Chapter 4

Circuit-Switching
Networks
Multiplexing
SONET
Transport Networks
Circuit Switches
The Telephone Network
Signaling
Traffic and Overload Control in Telephone Networks
Cellular Telephone Networks
Circuit Switching Networks
z End-to-end dedicated circuits between clients
z Client can be a person or equipment (router or switch)
z Circuit can take different forms
z Dedicated path for the transfer of electrical current
z Dedicated time slots for transfer of voice samples
z Dedicated frames for transfer of Nx51.84 Mbps signals
z Dedicated wavelengths for transfer of optical signals
z Circuit switching networks require:
z Multiplexing & switching of circuits
z Signaling & control for establishing circuits
z These are the subjects covered in this chapter
How a network grows
(a) A switch provides the network to a cluster of users, e.g.
a telephone switch connects a local community

Network Access
network

(b) A multiplexer connects two access networks, e.g. a high

speed line connects two switches
 A Network Keeps Growing
1*
a b
2 a b
4
(a) Metropolitan network A 3
viewed as Network A of A A
Access Subnetworks c d d
c

Network of
Metropolitan Access
(b) National network viewed
as Network of Regional Subnetworks
Subnetworks (including A)

zVery high-
speed lines α

Network of Regional National &

Subnetworks International
 Chapter 4
Circuit-Switching
Networks
Multiplexing
Multiplexing
z Multiplexing involves the sharing of a transmission channel
(resource) by several connections or information flows
z Channel = 1 wire, 1 optical fiber, or 1 frequency band
z Significant economies of scale can be achieved by combining
many signals into one
z Fewer wires/pole; fiber replaces thousands of cables
z Implicit or explicit information is required to demultiplex the
information flows.

(a) (b) Shared

A A A Channel A

B B B MUX MUX B

C C C C
Frequency-Division Multiplexing
z Channel divided into frequency slots

A
f
0 Wu
(a) Individual
signals occupy B z Guard bands
Wu Hz 0
f required
Wu
z AM or FM radio
C
stations
0 Wu
f z TV stations in
air or cable
(b) Combined z Analog
signal fits into telephone
channel A B C systems
bandwidth f
0 W
Time-Division Multiplexing
z High-speed digital channel divided into time slots

A1 A2 …
t
0T 3T 6T
z Framing
(a) Each signal B1 B2 …
t
required
transmits 1 unit 3T 6T z Telephone
0T
every 3T digital
seconds transmission
C1 C2 …
t
0T 3T 6T
z Digital
transmission in
(b) Combined backbone
signal transmits A1 B1 C1 A2 B2 C2 …
t network
1 unit every T
0T 1T 2T 3T 4T 5T 6T
seconds
 T-Carrier System
z Digital telephone system uses TDM.
z PCM voice channel is basic unit for TDM
z 1 channel = 8 bits/sample x 8000 samples/sec. = 64 kbps
z T-1 carrier carries Digital Signal 1 (DS-1) that
combines 24 voice channels into a digital stream:

1 1

2 MUX MUX 2
...

...
22 23 24 b 1 2 ... 24 b
24 Frame 24
Framing bit
Bit Rate = 8000 frames/sec. x (1 + 8 x 24) bits/frame
= 1.544 Mbps
 North American Digital
Multiplexing Hierarchy
1 DS1 signal, 1.544Mbps
.
. Mux
24

1 DS2 signal, 6.312Mbps

24 DS0 .
4 DS1 . Mux
4

1 DS3 signal, 44.736Mpbs

.
7 DS2 . Mux
7

1
DS0, 64 Kbps channel
.
z 6 DS3 . Mux
6
z DS1, 1.544 Mbps channel
z DS2, 6.312 Mbps channel DS4 signal
z DS3, 44.736 Mbps channel 274.176Mbps

z DS4, 274.176 Mbps channel

 CCITT Digital Hierarchy
z CCITT digital hierarchy based on 30 PCM channels
1 2.048 Mbps
.
30
. Mux

1 8.448 Mbps
64 Kbps .
4
. Mux

1 34.368 Mpbs
.. Mux
139.264 Mbps
1
z E1, 2.048 Mbps channel .
. Mux
z E2, 8.448 Mbps channel 4

z E3, 34.368 Mbps channel

z E4, 139.264 Mbps channel
Clock Synch & Bit Slips
z Digital streams cannot be kept perfectly synchronized
z Bit slips can occur in multiplexers
Slow clock results in late bit
arrival and bit slip

MUX t

5 4 3 2 1 5 4 3 2 1
Pulse Stuffing
z Pulse Stuffing: synchronization to avoid data loss due to slips
z Output rate > R1+R2
z i.e. DS2, 6.312Mbps=4x1.544Mbps + 136 Kbps
z Pulse stuffing format
z Fixed-length master frames with each channel allowed to stuff or
not to stuff a single bit in the master frame.
z Redundant stuffing specifications
z signaling or specification bits (other than data bits) are distributed
across a master frame.

