Side_1_360

Resource Sharing Networks

When the research began in 1969 “all” comput-

ers were large and expensive. The development

had the main purpose to study and develop re-

source-sharing networks. Resources to be shared

were both the “power” of the computers, pro-

grams and data of various types – information

for shared use. Further, and not least, it was im-

portant to create an environment where human

resources could co-operate and strengthen cre-

ativity and knowledge.

Communicating via Computers

The development of electronic circuit tech-

niques, perhaps the most conspicuous part of

computer hardware development, made so-called

microprocessors possible from about 1970.

Microprocessors are integrated circuits that com-

prise the main part of a computer. Small chips

the size of a fingernail can now comprise entire

computers. That development has continued fur-

ther and further and will probably continue –

nobody knows how far. Ultimate limits of com-

plexity have been repeatedly sought, predicted

and broken.

Today microprocessors with programs are used

as component parts in many devices, more or

less specialized.

Among the logical capabilities thus introduced

into an apparatus is the capability to execute the

rules – protocols – in Internet. That will enable

more and more telecommunications to proceed

as Internet type traffic. Telecom operators of the

world are now eagerly studying the implications.

Developing techniques and new types of telecom

market demands continue to emerge. We will

likely continue to see many great changes of

telecommunications and the use of information

in the years to come.

Advantage of the Internet

Compared to the traditional telecom network

two properties of Internet-technology are fore-

most: Flexible dynamical use of various carrier

media for different demand types and capability

to exploit the versatility of computers.

Flexibility concerns several factors. Primarily

it has the ability to meet differences in volume

of information often expressed as bandwidth of

offered traffic – measured as bits per second.

Large and small bandwidths may be mixed

dynamically, i.e. can be accommodated and opti-

mized while changing rapidly. Secondly there is

a great economic potential in exploiting combi-

nations of requirements such as urgency, free-

dom of error and reliability. Last, but not least,

Internet techniques are good at exploiting differ-

ent carrier media in combination – various cable

types, packet radio and satellite net, and proba-

bly many new concepts that were less practical

without these new techniques.

Some Further Details

National and international telecom networks

may be viewed as nodes interconnected by col-

lections of lines of varying properties. Each line

may have characteristics such as analog or digi-

tal, bandwidth (pulses per second), subject to

noise and other factors that determine the capac-

ity and quality of channels. Nodes typically have

switches capable of automatically setting up and

taking down telephone calls and facilities for

manually setting up and taking down other types

of channels.

All kinds of information may be represented dig-

itally. Natural values such as temperature, dis-

tance, weight, pressure, etc. are measured in spe-

cific units and the numberof units is the digital

representation of the value. Accuracy of such

representation can be as required. It is a matter

of the units and the measurement precision.

Depending on the requirements this may be eas-

ier said than done of course, but when a measure

has been digitized the accuracy of each value is

preserved thereafter, unharmed by noise and loss

as long as the number symbols can be recog-

nized, recovered and regenerated.

Computers and communications represent num-

bers in the binarysystem, by strings of the sym-

bols 0 and 1. These are binary digits – bits. Simi-

larly, people represent numbers in the decimal

system, by decimal digits – the symbols

0,1,2,3,4,5,6,7,8,9. In a transmission channel or

in a storage system symbols are distorted and

mixed with noise. In such deteriorated signals it

is easier to distinguish between two symbols –

than between ten. Therefore, binary representa-

tion of numbers is preferred in computers and

in digital transmission. Digital channels can be

made far more tolerant of noise than analog rep-

resentation. In practical terms symbols can be

restored to perfection, and hence preserve the

digitized information unchanged. Binary repre-

6 5 4 3 2 1 0

107 108 109 110 111 112 113

Measurement

time

Units

Digitizing an analog variable
means representing it by
measured values at fixed
sampling times