Voice over IP

Lecture# 5, 6
NETWORK EDGE
Asst. Prof. Muhammad Adil

END SYSTEM
The computers connected to the internet are referred to as
End Systems.
The internets End Systems include desktop computers
(PC), Macs , portable computers, PDA , Phones with
wireless internet , web TV , digital cameras , home
application etc.
They are also referred to as hosts because they run
(host) application programs like Web server programs, an
e-mail reader program or an e-mail sever program.
Hosts are also categorized in to two types. i.e. clients and
server.

END SYSTEM
Clients:
A client tends to be desktop and mobile PCs, PDAs
etc.

Servers:
A server is a more powerful machine that stores and
distributes web pages, streams video and audio, relays
e-mail etc.

END SYSTEM
Clients & Servers Applications:
In context of networking software, A Client program is a
program running on one end System that requests and receives
a service from a Server program running on another system.
Undoubtedly, the client-server model is the most powerful
structure for internet applications.
File transfer, e-mail, remote login, web, news groups and many
more applications have adopted this client-server model.
Client-server internet applications are called Distributed
Applications because the client program runs on one
computer and the server program runs on another computer.

CONNECTIONLESS AND CONNECTED SERVICES

TCP/IP Networks and in particular the internet

Developers creating an internet applications e.g.

Connection-Oriented Services
These services are based on connection-oriented transport i.e.
TCP.
Here, the client and server programs send control packets to
each other before sending packets with actual data to be
transferred, which is known as handshaking.
Handshaking procedure alerts the clients and server to be
prepared for transmission of data packets.
As soon as the handshaking procedure is finished, a connection
is said to be established between the two systems.
TCP is said to be connection-oriented because of handshake
process.
TCP is defined in RFC 793, RFC 1122, RFC 1323, RFC 2018 and
RFC 2581.

Connection-Oriented Services

Connection-Oriented Services
Connection-Oriented transfer is Reliable Data Transfer
because an application can rely on the connection to deliver
all of its data without errors and in proper order.
Reliability in internet is achieved by acknowledgements
and retransmissions.
For example, let an application has developed a
connection between end system X and Y. When Y receives
a packet from X, it sends an acknowledgement; this
makes X know that Y has received the sent packets,
definitely when X does not receive an Ack from Y, it
assumes the packet loss and thus retransmits the packet.

Connectionless Services
Here, there is no handshaking, so when one side
wants to transmit packets to other side it simply
sends the packets.
Due to non-handshaking prior to packet
transmission, data can be deliver sooner.
But there is no reliable data transfer, so a source
never knows which packets are lost.
There is no flow control or congestion control.

Connectionless Services
The internet connectionless services are called user
datagram protocol (UDP).
UPD is defined in RFC 768.
TCP is mostly used by application like Telnet
(remote login), SMTP (e-mail), FTP (file transfer) and
HTTP (for the web).
UDP is used by many applications also, like many of
the emerging multimedia applications such as
internet phone and video conferencing.

CIRCUIT-SWITCHED NETWORK
It consists of a set of switches connected by physical
links.
A connection between two systems is a dedicated
path made of one or more links. A link is divided
into n channels.
Each connection uses only one dedicated channel on
a link.
Each link is divided into channels using FDM or
TDM.

CIRCUIT-SWITCHED NETWORK

CIRCUIT-SWITCHED NETWORK
The end systems like phones or computers are directly connected to
switch.
Circuit switching takes place at the physical layer.
Communication in a circuit switched network requires 3-Phases.
Setup Phase: A dedicated circuit needs to be established before the
two or multiple parties can communicate. End systems are connected
by means of dedicated lines (channels) to switches, so connection
setup means creating dedicated channels between the switches. So
when a connection is established an Ack: from Y is sent back to X.
Data Transfer Phase: After the establishment of the dedicated circuit
(channels), the two parties can transfer data.
Teardown Phase: When one of the parties needs to disconnect, a signal
is sent to each switch to release the resources.

PACKET-SWITCHING
A network in which data are transmitted in form of
independent units called packets is called aPacketSwitching Network.
Data transmission using a packet-switch network is called
Packet-Switching.
Applications exchange messages over network.
Messages may perform control tasks like Handshaking
or may contain data e.g. an e-mail message, a GIF image,
an MP4 video file etc.
The source breaks long messages into smaller chunks of
data called Packets.

PACKET-SWITCHING
Each of these packets travel through communication links
called Packet Switches.
Two important type of devices are Routers and LinkLayer switches.
The rate at which packets are transmitted over a
communication link is equal to the full-transmission rate
of the link.
Most packet switches use Store-and-Forwarding
transmission at the inputs to the link. i.e. the switch must
receive the entire packet before it can begin to transmit
the first bit of the packet onto the outgoing link.

PACKET-SWITCHING
So a store-and-forwarding delay at the input to each
link is introduce.
If a packet consists of L bits and the packet is to be
forwarded onto a link of R bps then the store-andforwarding delay at the switch is L/R seconds.
This delay is proportional to the packet length in bits

PACKET-SWITCHING
OUTPUT BUFFER/QUEUE:
Each packet switch has multiple links attached to it, and for each
link there is an output buffer or output Queue, that stores the
packet that the router is about to send into that link.
Output buffer is very important. If a packet is to be transmitted
over an already busy link, then the newly arrived packet must wait
in the output buffer.
This also adds the output buffer Queuing delay. This delay is
variable and depends upon the level of congestion in the network.
As amount of buffer space is finite, so there is a possibility an for
arriving packet that the buffer is completely filled with other
packets waiting for transmission, that will result in packet loss;
Either the arriving packet or one of the already queued packet will
be dropped.

PACKET-SWITCHING

PACKET-SWITCHING
In this figure packets of equal size are to be transmitted over
the network.
Hosts A and B are sending packets to Host E.
At first the packets are sent along 10 Mbps Ethernet links to the
first packet switch.
The packet switch directs these packets to the 1.5 Mbps link.
Congestion will occur if the arrival rate of packets exceeds the
rate at which the switch can forward packets across the 1.5
output link.
If there are Q links between the two hosts, each of rate R bps.
Let queuing delay and end-to-end propagation delays are
negligible, and there is no connection establishment, then the
total delay for such a case is QL/R