Transcript 10-Circuit-Packet

Switching Networks
Long-distance transmission is typically
done over a network of switched nodes
Nodes not concerned with content of data
End devices are stations

Computer, terminal, phone, etc.
A collection of nodes and connections is a
communications network
Data routed by being switched from node
to node
Switching Nodes
Nodes may connect to other nodes only
(internal), or to stations and other nodes
Node-to-node links usually multiplexed

FDM or TDM
Network is usually not fully connected
Partially connected
 Some redundant connections desirable for reliability
Two different switching technologies
 Circuit switching
 Packet switching

A Simple Switched Network
Circuit Switching
Basic Concepts
Circuit Switching
Dedicated communication path provided between two stations
Three phases

Circuit establishment
Free channel must be allocated for each leg in the route

Data transfer
Data may be analog or digital, depending upon network type
Digital transmission for voice and data becoming dominant
Typically full-duplex

Circuit disconnect
As requested by one of two stations involved
Action propagated to deallocate the dedicated resources
Must have switching capacity and channel capacity between each
pair of switching nodes on path to establish connection
Switches must have intelligence to make allocations and device
route through network
Circuit Switching - Applications
Inefficient


Channel capacity dedicated for duration of connection
When no data, capacity wasted
Setup (connection) takes time

But once connected, transfer is transparent!
Developed for voice traffic (phone)


Now widely used for data traffic
Best-known example is public telephone network
Substantial data traffic from modems
Gradually being converted to digital network

Another example is private branch exchange (PBX)
Interconnect phones within a building or office
Public Circuit-Switched Network
Telecomm Network Components
Subscriber

Devices attached to network
Local Loop


a.k.a. subscriber loop or local loop
Connection to network
Exchange



Switching centers within the network
End office – switching center supporting subscribers
Intermediate switching nodes in-between
Trunks


Branches between exchanges
Carry multiple voice-frequency circuits via FDM or
synchronous TDM
Circuit Establishment
Elements of a Circuit-Switching Node
Digital Switch

Provide transparent signal path
between devices
Network interface


Functions and hardware to connect
digital devices (computers, digital
telephones) to network
Analog phones can also be attached if
interface logic includes converter
Control logic

Establishes connections on demand of
attached device
Handle and acknowledge requests
Must determine if destination free and
construct path through switch


Maintains connections
Terminates connections at request of
attached device or for its own reasons
Blocking vs. Non-blocking
Blocking



A network is unable to connect stations because all
paths are in use
A blocking network allows this situation to occur
Used on voice systems
Tolerance when assuming short-duration calls
Non-blocking


Permits all stations to connect (in pairs) at once
Used in some data applications
Space-Division Switching
Background

Originally developed for analog environment, but carried over
into digital realm
Principles same whether switch carries analog or digital signals



Separate physical paths (divided in “space”)
Each connection requires establishment of physical path through
switch dedicated to data transfer between 2 endpoints
Basic building block is crossbar switch
Crossbar switch



Number of crosspoints grows as square of number of stations
Loss of crosspoint prevents connection
Inefficient use of crosspoints
All stations connected, yet only a few crosspoints in use

Non-blocking
Crossbar Switch
Multistage Switch
Designed to overcome crossbar limitations



Scalability
Resilience to failed crosspoints
Utilization of crosspoints
Characteristics


Reduced number of crosspoints
More than one path through network
Increased reliability



But, more complex control scheme required
May be blocking
Can be made non-blocking by increasing number
and/or size of intermediate switches
At increased cost
Three-Stage Switch
NOTE: 48 crosspoints instead of 100 in crossbar switch
Time Division Switching
modern digital systems use intelligent
control of space & time division elements
use digital time division techniques to set
up and maintain virtual circuits
partition low speed bit stream into pieces
that share higher speed stream
individual pieces manipulated by control
logic to flow from input to output
Softswitch Architecture
Recent trend in circuit-switching technology is Softswitch

General-purpose computer running software to make it a smart
phone switch
Lower costs and greater functionality in addition to
handling traditional circuit-switching functions


Packetizing of digitized voice data
Allowing voice over IP
Most complex part of telephone network switch is
software that controls call processing



Call routing
Call processing logic
Typically running on proprietary processor
Difference with Softswitch




