Introduction to

Communication Systems
(EEEg3152)
Chapter V
Digital Representation and Transmission of
Analog Signals

Prepared by: Fisiha Abayneh

Email: [email protected]
 Chapter V
Digital Representation and
Transmission of Analog Signals

5.1 Introduction

5.2 Analog to Digital Conversion

5.3 Line Coding and Waveform Shaping

5.4 Overview of Digital Modulation Techniques

5.5 Time Division Multiplexing (TDM) and Digital

Carrier Systems

2
 5.1 Introduction

➢The use of analog signals in real world communication

is almost outdated due to the advancement of digital
communications systems which use digital signals.
❖ Digital systems are superior over analog ones in overcoming
noise.
❖Continuous development in integrated circuits (IC) make digital
systems more simple than analog circuits.

➢As a result, analog signals have to be converted to

Digital signals.
3
 5.2 Analog to Digital Conversion

➢Analog signal can be converted to digital signal in a

process which involves the following three steps:
1. Sampling
2. Quantization
3. Encoding
➢This process is also know as PCM (Pulse Code
Modulation).

➢The device which accomplish this task is called Analog

to Digital Convertor (ADC).

4
 5.2 Analog to Digital Conversion …

Analog Digital
Signal Signal

Sampling Quantization Encoding

5
 Sampling

➢Sampling is just a process in which instant values of an

analog signal are taken at fixed time intervals, say Ts ,
and all other values are ignored.
❖Ts is known as sampling interval.
𝟏
❖ 𝒇𝒔 = ; is called the sampling rate or sampling
𝑻𝒔

frequency.

➢This process is also referred as PAM (Pulse Amplitude

Modulation), and the result is a discrete time signal
with continuous values.
6
 Sampling …

. Sampling
Interval: Ts

Source: https://img.favpng.com/22/16/23/sampling-analog-signal-continuous-signal-digitization-png-favpng-
hnjd1GCDpqv0vXcjXu0B12k7u.jpg 7
 Sampling …

➢Practically, there are three sampling methods.

1. Ideal:
✓An impulse at each sampling instant (i.e: at nTs ; n=1,2,3,…)

2. Natural:
✓Short width pulse with varying amplitude (amplitude follows
the analog signal for the pulse duration)

3. Flat-top:
✓Short pulse with constant amplitude value, achieved by
sample and hold technique.

8
 Sampling …

9
Source: https://sipdtdevelopers.files.wordpress.com/2013/11/3.png
 Sampling …

➢The sampling frequency is determined based on

Nyquist theorem which states that the sampling rate
must be at least twice of the highest frequency of the
signal.
𝒇𝒔 ≥ 𝟐 × 𝒇𝒎𝒂𝒙

➢Notice that, prior to sampling, filtering the analog

signal to limit its maximum frequency is very important
as it affects the sampling rate.

10
 Sampling …

➢Under Sampling:-
❖Sampling below the Nyquist rate.
❖It causes a problem called aliasing, and as a result, the
original signal cannot be correctly reconstructed from the
samples.

➢Over Sampling:-
❖Sampling above the Nyquist rate.
❖It may result in a better quality, but it is associated with higher
bitrate which may mean unnecessary resource utilization and
higher cost.

11
 Sampling …

Source: https://www.researchgate.net/publication/327105431/figure/fig4/AS:661772210016258@1534790198987/Nyquist-
12
based-fixed-sampling-oversampling-and-undersampling-a-fixed-sampling-rate.png
 Sampling …

Source: https://www.researchgate.net/publication/327105431/figure/fig4/AS:661772210016258@1534790198987/Nyquist-
13
based-fixed-sampling-oversampling-and-undersampling-a-fixed-sampling-rate.png
 Sampling …

➢Example:- In a Telephone Communication system,

voice is digitized by assuming human voice is contained
below 4kHz. What would be the minimum sampling
rate required.

