Fundamentals of Digital Transmission
Fundamentals of Digital Transmission
Fundamentals of Digital Transmission
Fundamentals of Digital
Transmission
Baseband Transmission (Line codes)
Bit Value 0 1 0 1 1 1 0 0 1
ON-OFF
or 5V
Unipolar
(NRZ)
Non-Return-to-Zero 0V
Bit Value 0 1 0 1 1 1 0 0 1
5V
Polar
(NRZ)
0V
–5 V
Performance Criteria of Line Codes
Zero DC value
Inherent Bit-Synchronization
Rich in transitions
Average Transmitted Power
For a given Bit Error Rate (BER)
Spectral Efficiency (Bandwidth)
Inversely proportional to pulse width.
Comparison Between On-Off and Polar
Zero DC value:
Polar is better.
Bandwidth:
Comparable
Power:
BER is proportional to the difference between the two levels
For the same difference between the two levels, Polar
consumes half the power of on-off scheme.
Bit Synchronization:
Both are poor (think of long sequence of same bit)
More Line Codes
Bit Value 0 1 0 1 1 1 0 0 1
On-Off RZ 5V
Better synch.,
at extra
bandwidth
0V
Bit Value 0 1 0 1 1 1 0 0 1
Bi-Polar 5V
Better synch.,
at same
bandwidth
0V
–5 V
More Line Codes
Bit Value 0 1 0 1 1 1 0 0 1
Polar RZ
Perfect synch 5V
3 levels
0V
–5 V
Bit Value 0 1 0 1 1 1 0 0 1
Manchester
(Bi-Phase) 5V
Perfect Synch.
2 levels
0V or -5V
Spectra of Some Line Codes
1.2 On-Off
(NRZ)
1
0.8
Bipolar (NRZ)
power density
0.6
0.4
Manchester
0.2
0
0
2
0.2
0.4
0.6
0.8
1.2
1.4
1.6
1.8
-0.2
fT
Pulse Shaping
The line codes presented above have been
demonstrated using (rectangular) pulses.
There are two problems in transmitting such
pulses:
They require infinite bandwidth.
When transmitted over bandlimited channels
become time unlimited on the other side, and spread
over adjacent symbols, resulting in Inter-Symbol-
Interference (ISI).
Nyquist-Criterion for Zero ISI
Use a pulse that has the following characteristics
1 t 0
p (t )
0 t T b , 2T b , 3T b ,
t
-6Tb -5Tb -4Tb -3Tb -2Tb -Tb Tb 2Tb 3Tb 4Tb 5Tb 6Tb
0
-2 -1 0 1 2 3 4
-1
More on Pulse Shaping
The sinc pulse has the minimum bandwidth
among pulses satisfying Nyquist criterion.
However, the sinc pulse is not fast decaying;
Misalignment in sampling results in significant ISI.
Requires long delays for realization.
There is a set of pulses that satisfy the Nyquist
criterion and decay at a faster rate. However,
they require bandwidth more than Rb/2.
Raised-Cosine Pulses
1 b / 2 b
1 sin x
2 2x 2
b
P ( ) 0 x
2
b
1 x
2
P() x x
b/2
=/Tb
b/2 – x b/2 + x
Extremes of Raised-Cosine Spectra
P() x = b/2
x = 0
“Sinc”
b/2 b/2
b/2 b
=/Tb =2/Tb
Speech
Audio
Fax
Coloured Image
Video
Speech (PCM)
B = 3.4 kHz
Rs = 8000 samples/sec
Encoding = 8 bits/sample
Transmission rate = 64 kbps
Required bandwidth (passband) = 64 kHz
One hour of speech = 64000x3600 = 230.4 Mb
Audio
B = 16-24 kHz
Rs = 44 000 samples/sec
Encoding = 16 bits/sample
Stereo type = 2 channels
Transmission rate = 1.4 Mbps
Fax
Resolution 200x100 pixels/square inch
1 bit/pixel (white or black)
A4 Paper size = 8x12 inch
Total size = 1.92 Mb = 240 KB
Over a basic telephone channel (3.4 kHz, baseband) it
takes around 4.7 minutes to send one page.
Colour Image (still pictures)
Resolution 400x400 pixels/inch square
8 bits/pixel
3 colours/photo
A 8x10 inch picture is represented by
307.2 Mb = 38.4 MB !
Video (moving pictures)
Size of still pictures
15 frames/sec
307 Mb/frame x 15 frames/sec = 4605 Mbps =4.6 Gbps !!
Solutions
Compression
reduces data size
M-ary communication
Expands channel ability to carry information
M-ary Transmission
In the binary case one pulse carries one bit.
Let each pulse carry (represent) m bits.
Bit rate becomes m multiples of pulse rate
We need to generate 2m different pulses.
They can be generated based on:
Multiple Amplitudes (baseband and passband)
Multiple Phases (passband)
Multiple frequencies (passband)
Some combination (Amplitude and Phase).
Signal Constellation
Signal constellation is a convenient way of
representing transmitted pulses.
Each pulse is represented by a point in a 2-dimensional
space.
The square of the distance to the origin represents the
pulse energy.
The received signals form clouds around the
transmitted pulses.
A received points is decoded to the closest pulse point.
Multiple Amplitudes (PAM)
typical noise
4 “phase” 8 “phases”
2 bits / pulse 3 bits / pulse
2×B bits per second 3×B bits per second
Quadrature Amplitude Modulation (QAM)
QAM Bk 16 QAM Bk
Ak Ak
4 “levels”or pulses
2 bits / pulse 16 “levels” or pulses
2xB bits per second 4 bits / pulse
4xB bits per second
The Modulation Process of QAM
Ak x Yi(t) = Ak cos(c t)
cos(c t) + Y(t)
Bk x Yq(t) = Bk sin(c t)
sin(c t)
QAM Demodulation
Y(t) x LPF Ak
2cos(c t)
2cos2(ct)+2Bk cos(ct)sin(ct)
= Ak {1 + cos(2ct)}+Bk {0 + sin(2ct)}
x LPF Bk
2sin(c t)
2Bk sin2(ct)+2Ak cos(ct)sin(ct)
= Bk {1 - cos(2ct)}+Ak {0 + sin(2ct)}