Unit 3 Q.

Circuit Switching: Concept: Circuit

Q. Network layer services: The duty of the switching is a method of implementing a
network layer in TCP/IP is to provide the telecommunications network in which two
host-to-host delivery of datagrams. In this network nodes establish a dedicated
section we are going to discuss the services communication channel (Circuit) before the
that are expected from the network layer. At nodes communicate with each other.
the sending end, the network layer will Example: The simplest and the oldest
encapsulate the packet into datagram and telephone network is a circuit switched
will deliver the packet to the data link layer. network. In a telephone network, when a call
At the destination, exactly opposite process is made, from one telephone to the other, the
takes place. That means, at the destination switches within the telephone exchanges
the received datagram is decapsulated to creates a continuous circuit (wired) between
extract the packet from it and the packet is the two telephones as long as the call lasts.
delivered to the transport layer. Circuit switching is used in public telephone
1. Packetizing: Packetizing is the first duty networks. It was developed to handle voice
of the network layer in which it encapsulates traffic but it can also handle digital data.
the payload (data received from the. Using the circuit switching, a dedicated path
transport. layer) in a packet at network layer is established between two stations for
at the source. Then at the destination the communication.
decapsulation process takes place. 2. Packet Switching: In the previous section
2. Finding the logical address of the next we have discussed about routing and
hop: The datagram prepared with forwarding which implies that some kind of
packetizing contains the source and switching takes place at the network layer. A
destination addresses of the packet. The router is basically a switch which connects its
datagram is to be delivered to the next router. input. port with an output port or a number of
destination addresses in the datagram do not output ports. The two types of switching
give any information about the logical techniques used in data communication are:
address of the next hop. 1. Circuit switching and, 2. Packet switching.
3. Finding MAC address of next hop: Note But at the network layer, only the packet
that it is the duty of data link layer (and not switching is used because at this layer the unit
of network layer) to actually deliver the of data is packet. The data received from the
datagram to the next hop. 4. Fragmentation: upper layers is divided into packets of
The datagram at this stage may not always be manageable size and these packets are then
ready to be given to the data link layer. The sent through the network, one by one
LANs and WANs can carry the data of a sequentially.The destination computer
limited size in a frame. data is longer than receives these packets one
maximum size for LANs and WANs
Q. IPV4 Addresses: Each computer IP (Internet Protocol) is IPv4 whereas the
connected to the Internet should be identified advanced version is IPv6. The IPv4 address is
uniquely. The identifier used for this purpose a 32-bit address and it is used for defining the
is called as the Internet address or IP address. connection of a host or router to the Internet.
The hosts and routers on the Internet have Thus an IP address is an address of the
unique IP addresses. The currerit version of interface.
Q. Internet Protocol Version 4 (IPv4): We Q. IPv6 Packet & Protocol:
have already discussed the addressing 40 bytes upto 65,535 bytes
mechanism, delivery and forwarding for the Base header Payload
IP packets. Now we will discuss the format Each packet can be divided into two parts viz:
of IP packet in the next few sections. In the base header and payload. Base header is the
discussion we will see that an IP packet mandatory part and payload is an optional
consists of a base header and options which one. The payload follows the base header.
are sometimes useful in controlling the The payload is made up of two parts:
packet delivery. 1. An optional extension header and, 2. The
Position of IP: The main protocols upper layer data. The base header is 40 byte
corresponding to the network layer in the long whereas the payload consisting of the
TCP/IP suite as well as Internet layer are: extension header and upper layer data can
ARP, RARP, IP, ICMP and IGMP.Out of have information worth upto 65, 535 bytes.
these protocols IP is the most important Base header:
protocol. It is responsible for host to host 1. Version (VER): The contents of this 4 bit
delivery of datagrams from a source to field defines the version of IP such as IPv4 or
destination. IPv6. If VER = 6, then the version is IPv6.
