Chapter I

Transmission of data

Brief look on the course:

Industrial Networks Communication Systems is a very basic course on telecommunication

and network systems, this course is intended for third year bachelor of automatic. This course
will present fundamental knowledge on different types of computer network, such as LAN
(Local Area Network), MAN (Metropolitan Area Network) and WAN ( Wide Area Network).
This course presents also the most basic principles of telecommunication protocols and digital
modulation, for instance ASK, PSK and QAM, different codes used for telephone are also
given, especial NRZ code and Miller code. The course presents also the principle of modem
emission and transmission, time and length of message required between two stations.

I-Historical view of telecommunication

In the first stage, illustrated by Figure 1, the voice and data streams are of different
functional and physical nature. Each system has its own network. Note that the transmission
of data on the telephone network was prohibited in 1960. When this service was liberalized,
the authorized speed was first limited to 1,200 bit/s, then 2,400 bit/s in 1976 and 4,800 bit/s in
1980.
In the second step, the voice is digitalized. The physical flows are commonplace. They can
be transported by the same network (transport network). However, the access networks
remain functionally different and users still access services through separate channels.

Figure 1. Schematization of the evolution of telecommunications

The third stage pursues the digitalization of flows. The voice is not only digitalized, but the
different pieces of information are gathered in packets, like data. This is then referred to as
“packetized voice”, thus allowing identical end-to-end processing for the two streams. In this
approach, the transport protocol is identical, but the user protocols remain different. The user
now only needs a single physical access to the transport network (voice/data network). The
streams are separated by equipment (voice/data equipment) located at the user's premises and
are processed by different systems.

The fourth step consists of a complete integration, the terminal equipment has an identical
access interface but different application functionalities. Voice and data can not only coexist
on the same network, but collaborate in the final computer applications: this is native
computer telephony coupling. In this approach, the protocols used in the transport network
and those used in the user network are identical for the two types of flows.

II- Standardization

In telecommunications, standardization comes from various organizations. From

manufacturers' groups to international organizations, standardization covers all areas of
communication.
-ETSI (European Telecommunications Standard Institute) which standardizes public
networks and their means of access. The main manufacturer groups are:
–ECMA (European Computer Manufacturers Association), originally made up solely of
European manufacturers (Bull, Philips, Siemens, etc.), the ECMA now includes all the major
global manufacturers (DEC, IBM, NEC, Unisys, etc.). .).In terms of telecommunications,
– EIA (Electronic Industries Association) known, mainly, for recommendations RS232C, 449
and 442. The main national organizations in which manufacturers, administrations and users
participate are:
– AFNOR, French Association for NORmalisation,
– ANSI, American National Standard Institute (USA),
– DIN, Deutsches Institut für Normung (Germany), well known for its standardization of
connectors (DIN sockets);
– BSI, British Standard Institute (Great Britain). International organizations:
–ISO, International Standardization Organization, brings together around 90 countries. ISO is
organized into Technical Committees (TC) of around 200, divided into Sub Committees (SC)
which are themselves subdivided into Working Groups (WG);

III- Digitalization of information:

Digitalizing an analog quantity consists in transforming the continuous sequence of values

into a discrete and finite sequence. To this end, a sample of the signal is taken at significant
instants and its amplitude is expressed in relation to a finite scale (quantification).
The analog-to-digital converter converts analog voltages from the amplifier into an 8-bit, 12-
bit, or 16-bit digital signal. This operation is conditioned by three essential criteria, namely:

- The sampling frequency,

- Conversion time,
- The resolution.

1) Sampling:
The sampling frequency ideally must be very high for the signal to be perfectly represented,
but an infinitely high frequency is not possible for technical reasons (the conversion time is
limited), for this the minimum sampling frequency is given by Shannon's theorem expressed
by:

Fsample  2.Fsignal (max)

The sampling signal must be at least twice the largest frequency of the signal. Example a
physiological signal such as The ECG has a P wave having a frequency of 10 HZ and the Q,
R and S waves each having a frequency of 50 HZ and finally the T wave having a frequency
of 5 Hz, therefore the frequency minimum sampling rate is 2*50Hz = 100HZ.

2) Conversion time:

It is the time necessary for the converter to convert an analog value into a digital value, the
latter varies from 1 micro second up to 1 milli-second, for signals having a high frequency a
very short conversion time is required.
 Analog Voltage

Figure 2, Sampling
T: Sampling period, t: Conversion time
3) Resolution

The resolution of an analog to digital converter is the smallest value it can convert, a high
resolution converter is a converter that can convert the smallest value possible.

Example:
A converter of 8 bits and of maximum value has to convert from 5 V, with for resolution:
(5/255)=0,01960 V.
8 bits in binary (11111111) -----in decimal (255) in hexadecimal (FF).

For 5V---- (11111111) binary ------------- (255) decimal

X is the smallest value of voltage (00000001) binary --------------- (1) decimal

Then, X = (5/255) V

In general, the resolution = the maximum value to convert / the maximum number of binary
converted into decimal.

