UNIVERSITY OF LAGOS, AKOKA, LAGOS

FACULTY OF ENGINEERING

DEPARTMENT OF ELECTRICAL AND ELECTRONICS

ENGINEERING

TERM PAPER

ON

PACKET SWITCHING

BY

EZE CHINEDUM 179043023

LECTURER IN CHARGE: DR. OGBOI F. L

NOVEMBER, 2018.

1
 ABSTRACT

Packet switching is the basis for the Internet Protocol (IP). In packet switching, information

flows are broken into variable-sized packets. These packets are sent, one by one, to the nearest

router, which will look up the destination address, and then forward them to the corresponding

next hop. This process is repeated until the packet reaches its destination.

The routing of the information is thus done locally, hop-by-hop. Routing decisions are

independent of other decisions in the past and in other routers; however, they are based on

network state and topology information that is exchanged among routers using BGP, IS-IS or

OSPF. The network does not need to keep any state to operate, other than the routing tables. The

forwarding mechanism is called store and-forward because IP packets are completely received,

stored in the router while being processed, and then transmitted. Additionally, packets may need

to be buffered locally to resolve contention for resources.

If the system runs out of buffers, packets are dropped. With the most scheduling policies, such as

FCFS and WFQ, packet switching resources have contention when they have more

arrivals/requests that what they can process. Two examples are the outgoing links and the router

interconnects.

2
CONTENTS

ABSTRACT....................................................................................................................................2

1. INTRODUCTION....................................................................................................................4

2. TYPES OF PACKET SWITCHING....................................................................................5

3. PACKET FORMATS............................................................................................................8

4. PACKET SWITCHING PROTOCOL................................................................................9

5. PACKET SWITCHING MODE OF OPREATION:........................................................10

6. ROUTING IN PACKET NETWORK...............................................................................10

7. OTHER ROUTING APPROACHES.................................................................................12

7.1 FLOODING...................................................................................................................12

7.2 DEFLECTING ROUTING..........................................................................................14

7.3 SOURCE ROUTING....................................................................................................14

8. TRAFFIC MANAGEMENT AND QoS.............................................................................15

8.1. FIFO....................................................................................................................................16

8.2. HEAD-OF-LINE (HOL) PRIORITY QUEUEING............................................................16

9. ADVANTAGES OF PACKET SWITCHING...................................................................17

10. DISADVANTAGE OF PACKET SWITCHING...........................................................18

11. SUMMARY.......................................................................................................................18

REFERENCES..............................................................................................................................18

3
 1. INTRODUCTION

Traditional telephone network operate on the basis of circuit switching. A call setup process

reserves resources (time slot) along a path so that the stream of voice sample can be transmitted

with very low delay across the network. The resources allocated to the user can’t be used by

other users for the duration of the call. This approach is inefficient when the amount of

information transferred is small or if the information is produced in burst, as is the case in many

computer applications. In this paper we examine networks that transfer block of information

called packets. Packet switching network are better matched to computer application and can

also be designed to support real-time application such as telephony (Halabi, 1997).

Packet switching is similar to message switching using short messages. Any message exceeding

a network-defined maximum length is broken up into shorter units, known as packets, for

transmission; the packets, each with an associated header, are then transmitted individually

through the network. The performance of Packet Switching is called Best Effort performance. If

you transmit from sender to receiver, the entire network will do its best to get the packet to the

other end as fast as possible, but there are no guarantees on how fast that packet will arrive.

(Stevens, January 1997):

2. TYPES OF PACKET SWITCHING

There are two common packet switching methods in use:

4
 Datagram Packet Switching: This approach is connectionless and does not involve prior

allocation of resources. Each packet is forwarded independently from switch to switch

based on the destination address until the packet of information arrives its destination.

Routing decisions are made dynamically, so each packet may follow a different route and

thus the packages may arrive out of order and a message it sent back for its resen

(Telecom ABC, 2019) (Halabi, 1997)t.

In the datagram or connectionless packet switching approeach, each packet is routed

independently through the network. Each packet has an attached header t5hat provides all

the information required to route the packet to its destination. When a packet arrives at

packet switch, the destination address in the header is examined to determine the next

hop in the path to the destination. The packet is then placed in the queue to wait until the

given transmission line becomes available. By sharing the transmission line among

multiple packets, packet switching can achieve high utilization at the expense of packet

queuing delays. We note that routers in the internet are packet switch that operates in

datagram mode. (Telecom ABC, 2019)

Because each packet is routed independently, packets from the same source to the same

destination may transverse different paths through the network as shown in fig 1. For

example the route may change in response to a network fault. Thus packets may arrive

out of order; re-sequencing may be required at the destination.