Muxing of equal-rate signals Pulse stuffing

requires perfect synch
Wavelength-Division Multiplexing
z Optical fiber link carries several wavelengths
z From few (4-8) to many (64-160) wavelengths per fiber
z Imagine prism combining different colors into single beam
z Each wavelength carries a high-speed stream
z Each wavelength can carry different format signal
z e.g. 1 Gbps, 2.5 Gbps, or 10 Gbps

Optical Optical
λ1 MUX deMUX λ1

λ2 λ1 λ2 . λ2
λm

Optical
fiber
λm λm
Example: WDM with 16
wavelengths

30 dB
1540 nm

1550 nm

1560 nm
Typical U.S. Optical Long-Haul
Network
 Chapter 4
Circuit-Switching
Networks
SONET
SONET: Overview

z Synchronous Optical NETwork

z North American TDM physical layer standard for
optical fiber communications
z 8000 frames/sec. (Tframe = 125 μsec)
z compatible with North American digital hierarchy
z SDH (Synchronous Digital Hierarchy) elsewhere
z Needs to carry E1 and E3 signals
z Compatible with SONET at higher speeds
z Greatly simplifies multiplexing in network backbone
z OA&M support to facilitate network management
z Protection & restoration
 SONET simplifies multiplexing
Pre-SONET multiplexing: Pulse stuffing required demultiplexing
all channels

MUX DEMUX MUX DEMUX

Remove Insert
tributary tributary

SONET Add-Drop Multiplexing: Allows taking individual channels in

and out without full demultiplexing

MUX ADM DEMUX

Remove Insert
tributary tributary
SONET Specifications
z Defines electrical & optical signal interfaces
z Electrical
z Multiplexing, Regeneration performed in electrical
domain
z STS – Synchronous Transport Signals defined
z Very short range (e.g., within a switch)
z Optical
z Transmission carried out in optical domain
z Optical transmitter & receiver
z OC – Optical Carrier
SONET & SDH Hierarchy
SONET Electrical Optical Signal Bit Rate (Mbps) SDH
Signal Electrical Signal

STS-1 OC-1 51.84 N/A

STS-3 OC-3 155.52 STM-1
STS-9 OC-9 466.56 STM-3
STS-12 OC-12 622.08 STM-4
STS-18 OC-18 933.12 STM-6
STS-24 OC-24 1244.16 STM-8
STS-36 OC-36 1866.24 STM-12
STS-48 OC-48 2488.32 STM-16
STS-192 OC-192 9953.28 STM-64
STS: Synchronous OC: Optical Channel STM: Synchronous
Transport Signal Transfer Module
SONET Multiplexing

DS1 Low-speed
DS2 mapping
E1 function STS-1
51.84 Mbps
Medium
DS3 speed STS-1
44.736 mapping OC-n
STS-n
...
function
...

Scrambler E/O
STS-3c
High- STS-1 MUX
E4
speed STS-1
mapping STS-1
139.264 function
STS-3c
STS-1
High- STS-1
ATM or speed STS-1
POS mapping
function
SONET Equipment
z By Functionality
z ADMs: dropping & inserting tributaries
z Regenerators: digital signal regeneration
z Cross-Connects: interconnecting SONET streams
z By Signaling between elements
z Section Terminating Equipment (STE): span of fiber
between adjacent devices, e.g. regenerators
z Line Terminating Equipment (LTE): span between adjacent
multiplexers, encompasses multiple sections
z Path Terminating Equipment (PTE): span between SONET
terminals at end of network, encompasses multiple lines
 Section, Line, & Path in SONET
PTE PTE
LTE LTE
STE STE STE
SONET SONET
terminal MUX MUX terminal
Reg Reg Reg

Section Section Section Section

STS Line
STS-1 Path

STE = Section Terminating Equipment, e.g., a repeater/regenerator

LTE = Line Terminating Equipment, e.g., a STS-1 to STS-3 multiplexer
PTE = Path Terminating Equipment, e.g., an STS-1 multiplexer

z Often, PTE and LTE equipment are the same

z Difference is based on function and location
z PTE is at the ends, e.g., STS-1 multiplexer.
z LTE in the middle, e.g., STS-3 to STS-1 multiplexer.
Section, Line, & Path Layers in
SONET

Path Path
Line Line Line Line
Section Section Section Section Section Section Section
Optical Optical Optical Optical Optical Optical Optical

z SONET has four layers

z Optical, section, line, path
z Each layer is concerned with the integrity of its own signals
z Each layer has its own protocols
z SONET provides signaling channels for elements within a
layer
SONET STS Frame
z SONET streams carry two types of overhead
z Path overhead (POH):
z inserted & removed at the ends
z Synchronous Payload Envelope (SPE) consisting
of Data + POH traverses network as a single unit
z Transport Overhead (TOH):
z processed at every SONET node
z TOH occupies a portion of each SONET frame
z TOH carries management & link integrity
information
 STS-1 Frame z810x64kbps=51.84Mbps
810 Octets per frame @ 8000 frames/sec
90 columns

A1 A2 J0 J1
B1 E1 F1 B3
1
D1 D2 D3 C2
Order of
2 transmission H1 H2 H3 G1
9 rows B2 K1 K2 F2
D4 D5 D6 H4
D7 D8 D9 Z3
Special OH octets:
D10 D11 D12 Z4
A1, A2 Frame Synch S1 M0/1 E2 N1
B1 Parity on Previous Frame
(BER monitoring) 3 Columns of Synchronous Payload Envelope (SPE)
J0 Section trace Transport OH 1 column of Path OH + 8 data columns
(Connection Alive?)
H1, H2, H3 Pointer Action Section Overhead Path Overhead
K1, K2 Automatic Protection Line Overhead Data
Switching
 SPE Can Span Consecutive Frames

