Professional Documents
Culture Documents
3
Computer-Based Training in Ericsson's Training Program
Programmable PCM Multiplexer for the 30-Channel Hierarchy
1988
ERICSSON REVIEW
Number 3 1988 Volume 65
S u b s c r i p t i o n Peter Mayr
S u b s c r i p t i o n one year $ 20
Published in Swedish, English, French and Spanish with four issues per year
Contents
82 Ericsson's Strategies and Technologies for the 1990s
91 Fibre Optic Token Ring System ZAT8
100 Slotted Core Optical Fibre Cable
108 New Generation of Optical Fibre Line Systems for 140 Mbit/s
116 Computer-Based Training in Ericsson's Training Program
122 Programmable PCM Multiplexer for the 30-Channel Hierarchy
Cover
Optical fibre cables manufactured by Ericsson
Cables AB.
The picture shows, from the top, an armoured,
concentric cable and two slotted core cables, one
armoured and one unarmoured. In the slotted
core cables the secondary tubes are protected by
ridges along the central strength member
Ericsson's Strategies and Technologies
for the 1990s
Lars Ramqvist
After the divestments during 1987 and
early 1988 Ericsson can now focus on
Table 1
NUMBER OF INSTALLED MAIN LINES (X1000)
the business concept of providing so- 1986 IN THE TEN COUNTRIES WITH THE
phisticated systems, products and ser- LARGEST NUMBER OF INSTALLED LINES
vices for telecom networks and provid- US 10000
ing electronic defence systems. Erics- Japan 3200
son's primary ambition and objective is Soviet Union 2200
to be the leading international supplier Great Britain 1 800
West Germany 1 600
of advanced systems and services for France 1500
telecom networks. The justification for Italy 1300
this claim is Ericsson's extensive inter- Korea 1000
national expertise and broad operating Canada 750
Spain 700
base.
indicates an Ericsson market
This article will deal with
- Ericsson's market position in 1987
- international market and technology
development trends in the field of In a very short time Ericsson has made
LARS RAMQVIST telecommunications great progress in France, Great Britain
Lars Ramqvist was appointed Executive Vice Presi- - Ericsson's systems for telecom net- and the US. These countries represent
dent of Telefonaktiebolaget LM Ericsson in 1987
and is a member of the Corporate Executive Com- works the sixth largest, fourth largest and
mittee of Ericsson with special responsibility for the - cutting-edge technologies of impor- largest telecom markets in the world re-
coordination of technology, production and pur- tance to the development of Erics- spectively, table 1.
chasing. He is the chairman of Ericsson Compo- son's systems.
nents AB, a subsidiary of Ericsson. From October 1,
1988, Lars Ramqvist is also head of the Radio Com- In the UK, Ericsson received its first
munications Business Area and President of Erics- order for local exchanges in 1985 and
son Radio Systems AB Lars Ramqvist joined Erics- has since obtained a good share of the
son in 1980 as the head of the Information Systems
Ericsson's market position in market. Ericsson's factory in the UK has
Division In 1984 he was appointed President of the
subsidiary RIFA and head of the Components Busi- an annual capacity of 700 000 lines.
ness Area.
1987
Lars Ramqvist has studied at the universities of Ericsson's share of the public telecom- The breakthrough in France came in
Uppsala and Stockholm and has a PhD in solid state munications market has grown steadily 1987 when Ericsson, together with Ma-
physics and chemistry.
During the period 1962-1980 Dr Ramqvist held during the last three years. tra, acquired CGCT, which has 16% of
prominent positions in several major Swedish in- the French market. This company starts
dustrial enterprises, working with technical, busi- 1987 was a very successful year for AXE deliveries in 1988 and production is
ness and market development matters Ericsson. 2.5 million AXE lines were in- expected to reach 300,000 lines in 1991.
stalled worldwide and a further 4.8 mil-
lion lines on order, fig.1. Altogether In the US, Ericsson obtained its first test
telecommunication there are now 24 million lines already order in mid-1986. Hitherto, four region-
ISDN
t e l e c o m m u n i c a t i o n networks
installed or on order in 72 different al Bell companies have ordered equip-
cellular radio countries. ment for a total of 178,000 lines.
private t e l e p h o n e exchanges
Fig1
Total number of AXE lines worldwide, Julyl, 1988
(Excluding cellular systems)
Lines installed and on order
Lines installed
83
International trends
Markets
Plain ordinary telephone services,
POTS, will be the driving force in the
further development of telecom net-
works for many years. Today a total of
450 million main lines are connected to
telecom networks around the world,
fig. 4. The annual growth is approx-
i m a t e ^ . 3 % . Thisgrowth, together with
Fig 3 the need to replace old lines, creates a
Number of MD110 lines, installed and on order,
worldwide, December 1987 world market of approximately 40 mil-
lion lines per year.
Fig 4
454 million main lines were in service in 1987
throughout the world (4.3 %)
% = avpraae annual arowth rate
84
In addition to speech, data will be dis- order to realize these complex services;
tributed over different networks using specification, product development and
different techniques. Fig. 5 shows the establishment of networks. Broadband
growth of public data traffic in Western ISDN will provide subscribers with a
Europe. It may seem surprising that larger number of advanced services
fairly conventional technology - leased than those available now, and voice,
lines and public data networks - will data, text and image will admit of being
dominate the data field for the next 5 - 8 switched and distributed like today's
years. ISDN (Integrated Services Digital telephone calls.
Network) will play a relatively minor role
during this period. However, the development in the field of
cellular mobile telephony will become of
The real breakthrough for ISDN is ex- interest long before broadband ISDN.
pected when broadband services are in- During recent years the Nordic coun-
troduced, which will probably occur at tries have gained a world lead in the
the end of this century, fig. 6. We believe expansion of cellular networks with a
that ATM (Asynchronous Transfer telephone density of over 20 per 1000
Mode) will be the technology used. inhabitants. The density in the US and
Much work is needed before then in UK is five and in the rest of Europe less
Fig 5
Public data traffic in Western Europe, distributed
among different networks and technologies
Fig 6
Introduction of broadband ISDN
85
than one. The mobile telephone density phone system. The Pan-European sys-
in Japan is only a fraction of that in tem was standardized in 1987 and will
Scandinavia, fig. 7. provide service from 1991.
The market for mobile systems and tele- Mobile telephony is likely to play a domi-
phones is expected to expand dramat- nant role in digital telecommunications
ically during the 1990s, fig. 8, mainly be- in the early 2000s. Why be tied to cables
cause the leading European telecom ad- when you can move freely and still have
ministrations have decided on a stan- access to all communications services?
dard for a common digital mobile tele- The relevance of this question isempha-
Fig7
Number of subscribers connected to cellular
networks as a percentage of the population in
different countries,
January 1988
Fig.8
Development of the Personal Communications
market
Business Cordless Telephones
Telephones
Systems
86
Telephony
ISDN
Data c o m m u n i c a t i o n
FO-CATV B r o a d b a n d networks
TV d i s t r i b u t i o n f r o m
c o m m u n a l aerials
LAN
Fig 9, right
The broadband network will be built up through
Data t e r m i n a l
the integration of public telephone and data
communication
networks. Optical fibre cable-TV networks a n d
local data networks will be integrated
sized by the fact that the telephone sets, Further development of networks
the terminals, will be fairly cheap and of The broadband networks will grow from
convenient pocket size. present-day telephone and data com-
a munications networks, fig. 9. Existing
Which groups use mobile telephony? television and local data networks will
Ericssons experience indicates that the also be integrated. A future HDTV (High
first customers are executives. They are Definition Television) standard will be a
followed by professionals, such as law- milestone in the growth of broadband
b yers and doctors. Next come employees networks.
of large companies and finally private
persons. Every subscriber in a broadband net-
Fig 10 work will have access to all services, but
Functions in the broadband ISDN network will be In June 1985, users in the Nordic market the hardware and software necessary
distributed among the nodes were distributed among different cate- for the implementation of the services
3 All subscribers can have access to all services gories as follows: will be distributed among several nodes
No node needs to contain all services
Transport 25% in the network, fig. 10. This is possible
D Subscribers' view of the network
Construction industry 16% thanks to the high transmission speeds
Wholesale and retail 15% obtainable with optical fibre circuils;
Manufacturing 13% >2.4Gbit/s.