Internet Protocol (IP): The Internet 2. Priority: This 4 bit field contents defines
Protocol is the host to host delivery protocol the priority of the packet which is important
which belongs to the network layer and is in connection with the traffic congestion.
designed for the Internet. IP is used as the 3. Flow label: It is a 24 bit (3 byte) field
transmission mechanism by the TCP/ IP which is supposed to provide a special
protocols. That means the TCP or UDP handling for a particular flow of data.
packets are encapsulated in the IP packet and 4. Payload length: The contents of the 16 bit
the IP carries it from source to destination. IP or 2 byte length field are used to indicate the
is a connectionless datagram protocol with total length of the IP datagram excluding the
no guarantee of reliability. base header. That means it gives the length of
Datagrams: A datagram has two parts only the payload part of the datagram.
namely the header and data as shown. The 5. Next header: It is an 8 bit field which
length of datagram is not fixed. It varies from defines the header which follows the base
20 bytes to 65536 bytes. The length of the header in the datagram.
header is 20 to 60 bytes. The information 6. Hop limit: Contents of this 8 bit (1 byte)
necessary for the routing and delivery of the field have the same function as TTL (time to
datagram has been stored in the header. The live) in IPv4.
other part of the datagram is the data field 7. Source address: It is a 16 byte (128 bit)
which is of variable length. It is a custom in Internet address which corresponds to the
TCP/IP to show the header in 4-byte (32 bit) originator or source which has produced the
sections. datagram.
VER4bit Hlen Service Total length VER PRI Flow table
Ident cation16 Flag 3 bits Fragmentation Payload lengrh Next Header Hop limit
Source IP address Source address
Destination IP address Destination address
Options + Padding ( 0 – 40 bytes) Payload extension header + data packet from Up
Fig, IPv4 datagram Fig, IPv6 datagram
 Unit 4 Q. Types of Routing Algorithms: Routing
Q. Unicast and Multicast Routing: Routing algorithms can be divided into two groups.
can be broadly classified into three types: 1. Non-adaptive algorithms: For this type of
1. Unicast routing: In unicast routing there algorithms, the routing decision is not based
is a one to one relation between the source on the measurement or estimation of current
and the destination. That means only one traffic and topology. However the choice of
source sends packets to only one destination the route is done in advance, off- line and it is
The type of source and destination addresses downloaded to the routers. This is called as
included in the IP datagram are unicast static routing.
addresses assigned to the hosts. The concept 2. Adaptive algorithms: For these
of unicast routing is illustrated in algorithms the routing decision can be
1). Metric: A metric is defined as the cost changed if there are any changes in topology
assigned for passing through a network. or traffic etc. This is called as dynamic
2). Interior and exterior routing: An Internet routing. In the following sections we are
is so large that for one routing protocol it is going to discuss various static and dynamic
impossible to handle the task of updating the algorithms.
routing tables of all the routers. So an Q. Distance Véctor Routing Algorithm: In
Internet is divided into a number of this algorithm, each router maintains a table
Autonomous Systems (AS). An AS is group called vector, such a table gives the best
of networks and routers. known distance to each destination and the
3). Interior routing: The routing that takes information about which line to be used to
place inside an AS is called as Interior reach there. This algorithm is sometimes
routing. called by other names such as:
3. Multicast Routing: In multicasting a 1. Distributed Bellman-Ford routing
message from a sender is to be sent to a group algorithm. 2. Ford-Fulkerson algorithm
of destinations but not all the destinations in In distance vector routing, each router
a network. A process has to send a message maintains a routing table. It contains one
to all other processes in the group.For a small entry for each router in the subnet.
group it is possible to send a point-to-point Distance vector: In distance vector routing,
message. But this is expensive if the group is we assume that each router knows the identity
large. So we have to send messages to a well of every other router in the network, but the
defined groups which are small compared to shortest part to each router is not known. A
the network size. Sending message to such a distance vector is defined as the list of
group is called multicasting and the routing <destination, cost> tuples, one tuple per
algorithm used for multicasting is multicast destination. Each router maintains a distance
routing. Multicast routing is a special class of vector.
broadcast routing. Updation of router tables: A router
Broadcast Routing: In certain applications, periodically sends a copy of its distance
the host has to send packets to many or all vector to all its neighbours. When a router
other hosts. If the sender sends a packet to all receives a distance vector from its neighbour,
destinations simultaneously then it is called it tries to find out whether its cost to reach any
as broadcasting. Various methods of destination would decrease if it routed
broadcasting are as shown in through that particular neighbouring router.
