A COMPARATIVE ASSESSMENT OF WESTERN AND SOVIET DIGITAL SWITCHING TECHNOLOGY

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP92B00181R000300270030-6
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 94
• Document Creation Date: December 27, 2016
• Document Release Date: April 22, 2013
• Sequence Number: 30
• Publication Date: January 16, 1987
• Content Type: REPORT

Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
A COMPARATIO/E
ASSESSMENT OF
WESTERN AND SOVIET
DIGITAL SWITCHING
TECHNOLOGY (U)
REVISED 16 JANUARY 1987-
NOVEMBER 1985
Draft Report No. M2011
W. Thompson, ill, CSD
Prepared .by:
K. Bower ._.
R. Phnilps
Con#reci DAAH01.85-C-A093
We`s#ern Division
GTE Government Systems Corporation
1OC3::~erguson Drive
PO~ :Box 7188
Mountain View, CA 94039
ILLEGIB
Iltvni occ~~itrn
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
_, ,
?~
.GTE GOVERNMENT SYSTEMS CORPORATION - WESTERN DIVISION
P.O. Box 7188
Mountain View, California 94039
A COMPARATIVE ASSESSMENT OF WESTERN AND
SOVIET DIGITAL. SWITCHING TECHNOLOGY (U)
Draft Report No. M2011
Revised .16 January 1987
K. Bower
R. Phillips
W. Thompson, III, CSD
Prepared for
U.S. Army Missile Space Intelligence Center
Under Contract DAAHO1-85-C-A093
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
(U) TABLE OF CONTENTS
Faction
Page
BASIC C3 NETWORK STRUCTURE 2
2.1 Common Carrier Networks 2
2.1.1 Transmission 3
2.1.2 Switching 4
2.1.3 Control 6
2.2 User Networks 6
2.2.1 Transmission 7
2.2.2 Switching 8
2.2.3 Control g
3 SOVIET NETWORK
4 SWITCHING TECHNOLOGIES
4.1 Evolution of Switching 20
4.2 Present Technologies 23
4.2.1- Circuit Switching 27
4_.2.2 Packet Switching 29
4.2.3 Computer Switching 34
4.2.4 Components & Switching Systems 41
4.3 Future Technologies 48
4.3.1 Circuit Switching 51
4.3.2 Packet Switching 53
4.3.3 Computer Switching 54
4.3.4 Components & Switching Systems 55
5 TECHNOLOGY REQUIREMENTS 60
5.1 Issues
5.2 Soviet Developed
5.3 Soviet Acquired
6 EMP
7 CONCLUSIONS
8 REFERENCES
8.1 List 1
8.2 List 2
APPENDIX A
60
62
63
71
74
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
R 1
(U)? INTRODUCTION
-This report presents a description of current and future
digital switching technologies in the Western World and in
the Soviet Bloc countries, as they apply to command and
control networks. A complete set of references is provided
at the end of the report: Each reference is identified in
the text..
The first part of the report discusses the basic
structure of a command and control system. This part is
concerned with the structure of a common carrier network and
a .user network, and the transmission, switching and control
of each network.
The second part of the report discusses the particular
network that is the subject of this study. This
is followed by a general discussion of switching technology,
details of present and future switching technologies both in
the Western world and in the Soviet Bloc world. Three
specific areas of switching technologies that are discussed
are circuit switching, packet switching, and computer
switching. Lists of switching components are provided.
The final part of this report discusses the requirements
-and assessment of Soviet technology in the command and
control of their networks. This part describes the
requirements,. key technology issues, Soviet derived and
acquired technology for the next 10 years.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
(U) BASIC C3 NETWORK STRUCTURE
A command and control system consists of two basic
networks; namely, end user networks and common carrier
networks. Each end user network has its own related
communications network which in turn may rely on other common
carrier networks. The basic. end user network, whether data
or voice, involves the interconnection of multiple facilities
using either a point-to-point star or mesh configuration.
Rerouting of the information in the end user net-works is done
by switching at .the users facilities using dedicated
alternate routes and/or by switching and patching at the
common carrier facilities. Networks can consist of several
layers with one common carrier network a subset of one or
more other networks. The Soviet network, the subject of this
study, is an end user network which makes use of common
carrier facilities.
2.1 Common Carrier Networks
Common .carrier networks have evolved over the years from
manually routed analog systems to computer routed digital
systems. This evolution is still in process with most if not
all networks using hybrid systems. The driving factors that
cause the old style network to change vary from country to
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 l
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
a
country as well as between service companies. The two -main
technology areas, an all digital transmission system and a
computer based network control, are not directly related, and
can be implemented separately. For common carriers with a
heavy investment in an analog transmission system the drive
.would be to first automate its routing for network management
and reconstitution.. Where toll circuits operate over long
distances, digital transmission can be used to improve the
quality of the communications without having to change the
analog switches.
The change .from an analog transmission plant to a
digital one requires a considerable investment in hardware
[l~ It cannot be done in a casual way, because too many
transitions between analog and digital trunks will produce an
unacceptable noise level on the circuits. The main driving
force for digital communications is the better quality at
lower costs. Digital communications are also more easily
maintained and reconfigured than analog. Digital
communications are more likely to .be used on circuits that
are very long instead of short ones where the A/D noise on
the digital circuit can be more than that of the existing
analog circuits.
whether or not fiber optics is used in a network is
primarily a cost issue. If there is an installed base of
other types of media such as radio or coaxial cable, then the
M20i~
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 1
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
z
e t
tendency is to upgrade those first before installing fibers.
Within the United States there is a major move to install
fibers (except for AT&T which has a major investment in
radio relay systems that they-are upgrading to digital [2].)
The developing countries are using more fiber than radio
because they do not have a basic investment ih any major toll
systems. Radio systems are favored in some countries that
have remote or mountainous regions due to the problems of
maintaining long cable or fiber routes. However, the costs
of installing fiber is dropping and by the early 1990s most
all new toll transmission- systems will be fiber, with radio
and satellite systems confined to specialized markets [3],
[12]
There are presently few, if any, new long distance cable
systems being installed except in the bloc countries. In the
bloc countries new long distance fiber systems will likely
start appearing by 1990 as they develop the capacity to
produce the high data rate electronics required by the fiber
systems [30].
2.1.2 Snitching
The western world had their analog systems automated
before the digital movement began [6]. As an example, ATT
had the number 4A-ESS computer controlled switch, before the
number 4E-ESS was placed into service in 1976. The trend in
the bloc countries, particularly in the Soviet Union, is to
upgrade their existing analog systems to get an automated
ILLEGIB
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
control network in operation before moving to all digital
transmission.
The first part of a network usually to receive digital
matrix switches is the subscriber or.PABX level, while the
r
toll-(tandem) levels usually wait for the transmission to
become digital. Once a major part of the toll transmission
plant .has been converted to digital, it is practical to
convert the toll switches to a digital matrix instead of an
analog type. '
Today there are very few switches built in the western
world that are not digital. Most switch development work
presently is at the class 4 and 5 office (subscriber) level,
but there is very little difference between digital switch
types. The same basic digital switch is being used as a
PABX, central office and tandem switch [4] However, the new
tandem switches do not have the major control systems that
the older switches had. The new switches are moving toward a
separate network control system using standard hosts and
common channel signalling, thus reducing the amount of
control capability each toll switch has to have.
By the late 1990's the digital matrix tandem switch will
be combined with the multiplexers and distributed throughout
the transmission system. The entire network will then be
operated under a unified contrpl system [5].
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
~~
2.1.3 Control
The control of a switch, as well as of a network, was
very limited until computers were added to the systems [ 9 ] ,
[41]. Without the computers, functions were limited to
simple operations with manual- intervention when something
went wrong. Users also were required to have detailed
knowledge concerning the methods of routing their calls.
.Computer control of switches began with the analog matrix
type and-was carried over to the all digital switch [6].
Most of the control functions can be used with either an
analog or digital matrix switch. As discussed above the
trend now is to separate the control function from the
switch. This is being accomplished by establishing separate
signalling channels and control centers. This allows the
path of the control information to be different from that of
the users' communication [7], [39] . These new control
centers are computer serviced by distributed data bases which
permit the network to dynamically route calls based on such
factors as network loads, circuit- outages and customer
requirements.
2.2 User Networks
User networks have evolved from the PABXs which were
.interconnected by lines leased from the common carrier
networks. This technology remained much the same until the
introduction of stored program controlled switching allowed
these PABXs to control how the calls are routed as well as
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
the capability to control other switches in the common
carrier networks [8], [23]. The technology of the 1970s led
to the establishment of packet switched networks for data
[9], [40], and now in the late 1980s these two concepts
(stored program controlled switching and packet switching)
are being merged into the ISDN network [10], [11], [46]. In ,~
addition to the developments in the communications industry,
the computer industry has established its own set of data
routing (switching) systems [15]. These at first were only
for the local area around the main host, but now technology
is providing ways of interconnecting these local networks
into wide area ones through the common carrier networks [9],
[151, [21], [22] 1
2.2.1 Transmission
While the technologies in this area are the same as for
common carrier networks, the employment of the transmission
is somewhat different. The use of fibers is. much more
advanced in these networks because of the ease of
installation and bandwidths they provide. Fibers are being
used for building, campus and public subscriber networks
[14], [18]. Existing western technology is sufficient to
permit the use of fibers .and related digital modems and
multiplexers to be used for present users' requirements. Even
in the bloc countries fibe r is being used in these user
networks, but advanced networks are limited by the lack of
adequate microchips for the modems and multiplexers.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 r
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
2.2.2 Switching
The switching in the user networks is more diverse than
in the common carrier networks. It involves capabilities for
many different types of information. Three basic types of
switches are usually found in user networks. The first is
the standard telephone voice switch, the second is the packet
switching of data [8], [9], (13], [23], [40], and the third
type is the switch used by the computer industry [15], [28].
The technology trend for the voice switch is to change it
into a dynamic bandwidth switch which can handle all types of
information [37], [39]. The packet switches are being pushed
to higher data rates and more throughput. For the present,
the packet switch is used for data rates below 56 Kbits and
the circuit switch is used for data rates above 56 Kbits.
This is rapidly changing with the use of packet switching at
the 1.5 Nib it rate, but- the packet technology probably will
always be used at the lower data rates with respect to the
circuit switched technology [16], [17]. The ISDN efforts are
working to unify these two technologies and many of the end
user switches are a hybrid design using both the packet and
circuit switch technologies [36], [46]. While independent
computer related technologies are' still being developed,
there are -more efforts now to combine them with the
communications industry technologies [15].
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
2.2.3 Control
Control of user networks is rapidly developing as the
user and the communications supplier are integrating their
control systems [7] . Common channel signalling and ISDN are
part of this effort [38], [46]. When digital processors were
~.
added to the PABX the user became able to control the routing
of his own voice (and data) communications [23]. Private
packet switched networks have grown rapidly since the late
1970s which has led to the development of network control
systems for them [24], [39]. The trend for the user networks
is that of the common carrier network except that their
networks are not as large or diverse in requirements.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
(U) NETWORK MODEL
The main focus of this part of the study are the data
and related voice- networks which support the
command and control system. Both these
networks are of the end user type and employ various forms of
communications including common carrier circuits. The data
network appears to use low speed circuits in the 1200 baud
range. The use of these low speed circuits implies that
there is a considerable amount of front-end processing. The
use of data rates in the order of 1200 baud permits the data
to be routed over standard voice channels of the common
carrier network. These channels can be either the digital or
analog type. The basic user network structure is also not
well defined and could be of several types. The most basic
type of network would consist of dedicated circuits
interconnecting the various processing centers. This type of
configuration probably would rely on the computers at the
centers to route the information. The network also could use
some form of packet switching instead of the computers.
either case the common carrier system could easily reroute
the data circuits since they?would appear to the network as
standard telephone circuits using baseband modems. The
related voice communications would use PABX switches at the
user facilities and probably dedicated circuits between them.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
In this arrangement the PABX .switches would be responsible
for the routing of the traffic. They could also work with
-the common carrier network in the routing and reconstitution
functions.