Local authorities 12%
Fig 11 Finance and insurance 8% Physically the broadband networks will
Public networks of the future will be built up Others 11 % be realized by linking the ordinary tele-
through the interconnection of existing networks, phone network to mobile and data net-
such as public telephone and data networks. This
overlay preserves the value of existing invest-
works, fig. 11. This interconnection al-
ments and speeds up the introduction of the lows the investment in existing net-
service works to be utilized and subscribers will
quickly and easily have access to the
new services.
Fig 13
Architecture of an intelligent network. Feature
nodes containing databases and systems for the
management of services are connected to the
network via circuits that use CCITT system no. 7
for common channel signalling
SSP Service Switching Point
Fig 14
Functional representation of the architecture of
an intelligent network that makes it possible to
introduce new services quickly; utilizes existing
networks and permits the connection of all types
of subscriber equipment
SCP Service Control Point
SSCP Service Switching and Control Point
SMS Service Management System
VAS Value Added Service
ISDN and public data networks, Erics- To summarize - and this is certainly of
Table 5 son will supply PABXs, MD110, packet- great importance for our customers to
CUTTING-EDGE TECHNOLOGIES. ESSENTIAL
switched private data networks and sys- k n o w - the AXE technology will form
TO ERICSSON
tems for cellular mobile telephone and the basis for future Ericsson telecom-
- Microelectronics munications systems.
mobile radio networks, fig. 15. Erics-
- Computers
- Software and Artificial Intelligence son's overall Core Business Concept
- Fibre Optics comprises AXE, MD110and mobile tele-
- Radio phone systems together with cable,
- Networks
components and network engineering
- Standards
resources. Cutting-edge technology
Ericsson will continue to devote consid-
In order to meet the market require- erable resources, both human and eco-
ments forthe 1990s, future Ericsson net- nomic, to development in order to keep
works will most likely be designed as networks and systems at the forefront of
follows. The networks will contain ac- technology. Approximately 10% of the
cess nodes, transfer nodes, feature turnover is allocated to our R&D pro-
nodes and operations support nodes, grams, which are closely linked to our
fig. 16. The access nodes will ensure the market and product development.
flexibility of the networks through their
interfaces for such applications as The R&D programs cover many central
Fig 16 POTS and ISDN, basic systems for technical fields, table 5. The most impor-
Communications network of the 1990s
switching 64kbit/s data (or multiples of tant aspects are communications, VLSI,
AMX Connection via multiplexer
DCC Digital Cross Connect 64kbit/s) and signalling. fibre optics, systems and software.
FMS Facility Management System
90
Gunnar Nilsson
GUNNAR NILSSON
Ericsson Fiber Optics AB has developed equipment that makes it possible to Ericsson Business Communications AB
build fibre optic Token Ring networks. Token Ring is the name of local area
networks characterized by a ring structure and a special method of giving
computers etc. (nodes) access to the network. IEEE (Institute of Electrical and
Electronics Engineers) has prepared a standard for this type of network. IEEE
802.5. The equipment is transparent to electrical signals that conform to this
WAN (Wide Area Network) is used to in-
standard.
terconnect workstations and databases
The author describes how local data networks built up of optical fibre circuits that are widely spread geographically.
obtain such essential characteristics as high data security, reliability and long The equipment can be placed in dif-
distances between nodes. The architecture and performance of the networks are ferent cities and even different c o u n -
also described as well as various applications. tries.
The third network, LAN (Local Area Net- the box. The most common 802.3 net-
Box work), is used to connect workstations work is called Ethernet.
Token Ring
and databases locally, within an office
A Token Ring network connects a number of
nodes into a ring-shaped network. Each node in or a factory, fig. 1. LAN has several ad- IEEE 802.4 also describes a bus network,
the network receives signals, removes jitter, re- vantages: but with a capacity of 4 Mbit/s. The net-
generates the signals and passes them on to the - messages and files can be transferred
next node. The nodes can consist of computers
work is called Token Bus and uses the
such as PCs, workstations etc. The equipment
between workstations Token Passing method for giving the
is adapted to the network by means of a plug-in - databases and files can be utilized nodes access to the network - see the
board; in the case of a PC a Token Ring board. jointly by several users; updating is box.
In a "conventional" network the connections carried out in a single place; updates
between the nodes are metallic, but when ZAT8
is used they can be replaced by optical fibre
become accessible to all users simul- IEEE 802.5 describes a ring network
cables. taneously which also uses the Token Passing ac-
- hardware, such as printers, plotters, cess method. The network is called
Token Passing modems and multiplexers, can also Token Ring. It is supported by IBM,
A Token is a unique bit pattern which is trans- be utilized jointly by several users.
mitted throughout the ring. Each node detects which has also greatly influenced its de-
and passes the Token on. When a node has to This means savings resulting in the sign. It is the objective of IBM that the
transmit a message it waits for the next arrival of network paying for itself. majority of its equipment should be
the Token. The unique bit pattern is changed
and the node transmits the message. No other
compatible with Token Ring networks.
node can then send any messages since the The real breakthrough for LAN has been
Token has been distorted. The addressed node long in coming. In the US its use has These three standards, which describe
reads the transmitted message and passes it on. grown markedly during the last two the lowest level and parts of the second
When the transmitting node receives the end of
its own message it sends out a correct Token
years and it is expected that in Europe level in the OSI (Open Systems Intercon-
over the ring. Other nodes can then transmit the growth will start booming in 1988. nection) model, are linked to higher lev-
messages. The reason for the slow start is the lack els in the OSI model via a common stan-
of some prerequisites for the develop- dard, IEEE 802.2, fig. 2. It is therefore
CSMA/CD
ment of efficient, low-cost LAN. relatively easy to interconnect an Eth-
A node that wants to send a message senses the
bus to check whether another node is transmit- ernet and a Token Ring network. If two
ting. If this is the case it waits until the bus is free One of the problems has been the lack of nodes in a network are to be able to
and then transmits the message. The node con- communicate with each otherthey must
inexpensive and efficient workstations.
tinues to sense the bus during the transmission,
other nodes might have started transmitting si- This problem has been solved by the be compatible also at levels 3 - 7 in the
multaneously. If a collision has occurred, all rapid development of personal compu- OSI model. Acceptance of all these stan-
nodes will stop transmitting. Every node is ters, PC. IBM's personal computers dards has resulted in users now being
equipped with a random generator that con- have come to constitute a de facto stan- assured that there is and will be a large
trols a waiting time. After the shortest waiting
time has elapsed, the node in question again dard. Each office employee can now amount of equipment which can be con-
tries to transmit its message. Other nodes will have a personal workstation. nected to LAN.
then find the bus occupied when their random
waiting times have elapsed and they are again Standardization has also made it eco-
ready for transmission. Another problem has been the difficulty
of achieving, at a low cost, the high bit nomically viable to develop network cir-
rates necessary to obtain the desired cuits with a high degree of integration.
