ne

h a
Telephony -
a m
TelecommunicationTSwitching

ita
m
System

hu
ad
M
1
 ne
Simple telephone communication

h a
am
L

T
V

a
Microphone

t
Earphone

i
u m
□ One way communication (Simplex)

d h
□ Microphones and earphones are transducer.

a electrical signals.
□ Carbon microphones –
■ Do not give high fidelity signals

M
■ Gives strong
■ Acceptable quality
 ne
Simple telephone communication

h a
□ Microphone:

am
□ Microphone has carbon granules in a box.

T
□ One side fixed, other attached to diaphragm.

a
□ Resistance inversely proportional to density of granules.

it
□ Diaphragm vibrates with sound and resistance changes.

m
□ V applied across box.

r = max deviationh
u
□ ri = ro – r sin wt

d
ro = resistance without sound

a
□
□ in resistance.

M
□ r = instantaneous
i resistance
□ i = V/ {r – r sin wt}
o
 ne
Simple telephone communication

h a
m
□ i = V/ [ro {1 – (r/ ro) sin wt}

a
□ i = Io( 1 – m sin wt)-1

T
□ i = Io( 1 + m sin wt + m2 sin2 wt + m3 sin3 wt + …)

Carbon microphone acts i t a

□ m < 1.

m should be small to m
□ i = Io( 1 + m sin wt )

h u
□ as amplitude modulator.

d
□ avoid harmonic distortion.

a
□ Energizing current I (Quiescent current) is must.
o

M
 n e
Simple telephone communication

h a
Permits DC to flow from microphone m
□ Inductor :

a
□ Acts as high impedance element for voice.

Voice goes from microphone toT

□ and speaker.

t a
speaker .

i
□

u m
d h
a
M
 ne
Simple telephone communication

h a
m
□ Earphone:

Air gap between diaphragm and poles.a

□ Converts electrical to voice signal.

T
□ Electro magnate with magnetic diaphragm.

t a
□

i
m
□ Voice current through electro magnet exerts variable

u
force on diaphragm.

d h
□ Diaphragm vibrates and produces sound.

a
□ Condition for faithful reproduction:
Diaphragm displacement in one direction only.

M
□
□ Quiescent current provides this bias.
 ne
Simple telephone communication

h a
Instantaneous flux linking poles of electromagnet and

m
□

a
diaphragm:

φ = Constant flux due to quiescentT

□ φi = φo + φ sin wt

φ = instantaneous flux ta
i
□ o current

φ = max amplitudem
u
□ i

d h
□ of flux variation

a of diaphragm has little effect on air gap

□ Assuming

M
■ vibration
■ Reluctance of magnetic path is constant.
 ne
Simple telephone communication

h a
□ Instantaneous Force exerted on diaphragm is

a m
proportional to square of instantaneous flux.

T
□ F = K(φo + φ sin wt)2

t a
□ φ/ φ << 1

F = K φ (1 + K I sin wt)i
o

m
□ Expanding and neglecting second order terms..

h u
2
□ o 1 o

a d
□ Force exerted proportional to input voice signal.

M
 ne
Half Duplex telephone communication

h a
am
L Microphone

T
Microphone

t a
i
Earphone Earphone

An entity either sendsm

u
□ Signal travels in both directions but not simultaneously.

d h
□ or receives signal.

a
□ Speech of A is heard by B as well as A’s own earphone.
Audio signal heard by self earphone is called sidetone.

M
□
□ No sidetone: User tends to shout.
□ Too much sidetone: User tends to speak in too low volume.
□ Here entire speech intensity is heard as sidetone. Not
Desirable.
 e
Half Duplex circuit with Sidetone

an
h
□ At Transmitter:

m
Earphone connected through L L L a
□ ZL: Receiver load

T
□ ZB: Balancing load.

t a
□ .1 2 3

i
Transmitter current I reaches receiver.

m
□ 2

u
□ L very slightly different from L .

Currents divideh
1 2

d
□ Transmitter currents I and I in opposite direction.

a
1 2

□ in L and L such that very small resultant

1 2

M
field results.
□ Very small current induces in earpiece L . 3

□ Small sidetone.

10
 ne
h a
am
Half Duplex circuit with Sidetone

T
t a
i
□ At Receiver:

direction inducing m
u
□ Received current flows through L and L in same

h
1 2

d
additive field.

a signal in earphone.
□ Additive signal induces in L .
3

M
□ Strong received

11
 ne
h a
am
Local battery exchange

T
t a
i
um
□ Local battery installed at each telephone set.

dh
□ DC supplied to transmitter. Magneto is for signaling.

a
□ Switch brings magneto in circuit when required.
□ Ringer has high impedance, bridged across lines.

M
□ At “off hook”, switch closes, DC flows through Tr.
□ Sound waves striking Tr diaphragm produces pulsating current
through primary of induction coil ,inducing AC in secondary circuit.
□ Corresponding AC flows through line reproducing sound at remote
receiver.
 ne
Local battery exchange

h a
a m
Induction Coil/Transformer

T
□ Transformer separates transmitter and receiver ckts.

t a
□ Prevents DC of Tr to flow through receiver.

i of transmitter with line.

m
□ Transformer may step-up voltage on line.

h u
□ Coil matches impedance

d
□ Even one-to-one transformer will greatly increase percentage

a
change in resistance improving useful AC.

M
□ Capacitor is connected when number of LB sets are on same
line.
□ This ‘Sure-ring-condenser’ prevents off-hook receiver from
shunting low frequency ringing current because of high
reactance.
 ne
Central battery exchange

h a
am
T
t a
i
um
dh
a
M
□ Exchange supplies power to all phones from large
rechargeable central battery bank at exchange.
□ Subscriber lines terminated on jack mounted on switchboard.
□ One jack with light indicator for every subscriber line.
 ne
Central battery exchange

h a
□ As subscriber lifts handset, off-hook switch is closed

a m
causing current to flow through handset and lamp relay

T
coil.

a to subscriber through
Lamp relay operates .

t
□

i
□ Indicator corresponding to subscriber lights up.

u m
□ Operator establishes connect

d h
headset key and plug-ended cord pair.

a
□ Cord pair has two cords connected internally and
terminated with a plug each at external ends.

ToM
□ Plug mates with jack.
□ establishing contact, cord is plugged into subscriber
jack and keys corresponding to chosen cord is thrown in
position to connect headset.
 ne
Central battery exchange

h a
□ On verification that called number is free, operator

a m
sends ringing current using plug-ended cord pair.

T
□ Bell B with capacitor C are always connected to circuit.

t a
□ Capacitor allows AC ringing current from exchange to

i party is informed.
m
bell but prevents the loop direct current.

u
If called party busy, called

h
□

d
□ If called party answers, his indicator lamp lights up.

a party jack.
□ Operator connects both parties by plugging in cord

M switching, and releases connection after

pair to called
□ In manual exchange, operator enables signaling system,
performs
conversion.
 Signaling functions: establishing, maintaining andn
e
Signaling Tones-Automatic exchange

h a
Done using tones in automatic switchingm
□

a
releasing telephone conversations.

T
□ systems.

t a
□ Subscriber related signaling functions:

i
m
1. Respond to calling subscriber to obtain identification of

u
called party.

d h
2. Inform calling subscriber that call is being established.

a
3. Ring bell of called party.

M
4. Inform calling subscriber that called party is busy.
5. Inform calling subscriber that called party is
unobtainable.
 n e
Signaling Tones

33 Hz or 50 Hz or 400Hz(modulated with 25 Hz a
h
□ Dial tone: Exchange ready to accept dialed number.

a m
□ or 50 Hz)

Ringing tone sent to called party. T

continuous tone.

a
□ Ringing tone:

t
i 0.2s. Two sets separated by 2s.
□

Frequency is 133hz orm

Indicated to calling party by two short burst tones in a set

u
□

h
for 0.4s each separated by

a d
□ 400Hz.Busy Tone: burst width and gap
width both are same. 0.75s or o.375s

M
□ Number unobtainable:
□ 400 Hz continuous tone
□ Call-in-progress:
□ Burst duration 2.5s and off period of 0.5s.
□ Frequency 400 or 800Hz.
 n e
STROWGER SWITCHING

h a
□ Disadvantages:

Moving parts and contacts subject tom

a wear and tear.
□ Dependence on moving parts and contacts.