The common carrier network starting
in the late 1970s has been reconfigured into a system that is
similar to the United States toll network. It consists of a
unified network using standard analog voice circuits for
voice and data with rates up to 1200 baud. It is a multi-
layered system of switches and transmission facilities
extending across the entire country. Its structure basically
follows the CCITT guidelines for telecommunications network
architectures. Figure 1 shows the top two layers of this
network that is called the MOC network. There is a third
layer for the metropolitan (city) networks which access the
national network through the AMTS switches of the regional
network. _ A call placed through the three layers of this
system cannot go through more than twelve switches (or eleven
transmission segments) as shown in Figure 2.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
NATIONAL NETWORK
REGIONAL NETWORK
Q UAK i SWITCH ti UAK II SWITCH
Q AMTS SWITCH Q LOCAL SWITCH
Figure 1. (U) National/Regional Network
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
i
a' ~
~ O
U 3
O H
SUBSCRIBER
TERMINAL STATION
JUNTION CENTER
CENTRAL OFFICE
-~ AMTS
UAK II
UAK I
UAK I
AMTS
-~ CENTRAL OFFICE
JUNTION CENTER
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
,}
The has been divided into about 300 area
codes which are represented by three digit codes. These are
used with the subscriber numbers, .which are -all being
converted to seven digit numbers, to select a particular
subscriber. In Figure 1 these 300 areas are represented by
the AMTS switches. Each area will probably have one or more
AMTS switches within its boundaries. An AMTS switch will
serve many local exchanges within the regional and
metropolitan networks. Within the CCITT exchange structure
the AMTS similar to the group (primary center) is shown in
Figure 3. The AMTS switches are connected to one or more
national level (UAK) switches. There will be a primary
route, which is indicated by the solid lines of Figure 1,
and probably a secondary route which is indicated by the
dotted lines. Within, an area there may be special AMTS
switches for supporting the government's private exchanges.
These special AMTS switches will also be connected to a
respective UAK switch.
The national (YeASS) network is divided into 16 regions
which are served by a UAK II switch. These are
interconnected by a central hub which is made up of a number
of (possibly six) fully interconnected (UAK I) switches. The
UAK I switch does not serve a region, but only serves to
interconnect the UAK II switches. .However, the UAK II
switches can be directly interconnected (switches 1 and 2,
Figure 1) where traffic and geography permit. A UAK II
switch also can be connected to an alternate UAK I switch
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
(switch 4, Figure 1)~. A UAK switch will have one primary
route and up to four alternate routes to the selected region.
The national network transmission systems consist of both
radio relay and cable using, in most cases, analog
multiplexed channels. The control appears to be in-band
signalling between the UAK switches. The UAK (YeASS) switch
has been configured for standard voice and data up to 1200
baud.
The above user networks are carried in the national'
network as shown in Figure 4. Since this traffic is separate
from the commercial traffic it may not be switched by the UAK
centers (shown by the heavy vertical lines). It enters the
network in many cases at nodes that are only multiplexing
points. It is not necessary for it to enter through an AMTS
switch if the circuit uses a fixed route. However, if the
user's PABX or data switch has the capability of an AMTS
switch it~could control the. UAK switch directly. It would
make it easier to reroute these circuits if they were able to
be switched by the UAK centers. These centers have been
designed to provide the capability to reconfigure the network
due to traffic overloads and damage. While the system uses
analog transmission, it still can employ the newer common
channel signalling and control systems. This would greatly
enhance their ability to reconfigure the. network. The
present AMTS-4 switch is an analog crossbar switch with
inband signalling. It has a .stored program capability, but
all of its capabilities are not well understood. The newer
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
.switch for this function has gone to common channel
signalling and possibly an improved control system. It is
still an analog switch using reed relays.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
M2011
~ /
~ /
n
/// \`~
INDUSTRY STANDARD
(CCITT) DESIGNATION
SOVIET MOC
DESIGNATION
DISTRICT
(TERTIARY
UAK I
CENTER)
-ZONE
(SECONDARY
UAK II
CENTER)
GROUP
(PRIMARY
AMT S
CENTER)
LOCAL
(TERMINAL
ATS
EXCHANGE)
Figure 3. (U) Soviet MOC Network Hierarchy
17
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
NATIONAL- NETWORK
PRIVATE NETWORK
Q UAK i SWITCH ti UAK II SWITCH ~ NETWORK NODE
Q PABX SWITCH ~p PRIVATE DATA DEVICE
? COMMUNICATION MODE
Figure 4. (U) National/Private Network
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
(U) SWITCHING TECHNOLOGIES
For the purpose of this study switching technologies are
divided into the areas of circuit, packet and computer.
computer switching area is considered separately because
is being approached differently by the computer industry
[15]. The circuit and packet
telecommunication industry.
developments in transmission,
technologies are based on the
switching and control
United States telecommunications.
general
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
4.1 Evolution of Svritchinq
Telephone networks began with switchboards and copper
wire. Each line was a hard copper path for the information
which was changed by patch cords. The first change came with
the introduction of the electromechanical selector switch
controlled by electric pulses from the subscriber telephone.
The routing of the telephone call was controlled by the
numbers dialed by the subscriber. Since there was no
automatic route selection capability in those early days the
interconnection of these exchanges was accomplished by the
dialing of access codes The transmission system was also
limited to a pair of .wires per call. Later advances
introduced audio amplifiers in the circuits between
exchanges, but the subscriber had to be within a limited
distance from the exchange due to the direct current control
of the selector switches.
The analog system lasted well into the 1950's when
digital systems began to make their appearance in the control
of these very complex switching machines. The digitally
controlled systems used cross bar switches. instead of
selector switches and had automatic route selection which did
away with the access code. Each switch could check a number
of different routes for the subscriber. Interoffice lines
-were now using various analog multiplex schemes and had
developed an extensive toll network to interconnect the local
offices.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
The private users' network during this time had
developed in a manner very similar to the public network.
However, very few had dedicated long distance lines because
of their cost. Most of the private systems were confined to
a local area. The data networks by the 1950's were also
getting automatic. switching, but they were still limited to
the standard teleprinter systems. The 1950's also saw a
rapid growth in radio relay systems and the start of coaxial
cables. These items marked the start of good quality and
inexpensive long distance circuits. The information was
still carried in an analog format.
Digital electronics began to appear in the telephone
systems in the 1950's, but was limited in use to its control
of the switches. By the 1960's there were a few low capacity
digital transmission links. Technology in telecommunications
had reached a plateau. Digital circuits using vacuum tubes or
discrete solid state components were not cost effective.
The introduction of the microchip made it possible to do
away. with the old mechanical switch matrix and use an all
digital switching approach.- It also gave the control
processor more. power and capability. However, the digital
transmission was held back because of the low operating speed
of the early microchips. As the speed increased digital
multiplexers were able to operate at higher and higher speeds
and be produced at lower cost.
In. the 1970's digital technology was rapidly emerging
and the next problem was to integrate it into the existing
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
analog networks (49], [50]. This integration process. was not
uniformly achieved because each country and network are
actually-governed by different and various factors which are
not all technical. Such things as present investment,
overall costs, ,type of system philosophy and the politics are
factors that affect the potential conversion to a digital
network (the technical factors are the same across-all
networks).
Technically, the two factors to be considered in analog
versus digital disciplines are the digitizing noise and the
loss of the number of information channels compared to analog
multiplexing. The use of fiber optics and'the improved
modulation schemes such as 64 QAM reduced the number of
information channels lost, but the digitizing noise still has
to be considered on a case-by-case basis.
The digital switches first started to appear in the end
office level because of the increased features they offered
the subscriber. They were also smaller and permitted the
expansion of the capacity of these offices without having to
expand the buildings. The end office level also included the
PABX for private exchanges [8], (23].
Digital multiplexing began in two areas. One area
involved the trunks for the end offices and the other area
involved the very long distance circuits where digital
transmission improved the channel signal-to-noise ratio
because of its ability to be regenerated instead of
amplified.
22
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
. 9
The development of digital transmission proved to be
more important than fiber optics in improving the quality of
a network and reducing its operational costs. The use of
computers in network control also produced major changes in
services and network utilization [9]. Computer control can
be applied to both ana-log and digital networks [7], [39],
[41] .
By the 1990's there will be much more distinction
between the regional networks and the long distance ones
[37]. The regional networks will be moving in a direction to
integrate many new services into the basic telephone
offerings through ISDN [36], [38], [42], [43]. The local
carriers will vigorously pursue .the increased bandwidth
capabilities by handling data rates of 90 Mbits to every
subscriber using fiber optic lines [26]. Switching will
probably be distributed throughout the network and be of both
the packet and circuit switch types. Channels will be
allocated on both a time and bandwidth (or service) basis
(5], [45]. Network control will be handled separately and
will permit the customer many more options. The long
distance networks. are already developing separate network
control systems which allow the customers to configure their
service features. During the 1990's optical switches will
start replacing the digital switch matrixes [19], [27].
4.2 Present- Technologies
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Digital technology has created. new opportunities for
innovation including the integration of digital transmission
and switching, the combination of voice and data services in
one switching entity, and the design of switching systems
which are economical over a broad range of sizes.
In the strict sense, the term "digital switching" refers
to a system which establishes a message channel between two
terminals where information is represented in digital form.
In more common usage a digital switch usually contains a
time-divided network composed of logic gates and digital
memory to accomplish the switching function. Two major
techniques are involved. in implementing time-division
switching. These techniques are time-slot interchanging (T-
stage) and time-shared space division (S-stage) switching.
In constructing networks, the architecture may contain T-
stages, S-stages or a combination of both. Details are
described in McDonald (1983).**
For the Soviet Bloc, the procedure followed to assess
the current status of digital switching technology was first
to identify relevant documents in the Defense Technical
Information Center (DTIC) on-line data base. Secondly,
selected documents were located and/or obtained.- In general,
the identified documents were difficult to obtain; however,
personnel at the Defense Intelligence Analysis Center (DIAC)
provided for onsite viewing of several potentially relevant
documents.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Romanov (1982)** provides a trip report on a visit to
the "International Specialized Fair Systems and Means of
Communications - 'Svyaz'-81." He describes exhibits of
,digital transmission systems that were displayed by the
following countries and respective firms: Hungarian Peoples
Republic (HPR), (Vudavox Firm); German Democratic Republic
(GDR), RFT; Italy, Marconi Italiana; Finland, Telenokia;
France, CIT Alcatel; Federal Republic of Germany (FRG), AEG
Telefunken; YugoslaviA, Elektronska Industrija; Japan,
Nippon Electric Company.
On page 2 of the above reference Romanov states, "The
interest of the specialists was aroused by equipment of the
primary digital transmission systems with the use of single-
channel codes, realized on large-scale integrated circuits
(BIS) which is one of -the latest achievements in the
technology of digital transmission systems." This probably
indicates that the USSR does not have such large-scale
integration (LSI) available. On the other side,. the
existence of such a trade fair could mean that the Soviets
can purchase digital transmission equipment from the vendors.
Romanov (1982)** continues, "In the equipment of the
primary digital transmission system of a number of countries
(HPR, GDR, Finland, etc.) it is planned to use devices which
ensure the joint operation of.digital transmission equipment
with switch stations. The greatest interest in this respect
lies in the matching devices' unit DS30, which was developed
in Finland. As a result of its use in a processor for
25
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
control of the matching devices it ensures the. interaction of
PCM-30 equipment with switching stations of different types,
and the matching devices of the DS30 unit can operate
simultaneously on different algorithms with different
,signalling systems."
Additionally ROmdnC'V (1982)** describes optical digital
line circuit equipment (16, and 896 Mbit/sec.) using GaAs
diode lasers which were displayed by the GDR. An optical
digital line circuit with a transmission rate of 34 Mbit/sec.
was displayed by the Japanese.
The digital communications system DIKOS (FRG) was .of
interest. It ensures the bringing up to the subscriber of a
switching digital ,channel with a carrying capacity of 32, 64,
or 128 kbit/sec. The coding of the telephone signal is done
by delta-modulation with digital companding.
Romanov's final paragraph is particularly revealing as
to the status of digital switching technology in the Soviet.
bloc. Romanov (1982)** states, "In conclusion it should be
noted that the communications administrations of the leading
capitalist countries (USA, Japan, Great Britain, France,
Canada) have made the decision of the primary introduction of
digital transmitting and switching equipment in the
communications networks. The possibility of using single-
channel codes in digital transmission systems considerably
facilitates the supplying of the digital channel to the
subscriber, which in turn makes it possible to organize a
number of new communications services."~ This seems to
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
indicate that there is essentially no such service at the
time of his writing, which was around 1981-1982.
4.2.1 Cizcuit Switching
For circuit switching applications, ICs are replacing
discrete components in many applications. ICs, or chips, can
be placed into three broad categories: Logic chips, memory
devices, and microprocessors. Logic chips contain circuits
that process, convert and direct electronic signals in
predetermined patterns. These circuits are designed to
perform specified tasks and functions, upon request, by an
appropriate electronic signal. Peripheral-logic chips are
also used to manage interconnections between the various
elements of an electronic system. Memory-chips store data
and commands for processing activities. Microprocessors
combine logic and memory functions on a single chip.
Logic chips account for the largest segment of the IC
market. These chips are silicon-based circuits. The
greatest users of logic chip s are .the computer and
communications industries, which together consume about 40~
of the total world output. (Davidson, 1984.)**
Memory chips find wide application in message switching,
where information is transmitted from node to node in a block
message transfer approach.