performance for a LAN with many users, With the support of IBM, Texas Instru-
nodes; for example sufficiently short ments has developed five circuits for
waiting times. This problem has been Token Ring networks. The circuits have
solved through the development of made it possible to accommodate on a
semiconductors, which has led to fast normal plug-in PC board all the func-
integrated circuits with low power con- tions necessary to connect a PC as a
sumption. node in a Token Ring network. At the
same time the cost of such a connection
has become reasonable. Another con-
Perhaps the greatest obstacle to rapid sequence of the cooperation is that
growth has been the lack of standards. Texas Instruments' new circuits are
For more than ten years IEEE has work- compatible with IBM's own, and
ed on different standards for LAN, and through agreements with Tl several
during the last few years this work has companies have obtained the right to
resulted in a breakthrough with three buy these circuits. Printed board assem-
different standards - IEEE 802.3, IEEE blies for Token Ring networks can there-
802.4 and IEEE 802.5. fore be supplied by several manufac-
turers. Since all these boards meet the
IEEE 802.3 describes a bus network hav- requirements of IEEE 802.5 it is possible
ing a capacity of 10 Mbit/s. The access to mix boards f r o m different manufac-
method is designated CSMA/CD - see turers in nnp and the same network.
93
Ericsson has developed a system, ZAT8, Manufacturing industry often has prob-
that provides the data c o m m u n i c a t i o n s lems with interference in computer sys-
parts of Token Ring networks based on tems, for example f r o m electrical ma-
PCs The system gives a conventional chines and motors. A fault in the c o m p u -
Token Ring network the f o l l o w i n g im- tersystem can - as in b a n k s a n d finance
portant characteristics: companies - cause huge losses in a
- high reliability short time.
- high security against interception
- possibility of longer distances be-
tween nodes. Functional description of
ZAT8
The design of the c o m m u n i c a t i o n s net-
Choice of design principles
work greatly influences the possibilities
As has been mentioned already, ZAT8 is
of achieving high reliability and security
intended to provide the data c o m m u n i -
in a computer network, but other parts
cations parts in Token Ring networks
of the system must also have suitable
based on PCs. Some of the advantages
characteristics for the overall result to
of Token Ring networks are:
be satisfactory. For example, the termi-
- the methods of providing access to
nals must be non-radiative and the soft-
the networks for the nodes differ con-
ware must not allow unauthorized use.
siderably in networks in accordance
with IEEE 802.3 and IEEE 802.5. The
The characteristics of ZAT8 offer great
method used in IEEE 802.3 is called
advantages to many user groups, for ex-
CSMA/CD (Carrier Sense Multiple Ac-
ample
cess with Collision Detect) and in
- banks and finance companies
IEEE 802.5 Token Passing. The two
- the defence establishment
Fig. 2 methods are described in the box.
The figure shows the relationship between dif-
- manufacturing industry.
The Token Passing method guaran-
ferent IEEE standards and the OSI model
tees a value for the longest waiting
time f r o m the m o m e n t a transmission
requirement arises in a node until the
message can be transmitted. With
CSMA/CD the m a x i m u m waiting time
varies with the load on the network.
The time is often a very important pa-
rameter in real time applications, for
example in m a n u f a c t u r i n g industry -
since each node in a Token Ring net-
work f u n c t i o n s as a repeater it is easy
to introduce fibre optics in the f o r m of
optical transmitters and receivers. It is
much more c o m p l i c a t e d in a bus net-
w o r k of the Ethernet or Token Bus
type, where a certain amount of
94
Fig.5
Several LANs interconnected by means of
bridges
Fig. 3
The Token Ring topology and a practical config-
uration
95
cal structure of networks built up with shaped networks the cable is run from
ZAT8 is ring-star, fig.3 Up to eight node to node However, the star shape
nodes can be connected to a con- has several advantages w h i c h o u t w e i g h
centrator. FMAU (Fibre-optic Multi-sta- the disadvantage of the somewhat more
tion Access Unit) The FMAUs are most expensive cabling. An important advan-
suitable installed in a cross c o n n e c t i o n tage is that the cost of service and main-
r o o m , placed centrally in a b u i l d i n g or tenance is reduced. At the same time the
on each floor, fig. 4, 250 nodes can be availability increases since any fault can
connected to one and the same net- be repaired more quickly.
work, but limiting the network to less
than 100 nodes is preferable Otherwise Another reason for the greater avail-
the network will be considered slow if all ability of star-shaped networks is that
users try to send messages at the same the majority of the cables only connect
time. With a large n u m b e r of users it is the FMAU with one node each. If a fault
better to build several small networks occurs in a cable it will only affect one
and connect them by means of bridges, node. The remainder of the network
fig 5. functions as before, w h i c h w o u l d not be
the case in ring-shaped networks
One disadvantage of star-shaped net-
works is the relatively high cost of the Data security is essential in many ap-
cabling. A cable is run between each plications. An optical fibre network such
Fig. 4 node and the FMAU, whereas in ring- as ZAT8 increases the security level c o n -
An example of a Token Ring network in a building
96
Fig. 6
An FMAU with terminals for both optical fibre and
electrical connectors
Optical fibre communications link nect the node to the network a d.c. volt-
The optical fibre c o m m u n i c a t i o n s link age of approximately 5V, the phantom
between the FMAU and a PC consists of level, is added to the output signal. In the
three units: conventional network the voltage from
- a fibre optic plug-in board for the the Token Ring board controls - via the
FMAU, the FMAU board above-mentioned cable - a relay in
- a fibre optic plug-in board for the PC, MAU that performs the connection. The
the PC board phantom level provides a check that the
- a short cable that connects the PC cable between the PC and the MAU is
board with the Token Ring board in intact.
the PC.
The PC board converts the electrical sig-
The Token Ring board is a plug-in board nals to and from the Token Ring board to
for connecting a PC as a node in a Token optical signals. The electrical signals
Ring network. It does not form part of are transmitted to and from the PC
Fig. 10 board via a short cable connected to the
ZAT8 but can be obtained from several
SMbaud data signal with an added low-frequency
signal, f, = 2000 Hz, f2 = 5000 Hz suppliers. It handles the whole protocol rear of the PC, fig. 11. Optical transmis-
and determines how data is sent to and sion of the phantom level to the FMAU is
read from the network. The interface be- by means of a low-frequency signal
tween the Token Ring board and the net- w h i c h is added to the data signal, fig. 10,
work meets the IEEE 802.5 specification. in the same way as between the ring
In electrical networks the signals are adapters. The signal is shifted between
normally transmitted via a cable consist- 2000 Hz and 5000 Hz. The PC board acts
ing of two screened pairs surrounded by in the f o l l o w i n g way:
another screen. The output and input
signals to the board are balanced. In a
Electrical input signal
" c o n v e n t i o n a l " network the nodes are
from the Token Ring
connected to an MAU (Multi-station Ac-
board Optical output signal
cess Unit) which corresponds to the to FMAU
FMAU in ZAT8. High phantom level 5000 Hz and DATA
Low phantom level 2000 Hz and DATA
When the Token Ring board is to con- No signal 2000 Hz
Fig.1t
The positions of the Token Ring board and the PC
board in the personal computer
99
ANDERS LARSSON The properties of the cables vary some- According to the specification each type
KARIN NYGARD-SKALMAN of cable s h o u l d be available in versions
what d e p e n d i n g on whether the fibres
Telecom Cables Division
Ericsson Cables AB have just primary coating or tight or with 4, 8, 12, 24 and 36 fibres. In the
loose secondary coating. The manufac- purchase negotiations it was stated that
turing equipment and the prioritization the largest size considered was 12-fibre
of manufacturing engineering problems cables. However, the possibility of fu-
are also affected. ture orders including cables with up to
36 fibres greatly influenced the decision
to use a standardized cable body that
The Administration's could hold the larger numbers of fibres.