5
 Fig 1; Packet Switching: Datagram approach.

 Virtual Circuit Packet Switching: This approach is connection-oriented; it involves

setting up a connection across the network before the information can be transferred. The

setup procedures typically involve the exchange of signaling messages and the allocation

of resources along the path from the input to the output for the duration of the connection.

A route is set up prior to packets being sent. The packets will all follow this route. This

makes the routing through the network very easy and the packages will be received in the

correct order. Both approaches involve the use of switches or routers to direct packet

across the network. As in circuit switching, the call setup procedure usually takes place

before any packet can flow through the network as shown in fig 2

6
 .

Fig 2: Fig: Packet switching: virtual circuit approach

The connection setup procedure establishes a path through the network and then set

parameters in the switch along the path as shown in fig 3. The controller/processor in

every switch is involved in the exchange in the exchange of signaling message to set up

the path. As in the datagram approach, the transmission link are shared by packets in

many flows. All packet for the connection then follow the same path.

In datagram packet switching each must contain the full address of the source and

destination. In large networks, this address may require a large number of bits and result

in significant packet overhead and hence waste transmission bandwidth. One advantages

of virtual-circuit packet switching is that abbreviation header can be used.

7
 Fig 3. Delay in Virtual-circuit packet switching

The connection setup procedure establishes a number of entries in forwarding tables

located in the various switches along the path. At the input to every switch, the

connection is identified by a virtual-circuit identifier (VCI). When a packet arrives at an

input port, the VCI in the header is used to access the table. The table lookup provides

the output port to which the packet is to be forwarded and the VCI that is to be used at the

input port of the next switch. Thus the call setup procedure sets up a chain of pointer

across the network that directs the flow of packets in a connection (Huitema C., 1995).

The table entry for a VCI can also specify the type of priority that is to be given to the

packet by the scheduler that controls the transmissions in the next output port. (Stevens,

January 1997)

3. PACKET FORMATS

A packet contains three major fields: The header, the message, and redundancy check bits. Most

popular technique uses cyclic redundancy checks (CRCs) CRC is nothing more than a set of

parity bits that cover overlapping fields of message bits. CRC can detect small number of errors.

8
A header typically contains numerous subfields in addition to the necessary address field.

(Telecom ABC, 2019)

HEADER MESSAGE CRC

FROM TO OP.CODE SEQ BYTE

ADDRESS ADDRESS COUNT

Fig 4: typical packet format

4. PACKET SWITCHING PROTOCOL

Each data packet is transmitted individually and can even follow different routes to its

destination. After all the packets forming, the message arrive at the destination is recompiled into

the original message. (Huitema C., 1995)

Examples of packet switching protocol are:

 Modern wide area network (WAN)

 TCP/IP

 X.25, and frame relay

5. PACKET SWITCHING MODE OF OPREATION:

example a 2MB file would be broken up into chunks of 512 bytes in size.

 Before transmission the packet contains the ‘header’ that has network IP address that it

needs to arrive at and also details of the IP address from which it was sent.

 Moreover the header also allots a number to each packet and records how many packets

the data was split up into.

 After leaving the computer the packets start to head off in different directions taking the

least busy path at that instant.

 The router figures out which is the next fastest connection and sends each packet on its

way. This technique works extremely well, because if one branch gets busy then the

packets are automatically routed through another path.

 As on the transmitting side the packets were given a number so when the packets arrive at

their destination, they are put back together again in the right order before its finally sent

to the receiver’s side. (Zhang H, October 1995)

6. ROUTING IN PACKET NETWORK

A packet network consists of nodes (routers or switches) interconnected by communication links

in an arbitrary mesh like fashion as shown in fig 3. As suggested by the figure, a packet could

take several possible paths from host A to host B. For example, three possible paths are 1-3-6, 1-

4-5-6 and 1-2-5-6. However which path is the best one? Here the meaning of the best depends on

the objective function that the network operator tries to optimize. If the objective is to minimize

10
the number of hops, then path 1-3-6 is the best. If each link incurs a certain delay and the

objective function is to minimize the end-to-end delay, then the best path is the one that give the

end-to-end minimum delay. Yet a third objective function is selecting the path with the greatest

available bandwidth. The purpose of the routing algorithm is to identify the set of path that are

best in a sense defined by the network operator. Note that a routing algorithm must have global

knowledge about the network state in order to perform its task. (Halabi, 1997) (Jacobson, august

1988)

Fig 5: An example of a packet-switching network.