Memory products accounted for about one-third of the
total IC market in 1982. This market can be further
segmented according to the type of process technology used in
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
creating memory products and the type of memory product. The
three main technologies are also the three principal
processes used to produce-logic chips. The three techniques
are known as bipolar, MOS (metal oxide semiconductor), and
complementary-MOS or CMOS. The bipolar technique is the
original process used-to produce adjacent positive and
negative charged areas in a silicon substrate. Current
flowing from a positive "emitter" to a negative "collector,"
or vice versa, activates semiconductor circuits for logic or
memory purposes. The MOS approach differs in the use of a
"gate" between charged areas. When the current flow through
the gate reaches threshold levels, the circuit switches on.
MOS devices are superior to those produced by bipolar
techniques in terms of density and power requirements -- two
critical performance criteria for semiconductors. MOS
technology dominates the memory area, accounting for about
75~ of all memory devices.. CMOS technology permits the use
of a single charged area as both the emitter of one circuit
and the collector for another, and thereby dramatically
reduces power usage of the chip. CMOS chips are particularly
useful in battery-powered applications.
The four principal types of memory product applications
are: (1) random access memory (RAM), (2) read-only memory
(ROM), (3) programmable ROM or PROM, and (4) erasable PROM
or EPROM.
Microelectronic chips that contain nearly a half-million
circuit elements and which are hundreds of times faster than
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
currently available d'evi'ces are being developed for a wide
range of uses. The first Very High-Speed Integrated Circuits
(VHSIC) are made using photolithography and have device
geometries as small as 1.25 micrometers.
The importance of microprocessors lies in -their major
cost savings for digital switching. Equally important is the
role microprocessors play in making remote, unattended
switching possible. Finally, microprocessors are a major
driver in future distributed switching architectures.
4.2.2 Packet Switching
Packet switching was originally created to serve a class
of users primarily concerned with the transmission of data.
The characteristics of this user group is typified by bursty
transmissions to other data users or host computers with the
possibility of long delays before a response is returned. It
was not cost effective to establish a permanent virtual
circuit for this type of user. Therefore the X.25 protocol
was established for connecting asynchronous terminals to
other terminals or host computers over a network based on
packets. Because each packet contained address and protocol
information as well as the data the possibility of connecting
literally thousands of terminals and computers over a single
network exists. The user then only requires time on the
network for the length of his data request and when the
receiving station returns a response.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
The type of users and data now sent over packet switch
networks is increasing daily due to the higher packet through
put rates, the establishment and enhancement to the various
levels of protocols, and the proliferation of other digital
services that can be integrated into a packet switched
network.
The major impediment to packet switch networks was the
establishment and refinement of protocols at several layers
and the development of vendor hardware. The government
sponsored APRANET and private research have been successful
over the past decade. and a half at developing very
sophisticated. networking protocols and extensive, diverse
hardware. This has led to packet switch networks now serving
a wide array of users for every application imaginable.
Private companies, institutions, government agencies, and
telecommunications companies have established and operate
packet switched networks running over in-house local area
networks, regional teleco lines, and international networks.
The major protocols associated with packet switching
-X.25: network protocol
-X.3: packet assembly/disassembly (PAD)
-X.28: Data Terminal Equipment (DTE) to PAD protocol
-X.29: PAD to host protocol
-x.75: Intranet interconnection protocol for packet
switch networks
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
It is the interworking of the-above protocols that
allows for the wide variety of terminals and host computers
to be networked. Due to the variety of terminals and
computers it also took a very long time to establish and
refine the protocols so that the interconnections could take
place. Once the protocols had been established the hardware
vendors began to produce an extensive selection of equipment
for packet switching. Currently there are several options
for implementing packet switching:
-smart terminal with all protocols resident
-stand alone PAD serving one to several terminals
-PAD integrated into an on premise PBX/CBX
-PAD integrated into and offered by a regional
telephone company switch
-PAD and network offered and run as a value added
service by a third party
By having all protocols resident within a terminal a
user may access any available PSN through a standard modem
connection and with proper authorization. This consumes the
memory and power of a terminal and may not be the most
efficient implementation. This led to stand alone PADs that
interact with several terminals thus becoming the access
point for multiple users into the PSN. Eventually the stand
alone PAD was integrated into the switch, whether it be a PBX
or Class 5 regional switch. This allowed users simply to
call the PAD number on the PBX and gain access to the PSN.
This also allowed. access to multiple networks by having
31
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
numerous PADs on the switch, each serving different PSNs.
Finally, by integrating the X.75 protocol into major gateway
switches a user can reach practically any other user in the
western world who is also connected through a major gateway.
The advancement in protocols and hardware has freed
network designers from -the typical point to point circuits
and even from the number of users on the network. Network
designers can consider any public switch as part of his
network topology. This is primarily due to the high
reliability of packet transfer that has been established due
to network control. and packet accounting procedures that have
been developed. .Lowest cost routing algorithms have been
combined with node switch traffic loading algorithms to allow
for a free flow of packets across the, country by multiple
routes. This means that sequenced packets from one-user may.
not take an identical path enroute to the addressee. This
has eliminated single points of .failure within the system,
which, of itself has major military implications.
Future trends in packet switching will be towards better
integration of the various packet switched networks and in
-the streamlining and refining of protocols to allow for
easier terminal access to the network. Also the capacity and
processing speeds of the hardware used for packet switch
networks is rising rapidly. This will lead to other types of
services being offered or used over the PSNs. These services
will include packetized digital voice and packetized digital
slow scan video. The packetized voice is possible due to the
32
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
.high packet through put rates. This allows packets to be
transferred and reassembled at a rate fast enough to pass
voice traffic.
Soviet and East Bloc countries are currently researching
and implementing PSNs at a very basic level. Current efforts
may be equated to what the ARPANET was 10 or more years ago.
All though the mathematics of network control and
control/interface protocols are open source, it is the
hardware implementation that is and will remain to be a major
stumbling stone. To have a diverse, robust collection of
intraconnected PSNs requires sophisticated micro-processor
controlled hardware for implementation. No East Bloc country
has the plant capacity nor the advanced capability to produce
microprocessors in the quantity or quality necessary for
systems similar to the Western Nations. PSNs that are
established will be for specific users (Ministries, special
defense, KGB), be tightly controlled (security paranoia), and
be constrained in their network topology. An effort on the
level of the US's JTIDS is unforeseeable. Even if the
production and quality of micro-processors was significantly
upgraded the competition between user entities for these
components is so great that the communications faction within
the Eastern Bloc would still not obtain required hardware to
implement any major intraconnected networks.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
4.2.3 Computer Switching Technology
Communications between computers and other machines is
being developed using a set of standards formulated by the
International Standards Organization .(ISO). The standard
network model used is a local area network (LAN), defined as
a communication system to interconnect different computers,
terminals, and office machines in a geographically bounded
area for the purposes of resource and data sharing [15].
Communications hardware is reasonably standard an d?
presents few problems. However, when communication is
desired among heterogeneous (different vendors, different
models of the same vendor) machines, the software development
effort can be a nightmare. Different vendors use different
data formats and data exchange conventions. The task of
communicating in a truly cooperative way between different
applications on different computers is too complex to be
handled as?a unit. The problem, then, must be broken down in
manageable parts.
In 1977, the ISO established a subcommittee to develop a
layered. architecture. Communications functions are
partitioned into a hierarchical- set of layers. Each layer
performs a related subset of functions required to
communicate to another system.
Figure 5 defines in general terms the set of layers and
the services performed by each layer.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
1 Physical Concerned with transmitting unstructured
stream over physical link; involves such
parameters as signal voltage swing and bit
duration; deals with the mechanical,
electrical, and procedural characteristics
to establish, maintain, and deactivate
the physical link.
2 Data Link Converts an unreliable transmission
(U) The Seven Layers of the OSI Protocols
DFgCRTPT?~N/ egRVTr_a!
into a reliable one; sends blocks of data
(frames) with checksum; uses error
detection and frame acknowledgment.
3 Network Transmits data packets through a network;
packets may .be independent (dynamic
routing) or traverse a preestablished
network connection (virtual circuit);
responsible for routing and congestion
control.
Transport Provides reliable, transparent data
transfer between end points; provides end-
to-end error recovery and flow control.
Session
Provides means of establishing, managing,
and terminating connection (session)
,between two processes; may provide
checkpoint and restart service, quarantine
service.
Performs generally useful transformations
on data to provide a standardized
application interface and to provide
common communications services
(examples: encryption, text compression,
and reformatting).
7 Application
Provides services to OSI environment users
(examples: transaction server, file
.transfer protocol, and network
management).
35
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
M2011
A global effort has been underway to establish a digital
network to transport voice and nonvoice information as
needed. The ,architecture of this network closely follows the
OSI reference model and may be considered a prototype for the
evolution of the OSI standards. It is called the Integrated
Services Digital Network (ISDN). It provides end-to-end
digital connectivity to support a wide range of services
through a limited set of standard multipurpose user-network
interfaces.
Figure 6, a representation of an ISDN functional block
diagram, is provided for purposes of terminology and
understanding. Four interfaces (R, S, T, and U) are shown.
The "R" interface is between the non-ISDN terminals, such as
analog telephones or terminals that do not conform to ISDN
standard or terminal adapters. The "S" interface is between
ISDN- compatible terminals or terminal adapters. and the
Network Terminal (NT)-2 processor.. The NT2 will handle layer
1 protocols; i.e., getting the signal on the wire. The NT2
can act as a PABX-and many terminals can be attached to it.
The NT1 will handle the layer 2 and 3 protocols, with just
one connection between it and the NT2. The NT1 will not be
able to act as a PABX; it will go over the U interface to the
central office. -There can be several different network
configurations.
Companies involved in component-building for this type
of network are already selling them and setting up
fabrication factories in other countries. The same is true
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
of-other switch manufacturers. Refer to Section 4.3.4. ISDN
switches used as ordinary central office switches are
currently in place throughout the world, specifically in the
West and Japan. When ISDN truly becomes a reality, these
switches could be easily converted to a complete ISDN system.
The -ISDN cutover can be expected to occur very quickly.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
v
`~
U
ro
U
w
y~.,
~ ~
~ ~
~
~
~
C
H
~.,.{
w d
N O A
+~ ~ ~ U
+'~a ~ ~w
_ N
E '~ .~ ~ E ~
.i~ +1 V N e ~ r_
.~
.-1
~.i
v
b
~ N
~ .-. b
_ '"~~ 'd
q
~, ~ i0
z z ~ ~'
w ro
W
O d y
a
~co-= --
a
",
~
m
a
a
aaav, z
,
c
N H
~
W
~
U
H
p
J
.1
lfl
~
~
Z
~
~
w
~
~
q
~
w
~
a~
~
J..1
v
Fi
.r
~
p
iD
~
C7'
~
H
J.1
}~
u
a
rl
f3
-~
Lca
f.
~
~
w
'
~''
a
8
~'
~
a~
+'
1
ro
~
H
.
.
m
W
~
~
b
~
ri
~
ri
~
d
U
N rl
W W
E E ~~i
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Because the data will be digital, it can be very easily
encrypted. Data encryption standard algorithms are already
developed. Two methods of encryption that are referenced by
the Pacific Bell Company are the Product Cipher encryption
method and the Cipher text encryption method.
Two types of ISDN-switches are being designed: the
stimulus response switch and the functional response switch.
Northern Telecom is developing a stimulus response switch
wherein the network intelligence will be in the switch. The?
switch will have the ability to block calls, transfer calls,
and handle ISDN services. ATT is designing the functional
response switch. This will be a "dumb" switch, with the
intelligent part of the ISDN network to be located in other
ISDN equipment. The rationale for these two designs is
derived from the market sectors these two companies pursue.
Northern Telecom is mainly in the business of selling central
office equipment, while ATT primarily directs its business
activities in telephone sales and other forms of terminal
equipment.
The Japanese are very far ahead of the United States and
Europe in digital communications. They did not wait for the
CCITT to publish its recommendations to start work on their
digital communications system. -They first performed trials
in Singapore in 1985, and are performing trials in the Tokyo
area. Their first trial was to connect a suburb of Tokyo
with the downtown area. The communications technology worked
very well, and the problems they did have related more to
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
providing the proper services to the right people. Some
individuals received services they did not want, and others
were not given services they did need. The national plan is
to make digital communications available throughout Japan by
the year 2000, through the Information Network System (INS).
The published difference between INS and ISDN is that
ISDN focuses on the network structure while INS places more
emphasis on usefulness to the user. General requirements for
the- INS user-network ,interface are the same as those
identified for the ISDN: support for a wide variety of voice
and nonvoice services, support of point-to-point, multi-drop
and other multiple terminal arrangements, compatibility
between calling and called terminals, and easy plug-in, plug-
out operation. As currently configured, the INS model system
.serves approximately 400 digital telephone sets, 1,100
nonvoice digital terminals, and 9,000 analog terminals.
