 ) 2  zR
 
Ga and Gb have the same shape and live on the same frequencies.
This is also the case for Gs.
The spectrum of s(t) only depends on |P(f )|2.
Quadrature modulation

Example when p(t) = ideal low pass filter

Gs = z P( f − f 0 ) + P( f + f 0 ) 
 zR
2 2
 

− f0 f0

1 R 1 R
=R =R
T T
 b1 (t)

si
I/Q component i

Given a quadrature modulation,

let us consider its transmitted waveform: i b0 (t)

s(t) = a(t) + b(t) =

i(t) q(t)
I component (in phase) Q component (in quadrature)
Complex envelope
s(t) = i(t)cos (2 f0t )+ q(t)sin (2 f0t )

Complex envelope

i(t) = [n]p(t − nT) q(t) =  [n]p(t − nT)

n n

Complex symbol  [n] = [n] − j [n]

Complex envelope
 [n] = [n]− j [n]

b1 (t)

si
 i = i − j i
i

i b0 (t)

Quadrature constellation as a set of complex numbers

M =  i = i − j i i=1
m
Analytic signal

Analytic signal
4-PSK: characteristics

1. Band-pass modulation

2. 2D signal set

3. Basis signals p(t)cos(2f0t) and p(t)sin(2 f0t)

4. Constellation = 4 signals, equidistant on a circle

5. Information associated to the carrier phase

4-PSK: constellation

SIGNAL SET
M = {s1 (t) = Ap(t) cos(2 f0 t) , s2 (t) = Ap(t) sin(2 f0 t)
s3 (t) = −Ap(t) cos(2 f0 t) , s4 (t) = −Ap(t) sin(2 f0 t) }

If we write

Information associated to the carrier phase

4-PSK: constellation

SIGNAL SET M = {si (t) = Ap(t) cos(2 f 0t − i ) }i=1

4


i = (i −1)
2

b1 (t) = p(t) cos(2 f 0t)

Vectors
b2 (t) = p(t)sin(2 f 0t)

VECTOR SET

M = {s1 = (A,0), s 2 = (0, A), s 3 = (−A,0), s 4 = (0, −A) }  R 2

4-PSK: constellation
VECTOR SET

M = {s1 = (A,0), s 2 = (0, A), s 3 = (−A,0), s 4 = (0, −A) }  R 2

s2

s3 s1

s4
4-PSK: constellation
SIGNAL SET (with arbitrary starting phase)

M = {si (t) = Ap(t) cos(2 f 0t − i ) }i=1

4


i =  + (i −1)
2

si (t) = (Acosi ) p(t) cos(2 f0 t) + (Asin i ) p(t) sin(2 f0 t)

Vector set
Vectors
b1 (t) = p(t) cos(2 f 0t) M = {s i = (i , i ) }i=1
4
 R2
i = Acos i
b2 (t) = p(t) sin(2 f0t) i = Asin i

i =  + (i −1)
2
4-PSK: constellation
Example: =0

M = {s1 = (A,0) , s 2 = (0, A) , s 3 = (−A,0) , s 4 = (0, −A)}  R 2

4-PSK: constellation
Example: 
= −
4

M = {s1 = (− , − ) , s 2 = (+ , − ) , s 3 = (+ , + ) , s 4 = (− , + )}  R 2

A
=
2
4-PSK: binary labeling

Example of Gray labeling

e : H2  M
01

e(00) = s0
e(01) = s 1 11

00
e(11) = s 2
e(10) = s3 10
4-PSK: transmitted waveform
m=2→k =2 T = 2Tb R=
Rb
2
Each symbol has duration T
Each symbol component ( and β) lasts for T second

Transmitted waveform

i(t) q(t)
I component (in phase) Q component (in quadrature)
4-PSK: transmitted waveform
example for p(t) =
1
PT (t) f0 = 2Rb
T = T
1 1 0 1 0 0 1 0
1.2