The main ingredient of a good routing algorithm depends on the objective function that one is

trying to optimize. However in general a routing algorithm should seek one or more of the

following goals (Stevens, January 1997):

1. Rapid and accurate delivery of packets: A routing algorithm must operate correctly; that

is it must be able to find a route to the destination if it exists. In addition, the algorithm

should not take an unreasonably long time to find the route to the destination.

2. Adaptable to change in network topology resulting from node or link failure: In real

network equipment and transmission lines are subjected to failures. Thus a routing

11
 algorithm must be able to adapt to this situation and reconfigure the routes automatically

when equipment fails.

3. Adaptability to varying source-destination traffic loads: Traffic loads are quantities that

are changing dynamically. In a period of 24 hours, traffic loads may go into cycles of

heavy and light periods. An adapting routing algorithm would be able to adjust the routes

based on the current traffic loads

4. Ability to route packets away from temporarily congested links: A routing algorithm

should try to avoid heavy congestion links. Often it is desirable to balance the load on

each link.

5. Ability to determine the connectivity of the network: To find optimal routes, the routing

system needs to know the connectivity or reachability information.

6. Low overhead: A routing system typically obtains the connectivity information by

exchanging control messages with other routing systems. These messages represent an

overhead that should be minimized (Zhang H, October 1995).

7. OTHER ROUTING APPROACHES

In this section we look at three routing approaches.

7.1 FLOODING

The principle of flooding calls for a packet switch to forward an incoming packet to all

port except the one the packet was received from. If each switch performs this flooding

process, the packet will eventually reach the destination. Flooding is a very effective

12
routing approach when the information in the routing tables is not available, such as

during system startup or when survivability is required, such as in military networks.

However, flooding may easily swamp the network as one packet creates multiple packets

that in turn create multiples of multiple packets, generating an exponential growth rate as

illustrated in fig 4. Initially one packet arrives at node 1 triggers three packet to node 2,3

and 4. In the second phase node 2,3 and 4 send two, two and three packets respectively.

These packets arrive at nodes 2 through 6. In the third phase 15 more packets are

generated, giving a total of 25 packets after three phases. Clearly, flooding needs to be

controlled so that packets are not generated excessively. To limit such a behavior one can

implement a number of mechanisms. (Jacobson, august 1988)

 One simple method is to use a time-to-live field in each packet. When the source

sends a packet, the time-to-live field is initially set to some small number (say, 10

or smaller). Each switch decrements the field by one before flooding the packet.

If the value reaches zero, the switch discards the packet. To avoid unnecessary

waste of bandwidth, the time –to-live should ideally be set to the minimum hop

number between two further nodes (called the diameter of the network). In fig 4

the diameter of the network is two. To have a packet reach any destination, it is

sufficient to set the time-to-live to field to two. (Hashemi, November 1997)

 In the second method, each switch adds its identifier to the header of the packet

before it floods the packet. When a switch encounters a packet that contains the

identifier of the switch, it discards the packet. This method effectively prevents a

packet from going around a loop (Stevens, January 1997).

13
  The third method is similar to the second method in that they both try to discard

old packets. The only difference lies in the implementation. Here each packet

from a given source is identified with a unique sequence number. When a switch

receives a packet, the switch records the source address and the sequence number

of the packet. If the switch based on the stored source address and sequence , it

will discard the packet.

Fig 6: Flooding is initiated in node 1

7.2 DEFLECTING ROUTING

This approach requires the network to provide multiple paths for each source-destination

pair. Each switch first tries to forward a packet to the preferred port. If the preferred port

is busy or congested, the packet is deflected to another port

7.3 SOURCE ROUTING

14
 Is a routing approach that does not require an intermediate node to maintain a routing

table, but rather puts more burdens at the source hosts. Source routing works in either in

datagram or virtual-circuit packet switching. Before a source host can send a packet the

host has to know the complete route to the destination host in order to include the route

information in the header of the packet. The route information contains the sequence of

nodes to transverse and should give the intermediate node sufficient information to

forward the packet to the next node until the packet reaches the destination. Figure 5

shows how source routing works. (Jacobson, august 1988) (Huitema C., 1995)