The NADIR Project is a French pilot-project jointly
sponsored by the Ministry of PTT and the Ministry of
Industry. It aims at fostering the use of satellite
communications systems in the data transmission area. In
particular, this project uses the TELECOMI French satellite.
This satellite offers three types of.services: a military
service, a telephone and television service, and an
intracompany communication system service .? The NADIR Project
has several applications: a photograph broadcast application,
a distributed Videotex mailing application, a distributed
stock management application, and a -file transfer link
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
application. The photograph broadcast application is
operational and, the other applications are still under
development. The project ended in December, 1985, but was
continued under support of the Bureau d'Orientation de la
Normalization en Informatiqque (BNI).
In checking with various sources - GTE's library, Mr.
John Fields, a paper by Mr. Robert Golightly, and
conversations with Dr. Horst-Edgar Martin of Siemens - the
Eastern Bloc has done very little with respect to this type
of networking. These countries are primarily still concerned
with analog telecommunications. While the Hungarians and
Romanians have done some work, it is very little compared to
the rest of the world. The Soviets say they will conform to
CCITT recommendations for the first three layers when they do
get around to ISDN. They will use their own design for the
other four layers.[53,54]
4.2.4 Components and Sr+itchinq Systems
Table 2 is a list of the digital switching systems
manufactured by the nations of the west. In Table 2, the
system code, the country of manufacture, the type and source
of additional information (in the form of a paper number or
appendix identification) are given. The additional
information refers to papers and/or appendices in Amos
(1976) , (1981) . **
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Table 3 is a partial list of 4.Sth and 5th generation
PBX's (private branch exchanges) taken from Joel (1985A, p.
494) .
Table 4 is a list of three western military DSS.
-Table 5 is a listing of the top suppliers of western
central office switching equipment for the U. S. marketplace
in 1984.
Table 6 is a listing of the top suppliers of toll
switching equipment for the U. S. marketplace in 1984.
Table 7 is a listing of the world's leading
semiconductor companies as of 1982.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
TABLE 2 (U) LIST
SYSTEM
CODE
AFDT 1
AXE 10
D 1210
DCO
DMS 10
DMS 100
DMS 200
DMS 250
DMS 300
DS 1
DSC
DSS/1210
DTN 1
DTS
DTS 1
DTS 2
DTS 500
DX 100
DX 200
ENSAD
EWS-D
E10
E10
E10
E12
FETEX 150
FOCUS 5
GTD 5 EAX
HDX 10
IFS
ITS 4/IMAZ
ITS 4/5
ITS 5A
I2000
LCS 4/5
MSU
MT 20/25/35
NEAR 61
NO. 3 EAX
N0. 9 ESS
N0. 5 ES5
PROTEO
PRX-D
SPC 2
SXS
SX-2000
SYSTEM X
SYSTEM 1 (1210)
SYSTEM 12(1240)
TDDSS 1/2
TN 5
TP4/III
TROPICO
TSS 5
UT 10/3
UXD 5
1220/PCM-B
+ Electronic
* Electronic
COUNTRY
Italy
Sweden
U. S. A.
U. S. A.
Canada/U.
Canada/U.
Canada/U.
U. S. A.
Canada
Japan
U. S.
U. S.
Italy
U. S.
Japan
Japan
OF DIGITAL SWITCHING SYSTEMS +
S. A.
S. A.
S. A.
A.
A.
(Sudan)
A.
Netherlands
Finland
Finland
E. Germany
W. Germany
France
France
France
France
Japan
U. S. A.
U. S. A.
Japan
Switzerland
U. S. A.
U. S. A.
U. S. A.
Yugoslavia
U. S. A.
U. S. A.
France
Japan/U.
U. S. A.
U. S. A.
U. S. A.
Italy
TYPE
Local/Toll
Local/Toll
Local
Local/Toll
Local
Local/Toll
Toll
Tandem
Toll
Tandem
Local.
Local/Toll/Operator
Tandem
Tandem
Toll
Local
Tandem
Local/Tandem
Local
Local
Local/Toll
Local/Tandem
Local
Local
Toll
Local/Toll
Local
Local/Toll
Local
Local
Toll
.Local/Toll
Local
Local
Local
Local
Local/Toll
Local/Toll/Operator
Toll
Toll
Local
Local/Toll
Local/Toll
Local
Local
Local
Local/Toll
Local/Toll/Operator
Local
PAPER OR
APPENDIX +*
20
29/25*, B
B
7,A
8,A
13,A
12,A
B
B
B
B
9,A
B
B
28
B
B
B
B
23
22
20*/22*
16
B
17
27
B
14-
26
B
S,A
10,A
B
B
B
B
18
11,A
6
4/6*,A
15
19
21
B
B
B
25
9
29
Tandem B
Tandem B
Packet N/A
Local B
Local B
Local B
Local B
Rep. of China
Italy
U. S. A.
Brazil
U. S. A.
Italy
U. K.
Netherlands
India
France
Canada
U. IC.
U. S. A.
Belgium/U. S.
/W. Germany
Switching: Digital Central Office Systems of the
Switching: Central Office Systems of the World --
B
World -- IEEE Press 1961
IEEE Press 1976.
UNCLASSIFIED
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
TABLE 3. 4.5T8 AND 5TH GENERATION PBX'S
After Joel (1985A, p. 494)
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
TABLE 5 (U) TOP SUPPLIERS OF CENTRAL OFFICE SWITCHING
EQUIPMENT FOR THE UNITED
{From IEEE Spectrum,
COMPANY
AT&T Technologies, Inc.
Northern Telecom.
GTE Corporation
Plessey Telecommunications &
Office Systems Ltd.
ITT Corporation
CIT-Alcatel, Inc.
NEC Corporation
STATES MARKETPLACE IN 1984
Nov. 1985, Centerfold)
HEADQUARTERS
New York, New York, U.S.A.
Mississauga, Canada
Stamford, Connecticut,U.S.A.
Liverpool, England
New York, New York, U.S.A.
Reston, Virginia, U.S.A.
Kawasaki, Japan
* The first four companies contain the major market share.
UNCLASSIFIED
TABLE 6 (U) TOP SUPPLIERS OF TOLL SWITCHING
EQUIPMENT
FOR THE UNITED STATES MARKETPLACE IN 1984
COMPANY
Spectrum, Nov. 1985, .Centerfold)
AT&T Technologies, Inc.
Northern Telecom
GTE Corporation
DSC Corporation
ITT Corporation
HEADQUARTERS
New York, New York, U.S.A.
Mississauga, Ontario, Canada
Stamford, Connecticut, U.S.A.
Richardson, Texas, U S.A.
New-York, New York, U.S.A.
* The first two companies contain the major market share.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
TABLE 7.
LEADING WORLD SEMICONDUCTOR COMPANIES' SALES
(DISCRETE AND INTEGRATED DEVICES,
1982)
COMPANY
TOTAL
(S
MhLLIONI
Motorola ~
1310
Texas Instruments
-
1227
Nippon Electric
1220
Hitachi
1000
Toshiba
810
National Semiconductor
690
Intel
610
Philips
558
Fujitsu
475
Siemens
420
Matsushita
340
Signetics (Philips).
384
Mitsubishi
380
Mostek
335
Advanced Micro Devices
(Siemens)
282
Sanyo
260
AEG
196
Thomson-CSF
190
Sharp
155
SGS-AYES
150
Oki
125
SOURCE: Davidson (1984),, p. 103
While .the relative order of ranking may have changed,
-the list in Table 7 is useful for identification purposes.
It is important to note, however, that two of the world's
largest producers of semiconductors do not appear in the 1982
table. IBM, the world's largest semiconductor producer,
manufactured solely for internal usage. Western Electric,
with 1982 production valued at almost $400 million, also
produced solely for internal use at that time. General
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Electric, GTE Micro Circuits, General Motors, among others,
have significant production for internal use.
A review of Nhat is known about systems and
components in the Soviet Bloc countries follows.
Telefax and smart telephones have appeared in Hungary,
but service remains limited according to a news item in
Telephony, 23 September 1985, page 85.
.Pervyshin {1984) ** produced a book on the Soviet
information transmission industry.
.Kleinau, et al. (1981) **discussed digital telephone
networks. and their problems, and Prager, et al. (1981)**
authored a book on digital computer technology in
communication.
.A compendium of standard terms and definitions in
regard to Soviet integrated digital communication networks
was published in 1978 (Anon. 1978.)**
As far as the Soviet semiconductor industry is
concerned, W.H. Davidson, -who has recently .completed a
definitive study of the U.S. technorivalry with Japan,
states: "The United States is engaged in a two-front war
against specialized rivals. It competes with the Soviet
Union in the military arena and with Japan and other nations
in the industrial arena. The Soviet Union neglects- its
industrial activity to focus on military endeavors
(emphasis added) . [Davidson (1984) **, page vii .
The Soviet semiconductor industry is no exception.
According to Meldern et al. (1984)**, there are numerous
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
examples of Soviet block nations purchasing semiconductor and
IC producing equipment (for example, the 1975 bankruptcy
event of Hugle International (Melvern, et al., page 246))**.
4.3 Future Technologies
It is fortuitous, for the purposes of this report, that
the Institute of Electrical and Electronics Engineers (IEEE)
issued a special issue of their IEEE Spectrum in November
1985 on the future of telecommunication (edited by Edward A.
Torerro, 1985). In the issue, Kaplan, senior technical
editor, discusses the present and future of telecommunication
services in Great Britain, Japan, West Germany, France,
Italy, and China. These discussions include the potential
uses of digital switching systems.
What are some of the options available to would-be
designers of next-generation communications systems? Kulzer
and Montgomery** (1984) discuss several possibilities,
placing them in the spectrum shown in Figure 7. Techniques
toward the right end of the spectrum provide increasing
flexibility to handle variable rate information, but require
more processing. The region near the center of the spectrum
is labelled statistical switching and contains technologies
that can transport bursty information streams without the
full functionality of conventional packet switching.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
STATISTICAL SWITCHING
CIRCUIT
SWITCHING
MULTI-RATE
CIRCUIT
SWITCHING
FAST CIRCUIT FAST PACKET PACKET -
SWITCHING SWITCHING SWITCHING
FIXED BANDWIDTH ~ ~ VARIABLE BANDWIDTH
? MULTI-RATE CIRCUIT SWITCHING
- BANDWIDTH IN MULTIPLES OF BASIC RATE;
- BEST FOR SMALL BANDWIDTH VARIATIONS;
- NOT GOOD FOR BURSTY CHANNELS.
? FAST CIRCUIT SWITCHING
- HYBRID (PACKET/CIRCUIT);
- HANDLES BURSTY CHANNELS WELL;
- REQUIRES SUPER-RATING FOR HIGH SPEED CHANNELS.
? FAST PACKET SWITCHING
- STREAMLINED FORM OF PACKET SWITCHING;
- SIMPLE PROTOCOLS AND SWITCHING SYSTEMS;
- HANDLES VARYING BANDWIDTH ACROSS LARGE RANGE.
Figure 7. (U) Switching Technology Spectrum
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
The following paragraphs provide quotes from two Bell
Laboratories experts in the field of digital switching about
their views of the projections of digital switching
technology in the West through 1995. The title and/or
location of the persons quoted are at the time the statement
was made.
a. Matthew F. Slana, of Bell Laboratories, Naperville,
IL, stated:
"Digital switching makes use of the most advanced
semiconductor components. Hence, new technological
developments play a significant role in the development of
the three major elements required for a digital switch -
control, memory, and network fabric. Semiconductor
technology is undergoing rapid changes with substantial
effect on the design of these elements. For example, the
control portion of the system is proceeding more and more
towards a distributed architecture. Microprocessors are
providing more flexible and intelligent peripheral
controllers. Some form of central control will probably be
retained but it will become more like a system manager. More
of the diagnostic and maintenance functions may also be
distributed on a per frame basis. Some overall maintenance
and control functions will still be done by a central
processor.
"Memories are becoming less expensive and faster. One
problem which exists for time-division networks lies in the
configuration of semiconductor memory components. The trends
in chip architectures are toward larger word structures, such
as 256K~words by 1 bit, which are not useful for 128 or 256
time-slot systems. These systems need 256 x 8 or 256 x 10
memory structures. Technology trends toward faster memories
will lead to more time slots per channel.