Separate call processing from h/w switching function of switch
Physical switching performed by media gateway (MG)
Call processing performed by media gateway controller (MGC)
MG and MGC perhaps from different vendors; standard protocol
Traditional Circuit Switching
Softswitch
Packet Switching
Basic Concepts
Principles
Circuit switching designed for voice



Resources dedicated to a particular call
Fairly high utilization with talking
Shortcomings with data connections
Much of the time the line is idle

Inefficient use of resources
Data rate is fixed


Both ends must operate at same rate
Limits utility in interconnecting variety of host computers
Packet switching is far better for data
Basic Operation of Packet Switching
Data transmitted in small packets



e.g. packet length of 1000 octets
Longer messages split into series of packets
Each packet contains a portion of user data from
message, plus some control info
Control info

Routing (addressing) info
At each switching node


Each packet received, stored briefly (buffered), and
passed on to next node
Store and forward
Use of Packets
Advantages
Line efficiency


Single node-to-node link can be shared by many
packets over time
Packets queued and transmitted as fast as possible
Data rate conversion


Each station connects to local node at its own speed
Nodes buffer data if required to equalize rates
Packets accepted even when network is busy


Unlike blocking of calls in circuit switching
But, delivery may slow down (i.e. increased delay)
Priorities can be used

Useful when packets queued, as can send higherpriority packets first when output link available
Switching Technique
Station breaks long message into packets
Packets sent one at a time to the network
Packets handled in one of two approaches


Datagram
Virtual circuit
Datagram
Basic characteristics



Each packet treated independently
Packets can take any practical route
Packets may arrive out of order
Some get through faster than others

Packets may go missing
e.g. momentary crash of switching node may
cause all queued packets on it to be lost

Up to receiver to re-order packets and recover
from missing packets
Example:
Datagram
Virtual Circuit
Basic characteristics




Not a dedicated path
Preplanned route is established before any data
packets sent
First, call request and call accept packets establish
connection (handshake)
Then, each data packet contains a virtual circuit
identifier (VCI) instead of destination address
No routing decisions required for each packet

Clear request packet used to drop circuit
Example:
Virtual
Circuit
Virtual Circuit vs. Datagram
Virtual circuit

Network can provide sequencing and error control
By design, packets arrive in order
Retransmission request for missing packets

Data packets forwarded more quickly
No routing decisions to make

Less reliable
Loss of switching node loses all virtual circuits through that node
Datagram

No call setup phase
Better if few packets

More flexible
Routing with each data packet permits it to avoid congested or
failed parts of network
Packet Size
A virtual circuit from station X
through nodes a and b to station Y
Each packet contains 40 octets
and 3 octets of control information
Relationship between packet size
and transmission time

Smaller packets provide more
potential for concurrency in
network when multiple hops
Reduces msg. trans. time 

However, law of diminishing
returns applies as usual
At some point, too small a packet
starts to increase total transmission
time again 
Of course, other factors also in
play with change in packet size

e.g. smaller packets exhibit higher
ratio of overhead 
Circuit vs. Packet Switching
Delay factors in performance

Propagation delay
As we’ve seen, speed of signal through medium

e.g. 2x108 meters/sec through a wire
A function of distance and medium

Transmission time
As we’ve seen, time for transmitter to emit block of data
Clearly a function of packet size and data rate

Node delay
Processing delay of node to switch data
A function of technology and perhaps packet size
Event Timing
34
External and Internal Operation
Can consider datagrams vs. virtual circuits at two levels


Internal vs. external
They need not necessarily be the same at both levels
Interface between station and network switching node

Connection-oriented service
Station requests logical connection (virtual circuit)
All packets identified as belonging to that connection and sequentially numbered
Network delivers packets in sequence
External VC service
e.g. X.25
Different from internal VC operation

Connectionless service
Packets handled independently
External datagram service
Different from internal datagram operation

Basically, the logical behavior (VC or datagram) provided by service to upper
layer (“user”) is the external service (DG vs. VC)
External VC and Datagram
Operation inside the network fabric (i.e. cloud)
is not necessarily the same as external view.
Internal VC and Datagram
X.25
ITU-T standard for interface between host
and packet switched network
almost universal on packet switched
networks and packet switching in ISDN
defines three layers



Physical
Link
Packet
X.25 - Physical
interface between station node link
two ends are distinct


Data Terminal Equipment DTE (user
equipment)
Data Circuit-terminating Equipment DCE
(node)
physical layer specification is X.21
can substitute alternative such as EIA-232
X.25 - Link
Link Access Protocol Balanced (LAPB)