Tmaintenance.
□

a speedily accessible
Must be located at easilytand
i
□ Selector switches require regular

u m
□

h
locations.

switching. d
a
□ Problems in achieving above led to Crossbar

M
 ne
CROSSBAR SWITCHING


a
Principles of Common Control

h
□ Directorless system: Example

am
T
□
F

a
E

it
J

um
G

h
D

ad
C H

M
B I

A
 ne
CROSSBAR SWITCHING


a
Principles of Common Control

h
□ A to F – Two routes possible

m
Route 1 – A-B-C-J-F

a
■

T
■ Route 2 – A-I-H-G-F

t a
i
□ All outlets are numbered to identify the paths.

m
□ From EX OUTLET To EX

hu
□ A 01 B

d
□ A 02 I

a
□ B 04 C
□ C 03 J

M
□ I 05 H
□ H 01 G
□ G 02 F
□ J 01 F
 ne
CROSSBAR SWITCHING


a
Principles of Common Control

h
□ Phone number of F for A to call can be at least 4

m
types. e. g.

a
T
□ 02-05-01-02 A-I-H-G-F

a
□ 01-04-03-01 A-B-C-J-F

it
□ DIFFICULTIES:

User must know them

u
□ ID no. of subscriber is route dependent.

on exchanged
h
□ topology and outlet number.

a
□ Number and its size for a subscriber vary depending
from which call originated.

M
 ne
REMEDY: DIRECTOR SYSTEM

h a
□ Routing done by exchange.

a m
□ Uniform numbering scheme.

T
□ Number has two parts-

a
■ Exchange identifier

t
i and store the digits dialed.
m
■ Subscriber line identifier.

h u
□ Exchange must receive

d
a network.
□ Translate exchange identifier into routing digits.
□ Transmit routing and subscriber line identifier digits

M
to the switching
 e
Advantage of director-

an
Features of Common control system

h
□ Soon after translator digits are transferred,

a m
director free to process another call.

T
□ Not involved in maintaining the circuit for

a
conversation.

it
□ Call processing takes place independent of switching

u m
network.

d h
□ User assigned a logical number independent of

a
physical number used for establishing call

M
□ Logical address translated to actual physical address
for connection establishment by address translation
mechanism.
 ne
Control functions in Switching system

h a
m
□ Four broad categories.

a
T
□ Event monitoring

a
□ Call processing

i t
m
□ Charging

h u
□ Operation and maintenance

a d
M
 ne
Calling
Common control switching system

a
Called
subscriber

h
subscriber

m
Line unit Switching
Line unit

a
network

T
Register

t a
finder Common control

i
subsystem

m
Call processing

u
subsystem

dh
Operation

a
Digital receiver Charging
Register circuit control
And storage
sender
register

M Initial
translator
Final
translator
Maintenance
circuits
Event
monitor
 ne
Control subsystem- function I

h a
□ Event Monitoring

a m
□ Events occurring outside exchange are monitored by

T
control subsystem

t a
□ Where-at line units, trunk junctures and inter

i
m signals at line units.
exchange signaling receiver/sender units.

Call request, call u

ofh
□ Events-

Occurrenced
a
□ release
□ events signalled by relays.

M
 ne
Control subsystem

h a
□ Off-hook-

am
□ Event sensed,

T
□ calling location determined,

t a
i
□ free register seized

m
□ Identity of caller is used to determine line

Appropriate dialu
h
category (pulse/tone), class of service.

a d
□ tone sent to caller.
Waits for dialled number.

M to identify exchange.
□
□ Initial digits received and sent to initial
translator
 n e
Off-hook-contd.

Initial translator determines route for calla

h
□ Remaining digits received.

m
aof service as----.
□ through

T
network.

t a
□ Puts through call depending on class

i overload only priority

call subscribers arem
□ Call barring – STD, ISD

Call charging –h
u
□ Call priority – during network

d
put through.

No dialinga
□ Various schemes available.

M
□ calls – hot-line connections.
□ Origin based routing -Emergency call routed to
nearest emergency call center.
□ Faulty line – alternate route chosen
 ne
Off-hook-contd.

h a
□ Initial translator also called office code translator

m
or decoder marker.

a
T routing digits,
□ ‘marker’ because desired terminals were ‘marked’ by
applying electrical signals.

passes to register sender.a

i t
□ Out-of-exchange calls- IT generates

external exchange. m
h u
□ Added to subscriber identification digits and sent to

□
a d
Within-exchange calls – final translator converts

M
subscriber identification digits to equipment number
called.
□ All above can be done by single translator also.
 n e
Control subsystem- function II

a
h by
binary data, defining the path. am
□ Controlling operations of switching network
□ Marks switching elements to be connected

Commands actual connection T

t a
i
□ of the path.

u m
□ Path finding done

d h
□ At Common control unit – map-in-memory

a
□ Or at switching network – map-in-network.

M
 ne
Control subsystem- function II

h a
□ map-in-memory –control unit supplies complete

a m
data defining the path.

Tdetermined by
□ Done in Stored program control.

t a
map-in-network – Actual path

i inlet-outlet to be
□

m
switching network.

h u
□ Control unit only marks

a d
connected.
□ More common in crossbar exchanges.

M
 n e
Control subsystem- function III

a
h into
□ Administration of telephone exchange-

a m
□ Putting new subscriber lines and trunks

Modifying subscriber serviceT

service.

t a on N/W status.
entitlements

i
□

m
□ Charging routing plans based

h u
ad
M
 n e
Control subsystem- function IV

h a
□ Maintenance of telephone exchange-

a m
□ Supervision of proper functioning of the

T of different
exchange equipment, subscriber lines and trunks.

t a
Performs tests and measurements

i elaborate testing.
□

m
line parameters.

h u
□ Aids Fault tracing without

a d
M
 n e
TOUCH-TONE DIAL TELEPHONE

Takes 12 seconds to dial a 7 digit number. a

h
□ Disadvantages of rotary dial telephone:

m
aexchange can not
□

T
□ Faster dialing rate not available.

a of call.
Step-by-step switching of strowger

t
□

i
respond to more than 10-12 pulses/s.

Pulse dialing limitedm

and exchange. u
□ Exchange tied-up for duration

d h
□ to signaling between subscriber

possible. a
□ No end-to-end (subscriber-subscriber) signaling

□
M
Limited to 10 distinct signals.
 ne
TOUCH-TONE DIAL TELEPHONE

h a
□ Advantages of Touch Tone telephone:

am
□ Faster dialing rate feasible.

T
□ Common equipment not tied-up for the duration

t a
of the call.

i
m
□ End to end signaling feasible using voice

h u
frequency bands.

a d
□ Higher number of signaling capability.

M
□ More convenient method of signaling, using push
button keyboard.
 ne
TOUCH-TONE DIAL TELEPHONE

h a
m
1209 1336 1477

a
697 1 2 3

T6
t a
i
4 5

m
770

hu
852 7
ad 8 9

941 M
* 0 #
 n e
TOUCH-TONE DIAL TELEPHONE

h a
□ Touching a button generates a tone.

Called Lower band and upper band m

a
□ Each tone is a combination of 2 frequencies.

T
□ frequencies.

Speech signals may be t a for touch tone

i
□ PROBLEMS:

signals – talk-off. m
u
□ mistaken

d h
a may interfere with touch tone
□ unwanted control actions may occur.

M
□ Speech signals
signaling attempted together.
 ne
TOUCH-TONE DIAL TELEPHONE –Design considerations

h a
□ Protection against talk-off

am
□ Choice of codes

T
□ Band separation

t a
i
□ Choice of frequencies

u m
□ Choice of power levels

d h
□ Signaling durations

a
□ Human factors and mechanical aspects

M
 n e
TOUCH-TONE DIAL TELEPHONE –Design considerations

h a
□ Choice of codes:

m
a to easy
□ Imitation of code signals by speech and music should

imitation as occurring in speechT

be difficult.

t a
□ Single frequency structures are prone

i
and music.

Done by selecting Nm
u
□ Multi frequency code required.

d h
□ frequencies

a
□ Tested for presence/absence.

M
□ 2 combinations
N
using N frequencies .
□ Avoid single frequency combinations.
 ne
TOUCH-TONE DIAL TELEPHONE –Design considerations- Choice of codes

h a
□ Number of frequencies to be transmitted

am
simultaneously should be small to save BW.

T
□ Advantageous to keep fixed number of frequencies to

a
be transmitted simultaneously.

i t
□ Hence P-out-of-N code.

Old multi-frequencym
u
□ P frequencies at a time, out of N.

Inadequate d
h
□ key pulsing (MFKP) with 2/6

a
code gave talk-off less than 1/5000.
□ for subscriber level signaling .

M
 ne
a
TOUCH-TONE DIAL TELEPHONE –Design considerations- Choice of codes

h
□ Hence

a m
□ P is 2 and N s 7 or 8 depending on requirement.

One from lower and one fromT

a
□ Frequencies divided into 2 bands.

t
i spaced frequencies.
□ upper band chosen.

u m
□ Speech contains closely

d h
□ Codes can not be confused with speech.

a
□ Band separation reduces this probability.