Fig 7: Example of source routing

8. TRAFFIC MANAGEMENT AND QoS

Traffic management is concerned with the delivery of QoS to specific packet flows. Traffic

management entails mechanisms for managing the flows in a network to control the load that is

applied to various link and switches. Traffic management also involves the setting of priority and

scheduling mechanisms at switches, routers and multiplexers to provide differentiated treatment

for packets and cells belonging to different classes, flows or connections. It also may involve the

policing and shaping of traffic flows as they enter the network. (Jacobson, august 1988)

15
8.1. FIFO

The simplest approach to managing a multiplexer involves First-in, First-out (FIFO) queueing

where all arriving packet are placed in a common queue and transmitted in order of arrival as

shown in fig 8a. Packets are discarded when they arrives at a full buffer. The delay and loss

experience by packets in a FIFO system depends on the inter-arrival times and on the packet

lengths. AS inter-arrivals become bustier or packet length more variable, performance is not

possible to provide different information flows with different qualities of service. FIFO systems

are subjected to hopping, which occurs when a user sends packets at a high rate and fills the

buffer in the system, thus depriving other users of access to the multiplexer (Stevens, January

1997).

Fig 8(a): FIFO queueing. (b) FIFO queueing with discharge priority.

8.2. HEAD-OF-LINE (HOL) PRIORITY QUEUEING

The second approach involves defining a number of priority classes. As shown in fig 8b, each

time the transmission line becomes available the next packet for transmission is selected from the

16
head of the line of the highest priority queue that is not empty. For example packets requiring

low delay may be assigned a high priority, whereas packets that are not urgent may be given

lower priority. The size of the buffer for the different priority classes can be selected to meet

different levels of service to the different classes, it still has shortcomings. For examples it does

not allow for providing some degree of guaranteed access to transmission bandwidth to the lower

priority classes. Fairness problem may arise here when a certain users hogs the bandwidth by

sending an excess number of packets (Zhang H, October 1995).

Fig 9. HOL priority queueing.

9. ADVANTAGES OF PACKET SWITCHING

 Line efficiency: There is better use of communication lines. 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 the local node at its own speed.

Nodes buffer data if required to equalize rates

17
  Packets are accepted even when network is busy: Delivery may slow down but

priorities can be used (Telecom ABC, 2019)

 Priority scheme are implemented to reduce congestion on the network

 Packets are rearranged when received and are delivered in order.

10. DISADVANTAGE OF PACKET SWITCHING

 There is quality of service issue because the links are shared by many packets.

This will bring congestion and packet may drop once it elapses the time-to-leave

(TOL)

 It is complex because it needs to add address in switching and routing till it gets to

its destination.

11. CONCLUSION

Packet-switched networks were originally designed to overcome the inherent weakness of

sending data over the analogue circuit switched network. Circuit switching is not very efficient

for small messages due to the lack of efficiency in bandwidth utilization and the analogue

circuits make the data subject to noise and errors. The biggest packet switched network is the

internet. The internet uses the datagram packet switching method X.25 which is based on virtual

packet switching. (Telecom ABC, 2019)

Packet switching is similar to message switching using short messages. Any message exceeding

a network-defined maximum length is broken up into shorter units, known as packets, for

18
transmission; the packets, each with an associated header, are then transmitted individually

through the network. The performance of Packet Switching is called Best Effort performance. If

you transmit from sender to receiver, the entire network will do its best to get the packet to the

other end as fast as possible, but there are no guarantees on how fast that packet will arrive.

19
 REFERENCES

1. Bertsekas, D. a. (1992). Data Networks. prentice-Hall, Englewood Cliffs, New Jersey.

2. Halabi, B. (1997). Internet Routing Architectures. Indianapolis, indiana: New Rider

publishing .

3. Hashemi, M. a.-G. (November 1997). implementing and scheduling schame using a

sequence circuit . Dallas: SPIE .

4. Huitema C. (1995). Routing in the internet . New Jersey : computer communication

review.

5. Jacobson, V. (august 1988). congestion aviodance and control. New York.

6. Stevens, W. (January 1997). Fast retransmit and recovering algorithm. RFC 2001.

7. Telecom ABC. (2019). telecomabc.com. Retrieved OCT 20, 2018, from

http://www.telecomabc.com/p/packet-switching.html

8. Zhang H, .. (October 1995). Service discipline and Guarranteed Performance in Packet

Switching Networks. chencheng: proceding of IEEE.

20