Telecommunications 1

• Week 1
Agenda

• Welcome
• Recap P1
• Recap encoding
• Digital modulation
– Pulse
– ASK, FSK, PSK
– QAM
• Other Order of Business
Thank you
Contract - how do we feel safe with each other
• Safety feeling enables learning
• Phones away
• There are no stupid questions
• Be on time
• Optional breaks
• Upload Canvas upfront
• Patience => slow down
• Optional assignments: Presentations protocols
Converting From Analog To Digital

Recap P1
Bandwidth / Representation
2000 bps
Increasing bandwidth improves the
representation of the data signal.
B=500 Hz
500Hz too low to reproduce the signal.
Want to maximize the capacity of the
B=1000 Hz available bandwidth.

B=1700 Hz

B=4000 Hz
Telecommunication block diagram
Steps Required to Generate PCM Streams

• Sampling: periodic measurement of the analog signals at

regular intervals
• Quantizing: assigning discrete values to samples
• Coding: assigned binary codes to samples using what is
known as the PCM code word
Sampling

(a)

(b)

(c)
Figure 2.2 : creating a PAM wave for a single sinusoid.
(a) is a sinusoid signal, (b) a pulse train, (c) the result of
passing (a) and (b) through a point by point multiplier.
Sampling

• Sampling rate: how often should we take measurements

of the analog signal
• at least at twice the rate of its highest frequency
component
• For a voice channel with a frequency range between 300
Hz and 3400 Hz (bandwidth of 3100 Hz) we need to take
a sample at least at a rate of 2 X 3100 = 6200 Hz or
every 1/6200 second
Quantization

• To represent samples by a fixed number of bits

• For example if the amplitude of the PAM signal range
between -1 and +1 there can be infinite number of values.
For instance one value can be -0.2768987653598364834634
• For practicality, we may use 20 different discrete values
between -1 and +1 volts
• Each value at a 0.1 increment
Quantization: the binary world

• total number of discrete values: binary number multiple

(i.e., 2, 4, 8, 16, 32, 64, 128, 256, and so on)
• facilitates binary coding
• For instance, if there were 4 values they would be as
follows: 00, 01, 10, 11 – this is a 2-bit code
Quantization: 16 coded quantum steps
+1
• Between -1 and + 1 volts signal 15 : 1111
14: 1110

• 16 discrete steps Range of standard

13: 1101
12: 1100
11: 1011
• each step at 0.125 volts increment values (V)
0
10: 1010
9: 1001

or decrement from the adjacent step 8: 1000

7: 0111
6: 0110
0 0000 0v 5: 0101
4: 0100
3: 0011
1 0001 0.125v 2: 0010
-1 1: 0001
2 0010 0.25v 0: 0000
8 9 10 1112 13 12 11 10.. 6 .........

3 0011 0.375v Coded values

4 0100 0.500v
Figure 2.4: quantization and resulting coding
5 0101 0.625v using 16 quantizing steps
Quantization Distortion
• Quantization error is the different between the quantum value and the true
value
• More steps reduce quantizing distortion in linear quantization
• This will require higher bandwidth, since we need more bits for each code
word
• Voice represent a problem because of the wide dynamic range, the level
from the loudest syllable of the loudest talker to the lowest syllable of the
quietest talker
• S/D = 6n + 1.8 dB EX: 7 bit PCM cod 6.7 + 1.8 = 43.8
• practical system S/D = 30 - 33 dB
Companding algorithms
Segment # Voltage levels

• Compression/Expanding
• Non-linear
5.0

• The voltage level between 3.0

the loudest and the lowest 2

1.5
is segmented in non-linear 3
0.5
manor 4
5
0

• The voltage range of each 6

segment varies according to 7

the level of the voltage

8

Figure2_5: Nonlinear quantization using 8 segments with each

segment assigned two steps (two coded words)
Modulation - Digital Data, Analog Signal