➢Solution:-
𝑓𝑠 ≥ 2 × 𝑓𝑚𝑎𝑥 = 2 × 4𝑘𝐻𝑧 = 𝟖𝒌𝑯𝒛

➢Therefore, the minimum sampling rate is 8000 samples

per second

14
 Quantization

➢During sampling, the continuous time analog signal is

changed to a discrete time signal, but its amplitude still
takes any random value.
❖i.e: The amplitude values of the sampled signal are
infinite between the min. and max. values of the
sample.

➢Quantization is process of mapping these amplitude

values in to a pre-determined finite set of values.

15
 Quantization …

➢The 1st step in quantization is to set a parameter called

quantization level, denoted by ‘L’.
❖L is a unitless value, commonly chosen to be in the powers of 2.
(… to be justified in the next sub-section)

➢Then, the quantization step size (also known as

quantization zone width) has to be determined.
❖It is denoted by Greek letter ′∆′ and can be determined by
dividing the difference between the maximum and minimum
sample value with L.

max − 𝑚𝑖𝑛
∆=
𝐿 16
 Quantization …

➢Next, the distance between the min. and max. sample

value is divided in to L quantization steps (or zones),
each with a size of ′∆′.

➢The mid-points of these quantization steps become the

new finite set of quantized values.
❖Any sampled value which fall in a specific quantization zone
will be approximated by the mid point value of that zone.

17
 Quantization …

Source: https://www.rohm.com/electronics-basics/ad-da-converters/what-are-ad-da-converters 18
 Quantization …

➢The process of quantization introduces a new type of

error which is known as quantization error.
❖This due to the approximation of actual values by new set of
defined values.

➢Quantization error can be reduced by increasing the

value of L.
❖Which means decreasing the size of quantization zones.
❖But, increasing L also means increased bit rate, which is not
desirable. So, finding the tradeoff will be the problem to handle.

19
 Encoding

➢It is a process of assigning binary code for each of the

new quantized values.

➢The number of bits required to encode a single

quantized value (denoted by nb and referred as number
of bits per sample) can be determined as:-
𝑛𝑏 = log 2 𝐿
❖So, this is why L is commonly chosen to be in the powers of 2,
but practically there are different ways of encoding too.

20
 Encoding …

➢Once the number of bits per sample is known, each

quantized value is mapped to a binary code of nb bits.
❖Note that the encoding scheme (which quantized value to which
code?) should be known to both transmitter and receiver.

➢After encoding, the signal is fully digital (stream of

bits), and it is also called a PCM signal.

➢The bit rate of a PCM signal can be calculated as:-

𝐵𝑖𝑡 𝑅𝑎𝑡𝑒 = 𝑛𝑏 × 𝑓𝑠

21
 5.2 Analog to Digital Conversion …

22
Source: https://blogs.synopsys.com/vip-central/2015/04/28/1787/
 5.2 Analog to Digital Conversion …

➢Example:- Suppose it is required to digitize an analog

signal which has a maximum frequency of 4kHz, and
after sampling, the minimum and maximum sampled
amplitudes are -15 and 25, respectively. Assuming
quantization level of 8:-
1. Calculate quantization step size.

2. Determine all quantization zones and quantized value set.

3. Calculate the required number of bits per sample.
4. Calculate the bit rate of the digital signal.
5. Develop a sample code for all quantized values. 23
 5.2 Analog to Digital Conversion …

Solution:-

1. Quantization step size:-

max − 𝑚𝑖𝑛 25 − (−15) 40

∆= = = =𝟓
𝐿 8 8
2. Quantization Zones and Quantized Value set.
❖This is determined by dividing the distance between the min.
and max.(-15 and 25) in to 8 zones each having a distance of ∆=
5.

❖Then, the mid points will be the set of quantized values.