M
 n e
TOUCH-TONE DIAL TELEPHONE –Design considerations- Choice of codes

N1 and N2 are number of frequencies in h

a
□ Number of valid combinations = N1 X N2

a m
□ lower and

T
upper band.

t a
□ With 7 frequencies ( 4:3) 12 distinct signals by push

i 16 distinct signals by push

m
buttons.

h u
□ With 8 frequencies ( 4:4)

d
buttons. –Special applications only.

aMulti-frequency Frequency.
□ Hence Called DTMF

M
□ Dual Tone
 n e
TOUCH-TONE DIAL TELEPHONE –Design considerations- Band separation

At receivers, band separations can beh

a
m
□ Advantages of band separations:

a
□ done first

Tbe amplitude
to ease frequency determination.

regulated separately. ta
i
□ Each frequency component can

u m
d h
□ Speech interference can be reduced by using

a
extreme filters for each frequency.

M
 ne
S1 D1 LBF1
Receiver

h a
m
S2 D2 LBF2

a
LA

T
S3 D3 LBF3

t a
i
S4 D4 LBF4

m
Band

u
Separation

h
filter

d
S5 D5 HBF1

a
S6 D6 HBF2

M
LB
S7 D7 HBF3
L – Limiter
S- selector circuit
D – detector S8 D8 HBF4
 After band filter, only one valid frequencyn
e
Receiver

h a each

m
□

a valid
side.

T frequencies.
□ If mixed, limiter receives one strong

t a and further
i
frequency and other invalid weak

u m
□ Limiter peaks strong signal

h
attenuates weak signal.

d
a full o/p and neither signal dominates.
□ If both signals have same strength, limiter o/p is

M
much below
 ne
Choice of Frequency

h a
□ Choice of frequency for touch tone signaling

am
depends on-

T
□ Attenuation characteristics

t a
i
□ Delay distortion characteristics

u m
□ In band 300hz to 3400Hz.

d h
□ Required-

a
□ A flat amplitude response with very low

M
attenuation.
□ A uniform delay response with low relative delay
values.
 ne
Choice of Frequency

h a
m
Delay (ms)

a
Attenuation (dB)

T
4

a
4

it
3 3

um
2 2

dh
1

a
1

1 2 3 4 f (KHz)

M
1 2 3 4 f (KHz)

Best choice- 700 Hz to 2200 Hz

Actual range – 700 Hz to 1700 Hz
 ne
Choice of Frequency

h a
□ Actual range – 700 Hz to 1700 Hz

a m
□ Spacing depends on detection accuracy.

T are to be
□ Minimum spacing chosen more than 4%.

t a
i
□ 1:2 or 2:3 such harmonic relationship

m
avoided-

h u
□ between two adjacent frequencies of same band.

a d
□ between pairs of frequencies in different bands.
□ Improves talk-off performance.

M
□ Chosen frequencies almost remove talk-off.
 n e
Signal power

h a
□ Only two frequencies.

a m
□ Hence signal power can be as large as possible.

T
□ 1dB above 1mW nominal value.

Worst case attenuation a

t
Attenuation increases with frequency.

i
□

Hence upper bandm

□ in 697-1633 can be 4dB.

h u
frequencies powers are 3dB

d for output power are –

□

Nominal a
higher than lower band frequencies.

M
□ values
□ Lower band power = -3.5dBm
□ Higher band power = - 0.5dBm
 n e
Signaling Duration

check for presence of a frequency is h

a
m
□ The probability of talk-off can be reduced if

a
done for a

T button pressed
longer time.

t a
i
□ This requires subscriber to keep

But with efficientm

u
for long time than normal.

durations canh
d
□ circuit designs, lower

a pauses for 200ms between digits.

be tested.

M
□ Fast dialer
□ In normal practice tone duration 160ms and inter
digit gap 350ms followed.
 Data-in-voice-answer is a major advantage ofn
e
DIVA – an advantage

to-end signaling using touch tone dialing.h

a
m
□ end-

a
T where operator
□ Examples-

t a
i
□ Fault lodging in telephone services

m
sends voice message and user sends digits

h u
corresponding to answers.

d
a to operator’s voice message.
□ Airline and railways services where user dials digits
to opt for various services (information, reservation)

M
in response
□ Best example of dialing and voice conversation
together.
 e
STORED PROGRAM CONTROL

an
h
□ Program or set of instructions to the computer are

a m
stored in its memory.

T
□ Instructions executed automatically one by one by

a
the processor.

it
Programs are for Telephone exchange switching

Hence called SPC. m

□

u
control functions.

d h
□

a
M
53
 e
FEATURES

an
h
□ Full scale automation of exchange functions.

am
□ Common channel signaling

T
□ Centralized maintenance

t a
i
□ Automatic fault diagnosis.

m
□ Interactive human machine interface.

h u
Telephone d
!

a
□ REQUIREMENTS FROM COMPUTER-
exchange must operate without

Mto come.
□
interruption 24 hrs a day, 365 days a year. for
years
□ And hence the computers.
54
 e
Centralised SPC

an
h
□ Control equipments must be replaced by a single

a m
powerful computer.

T for redundancy.
□ Must be capable of processing more than 100 calls

ato all exchange

per second along with other tasks.

t
i programs.
□ May use more than one processor

Each processor u
m
□ Each processor has access

functions. dh
resources and function

a
□ capable of executing all control

M
55
 e
Centralised SPC – no redundancy

an
h
▪

m
To lines From lines

a
T
Signal

a
Scanners

t
Distributor

i
um
h
Processors

ad Secondary

M
Storage: Call
Maintenance Recording, Program
Memory Storage etc
Console

56
 e
Centralised SPC – With redundancy

an
h
m
!

a
R1 R2 Rt Resources
!

T
!

t a
P1 P2 Pp Processors

i
!

m
Function

u
! programs

h
F1 F2 Ft

d
!

a
□ Redundancy at the level of processors, exchange
resources and function programs .

M
57
 ne
h a
□ Practically, resources and memory modules are

am
shared by processors.

T
□ Each processors may have dedicated path to

t a
exchange resources.

i
m
□ Each processors may have its own copy of

h u
programs and data in dedicated memory

a d
modules.
□ Two Processor configuration is most common.

M
58
 e
Modes of Dual Processor Architecture

an
h
□ Standby mode

am
□ Synchronous mode

T
□ Load sharing mode

t a
i
um
dh
a
M
59
 e
Standby mode


an
h
am
Exchange

T
Environment

t a
i
um
dh
P1 P2

a
M Secondary Storage

60
 e
Standby Mode

n
□H/W and S/W of one processor are active. a
h
□Simplest

a m
T
□Other is standby.

a
□Standby processor shouldtbe able to reconstitute the
□Standby processor is brought to line only when active

i
processor fails.

u m
h
d connected through the network.
state of exchange system during takeover.

■Whicha
■Which subscriber or trunk are busy or free.

M
paths

61
 e
Standby Mode

n
■By scanning all status signals during takeover.a
h
□ Small exchanges:

a m
T
■Only the calls being established at time of failure are

a signals within reasonable

disturbed.

i t
□Large exchanges:

■Active processoru
m
■Not possible to scan all status

periodically d
h
time.

a updates are taken by standby at takeover.

copies the status of the system
into secondary storage.

M
■Most recent
■All calls which changed status after last updates are
disturbed.
62
 e
Synchronous duplex mode

an
h
am
Exchange

T
Environment

t a
i
um
dh
P1 C P2

a
M
M1 M2

63
 e
Synchronous duplex mode

an
h
□ Both processors execute same instructions.

am
□ Results compared continuously.

Tin its memory.

□ During fault, comparator results mismatch.

t a
icontrols .
□ Each processor have same data

m
□ Each receive same information from exchange.

h u
□ One processor actually

a d
□ Other synchronises but does not participate.

M
64
 e
Synchronous duplex mode

an
h
□ During fault:

a m
□ P1 & P2 decoupled

T
□ Run checkout program in each machine.

t a
Call processing suspended temporarily without

Good processor takes i

□

m
disturbing the current call.

h u
□ control.

d
a is enabled.
□ Once repaired, other processor copies contents of
active processor in its memory.

M
□ Comparator

65
 e
Load Sharing mode

an
h
am
Exchange

T
Environment

t a
i
um
dh
P1 P2

a
M
ED

M1 Exclusion Device M2

66
 e
Load Sharing mode

a n
h
□ Both processors are active simultaneously.

a m
□ Both share the load and resources dynamically.

T
□ Both processors have access to entire exchange.

t a
Incoming call is assigned randomly

i
□

m
□ or in a predetermined order to one of the

completion. h
u
processors.

d
aseparate memories for storing temporary
□ Assigned processor handles the call through

M
□ Both have
call data .