• Public telephone system

– 300Hz to 3400Hz
• Guardband from 0-300, 3400-4000Hz
– Use modem (modulator-demodulator)
• Amplitude shift keying (ASK)
• Frequency shift keying (FSK)
• Phase shift keying (PSK)
Modulation recap
(analog carrier,
analog mod wave)

em = Em cos mt

ec = Ec sin (ct + )
Amplitude Modulation and ASK
• Values represented by different
amplitudes of carrier
• Usually, one amplitude is zero
– i.e. presence and absence of
carrier is used
• Susceptible to sudden gain
changes
• Inefficient
• Up to 1200bps on voice grade
lines
• Used over optical fiber
Frequency Modulation and FSK
• Values represented by different
frequencies (near carrier)
• Less susceptible to error than
ASK
• Up to 1200bps on voice grade
lines
• High frequency radio
• Even higher frequency on
LANs using co-ax
Phase Modulation and PSK

• Phase of carrier signal is

shifted to represent data
• Differential PSK
– Phase shifted relative to
previous transmission rather
than some reference signal
Sending Multiple Bits Simultaneously
• Each of the three modulation
techniques can be refined to send more
than one bit at a time. It is possible to
send two bits on one wave by defining
four different amplitudes.
• This technique could be further refined
to send three bits at the same time by
defining 8 different amplitude levels or
four bits by defining 16, etc. The same
approach can be used for frequency
and phase modulation.
• In practice, the maximum number of bits
that can be sent with any one of these
techniques is about five bits.
The solution is to combine
modulation techniques.
Constellation Space
• Create 2-axis (e.g. sine and cosine) actually it could be a n-dimensional
hyper-plane
• Express digital modulation alphabet as points in the hyper-plane. The farther
apart the points are in the space, the more immunity there is against noise
and interference.
• More distance, better error performance. Keep this in mind.
• The maximum power is the length of the longest vector. The average
transmitter power is the average distance squared of all the points.
Analog I/Q Modulation
Baseband Input

iti(t)
(t) it (t)
t
cos2 f 0t  1t
2cos(2f
yt (t)
)sin2 f0t 

qq(t)
t(t)

t qt (t)

• Analog signals take on a continuous range of values (as viewed

in the time domain)
• I/Q signals are orthogonal and therefore can be transmitted
simultaneously and fully recovered

Lecture 9 Fall 2006

Polar View of Analog I/Q Modulation

it (t)  i(t )cos2 fot  0

i(t)
it(t) it (t)
qt (t)  q(t )cos2 fot  90   q(t )sin2 fot 

cos2 f 0t 
yt (t)
yt (t)  i 2(t )  q2(t ) cos2  f ot  (t)  sin2 f0t 

q(t)
qt(t)
where (t )  tan1 q(t )/ i(t) qt (t)

180    180

Lecture 9 Fall 2006

Polar View of Analog I/Q Modulation (Con’t)
• Polar View shows amplitude and phase of it(t), qt(t) and yt(t)
combined signal for transmission at a given frequency f.
• Magnitude of i(t) and q(t) vary with time, representing information in
the analog domain.

Q Q

y(t) i2 (t)  q2 (t)

q(t)
(t) i(t)
I I
i(t) q(t)
y(t) (t)
i2 (t)  q2 (t)

Lecture 9 Fall 2006

Frequency Domain View of Analog I/Q
Modulation 1 1
Transmitter Output
f
II(f)
t(f -fo 0 fo It(f)
)1 1 1
i(t)
it(t)

f f
-fo 0 fo
0 2cos2 f 0t  y(t)
Q(f)
Qt(f 2sin2 f0t 
1 j Y
Qqt(f)
(f)
q(t)
fo
f f
0 j -fo 0
ffo1 -j
f
-f1o 0
-j

• Takes advantage of coherent receiver’s sensitivity to

phase alignment with transmitter local oscillator
– We have two orthogonal transmission channels (I and Q)
available to us
– Transmit two independent baseband signals (I and Q) with two
sine waves in quadrature at transmitter
Lecture 9 Fall 2006
Analog I/Q Modulation-Transceiver
Baseband Input Receiver Output

Lowpass
it(t) ir(t)

t t
2cos2 f0t  2cos2 f0t 
2sin2 f0t  2sin2 f0t 
Lowpass
qt(t) qr(t)
t t

• I/Q signals take on a continuous range of values (as

viewed in the time domain)
• Used for AM/FM radios, television (non-HDTV), and the
first cell phones
• Newer systems typically employ digital modulation
instead