Pointer First octet

Frame 87 Columns
k

Synchronous
9 Rows
payload
envelope
Pointer Last octet
Frame
k+1
First column is path overhead

z Pointer indicates where SPE begins within a frame

z Pointer enables add/drop capability
Stuffing in SONET
z Consider system with different clocks (faster out than in)
z Use buffer (e.g., 8 bit FIFO) to manage difference
z Buffer empties eventually
z One solution: send “stuff”
z Problem:
z Need to signal “stuff” to receiver

FIFO
1,000,000 bps 1,000,001 bps
 Negative & Positive Stuff

Frame Frame
k Pointer k Pointer
First octet First octet
of SPE of SPE

Stuff byte Stuff byte

Frame Frame
k+1 Pointer k+1 Pointer
First octet First octet
of SPE of SPE

(a) Negative byte stuffing (b) Positive byte stuffing

Input faster than output Input is slower than output
Send extra byte in H3 to catch up Stuff byte to fill gap
Synchronous Multiplexing
z Synchronize each incoming STS-1 to local clock
z Terminate section & line OH and map incoming SPE into a new
STS-1 synchronized to the local clock
z This can be done on-the-fly by adjusting the pointer
z All STS-1s are synched to local clock so bytes can be
interleaved to produce STS-n

STS-1 STS-1 STS-1 STS-1

Map
STS-1 STS-1 STS-1 STS-1 Byte STS-3
Map Interleave
STS-1 STS-1 STS-1 STS-1
Map

Incoming Synchronized new

STS-1 frames STS-1 frames
Octet Interleaving

Order of
transmission 1

A1 A2 J0 J1
2
A1 A2 J0 J1
3 B1 E1 F1 B3
A1 A2 J0 J1
B1 E1 F1 B3
D1 D2 D3 C2
B1 E1 F1 B3
D1H1 D2 D3 C2
H2 H3 G1
D1H1 D2 D3 C2
H2 H3 G1
H1 B2 K1 K2 F2
H2 H3 G1
B2 K1 K2 F2
D4 D5 D6 H4
B2 K1 K2 F2
D4 D5 D6 H4
D7 D8 D6 H4
D9 Z3
D4 D5
D7 D8 D9 Z3
D10 D11 D12 Z4
D7 D8 D9 Z3
D10 D11 D12 Z4
S1 M0/1 E2 N1
D10 D11 D12 Z4
S1 M0/1 E2 N1
S1 M0/1 E2 N1
 Concatenated Payloads
Concatenated Payload OC-Nc z Needed if payloads of interleaved
frames are “locked” into a bigger
zN x 87 columns unit
z Data systems send big blocks of
J1
information grouped together, e.g.,
B3 a router operating at 622 Mbps
C2 z SONET/SDH needs to handle
G1 these as a single unit
F2 z H1,H2,H3 tell us if there is
H4 concatenation
Z3 z STS-3c has more payload than 3
Z4
STS-1s
N1
z STS-Nc payload = Nx780 bytes
z OC-3c = 149.760 Mb/s
z OC-12c = 599.040 Mb/s
87N - (N/3) z OC-48c = 2.3961 Gb/s
(N/3) – 1
columns of columns of z OC-192c = 9.5846 Gb/s
fixed stuff payload
 Chapter 4
Circuit-Switching
Networks
Transport Networks
Transport Networks
z Backbone of modern networks
z Provide high-speed connections: Typically STS-1 up to OC-192
z Clients: large routers, telephone switches, regional networks
z Very high reliability required because of consequences of failure
z 1 STS-1 = 783 voice calls; 1 OC-48 = 32000 voice calls;

Telephone
Switch

Router
Router

Transport Network

Telephone
Switch
Telephone
Switch
Router
SONET ADM Networks

MUX ADM DEMUX

Remove Insert
tributary tributary

z SONET ADMs: the heart of existing

transport networks
z ADMs interconnected in linear and ring
topologies
z SONET signaling enables fast restoration
(within 50 ms) of transport connections
Linear ADM Topology
z ADMs connected in linear fashion
z Tributaries inserted and dropped to connect clients

1 2 3 4

z Tributaries traverse ADMs transparently

z Connections create a logical topology seen by clients
z Tributaries from right to left are not shown
2

1 3

4
1+1 Linear Automatic Protection
Switching T = Transmitter W = Working line
R = Receiver P = Protection line

W
T R

Bridge Selector

T R
P

• Simultaneous transmission over diverse routes

• Monitoring of signal quality
• Fast switching in response to signal degradation
• 100% redundant bandwidth
1:1 Linear APS
Switch Switch
W
T R

APS signaling

T R
P

• Transmission on working fiber

• Signal for switch to protection route in response to
signal degradation
• Can carry extra (preemptible traffic) on protection line
1:N Linear APS
Switch Switch
W1
T R

W²
T R

…
…
…

…
Wn
T R

P
T R

APS signaling
• Transmission on diverse routes; protect for 1 fault
• Reverts to original working channel after repair
• More bandwidth efficient
SONET Rings
z ADMs can be connected in ring topology
z Clients see logical topology created by tributaries