Q. Link state routing: Now we will discuss Q. Path Vector Routing: It is different from
the development of routing table in link state both distance vector routing and link state
routing. Here the term link state is used for routing.Each entry in the routing table will
defining the characteristic of a link or edge, have the information about the destination
which represents a network in the Internet. network, the next router and the path to reach
The cost associated with each link is the destination.
important. The links having lower costs are Path Vector Messages: The autonomous
preferred to the links having higher costs. For boundary routers participate in path vector
the entire Internet, there is only one LSDB routing. Their job is to advertise the
and its copy is available with each node. reachability of networks present in their A.S.
Each node uses it to create the least cost tree. to the neighbour autonomous boundary
The example of LSDB. A nonexisting or router. Each router that receives a path vector
broken link is indicated by ano cost. In this message verifies whether or not the
method, each node must have a complete advertised path is according to its policy.
map of the network. This LSDB is same for Loop Prevention : When a message is
each node which shows the whole map of the received, a router checks it to see if its
internet. That means a node can use the autonomous system is in the path list to the
LSDB to make the whole map of the Internet. destination.
Path Attributes: The path is specified in
Q. Dijkstra's Algorithm: is used for terms of attributes. Each attribute gives some
computing the shortest path from the root information about the path. Hence the list of
node to every other node in the network. The attributes helps the receiving router to make a
root node is defined as the node better decision about when to apply its policy.
corresponding to the router where the
algorithm is being run. The total number of Q. Routing Protocols: Routing protocols are
nodes are divided into two groups namely the designed on the basis of the demand for
P group and T group. In T group the dynamic routing tables. The router in an
remaining nodes are placed. The path to Internet are supposed to inform each other
every node in the T group should be about changes. Routing protocols combine
computed from a node which is already the rules and procedures which allow the
present in group P. We should find out every routers to exchange information about these
possible node by a one hop path from a node changes between themselves. We can divide
which is already present in P and choose the the routing protocols into two categories,
shortest of these paths as the path to the interior protocols and exterior protocols. We
desired node.The algorithm then repeats the can define an interior protocol as the one that
following steps: 1. Start from the desired handles the intradomain routing. Similarly an
node say p. Write p in the P set. 2. For this exterior protocol is defined as the one which
node p, add each of its neighbours n to T set. handles the interdomain routing.
The addition of these nodes in T will have to
satisfy the following conditions:
1. If the neighbouring node (say n) is not
there in it with the cost to reach it through p
and p's ID.2. Stop when T is empty.
Q. Routing Information Protocol(RIP): Q. IGMP (Internet Group Management
RIP is used for updating the routing tables. Protocol): IGMP is a necessary but not
The routing updates are exchanged between sufficient protocols used in multicasting
the neighbouring routers after every 30 environment. It is always used along with IP.
seconds with the help of the RIP response 1. Group management: In the multicasting
message. These messages are also known as environment we need to use the multicast
the RIP advertisements. These messages are packets. So in an Internet we have to use the
sent by the routers or hosts. They contain a routers which can route multicast packets. A
list of multiple destinations within an. multicast routing protocol should be used to
Autonomous System (AS). RIP is an interior update the routing tables of these routers. But
routing protocol used inside an Autonomous note that IGMP is not a multicasting routing
System (AS). Its operation is based on protocol. Instead its job is to manage the
distance vector routing. In the distance vector group membership.
routing each router periodically shares its 2. Maximum response time: This is the next
knowledge about the whole Internet with its 8-bit field which defines the amount of time
neighbours. As stated earlier, RIP is a very allowed to answer a query. The value in this
simple intradomain or interior routing field shows the maximum response time in
protocol which works inside an Autonomous tenths of seconds.