"Similar trends help the network fabric. Faster logic
gates will allow more time slots in the .system. This will
lead to smaller systems and lower propagation delays, which
again leads to faster systems. Systems will be fabricated in
smaller and smaller spaces which will continue to handle the
same number of lines or trunks. The advent of high-speed,
economical fiber optic links will alleviate the problems
associated with communication between frames, and may also
50
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
lead to new architectures, such as distributed or remote
network components." [McDonald (1983) p. 173) **
b. John S. Mayo, executive vice president, Network
Systems, Bell Laboratories, Murray Hill, NJ, stated:
"The .ability of communications networks to deal
with data is well advanced in the telephone network. The
thousands of existingspace division electronic switching
systems in the telephone network can be equipped to switch a
56 kbit/sec. signal on each voice circuit. Time division
.switching systems are also being intrgduced, and they readily
switch 56 kbit/sec. on each voice channel." Mayo (1982).**
The twelve papers ,in the Special Issue on Serving the
Business Customer Using Advances in Switching Technology,
published in July 1985 (Joel, 1985A)**, show that
telecommunication switching. technology and the system
architecture it supports now encompasses and integrates
switching and data processing. The result has been that
these twelve papers not only view some of the current
systems, but also provide a glimpse int o the future,
including the extension of telecommunications into office
automation.
4.3.1 Circuit Switching - Future Technology
To those countries -with a background in telephony,
Multi-Rate Circuit Switching (MRCS) seems a natural choice
for new communications systems. MRCS provides connections
having bandwidths equal to an integer multiple of some basic
rate, such a 8 or 64 kbs. The user specifies a transmission
speed when the call is set-up and the network provides enough
channels to satisfy the request. As Kulser and Montgomery
point out, there are several problems with MRCS. First is
51
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
the choice of the basic rate. Many services require a low
rate such as 1 kbs, but this can imply a long delay due to
the large frame size required for time-divisian multiplexing.
It also implies a large overhead for establishing high speed
connections, since these must be implemented using multiple
channels, each of which must be set-up individually. Even
with a fairly large basic rate of 1 Mbs, a video connection
might require the establishment of 100 channels. MRCS is
also ill-suited to applications with bursty transmission
characteristics. Applications such as remote file access
require occasional transfer of bursts of data at high rates
such as 10 Mbs. Dedicating a high speed connection to such
applications is costly and inefficient. Using a lower speed
channel yields efficiency, but only by sacrificing
performance.
If we take the next step to the right on the spectrum in
Figure 1, we come to Fast Circuit Switching (FCS) . As with
MRCS, customers request connections having bandwidth equal to
some integer multiple of a basic rate . In FCS however, the
system does not allocate the required channels until the user
has some information to send. Thus FCS, allocates bandwidth
dynamically among a group of users, allowing efficient
sharing of the transmission facilities. Of course, there may
be occasional peak traffic periods when the network cannot
satisfy all user's requests. When this happens, one or more
requests are denied. This kind of switching has been termed
S~
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
"burst switching" by Amstutz (1983) and Haselton (1983),
(1984) .
4.3.2 Packet Sv+itchiaq - Future Technology
Fast Packet Switching (FPS) is the next option on the
spectrum. As in conventional packet switching, FPS uses the
transmission facility as a "digital pip," which carries short
packets of information one after another. Information in the
header of each packet identifies which of many logical
connections the packet belongs to. With this multiplexing
scheme, .connections of arbitrary bandwidth are accommodated
in a simple and natural way. A key aspect of FPS is the
recognition that the high speed and low error rate of modern
digital transmission facilities allow simplification of the
communication protocols used in conventional packet switches.
These simplifications make possible the- construction of
hardware protocol processors. High speed transmission
facilities. also dramatically reduce the queueing delays
.inherent in packet switching. Another key element is the
observation that high speed computer interconnection networks
originally designed for large parallel computer systems, as
in Feng (1981)**, are ideally suited to large high
performance packet switching systems. FPS has been developed
by a group at Bell Laboratories and is described in Hoberect
(1983), Jeng ..(1983), Kulzer and Montgomery (1984) and
Montgomery (1983)..
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
4.3.3 Computer Switching - Suture Technology
Joel (1985A, p. 495)** concludes that the state of
switching technology to serve the business customer is
advancing rapidly. It is in this area of application where
there is the greatest current need for combined voice and
data service to the same stations or terminals. This
technology is in all likelihood a precursor of the future
central-office switches for the so-called -ISDN-Integrated
Service Digital Network.
M. Schwartz, ? director of the Center for Communications
Research at Columbia University in-New York City, was quoted
on the state of ISDN. "It's the future of
telecommunications, once in place, ISDN will offer
ubiquitous, coexisting voice, data, and video networks."
(IEEE Spectrum, Nov. 1985, p. 83, Fischetti (1985)).
Siemens A G Telefunken estimated that, by the year 1990,
there would be 600 million subscribers to analog circuits
worldwide and 125 million subscribers to ISDN. By the year
2000, the company envisions a reversal with only 250 million
subscribers to analog and 750 million to ISDN. (IEEE
Spectrum, Nov. 1985, p. 83).
Roy Weber, AT&T Bell Laboratories is quoted as follows:
"ISDN today (Oct. 1985) is where we were 10 years ago with
common channel signalling." "Now we can't live without it.
. and flexible signalling is the heart of ISDN. .Every
(AT&T) No. 4 ESS will have the primary rate interface by 1987
54
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
- not a trial but a service." [Telephone/Oct. 28, 1985, p.
83.]**
Moving to ISDN is not a purely technical issue and will
take the joint efforts of regional operating companies,
interexchange carriers, state and federal regulators and
equipment manufacturers. In addition, ISDN requires a
receptive marketplace ready to buy.
"The major trend appears likely to be the continuing
convergence of computers and communications." Bell (1985),
IEEE Spectrum, Nov. 1885, p. 111.)**
"The ultimate blueprint for world telecommunications by
the turn of the century is a new socket in the wall of every
home and public place, as standard as the sockets for
electrical current. Into this new socket a person could plug
a telephone, television. set, computer, or other terminal from
any manufacturer to instantly send or receive voice
communications, video images, or high-speed data." (Bell,
1985, p. 111.)** Such universal communications over an ISDN
depend on three key prerequisites: large bandwidth, digital
processing,, and protocol standards.
4.3.4 Components and Switching Systems - Future
Technology
Recognizing, the ISDN as the wave of the future, this
section gives a list of the major ISDN siaitch manufacturers
and some additional information about their switches. In
general, these are the companies involved with completing the
last .four layers of the OSI protocols.
55
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
ITT sells System 12. Although they have dropped-out of
the U.S. marketplace, they are still building and selling
these switches abroad. -ITT has performed many switch trials
throughout Europe. Their -most responsive customer is the
Deutsche Bundespost in West Germany. According to ITT, ISDN
equipment will be regularly introduced into the Bundespost's
network by 1988, and within a few following years, ISDN will
be available throughout Germany.
GTE sells the GTD-5EAX. GTE, together with Mountain
Bell, is involved in trials in Phoenix, Arizona. GTE has a
:division preparing for the International Switching Symposium
in March 1987.
Northern Telecom has two switches involved in ISDN
trials: the DMS-100 and the DS-10. .They are working with
several U.S. telephone companies.
Ericsson sells the AXE-10. They are active in Europe
and have spent $70 million to break into the United States
market. Ericsson ran trials in Sweden and Italy starting in
1984. They recently won a contract to supply Mexico with
central office switches. Ericsson is also working to connect
all of Sweden with digital communications by 1995.
ATT is selling the ~~, and is involved in a trial with
McDonald's and Illinois Bell. The company is constructing a
.functional response switch, but the software is taking longer
to prepare than first estimated.
Siemens offers two switches, the DES and the EWSD. The
EWSD is their top-of-the-line ISDN switch. Siemens is
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
currently involved in trials with several U.S. telephone
companies and is also entering the United States market. GTE
is forming a business venture with Siemens.
,NTT, with its D-5 switch, is conducting trials in
Japan. The .company plans to make digital communications
available throughout Japan by 1995.
Fujitsu has the FETE switch. They are very advanced
.insofar as ISDN technology is concerned, having conducted a
trial in Singapore and Tokyo in 1984. They are supplying the
ISDN terminals for the McDonald's trial in Illinois, and
working with ,GTE to develop and market PBX's. As the only
company offering ISDN features as part of their
communications system, they won the contract to upgrade
communications for Texas.
The following list is provided to show which companies
have been in the news for selling digital terminals or .ISDN
equipment. This list is a starting point for locating ISDN
terminal equipment vendors.
a. NEC -- Terminals for Pacific Bell's trial
b. IBM -- ISDN compatible terminals
c. AMBI -- Digital Communications Device
d. DAVOX --IBM 3270 compatible Integrated Voice and Data
Terminal (IVDT)
e. MITEL of Kanata, Ontario, Canada
f. COMPAQ Telecommunications of Dallas, Texas
g. TELEVIDEO Systems, INC.
h. SIEMENS AG
57
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
,.
i. NORTHERN TELECOM
j. GTE TELENET -- Digital Telephone for digital PABX
system
In regards to chip components, the principal ISDN chip
manufacturers are ATT, Advanced Micro Devices, Siemens, and
Northern-Motorola.
ATT has been selling ISDN chip sets for some time. The
company is currently making chips called jJ~ITE, a transceiver
conforming to CCITT I.430 specifications. The transceiver
will support the basic data rate interface standard for
transmitting voice, data, and video over two pairs of
telephone wire . Commercial production of this chip will be
available in the first quarter of 1987. ATT is also working
on a U interface hiD,
Advanced Micro Devices will put its four primary ISDN
chips into production by the first part of 1987. These are
the 793X family - the 79C30 Digital Subscriber Controller
(DSC), the 79C31 Digital Exchange Controller (DEC), the 7935
Subscriber Power Controller (SPC), and the 7938 Quad Exchange
Power Controller (QEPC).
Siemens is the foremost ISDN developer. The company is
trying to design the fewest number of standard chips possible
to be used to build different types of ISDN telecom
equipment. The company has designed what it believes is the
optimum set of generic subfunctions: the Siem nc od ~
filter ( TC'[~FT1 , the ISDN COmmuni rat i nnc
on rnllPr (T 1_
-the S-BLS in Arfara i rnit (SB 1~ the ISDN hn-
58
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
cancellation circLit (IEC), and the ISDN bnrsr _on collar
{IBC)
Intel, National Semiconductor, Rockwell, and Northern-
Motorola are expected to have chips soon. By late 1987,
Motorola will be producing its-ISDN ~n~v_rsal digital loop
trans Piv r ( TDr,TI, the 245471, and the S/T tranSC^_eiVPrf the
25474. A universal data-link controller for the ISDN data
channel, the 1588, should be in production late in 1987.
Competitive pressures have contributed and will continue
to contribute to an acceleration of memory technology. The
Japanese presence in many market segments has led to a
dramatic reduction in prices and profit margins. (Davidson,
1984, p. 239.)** The 64-k_RAM reached commercial volumes in
late 1981, but pilot-production of 256=k RAMs commenced in
late 1982. The 512-k RAM was being used in laboratory
applications in 1984. (Davidson, 1984, p. 239.)**
Shott and Meindl (1985)** believe that ultra-large scale
integration (ULSI) technology will be successfully coupled to
advanced system architectures by the second half of the
1990s.
59
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
(U) REt,~UIREMENTS i'~2JD ASSESSMENT OF SOVIET TECHNOLOGY
The following assessment of Soviet technology is based
on a continuing evaluation of their telecommunications and
computer developments. It is very difficult in a closed
society to identify the exact impediments to research and
development efforts, but they should have to deal, one way or
another, with the -same technology problems as the .United
States did. There is not any- good single source as to what
the issues were in the United States as the frontiers of
communications networks were explored. The information is
available in the years of published articles which is beyond
the scope of this effort. However, much information can be
gained by?talking with the technologists that explored those
frontiers.
5.1 Key Technology Issues
Unless one can produce good quality microchips at
reasonable cost,- it is not cost effective to go to digital
communications [42], [43], [44]. Both switches and
multiplexers are practical only if microchips are used.
Also, high-data rates can be implemented only by using
microchip technology [35]. The actual circuit technology is
not a major problem since most of the circuit information,
except for that directly related to the manufacturing of the
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
microchips, is published in engineering journals or
standards. All the major western countries are manufacturing
digital switches, and multiplexers with rates up to 140
Mbits/second. The United States,- Japan, Germany and France
have capabilities at 560 Mbits/second and are working on
systems above one Gigabit/second. The data rates will
probably level off for the next several years at 560
Mbits/second because of the need for GaAs technology for the
higher data rates. The 1 to 2 Gigabit per second rates will
begin to appear in the early 1990's.
