Module-4 (part I)

TIME DIVISION SWITCHING: Introduction, space and time switching, Time switching networks,
Synchronisation.

Text Books: 1. Telecommunication and Switching, Traffic and Networks - J E Flood: Pearson
Education, 2002.

TIME DIVISION SWITCHING

1.Introduction:

• The first application of digital time-division switching was to provide tandem switching
of PCM (Pulse Code Modulation) junctions and trunk circuits.
• Examples of these systems are Bell ESS (Electronics Switching System) and the
French E 12 system.
• Since a tandem exchange is similar to the route switch in a local exchange, local systems
were developed by adding reed-relay space division concentrators.
• examples of such system include the initial version of System X and the AXE 10 and
E 10 systems.
• The developments in the technology of solid-state integrated circuits eventually solved
the BORSCHT problem and enabled TDM concentrators to replace space division
concentrators. This is evolution of integrated digital network.
• Finally extension of digital transmission over customers’ lines has enabled integrated-
services digital network (ISDN).
• It provides the customer with a wide variety of services, based on 64 Kbits/sec digital
transmissions, over a single line from a local exchange.

The architecture of a system X local exchange is shown in figure 1.

• The digital switching subsystem (DSS) corresponds to the route switch of Digital
switch.
• The line terminating units of PCM junctions are connected to it directly.
• Voice-frequency junctions are connected to it directly.

1
• Voice-frequency junctions are connected via a signalling internetworking subsystem
(SIS) and an analog line terminating subsystem (ALTS) that provides analog/digital
and digital/analog conversions.
• Analog and Digital customers’ line are connected to the DSS via a concentrators,
known as the digital subscribers switching subsystem (DSSS). These concentrators may
be in the main exchange or located remotely.
• The processor utility subsystem uses software built largely from modules
corresponding to the hardware subsystems that they control as shown in figure 1.
• The system has a capacity for 60,000 lines and 10,000 E of traffic. A tandem or trunk
exchange uses a similar DSS, but it has no concentrators.

Figure 1: system X local exchange

2
2.space and time switching:

A tandem switching centre, or the route switch of a local exchange, must be able to connect
any channels on one of its incoming PCM highways to any channels of an outgoing PCM
highway. A connection will occupy different time slots on the incoming and outgoing
highways.

2.1 Space switches:

• Connections can be made between incoming and outgoing PCM highways by means of
a cross point.
• However different channels of an incoming PCM frame may need to be switched by
different cross points in order to reach different destinations.
• The cross point is therefore a two input AND gate. Thus the switching network must
be able to receive PCM samples from one time slot and retransmit them in a different
time slot. This is known as time switching.
• Simple time division switching network make connections between channels on
highways carrying a primary multiplex group, i.e. they operate at 1.5 M bits/sec or
2Mbits/sec.
• 2Mbits/sec line system has 32 timeslots. It uses 30 speech channels, time slot 0 is used
for frame alignment and 16 for signalling.
• One input is connected to the incoming PCM highways and other to a connection store
that produces a pulse at the required instants.
• A group of cross point gates can be implemented as an integrated circuit, for example
by using a multiplexer chip figure 2 shows a space switch with ‘k’ incoming and ‘m’
outgoing PCM highways, each carrying ‘n’ channels.
• The connection store for each column of cross points is a memory with address location
for each time slot, which stores the number of cross point to be operated in that time
slot.

3
Figure 2 space switch

4
2.2 Time switches:

Figure 3: Time switching a) Time switch b) Space-division equivalent

• The principle of Time switch is as shown in figure 3(a). It connects incoming n-channel
PCM highways to outgoing n-channels PCM highways. Since any incoming channels
can be connected to any outgoing channels.
• It is equivalent to a space division cross point matrix with “n” incoming and “n”
outgoing trunks as shown in figure 3 (b).
• Time slot interchange is carried out by means of two stores, each having a storage
address for every channels of PCM frame.
• The speech store contains the data of each incoming time slots at a corresponding
address.
• Each address of the connection store corresponds to a time slot on the outgoing
highways. Information is read into the speech store cyclically in synchronism with the
incoming PCM systems; however, random access read out is used.

5
 • The connection store has cyclic read-out, but writing is non cyclic.
• To establish connection, the number (X) of time slot of an incoming channels is written
into the connection store at the address corresponding to selected outgoing channel(Y).
• During the cyclic scan of speech store, the incoming PCM sample from channel X
written into address X.
• During each cyclic scan of connection store, the number X read out at the beginning of
the time slot Y.
• This is decoded to select address X of speech store, whose contents are read out and
sent over the outgoing highways.