requirements specification
The new requirements specification is-
sued by the Swedish Telecommunica- Choice of design
Fig. 1, below tions Administration comprises four The guiding principles and previous ex-
The structure of concentric and slotted core types of cable: perience on w h i c h the development has
cables - cable to be run in ducts been based can be summarized as fol-
- armoured cable to be buried lows:
Fig. 2. far below
The core of a slotted core cable. Each of the six - submarine cable - the same core s h o u l d be used for
slots holds a multifibre tube containing up to six - indoor cable. duct, armoured and submarine cable,
fibres
if possible
- materials with well d o c u m e n t e d prop-
erties should be used in order to avoid
compatibility problems
- different designs must be evaluated
by means of full mechanical testing
- know-how acquired in the f o l l o w i n g
areas should be utilized:
volume p r o d u c t i o n
management of excess fibre in the ca-
ble
installation and splicing of optical
fibre cable.
rotate. Both cases require a controlled This requirement must be met during
pay-off unit for each item placed in the manufacture, installation and opera-
slots. Thus, a slotted core cable with 24 tion. In the slotted core cable the need
fibres having primary coating or tight for radial protection is met partly by the
secondary coating needs 24 pay-off secondary coating of the fibres and
units. If multifibre tubes with four fibres partly by the cable structure. All optical
are used instead, the same cable will cables are provided with a metallic or
only need six pay-off units. Such a re- non-metallic strength member in order
d u c t i o n of the number of c o m p o n e n t s in to protect the fibres from tensile forces.
the cable is desirable, not least from the The strength members have high me-
point of view of quality and p r o d u c t i o n chanical strength and low temperature
engineering. coefficient.
set. A multifibre tube is placed in each radial forces and must not affect the op-
slot. This applies to duct, armoured and tical characteristics of the fibre when
submarine cable. In indoor cable two the temperature varies within a given
fibres with tight secondary coating are range Ericsson has chosen loose sec-
placed in each slot, fig. 3. ondary coating in cables w h i c h are ex-
posed to great external stress.
Secondary coating
The secondary coating protects the Tight secondary coating is used for
fibre from both axial and radial forces. fibres in indoor cables. The resultant
The material chosen for the secondary high flexibility and small dimensions are
coating is polyamide, because of its advantageous in cable installation.
high modulus of elasticity.
Moisture barrier
In loose secondary coating, the fibre The secondary tubes are filled with a
Fig. 5
O p t i c a l cable for i n d o o r i n s t a l l a t i o n
can remain free from tension when the thixotropic gel w h i c h serves as mechan-
cable expands or contracts long- ical protection as well as a moisture
Temperature range itudinally. The inner diameter of the tube barrier. The advantages of a t h i x o t r o p i c
operation - 3 0 to +70 C
storage - 4 0 to +70 C is larger than the fibre diameter. In a gel compared with petroleum jelly or
installation - 1 0 to +50 C cable with the secondary tube placed in polyisobutenes for optical cables are:
Smallest bend radius 15 x cable diameter
Diameter 14.5 mm a helix, spiralling round the strength - relatively constant viscosity
Net weight 180 kg km
Maximum tensile member, the fibre is nominally posi- t h r o u g h o u t large temperature ranges
torce 1.0 kN tioned in the centre of the tube when the - no melting point
Design cable is not under stress. When the ca- - the gel can be p u m p e d at room tem-
1 Fibre with primary and tight secondary coating
2 Core, dielectric ble is extended, t h r o u g h pulling or tem- perature
3 PE insulation perature changes, the fibre will move
4 Sheath: PVC, green, nominal diameter 14,5mm - the gel is soft and flexible even at low
Colour coding of the secondary coating: RED BLUE in within the tube towards the centre of the temperatures.
each slot
cable, fig. 4. When the cable contracts,
the fibre will move away from the centre. Petroleum jelly is used in the slots
This movement, w h i c h can be deter- around the secondary tubes, where it
mined theoretically and w h i c h is called provides good protection against pen-
the expansion w i n d o w , is important to etration by moisture. Ericsson cables
the behaviour of the cable during tem- meet the requirements for longitudinal
perature cycling and elongation. The watertightness in a c c o r d a n c e with IEC
size of the w i n d o w is determined by the specification 794-1-F5.
inner diameter of the secondary tube,
the distance to the cable centre, and the Sheath
cable pitch. The size of the expansion A sheath of LLDPE, linear low-density
w i n d o w can easily be verified em- polyethene, is extruded around the slot-
pirically t h r o u g h tensile tests of the ca- ted core. Linear polyethene gives a me-
ble with simultaneous measurements of chanically better sheath than LDPE,
the fibre length. low-density polyethene.
Loose secondary coating gives better The cores of all three o u t d o o r versions
mechanical performance than tight are identical out to and including the
coating. Materials for tight secondary first sheath. A r m o u r e d and submarine
coating must be able to withstand high cables contain a r m o u r i n g wires be-
tween the inner and outer sheaths, and
Fig. 6 the submarine cable is also equipped
Optical duct cable with an inner lead sheath.
Temperature range
operation - 3 0 to +70 C
storage - 4 0 to +70 C
installation - 1 0 to +50 C
Smallest bend radius 15xcable diameter
Diameter 13.5 mm
Net weight
Maximum tensile
160 kg/km Material properties
force 1 OkN
Compatibility
Design It is essential that the materials used in
1 Fibre with primary coating, red-blue-white-green
2 Loose secondary coating, polyamide the cable are compatible. For example,
3 Core, dielectric
4 PE insulation
the gel in the cable must not break d o w n
5 Filler the plastic and cause stress c r a c k i n g
6 Filler
7 chaaih nrtluthpnfi hlack nominal diameter 13.5mm and reduced strength.
Fig. 7, right
Armoured optical cable
Temperature range
operation - 3 0 to +70 C
storage - 4 0 to +70 C
installation - 1 0 to +50 C
Smallest bend radius 20 x cable diameter
Diameter 20.5 mm
Net weight 670 kg/km
Maximum tensile
force 3.5 kN
Design
1 Fibre with primary coating, red-blue-white-green
2 Loose secondary coating, polyamide
3 Core, dielectric
4 PE insulation
5 Filler
6 Filler
7 Inner sheath: polyethene, black
8 Armouring: galvanized steel wire
9 Outer sheath: Polyethene, black, nominal diameter
20.5 mm
Fig. 8, far right Ericsson Cables AB has studied the - Interstitial hydrogen. The effect is
Optical submarine cable
compatibility of polyethene and poly- fully reversible and has a limited
Temperature range
operation - 3 0 to + 70 C amide with t h i x o t r o p i c gel and pe- effect at the wavelengths 850, 1300
storage - 4 0 to + 70 C troleum jelly. There are different ways of and 1550nm
installation - 1 0 to +50 C
Smallest bend radius 20 x cable diameter carrying out such an evaluation. Poly- - Absorption peak caused by OH
Diameter 40 mm
Net weight 4100 kg/km ethene is studied by measuring the ox- groups formed in chemical reactions
Maximum tensile idation induction time, OIT, i.e. the time between the hydrogen and different
force 25 kN
it takes oxygen to break d o w n a material dopants in the glass, e.g. germanium.