Without the microchip technology Soviet and Eastern bloc
data rates are going to be limited to the 34 Mbit/second
level, which also is too low to -make fiber installations
practical [30]. A review of the brochures of their present
equipment- [29], [30] indicates that most of the equipment
does not use microchip technology, but employs thick and thin
film processes. This level. of technology will not support
the deployment of multiplexers much above 34 I"lbits/second and
switches with large digital matrixes. While they have some
capability at 140 Mbits/second, they probably will not have
an operational capability until the early 1990's [30], [32].
Computer control of switches, whether it is part of the
switch or a separate host, is very software intensive [39].
To develop the programs to control networks and switches,
software developers with special backgrounds [41], [47] are
required. This key area has been a major impediment for the
bloc countries. Software development and quality control is
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
a key technology required by them to permit them to manage
-their network resources effectively. Much of the basic
information about their technology is gained from articles
published in technical journals [33] and also by foreign
nationals attending school and working in the West [25].
The above two technology areas (microchips and computer
control) are also concerns for packet and computer
switching. However, microprocessor development is an
additional technology which affects these systems. The data
rates and the throughput of these switches are directly
related to the power of the microprocessor used. This is not
as critical for the circuit switch since its information is
not switched through the microprocessor.
5.2 Soviet Developed Technology
The: Soviets do not appear to be ,able to satisfy the
three critical technologies of:(1) developing microchips; (2)
providing software control; and (3) providing
microprocessors, without outside assistance. They have or
could have the technical data to design the required
switching systems. They have the MT20 switch from the French
which, should give them the required technical data for stored
program controlled PABX switches [29], [48]. This would
include both the hardware and the software, but there is some
question as to their capability to manufacture the switch.
However, the software will give them a base to build their
control technology. In the area of packet switching, the
62
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Indians designed a packet switching system using their own
indigenous computer and software [33]. This information is
available. Also, information on a standard computer based
packet switch can be obtained through legal imports of X.25
boards for hosts [34] and the published information on X.25
protocols. With the control of switches going to standard
hosts, the software becomes more available and the Soviets
could develop their own designs. However, software designed
to run on a specially .designed processor, as has been done?
with switches like the MT20, is harder to adapt to different
operating environments.
It is very probable that the network control software
will become a Soviet developed technology by the early 1990s.
It will find its roots in information and software that has
been obtained from the west, but due to the uniqueness of
each network control situation, they will have, to develop
their own capability. It is more, than likely that it is one
of their higher priorities since it can significantly help
their present analog switches and network (31].
5.3 Soviet Acquired Technology
Of the above three critical technologies the one of
greatest need for the Soviets appears to be the acquisition
of the knowledge of how to manufacture high speed microchips.
The manufacturing knowhow has always been their biggest
problem because of their quality control. If they can ever
get an adequate microchip production capability in place, it
63
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
will impact on far more than just .their telecommunication
problems. The actual purchase of the required microchips is
not a very good solution because of the quantities required
and the special designs which are not commercially available.
Samples of the microprocessor chip probably will have to be
,acquired in order for the Soviets to study and learn its
physical design Qn~e they_ establish a a~abil;ty_~
manLfact?re i t ,
64
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
EMP, as a propagated signal, can damage equipment or
cause RF interference. :A surface burst generates an EMP due
to the rapid ionizing radiation effects. The EMP can be
divided into -two regions; i.e., a'source region which is
associated with the air volume and ground current returns
where the EMP is generated, and the radiated region. Within
the source region, which extends to 3-5 kilometers, electric
fields may exceed 100 kV/meter. The EMP energy coupled to a
system must be treated as a survivability issue because
permanent damage may result. Beyond this 3 to 5 kilometer
distance, the radiated. fields are less intense, and less than
those fields due to high altitude EMP. High altitude EMP is
generally specified as 50 kilovolts/meter from a system
survival requirement viewpoint. ?The source of high altitude
EMP is ionization at the top of the atmosphere caused by X-
rays and GAMMA rays moving downward from the nuclear burst
points. Residual GAMMA radiation can travel over 2000 miles.
The Soviets have done a number of studies on radiation
hardness. of optical fibers and components. Degradation of
fiber optics transmission systems due to nuclear radiation is
of major concern. While fiber optics are attractive for use
in High Altitude EMP mitigation, it must be made clear that
fiber optic media is not completely insensitive to EMP. It
is, however, less sensitive than metallic or atmospheric
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
transmission media. Single mode fiber exhibits the best
recovery; with .additional attenuation (path loss) being 3 dB
to 10 dB over benign levels 10 to 15 .minutes after EMP
exposure. Multi-mode and plastic fiber is much worse and may
be destroyed by .the .intense heat deteriorating the light
guide capabilities of these fibers. [51,52]
As far- as digital switching components are concerned,
solid state components are susceptible to EMP. Different
radiation hardening techniques are being considered. These
alternatives or combination of alternatives are listed below:
1. Use of switching materials .(GaAs instead of silicon)
and. shielded enclosures to make the equipment
resistant to EMP;
2. Creation of backup power sources for circuits and
terminal equipment;
3. Reduction of the number of switches employed in a
network;
4. Employment of widely dispersed switches and circuits,
away from known targets;
5. Improved methods of testing to better quantify
baseline problems and design improvements.
During the next 10 years, it is expected that the number
of switches will decrease significantly. This expectation is
.based on the observation that switching and computing devices.
are. becoming so efficient that the number of switches used by
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
public telephone companies has already decreased in a steady
manner (the trend has been established).
Another area of improvement will be the standardization
in the testing for EMP effects. Currently, different
laboratories frequently ,come up with divergent results making
it difficult (i'f not impossible) for the engineering
community to devise uniform proposals for improving the
resistance of switching components and networks.[55]
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
(U) CONCLU3ION3
The Soviets are at the threshold of deploying digital
systems. They have implemented it in small steps, but appear
not to have the capability to make major installations. They
have made low capacity digital transmission systems and
appear to have produced their first all digital exchange for
end office use [29], [30J, [48]. They have also implemented
simple, but effective automated network control schemes [31],
[32J. The major factors that appear to be holding them back
are the ability to produce, quality microchips in the numbers
required .and the sophisticated software to make the robust
network and switching control systems that the new networks
require . They appear to be where the United States was 15
years ago',, but should be able to implement this technology
within the next 10-20 years.
The dynamics of the development and implementation of
high capacity digital transmission systems, with throughput
capacity equivalent to existing and developmental analog
systems carrying tens of thousands of channels for mainline
communications links, provide no basis to assume that the
entire primary mainline network will be switched over to
digital transmission methods within the next 20 years.
However, the creation of a mainline primary digital network
with limited throughput capacity (based on switching medium-
capacity balanced cables, coaxial and radio relay lines over
68
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
M2011
to digital transmission systems), in parallel with the
existing high capacity analog network, can solve the problem.
Two of the greatest needs for the Soviets are : (1) the
ability to manufacture good quality microchips needed for
high bandwidth digital switching; and
(2) the ability to
develop network control software that will lead to high
bandwidth network communications management. By the early
1990s the Soviets should be able to acquire sufficient
production capability to manufacture a digital circuit switch
of the MT20 technology level and switches of the packet and
computer type equivalent to the early 1980 .western
technology. Network control software will become Soviet
developed in the early 1990s. Early development efforts will
be spurred on by acquisitions from the western world.
The future of telecommunications for both the West and
Soviet bloc is the merging of computer and communications
technologies. The western countries are beginning to
experiment with networks of the ISDN type, with Japan in the
lead. The Soviets have announced that they are developing a
limited version of the ISDN layered concept-. Since digital
switching technology is?an essential part of any ISDN, then
digital switching technology must be considered a critical
technology for Soviet and Western telecommunication advances.
As far as fiber optics is concerned, it is expected
that by the early 1990s all new toll transmission systems
will be of the fiber optics type. In the Soviet and bloc
countries, new long distance fiber systems will likely
69
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
appear by 1990. EMP can affect fiber optics and solid state
components' transmission quality and signal intensity, and
the Soviets are conducting research in this area.
70
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
M201i
Section 8
(U) REFERENCES
This section is divided into tNO lists of references.
The first list is given by number and these numbers are
referenced in the text. The second list which follows the
first list is referenced by author and year and each
reference in this second list is flagged by a double asterisk
in the text.
8.1 LIST 1 (REFERENCES)
1. Recent Discussions between GTE Government Systems and
SPRINT personnel regarding hardware investment and expendi-
tures (R. Phillips and R. Cassum)
2. "ATT to Upgrade Their Radio Relay Systems", Electron~_G,
-Sept. 18, 1986.
3. Recent Discussions between GTE Government Systems and
Sprint personnel regarding KMART Space Network .(R. Phillips
and R. Cassum)
4. GTD-5' EAX Brochure, GTE Communications Systems Corp.
5. J. Gechter, "Next-Generation Switching: System Character-
istics and Industry Implications, (See Appendix A).
6 . A. E . Joel, Jr . , Editor (197 6) , F:7 ect ron; Swi h~ ng~
C__entra~ O f;rP ystems of the World, IEEE Press, New York.
7. "Dynamic Network Control Systems (DNC-500)," Northern
Te~ecom Broch~r , 1985.
8. .Edwin E. Mier, "PABX Trends and Technology Update:
Following the Leaders," Data Co mnniratinnc, Sept. 1985.
9. F. W. Ellersick and D. L. Schilling, "Special Issue on
Progress in Military Communications - II," IFEF JOLrnal on
S ~ d A Pa in o mnni~afinna, v01. SAC-4, n0. 2, Mar.
1986.
10. Special Issue on PABX/LAN Connectivity, T Omm,n;-
cations Maq ;na, vol. 24, ~no. 12, Dec. 1986.
11. J. Robert Lineback, "New Tests Will Try to Get Stalled
Digital Net Moving," F.lec ron; , Feb. 10, 1986. "
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
12. Trudy E. Be11, "Communications," 7 ..F SpPCtr,~m, Jan.
1986.
13. Newsfront, "France Ugrades Its Trans pac Network," Data
o m ,ni _a i ons, Nov. 1984 .
14. Robert Rosenberg, "Next Step for Fiber Optics: The Local
Loop,".Electronics, Nov._27, 1986.
15. John Voelcker, "Helping Computers Communicate," TFF.F.
~ectrum, March 1986.
16. "T-1 (Packet) MUX Unveiled," Data .o mvnicatinnG,
17. Discussions between R. Phillips (GTE} and Roy Rossner of
TELENET, 1986.
18. John Gosch, "Philips Local-Area Net Employs Fiber
Optics," F.lec roni-G, Dec. 23, 1985. (lan fo)
19.Bernard C. Cole, "Optical ICs: The New Alternative,"
E1ec roni_s, Nov. 18, 1985..
20. Clifford Barney, "Siliconix Joins Class 1 Clean-Room
Elite,." Elec roni s, Nov. 18, .1985.
21. Electronics Newsletter, "Vitalink Offers a Wide-Area
Link for Local Nets," Electronics, Nov. 27, 1986.,
22. J. Robert Lineback, "IBM Bridging Scheme Roils the Local
Net Community, " F1 ec xoni _a, Nov. 27, 1986.
23. A. E. Joel, Jr., "Special Issue on Serving the Business
Customer Using Advances in Switching Technology," IEEE
Jo ,rnal on ~ Pf,tPr~i ~YpAC in o m ,ni cat i nnc, vol . SAC-3,
no. 4, July 1986.
24. "Network Management & Control System: TPS/II," Telenet
Brochure, 1985.
25. Vyvyan Tenorio, "India Strives to Join High-Tech World,"
F.lec roni G, Sept. 2, 1985.
26. M. Gregory, Project Eureka, "Wide Band Switching,"
French Advance in S ; ncP & T ~chnolnew vol. 1, no. 2,
Fall 1986.
27. Robert T. Gallagher, "Fully Optical Switch Looms for
Video," Electron; s, Feb. 10, 1986.
28. "Networking Technical Notes," Digital Equipment Corp.,
1980.
~~
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
29. "Electronic Automatic Telephone Exchange: MT-20/25,"
Soviet Brochure, 1986.
30. Selected Material from Soviet Communication Exhibits,
Moscow, 1986
31. Fanya B. Bakaleyshchik, "Planning and Design of
Automatic Long-Distance Telephone Exchanges," JPRS L/1_0354;
Foreign Broadcast Information Service Translation, Feb. 26,
1982 ,
32. T. A. Vladimirova, et. all., "Terms of the Unified
Automated Communications Network (UACN)," Elektrosvya~,,,
no. 1, 1979.
33. Kishore K. Oka, "Indian Plans for a Public Data
Network," (See Appendix B - to be supplied)
34. "ECCN 1565, Advisory Note 9," LT.S. Exnorr Reg~lat;nnG,
Dept . of Commerce, Oct . 1, 1986.
35. Robert T. Gallagher, "Single-Chip Card Does Switching,"
Elec on; s, Feb. 18,1985..