24
 5.2 Analog to Digital Conversion …

Solution …
Quantization Zone Quantized Value
-15 to -10 -12.5

-10 to -5 -7.5

-5 to 0 -2.5

0 to 5 2.5

5 to 10 7.5

10 to 15 12.5

15 to 20 17.5

20 to 25 22.5

25
 5.2 Analog to Digital Conversion …

Solution …

3. Required number of bits per sample:-

𝑛𝑏 = log 2 𝐿 = log 2 8 = 𝟑 𝒃𝒊𝒕𝒔/𝒔𝒂𝒎𝒑𝒍𝒆

4. Bit Rate of the Signal:-

𝑓𝑠 = 2 × 𝑓𝑚𝑎𝑥 = 2 × 4𝑘𝐻𝑧 = 8𝑘𝐻𝑧
𝐵𝑖𝑡 𝑅𝑎𝑡𝑒 = 𝑛𝑏 × 𝑓𝑠 = 3𝑏𝑖𝑡𝑠 𝑝𝑒𝑟 𝑠𝑎𝑚𝑝𝑙𝑒 × 8 𝑘𝐻𝑧 = 𝟐𝟒 𝒌𝒃𝒑𝒔

26
 5.2 Analog to Digital Conversion …

Solution …

5. Sample Code
Quantization Quantized
Code
Zone Value
-15 to -10 -12.5 000

-10 to -5 -7.5 001

-5 to 0 -2.5 010

0 to 5 2.5 011

5 to 10 7.5 100

10 to 15 12.5 101

15 to 20 17.5 110

20 to 25 22.5 111 27
 Chapter V
Digital Representation and
Transmission of Analog Signals

5.1 Introduction

5.2 Analog to Digital Conversion

5.3 Line Coding and Waveform Shaping

5.4 Overview of Digital Modulation Techniques

5.5 Time Division Multiplexing (TDM) and Digital

Carrier Systems

28
 5.3 Line Coding and Waveform
Shaping
➢In analog to digital conversion, the analog signal is
converted to a digital signal or stream of bits (‘1’s an
‘0’s).

➢ But, if transmission is considered, the above operation

will not be enough.

➢The ‘1’s and ‘0’s have to be represented by some sort of

signal which is compatible for transmission through a
medium, or to facilitate any operation related to
transmission.
29
 5.3 Line Coding and Waveform
Shaping …
➢This process of changing the bit stream in to a
compatible signal is known as digital to digital
encoding, or simply line coding.

➢This process will also affect the bandwidth requirement

of the digital transmission.

➢The primary line encoding methods include:-

1. Nonreturn to Zero (NRZ) encoding
2. Return to Zero (RZ) encoding
3. Bi-phase encoding
4. Bipolar encoding
30
 1, Nonreturn to Zero (NRZ) Encoding

➢The basic feature of NRZ encoding is that the signal

level remains on the fixed voltage level assigned for that
bit for the entire bit duration.

➢The frequency of pulses is relatively smaller in this

encoding method, which results in smaller bandwidth
requirement.

➢If similar bits appear successively, it will be difficult for

the receiver to identify where the bits begin and end.
❖So, this method is not suitable for synchronous transmission.

31
 1, Nonreturn to Zero (NRZ) Encoding

➢There are two NRZ forms:-

❖Unipolar NRZ:-
✓Bit ‘1’ is represented by positive pulse
✓Bit ‘0’ will not have a pulse

❖Polar NRZ:-
✓Bit ‘1’ is represented by positive pulse
✓Bit ‘0’ by negative pulse.

32
 1, Nonreturn to Zero (NRZ)
Encoding…

➢Unipolar NRZ:-

Source: https://www.tutorialspoint.com/digital_communication/images/nrz.jpg 33
 1, Nonreturn to Zero (NRZ)
Encoding…

➢Polar NRZ:-

Source: https://www.tutorialspoint.com/digital_communication/images/polar_nrz.jpg 34
 2. Return to Zero (RZ) Encoding

➢In RZ encoding, the voltage level assigned to the bit

returns to zero sometime during the bit duration.

➢Like the NRZ, RZ also has unipolar and polar forms.

❖Unipolar RZ:-
✓+ve pulse for ‘1’ and nothing for ‘0’.
❖Polar RZ:-
✓+ve pulse for ‘1’ and –ve pulse for ‘0’.
✓Every bit has easily identifiable pulse.
✓This makes polar RZ preferable for synchronous
transmission compared to NRZ or unipolar RZ.

