Tài liệu Lecture 26: Switching pdf

Virtual circuit switching - a compromise between circuit switching and packet switching.. Crossbar switches: circuit switching A crossbar or crosspoint switch is shown in Fig.2-38, Tane

(Ref: Ch-9 of your Text book)

Three different switching techniques are used inside the telephone system:

1 Circuit switching

2 Message switching (using store-and-forward network)

2.1 Packet switching

3 Virtual circuit switching - a compromise between circuit switching and packet

switching The actual service offered is connection oriented, but it is implemented internally with packet switching

Circuit switching:

In circuit switching, a physical connection established by the switching mechanism

Once a call has been set up, a dedicated path between both ends exists and will continue to

exist until the call is finished

Fig 2-34, 2-35 Tanenbaum

An important property of circuit switching is the need to set up an end-to-end path before any

data can be sent

As a result of the established path, there is no danger of congestion, but wastes bandwidth

Message switching:

In message switching no physical path is established in advance between sender and receiver

Instead, it is sent and stored in the first switching office (i.e., router) and then forwarded later,

one hop at a time Each block is received in its entirety, inspected for errors, and then

retransmitted

Telegrams were message switching

This technique needs routers with big buffers and also a single block may tie up a

router-router line for minutes, useless for interactive traffic Packet switching gets round this problem

Packet switching:

In Packet switching no physical path is established in advance between sender and receiver

Instead, it is sent to the first switching office (i.e., router) and then forwarded to the next, one

hop at a time Each block is received in its entirety, inspected for errors, and then

retransmitted The only difference from the message switching is that, the block size is small,

so no need to store, packets can be buffered in router's main memory instead of on disk

This limit on the packet size makes sure that no user can monopolize any transmission line for

a long time And better for interactive traffic compared to message switching

Item Circuit switched packet switched

Potentially wasted bandwidth

Yes No

Store-and-Forward transmission

No Yes

Each packet follows the same route

When can congestion occur?

At setup time On every packet

The switch hierarchy:

The routing between switches in the current circuit-switched telephone system is done through the five classes of switching offices

It can be 5-4-5 (end office - toll office - end office)

It can be 5-4-3-4-5 (end office - toll office - primary office - toll office - end office)

It can be 5-4-3-2-3-4-5 (end office - toll office - primary office - sectional office - primary

office - toll office - end office)

It can be 5-4-3-2-1-2-3-4-5 (end office - toll office - primary office - sectional office - regional

office - sectional office - primary office - toll office - end office) Fig.4-5, Fitzgerald

Fig.2-37, Tanenbaum

To avoid congestion between busy routes, telephone companies simply install direct trunks

How switches work inside?

Crossbar switches: (circuit switching)

A crossbar or crosspoint switch is shown in Fig.2-38, Tanenbaum Here an input and output line may

be connected by a semiconductor switch

For 1000 lines of inputs and 1000 lines of output we need 1000,000 [= n x n] crosspoints It is possible

in a VLSI chips but 2000 pins on the chip is not This is the problem with a crossbar switch is that the

number of crosspoints grows as the square of the number of lines into the switch

The solution is space division switch

A Crossbar switch

Space Division Switch: (circuit switching)

By splitting the crossbar switch into small chunks and interconnecting them, it is possible to build multistage switches with fewer lines and cross points These are called space division switches

A Crossbar switch with fewer lines

A Space Division switch with fewer lines Fig.2-39, Tanenbaum

In the space division switch with three stages the number of cross points = 2kN + k(N/n)2

For N=1000, n=50 and k=10, we need only 24,000 crosspoints instead of 1000,000 required by a

1000x1000 single stage crossbar

Stage 2 has a k(N/n) = 10(1000/50) = 200, so a maximum of 200 calls can be connected at once

The larger k is, the more expensive the switch and the lower the blocking probability When k = 2n-1, the switch will never block

Theoretically a switch with fewer k can block, but in practice it rarely blocks

With time division switching, the n input lines are scanned in sequence to build up an input frame with n slots Each slot has k bits

For T1 switches, the input lines are 24, the slots are 8 bits, with 8000 frames processed per second That results the speed of a T1 circuit in 1.5 Mbps [=(24x8+1)x8000 = 193x8000]

Fig.2-40, Tanenbaum

• The heart of the time division switch is the time slot interchanger, which accepts input frames and produces output frames in which the time slots have been reordered

In this example, the switch has moved a byte from input line 4 to output line 0, and so on, that is how the switching happens

The whole arrangement is a circuit switch, even though there are no physical connections

• Time division switches use mapping tables that are linear in the number of lines, rather than

quadratic

Time division switches use RAM buffers to reorder the slots, so the number of lines a switch can handle is limited by the memory access time, for 100ns it is 625 lines (2nT=125)

It is possible to device multistage switches to overcome the limitations

ISDN Services: (circuit switching)

The circuit switched telephone system was designed for analog voice transmission and is inadequate for modern communication needs

The world's telephone companies got together under the auspices of CCITT and agreed to build a new,

fully digital, circuit-switched telephone system, called ISDN (Integrated Services Digital Network),

goal was to the integration of voice and nonvoice services

Services:

• Couple of instant call setups,

• Caller's Telephone number, name and address display,

• Call forwarding,

• Call conference

• Electricity meter reading, medical, burglar and smoke alarms

ISDN System Architecture:

The key idea behind ISDN is that of the digital bit pipe, support multiple independent channels by time

division multiplexing

Two principal standards for the bit pipe have been developed,

1 A low bandwidth for home use and

2 A higher bandwidth for business use

Basic rate: 2B (64kbps digital PCM) + 1D (16kbps digital)