Design
1 Fibre with primary coating: red-blue-white-green at a certain temperature. Measurements The increase in attenuation is irrever-
2 Loose secondary coating, polyamide are made before and after ageing in pe- sible and clearly noticeable at 1300
3 Core, dielectric
4 PE insulation troleum jelly and t h i x o t r o p i c gel for 14 and 1550nm but negligible at 850 nm
5 Filler
6 Filler days at a temperature of 60C. HDPE, - UV absorption caused by defects in
7 Inner sheath: polyethene high-density polyethene, and LLDPE the glass that develop in interaction
8 Lead sheath
9 Wrapping: bituminous crepe paper, jute yarn, filler have the best properties as regards sta- between hydrogen and glass. UV ab-
10 Armouring: galvanized steel wire
11 White wash: nominal outer diameter 40mm bility. Gel impairs the tensile strength of sorption occurs at all wavelengths.
LDPE. LLDPE is not affected to the same The increase in attenuation is greatly
extent. dependent on temperature and barely
measurable at room temperature.
The mechanical properties of the mate-
rial used for the secondary tubes, poly- Hydrogen can occur in optical cables as
amide, is only marginally affected by the a result of
gel. - diffusion from the materials used
- corrosion of metal in the presence of
The results of the compatibility test water
show that stability problems may occur - electrolytic phenomena.
if thixotropic gel is allowed in direct con-
tact with polyethene. This is one of the In order to avoid attenuation increases
reasons for choosing loose secondary caused by hydrogen it is necessary to
coating. - minimize the sensitivity of the fibre to
hydrogen
Hydrogen - choose the cable design and mate-
Hydrogen diffuses quickly t h r o u g h all rials so that the amount of hydrogen
known materials. The diffusion constant in the cable is low.
for glass has been calculated to be
1.5 x 10 11 cm 2 /s at 20C. When a fibre is Ericsson Cables AB has studied the gen-
surrounded by a hydrogen atmosphere, eration of hydrogen from all relevant
hydrogen will diffuse into the fibre until materials, at room temperature as well
equilibrium with the partial pressure of as at temperatures up to 150C. The
the environment is established. Hydro- amount of hydrogen generated is al-
gen in a fibre increases its attenuation in ways very low and at room temperature
different ways: it is not measurable; less than 0.5|il/g.
Fig. 9
Apparatus for tensile test in accordance with IEC
794-1-E1
Fig. 10
The results of tensile tests on armoured cables
The change in attenuation as a function of tensile
force for concentric and slotted core cables. The
slotted core cable shows no increase in attenua-
tion for forces up to 10 kN
Fig. 11
Equipment in accordance with IEC 794-1-E1 for
performing crush tests on cables. The cable is
pressed between two circular plates having a
diameter of 100mm. In a variant of the test the
cable is pressed between a circular plate and a
i i ,: n /<omotar t\t OCmm
106
Fig. 12
The change in attenuation for duct cables as a
function of the pressure. Circular plates in accor-
dance with fig. 11 were used for this test
Requirements and test results Figs. 1 2 - 1 3 show the results for duct
In all crush tests the increase in attenua- cable, both slotted core and concentric.
tion must be less than 0.1 dB at the spec- The slotted core cable meets the re-
ified force. This is 3000N for duct cable quirements with a g o o d margin, which
and 6000N for armoured cable when confirms the theoretical assumption
plates with a diameter of 100 mm are that the slotted core design has good
used. These requirements are more resistance to pressure.
stringent than the c o r r e s p o n d i n g ones
in the previous specification. Temperature test
The Telecommunications Administra-
When a 25 mm rod is used, duct cable tion's specification states that outdoor
must be able to withstand 1000N and cables must be able to function
armoured cable 2000N. No such re- t h r o u g h o u t the temperature range - 3 0
quirements were stipulated before 1988. to +60C w i t h o u t the attenuation chang-
Fig. 13
Change in attenuation and deformation of duct
cables as a function of pressure. The test consis-
ted in pressing the cables against a 2 5 m m rod
Deformation
Fig. 14
Typical results of temperature cycling tests on
Ericsson's slotted core cables. In this case the
change in attenuation was measured on a duct
cable
ing more than 0.1 d B / k m . The cables un- Administration had installed approx-
dergo temperature cycling ten times in a imately 6000 km of optical fibre cable.
temperature chamber between - 4 0 and The first one, a m u l t i m o d e fibre cable,
+70C. The fibre attenuation is mea- was laid in S t o c k h o l m in 1979 1 and the
sured d u r i n g the cycling. Each fibre is first single-mode fibre cable in 1984 2 .
measured individually and a mean value
for six fibres is calculated. Fig. 14 shows The first installation of slotted core ca-
a typical result of temperature cycling; ble was a 22 km route in southern Swe-
in this case a duct cable. The cable den, laid in March 1988. The cable was a
meets the A d m i n i s t r a t i o n s require- duct cable from Ericsson Cables AB and
ments with a good margin. installation went according to plan. Dur-
ing 1988 Ericsson will supply approx-
Conclusion imately 1500 km of slotted core cable to
Test results show that the slotted core the Swedish Telecommunications Ad-
cable has very g o o d mechanical proper- ministration. The majority will consist of
ties and meets the requirements of the 12-fibreduct cable, w h i c h is often run in
Swedish Telecommunications A d m i n - plastic tubes. A r m o u r e d cable is nor-
istration very satisfactorily. The slotted mally buried t h r o u g h p l o u g h i n g , fig. 15.
core cable withstands external pressure
better than concentric cables. Other- During the next few years the Swedish
wise the t w o types of cable design are Telecommunications Administration
comparable. will lay approximately 2500 km of cable
per year. All of these cables will be of the
slotted core type and contain between 4
Installation of slotted core and 36 fibres. Approximately 9 0 % of the
cable cable laid in 1988 will be 12-fibre cable.
The optical fibre network in Sweden has 24-fibre cable is mainly used in urban
been built up very quickly. By the end of areas. The Administration plans to start
1987 the Swedish Telecommunications using 36-fibre cable in 1989.
References
1. Gobi, G. and Hogberg, S.: Field Trial
with Optical Communication. Ericsson
Review 57(1980):3, pp. 109-116.
2. Hellstr6m, B., Hogberg, S. and Stens-
land, L.: Field Trial with 140 Mbit/s Sys-
tem over Single-Mode Optical Fibre.
Ericsson Review 62 (1985):1, pp. 2 8 -
40.
New Generation of Optical Fibre Line
Systems for 140 Mbit/s
System structure
The system, fig. 1, consists of
optical fibres A new generation of 140 Mbit/s systems, - line terminals w h i c h encode the sig-
telephone lines ZAM 140-4, has been developed. The nal, attenuate jitter, supervise the line
maintenance e n g i n e e r i n g new system is more compact and re- system and initiate alarms if a fault is
design e n g i n e e r i n g
quires less power than the earlier type detected
and operates at a wavelength of either - locally powered, two-way intermedi-
1300 or 1550nm. The m a x i m u m repea- ate repeaters w h i c h are installed
ter span is 50 and 80 km respectively. In along the cable in o r d e r t o regenerate
addition to a fault location channel and the optical signal.
a service channel, ZAM 140-4 can trans-
mit two data channels with interface The basic version of the line terminal,
RS422. The electrical interface of the fig. 2, contains four units:
Fig. 1
system meets CCITT Recommendation - transmit unit
Two-way optical fibre line system ZAM 140-4 with G.703.9. Ericsson's fault location sys- - receive unit
supplementary equipment for fault location and tem ZAN 201 can be used in c o n j u n c t i o n - alarm unit
service channel with ZAM 140-4. The line system can be - d.c./d.c. converter.