(a) (b)
a a

OC-3n
OC-3n

b c
c
OC-3n
Three ADMs connected in Logical fully connected
physical ring topology topology
SONET Ring Options
z 2 vs. 4 Fiber Ring Network
z Unidirectional vs. bidirectional transmission
z Path vs. Link protection

z Spatial capacity re-use & bandwidth

efficiency
z Signalling requirements
Two-Fiber Unidirectional Path
Switched Ring
Two fibers transmit in opposite directions
z Unidirectional
z Working traffic flows clockwise
z Protection traffic flows counter-clockwise
z 1+1 like
z Selector at receiver does path protection
switching
UPSR
1

4 2

W = Working Paths

zNo spatial re-use P = Protection Paths

Each path uses 2x bw 3
UPSR path recovery
1

4 2

W = Working line
P = Protection line
3
UPSR Properties
z Low complexity
z Fast path protection
z 2 TX, 2 RX
z No spatial re-use; ok for hub traffic pattern
z Suitable for lower-speed access networks
z Different delay between W and P path
Four-Fiber Bidirectional Line
Switched Ring
z 1 working fiber pair; 1 protection fiber pair
z Bidirectional
z Working traffic & protection traffic use same route
in working pair
z 1:N like
z Line restoration provided by either:
z Restoring a failed span
z Switching the line around the ring
4-BLSR
1

Equal
W
delay

P
Standby
bandwidth 2
4
is shared

Spatial
Reuse

3
BLSR Span Switching
1
W
Equal
delay

P
zSpan
Switching 2
4
restores
failed line

Fault on
working
links
3
BLSR Span Switching
1
W
Equal
delay

P
zLine
Switching 2
4
restores
failed lines

Fault on
working and
protection
links

3
4-BLSR Properties

z High complexity: signalling required

z Fast line protection for restricted distance
(1200 km) and number of nodes (16)
z 4 TX, 4 RX
z Spatial re-use; higher bandwidth efficiency
z Good for uniform traffic pattern
z Suitable for high-speed backbone networks
z Multiple simultaneous faults can be handled
Backbone Networks consist of
Interconnected Rings

Regional UPSR
Metro OC-12
ring ring Interoffice
rings

BLSR
OC-48,
OC-192

UPSR or
BLSR
OC-12,
OC-48
The Problem with Rings
z Managing bandwidth can be complex
z Increasing transmission rate in one span
affects all equipment in the ring
z Introducing WDM means stacking SONET
ADMs to build parallel rings
z Distance limitations on ring size implies many
rings need to be traversed in long distance
z End-to-end protection requires ring-
interconnection mechanisms
Managing 1 ring is simple; Managing many
rings is very complex
Mesh Topology Networks using
SONET Cross-Connects
z Cross-Connects are nxn switches
z Interconnects SONET streams
z More flexible and efficient than rings
z Need mesh protection & restoration
Router

B A
C

D
Router F Router

G E

Router
 From SONET to WDM
SONET WDM
z combines multiple SPEs z combines multiple wavelengths into a
into high speed digital common fiber
stream z Optical ADMs can be built to insert and
z ADMs and drop wavelengths in same manner as
crossconnects in SONET ADMS
interconnected to form z Optical crossconnects can also be built
networks z All-optical backbone networks will
z SPE paths between provide end-to-end wavelength
clients from logical connections
topology z Protection schemes for recovering
z High reliability through from failures are being developed to
protection switching provide high reliability in all-optical
networks
Optical Switching

…
Optical
… fiber switch

…
DeMUX
MUX

Output Input

WDM
…
WDM

Wavelength
cross-connect

WDM
…
…
WDM

Dropped Added
wavelengths wavelengths
 Chapter 4
Circuit-Switching
Networks
Circuit Switches
 Network: Links & switches
z Circuit consists of dedicated resources in sequence
of links & switches across network
z Circuit switch connects input links to output links
zSwitch
zNetwork
Control

Link Switch
1 1
2 2
User n 3 Connection 3
of inputs
User n – 1 …
to outputs

…
User 1

N N
 Circuit Switch Types

z Space-Division switches
z Provide separate physical connection between
inputs and outputs
z Crossbar switches
z Multistage switches
z Time-Division switches
z Time-slot interchange technique
z Time-space-time switches
z Hybrids combine Time & Space switching
 Crossbar Space Switch
z N x N array of
crosspoints
1
z Connect an input to
2
an output by closing

…
a crosspoint
z Nonblocking: Any N
input can connect to
idle output …
N –1 N
1 2
z Complexity: N2
crosspoints
 Multistage Space Switch
z Large switch built from multiple stages of small switches
z The n inputs to a first-stage switch share k paths through intermediate
crossbar switches
z Larger k (more intermediate switches) means more paths to output
z In 1950s, Clos asked, “How many intermediate switches required to
make switch nonblocking?”
2(N/n)nk + k (N/n)2 crosspoints

n×k N/n × N/n k×n

1 1
1
n×k k×n
2 2 N
N N/n × N/n
inputs n×k 2 k×n outputs
3 3

…
…
…

n×k k×n
N/n N/n
N/n × N/n
k
 Clos Non-Blocking Condition:
k=2n-1
z Request connection from last input to input switch j to last output in output switch m
z Worst Case: All other inputs have seized top n-1 middle switches AND all other
outputs have seized next n-1 middle switches
z If k=2n-1, there is another path left to connect desired input to desired output

nxk N/n x N/n kxn

1 1
1
…

…
n-1
busy N/n x N/n
Desired nxk n-1 kxn Desired
j
input m output
n-1
N/n x N/n
n+1 busy
…