Primary rate: 23B + 1D (US & Japan, to match T1, 1.54Mbps) or

30B + 1D (Europe, to match 2.048Mbps) Hybrid: 1A (4kHz analog) + 1C (8 or 16kbps digital)

Prospect of ISDN:

The standardization process took years to finalize by this time technology moved fast and made it obsolete

Offering 64kbps service to business in the 1980s was a serious proposition, in the 1990s it is a joke Many Internet service providers support 144kbps (2B+1D) over a fully digital link

Broadband ISDN and ATM: (packet switching)

ISDN did not work, so CCITT proposed another standard, the broadband ISDN (B-ISDN), basically a

digital virtual circuit for moving fixed-size packets (cells) at 155 Mbps

• Broadband ISDN is based on ATM (Asynchronous Transfer Mode) technology

• ATM is fundamentally a packet-switching technology In contrast, both the existing PSTN and

narrowband ISDN are circuit-switching technologies

• B-ISDN can not be sent over existing twisted pair wiring for any substantial distance

• Space division and time division switches can not be used for packet switching

• The telephone companies will adopt this technology sooner otherwise some body will, like cable television companies

Virtual Circuit: (a compromise between circuit switching and packet switching)

The basic B-ISDN service is a compromise between pure circuit switching and pure packet switching The actual service offered is connection oriented, but it is implemented internally with packet

switching

Two kinds of connections are offered:

1 Permanent virtual circuits - typically remain in place for months or years (always hold table

entries)

2 Switched virtual circuits - set up dynamically as needed like telephone calls (dynamic table

entries)

In a virtual circuit network, like ATM, when a circuit is established, what really happens is that the

route is chosen from source to destination, and all the switches (routers) along the way make table entries so they can route any packets on that virtual circuit

Fig.2-43, Tanenbaum

Evolved from standardization process of ISDN Frame relay is designed to eliminate overhead that X.25 packet-switching network imposes on end user systems

Frame relay can best be thought of as a virtual leased line compared to actual leased line of ISDN The difference between an actual leased line and a virtual leased line is that with an actual one, the user can send traffic all day long at the maximum speed With a virtual one, data bursts may be sent at full speed, but the long-term average usage must be below a predetermined level In return, the carrier charges much less for a virtual line than a physical one

Transmission in ATM Networks:

ATM stands for Asynchronous Transfer Mode This can be contrasted with the synchronous T1 carrier

in Fig.2-44

One T1 frame is generated precisely every 125 μsec This rate is governed by a master clock Slot k of each frame contains 1 byte of data from the same source T1 is synchronous

In ATM, cells arrive randomly from different sources ATM does not standardize the format for

transmitting cells Cells may be encased in a carrier such as T1, T3, SONET

Primary ATM rate is 155.52 Mbps, the additional rate is 622.08 Mbps (4 times the primary rate) The transmission medium for ATM is normally fiber optics and can run many kilometers, but coax and category 5 twisted pair are also acceptable for runs under 100 meters

Each point-to-point link is unidirectional For full-duplex operation, two parallel links are needed

ATM Switches:

The general model for an ATM cell switch is shown in Fig.2-45 Tanenbaum

Cells arrive at ATM speed about 150 Mbps That means roughly 360,000cells/sec (150000000/424), which means the cycle time of the switch has to be about 2.7 μsec The cells are 53 bytes (424 bits)

A switch with 1000 input lines has to process 1000 cells every 2.7 μsec This small cell size makes it possible

All ATM switches has two common goal:

1 Switch all cells with as low a discard rate as possible - it is permitted to drop a cell in

emergencies (1 in 1012)

2 Never reorder the cells on a virtual circuit (cells arriving in certain order must depart in that order)

To reduce the collision as a result the discard rate queues are used in the input or output

There are different ATM switch designs The Batcher-Banyan Switch is one of them In figure 2.49 (Tanenbaum) we have an 8x8 three-stage banyan switch

Routing is done by looking up the output line for each cell (based on the virtual circuit information and the routing tables) A three-bit binary number is then put in front of the cell, as it will be used for routing through the switch

Each of the 12 switching elements in the banyan switch has two inputs and two outputs

When a cell arrives at a switching element, 1 bit (MSB) of the output number is inspected, and based

on that, the cell is routed either to port-0 (the upper one) or port-1 (the lower one) The middle bit is used to route through the stage-2 switching elements and finally the least significant bit (LSB) is used

to route through the stage-3 switching elements

Control Signaling:

In a circuit -switched network, control signals are the means by which the network is managed and by which calls are established, maintained, and terminated

Both call management and overall network management requires that information be exchanged

between subscriber and switch, among switches, and between switch and network management center Examples of control signals are: dial tone, ringing tone, busy tone, etc

Common Channel Signaling:

Traditional control signaling in circuit-switched networks has been on an inchannel basis With

inchannel signaling, the same channel is used to carry data and control signals

There are two different forms of inchannel signaling:

1 inband - transmit control signals in the same band of frequencies used by the voice signals

2 out-of-band - transmit control signals over the same facilities as the voice signal but a different

part of the frequency band

Problems associated with inchannel signaling:

With inband signals - The voice channel being used is only available for control signals when

there are no voice signals on the circuit

With out-of-band signals - A very narrow bandwidth is available for control signals

In Common Channel Signaling these problems has been overcome by transmitting control signals

over dedicated channels and are common to a number of voice channels

e.g., Signaling System No.7 (SS7) was designed with ISDN in mind