67
 e
Load Sharing mode

an
h
□ Both are in mutual coordination through inter

am
processor link.

T
□ If information exchange fails, healthy processor

a
takes over.

i t
Exclusion devise prevents both to be active

m
□

u
together.

d h
▪ Current calls are transferred.

a
□ Calls being established are lost.

M
68
 e
Load Sharing mode

an
h
□ Traffic sharing depends on the conditions of the

a m
processors and their requirements.

T
□ During testing on one, other can take more traffic.

t a
□ Gives much better performance during traffic

i
overloads.

u m
It’s a step towards distribution control.

h
□

a d
M
69
 ne
Availability of the single processor system


h a
□ Main purpose of redundant configuration is to

a m
increase availability.

T
□ Availability of single processor:

t a
□ A = MTBF / (MTBF + MTTR)

MTTR = mean time to i

m
□ MTBF = mean time between failure

Unavailability U u
/h
□ repair

= 1- {MTBFd
MTTR /a
□ = 1-A
□ (MTBF + MTTR)}

M
□ (MTBF + MTTR)
□ If MTBF>>MTTR
□ U = MTTR / MTBF
70
 ne
Availability of the Dual processor system


h a
□A dual processor is said to have failed only when both

a m
the processors fail.

T
□System is totally unavailable.

t a
□Condition – One processor has failed.

i second fails during MTTR

m
■ Other also fails before first is repaired.

u
h can be given in terms of
□Conditional probability that

□MTBF of duald
a of single processors as-
of first.
processor

M
MTBF and MTTR
□ MTBF = (MTBF) / 2MTTR -using conditional
D
2

probability.
71
 ne
Availability of the Dual processor system


h a
□ AD = MTBFD/ (MTBFD+ MTTR)

am ]
/ [(MTBF) + 2(MTTR)
T
□ AD = (MTBF)2 2 2

t a+ 2(MTTR) ]
□ UD = 1- AD

i
m
/ [(MTBF)
u
□ = 2(MTTR)2 2 2

d h
□ If MTBF>>MTTR

□
a
U = 2(MTTR) / (MTBF)
2 2

M
D

72
 ne
Assignment

h a
□ Given that MTBF = 2000 hours and MTTR = 4 hours,

am
calculate the unavailability for single and dual

T
processor system.

t a
!

i
□ U = 4/2000 = 2 X 10-3

u m
525 hours in 30 years.

h
□

d
2.1 hoursa
!
□ U = 2 X 16/2000
D = 8 X 10
-6

M
□ in 30 years

73
 ne
Functions of control subsystem

h a
□ Event monitoring

am
□ Call processing

T
□ Charging

t a
□ Operation and maintenance

i
m
□ Grouped under 3 levels

hu
d
Event monitoring

a
Level 3
and distribution
Real time
constraint

M
increases
Level 2 Call processing

Level 1 O & M and charging

74
 ne
Functions of control subsystem

h a
□ Event monitoring has highest priority, O&M and

a m
then charging the least.

T
□ Real time constraint asks for priority interrupts.

t a
□ If an EVENT occurs during O&M, it will be

i
interrupted and event will be handled.

Nesting interrupt m
u
Then O&M will be resumed.

h
□

d
□ to suspend low level functions and

a
take up higher level functions.

M
75
 ne
Functions of control subsystem

h a
□ Interrupt processing

am
T
a
Level n process

it
Suspend level n

um
h
Take up level n+x

ad
Suspend level n + x

M
Take up level n+x+y
Level n + x + y complete
Resume level n + x
Level n + x complete
Resume level n 76
 ne
Functions of control subsystem

h a
□ When an interrupt occurs, program execution is

am
shifted to an appropriate service routine address in

T
memory through branch operation.

t a
□ Non-vectored interrupt:

i scans interrupt signals

m
□ Branch address fixed.

h u
□ Interrupt service routine

d
and decides on appropriate routine to service.

a
M
77
 ne
Functions of control subsystem

h a
□ Vectored interrupt:

am
□ Branch address not fixed.

T
□ Branch address supplied to processor by

t a
interrupting source.

i directly, without full

m
□ Set of address called interrupt vector.

h u
□ Faster as can be addressed

a d
scanning

M
78
 e
DISTRIBUTED SPC

an
h
□ Control functions shared between many processors

a m
within the exchange.

T
□ Low cost microprocessors offer better availability

a
and reliability than centralised SPC.

i t
Exchange control functions decomposed

m
□

u
horizontally or vertically.

d h
a
M
79
 ne
DISTRIBUTED SPC – Vertical decomposition


h a
□ Exchange divided into blocks.

a m
□ Each block assigned to a processor.

T
□ Performs all control functions related to that block

t a
of equipments.

i
m
□ Total control system consists of several control

u
units coupled together.

redundancy. h
d
□ Processor in each block may be duplicated for

ain any of three dual processor modes as

M
□ Operates
explained earlier.
□ Modular so that more can be added when exchange
is expanded. 80
 ne
DISTRIBUTED SPC – Horizontal decomposition


h a
□ Each processor performs one or some of exchange

a m
control functions.

T
□ Chain of processors for 3 functions.

t a
□ Entire chain may duplicate for redundancy.

i
u m
d h
Event monitoring

a
Level 3
and distribution
Real time
constraint

M
increases
Level 2 Call processing

Level 1 O & M and charging

81
 ne
DISTRIBUTED SPC – Horizontal decomposition


h a
□ Entire chain may duplicate for redundancy.

am
T
Exchange environment

t a
i
um
EM & DP EM & DP Level 3

dh
a
CP CP Level 2

M
O & MP O & MP Level 1

EM & DP-Event monitoring and distribution 82

 ne
DISTRIBUTED SPC – Level 3 processing

h a
□ Handles scanning, distribution and marking

a m
functions.

T
□ Operations simple, specialised and well defined.

t a
□ Sets or senses binary conditions in F/F or registers.

i WORD
m
□ Achieves control by sensing or altering binary

u
conditions using CONTROL

d h
□ Hard wired or micro programmed device.

control. a
□ Compare micro programmed control to Hard wired

M
83
 ne
DISTRIBUTED SPC – Level 3 processing

h a
□ Set of control words stored in memory and read

am
one by one.

T
□ Horizontal control - One bit per every control

t a
signal.

i - depends on number of
m
□ Flexible and fast.

h u
□ Expensive as large width

d
signals.

word. a
□ Vertical control – Each signals binary encoded as a

□
□
M
Time too large as at a time only one signal.
Mid approach chosen.
□ Control word contains group of encoded words. 84
 ne
DISTRIBUTED SPC – Level 2 processing

h a
□Processors for call processing.

am
□Called switching processors.

T
□Instructions designed to allow data to be packed more

t a
tightly in memory without increasing access time.

i
m
□Processor designed to ensure over 99.9% availability,

hu
fault tolerance and security of operation.

d
acontrolled data transfer .
□I/O data transfer order of 100 kilobytes per s.
□I/O technique:

■M
■Program
Direct memory access.

85
 ne
DISTRIBUTED SPC – Level 2 processing

h a
□ Traffic handling capacity of control equipment

a m
limited by capacity of switching processor.

T
□ Load on switching processor measured by occupancy

a
t.

processor is occupied. t
i
Occupancy: Fraction of unit time for which

m
□

h u
d
□ t = a + bN

a
□ a = fixed overhead depending on exchange capacity

M
and configuration
□ b = average time to process one call.
□ N = number of calls per unit time.
86
 ne
DISTRIBUTED SPC – Level 2 processing

h a
a depends on scanning workload which depends on

m
□

a
number of subscriber lines, trunks and service circuits

T execution time per

in exchange.)

a
■a estimated by knowing total lines, instructions

t
i process.
required to scan one line and average

m
instruction.

u
calls etc.. dh
□ b depends on type of call

■Results a
■Incoming call process time less than outgoing or transit

M
of party busy or no answer etc also affect.
■ Type of subscribers (DTMF/rotary dial) also affect as
grouping PBX lines change.
87
 n e
DISTRIBUTED SPC – Level 1 processing (O&M)

h a
□ Administer the exchange H/W and S/W.

a m
□ Add, modify and delete information in translation

T
table.

a
Change subscriber class of service.

t
□

i
□ Put a new trunk or line into operation.

Monitor traffic. m
u
□ Supervise operation of the exchange.

d h
□

a and test programs.

□ Detect and locate fault and errors.

!M
□ Run diagnostic
□ Man-machine interaction.

▪ 88
 ne
DISTRIBUTED SPC – Level 1 processing (O&M)

h a
□ Less subject to real time constraint.

a m
□ Less need for concurrent processing.