System (AS). RIP implements the distance 3. IGMP checksum : The checksum is the
vector routing with the following 16-bit one's complement the one's
considerations: 1. in an A.S. it has to deal complement sum of the 8-byte IGMP
with routers and networks (links) and not the message. When the checksum is computed,
nodes. 2. Now the destination in a routing the checksum field should first be cleared to
table is a network. That is why, the network 0.
address in defined in the first column. 3. The 4. Group address: This is a 32-bit field and
metric used in RIP is called as the hop count its value depends on the type of message. the
and it is very simple. It is defined as the value of this field is zero for a general query
number of links a packet has to travel to message. The value in this field defines the
reach its destinati multicast address of the group called groupid,
in the other three types of messages.
Q. Border Gateway Protocol (BGP): BGP Q. OSPF: 1. Type of service routing: It is
is an exterior routing protocol. It is a unicast possible to configure different routers to
routing protocol. It is used for the support different types of service
interautonomous system routing i.e. routing requirements. For while the other one is
among different ASs. It was introduced in configured to support 2. Load balancing:
1989 and has four versions. BGP operation When multiple routes are available, traffic
takes place on the basis of the routing method can be evenly distributed over the routes.
called path vector routing. This principle is 3. Subdivision of autonomous systems: It is
used because the distance vector. routing and possible to further divide the system into
link state routing do not prove to be much. logical areas. 4. Security: The data
suitable for intracutaneous system routing. exchanges in OSPF are authenticated.
Inadvertent or malicious.
5. Host: OSPF supports specific routing.
 Unit 5 Q. Transmission Control Protocol (TCP):
Q. Transport Layer Process delivery The TCP provides reliable transmission of
service parameters: Transport layer is data in an IP environment. TCP corresponds
meant for the process to process delivery and to the transport layer (Layer 4) of the OSI
it is achieved by performing a number of reference model.TCP has been revised
functions. multiple times in last few decades. With
1. Packetizing: The transport layer creates stream data transfer, TCP delivers an
packets with the help of encapsulation on the unstructured stream of bytes identified by
messages received from the application sequence numbers.Instead, TCP groups bytes
layer. Packetizing is a process of dividing a into segments and passes them to IP for
long message into smaller ones. These delivery.Bytes not acknowledged within a
packets are then encapsulated into the data specified time period are retransmitted. The
field of the transport layer packet. reliability mechanism of TCP allows devices
2. Connection control: Transport layer to deal with lost, delayed, duplicate, or
protocols are divided into two categories: misread packets.TCP supports a full-duplex
1. Connection oriented, 2. Connectionless. operation means that TCP processes can both
Connection oriented delivery: A send and receive at the same time.
connection oriented transport layer protocol Q. TCP Services: Following are some of the
establishes a connection i.e. virtual path services offered by TCP to the processes at
between sender and receiver. This is a virtual the application layer :
connection. The packet may travel out of 1. Stream delivery service.
order. 2. Sending and receiving buffers.
2.Connectionless: A connectionless 3. Bytes and segments.
transport protocol will treat each packet 4. Full duplex service.
independently. 5. Connection oriented service.
3. Addressing: The client needs the address 6. Reliable service.
of the remote computer it wants to 7. Process to process communication.
communicate with.
4. Flow and error control: For high Duties of transport layer
reliability the flow control and error control
should be incorporated. Packetizing
Flow control: We know that data link layer
can provide the flow control. Similarly Connection control
transport layer also can provide flow control.
Congestion control and QoS: The Addressing
congestion can take place in the data link,
Flow Control
network or transport layer. But the effect of
congestion is generally evident in the Error Control
transport layer. Quality of Service (QoS) can
be implemented in other layers but its actual Fig, Transport Layer
effect is felt in the transport layer. The
transport layer enhances the QoS provided
by the network layer.