IV- Different types of transmission:

The transmission of information between two correspondents can be unidirectional (the

exchange only takes place in one direction), we then speak of a simplex link (Figure 3). Each
correspondent fulfills only one function, he is a transmitter (source) or a receiver (sink or
collector). If the correspondents can alternately fulfill the functions of sender and receiver, the
link is said to be: link to the alternative or half duplex. The time taken by systems to switch
from one function to another is called turnaround time. This time can be significant, up to 1/3
of a second.
When the exchange can take place simultaneously in both directions, on separate channels or
on the same channel using special techniques such as frequency multiplexing2, the link is
called full bidirectional or full duplex.

Figure 3. Different types of links

IV-1 Parallel serial transmission:

Parallel transmission (Figure 4) is characterized by a simultaneous transfer of all the bits of

the same word. It requires as many conductors as there are bits to be transmitted and a
common conductor (asymmetric link) or as many pairs of wires if the ground is not common
(symmetrical link). Parallel transmission is very efficient in terms of throughput. It is used for
links between a computer, its peripherals and its slave calculation units. For example, the
HiPPI (High Performance Parallel Interface) interface which defines a mode of transmission
between a computer and its peripherals offering a speed of 800 Mbit/s. It uses a 50-pair cable
of which 32 are used for data transmission (parallel transmission per 32-bit word), HiPPI is
limited to 25 m.

Figure 4. Serial parallel transmission

In serial transmission, all the bits of a word or message are transmitted successively on the
same line. In computers, data (bits) are processed in parallel (bus). Serial transmission
requires a conversion interface to serialize the bits in transmission (parallel/serial conversion)
and rearrange them in reception (serial/parallel conversion). Serial transmission is used, for
data transmission, as two conductors. At a lower cost, it is suitable for transmissions over long
distances.

IV-2 Synchronous and asynchronous transmission:

The bits are transmitted on the line at a certain rate. This rate is defined by a clock called the
send clock. To correctly decode the sequence of bits received, the receiver must examine what
is happening to it at a rate identical to that of the transmission of the bits on the medium. The
receiver and transmitter clocks must "beat" in harmony.
It is not enough for the clocks to beat at the same rate, it is also necessary that the
instants of analysis of the electrical levels of the line are the same for the two elements, they
are said to be in phase. The operation of enslaving the receive clock to the transmit clock is
called synchronization.
In asynchronous transmissions the clocks are independent; on the contrary, in synchronous
transmissions, a strict phase relationship is permanently maintained between the send and
receive clocks.

Asynchronous transmission: In asynchronous transmissions, the sender and receiver clocks

are independent. To ensure the synchronization of the clocks, a specific servo signal is sent
before any significant binary sequence. After this operation, the receiving clock is free, it
drifts. The time interval, during which the drift is tolerable and allows correct decoding of the
binary sequence, is small. This time interval only allows the transmission of a short binary
sequence: the character.

In asynchronous transmission, figure 5, the characters transmitted are preceded by a

synchronization signal: the start bit. Between each character, to guarantee detection of the
next start bit, the line is reset to the zero state. This minimum idle time varies from 1 to 2 bit
times, it constitutes the stop bit or bits. The resting level of the line or zero level is fixed at a
certain potential (V) and not at the electrical zero so as not to confuse a binary zero with a
break in the line, figure 6. This resting voltage signals to the systems that the terminals are
active.

Figure 5. Asynchronous Transmission

Figure 6. Character mode and block mode

Synchronous transmission In synchronous transmission, the synchronization of the

transmission and reception clocks is maintained throughout the transmission by a particular
signal: the synchronization signal, figure 7. It is then possible to transmit large blocks.
However, between each transmitted block, the reception clock is no longer controlled and
drifts. Each block transmitted is therefore preceded by a synchronization sequence which will
also be used to delimit the start and end of the block.

Synchronisation Commande 8 bits Blocs de caractères Contrôle 8 bits

8 bits de données

Figure 7. Synchronous transmission structure

V- Bandwidth of a transmission:

Therefore, to study the behavior of a system in a non-sinusoidal regime, it can be studied for
each of the components of the signal to which it will be subjected. The frequency response of
this system is obtained by using a generator whose frequency is varied at constant voltage
(frequency generator). The measurement of the power at the output of the system makes it
possible to draw a curve, called the frequency response curve (Figure 8).

Figure 8. Bandwidth of a transmission

The transmission systems (lines, amplifiers, etc.) do not transmit all the harmonics of the
signal in the same way. Signals are transmitted with low distortion up to a certain frequency
called the cutoff frequency. Above this frequency, all harmonics are strongly attenuated. We
call bandwidth Figure 8) the space of frequencies such that any signal belonging to this
interval, undergoes, at most, only one attenuation determined with respect to a reference level.
The attenuation, expressed in decibels (dB), is given by the relationship:

A = 10log10 (P1/P0)

P1: output signal power

P0: reference signal power Bandwidth is typically set for half power attenuation, which is –3
dB.
The bandwidth of a signal corresponds to the minimum bandwidth that the system must have
to correctly restore the information. Thus, the bandwidth that qualifies the system is the
frequency bandwidth with different signals.