3.Time switching networks:

3.1 Basic network:

Figure 4: space-time-space (STS) switching network m=number of PCM highways, n=number of time slots

• Figure 4 shows the space-time-space (S-T-S) switching network.

• Each of m incoming PCM highways can be connected to “k” links by cross points in
the A switch and the other ends of the links are connected to “m” outgoing PCM
highways by cross points in the C switch.
• Each link contains a time switch.
• To make the connection between time slots X of an incoming highways and time-slot
Y of an outgoing highways, it is necessary to select a link having a address X free in its
speech store and address Y free in its connection store.
• The time switch is then set to produce a shift from X to Y.
• The connection is completed by operating the appropriate A switch cross point at time
X and appropriate C-switch cross point at time Y in each frame.
6
Time-space- time switch (T-S-T):

Figure 5: Time-Space-Time (T-S-T) switching network m=no. of PCM highways, n=no. of time-slots

• Figure 5 shows T-S-T switching network.

• Each of the m incoming and m outgoing PCM highways is connected to a time
switch. The incoming and outgoing time switches are connected by space switch.
• To make connection between time slot X of an incoming highways and time slot Y of
an outgoing highway, it is necessary to choose a time slot Z which is free, n the
connection store of the incoming highway and the speech store of the outgoing
highway.
• The connection is established by setting timer switch to shift from X to Z, setting
outgoing time switch to shift from Z to Y and operating appropriate cross point at
time Z in each frame.

7
3.2 Bidirectional paths:

Figure 6: Bidirectional transmission through time-division switching network

• The switching network, S-T-S and T-S-T are unidirectional transmission.

• Since PCM transmission systems use four wire circuits, it is necessary to provide
separate paths for the ‘send’ and ‘receive’ channels.
• One way is provide separate switching network for each direction of transmission.
However, this may be avoided by connecting ‘send’ highways of both incoming and
outgoing circuits to one side of the switch and ‘receive’ highways to other side, as
shown in figure 6.
• In an S-T-S network the same speech-store address in the time switch may be used for
each direction of transmission.
• For a connection between time-slot X on one trunk and channel Y on another, for one
direction of transmission, the contents at address are written at the end of time slot X
and at the beginning of time slot y.
• For the opposite direction of transmission, they are written at the end of same time-slot
Y and read at the beginning of next time-slot X.
• In T-S-T network, speech in the two directions must be carried through space switch
using different timeslots.
• In order to simplify through space switch using different time-slot for the two-directions
of transmission have a fixed time difference.
• Usually, the time-slots have phase difference of 180 degrees.
• In a 32 channels system, if time slot 12 is used for one direction of transmission, then
time slot (12+16) =28 is used for reverse direction.

8
3.3 Concentrators:

• Concentrators connect to a PCM highway a number of customer’s line units greater

than number of timeslots on the highway.
• In a simple concentrator, the customer codec are all connected to the common highway
and each may use any timeslot.
• A codec is operated in the required time slot by means of a connection store. This
method is used in the AXE system. Alternatively, a group of codec’s equal to the
number of available time-slots (e.g. 24 or 30) uses fixed channels times on a highway.
• The control function of concentrator may be enhanced to enable it to connect calls
between its own customers if the PCM link fails.
• Facilities must be added to receive and analyse address signals, generate tones and
make cross switch connections between customer’s lines. This unit is then known as a
remote switching unit.
• Mark 1 Digital Switching Subsystem DSSS of system X is Shown in figure 7.

Figure 7: Mark 1 digital Switching System of System X. DLT= Digital Line Terminating unit.

9
3.4 PBX switches:

Figure 8: Trunking of a digital PBX

• A large PBX may use a switching network similar to public exchange.

• A small PBX may only generate sufficient traffic for all its connection to be made over
a single highway.
• All its ports i.e. those for extension lines, exchange lines and the operator’s position,
have codecs connected to a common highway, as shown in figure 8.
• The codecs are operated in the required timeslots by a connection store. In order to
increase line capacity of PBX, the number of timeslots on the common highways may
be increased by using 8-bit parallel transmission instead of serial transmission.
• Two time slots are used to provide two- way communications over same highway.

3.5 Digital cross connect units:

• For a telephone call, a connection is made through a digital switching network at the
start of a call and cleared down as soon as the call ends.
• However, a similar digital switching network may be used for semi-permanent
connections. It is controlled manually from an operating terminal instead of
automatically by processor of exchange.
• Such digital switching network is called Digital cross connect units. It performs a
function for digital circuits similar to that of a distribution frame for analog circuits.
• It is some time called a ‘slow switch’, in contrast to a ‘fast switch’ used to connect
telephone calls.