Q. Quality of Service (QoS) : As mentioned Q. User Datagram Protocol (UDP) : The
earlier, the QoS parameters are as follows: User Datagram Protocol is a very simple
1. Connection establishment delay: The protocol. It adds little to the basic
time difference between the instant at which functionality of IP. Like IP, it is an unreliable,
a request for transport connection is made connectionless protocol. You do not need to
and the instant at which it is confirmed is establish a connection with a host before
called as connection establishment delay. exchanging data with it using UDP, and there
2. Connection establishment failure is no mechanism for ensuring that data sent is
probability: Sometimes the connection may received. The concept of port numbers is
not get established even after the maximum common to both UDP and TCP. The port
connection establishment delay. numbers identify which protocol module sent
3. Throughput: It is defined as the number (or is to receive) the data. Most protocols
of bytes of user data transferred per second, have standard ports that are generally used for
measured over some time interval. this. For example, the Telnet protocol
Throughput is measured separately for each generally uses port 23. The Simple Mail
direction. Transfer. Protocol (SMTP) uses port 25. The
4. Transit delay: It is the time duration use of standard port numbers makes it
between a message being sent by the possible for clients to communicate with a
transport user from the source machine and server without first having to establish which
its being received by the transport user at the port to use.It is also used by protocols like the
destination machine. Simple Network Management Protocol
5. Residual error ratio : It measures the (SNMP).
number of lost or garbled messages as a Q. TCP Connection Establishment:
percentage of the total messages sent. Ideally Connection establishment is performed by
the value of this ratio should be zero and using a three-way handshake mechanism.To
practically it should be as small as possible. make the transport services reliable, one
6. Protection: This parameter provides a another.This is necessary so that packets are
way to protect the transmitted data against not transmitted or re-transmitted during
reading or modifying it by some session establishment or after session
unauthorised parties. termination. Each host randomly chooses a
7. Priority: Using this parameter the user sequence number used to track bytes within
can show that some of its connections are the stream it is sending and receiving. and the
more important (have higher priority)than client's initial sequence number (x). The
the other ones. server (HOST B) responds with its own SYN
Q. TCP Congestion Control: We have segment containing the server's initial
already discussed the reasons of congestion sequence number (y). The server also
in networks and the Internet is no exception. acknowledges the client's SYN by
So there are congestions occurring on acknowledging the client's SYN plus one (x
Internet too. The network layers detects the + 1). A SYN consumes one sequence number.
congestion by looking at the growing queues The client must acknowledge this SYN from
at the routers and tries to manage it by the server by acknowledging the server's
dropping packets.In the Internet, TCP plays a SYN plus one. (SEQ. = x + 1, ACK = y + 1).
major role in controlling congestion. This is how a TCP connection is established.
Q. Leaky Bucket Algorithm: Leaky bucket queue length. Packets which are put in the
algorithm is used to control congestion in buffer when buffer is full are thrown away.
network traffic. As the name suggests it's
Q. Token Bucket Algorithm: This algorithm
working is similar to a leaky bucket in real is similar to the leaky bucket but it is possible
life. The principle of leaky bucket algorithm to vary output rates. This is useful when
is as follows: Leaky bucket is a bucket with larger burst of traffic is received. It enforces
a hole at bottom. If bucket is full, any a long-term average transmission rate while
additional water entering in the bucket is permitting bounded bursts. In this approach,
thrown out (Packets are discarded). Same a token bucket is used to which manages the
technique is applied to control congestion in queue regulator that ultimately controls the
network traffic. Every host in the network is rate of packet flow into the network.If the
having a buffer (equivalent to a bucket) with token bucket gets full then the extra tokens
finite queue are discarded.

Unit 6 Q. HTML: The web pages are created by

Q. Domain Name System (DNS) Work : To using a language called HTML. It uses
map a name onto an IP address, an certain marks to format the text. For example
application program calls a library procedure if a part of text is required to be "boldface"
called the resolver. The name is passed on to then we can use the beginning and ending
the resolver as a parameter. The resolver bold face tags (marks) in the text as shown
sends a UDP packet to a local DNS server below:
which looks up the name and returns the <B> -Beginning of boldface </B>-End of
corresponding IP address to the resolver. The boldface.