Lecture 9 Fall 2006

I/Q Transceiver Frequency Domain View
1 1 1 1
Transmitter Output Receiver Output
f f
It(f) -fo 0 fo Yi(f)
-fo 0 fo Ir(f)
1 1 1 Lowpass 2
it(t) ir(t)

f f f
-fo 0 fo
0 2 cos2 f 0t  y(t) y(t) 2 cos2 f 0t  0
Qt(f) 2 sin2 f0t  2 sin2 f0t  Lowpass
Qr(f)
1 j Yq(f) 2
qt(t) qr(t)
fo
f f f
0 j -fo 0 j 0
ff1o -j ffo1
f f
-f1 0 -f1 0
-f o -j
-fo -j

• Demodulate using two sine waves in quadrature at

receiver
– Must align receiver LO signals in frequency and phase to
transmitter LO signals
• Proper alignment allows I and Q signals to be recovered as shown
Lecture 9 Fall 2006
Impact of 90 Degree Phase Misalignment
1 1 j
Transmitter Output f1 Receiver Output
f f
It(f) -fo 0 fo Yi(f) -f1 0 Ir(f)
1 1 1 -j 2
it(t) ir(t)

f f f
-fo 0 fo
0 2cos(2f 2t) y(t) y(t) 2sin(2f 1t) 0
Qt(f) Qr(f)
2sin(2f2t) -2cos(2f1t)
1 j Yq(f)
qt(t) qr(t)
fo
f f f
0 j -fo 0 0
f1 -j -fo fo -2
f f
-f1 0 0
-j -1 -1

• I and Q channels are swapped at receiver if its LO signal

is 90 degrees out of phase with transmitter
– However, no information is lost!
– Can use baseband signal processing to extract I/Q signals
despite phase offset between transmitter and receiver

Lecture 9 Fall 2006 31

Digital I/Q Modulation

Baseband Input
io
it(t)
i(t) it (t)
io
t yt (t)
2cos(2fot)
t 2sin(2 fot)
qo

qq(t)
t(t)
qo qt (t)
t
t

• I/Q signals take on discrete values at discrete time instants

corresponding to digital data

Lecture 9 Fall 2006 32

Polar View of Digital I/Q Modulation
Polar View shows amplitude and phase of it(t), qt(t) and yt(t) combined signal for
transmission at a given frequency f.
i(t) and q(t) have discrete values. In this case, binary values. ±io ±qo

it (t )  io cos2 fot 

it(t) qt (t)  qo sin2 fot 

qt(t) Q
Q
 qo
io io I
  I
 qo

yt (t)  io2  qo 2 cos 2  f ot  (t) 

qo
where (t)  tan1
io

Lecture 9 Fall 2006 33

Polar View of Digital I/Q Modulation (cont’d)
it(t)
i(t ) it (t ) Q

t  
2cos(2fot) yt (t )
2sin(2fot)
io2  q o2
qq(t 4 QAM
t(t)
) 45
I Quadrature
q t (t ) Amplitude
t Modulation
Given io  qo  1
 
yt (t)  2
(t ) can have 4 values
45, 135,  45,  135

Transmission signal is sine wave at frequency f0 with information

encoded in discrete values of amplitude and phase.

Lecture 9 Fall 2006 34

Digital Modulation: 16-QAM
Baseband Input
it(t)

t 2cos(2f1t)
2sin(2f1t)

qt(t)
t

• I/Q signals take on discrete values at discrete time instants

corresponding to digital data
• I/Q signals may be binary or multi-bit
• – Multi-bit shown above (4 levels each)
Lecture 9 Fall 2006 35
Constellation Diagram:16-QAM
Q
00 01 11 10

00

Decision 01
Boundaries I
11

10

Decision
Boundaries

• We can view I/Q values at sample instants on a two-dimensional

coordinate system
• Decision boundaries mark up regions corresponding to different data
values
• Gray coding used to minimize number of bit errors that occur if
wrong decisions made due to noise
• Effectively combining phase and amplitude modulation
Lecture 9 Fall 2006 36
Case Study 1: ASK
• If m(t) = {0,1} and we amplitude modulate a carrier with m(t)
then the modulation is called on/off keying (OOK) or 2-amplitude
shift keying (2-ASK)