TR Transmit unit
FDU Fault detector unit
REC Receive unit
FLM Fault location magazine
ST Service telephone
PIERLUIGI CIACCI When necessary, the line terminal can The basic version of the intermediate
FATME S.p.A, Italien
EVA SCHERP
be supplemented with a fault detector repeater, fig. 9, contains five units:
LARS-GOTE SVENSON unit. Each unit comprises one printed - two transmit units
Ericsson Telecom AB board assembly - two receive units
- d.c./d.c. converter.
Three different types of transmit unit are The intermediate repeater can also be
available. Two are designed for long- equipped with afault detector unit when
haul transmission at 1300nm and required.
1550nm respectively and the third for
short-haul full dynamic systems. The The new magazine construction prac-
same type of receiver is used in all cases. tice BYB88 was used for both the line
Fig. 2
Block diagram of the line terminal
D4 Digital electrical interlace for 140 Mbit s in accor-
dance with CCITT Recommendation G.703
F4 Optical interface. 178 Mbauds
DCH Two data channels (RS422). accessible at the
front of the receiver board
Fig. 3
Line terminal magazine equipped with two line
110
Fig. 8
CMOS standard cell circuit
The majority of the digital functions have been
assembled in CMOS circuits built up of standard
cells. The circuit includes a ROM in which the
code tables for the line code are stored. A 2 \.im
comirnnriiictor Drocess was used
112
Fig. 9
Block diagram of the intermediate repeater
the F4 timing, is used. It can be utilized The fault detector board links the line
to oversample a low-frequency signal or system to fault location system ZAN 201.
to transmit one or more service chan- The instructions issued by the fault loca-
nels which are synchronous with the F4 tion magazine, FLM, in ZAN 201 are pro-
timing. cessed and acknowledged by FDU. The
processing consists of FDU, when re-
Alarm unit quested, carrying out such operations
The alarm unit will as bit error rate measurement for the
- collect primary alarms from the other receiver and sending the result to FLM.
boards Communication within system ZAN201
- generate secondary alarms is in the form of 300 baud FSK signals. In
- control the AIS generation. order to make it possible to send fault
location, alarm and service telephone
The traffic is supervised by the transmit- signals on a single service channel, the
ter and the receiver. Alarms for loss of three signals are allowed to frequency
input signal to the transmitter and re- modulate a 30 kHz carrier, which is then
ceiver are fed to the alarm unit, which inserted using the service channel
controls the transmission of alarm in- clock. The high sampling rate of
dication signals (AIS) in the relevant di- 1857 kHz ensures low quantizing noise
rection. An alarm is also sent to the al- on the demodulated signal.
arm unit if the bit error rate in the re-
ceiver exceeds 1x10 6 and 1x10 3 re- Intermediate repeater
spectively. In addition, the laser bias The two-way intermediate repeater,
current is supervised and an alarm is fig. 9, regenerates the optical 178
initiated when the laser diode has de- Mbaud signal. Each repeater contains a
graded to such an extent that the bias receive unit and a transmit unit for each
current is 50% above its initial value. In direction. The optical line signal is con-
the case of a power failure an alarm is verted to an electrical signal which is
initiated through a make function to regenerated and decoded. The service
earth. These alarms, the primary alarms, channels are extracted, processed in
are accessible at the front of the alarm the repeater and inserted again whereas
unit and will also result in secondary the traffic-carrying channel isfed outvia
alarms such as system and bay alarms. the D4 interface. The signal is then en-
coded and converted to an optical sig-
Fault detector unit nal. The intermediate repeater also in-
The fault detector unit, FDU, will cludes a d.c./d.c. converter and a fault
- adapt fault location, service tele- detector unit equipped with an interface
phone and alarm signals for transmis- for a service telephone.
sion via the line terminal service chan-
nels over the fibre
Fig. 10 - measure the bit error rate at the line Duplex communication over
Duplex communication over one fibre terminal.
one fibre
ZAM 140-4 can be used for simultaneous
duplex communication over one fibre,
fig. 10. The receiver and transmitter at
each end of the circuit are then con-
nected to the transmission fibre via
single-mode couplers. These couplers
introduce additional attenuation which
reduces the maximum transmission dis-
tance.
- LEDs that indicate alarms from faulty teristics that are important to the user: it
Technical data units and line sections as required by is compact and robust and offers ease of
Electrical interface D4
CCITT handling and installation. This has been
Bit rate 139.264 Mbit's
Code CMI - two splitting coaxial links at the D4 achieved t h r o u g h
Impedance 75 ohms, interface. - using k n o w - h o w from the design and
unbalanced installation of other digital line sys-
Pulse amplitude 0 5V
Return loss
IEC symbols are used t h r o u g h o u t to in- tems such as ZAM 140-2 and ZAM 34-4
7-210MHZ >15dB dicate the various functions. - simulating and optimizing the perfor-
Permissible cable attenuation mance of the system and the f u n c t i o n
at 70MHz <12dB All connections between sub-racks and of its circuits by means of computer
Optical interface F4 between a sub-rack and the exchange programs
Symbol rate 178.2579 Mbauds cable are made at the rear of the sub- - collecting all digital functions in
Code 5B6B, scrambled
racks using three types of plug-in con- ASICs
Modulation format . NRZ
nector: - building up a compact structure with
1300 nm
- an optical connector, fig. 13 all transmission functions assembled
wavelength 1285-1330nm
spectral width (FWHM) <5 nm - a 1.6/5.6 coaxial plug on two circuit boards, one for each
output power ?-4dBm - a fifty-pin wirewrap connector for direction of transmission
1300 nm. full dynamic system low-frequency signals. - placing all sensitive low-level signals
wavelength at 25X 1285-1330nm in a screened PIN-FET hybrid
spectral width (FWHM) <5 nm The intermediate repeater can also be - transmitting fault location and ser-
output power 5>-14dBm
mounted in a slim rack and used in- vice line signals over the actual fibre
1550 nm doors. In o u t d o o r installations it is and using Ericsson's general-pur-
wavelength 1530-1570nm
placed in a suitable container. pose fault location system Z A N 2 0 1 .
spectral width (FWHM) <0.25 nm
spectral width w h i c h permits fault detection d u r i n g
(15 dB off max.) <1 nm operation
sidemode suppression >30 dB Summary - high service channel capacity
output power ?-4dBm
The optical fibre line system ZAM 140-4 - designing printed board assemblies
Input sensitivity <-38 dBm has g o o d technical performance and so that the radiated electromagenetic
Max. input signal >-12 dBm
high reliability. It has several charac- interference is reduced to a m i n i m u m .
Connector losses <1.0 dB
Maximum permissible system
attenuation S-R in accordance
with CCITT at BER = 1x10 9
(guaranteed throughout system
life)
High-performance laser 33 dB
Full dynamic system 23dB
Transmission medium
Optical fibre single-mode
External diameter 125 2 (im
Mode field diameter 5 0.5 nm
Cut-off wavelength 1100-1270 nm
Power supply References
Battery -20 to -60 V
1. Silvergran, U. and Woldegiorgis, K.:
Power consumption Fault Location System ZAN201. Erics-
Line terminal including fault son Review 61 (1984):4 pp 162-169.
detector and alarm unit <17W 2. Kvarnemo, A. and Landberg, C : Opti-
Environment cal Fibre Line System for 8Mbit/s.
Temperature
Ericsson Review 64 (1987):1 pp. 3 0 -
Line terminal 0 to +45C
34.
Intermediate repeater -20 to +55C 3. Stjernlof. B. and Wallers, A.: Optical
Fibre Line System for 140 Mbit/s. Erics-
Humidity son Review 60(1983):3 pp. 144-150.
Line terminal <90% 4. Hellstrom, B., Hogberg, S. and Stens-
Intermediate repeater <95% land, L.: Field Trial with 140 Mbit/s Sys-
Fault location
tem over Single-Mode Optical Fibre.