resolver then sends this address to the caller. Here <B> and </B> are the instructions for
Then the program can establish a TCP the browser. The browser will make the part
connection with the destination or sends in of the text between these tags bold. HTML
the UDP packets. lets the user to use only ASCII characters for
Q. DHCP & Short Note: DHCP (Dynamic the main text as well as for formatting
host configuration protocol) is the first client instructions. So every computer can receive
server application program that is used after the whole document as an ASCII document.
a host is booted. Thus it works as a bootstrap The formatting instructions are used by the
when the host is booted and is to be browser to format the data
connected to the Internet, but does not know Q. BOOTP (Bootstrap Protocol) & Short
its IP address. A computer that makes use of Note: The RARP is reverse address
the TCP/IP suite must know its IP address. resolution protocol. It uses a computer's
Alongwith its IP address it must also know hardware address to identify machine, it
the following information: 1. Subnet mask of cannot be used on networks that addresses.
the computer 2. IP address of the router, so The RARP protocol has three drawbacks,
that it can communicate with other networks. first, RARP operates at low level. Second,
3. IP address of the name server so that it can client machine and a computer that answers
use the names instead of addresses. This is its request and third, it uses a computer
known as host configuration process. hardware address to identify machine.
Q. Electronic Mail: One of the most popular Q. FTP protocol & Short Note: A standard
network services is electronic mail (e-mail). mechanism provided by the Internet which
Simple Mail Transfer Protocol (SMTP) is the helps in copying a file from one host to the
standard mechanism for electronic mail in other is known as the File Transfer Program
the internet. The first e-mail systems simply (FTP). Some of the problems in transferring
consisted by file transfer protocols. But some files from one system to the other are as
of the limitations of this system were as follows:
follows: 1. Two systems may use different file name
1. It is difficult to send a message to a group conventions.
of people. 2. Two systems may represent text and data
2. Message did not have any internal in different ways.
structure. So its computer processing was 3. The directory structures of the two systems
difficult. may be different. FTP provides a simple
3. The sender never used to know if a solution to all these problems. FTP
message arrived or not. establishes two types of connections between
4. It was not easy to handover one's e-mails the client and server. One of them is used for
to someone else for the purpose of managing data transfer and the other is for the control
them when one is out of town or country for information. The fact that FTP separates
sometime. control and data makes it very efficient. The
5. The user interface with the transmission control connection uses simple rules of
system is poorly integrated. So more communication.
elaborate e-mail systems were proposed.
ARPANET e-mail proposals were published Q. Internet Protocol: Internet Message
as RFC 821. These are used in Internet. Access Protocol (IMAP) is a protocol for
Q. Telnet Protocol & Short Note: Telnet is accessing email or bulletin board messages
a protocol that allows you to connect to from a (possibly shared) mail server or
remote computers (called hosts) over a service. IMAP allows a client e-mail program
TCP/IP network (such as the internet). Using to access remote message stores as if they
telnet client software on your computer, you were local. Email stored on an IMAP server
can make a connection to a telnet server (that can be manipulated from a workstation at the
is, the remote host). Once your telnet client office, a desktop computer at home or a
establishes a connection to the remote host, notebook computer while traveling, without
your client becomes a virtual terminal, requiring the transfer of messages or files
allowing you to communicate with the back and forth between these computers.
remote host from your computer. In most Details of the IMAP specification can be
cases, you'll need to log into the remote host, found at www.imap.org.
which requires account on that system. you
can log in as guest or an account. Telnet
clients are available for all major operating
systems. Command-line telnet clients are
built into most versions of macOS, Windows,
Unix, and Linux. To use these clients, go to
their respective command lines
Q. WWW short Note:- Q. Basic Function E-Mail:
-People have become aware of the power of An e-mail system consists of two subsystems
Internet through WWW. 1. User agents.
-HTTP is a file transfer protocol which is 2. Message transfer agents.
specifically designed to facilitate access to User agents: They enable users to read and
the WWW. send e-mail.