• 2-ASK, (points are at (0,0), and (0,1), in the 2 dimensional (sine,

cosine plane). Minimum distance between points is 1 for 1 unit of
power, and 1 bit per symbol.
• Distance between points corresponds to error performance
Case Study 2: Multi-Level ASK
•If maximum power is normalize to 1 then points are at
(0,0), (0,1/3), (0,2/3), (0,1). Distance is reduced from 2-ASK and performance is
worse. Requires 3x or 9x power to maintain 1 unit of distance.
• From Shannon, as we add more information in a fixed bandwidth, it becomes
increasingly expensive in terms of SNR to add more data.
Case 3: Orthogonal FSK
•Frequencies are chosen so that the waveforms are orthogonal
over the period of the bit T.
• Points are at (0,1) and (1,0) for 2-FSK. Distance is sqrt(2). Error
performance better than 2-ASK but not as good as others.
Case 4: QPSK and PSK
y(t)
y(t)

-A A x(t)
-A A x(t)

-A

Example signal constellation

diagram for BPSK signal.
Higher Order Modulations Very Inefficient in
terms of Power
Performance of Digital to Analog Modulation
Schemes
• Bandwidth
– ASK and PSK bandwidth directly related to bit rate
– FSK bandwidth related to data rate for lower frequencies, but to offset
of modulated frequency from carrier at high frequencies
– (See Stallings for math)
• In the presence of noise, bit error rate of PSK and QPSK
are about 3dB superior to ASK and FSK
Pulse Amplitude Modulation !remember

Analog voice data must be translated into a series of binary digits

before they can be transmitted.
With Pulse Amplitude Modulation, the amplitude of the sound wave
is sampled at regular intervals and translated into a binary
number.
The difference between the original analog signal and the translated
digital signal is called quantizing error.
Pulse Code Modulation
Pulse Code Modulation is the most commonly used technique in the
PAM family and uses a sampling rate of 8000 samples per
second.
Each sample is an 8 bit sample resulting in a digital rate of 64,000
bps (8 x 8000).
Sampling Theorem: If a signal is sampled at a rate higher than twice
the highest signal frequency, then the samples contain all the
information of the original signal.
E.g.: For voice capped at 4Khz, can sample at 8000 times per
second to regenerate the original signal.
Pulse Code Modulation(PCM) (1)
• If a signal is sampled at regular intervals at a rate higher than twice the
highest signal frequency, the samples contain all the information of the
original signal
– (Proof - Stallings appendix 4A)
• Voice data limited to below 4000Hz
• Require 8000 sample per second
• Analog samples (Pulse Amplitude Modulation, PAM)
• Each sample assigned digital value
Pulse Code Modulation(PCM) (2)

• 4 bit system gives 16 levels

• Quantized
– Quantizing error or noise
– Approximations mean it is impossible to recover original exactly
• 8 bit sample gives 256 levels
• Quality comparable with analog transmission
• 8000 samples per second of 8 bits each gives 64kbps
Delta Modulation

• Analog input is approximated by a staircase function

• Move up or down one level () at each sample interval
• Binary behavior
– Function moves up or down at each sample interval
Delta Modulation - example
Delta Modulation - Operation
Delta Modulation - Performance

• Good voice reproduction

– PCM - 128 levels (7 bit)
– Voice bandwidth 4khz
– Should be 8000 x 7 = 56kbps for PCM
• Data compression can improve on this
– e.g. Interframe coding techniques for video
Advantages of Digital Modulation

• Allows information to be “packetized”

– Can compress information in time and efficiently send as packets
through network
– In contrast, analog modulation requires “circuit-switched” connections
that are continuously available
• Inefficient use of radio channel if there is “dead time” in information flow
• Allows error correction to be achieved
– Less sensitivity to radio channel imperfections
• Enables compression of information
– More efficient use of channel
• Supports a wide variety of information content
– Voice, text and email messages, video can all be represented as digital
bit streams

Lecture 9 Fall 2006 51