Fault location system ZAN 201 Ericsson Review 62 (1985):1 pp. 2 8 -
Transmission medium the optical fibre 40.
5. Johansen, B. and Stjernlof, B.:
Service telephone 565 Mbit/s Optical Fibre Line System.
Type four-wire Ericsson Review 62(1985) :3 pp. 130-
Transmission medium the optical fibre 137.
6. Hansson, A-K. and Linden, K.: Optical
Fibre Line System for 4x140 Mbit/s, a
New 565 Mbit/s Application. Ericsson
Review 64 (1987):3 pp, 102-109.
Computer-Based Training in Ericsson's
Training Program
Fig 1
Computer graphics is used in CBT programs. The
student gives an answer in the space indicated. In
this case the program control functions are
implemented by means of function keys F1, F2,
F6 and F9
117
PER-OLOF NERBRANT
Ericsson Telecom AB
Examples Box 3
DIARMUIDO COLMAIN The following case studies describe STUDENT/COMPUTER INTERFACE
Ericsson Expertise Ltd. Ireland some p r o d u c t s for computer-based The interface can be implemented in both hard-
training and the training problems they ware and software.
solve. Box 7 lists the CBT products hith- Input/output devices
erto developed by Ericsson and now A keyboard is used to enter answers, put ques-
ready for use. tions, select menu items
Subscriber Addressing
Processor Structure Hardware Group switch APZ I/O systems Cabinets
switch method
Fig 4
Hierarchic diagram of the CBT course AXE 10
summary"
Product description Design of course material
The simulator is implemented on micro- The design process follows the work
computers. It consists of a student pre- flow shown in fig. 3 and is based on an
sentation part and a development en- analysis of the target group. The docu-
vironment. The latter is designed so that mentation produced initially includes a
no programming expertise is required general course description and an over-
for the development of new exercises. all objective. The subjects to be covered
Essentially, AXE commands are simul- are structured into a hierarchic diagram,
ated in such a way that the responses fig.4.
from the simulator are exactly the same
as those obtained from an AXE system. Objectives are specified for each part of
In addition, the designer can add infor- the diagram, down to lesson level. Tests
mation, advice etc. to aid the student. are then devised for these objectives
The simulator has been designed by and the content of the lessons defined.
ETX.4 Once the content has been established
the subjects that require development of
computer-based material are defined.
What do we mean by
interaction? The computer-based material is pre-
The concept of interaction is fundamen- pared, which results in the following
tal to all media-supported, self-instruct- documents for each lesson, fig. 3:
ing training. Such terms as student in- - objective of the lesson
teraction, interactive video etc. are com- - a list of tests with answers
mon. But what is meant by interaction? - a sequential list of items to be learned
Generally speaking it means that stu- - layout of the display for each item to
dents can adapt training to suit their be learned
needs and capabilities. The student be- - an answer analysis sheet for each
comes an active participant instead of a question; it lists all the answers that
passive observer. Box 2 describes the can be expected, together with the
characteristics of some types of interac- program's response to each answer
tion or dialogues. Interaction requires a - a lesson flow chart showing how the
means for the student to communicate different displays are linked together.
with the training system, i.e. a student/
computer interface. Some methods of The documentation for each lesson is
implementing such interfaces are sum- examined by other course designers
marized in box 3. and subject matter experts, and any nec-
Fig 2 essary modifications are made before
Work environment for the course "Introduction to
programming starts.
AXE"
Development of computer-
based training Programming of CBT material
The following method has been used by When a lesson has been designed and
EXI in a number of projects in which approved, it is programmed; if possible
courses for computer-based training by the person who designed it. Program-
have been developed. The projects have ming can be done in a general high-level
not included any video production, language or by means of an "authoring
which would require a number of addi- language". The latter is explained in
tional stages. Ericsson has now ac- box 4.
quired considerable experience
through using this method in the design When the programming has been com-
of courses. pleted, each lesson is tested by the pro-
Box 6 first large "expert system" was developed at
BRIEF HISTORY1 Stanford University. The system, named MYCIN,
Two concepts in particular have influenced the is a computer program designed to provide at-
development of computer-based teaching: the tending physicians with advice comparable to
teaching machine and programmed instruc- that which they would otherwise get from a con-
tion. sultant specialist in infections.
The teaching machine, developed during the Expert systems can be expected to influence
1920s by the psychologist Sidney Pressey, pre- training in at least two ways. First, the pos-
sented multiple choice questions and kept sibility for a user to study the underlying rules
track of the students' answers. and facts used by an advisory system to reach
its conclusion will in itself result in learning.
By the late 1950's two basic approaches to pro-
grammed instruction could be identified. The Second, the possibility of building intelligent
first, often associated with the educational psy- tutoring systems modelling the student's learn-
chologist B.F. Skinner, organized the material ing characteristics and consequently adapting
in a linear sequence of small segments called his training. Expert systems are briefly de-
frames. Transition from one frame to another scribed in box 5.
was simple and contained minimal additional
information. Because of the simplicity of each
frame the students were expected to be able to
complete the instruction on their own.
Box 7
The second approach is typified by N.E. CBT PRODUCTS MARKETED BY ERICSSON
Crowder and his "scrambled textbook". This Products developed by EXt (in collaboration
approach was based on a branching model that with ETX)
Fig 5
allowed different paths through the material. PCM principles
The structure of a tutorial dialogue. Drill and
The path taken by a student depended on his AXE Telecom Concept
practice is the special case when no new informa-
answers to multiple choice questions. This AXE Introduction
tion is presented and the student uses previously
made it possible to adapt training to the pre- 64 K Group Switching System
acquired knowledge to answer questions
vious knowledge and learning profile of each AXE Addressing Principles
student. The computer became the ideal tool AXE Subscriber Switching Subsystem
for implementation and refinement of both the- AXE Dictionary of Abbreviations
se approaches. ISDN Overview (under development)
Telecommunications Networks Overview
Efforts to computerize instruction began (under development)
around 1960. Basic research and key innova- Modems: Theory and Practice (under
tions by universities paved the way for commer- development)
cial computerized educational systems that
started to appear in the late 1960s. The intro- Products developed by ETX
duction in 1975 of the microcomputer has had SIMAX (a program for simulation of AXE
greater impact on education than any pre- commands)
viously available type of machine. Language practice (the program is under
development but is used internally)6
Box4 Products developed by EBC
The first demonstration of the optical long-play
AUTHORING SYSTEM videodisc was made in 1972. This event was the MD110 Data Switching
Coding the computer control program in accor- start of a very interesting development which MD110 Call Information logging
dance with the design documents is a major has also affected training. The videodisc, see MD110 Traffic Recording
piece of work. Any suitable high level computer figs. 7 and 8, is characterized by short search MD110 Paging Facility
language can be used. This makes for great times, high capacity and high quality. It can be MD110 New Commands in UPA7
flexibility in the programming to tackle a variety used to enrich training with concrete examples. MD110 Multinet
of design problems. One disadvantage is that MD110 CAC 3274 and CAC 3278
programming knowledge is required and that John McCarthy of Stanford University coined X.25 Survey Course
programmers often do not have the pedagogi- the term Artificial Intelligence in 1956 to de- MD110/FS Dealer System
cal knowledge essential for a good result. scribe an emerging field of research. In 1970 the Ericsson Hotel System
Moreover, media-based training offers Research has started into how intel-
several value added benefits whose eco- ligent t u t o r i n g systems (ITS) should be
nomic value is difficult to assess. One designed. This work falls within the
example is the fact that media-based scope of artificial intelligence. Such sys-
training methods can eliminate the wait- tems are characterized by their integrat-
ing time for essential training. ing learning strategies - w h i c h state
how and when information should be
presented to the student - with a model
Future prospects of the student that reflects the student's
Computer-based training has proved to capability. The systems tend to become
be an efficient form of teaching. B o x 6 very large and require powerful c o m p u -
gives a brief history. The next major step ters for their execution.
in the development will be the use of
videodiscs. A pilot project has been
started by EXT.