36. Robert T. Gallagher, "EC Strives for Telecom Standards,"
Rlec ron;~~, Feb. 18, 1985.
37. Karen Berney, "GSA Seeks Fast Phone Net," Elect ron; ,
Feb. 18,1985.
38. Donald Marsh,, "ISDN Evolution Possibilities in the
U.S.," Jo ~rnal O T 1 nmmnnicatinn To*?-+r+~-y~~ VO1. 1, nO. 2,
Summer 1982.
39. DSS Overview, DAXCON Network_ ,p o t y Pm, Microtel
Limited, May 1986.
40. C. R. Dhas and V. K. Konangi, "X.25 an Interface to
Public Packet Networks, " TF.F.F Commi~ni cat i nnc Magazi nP, vO1 .
24, no. 9, Sept. 1986.
41. Kenneth J. Macleish, Steven Thiedke and David
Vennergund, "Expert Systems in Central Office Switch
Maintenance, " TF.F.F. Comm ~ni cat; nn5 Maga2i na, vol. 24, no. 9,
Sept. 1986.
42. Robert Rosenberg, "Users Face a Risky Choice in ISDN
Chips Sets," Elec on; a, Oct. 2, 1986.
43. "Bullish .on ISDN, Intel Launches Its Chip Set,"
F.1 ect ron; a, Oct 2, 198 6 .
44. "One-Chip Bell 202 Modem Sell for Under $10,"
F~ectron;c-s, pct 2, 1986. .
~~
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
45. Jonathan S. Tuner , "New Directions in Communications
(or which Way to the Information Age?)," o min; a ionG
Magazine, vol. 24, no. 10, Oct 1986.
46. "Special Issue on Integrated Services Digital Network:
Technology and Implementations-II," Sele d A_r as in
CommLnications, vol. SAC-4, no. 8, Nov. 1986.
47. Tobias Naegele, "ITT Reins in Its Troubled System 12,"
Fle Toni s, Feb. 24, 1986.
48. "Leningrad Exchange Opens," In na ;onat T 1 om_muni-
cations Intelligence, Jun. 27,1986. (Same exchange as
reference [29].)
49. CCITT (1984), .Economic and Techni al A pests of th
Transition from Analog to D; g; tal Co m ~n i a ; onG,
International Telecommunications Union, Geneva.
50. ITU (1984), Fronom;c and Technical AGpects of the
Transition from nalogue *o Digit-a1 T 1 om_m ~n; at; nn
Networks, International Telecommunications Union, Geneva.
51 Fiber Optic and Digital Switching Needs for a Network,
October 1986, A T R Project.
52. .National Security Telecommunicaions Advisory Committee
(NSTAC) Electromagnetic Pulse (EMP), Final Re or , July 1985.
53. Advanced Technology Planning (ATP), Project 1611,
?ntearated Serv;_cPS Dicrital N wo k (T DN) o muni a inn
Trend Surv~v, December 1986, GTE Government Systems
Corporation, Mountain View, CA
54 . SLlitmary of SOVi P and W ~ n D' gi ta1 wi h; n~
Technology. November, 1985, GTE Communications Systems
Department, GTE Government Systems Department, Mountain
View, CA.
55. T?,F SnectrLm, Special Issue on Telecommunications,
November, 1985; Safeguarding the National Security.
8.2 LIST 2. REFERENCES (REFERENCED BY DOUBLE ASTERISK
IN TEXT, AND BY AUT$OR/YEAR.) ,
Amstutz, S. R. (1983), "Burst Switching - An Introduction,"
IF.F.F. Co mien i rat i nn c Mag , NOV . , pp , lA-7A.
Anon. (1978). "Integrated Digital Communication Network,"
Soviet- ~ n -rarprj Cranr-~arri for T rma and D ; nit- i nnG,
English Translation avail., AD-B072 514.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Anon. (1982), "Effect of Future Technology on Central Office
Switching," GTE Laboratories, Inc., Waltham, MA.
-Anon. (1984). "The Military Critical Technologies List,"
Office of the Under Secretary of Defense, Research and
Engineering, Washington, D. C., AD-A146 998.
Anon. (1985A), "ABM 301 SPC Field Telephone Exchange,"
Fr~cGSOn R_v~Pw, Vol. 62, No.2, pp. 66-71.
Anon.' (19858), "Digital PBXs in Europe -- The Next Five
Years," R~Dort by Logica onai~lr;_ng Ltd,, 64 Newman
St., London, WIA 4SE, England.
Anon. (198SC). "Soviet ,Acquisition of Militarily Significant
Western Technology: An Update," No Corporate Source.
Aripov, M. N. (1980), "Transmission of Digital Information
Along Low-Speed Communication Channels," English
Translation avail. AD-8062 495L.
Bakaleyshchik, Fanya B., "Planning and Design of Automatic
Long-Distance Telephone Exchanges," JPRS L/10~54,
Foreign Broadcast Information Service Translation,
Feb. 26, 1982.
Beam, A. (1985), "Russia Gropes for a Way to Enter the High-
Tech Age," B~G~n ~ w k, 11 Nov., p. 98-ff.
Blackman, .J. A. (1979), "Switched Communications for the
Department of Defense," IEEE TranG= Comm ,n ,, Vol.
COM-27, No. 7, pp. 1131-1142.
Carlson, H. V. (1983), "A Look at Switching, Past to Future,
An Interview with Amos E. Joel, Jr.", Bell r,abora_tc_,ria_G
Record, Vol. 61, No. 1, pp, 20-24.
CCITT (1984) , Econom; c and Tar~hn i ~^~ 1 A Re~-r S of h
Transition from Arialo~ t0 Digital CO munieatinnc~
International Telecommunication Uniori, Geneva.
Chinchikas, G. V., A. S. Yarutis (1983), "The Application of
Microprocessors in the Synchronization of Communi-
cations Network Clock Generators," Radioelektronika
(Kasai., Vol. 19, Iss. 5, pp, 91-96. (DS29.)
Chorafas, D. N. (1985), The Han t~nnk o Da a .o muni~atinnc
and Comt~uter Netwo k , Petrocelli Books, Princeton, NJ.
Clay, A. W, and R. D. Wirth (1985), "Communications Software:
Technology Map to the 1990's," 7FFF GT,OBE COM '85,
Paper No.
75
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Davidson, W. H. (1984), The Amazing Ra Winning th _
Technorivalry with Japan, John Wiley & Sons, New York.
Gallagher, E. F. (1979), "Digital Time Division Switching for
Military Communications," IEEE Trans. Common., Vol.
COM-27, No.7, pp. 1138-1143.
Gerla, M. (1985), "Packet, Circuit, and Virtual Circuit
Switching, " chap . 1 6 i n ComnLter o m ~n i .at i nn G ~ Vol .
II, Systems and Applications, edited by W. Chou,
Prentice-Hall, Inc., Englewood Cliffs, NJ.
Goldenberg,- L. M., B. D. Matyushkin, and M. N. Polyak (1980),
"Digital Units of Coupling (Transmultiplexers) in the
Communicating Systems," Elektrosvay~, N. 1, pp. 28-32,
English Translation'Avail. AD-BOSS 489L.
Goodman, S. (n.d.), "Soviet Computing and Technology
Transfer: An Overview," world Posit-ics, pp. 539-570,
Princeton Univ. Press. [Cited in Melvern et al. (1984),
p. 287.)
Haselton, E. F. (1984), "New Switching Concept Integrates
Voice and Data Bursts," ,LOB. .OM '84, pp. 126-129.
ITU (1984) .
ITU (1984), "Economic and Technical Aspects of the Transition
from Analogue to Digital Telecommunication Networks,"
International T 1 o muninat-inn 7nion, Geneva.
James, J. W. and M. Marques (1984), "Beefing up the
Processing Power of the 4ESS(TM) Switching System,"
AT&T Bell Labora nr'1PS` R ord, Feb., pp. 17-20.
Joel, A. E., Jr., Editor (1976), E7ec Toni Sw~ _hing~
Ventral Off; CP y~tems of h Wo 1 ci, IEEE Press,
New York.
Joel, A. E., Jr., Editor (1981), F-_hc ron~ ~w; hing?
Digital _ n ral pf i-P SyGtPmS of the World, IEEE
Press, New York.
Joel, A. E., Jr., Editor (1985A), "Special Issue on Serving
the Business Customer Using Advances in Switching
Technology, " IEEE Trans 1 PPS nrPa a in o min ,
Vol. SAC-3, No. 4, Whole Issue.
Joel, A. E., Jr. (1985B), "Time Division Digital Switches:
A Progress Report on Installations in -North America,"
Telephony, Sept. 16, pp. 68-71.
Kaplan, G. (1985), "Invasions and Counterinvasions," IEEE
Spec gym, Vol. 22, No. ll, pp. 64-69.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
Kassel, S. (1985), "Outlook for Soviet Advanced
Technologies," pp. 53-69 in P_roc. Gonf= ; n and
Technology in th_ Sovi lnion, Stanford, CA, AD-
A152828.
Mayo, J. S. (1982A), "Evolution of the Intelligent
Telecommunications Network," S-~ n , Vol. 215, No.
4534, pp. 831-837.
.Mayo, J. S. (1982B), "Technology Requirements of the
Information Age," Bell Labo atnriPS R ord, Vol. 60,
No. 3, pp. 55-59.
McDonald, J. C., Editor (1983A), Fundamen a1c of Dig; al
Switchingr, Plenum Press, New York.
McDonald, J., Editor (1983B), "Digital Switching," Special
TSS17P Of TF. .F. C'~mmuni r?at i nnc Maga2i nP, VOl. 21, NO. 3,
Whole Issue.
Melvern, L., N. Anning and D. Hebditch (1984), Techno-
Bandits, Houghton Mifflin Co., Boston.
Misulovin, L. A., V. V. Makarov, Uy. A. Baklanov (1981),
"Yess ATS 'ISTOK' Integral Analog-Digital Communi-
cations System Test Results," Elektrosvyaz (T^SR~,
SEP., pp.4-10, English Translation Avail. AD-B064 002.
OTA (1985), "Information. Technology and R&D: Critical Trends
and Issues," ~x~o,-t No. ETA- .TT- 68, U. S. Congress,
Office of Technology Assessment, Washington, D. C.
Pervyshin, E. K. et al. (1984), The Information Transmission
Industry, 1n RLSSian, Radi_~ ; vy~, Moscow (DSSS) .
Phillips, B. H. and J. P. Hurrell (1985), Military
F.1 ectroni a .
Popov (1978), G. M. (1978), "A Small Computer as a Message
Switching Center," iTDravlyayLShchiye ;StPmy ; Mash;ny
(USSR), N.4, pp. 41-45, English Translation Avail.,
AD-B052 841L.
Rinearson, L. E. (1985), "Software Development for the ZODIAC
Tactical Switch, " ?F.F_.F. MIT,CC)M ' 8 , Vol . 2, pp. 339-345 .
Romanov, V. D. (1982), "Digital Transmission Systems,"
Elektrosvy~, No. 4, pp. 53-55, English Translation
Avail. AD-B069 448.
Sharashenidze, A. I., L. Y. Misulovin, I. I. Panin, L. A.
Kovalev, and S. S. Kodner (1982), "Electrical Communi-
cations - .Selected Articles," Elektrosvyaz ( S R),
N4, pp. 16-25, 47-63, English Translation Avail.
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
M2011
Shvartsman, V. O. and V. G. Osipov (1980), "Data Transmission
on the Telephone Switched Exchange," F.lektrosvya~_
.(USSR), N.1, pp. 25-27, English Translation Avail.,
AD-B055 S11L.
Shott, J. D. and J. D. Meindl (1985), "Wonder Chips: How do
you Connect 10 Million Components on a Dime?" Sri nr
$,~, Nov., pp. 115-T17.
Slana, M. F. (1983), "Ti:me-Division Networks," Chay~ter 5 in
Fundamentals of Digital Sw; c-h;na, J. C. McDonald,
Editor, Plenum Press, New York.
Torerro, E. A., Editor (:1985), "Hello Again -- The Future of
Telecommunications," TEES Spectrum, Vol. 22, No. 11,
Whole Issue.
Waters, W. E.
(1983),
Rl_ectr;cal Ind~.tinn from D~Stant
~ n
o~
Prentice-Hall
Inc
Englewood
Cliffs,
NJ.
,
.,
Wood, D. C. (1985), "Computer Networks: A Survey," Chap. 14
iIl Gom~uter o m~ni~atinna~ VO1. II, Systems and
Applications, Edited by W. Chou, Prentice-Hall, Inc.,
Englewood Cliffs, NH.