Helps diagnosis of many from T

□ Single O&M computer caters to many exchanges.

t a
one location.

i
□

u m
h
Operator Maintenance Personal

a d O&M

M
Computer

Exchange 1 Exchange 2 Exchange P

89
 ne
SINGLE STAGE NETWORKS

h a
□ No of cross points will be 10 x 10 = 100.

a m
□ Fully connected so no blocking.

T
□ Used for 10-25% time on average.

t a
Remains idle. Waste of infrastructure.

i
□

u m
d h
a
10 outputs

M
10 inputs

90
 ne
TWO – STAGE NETWORKS

h a
□ N X N two stage network with K simultaneous

am
connections-

T
!
N N

a
NXK KXN

Full connectivity to K i
t
!

m
!

Blocking after K.u

h
□ simultaneous calls.

Each staged
a
□
□ has NK switching elements.

M elements each stage = N /16

□ Assuming 0nly 10% connectivity K can be N/16.safe
□ Switching 2

□ Total switching elements = N2/8.

91
□ If N = 1024, switching elements = ?
 n e
TWO – STAGE NETWORKS

h a
□ For large N, N X K is unrealizable.

a m
□ Remedy: Using smaller size switching matrices.

T
t a
i
um
d h
a
M
92
 ne
TWO – STAGE NETWORKS

h a
am
p 1 s

T
p 2 s r 1 q

t a
i
m
M inlets N outlets

hu
d
p r-1 s r s q

a
p r s

M
93
 ne
TWO – STAGE NETWORKS

h a
□ M = pr. ( p inlets per p blocks)

am
□ N = qs.

T
□ Full availability: Atleast one outlet from each

t a
block in 1st stage must be connected to inlet of

i =S=
every block in 2nd stage.

S = psr + qrs m
u
No. of Switching elements

h
□

d
□

a
□ S = Ms + Nr

M
□ No of simultaneous calls – switching capacity SC
=
□ SC = sr
□ Condition: 94
 ne
TWO – STAGE NETWORKS

h a
□ These N/W are blocking.

a m
□Under 2 conditions:

T
■If calls are uniformly distributed, (rs + 1)th call arrives.

t a
■Calls are not uniformly distributed.

i in I block is active = β .
m
□Probability that given inlet in I block is active = α .
th

h u
□Probability that given outlet th

block. a
d
□ β is
■inversely proportional to number of outlets in each

M
■Directly proportional to number of inlets in each block.
□ β = p α/s
□Probability that another inlet becomes active and asks
95
 ne
TWO – STAGE NETWORKS

h a
□ Blocking means –

am
□ All outlets are already active, and no free outlets.

T
□ The probability that an already active outlet is sought =

a
= probability that the particular outlet is active AND

t
□

i
□ other outlets are not sought.

um
□ PB = p α/s[1-{(p-1) α/(s-1)}]

d h
□ If p = M/r

a – ((M/r) –1) α} / {rs(s-1)}

!
P α(s-1)

M
□ B = {M

96
 ne
THREE – STAGE NETWORKS

h a
pxs sxp

am
rxr
p 1 s s 1 p

T
t a
p 2 s r 1 q s 2 p

i
um
N N

h
inlets outlets

ad
p r-1 s r s q s r-1 p

M
p r s s r p

97
 ne
THREE – STAGE NETWORKS

h a
□ N inlets = r blocks of p inlets each.

am
□ Same for p outlets.

T
□ Stage 1 --– p x s.

a
□ Stage 2 --– r x r.

it
□ Stage 3 --– s x p.

um
□ No of switching elements = S =

h
□ rps + sr2 + srp

ad
□ = 2Ns + sr2 .
□ = s(2N + r2 ).

M
98
 ne
TIME DIVISION SPACE SWITCHING

h a
m
1 1

a
2 2

T
t a
i
um
h
N-1 N-1

d
N N

a
BUS

M
k – to 2k Modulo – N
decoder counter Clock

Cyclic control

SWITCHING STRUCTURE 99
 ne
TIME DIVISION SPACE SWITCHING

h a
Two – stage equivalent circuit
□

am
!1

T
1
!

t a
i
!

um
!N N

dh
!

a
!

M
!
□ N X 1 and 1 X N switching matrix for 1st and 2nd
stage.
□ 1 interconnecting link.
100

□ Speech in PAM analog - analog time division

 ne
TIME DIVISION SPACE SWITCHING

h a
□ Inlet-outlet pair connected to bus through control mechanism.

am
□ Number of simultaneous conversations SC = 125/ts

T
□ ts is time in µs to set up a connection.

t a
□ Inlet-outlet selection dynamic.

i
□ Simplest is cyclic. ( i connected to i.)

um
□ Hence no switching.

dh
□ Hence lacks full availability.

a
□ Inlet or the outlet control can be memory based to achieve
switching as…

M
101
 ne
Input controlled TIME DIVISION SPACE SWITCHING

h a
□

m
1 1

a
2 2

T
t a
N-1 N-1

i
N N

m
BUS

hu
d
Modulo – N

a
k – to 2k Address
counter- decoder-
cum-MAR decoder
cum-MDR
Cyclic control

M
7
4
1
5

102
 ne
Input controlled TIME DIVISION SPACE SWITCHING

h a
□ Sequence required is stored serially in memory

a m
address register at outlet side.

T
□ Input serially.

t a
□ 7-4-1-5 stored in locations 1, 2, 3 and 4.

i
m
□ Inlet 1 connected to outlet 7……

depending on h
u
□ Full availability.

d
a has N words for N inlets.
□ Called inlet or input controlled as outlet is chosen
inlet being scanned.

M
□ Control memory
□ Width of log N bits. (Stored in binary.)
2

□ Cyclic control means all subscriber scanned whether

active or not. 103
 ne
Input controlled TIME DIVISION SPACE SWITCHING

h a
□ Decoder o/p enables proper outlet to be connected

am
to bus.

T
□Sample signal is passed from inlet to outlet.

a
□Any inlet I can be connected to any outlet k.

it
□Full availability.

■Memory locationm
u
□If inlet inactive-

d h
has null value.

a
■Address decoder does not enable any outlet

M
line.
□Bus–single switching element–time shared by N
connections.
□All can be active simultaneously.
104
□Physical connection established between inlet and
 ne
Output controlled TIME DIVISION SPACE SWITCHING

h a
m
1 1

a
2 2

T
t a
N-1 N-1

i
N N

m
BUS

hu
d
Modulo-N

a
Decoder
Decoder counter
Cyclic control
CLK

M
7
4
1
5

105
 ne
Output controlled TIME DIVISION SPACE SWITCHING

h a
□ Sequence required is stored serially in memory address

am
register at inlet side.

T
□ Output serially.

a
□ 7-4-1-5 stored in locations 1, 2, 3 and 4.

it
□ Outlet 1 connected to inlet 7……

m
□ Full availability.

hu
□ Called outlet or output controlled as inlet is chosen depending

d
on outlet being scanned.

a
□ For active outlet i, inlet address stored in location i.

M
□

106
 ne
Output controlled TIME DIVISION SPACE SWITCHING

h a
□ SC = N = 125/( ti + tm + td + tt)

am
□ ti = Time to increment the modulo-N counter.

T
□ tm = Time to read the control memory

t a
□ td = Time to decode address and select inlet or outlet.

i
m
□ tt = Time to transfer sample value from inlet to outlet.

u
All times in µs.

h
□

d
□ Clock rate 8 X N KHz

a
M
107
 ne
Some more on TIME DIVISION SPACE SWITCHING

h a
□ For two direction data transfer-Two independent buses.

am
□ Simultaneous data transfer on two buses.

T
□ Or single bus with time sharing two directional traffic.

a
□ All lines scanned irrespective of active or inactive.

it
□ Waste as only 20% are active.

um
□ Hence control on memory on both sides more useful.

h
□ Hence memory-controlled time division space switching.

ad
M
108
 ne
Generalised TIME DIVISION SPACE SWITCHING

h a
m
1 1

a
2 2

T
t a
N-1 N-1

i
N N

m
BUS

u
Decoder Decoder

dh
a
MDR Data input
CLK

M
7
Modulo-SC
4
counter
1
5 Read/Write

MAR
109
 ne
Generalised TIME DIVISION SPACE SWITCHING

h a
□ Control memory word has two address.

a m
□ Inlet and out let address.

T
□ Word width is 2[log2N] bits.

t avia data input.

□ Operation:

The Location then m

i
□ Inlet k and outlet j addresses entered into free

u
location of control memory

d h
□ marked busy.

a
□ Modulo – SC counter updated at clock rate.