1.0
01
0.8

0.6 11

00
0.4

0.2 10

0.0

-0.2

-0.4

-0.6

-0.8

-1.0

-1.2
0 2 4 6 8
t/T
4-PSK: analytic signal

i(t) q(t)

 [n] = [n]− j[n]

4-PSK: analytic signal

 [n] = [n] − j[n]

M = s1 = (a − ja), s2 = (−a − ja), s3 = (−a + ja), s4 = (a + ja),

4-PSK: bandwidth and spectral efficiency
Transmitted waveform

Each symbol [n] and β[n] has time duration T = 2Tb

4-PSK: bandwidth and spectral efficiency

Case 1: p(t) = ideal low pass filter

− f0 f0

1/RRT 1/RRT

Total bandwidth Rb
Bid = R =
(ideal case) 2

Spectral efficiency R
id = b = 2 bps / Hz
(ideal case) Bid
4-PSK: bandwidth and spectral efficiency

Case 2: p(t) = RRC filter with roll off 

−f0 f0

1 R(1+) 1R(1+)
(1+  )
(1+  ) T
T
Total bandwidth Rb
B = R(1+  ) = (1+  )
2

Spectral efficiency  = Rb = 2
bps / Hz
B (1+  )
Exercise

Given a bandpass channel with bandwidth B = 4000 Hz, centred around f0=2
GHz, compute the maximum bit rate Rb we can transmit over it with a
4-PSK constellation in the two cases:

• Ideal low pass filter

• RRC filter with =0.25
4-PSK: modulator
4-PSK: demodulator
4-PSK: Eye diagram

4-PSK constellation with RRC filter (=0.5)

Canale I Canale Q

33
4-PSK: intepretation
The 4-PSK vector set can be viewed as
the Cartesian product of two 2-PSK constellations

( A, A)
( A)

( A)
4-PSK: intepretation
This is also true for the binary Gray labeling
(first bit = I component, second bit = Q component)

01 11
1

0
00 10

0 1
4-PSK: intepretation
The AWGN channel adds two Gaussian components which are
statistically independent
 4-PSK: intepretation
The Voronoi regions of 4-PSK signals are the Cartesian product of the Voronoi
regions of the constituent 2-PSK constellations
 4-PSK: intepretation
The Voronoi regions of 4-PSK signals are the Cartesian product of the Voronoi
regions of the constituent 2-PSK constellations

Given the received vector

(ρ1 [n] , ρ2 [n])

The sign of the first component

0
ρ1 [n] determines the first received
bit

The sign of the second component

ρ2 [n] determines the second
1 received bit
4-PSK: intepretation
The 4-PSK modulation can be viewed as
the Cartesian product of two 2-PSK constellations
transmitted over two independent channels

01 11
1

0
00 10

0 1
4-PSK: modulator
4-PSK: demodulator
4-PSK: intepretation

The Cartesian product

interpretation clarifies why 11
01
a 4-PSK constellation 1

1. Has the same BER 0

performance of a 2-PSK 00 10

2. Has double spectral

efficiency (two sequences
with half bit-rate transmitted 0 1
on the same frequencies)
4-PSK: error probability
1 E
BER = erfc b BER
2 N0
1
0.1
0.01
1E-3
1E-4
1E-5
1E-6
BER

1E-7
1E-8
1E-9
1E-10
1E-11
1E-12
1E-13
1E-14
-2 0 2 4 6 8 10 12 14 16
Eb/N0 [dB]
4-PSK: applications

Probably the most used digital modulation

• Satellite links
• Terrestrial radio links (with low spectral efficiency)
• GPS/Galileo
• UMTS
• ...
m-PSK: characteristics

1. Band-pass modulation

2. 2D signal set

3. Basis signals p(t)cos(2π f0t) & p(t)sin(2π f0t)

4. Costellation = m signals, equidistant on a circle

5. Information associated to the carrier phase

m-PSK: constellation
SIGNAL SET M = {si (t) = Ap(t) cos(2 f 0t − i ) }i=1
m

i =  + (i −1) 2m

Information associated to the carrier phase

m-PSK: constellation

si (t) = Ap(t) cos(2 f 0t − i ) 2

i =  + (i −1) m

We can write

si (t) = ( Acosi ) p(t) cos(2 f 0 t) + ( Asin i ) p(t) sin(2 f 0 t)