10
 • Two functions that can be performed by digital cross connect units are grooming and
consolidation.
• In grooming, 64 k bit/s channels on a common PCM dearer are separated for routing
to different destinations.
• For example, a line from a customer’s PBX may carry a mixture of PCM channels,
some to the public and some to the public exchange and some to other PBXs in the
customer’s private network.
• In consolidation, channels onto a smaller number of bearers, thereby improving the
utilization of PCM systems.

4. Synchronization.:

4.1 Frame alignment:

Figure 8: Frame alignments of PCM signals entering a digital exchange

• For correct operation of a time division switching network the PCM frames on all the
incoming highways must be exactly aligned.
• However, since incoming PCM junctions come from different places, their signals are
subjected to different delays.
• To solve this problem, the line terminating unit of a PCM junctions store the incoming
digits in a frame alignment buffer as shown in figure 8.
• Digits are read into this buffer at a rate fa of incoming line, beginning at the start of
PCM frame of the exchange.
• A frame alignment buffer caters perfectly constant misalignment.
• The fill of buffer is constant and its level depends on the phase difference between
incoming line system and the exchange.

11
 • It will also cope with a misalignment that changes slowly between limits (e.g. due to
temperature changes in cables)

4.2 Synchronization network:

• In a Synchronous digital network just one or two atomic reference clocks control the
frequencies of clocks, of all exchanges in the network. This is called despotic control.
• Synchronizing links may be unilateral or bilateral. In the first case, there is master-slave
relationship; the clock frequency of the exchange influences the frequency of the other.
• For this purpose, synchronizing network is added to the PSTN in order to link the
exchange clocks to the national reference standard.
• The local clock in each exchange is provided by a crystal oscillator whose frequency
can be adjusted by a control voltage.
• This control voltage is derived from the incoming digit stream on a synchronizing link,
which is used to determine whether the exchange clock rate should be increased,
decreased or left unchanged.
• This ensures that exchanges maintain the same long - term average frequency, although
short term variation may occur. This is known as mesochronous working.
• clock frequency of the exchange influences the frequency of the other.
• In the second case, there is a mutual relationship; each exchange influences the
frequency of the other.

The principles of these methods are shown in figure 9.

• A unilateral sync system is shown in figure 9(a) exchange A is the’ master ‘ and
exchange B is the ‘slave ’.
• Exchange B determines the phase difference between its own clock and that of the
exchange A by fill of the aligner buffer on the incoming link.
• If there is a more than one sync network link into exchange B, its correction is based
on a majority decision.
• In a single ended bilateral sync link, as shown in figure 9(b).
• The above decision process is made at each end of the link. As a result, both exchange
clocks achieve the same average frequency.
• In a mesh of such synchronous nodes, the exchange would mutually agree on a common
frequency without being controlled by an overall master clock.

12
 • A disadvantage of single ended unilateral and bilateral synchronous system is that phase
comparators are unable to distinguish between phase changes due to frequency drift
and those due to changes in propagation time.
• This disadvantage of single ended unilateral and bilateral synchronous system is
overcome by double ended system as shown in figure 9(c) and (d).
• Synchronizing links may be unilateral or bilateral. In the first case, there is master-slave
relationship; the This eliminates the influence of propagation-delay variations by
subtracting the change in phase determined at the end of link from that determined at
the other end.

Let the phase error detected at exchange A be (A-B) + T, where (A-B) is the phase
change due to discrepancy between the clocks and T is that due to a change in propagation
time.

Then the phase change detected at exchange B is (B-A) + T. Since (B-A) =-(A-
B),

the difference between the two measurements is 2(A-B) and the T is eliminated.

A signalling channel is required to carry the result of phase comparison to the other end of link
in order to make subtraction.

For a unilateral network, as shown in figure 9(c), this channel is needed in only one direction.
For a bilateral link, a signalling channel is needed in each direction as shown in figure 9(d).

13
 Figure 9: Exchange synchronization system

A synchronizing network for an integrated Digital Network (IDN) is shown in figure 10.

• Since this auxiliary network must link all exchanges in the IDN.
• The synchronous network has same nodes and the same hierarchical structure as the
parent public switch telephone network (PSTN).
• The synchronous links are provided by PCM systems that carry normal traffic between
the exchanges.

14
Figure 10 synchronization hierarchy of an integrated digital network.

15