M
□ Control memory word read out one by one.
□ Addresses are used to connect respective inlet and
outlet.
□ Sample transferred from inlet to outlet.
□ Clock updates counter.
110
 ne
Generalised TIME DIVISION SPACE SWITCHING

h a
□ Busy / free information stored in bit vector.

am
□ 1 bit per location.

T
□ Bit set – busy.

t a
□ SC = 125/ts

i
m
□ Clock rate = 8 SC kHz.

u
Ts = ti + tm + td + tt

h
□

d
□ If is tm dominant, control memory busy throughout 125 µs.

a
□ One write cycle reserved for input purpose in every 125 µs.

M
111
 ne
TIME DIVISION TIME SWITCHING

h a
□ Memory block in place of bus.

a m
□ PCM samples.

T out of memory.
□ Serial data taken in and out.

Serial/ parallel convertera

t
But parallel data written and read

i
□

m
□ at inlet and vice versa.

h u
□ MDR is a single register.

No physicald
a
□ Gating mechanism to connect inlet and outlet.
□ connection between inlet and outlet.

M
□ Information not transferred in real time.
□ Data first stored in memory, then transferred to
outlet.
112
□ Hence called TIME DIVISION TIME SWITCHING.
 ne
TIME DIVISION TIME SWITCHING

h a
MDR
▪

m
In D D O

a
1 g at ut 1
at
a g

T
at a Data
e S/P in memory N o P/S at

a
words of 8 ut e

t
bits each

i
N N

um
h
MAR

ad
Modulo-N
counter
MDR Data

M
in

M Control memory N
A words of log2N bits
R each

113
 ne
TIME DIVISION TIME SWITCHING

h a
□Equivalent circuit-

am
!

T
!1 1

a
NX1 Delay 1XN

t
!N

i
N

m
!

u
□Types:

dh
□Sequential write/random read

a
□Random write/sequential read

M
□Random input/random output
□Inlets and outlets and control memory scanned
sequentially.
□Data memory read/written sequentially/random.
114
□Three forms can operate in any of two modes:
 ne
TIME DIVISION TIME SWITCHING – Phased operation

h a
□ Two phases.

am
□ Sequential write/random read – Phase one –

T
□ Inlets scanned sequentially 1, 2, …N.

a
□ Data stored in Data memory sequentially 1, 2, …N.

it
□ Control memory stores inlet addresses as required by outlets.

um
□ Inlet numbers 5, 7, 2, … for outlets 1, 2, 3, …

h
□ Phase Two –

ad
□ Outlets scanned sequentially 1, 2, 3, ….
□ Data read from data memory randomly 5, 7, 2,….

M
□ Data reading controlled by control memory.

115
 ne
TIME DIVISION TIME SWITCHING – Phased operation

h a
□ First phase - one memory write per inlet (total N)

am
□ Second phase– one control memory read + one data memory

T
read per outlet.

a
□ Total time taken = in µs

it
□ ts = Ntd + N(td + tc)

um
□ td= read/write time for data memory

dh
□ tc= read/write time for control memory

a
□ If td = tc = tm ,
ts = 3N tm

M
□
□ Number of subscribers = N = 125/3 tm
▪

116
 ne
TIME DIVISION TIME SWITCHING – Phased operation

h a
□ Number of subscribers can be increased

am
□ By overlapping read cycle of data memory and

T
control memory.

t a
! Phase 1 Phase 1

i
a
1 a2 N a

m
!

u
DM write DM read
!

dh
!

a
1 2 N
!
CM read

M
!
□ Last cycle of phase 1, memory write coincides with
□ -first location of control memory read having inlet
address.
117
□ Gives out data 1 and reads next control
 ne
TIME DIVISION TIME SWITCHING – slotted operation

h a
□

m
125µs

a
!

T
1 2 N

t a
i
! DM read

m
DM write

u
CM read
!

h
Operationsdin each sub periods:
!

a
□ Sub periods i = 125/N µs.
□

M i.
□ 1. Read inlet i and store data in data memory
location
□ 2. Read location i of control memory and read
118
address say j.
 n e
TIME MULTIPLEXED SPACE SWITCHING

h a
□ Time division switches means:

a m
□ An inlet or an outlet corresponds to single

T
subscriber

Used in local exchanges. a

t
□ with one sample speech appearing every 125 µs.

i means:
m
□

Used in transit u
h
□ Time multiplexed switches

a d
□ exchanges.
□ Inlet and outlets are trunks carrying TDM data.

M
□

119
 ne
TIME MULTIPLEXED SPACE SWITCHING

h a
□

m
M 2 1 M 2 1
1 1

a
2 2

T
t a
i
m
N-1 N-1

hu
N N

d
M 2 1 M 2 1

a
Decoder

M
Cyclic
Control
!
CM
MAR
!
MN words
120
 ne
TIME MULTIPLEXED SPACE SWITCHING

Each carry a TDM stream of M samplesh aframe.

□ N incoming trunks and N outgoing trunks.

a m
□ per

One frame time – MN samples.T

□ Frame time 125 µs.

One time slot = 125 µs a

t
i are switched.
□

m
□

u
scanned. dh
□ One time slot – N samples

a to each outlet, M locations in control

□ Output controlled switch - Output Cyclically

M
□ Corresponding
memory.
□ M blocks of N words each.
□ Two dimensional location address . 121
 ne
TIME MULTIPLEXED SPACE SWITCHING

h a
□ Block address i corresponds to time slot i.

am
□ Word address j corresponds to outlet j.

T
□ First N locations corresponds to first time slot.

a
□ And so on.

it
□ If inlet address k is present in location (i,j)- ( output controlled)

um
□ Means inlet k is connected to outlet j during time slot i.

h
□ Number of trunks supported =

ad
□ N = 125/Mts
□ ts= is switching time including memory access time per inlet-

M
outlet pair.
□ Physical connection provided between inlet and outlet.

122
 ne
TIME MULTIPLEXED SPACE SWITCHING

h a
□ Cost of switches = No of switches + no of memory

a m
words

T
□ = 2N + MN

t a
□ Cost of equivalent single stage switch = (MN) .
2

i
m
!

h u
!

a d
!
▪ ASSIGNMENT:

M
▪ Calculate number of trunks that can be supported
on a time multiplexed space switch, given that
▪ a) 32 channels are multiplexed in each stream.
b) Control memory access time is 100 ns.
123
▪
 ne
SOLUTION

h a
□

am
□ M = 32

T
□ ts = 100 + 100 = 200 ns

t a
!

i
□ N = 125/M ts = 20

um
dh
a
M
124
 n e
TIME MULTIPLEXED TIME SWITCHING

h a
□ Time switch does not give physical connection.

a m
□ Data stored and then transferred during another

T
slot.

t a
□ Delay.

i
m
□ Employs TIME SLOT INTERCHANGER.

h u
!

a d
!
□

M
125
 ne
Time Slot Interchanger

h a
□ Let one incoming trunk and one outgoing trunk.

am
□ M channels multiplexed in 125 microseconds.

T
□ Sequential write / random read

a
□ Time slot duration tTS= 125/M

it
□ MtTS= 125

um
▪

dh
a
M
126
 ne
Time Slot Interchanger

a
frame

h
frame 1 2 3 M O/P slot number
▪

am
M 2 1 38 42 51 19 Control data
memory location

T
t a
1 38

i
2 13 1 1

m
3 76 2 7

u
4 51 3 4

h
5 19 4
6 5

d
26
6

a
DM
CM

M
M 27
M 27

Time slot
counter
127

CTS
 n e
Time Slot Interchanger

h a
□Clock runs at time slot rate.

a m
□Time slot counter incremented by one at end of each

T
slot.

t a
□Counter contents provide

■locations addressesifor control memory .

m
■locations addresses for data memory .

h u
simultaneouslyd
□Content ofa
□Data memory and control memory access
at beginning of time slot.

M data read to output trunk.

CM used as address of data memory.
□Respective

128
 e
Operation is a Time Slot

an
h
▪

am
T
tTS

t a
i
t t

um
Read input data; Read DM;

h
Write into DM; Write data

d
Read CM. to output

a
M
t

129
 n e
Time Slot Interchanger

Data ready for writing on O/P at end ofh a slot.

□ I/P data available to read at beginning of time slot.

a m
□ time

T
□ Storage action.

t a
Hence delay of minimum one time slot even if no

Output delayed by t imicrosecond.

□

m
time slot interchange.

h u
□ TS

a d
M
OS1 OS2

IS1 IS2

0 tTs 2tTs

130
 Delay depends on to which output slot, inputn
e
Time Slot Interchanger

□
h a
m
slot is

a
switched.