…
# internal links =
N/n x N/n 2x # external links
2n-2
nxk kxn
N/n
Free path N/n2n-1
x N/n Free path N/n
Minimum Complexity Clos Switch
C(n) = number of crosspoints in Clos switch

= 2Nk + k( N )2 = 2N(2n – 1)+(2n – 1)( N )2

n n
Differentiate with respect to n:

0= δC
δn
2N 2 2N 2
= 4N – 2 + 3 ≈ 4N –
n n
2N
n2
2
==> n ≈ √ N2

The minimized number of crosspoints is then:

N2 N
C* = (2N + N/2 )(2( 2 )1/2 – 1) ≈ 4N √ 2N = 4 √ 2N1.5

This is lower than N2 for large N

 Example: Clos Switch Design
z Circa 2002, Mindspeed offered a Crossbar
chip with the following specs:
z 144 inputs x 144 outputs, 3.125
Gbps/line 8x16 144×144 16x8
1 1
z Aggregate Crossbar chip throughput: 1
450 Gbps

1152 outputs
8x16 16x8

1152 inputs
2 2
144x144
z Clos Nonblocking Design for 1152x1152 8x16 2 16x8
switch 3 3

…
z N=1152, n=8, k=16

…
…
z N/n=144 8x16 switches in first stage
8x16 16x8
z 16 144x144 in centre stage
144 N/n
z 144 16x8 in third stage 144x144
16
z Aggregate Throughput: 3.6 Tbps!

z Note: the 144x144 crossbar can be

partitioned into multiple smaller switches
 Time-Slot Interchange (TSI)
Switching
z Write bytes from arriving TDM stream into memory
z Read bytes in permuted order into outgoing TDM stream
z Max # slots = 125 μsec / (2 x memory cycle time)

1 a
Read slots
2 b
according to
3 connection
d c … b a zzz permutation b a … d c
24 23 2 1 24 23 2 1
Write
22
slots in
order of 23 c
zIncoming arrival zOutgoing
TDM 24 d TDM
stream stream

Time-slot interchange
Time-Space-Time Hybrid Switch
z Use TSI in first & third stage; Use crossbar in middle
z Replace n input x k output space switch by TSI switch that takes n-slot
input frame and switches it to k-slot output frame

nxk N/n x N/n kxn

1 1
1

nxk
N 2
Input TDM Output TDM
inputs
nxk frame with frame with k
n slots 1
3 slots
2

zzz
…

n … 2 1 k … 2 1

nxk n
N/n

Time-slot interchange
Flow of time slots between
switches
First slot First slot
n×k N/n × N/n k×n
1 1 1

n×k k×n
2
2 N/n × N/n
2

…
…
…

n×k k×n
N/n
N/n N/n × N/n
kth slot k kth slot

z Only one space switch active in each time slot

Time-Share the Crossbar Switch
TSI stage Space stage TSI stage

TDM nxk TDM TDM kxn

n slots 1 k slots k slots 1

n slots nxk kxn

N 2 N/n x N/n 2 N
inputs Time-shared outputs
n slots nxk space switch kxn
3 3
…

…
n slots nxk kxn
N/n N/n

z Interconnection pattern of space switch is

reconfigured every time slot
z Very compact design: fewer lines because of TDM
& less space because of time-shared crossbar
 Example: A→3, B→4, C→1, D→3
(a)
A C
B
A
z3-stage
Space
Switch
C D
D B

(b)
B2 A2 B1 A1 B1 A1 C1 A1 A1 C1
2x3 3x2
1 1
zEquivalent
TST Switch

D1 B1 B1 D1
D2 C2 D1 C1
2x3
D1 C1 3x2
2 2
Example: T-S-T Switch Design
For N = 960
z Single stage space switch ~ 1 million crosspoints
z T-S-T
z Let n = 120 N/n = 8 TSIs
z k = 2n – 1 = 239 for non-blocking
z Pick k = 240 time slots
z Need 8x8 time-multiplexed space switch

For N = 96,000
z T-S-T
z Let n = 120 k = 239
z N / n = 800
z Need 800x800 space switch
Available TSI Chips circa 2002
z OC-192 SONET Framer Chips
z Decompose 192 STS1s and perform (restricted)
TSI

z Single-chip TST
z 64 inputs x 64 outputs
z Each line @ STS-12 (622 Mbps)
z Equivalent to 768x768 STS-1 switch
Pure Optical Switching
z Pure Optical switching: light-in, light-out,
without optical-to-electronic conversion
z Space switching theory can be used to
design optical switches
z Multistage designs using small optical switches
z Typically 2x2 or 4x4
z MEMs and Electro-optic switching devices
z Wavelength switches
z Very interesting designs when space switching is
combined with wavelength conversion devices
 Chapter 4
Circuit-Switching
Networks
The Telephone Network
Telephone Call
z User requests connection
z Network signaling
establishes connection
z Speakers converse
z User(s) hang up
z Network releases
connection resources
Source
Signal
Go Message
Signal ahead Release

Signal
Destination
 Call Routing
z Local calls routed
(a) 4
C D through local network
(In U.S. Local Access &
2 3 Transport Area)
5