Subset of HDLC
see chapter 7
provides reliable transfer of data over link
sending as a sequence of frames
X.25 - Packet
provides a logical connections (virtual
circuit) between subscribers
all data in this connection form a single
stream between the end stations
established on demand
termed external virtual circuits
X.25 Use of Virtual Circuits
Virtual-Circuit Service
Logical connection between two stations

External virtual circuit
Specific preplanned route through network

Internal virtual circuit
Typically one-to-one relationship between
external and internal virtual circuits
Can employ X.25 with datagram-style network
External virtual circuits require logical channel

All data considered part of stream
User Data and X.25 Protocol
Control Information
Issues with X.25
key features include:



call control packets, in band signaling
multiplexing of virtual circuits at layer 3
layers 2 and 3 include flow and error control
hence have considerable overhead
not appropriate for modern digital systems
with high reliability
PSN Example: Frame Relay
Designed to be more efficient than X.25


Developed before ATM
Larger installed base than ATM
Key features of X.25



Call-control packets, in-band signaling
Multiplexing of virtual circuits at layer 3
Layer 2 and 3 include flow and error control
Weaknesses


Considerable overhead
Not particularly appropriate for modern digital
communications systems with high reliability at high
speed
Frame Relay – Key Differences
Streamlines the communications process
Call-control signaling carried on separate logical
connection from user data

Intermediate nodes in fabric relieved of much work in
maintaining state tables and processing
Multiplexing and switching is at layer 2

Eliminates one layer of processing
No hop-by-hop error or flow control

End-to-end flow and error control (if used) performed by higher
layer
Single user data frame sent from source to destination
and ACK carried back

No hop-by-hop exchanges of data and ACK frames
End-to-End vs Hop-by-hop
Advantages and Disadvantages
Lost link by link error and flow control

Increased reliability makes this less a problem
Streamlined communications process


Lower delay
Higher throughput
Better solution for higher data rates,
reliable links
Protocol Architecture
Control Plane
Between subscriber and network
Separate logical channel used

Similar to common channel signaling for circuitswitching services
Data link layer





LAPD (Q.921)
Reliable data link control
Error and flow control
Between user (TE) and network (NT)
Used for exchange of Q.933 control signal messages
User Plane
End-to-end functionality
Transfer of info between ends
LAPF (Link Access Procedure for Frame Mode
Bearer Services) Q.922








Called “LAPF Core” for minimum-function protocol
Frame delimiting, alignment and transparency
Frame mux and demux using addressing field
Ensure frame is integral number of octets (zero bit
insertion/extraction)
Ensure frame is neither too long nor short
Detection of transmission errors
Congestion control functions
LAPF Core is similar to LAPB, LAPD, HDLC, etc.
 Except no control field!
Frame Relay Data Link
Connections
logical connection between subscribers
data transferred over them
not protected by flow or error control
uses separate connection for call control
overall results in significantly less work in
network
User Data Transfer
only have one frame type which

carries user data
no control frames means


no inband signaling
no sequence numbers
flag and FCS function as in HDLC
address field carries DLCI
DLCI (Data Link Connection Identifier) has
local significance only
User Data Transfer
The information field carries higher-layer data.
If the user selects to implement additional data
link control functions end to end

then a data link frame can be carried in this field.
A common selection will be to use the full LAPF
protocol (known as LAPF control protocol), to
perform functions above the LAPF core functions.
Note that the protocol implemented in this fashion
is strictly between the end subscribers and is
transparent to the frame relay network.
55
LAPF-Core Formats
56
LAPF-Core Address Formats
The address field has a default length of 2 octets

may be extended to 3 or 4 octets.
It carries a data link connection identifier (DLCI) of 10, 16,
or 23 bits.
The DLCI serves the same function as the virtual circuit
number in X.25
It allows multiple logical frame relay connections to be
multiplexed over a single channel.
As in X.25, the connection identifier has only local
significance:

Each end of the logical connection assigns its own DLCI from the pool of
locally unused numbers, and the network must map from one to the other.
The alternative, using the same DLCI on both ends, would
require some sort of global management of DLCI values.
LAPF-Core Address Formats
The length of the Address field, and hence
of the DLCI, is determined by the Address
field extension (EA) bits.
The C/R bit is application specific and not
used by the standard frame relay protocol.
The remaining bits in the address field have
to do with congestion control
58