Clearly, we have two vectors

b1 (t) = p(t) cos(2 f 0t)

b2 (t) = p(t) sin(2 f 0t)
m-PSK: constellation
SIGNAL SET i =  + (i −1) m
2
M = {si (t) = Ap(t) cos(2 f 0t −  i ) }i=1
m

VECTORS b1 (t) = p(t) cos(2 f 0t)

b2 (t) = p(t)sin(2 f 0t)

VECTOR SET M = {s i = (i , i ) }i=1

m
 R2
i = Acos i
i = Asin i
2
i =  + (i −1) m
Example
8-PSK =0
Example
16-PSK =0
m-PSK: binary labeling

e : Hk  M

It is always possible to build Gray labelings

011 0100 0101 0111

111
001 0011
1100
0001
101 000 1101
0000
1111
0010
1011
100 010
1001 0110
110 1000 1010 1110
m-PSK: transmitted waveform
k = log2 m T = kTb R=
Rb
k
Each symbol has duration T
Each symbol component ( and β) lasts for T second

Transmitted waveform

i(t) q(t)
I component (in phase) Q component (in quadrature)
m-PSK: analytic signal

i(t) q(t)
m-PSK: bandwidth and spectral efficiency
Transmitted waveform

Gs ( f ) = z  P( f − f 0 ) + P( f + f 0 )  zR
2 2

 

Each symbol [n] and β[n] has time duration T = kTb

 m-PSK: bandwidth and spectral efficiency

Case 1: p(t) = ideal low pass filter

− f0 f0

1/RT 1/RT

Total bandwidth Rb
Bid = R =
(ideal case) k

Spectral efficiency Rb
id = = k bps / Hz
(ideal case) Bid
 m-PSK: bandwidth and spectral efficiency

Case 2: p(t) = RRC filter with roll off 

−f0 f0

1 R(1+) 1R(1+)
(1+  )
(1+  ) T
T
Rb
Total bandwidth B = R(1+  ) = (1+  )
k

Spectral efficiency  = Rb = k bps / Hz

B (1+  )
Exercise

Given a bandpass channel with bandwidth B = 4000 Hz, centred around f0=2
GHz, compute the maximum bit rate Rb we can transmit over it with an
8-PSK constellation or a 16-PSK constellation in the two cases:

• Ideal low pass filter

• RRC filter with =0.25
m-PSK: modulator

FOR m > 4 NOT CARTESIAN PRODUCT

m-PSK: demodulator

FOR m > 4 NOT CARTESIAN PRODUCT

Voronoi regions = plane sectors
m-PSK: eye diagram
8-PSK constellation with RRC filter (hệ số cuộn=0.5)

[ and β components = 0.924,0.383,-0.383,-0.924]

Channel I Channel Q
 m-PSK: eye diagram
16-PSK constellation with RRC filter (hệ số cuộn=0.5)

[  and β components = 0.981,0.832,0.556,0.195,-0.195,-0.556,-0.832,-0.981]

Channel I Channel Q
 m-PSK constellation: error probability

By applying the asymptotic approximation we can obtain

The performance decrease for increasing m

(minimum distance decreases)

m-PSK constellation: error probability

-3.6 dB with respect to 4-PSK

-4.6 dB with respect to 8-PSK

No one uses m-PSK for m > 16: very poor BER performance
m-PSK constellation: error probability
2-PSK,4-PSK
1 8-PSK
0.1 16-PSK
0.01
1E-3
1E-4
1E-5
1E-6
1E-7
BER

1E-8
1E-9
1E-10
1E-11
1E-12
1E-13
1E-14
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Eb/N0 [dB]