A B
1 z Long distance calls
routed to long distance
service provider

(b)
Net 1

Net 2

LATA 1 LATA 2
Telephone Local Loop
Local Loop: “Last Mile”
z Copper pair from telephone to CO
z Pedestal to SAI to Main
Distribution Frame (MDF)
z 2700 cable pairs in a feeder cable
z MDF connects

Pedestal z voice signal to telephone switch

z DSL signal to routers

Serving
area Local telephone office
interface

Distribution frame
Distribution
cable Switch

Serving
area Feeder
interface cable
zForinteresting pictures of switches & MDF, see
zweb.mit.edu/is/is/delivery/5ess/photos.html
www.museumofcommunications.org/coe.html
 Fiber-to-the-Home or
Fiber-to-the-Curve?
Table 3.5 Data rates of 24-gauge twisted pair z Fiber connection to the
home provides huge
Standard Data Rate Distance amount of bandwidth,
T-1 1.544 Mbps 18,000 feet, 5.5 km but cost of optical
DS2 6.312 Mbps 12,000 feet, 3.7 km modems still high
1/4 STS-1 12.960 4500 feet, 1.4 km z Fiber to the curve
Mbps
(pedestal) with shorter
1/2 STS-1 25.920
Mbps
3000 feet, 0.9 km distance from pedestal
to home can provide
STS-1 51.840 1000 feet, 300 m
Mbps high speeds over
copper pairs
Two- & Four-wire connections
z From telephone to CO, two wires carry signals in both directions
z Inside network, 1 wire pair per direction
z Conversion from 2-wire to 4-wire occurs at hybrid transformer in
the CO
z Signal reflections can occur causing speech echo
z Echo cancellers used to subtract the echo from the voice signals

Original Transmit pair Received signal

signal

zFour Wires Echoed Hybrid

signal transformer

zTwo Wires
Receive pair
 Integrated Services Digital
Network (ISDN)
z First effort to provide end-to-end digital connections
z B channel = 64 kbps, D channel = 16 kbps
z ISDN defined interface to network
z Network consisted of separate networks for voice, data, signaling

Circuit-
switched
network
Basic rate interface
(BRI): 2B+D
Private
BRI channel-
switched BRI
network

Packet-
PRI switched
networks PRI

Signaling Primary rate interface

network (PRI): 23B+D
 Chapter 4
Circuit-Switching
Networks
Signaling
Setting Up Connections
Manually Automatically
z Human Intervention z Management Interface

z Telephone z Operator at console sets

up connections at
z Voice commands &
various switches
switchboard operators
z Transport Networks z Automatic signaling
z Request for connection
z Order forms &
generates signaling
dispatching of
messages that control
craftpersons
connection setup in
switches
Stored-Program Control Switches
z SPC switches (1960s)
z Crossbar switches with crossbars built from relays
that open/close mechanically through electrical control
z Computer program controls set up opening/closing of
crosspoints to establish connections between switch
inputs and outputs
z Signaling required to coordinate path set up across
network
SPC
Control Signaling Message
Message Signaling
z Processors that control switches exchange signaling
messages
z Protocols defining messages & actions defined
z Modems developed to communicate digitally over
converted voice trunks

Office A Trunks Office B

Switch Switch

Processor Modem Modem Processor

Signaling
 Signaling Network
z Common Channel Signaling (CCS) #7 deployed in 1970s to control call setup
z Protocol stack developed to support signaling
z Signaling network based on highly reliable packet switching network
z Processors & databases attached to signaling network enabled many new
services: caller id, call forwarding, call waiting, user mobility
Internodal Signaling
Signaling System 7
Access Signaling
Dial tone STP STP SCP

STP STP
SSP SSP
Signaling Network

Transport Network

SSP = service switching point (signal to message)

STP = signal transfer point (packet switch)
SCP = service control point (processing)
 Signaling System Protocol Stack
z Lower 3 layers ensure
Application layer
delivery of messages to
signaling nodes
Presentation layer TUP TCAP ISUP z SCCP allows
messages to be
Session layer
directed to applications
Transport layer SCCP z TCAP defines
messages & protocols
Network layer MTP level 3
between applications
z ISUP performs basic
call setup & release
Data link layer MTP level 2
z TUP instead of ISUP in
Physical layer MTP level 1 some countries

ISUP = ISDN user part MTP = message transfer part

SSCP = signaling connection control part TCAP = transaction capabilities part
TUP = telephone user part
Future Signaling: Calls, Sessions,
& Connections
Call/Session Connection
z An agreement by two end z Allocation of resources to
parties to communicate enable information transfer
z Answering a ringing between communicating
phone (after looking at parties
caller ID) z Path establishment in

z TCP three-way telephone call

handshake z Does not apply in

z Applies in connection-less & connectionless networks

connection-oriented z ReSerVation Protocol
networks (RSVP) provides for resource
z Session Initiation Protocol reservation along paths in
(SIP) provides for Internet
establishment of sessions in
many Internet applications
Network Intelligence
z Intelligent Peripherals provide additional service capabilities
z Voice Recognition & Voice Synthesis systems allow users to access
applications via speech commands
z “Voice browsers” currently under development (See: www.voicexml.org)
z Long-term trend is for IP network to replace signaling system and provide
equivalent services
z Services can then be provided by telephone companies as well as new
types of service companies