V-1 Filter concepts

The low-pass filter can only be "crossed" by the low frequencies, it attenuates the high
frequencies, figure 9. Conversely, the high-pass filter attenuates the low frequencies. In
principle, a transmission system is presented as both a low-pass filter and a high-pass filter, it
lets a certain frequency band pass, it is a band-pass filter. The signal to be transmitted must
take these characteristics into account.
 Figure 9, different types of filters

Two data characterize a filter:

– The cutoff frequency (fc), or frequency from which it is considered that all the higher and
(or) lower frequencies are attenuated by a given amount (generally –3 dB).
– The slope of the attenuation curve which is expressed in dB per octave.

V-2 Characteristic impedance

A transmission line consists of 2 copper conductors separated by an insulator. Figure 10

models an element of a line by materializing its physical components. It presents to the
electric current a resistive effect (R) responsible for signal attenuation, reactive effects which
break down into the salt effect (L) and the capacitive effect (C), and finally the conductance
(G) which expresses the loss by resistive effect between the two conductors (generally
negligible).

Zc Zr

Figure 10, Input Impedance Model

The impedance (Z) of the line element of length dl is the du/di ratio. The notion of alternating
current impedance covers a notion similar to that of direct current resistance, it is expressed in
ohms (V). The du/di ratio for a line assumed to be of infinite length is called the characteristic
impedance noted Zc:

√ √

In the ideal case, Ideal transmission we have Zc = Zr

V-3 Bandwidth:

The signal to be transmitted must best be adapted to the physical constraints of the
transmission system. Two types of adaptation or transmission techniques are possible (Figure
11): – The first consists of slightly modifying the signal, it is essentially intended to reduce
the DC component. However, the high-frequency components being strongly attenuated, the
transmission will be limited in distance: this is baseband transmission. – The second translates
the spectrum of the signal to be transmitted into a frequency band better accepted by the
system, this is broadband transmission.

Figure 11, Modulation baseband and Large band

V-3-1 Definition:

We qualify as baseband transmission systems those systems which do not introduce a

frequency difference between the signals transmitted and those received, Figure 12. This
definition in no way excludes modifications of the signal to better adapt the adapter. to the
characteristics of the transmission medium. Coding is the operation that matches a symbol
belonging to an alphabet to a binary representation (source coding). Transcoding, or online
coding, is the operation which consists of replacing the digital signal (binary representation)
with an electrical signal better suited to transmission. This transformation is carried out by an
encoder/decoder called Baseband Transmitter/Receiver.

Figure 12, Coding and decoding of information signal

V-3-2 Principal Codes:

NRZ Code: By symmetrizing the signal with respect to the reference potential (0 volt), the
DC component is reduced, figure 13.

Figure 13, NRZ code

To do this, we represent the 1s (or 0s) by a value + V and the 0s (oules1) by – V. This
elementary coding known under the name of NRZ code (No Return to Zero) is at the basis of
all codes. However, the spectrum of this signal is relatively broad. It presents a maximum
power at zero frequency, which corresponds to a significant DC component
The NRZ coding symmetrizes the value 1 and the value 0 with respect to a zero potential
level. However, this coding has a non-zero continuous component and does not present any
transition during long sequences of 0 or 1.

Manchester code: With a transition in the middle of each bit time, Manchester coding
remedies the absence of synchronization information. The transition is increasing for 0s,
decreasing for 1s, figure 14.

Figure 14, Manchester code

The direction of the transitions is significant, which poses problems in the event of inversion
of the connecting wires. Multiplying transitions, Manchester coding has a very broad
spectrum, it is used in local networks such as Ethernet over coaxial cable. The bandwidth of
the support is high and free and wire reversal is impossible.

Miller code: To reduce the spectrum we can, from simple Manchester, delete every second
transition, whether this is significant or not, we then obtain the so-called Delay Mode or
Miller code. By applying this rule, we see that the 1s have a transition in the middle of the bit
time and the 0s have no transition. But a 0 followed by a 0 has a transition at the end of the bit
time.

Figure 15, Miller code

Summary of the codes: Figure 16 presents a summary of the different codings.

Figure 16, summary of codes

VI- Notions of modulation speed

A transmission line or channel behaves like a low-pass filter, the different components are
attenuated (amplitude distortion) and delayed (phase distortion). One of the most visible
consequences is the spreading of the signal, figure 17. Under limit conditions, this spreading
has the consequence that the end of a transmitted pulse merges with the beginning of the next
one.
We call modulation speed the following relationship:

Rmax= 2 Fmax

Also,

Rmax < 2 BP

Digital Modulation of 8 states Digital Modulation of 4 states

Figure 17, Digital Modulation