TDM address.
□ Previous cycle, all DM is filled/ written in.

t a
ibe switched only in this case.
□ In current cycle, CM is read for

u m
□ CM1 =1, contents of DM1 switched to O/P1.

d h
□ Current contents can

CM2=7, a
□ Delay t microseconds.
TS

M
□ contents of DM7 switched to O/P 2.
□ Delay = [(M-7)+2+1] t = (M-4) t microseconds.
TS TS

□ CM3=4, contents of DM4 switched to O/P 3.

□ Delay = [(M-4)+3+1] tTS = M tTS microseconds.
131
 ne
Time Slot Interchanger

Two sequential memory access per timeh

a
□ We have MtTS= 125

a m
□ slot.

T
□ t =2t

No switching elements. ta
TS m

i
□ 2 t M = 125
m

m
u of CM and DM.
□

M locations inh
d
□ Cost equal to number on memory elements.

a
□ each

M
□ C = 2M units.

132
 ne
Assignment

h a
□ Calculate the maximum access time that can be

am
permitted for the data and control memories in a

T
TSI switch with a single input and output trunk

a
multiplexing 2500 channels. Also estimate cost of

it
the switch and compare it with single stage space

u m
division switch.

h
□ 2 tm M = 125

d
a = 5000 units.
□ t = (125 X 10
m )/(2500
3
X2) = 25ns

M
□ C = 2 X 2500

133
 e
Traffic Engineering

an
h
□ Provides basis for design and analysis of

m
a for which
telecommunication networks.

T
□ Blocking probability is major issue for design.

a
Blocking probability depends on time

t
□

i
following are busy –

u m
□ Subscriber

d h
□ Digit receiver

a
□ Inter stage switching links
Call processors

M
□
□ Trunk between exchanges

134
 e
Traffic Engineering

an
h
□ Traffic pattern on the network varies throughout

am
the day.

T
□ Traffic engineering provides a scientific basis to

a
design cost effective network taking all above into

it
account.

m
□ It helps to determine ability of network to carry a

hu
given traffic at a particular loss probability.

ad
□ Provides a means to determine quantum of common
equipment required to provide a particular level of

M
service for a given traffic pattern and volume.

135
 e
Traffic load and parameters

an
h
□ Typical traffic load of a day

am
Number oh calls

T
In the hour

t a
i
um
dh
a
M
1 13 Hour of the day 24 136
 e
Traffic load and parameters

an
h
□ Traffic pattern varies for domestic and official

a m
areas.

T
□ Varies for working and non-working days.

t a
□ Busy hour- 1 hour interval lying in time interval

i
m
concerned in which traffic is highest (Max call

u
attempts).

d h
□ Peak busy hour- The busy hour each day.

which is a
□ Time consistent busy hour- particular 1 hour period

M
peak busy hour each day over the days
under consideration.

137
 e
Traffic load and parameters

an
h
□ CCR Call Completion Rate – ratio of number of

a m
successful calls to number of call attempts.

Designed to provide overall T

□ Used in dimensioning the network capacity.

CCR=0.75 considered t a
i
□ CCR of 0.70.

Higher CCR is notm

u
□ excellent.

d h
□ cost effective.

a
M
138
 e
Traffic load and parameters

an
h
□ BHCA Busy Hour Call Attempts – Number of call

am
attempts during busy hour.

T
□ It is an important parameter in deciding the

a
processing capacity of common control or stored

it
program control in an exchange.

m
□ Busy hour calling Rate – average number of calls

h u
originated by a subscriber during the busy hour.

d
a– 0.2 typical
□ It is useful in sizing the exchange to handle peak
traffic.

M
□ Rural area
□ Business area – 3 typical

139
 e
Traffic load and parameters

an
h
□ Example: An exchange serves 2000

a m
subscribers. If the average BHCA is 10000 and

Only 60% of total attemptsT

CCR is 60%, calculate the busy hour calling rate.

t a= 10000 X 0.6 = 6000

are successful.

i
□

Busy hour calling m

□ Average busy hour calls

= 3 calls perh
u
d
□ rate = 6000/2000

a
□ subscriber.

M
140
 e
Traffic load and parameters

an
h
□ Day-to-busy hour traffic ratio – ratio of busy

am
hour calling rate to average calling rate for the

Ttraffic is carried
day.

t a
Gives how much of day’s total

Business area - 20 i
□

m
in busy hour.

Rural area -h u
□

a d
□ 6-7

M
141
 n e
Traffic load and parameters

h a
Traffic intensity A – Ratio of period for which a

m
□ o

a
server is occupied to total period of observation.

T
□ Server includes all common equipments irrespective

t a
of locations.

i in the network.
m
□ This gives traffic on the network in terms of the

Generally periodu
h
occupancy of the servers

a d
□ of observation is 1 hour.
□ A is dimensionless.

Mof traffic – servers occupied for entire

o

□ Called erlang (E) in honour of scientist.

□ 1 erlang
period of observation.
 e
Traffic load and parameters

A group of 10 servers, each is occupied forn

minutes in an observation interval of 2h a 30

Calculate the traffic carried by them

□

a
hours.

T
group.

t a
□ Traffic carried per server = 30/120

Total traffic carried i

m
□ = 0.25E

h u
□ by the group = 10 X 0.25

a d
□ = 2.5E
Erlang measure indicateds average number of

Min driving average number of calls put

□
servers occupied .
□ Useful
through during period of observation
 A group of 20 servers carry a traffic of 10n
e
Traffic load and parameters

a
If the average duration of the call is 3hminutes,
erlang.

m
□

a
T
calculate number of calls put through by a single

t a
server and the group as a whole in one hour period.

i total period.
m
□ Traffic per server = 10/20 = 0.5 E

Hence a server u
h
□ Server busy for 0.5 of

a d
□ busy for 0.5 * 60 = 30 minutes
□ Total number of calls/server = 30/3 = 10 calls.

M
□ Total number of calls by group = 10*20 calls.
□ =200 calls
 Traffic intensity also measured in CCS n
e
Traffic load and parameters

a
h product.
m
□

Valid only in telephone circuits. a

□ Centum Call Second represents call time

T or 100 call
a
□

for 1s duration or any t

i
□ 1 CCS can be 1 call for 100s duration

u m
other.

d h
□ Total duration same = 100s.

a = 3600 CS = 60 CM
□ Some times CM or CS are used to measure TI.

1EM
□ 1E = 36CCS
□ means busy full duration of 60 CM.
□ 100CS = 1CCS
 n e
Traffic load and parameters

duration in a 1-hour period. Calculate h

a
m
□ A subscriber makes 3 phone calls 3m, 4m and 2m

a
subscriber

Subscriber traffic in erlangT

traffic in erlang, CCS and CM

t a= 9/60 = 0.15E
i
□ =

u m
□ =busy period/total period

Traffic in CMh= 60* 0.15 = 9

d
□ Traffic in CCS = 36*0.15 = 5.4 CCS

a in CM = 3+4+2 = 9
□

M
□ Or Traffic
 Traffic intensity is a call-time product. n
e
Traffic load and parameters

h a
m
□

a
□ Parameters –

Average holding time per callT

a
□ Average call arrival rate C

t
i = A = Ct
□ t

m
h

u
Load offered to network

h
□ h

a d
M
 ne
Traffic load and parameters

h a
□ Assignment: Over a 20 minute observation interval,

am
40 subscribers initiate calls. Total duration of the

T
calls is 4800s. Calculate the load offered to the

a
network by the subscribers and average subscriber

Average holding time i

t
traffic.

m
□ Average call arrival rate = 40/20 = 2 calls/m

h u
□

d
Average a
□ = 4800/40 =120s = 2m/call
□ Offered load = 2*2 = 4E

M
□ subscriber traffic = 4/40 = 0.1E
 n e
Action during overload

Loss system-Overload traffic may beh

a
m
□ Two options.

a NW facilities
□ rejected.

T
a
□ Delay system – Held in queue until

Conventional automatictexchanges are based on

i
are made available again.

u m
□

d h
loss system.

a
□ User has to retry.

M
 n e
Grade of service and blocking probability

h a
a m
□ In loss system, traffic carried by NW is lower

T
than actual traffic offered to NW.

Grade of service –GOS a

t
□ Overload traffic is rejected.

i of service.
m
□

h u
□ Is an index of quality

Is ratioaofd
□ Is amount of traffic rejected by network.

M
□ lost traffic to offered traffic.
 ne
Grade of service and blocking probability

h a
am
□ Offered traffic- A = Cth

T
□ C -- Average number of calls generated by the

t a
user.

Carried traffic – m
i
□ th– average holding time per call.

h u
d
□ actual traffic carried by NW.

is periodafor which a server is occupied out of

□ Is average occupancy of server.