-The World Wide Web is an architectural Message transfer agents: They move the
framework for accessing documents which messages from the sender to the receiver.
are spread out over a number of machines Basic functions: E-mail systems support five
over Internet. basic systems which are as follows:
-It has a colourful graphical interface which 1. Composition: The system can also provide
is easy for the beginners to use. assistance with addressing and a number of
-It provides information on almost every header fields attached to each message.
subject. The web (also known as WWW) 2. Transfer: This includes establishment of a
began in 1989 at CERN the European center connection from sender to destination or
for nuclear research. some intermediate machine, transferring the
-The web was designed basically to connect message, and breaking the connection.
scientists stationed all over the world. The 3. Reporting: The reporting system is
web is basically a client-server system. designed to tell the sender about whether the
-The web pages are written in the languages message was delivered or rejected or lost.
HTML and Java. 4. Displaying: It is the process of displaying
-The growth of the World-Wide Web (WWW the incoming messages so that it can be read
or simply Web) today is simply phenomenal. by the user.
-Each day, thousands of more people join the 5. Disposition: This is concerned with what
Internet (above 100 million users at recent the recipient does with the received message.
estimates). Disposition is the final step in e-mail system.
Some of the possibilities are as follows:

Advanced features of E-mail systems:

Some of the advanced features included in
4. Sending a message to a large group of addition to the basic functions are as follows:
people using the idea of mail list. 1. Forwarding an e-mail to a person away
5. To provide the facility of registered e-mail. from his computer.
6. Automatic notification of undelivered e- 2. Creating and destroying mailboxes to store
mails. incoming e-mail.
7. Carbon copies. 8. High priority E-mail 3. Inspecting contents of mailbox, insert and
(setting the priority of E-mail). delete messages from the mailboxes.
 Circuit Switching Datagram Packet Virtual-Circuit
Switching Packet Switching
Dedicated Transmission path No dedicated path No dedicated path
Continuous transmission of data Transmission of packet Transmission of packet

Fast enough for interactive Fast enough for interactive Fast enough for interactive
Messages are not stored Packet may be stored until Packet stored until delivered
delivered
The path is established for entire Rout established for each Rout established for entire
conversation packet conversation
Call setup delay; negligible Packet transmission delay Call setup delay; packet
transmission delay transmission delay

IPv4 IPv6
1. In IPv4 there are only 232 possible ways to 1. IPv6 there are 2128 possible way (about 3.4 x
represent the address (about 4 billion possible 1038 possible addresses).
addresses).
2. The IPv4 address is written by dotted-decimal 2. IPv6 is written in hexadecimal and consists of
notation. e.g. 121.2.8.12 8 groups, containing 4 hexadecimal digits or 8
groups of 16 bits each. e.g. FABC: AC77:
7834:2222:FACB: AB98: 5432:4567.
3. The basic length of the IPv4 header comprises 3. The IPv6 header is a fixed header of 40 bytes
a minimum of 20 bytes (without option fields). in length, and has only 8 fields
4. IPv4 header has a checksum, which must be 4. IPv6 has no header checksum because
computed by each router. checksums are, for example, above the TCP/IP
protocol suite, and above the Token Ring,
Ethernet
5. IPv4 contains an 8-bit field called Service 5. The IPv6 header contains an 8-bit field called
Type. The Service Type field is composed of a the Traffic Class Field.
TOS
6. The IPv4 node has only stateful auto- 6. The IPv6 node has both a stateful and a
configuration. stateless address autoconfiguration mechanism.
7. Security in IPv4 networks is limited to 7. IPv6 has been designed to satisfy the growing
tunneling between two networks. and expanded need for network security.
8. Source and destination addresses are 32 bits 8. Source and destination addresses are 128 bits
(4 bytes) in length. (16 bytes) in length.
 Intra-Domain Inter-Domain
Routing within an Autonomous System (AS routing between AS's
Ignores the Internet outside the AS assumes that the Internet consists of a collection
of interconnected AS's
protocols for Intradomain routing are normally, there is one dedicated router in each
collectively called Interior Gateway Protocols AS that handles interdomain traffic.
or IGP s.
popular protocols are: protocols are collectively called Exterior
1. RIP (simple, old) Gateway, Protocols or EGP's.
2. OSPF (better) -popular protocols are: Border Gateway
Protocol (BGP) v4 current

Q. Direct Vs Indirect