~Q
Declassified in Part - Sanitized Copy Approved for Release 2013/04/23: CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
APPENDIX A
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
Next-Generation Switching:
System- Characteristics and Industry Implications
by Jerry Gechter, GTE Laboratories Inc.
Introduction
GTE, as both manufacturer and provider of a wide variety of network services, has been
particularly well-placed to observe the evolution and implications of switching
technology.This paper reviews results of a recent corporate tasks force study of the
characteristics and implications of next-generation switching. We feel the subject is of
particular relevance for this conference, since the conference itself will influence the
directions the industry will take. '
Briefly stated, the conclusions we reached are the following:
1) By the 1990's network service providers will require network capabilities going
beyond the potential of current digital switching systems. Motivations include new
service options as well as cost-effectiveness of network operations, administration, and
maintenance.
2) Switching implemerrtation issues are increasingly bound up with other aspects of the
network service provider's business. Traditional switching control is only one part of
the real-time data management necessary for automated provisioning of network
services in the 1990's.
3) In the future switching control and information transport structures will be
conceived and evolved separately. Benefits will include enhanced service capabilities,
easier introduction of new switching technology. A further consequence will be reduced
dependence of network service providers on switching manufacturers for
implementation of services.
This paper will discuss the motivations and technical capabilities on which these
conclusions are based.
Motivations for Next-Generation Switching
For our study we viewed network service provider needs as the high-level definition of
next-generation switching. We put these needs in two categories:
1) New switching capabilities
2) Cost and performance improvements
In the first category we placed distinctly new switching functions which cannot be
provided by evolved versions of existing central office technology except by overlay. In
the second were next-generation system advantages for the implementation of current as
well as future services.
As noted in Figure 1, the emphasis in new switching capabilities was on the h~~iggh~~-
bandwidth side. As is also eviderrt from the figure, lt is difficult to specify upper boun~C s
on the bandwidth needs of switching systems required in the 1990's. Furthermore, the
bandwidth will be used in applications ranging from long holding time circuit
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
,,
connections (entertainment video) on one hand, to interactive bulk transfers (CAB/CAM
applications) on the other. Our technology assessment concluded, finally, that braodband
switching will not remain an overlay add-on as it is today, but will be fundamental to the
capabilfies of a central office or PBX in the 1990's.
For cost and performance improvements, our emphasis was on advantages of integrated
systems. While such systems have been described in conference and trade publications
by GTE, ATB~T and others, the intent in our study was to be as specific as possible about
the advantages they bring to network service providers.
Figure 2 gives an indication of the degree of complexity which will be implicitly or
explicitly present with evolved versions of current technology by the late 1980's. Some
of these network switching functions are already present in carriers' networks; others
are only in the planning stages. Hence it is somewhat difficult today to quantify the cost
of complexity in this proliferation of networks.
Our first conclusions are given in Figure 3. The figure identifies five specific areas of
benefit: service software, switching hardware, OAM costs, networking capabilities and
economies, and universal interfacing. The benefits which we ident'rfied in these areas
are also described and are not necessarily the ones which first come to mind. Expanding
on the text in the figure, they are as follows:
Service software: Whatever other benefits may arise from new software structures in
next-generation systems, we feel the strongest demonstrable motivation will come from
integrated services. Coordination problems in multi-media services will be far greater
in non-integrated networks, translating to cost and responsiveness advantages for
integrated systems.
Switching hardware: While we felt it was difficult to demonstrate cost advantages for
integrated systems on sport-for-port basis, we did see a significant provisioning
advantage for integrated systems, particularly in an environment where new services
will be introduced with uncertain demand. In the early part of the 1990's one would
like to be able to handle the entire service mix with switching capabilities sized
incrementally over voice requirements.
Operations. administration= maintenance: Integgration of networks translates
immedaitely to a reduction in staff requirements for OAM.
Network capabilities and economies: The greatest networking advantage of integrated
system appeared to us not to be in transmission saving, but in support to distributed
processing aapppplInations. The point, briefly stated, is that once one folds control message
communication into the common transport network, then the system software of an
integrated switch becomes available as support for more general distributed applications
on the network Such applications will be typical in the 1990's operating environment.
Universal interfacing: Current access interfaces, including the ISDN interfaces, will not
exhaust the requirements for switches introduced in the 1990's. Switching approaches
chosen today must impose as few constraints as possible on system ability to adapt to
currently unspecified interface standards. An integrated system, in that it is can
uniformly handle all information types, imposes fewer constraints on the interfaces it
can economically aooept.
With this classification of potential advantages of integrated system, we attempted some
quantification of advantages on the basis of available data, current and projected. The
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
result was the list of qualitative conclusions given in Figure 4. As noted, the two major
items were oortsidered to be the OAM and networking advantages, in that these impacted
total networ>F operating cost, quality, and responsiveness.
Next-Generation System Characteristics
We next attempted to further define the nature of next-generation switching by looking
in detail at carrier business objectives, at characteristics of next-generation system
options, and at various future scenarios. The results were as follows:
1) Business objectives
Figure 5 gives a list of eight fundamental architectural requirements deduced from
career business objectives. The descriptions are largely self-explanatory, and the
ooncuusions which we drew from the list are indicated in the following figure. The main
message is that these objectives have both control and transport implications, and in fact
that any technological solutions must deal separately with the control and transport
structures necessary to meet the objectives.
Proceeding further, we came to the next-generation network model given in Figure 7.
Here we have formalized the control/transport boundary. This allows the transport
network to evolve with hardware technology and service-bandwidth demand, while
retaining the control capabilities in the higher layer. Moreover, it enforces a common
context on all control considerations in the network, incuding call processing, customer
input, and network administration.
It was noted finally that such a formal boundary, analogous to what exists for LAN's
today, has consequences for the business relationship between switch vendors and
network service providers. Today the switch vendor produces a bundled offering of
transport and control hardware and software. The formal boundary allows feature
software to exist as an independent product: separate from transport hardware and
governed by standards arising in the data processing area. The introduction of common
channel signaling, for current generation switching systems, should hasten such an
evolution in the industry.
2) Next-generation system options
Our analysis of next~eneration options considered a number of proposed systems from
our own corporation and others. In qualitative terms we concluded that it appears
possible to develop a system with the fdlowing combination of properties:
- Full integrated system advantages
-Minimal limitations on total throughput capacity
- Satisfies currently-formulated central office requirements, including delay
pertormance
- Allows functional arcuit and packet approaches to video and other broadband services
We feel that these capabilities are highly desirable in economically meeting future needs
of network service providers. -
3) Scenario variables
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
Our treatment of scenario variables identified certain other evolvability or flexibility
requirements which had not been identified from the other points of view. Examples
included
- VLSI to optical component evolvability
- Efficient switching and multiplexing capabilities
-Phased system implementation
These, added to the condusions from other areas, gave us a reasonably dear idea of what
a next-generation system should be. Looking from the network service provider's and
from the technologist's point of view, what we see is a system that is most easily
imagined as an "enhanced" local area network
Like a LAN it has a formal boundary between control and transport functions, thereby
turning control software into a separately sourceable product in a data processing.
context. Also like a LAN, the transport function handles all its applications within the
same information transfer mechanism.
The required enhancements to LAN operation apply to both transport and control aspects.
The transport system must be so conceived as to carry voice, data, and video subject to
the performance requirements of each. Further tbe~h should be no geographic limitation
on the extent of the network -PFcrc
The control requirements include extended system-software capabilities for reliable,
distributed, real-time operation, as well as the necessary application processes for
telephony and other services. Further the entire system must be testable and
maintainable on afully-distributed basis. We emphasize that while these control
capabilities represent a considerable advance over the current state of the art, the
problems are not unique but belong to the mainstream of distributed data processing.
Concluslor~s
The message we would like to leave is that there is such a thing as next-generation
switching, defined by applications of evolving technology to the perceived needs of
network service providers.
For carriers such as ourselves the consequence is dear: the industry should evolve so as
to provide the kind of next-generation systems we see as necessary. This has
implications most directly in standards. The control/transport boundary should be
formalized and control software should join the broad range of distributed computing
.applications.
Beyond that we see a need to encourage the develop of independerrt, application-specific
soitnrare vendors. As communications services bec~mme ever more dosely entwined with
our own and our customers businesses, the industry must provide these services with
the responsiveness and flexibility the customer expects in all areas of his business. For
ISDN and for services beyond ISDN, the industry will only be successful if it is also
entrepreneurial.
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
We at GTE intend to encourage such an evolution through our commitment to the
standards process, through an open system approach to our own network architecture,
and through presentation of our views at public fora such at this one.
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
New Switching Capabilities
CAPABAITIES OF FULL --
SERVICE PROVIDER
- voICE f---cAPAenrrlES of
-DATA PRESENT NETWORK
- VIDEOTEX
- TELECONFERENCING .
N
- ENTERTAINMENT oWLITY TV
- HIGH DEFINITION TV
- STEREOSCOPIC VIDEO
- HOLOGRAPHIC VIDEO ~3D1
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
Candidates for Network Integration
Circuit switching
Packet switching
Wideband switching/cross?connect
Broadband (video) switching
D-channel message handling
CCS message handling
DSI switching (speech spurts)
Channel rate adaptation (voice and data)
Remote unit control-message handling
OAM message handling
Billing message handling
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
BENEFITS OF INTEGRATION:
COSTS AND CAPABILITIES
BENEFIT
VALUE
Service software ~
Reduced system complexity for integrated services
Switching hardware
Integrated switch sized for fatal demand
Operations. Administration, Maintenance
Single network to administer, manage, and maintain
Network capabilities and economies
Coordination and flexible allocation of control resources
Universal interfacing
Adaptable to interface requirements beyond ISON
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
INTEGRATION -PRIORITIZING OF BENEFITS
BENEFITS
~ ~MpACT
Service software
Incremental impact: Advanced services devektpment
Switch hardware
Incremental impact: Advanced services deployment
Operations, Administration, Maintenance
PRIMARY DRIVER: Impacts total network operating cost
Network capabilities and economies
PRIMARY ORIVER: Impacts total network quality and responsiveness
Universal interfacing
Depends on market evolution; Wideband Packet may be a required interface
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
SYNTHESIS OF CUSTOMEp FOCUS AND COST
MANAGEMENT VISIONS YIELDS EIGHT FUNDAMENTAL
ANCHITECTURAL flEpU1REMENTS
? Unified Control Structure
? Automated Allocation
of Resources
? Unified OAbM Systems
? Universal Interface
Support
? Simplification of
Network Operations
? Flexible Resource
Allocation
? HighNariable Bandwidth
? Modular Network
Functionality Growth
Allows multiple processing systems to work together and share
resources
Capability to automatically assign transport resources to satisfy
network and customer requirements ,
Ability to operate and maintain the network as a single entity
Ability to accept diverse standard -and customerspecific inter-
face protocols
Reduction in the overall complexity of the network as seen by
the network operator
Capability to assign bandwidth to fulfill customer and network
requirements in whatever granularity requested
Defines the upper limit of allocatable bandwidth
Ability to add processing capabilities switch attempts, services)
in a graceful and economic fashioa
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
ARCHITECTURAL REQUIREMENTS:
CONTROL AND TRANSPORT
~ ~
? UNIFIED ? AUTOMATED ~ ? UNIVERSAL ~ ? fLEX18LE
CONTROL ALLOCATION I INTERFACE 1 RESOURCE
STRUCTURE OF RESOURCES ~ SUPPORT ~ ALLOCATION
? UNIFIED ? MODULAR
OAiM NETWORK
~ ? SIMPLIFICATION
gYSTEM fUNCT10NAUTY ~
A
GROWTH 1
IMPACTS ARCHITECTURE
CONTROL STRUCTURE
? HICHNARIABLE
SANOWIOTH
IMPACTS ARCHITECTURE
TRANSPORT STRUCTURES
IMPLICATIONS NETWORK ARCHITECTURE TECHNOLOGY PLANNING DEALS WITH TWO
INDEPENDENT. FUNDAMENTAL GTRIICTUREa: THE CIaAfTROI STRUCTURE
Tf~ TRAMSPgAT STRUCTURE
OF NETWORK
OPER
TIONS
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6
 NEXT-GENERATION NETWORK MODEL
T~a~sper:
? Provides basic transport functions
? Transport of voice. data. video. and control
? Emphasis on integration of switching and transmission
? Emphasis on unified transport structure
? Nonstandard protocol conceivable
Acceu .
? Accommodates standard access interfaces
? Provides interface to network iunctions and resources
? ~ Emphasis on unification of access
Control
? Provides all network signaling. high-level functions. feature logic. and ~terface to information pracessing
? Interprocessor communication through the transport structure
? System functions distributed across transport network
? Control structure is independent of transport and access
Declassified in Part -Sanitized Copy Approved for Release 2013/04/23 :CIA-RDP92B001818000300270030-6