External
Database
Signaling Intelligent
Network Peripheral
SSP
SSP

Transport Network
 Chapter 4
Circuit-Switching
Networks
Traffic and Overload Control in
Telephone Networks
Traffic Management & Overload
Control
z Telephone calls come and go
z People activity follow patterns
z Mid-morning & mid-afternoon at office
z Evening at home
z Summer vacation
z Outlier Days are extra busy
z Mother’s Day, Christmas, …
z Disasters & other events cause surges in traffic
z Need traffic management & overload control
Traffic concentration

Many Fewer
lines trunks

z Traffic fluctuates as calls initiated & terminated

z Driven by human activity
z Providing resources so
z Call requests always met is too expensive
z Call requests met most of the time cost-effective
z Switches concentrate traffic onto shared trunks
z Blocking of requests will occur from time to time
z Traffic engineering provisions resources to meet blocking
performance targets
Fluctuation in Trunk Occupancy
zNumber of busy trunks

N(t) All trunks busy, new call requests blocked

For Evaluation Only.

Copyright(C) by Foxit Software Company,2005-2008
Edited by Foxit Reader
t

zactive
1
2 zactive
Trunk number

zactive
3
zactive zactive
4
zactive
5
6 zactive zactive

7 zactive zactive
Modeling Traffic Processes
z Find the statistics of N(t) the number of calls in the system

Model
z Call request arrival rate: λ requests per second
z In a very small time interval Δ,
z Prob[ new request ] = λΔ
z Prob[no new request] = 1 - λΔ
z The resulting random process is a Poisson arrival process:

(λT)ke–λT
Prob(k arrivals in time T) =
k!

z Holding time: Time a user maintains a connection

z X a random variable with mean E(X)

z Offered load: rate at which work is offered by users:

z a = λ calls/sec * E(X) seconds/call (Erlangs)
Blocking Probability & Utilization
z c = Number of Trunks
z Blocking occurs if all trunks are busy, i.e. N(t)=c
z If call requests are Poisson, then blocking probability
Pb is given by Erlang B Formula
ac
c!
Pb = c
∑ ak
k=0 k!
z The utilization is the average # of trunks in use

Utilization = λ(1 – Pb) E[X]/c = (1 – Pb) a/c

Blocking Performance

za

To achieve 1% blocking probability:

a = 5 Erlangs requires 11 trunks
a = 10 Erlangs requires 18 trunks
Multiplexing Gain
Load Trunks@1% Utilization
z At a given Pb, the
1 5 0.20
2 7 0.29
system becomes more
3 8 0.38 efficient in utilizing
4 10 0.40 trunks with increasing
5 11 0.45
system size
6 13 0.46
7 14 0.50 z Aggregating traffic
8 15 0.53
flows to share centrally
9 17 0.53
10 18 0.56
allocated resources is
30 42 0.71 more efficient
50 64 0.78
60 75 0.80
z This effect is called
90 106 0.85 Multiplexing Gain
100 117 0.85
Routing Control
z Routing control: selection of connection paths
z Large traffic flows should follow direct route because they are
efficient in use of resources
z Useful to combine smaller flows to share resources
z Example: 3 close CO’s & 3 other close COs
z 10 Erlangs between each pair of COs
Trunk
(a) (b)
Tandem group Tandem
A D
switch 1 switch 2

B E

C F A B C D E F
10 Erlangs between each pair 90 Erlangs when combined

17 trunks for 10 Erlangs

9x17=153 trunks 106 trunks for 90 Erlangs
Efficiency = 90/153=53% Efficiency = 85%
Alternative Routing
Tandem
switch

Alternative route

Switch Switch
High-usage route

z Deploy trunks between switches with significant traffic volume

z Allocate trunks with high blocking, say 10%, so utilization is high
z Meet 1% end-to-end blocking requirement by overflowing to
longer paths over tandem switch
z Tandem switch handles overflow traffic from other switches so it
can operate efficiently
z Typical scenario shown in next slide
Typical Routing Scenario

Tandem Tandem
switch 1 switch 2

Alternative routes
for B-E, C-F

Switch D
Switch A
Switch B Switch E
High-usage route B-E
Switch C Switch F
High-usage route C-F
Dynamic Routing

Tandem Tandem Tandem

switch 1 switch 2 switch 3

Alternative routes

Switch A Switch B
High-usage route

z Traffic varies according to time of day, day of week

z East coast of North America busy while West coast idle
z Network can use idle resources by adapting route selection
dynamically
z Route some intra-East-coast calls through West-coast switches
z Try high-usage route and overflow to alternative routes
Overload Control

