Digital Communication Transmitter: T N M 2 X N 1

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

Digital Communication

Transmitter
Introduction to Digital „ A simplified block diagram:
Modulation antenna

waveforms
bits
interleaver (symbols)
0110
RF
Instructor: M.A. Ingram User
Interface Encoder Modulator Front
more bits End
ECE4823 011001
Frequency
up-conversion
and amplification

Digital Communications
Receiver Symbols
„ A simplified block diagram: „ In each symbol period, Ts, a digital
antenna
waveforms modulator maps N coded bits word to a
(faded symbols
plus noise) bits
transmitted waveform from a set of
deinterleaver
RF 0110
User M=2N possible waveforms
Front Demodulator Decoder Interface
End
“soft detections”
„ Each waveform corresponds to an
Low-noise amplification of coded bits
011001
information symbol, xn
and frequency
down-conversion
„ For Binary symbols, N=1

Detection Definitions
„ The job of the receiver is to determine „ Bit Rate (bits per sec or bps)
which symbols were sent and to R = N / TS
reconstruct the bit stream that created „ Bandwidth Efficiency (bps/Hz)
them
ηB = R / B
where B is the bandwidth occupied by
the signal

1
Shannon Theorem Pulses
„ In a non-fading channel, the maximum „ A symbol period, Ts, suggests a
bandwidth efficiency, or Shannon localization in time
Capacity is „ Localization in frequency is also
necessary to enable frequency division
ηB MAX
= log 2 (1 + SNR ) multiplexing
„ Regulatory agencies provide spectral
SNR = signal-to-noise ratio masks to limit the distribution of power
in the frequency domain

Example: Linear Modulation Rectangular Pulses


„ s(t) is the output of the modulator „ Suppose p(t) is a rectangular pulse
„ g(t) is the complex envelope „ This pulse is not used in practice, but is OK
„ p(t) is the basic pulse for illustration
„ xn is the nth symbol
„ A is the amplification in the transmitter
p (t )
2
{
s (t ) = Re g (t )e j 2πf c t } TS

g (t ) = A∑ x n p (t − nTS ) t
0 TS
n

Binary Phase Shift Keying


(BPSK) BPSK-Modulated Carrier
„ For BPSK, each symbol carries one bit „ The information is in the phase of the
of information x n ∈ {− 1,1} carrier
g (t ) s (t )
2 1 0 0 1 0 1 1 2 1 0 0 1 0 1 1
A A
TS TS
2TS L L
t t
0 TS
2 2
−A −A
TS TS

2
Power Spectral Density (PSD)
of Linear Modulation Bandwidth Properties
„ Assume that the symbol sequence {xn} is iid „ The RF bandwidth of the modulated
and zero mean. Then, the PSD of s(t) is carrier is two times the baseband
bandwidth of
1
[ ]
Ss ( f ) =
2
S g ( f − f c ) + S g (− f − f c )
Sg ( f ) =
{ } P( f )
A 2 E xn
2
2
where
Sg ( f ) =
2
A E xn { } P( f )2
2
2TS

2TS which is clearly seen to depend on the


and P ( f ) is the Fourier Transform of p (t ) bandwidth of the pulse

Fourier Transform of the PSD for BPSK and the


Rectangular Pulse Rectangular Pulse
P ( f ) = 2TS sinc 2 ( fTS ) Ss ( f )
2
„

TS
2
„ |xn| =1
+∞
2
P( f ) = ∫ p (t ) exp (− j 2πft )dt = ∫ TS
exp (− j 2πft )dt
A 2 E xn { } P( f )
2
2 − fc fc
−∞ 0
Sg ( f ) =
2TS
2 exp (− j 2πft ) 2 exp (− j 2πfTS ) − 1
TS

= =
TS − j 2πf TS − j 2πf = A 2 sinc 2 ( fTS )
0

exp (− jπfTS ) 2
=
πf TS
sin (πfTS ) = exp (− jπfTS ) 2TS sinc ( fTS ) Ss ( f ) =
A2
2
[
sinc 2 ([ f − f c ]TS ) + sinc 2 ([− f − f c ]TS ) ]

Received Signal BPSK Demodulator


„ The received signal has been attenuated by path loss „ The output of the correlator is a sequence of
and multipath fading and has added noise noisy versions of the transmitted symbol
sequence
r (t ) = a (t ) A∑ xn p (t − nTS ) cos( 2πf c t + θ c ) + n (t ) demodulated
n
received symbols
.95,-1.03,-1.02,.93,-.85,.81,.74
(• )dt
nTS
2 1 1
r(t) ∫( n −1)TS
a (t ) A 0 0 1 0 1
TS To deinterleaver and decoder

Channel L
Gain t
1 2
cos (2πf c t )
Aa (t ) TS

Automatic Gain Control

3
Integrate and Dump Output Conditional PDFs
„ Consider one symbol period and the „ Recall R = xn ε b + vn
rectangular pulse: channel gains and TX
power lumped into ε b

 2ε b  2  
TS TS
2  
R = ∫ r (t )
TS
cos(2πf ct )dt = ∫  x n
TS
cos(2πf c t ) + n(t )  cos(2πf ct )dt
f R| xn = −1 (r ) =
1
exp −
(
 r+ ε 2
b)  1 ( )
 r− ε 2
 TS f R|xn =1 ( r ) = exp − 
b
0 0  N   2 N 0    N 
2π  0    N   2 0 
2π  0    
  2  
TS TS
xn 2  2    2 
= ε b ∫ 1 dt + ∫ n(t ) cos(2πf t )dt = x ε
c n b + vn  2   
TS 0
TS 0
r
„ vn is a zero-mean Gaussian RV with variance − εb εb
N0
σ v2 =
2

Conditional Probability of Bit Conditional Probability of Error


Error Expression
„ If xn is -1, then an error happens if R>0  
+∞ 
f R| xn = −1 (r | xn = −1)dr = Q
0 − − εb ( ) = Q 2ε b 

f R| xn =1 (r | xn = −1) ∫  N0   N0 
0
   
r
 2 
− εb εb
„ Same for other kind of error, by symmetry
P (error | xn = −1) = P ( R > 0 | xn = −1)
+∞
= ∫f
0
R| xn = −1 (r | xn = −1)dr

Unconditional Probability of Bit How to Improve System


Error Performance
„ Assume the two possible values of xn „ Increase symbol energy ε b
are equally likely „ Decrease average noise power N 0 / 2
1 1
P (error ) = P (error | xn = −1) + P (error | xn = 1)
2 2  2ε b 
P(error ) = Q 

 2ε b 
= Q   N0 

 N0 

4
References
„ [Rapp, ’02] T.S. Rappaport, Wireless
Communications, Prentice Hall, 2002
„ [Stuber, ’01] Gordon Stuber, Principles
of Mobile Communication, 2nd ed,
Kluwer Academic Publishers, 2001.

You might also like