M
□
total observation time.
 ne
Grade of service and blocking probability

h a
am
□ GOS = (A-Ao)/A

T
□ A = offered traffic

t a
i
□ Ao = carried traffic

u m
□ A - Ao= lost traffic

dh
□ GOS as small as possible for better service.

a 1000 calls.
□ Recommended value = 0.002

M
□ 2 calla per
 n e
Blocking probability PB- Loss system

Defined as probability that all servers a

h
a m
□ in system

T
are busy.

t a
□ Any new arrival is blocked.

i number of servers and

m
□ Not same as GOS.

subscribers- u
h
□ If an exchange has same

□
a d
GOS is zero.

M
□ Blocking probability non zero.
 n e
Blocking probability PB

h a
a m
□ GOS is a measure from subscriber point of view.

T
□ Blocking probability is a measure from network

t a
or switching system point of view.

rejected subscriber icalls.

m
□ GOS is arrived at by observing number of

h u
a d
□ Blocking probability is arrived at by observing
the busy servers in switching system.

PM
□ GOS called call congestion/loss probability.
□ B called time congestion
 n e
Blocking probability -Delayed system

All calls are put through the network ash

a
□ In system, traffic carried = load offered.

a m
□ and when

T
NW facilities are available.

t a
GOS always zero.

i
□

m
□ Delay probability- prob that a call experiences a

If input rate faru

h
delay.

d
Unstablea
□ exceeds NW capacity, undesirably
long queue and delay.

M
□ as never recovers.
 If queue size more that acceptable level-n
e
Flow control -Delayed system

Made to act as loss system till queueh

a
m
□

a
□ size below

T
acceptable level.

t a
i
u m
d h
a
M
 ne
FASCIMILE

h a
m
□ Means exact reproduction.

Band width required very small.a

□ Of a document or a picture.

Ttelephone lines.
a
□

USES: Transmissionioft
□ Suitable for transmission over

photograph. um
□

Document,d h maps etc..

□

a
□ weather

M
□ Language texts for which tele-printer is not
available
 ne
FASCIMILE -Sender
Message—
h a
m
□

a
□ A single page

T
□ Narrow continuous tape.

t a
i
□ Continuous sheet paper.

u m
□ Scanning methods—

d h
□ Optical scanning-light spot traverses the message.

a scanning-character of message offers

□ More common.

M
□ Resistance
varying resistance,
□ Brought into circuit using a stylus.
 e
FASCIMILE-Sender

an
h
m
□ Cylindrical Scanning

a
T
t a
i
um
dh
a
M
 n e
Cylindrical scanning
Message fixed around drum.
a
halong axis
m
□

a
□ Drum rotated about its axis and moves

Moves below a fixed scanningT

simultaneously.

t a
i
□ spot.

m
Reflected light focused on photo cell .

u
□

h
Photo cell converts light to electrical signal.

d
□

a very small using mask or lenses.

□ Solid state amplifiers amplify signal.

M
□ Spot made
□ Spot follows spiral path.
 ne
Cylindrical scanning

h a
m
□ Uncommon alternate arrangements-

Drum stationary, spot moves. a

□ Scan in series of closed rings.

T
a
□

t
i inch per second
–m
!

Rotation speedu
□ Traversing speed – 1/100

d h
□ 60 rpm

picture.a
□ 100 scanning lines on each 1 inch width of

M
 ne
Tape scanning

h a
am
T
t a
i
um
dh
a
M
 n e
Tape scanning

a
h across
m
□ Message taken directly off a printed tape.

a
□ Scanning beam falls from top and travels

T
the rape.

t a beam to travel across

i
□ Achieved using hexagonal prism.

u m
□ Prism rotates and deflects

New trace athstart of each face of prism.

d
the tape.

a
□

M
 n e
Scanning spot

h a
m
□ Shape of scanning spot determines wave shape

a gap or
of signal output.

T
□ Preferred- Rectangular shape without

t a with little overlap.

i
overlap.

Average width ofm

u
□ Less preferred – Trapezoidal

d h
□ top and bottom widths is P.

a
M
 ne
Scanning spot

h a
am
T
t a
i
um
dh
a
M
 Equipment used is identical but process isn
e
Facsimile Receiver- photographic reception

h a
m
□

a
reverse.

T intensity of
□ Input is electrical and output optical.

t a
i
□ Received electrical signal varies

u m
light beam.

d h
□ Light beam falls on photographic material.

a
M
 ne
h a
am
T
t a
i
um
dh
a
M
 ne
Facsimile Receiver- photographic reception

h a
□ Small coil of fine wires suspended in strong magnetic

m
a the mirror as
field.

T
□ Small mirror is mounted in coil.

t aat no signal.
□ Electrical signal through coil deflects

i
per its strength.

u m
□ Mirror is kept off center

h
Large signalddeflects mirror more and larger light
□ Small signal deflects mirror less and less light passes

a aperture.
through aperture.
□

M positive image on photographic plate.

passes through
□ Provides
 n e
Facsimile Receiver- photographic reception

a
hhelium.
m
□ Alternate method.

Glows when voltage is applied. a

□ Crater lamp containing neon, argon, or

T voltage .
a
□

it
□ Intensity of light changes with

Output light is u m to fall on a photographic

□ Signal is applied to lamp.

d h
□ made

aefficient and responsive.

plate.

M
□ Not very
 n e
Facsimile Receiver- Direct Recording reception

h a
Highly absorbent chemically treated paper is used.

m
□

a
□ Electrolyte held in paper disassociate when voltage is

T
applied.

t a
□ Signal voltage applied via a metal stylus.

Produces a mark onm

i
□ Metallic salt so produced reacts with colour chemical

u
on paper.

d h
□ paper.

a
□ Intensity of mark depends on amount of
disassociation.

Mis damp and must be kept sealed

□ Hence depends on electrical signal.
□ Paper
 Paper is damp and must be kept sealed n
e
Facsimile Receiver- Direct Recording reception

h a
m
□

a
□ Cheap.

T
a
□ Tonal range less.

it
□ Suitable for low grade applications.

u m
d h
a
M
 n e
Synchronization

h a
m
□ For documentary, need for synchronization is not

a motor at both
severe.

Tcontrolled mains.
□ Can be achieved using synchronous

t a be synchronized with
i
ends, operated off frequency

u m
□ For picture, receiver must

h
transmitter.

Sender a
d
□ By sending synchronizing signals at 1020Hz.

M
□ speed bears known relation to 1020Hz
□ Receiver speed adjusted using stroboscope.
 n e
Synchronization

h a
With carrier transmission, carrier is sent along

m
□

a
with USB.

T this 1020Hz.
□ Carrier helps in recovering 1020Hz.

t aspeed error, picture

would be distorted. i
□ Speed of receiver adjusted with

m
If receiver has constant

u
□

d h
a
□ Phase error breaks the picture.
Can be avoided by sending 1020Hz pulsed

Mreleases the switch holding the receiver

□
momentarily to indicate start of the transmission.
□ Pulse
drum.
 ne
Synchronization
h a
am
T
□ No error

t a
i
um
Constant speed

h
error

ad
M
Phasing
error
 ne
Index of cooperation

transmitted and received pictures. h

a
m must be
□ Height/width ratio must be same for

Hence scanning pitch and drum a

T
□ diameter

t a
i
same at both ends.

u m
d h
a
M
 ne
D- sending drum Diameter 


a
d – receiving drum Diameter


h
P – Sender scanning pitch


am
p – Receiver scanning pitch


T
n – number of lines scanned

t a
i
um
dh
a
M
 ne
Index of cooperation

h a
m
□ Width of transmitted picture – nP

aproportional to
□ Width of transmitted picture – np

T
a is proportional to d
□ Height of transmitted picture is

it
D.

u m
□ Height of received picture

d h
with same constant of proportionality.

a
□ For correct height/width ratio-

M
□ D/nP = d/np
□ D/P = d/p
 ne
Index of cooperation

h a
Called Index of cooperation. am
□ Ratio of diameter to scanning pitch should be
same at both ends.

Tstroke length and

□

t a
i
□ IEEE defines it as product of

For drum scanner,m

u
scan density.

Scan densityh
d
□ stroke length is ΠD

a = ΠD /P
□ is lines per unit length = 1/P

M = D/P
□ IOC(IEEE)
□ IOC(CCITT)
 Effect of different index of cooperation.n
e
Index of cooperation

h a
m
□

a
T
t a
i
u m
d h
a
M
 n e
Index of cooperation

a
h is
ism
□ Assignment:

70.4mm, and the scanning pitcha

□ The drum diameter of a facsimile machine

T
0.2 mm per

t a
i
scan. Calculate IOC

um
□ IOC(IEEE) =

IOC(CCITT) h
d
□ 1106

a
□ =

M
□ 352