35
 2. Return to Zero (RZ) Encoding…

➢Unipolar RZ:-

36
Source:- https://www.tutorialspoint.com/digital_communication/images/unipolar_return_to_zero.jpg
 2. Return to Zero (RZ) Encoding…

➢Polar RZ:-

Source:- https://www.tutorialspoint.com/digital_communication/images/polar_rz.jpg 37
 3. Bi-phase Encoding

➢In bi-phase encoding, the binary bits are represented

not by +ve or –ve pulses, but by the transition of pulses.
✓There is always a transition at least at the middle of bit
duration.

✓This high frequency of pulses results in higher requirement

of bandwidth.

➢The two most common bi-phase encoding methods are:-

✓Manchester encoding (also known as bi-phase Manchester)
✓Differential Manchester encoding.

38
 3. Bi-phase Encoding…

Manchester encoding:-

➢The transition of a pulse at the middle of bit

duration determines bit representation.
✓Bit ‘0’ is represented by transition from high to low.
✓Bit ‘1’ is represented by transition from low to high.
✓NB:- this is based on IEEE 802.3 standard. But
opposite representation scheme is also available.

39
 3. Bi-phase Encoding…

40
Source:- https://media.geeksforgeeks.org/wp-content/uploads/ETHERNET_1.jpg
 3. Bi-phase Encoding…

Differential Manchester encoding:-

➢The transition of a pulse at the beginning of bit

duration determines bit representation.
✓Bit ‘1’ is represented by an absence of transition at the
beginning of the bit duration.
✓Bit ‘0’ is represented by a presence of transition at the
beginning of the bit duration.

41
 3. Bi-phase Encoding…

42
Source:- https://www.sharetechnote.com/image/Communication_Digital_Encoding_Manchester_01.png
 4. Bipolar Encoding

➢In bipolar encoding:-

❖Bit ‘0’ is not represented by any pulse. (just a zero level.)
❖Bit ‘1’ is represented by alternating +ve and –ve pulses.
✓i.e:- if the 1st bit ‘1’ takes +ve pulse, then the 2nd bit ‘1’ will
have –ve pulse, and the next bit ‘1’ will be +ve pulse …etc.

➢The alternating polarity of bit ‘1’ prevents the build up

of DC voltage level in the transmission line.
❖The +ve and –ve pulses approximately cancel each other, and
make the transmission line DC-balanced.

43
 4. Bipolar Encoding…

➢Bipolar encoding can have both NRZ or RZ forms.

➢The best example of bipolar encoding is AMI

(Alternative Mark Inversion).

➢The name AMI had an origin in the context of T-carrier

systems, where:-
❖“Mark” is used to refer binary bit ‘1’, and
❖“Space” for binary bit ‘0’.

44
 4. Bipolar Encoding…

45
Source: https://www.tutorialspoint.com/digital_communication/images/bipolar_signaling.jpg
 Waveform Shaping

➢The line encoded digital signal can be directly

transmitted using a transmission line (cable).
❖The best example for this is the telephone line.

➢But, if the intension is to transmit the line encoded

signal wirelessly using RF signal form, two further
processing steps are required.
❖Waveform shaping (best known as pulse shaping)
❖Analog modulation

46
 Waveform Shaping…

➢Pulse shaping is a process of changing waveform of

pulses to be transmitted.
❖The purpose is to make the signal better suited to
the transmission channel by limiting the signals
effective bandwidth.
❖This in turn helps to avoid ISI (Intersymbol
Interference).

➢Pulse shaping is achieve by filtering the pulses using

selected filters.
47
 Waveform Shaping…

➢Commonly used pulse shaping filters include:-

❖Sinc shaped filter
❖Raised-cosine filter
❖Gaussian filter

Amplitude response of
raised-cosine filter with
different roll-off factor

48
Source:- https://upload.wikimedia.org/wikipedia/commons/thumb/a/a1/Raised-cosine_filter.svg/400px-Raised-cosine_filter.svg.png