Overload Situations
z Mother’s Day, Xmas

z Catastrophes
Network capacity
z Network Faults
Carried load

Strategies
z Direct routes first

z Outbound first

Offered load z Code blocking

z Call request pacing

 Chapter 4
Circuit-Switching
Networks
Cellular Telephone Networks
Radio Communications
z 1900s: Radio telephony demonstrated
z 1920s: Commercial radio broadcast service
z 1930s: Spectrum regulation introduced to deal with
interference
z 1940s: Mobile Telephone Service
z Police & ambulance radio service
z Single antenna covers transmission to mobile users in city
z Less powerful car antennas transmit to network of
antennas around a city
z Very limited number of users can be supported
Cellular Communications
Two basic concepts:
z Frequency Reuse
z A region is partitioned into cells
z Each cell is covered by base station
z Power transmission levels controlled to minimize inter-cell
interference
z Spectrum can be reused in other cells
z Handoff
z Procedures to ensure continuity of call as user moves from
cell to another
z Involves setting up call in new cell and tearing down old one
Frequency Reuse
z Adjacent cells may not
2 use same band of
frequencies
7 3 z Frequency Reuse
1 Pattern specifies how
frequencies are reused
6 4 z Figure shows 7-cell
2
reuse: frequencies
5
divided into 7 groups &
2 7 3 reused as shown
z Also 4-cell & 12-cell
7 3 1 reuse possible
1 6 4 z Note: CDMA allows
adjacent cells to use
6 4 5 same frequencies
5
(Chapter 6)
 Cellular Network
Base station
z Transmits to users
on forward channels
z Receives from users
on reverse channels
BSS BSS Mobile Switching
MSC Center
STP SS7
HLR z Controls connection
VLR Wireline
EIR terminal setup within cells &
AC PSTN to telephone network

AC = authentication center MSC = mobile switching center

BSS = base station subsystem PSTN = public switched telephone network
EIR = equipment identity register STP = signal transfer point
HLR = home location register VLR = visitor location register
Signaling & Connection Control
z Setup channels set aside for call setup & handoff
z Mobile unit selects setup channel with strongest signal &
monitors this channel
z Incoming call to mobile unit
z MSC sends call request to all BSSs
z BSSs broadcast request on all setup channels
z Mobile unit replies on reverse setup channel
z BSS forwards reply to MSC
z BSS assigns forward & reverse voice channels
z BSS informs mobile to use these
z Mobile phone rings
Mobile Originated Call
z Mobile sends request in reverse setup channel
z Message from mobile includes serial # and possibly
authentication information
z BSS forwards message to MSC
z MSC consults Home Location Register for
information about the subscriber
z MSC may consult Authentication center
z MSC establishes call to PSTN
z BSS assigns forward & reverse channel
Handoff
z Base station monitors signal levels from its mobiles
z If signal level drops below threshold, MSC notified &
mobile instructed to transmit on setup channel
z Base stations in vicinity of mobile instructed to
monitor signal from mobile on setup channel
z Results forward to MSC, which selects new cell
z Current BSS & mobile instructed to prepare for
handoff
z MSC releases connection to first BSS and sets up
connection to new BSS
z Mobile changes to new channels in new cell
z Brief interruption in connection (except for CDMA)
Roaming
z Users subscribe to roaming service to use service
outside their home region
z Signaling network used for message exchange
between home & visited network
z Roamer uses setup channels to register in new area
z MSC in visited areas requests authorization from
users Home Location Register
z Visitor Location Register informed of new user
z User can now receive & place calls
GSM Signaling Standard
z Base station
z Base Transceiver Station (BTS)
z Antenna + Transceiver to mobile
z Monitoring signal strength
z Base Station Controller
z Manages radio resources or 1 or more BTSs
z Set up of channels & handoff
z Interposed between BTS & MSC

z Mobile & MSC Applications

z Call Management (CM)
z Mobility Management (MM)
z Radio Resources Management (RRM) concerns
mobile, BTS, BSC, and MSC
Cellular Network Protocol Stack
CM Um Abis A CM

MM MM

RRM RRM RRM RRM

SCCP SCCP

MTP MTP
Level 3 Level 3
MTP MTP
LAPDm LAPDm LAPD LAPD Level 2
Level 2

Radio Radio 64 64 64 64
kbps kbps kbps kbps
Base Base
Mobile station MSC
transceiver station
station controller
Cellular Network Protocol Stack
CM Um
Radio Air Interface (Um)
MM z LAPDm is data link control
adapted to mobile
RRM RRM
z RRM deals with setting up of
radio channels & handover

LAPDm LAPDm LAPD

Radio Radio 64
kbps
Base
Mobile station
transceiver
station
Cellular Network Protocol Stack
Abis
Abis Interface
z 64 kbps link physical layer

RRM RRM z LAPDm

z BSC RRM can handle

SCCP
handover for cells within its
MTP control
Level 3
MTP
LAPDm LAPD LAPD Level 2

Radio 64 64 64
kbps kbps kbps
Base Base
transceiver station
station controller
 Cellular Network Protocol Stack
CM A CM
Signaling Network (A)
MM MM
Interface
z RRM deals handover
RRM RRM RRM involving cells with
SCCP
different BSCs
SCCP
z MM deals with
MTP
Level 3
MTP mobile user location,
Level 3
MTP
authentication
LAPDm LAPD Level 2 MTP
Level 2 z CM deals with call

Radio 64 64 64 setup & release

kbps kbps
Base
kbps
using modified ISUP
Mobile station MSC
station
controller
What’s Next for Cellular
Networks?
z Mobility makes cellular phone compelling
z Cell phone use increasing at expense of telephone
z Short Message Service (SMS) transfers text using
signaling infrastructure
z Growing very rapidly
z Multimedia cell phones
z Digital camera to stimulate more usage
z Higher speed data capabilities
z GPRS & EDGE for data transfer from laptops & PDAs
z WiFi (802.11 wireless LAN) a major competitor