References
1 Introduction to Computer-Based Edu-
cation Digital Equipment Corporation,
1985.
2 Weller, H. G.: Interactivity in Micro-
computer-Based Instruction: Its Es-
sential Components and How It Can Be
Enhanced. Educational Technology,
February 1988.
3 Kearsley, G.: Costs. Benefits, & Prod-
uctivity In Training Systems. Addison-
Wesley Publishing Company, 1982.
4 Mattsson, H. Ericsson Telecom AB
5 Ringh, B. Ericsson Business Com-
munication
6 Nickson, P. Ericsson Telecom AB
Programmable PCM Multiplexer for the
30-Channel Hierarchy
Fig.1
Point-to-point connection between analog ex-
changes
Digital path
Fig. 2
The multiplexer provides flexibility in the connec-
tion of analog exchanges to the digital network
123
Fig. 4
Multiplexer system structure
The second control unit is used in special ap-
plications, tor example drop/insert
QucYpm b u s
124
The possibility of free allocation of the means that, in the most frequent cases,
time slots, fig. 5, is very important. It speech and signalling are handled by
means that more channels can be con- the same channel unit. The channel unit
nected to the multiplexer than the ca- for 4-wire speech and E&M signalling
pacity of the 2Mbit/s stream would operates in this way, as does the chan-
otherwise permit. The logic connection nel unit for 2-wire speech with d.c. sig-
of the desired channel to a time slot is nalling.
arranged by means of simple com-
mands. Consequently, many changes in The control unit includes a dedicated
the traffic need can be handled without processor for signal conversion. It is ca-
having to change hardware. pable of handling several different sig-
nalling diagrams simultaneously. The
operation of the signalling processor is
Signalling supervised with the aid of well-designed
The analog side of the multiplexer can algorithms. Software adapted to the re-
be connected to several exchanges, quirements of different countries is
each of which may have different sig- available. The selection of signalling di-
nalling requirements. agram in the processor for a certain
channel is made automatically on the
Signal matching has been integrated basis of information from the associated
with other functions in order to make the channel unitorby meansof simple com-
equipment rational and compact. This mands.
Fig. 5
Allocation of time slots
At time t, the channels transmitted in the time
slots are no. 1 from channel unit A,, no.6 from
channel unit A, etc.
At time t2 other channels have been allocated to
the time slots
Time slots
125
Fig. 6
Drop and insert
Drop and insert takes place in multiplexer B.
where 17 time slots are terminated. 14 time slots
are through-connected from A to C and 5 time
slots are inserted in B
a
Fig. 7
Drop and insert is used to increase the security of
routes
a Without drop and insert
h With a direct route and drop and insert
126
Fig, 8
The magazine holds two separate multiplexers
a The two systems are used separately
b The system buses in the two systems have been
connected together. Many channel units can be
connected to the system in this configuration
it, fig. 7. The circuit between A and B can cuits developed by Ericsson has con-
also be used as a transverse route of tributed to the reduction of volume, but
course. the most important factor has been the
integration of the attenuation, filtering
Structure and A/D conversion in a CMOS circuit.
The BYB packaging structure is used for
the multiplexer. A magazine of
244 x244 x220 mm holds two multiplex- Operation and maintenance,
ers and a d.c./d.c. converter. The maga-
zine is a mechanical and functional unit handling, lining up
that can be installed in the various racks The handling facilities in connection
and cabinets of the packaging structure with lining up, operation and mainte-
nance of the new PCM multiplexer have
In the most common applications the been improved considerably. The sys-
magazine holds equipment for two inde- tem includes a large amount of software
pendent systems. There is also space for for handling functions. Some of these
extra channel units so that the free time functions are controlled from a terminal
slot allocation facility can be utilized connected to the multiplexer. This has
efficiently. almost wholly eliminated trimming by
means of manual straps. The functions
The two independent buses in the maga- controlled via the terminal are parame-
zine can be interconnected, fig. 8. In this ter setting and maintenance.
way the system is adapted for applica-
tions in which many channel units need Parameter setting is used during lining
to be connected. up, reconfiguration of equipment and
check of set parameters.
Technology The following functions can be per-
In spite of the greater functional content formed via the terminal during config-
and improved handling, the volume and uration or test of the multiplexer:
power consumption of the new multi- - channel allocation; the time slots can
plexer have been reduced considerably. be freely allocated
A common version is built up of four - channel setting; the level and balance
printed board assemblies, one of which can be set individually for each chan-
is a d.c./d.c. converter. In comparison nel
the previous generation comprised ten - signalling; the appropriate program
boards plus d.c./d.c. and signalling con- can be selected
version equipment. Consistent use of - timing selection; internal or external
surface-mounting and standard cell cir- r o f o r o n r D r\r\ri< ran h o Qplpcted
Fig. 9
The magazine equipped with two separate multi-
plexers. The hand-held terminal is used for pa-
rameter setting and maintenance
Fig. 10
Putting the new multiplexer into operation
mon core. It consists of a standard cir- - 4-wire speech with E&M signalling,
cuit, COMBO, w h i c h carries out level ad- A-channel, 15 channels/board
justment, filtering and e n c o d i n g . It also - 2 or 4-wire speech with E&M signal-
provides the balance in two-wire ap- ling, A + B channel, 10 channels/
plications. A standard cell circuit routes board
the signalling information to the correct - 2-wire speech with loop signalling,
channel. d.c. voltage, 5 channels/board
- 64 kbit/s codirectional or contradirec-
The equipment specific to each type of tional interface in accordance with
channel board consists of transformers, CCITT G.703, 5 channels/board
Technical data interfaces towards the VF side and cir- - 64 kbit/s access to time slot 16, con-
Nominal bit rate 2048kbit/s cuits for signalling conversion. The tradirectional interface.
Performance and complexity and power requirements are
interfaces CCITT G.703, G.711,
dependent on the number of channels Summary
G.712, G.713. G.714,
G.732 accommodated on a board. Different The simplified handling and greater
Number of systems units for subscriber connection are functional content of the new PCM mul-
per magazine 1-2 being developed, resulting in a sub- tiplexer, together with its small volume,
Magazine dimen- scriber multiplexer (RSM). In addition,
sions
low power c o n s u m p t i o n and high re-
HxWxD 244 x244 x220 mm
the following variants are available: liability, make it a very attractive module
- 4-wire speech without signalling, 15 in telecom networks, now as well as in
Comparison of Performance channels/board future.
30-CHANNEL MULTIPLEXERS WITH SIGNAL
CONVERSION
Present New
gener- gener-
ation ation
Functions
internal supervision - yes
CRC-4 yes
function trimming straps program
control-
led
Number of boards when
equipped for
E&M signalling 13 4 References
d.c. signalling 21 9 1. Einarsson, G.: An Experimental 24-
Number of systems per Channel System with Pulse Code Mod-
rack when equipped for ulation. Ericsson Review 43 (1966):2,
E&M signalling 12 32 pp. 42-48.
d.c. signalling 8 16 2. Davidsson S., Ohlsson E. and Johans-
Power consumption from son L.: Subscriber Multiplexers in the
battery when equipped Swedish Network. Ericsson Review 65
for E&M signalling, W 22 11 (1988):2, pp. 51-55.
ERICSSON