Figure 2.2 shows details of TCP/IP
Layers. Each layer stacks on top of each other to form the basic TCP/IP
structure and functionality.
A)
Application
Layer
The top layer is an Application Layer. It is
used by most applications for network communication such as Telnet, Hypertext
Transfer Protocol (HTTP), File Transfer Protocol (FTP), Simple Mail Transfer
Protocol (SMTP), Domain Name System (DNS), Routing Information Protocol (RIP)
and Simple Network Management Protocol (SNMP). Figure 2.3 shows the interaction
between user and HTTP Server. HTTP is used for viewing pages on the internet.
Figure 2.4 shows the outbound mail
and inbound mail through SMTP. SMTP is used for mail transfer from a mail
server to a client computer.
DNS is a naming system for a computer
that is connected to the internet. It finds the nearest DNS server and requests
the IP address of the website before it is able to connect into that website.
As an example; facebook.com turns into 69.63.189.11. This naming system used
because facebook.com is far easier to remember than 69.63.189.11.
RIP is used by router to exchange
routing information with each other. This gives computers network information
as to where to route the packets through. This gives a simple easy path for the
data to go through, instead of the computer sending the packet to every other
computer on the internet. SNMP enables user to manage server remotely from
other devices in a network.
B)
Transport
Layer
The second layer is Host-to-Host Transport
layer. Application layer will inform this layer that it has some data to be sent.
There are two protocols which are Transmission Control Protocol (TCP) and User
Datagram Protocol (UDP). TCP is a reliable one-to-one connection service. This
connection oriented protocol ensures data transmitted accurately. This layer
also manages errors such as packet loss, sequencing and acknowledgement of how
many packets were sent. It negotiates and maintains a connection between two
hosts. It uses the connection to ensure the receiver is ready to receive a
packet. It acknowledges sender that it accepts the packet and retransmit a
packet that is not acknowledged by the receiver. UDP is a one-to-one or one-to-many
communication service. It is very unreliable and connectionless because UDP
cannot guarantee delivery of packet data. UDP is used when the data want to be
sent is very small (one packet) and do not a TCP connection or when the
applications provide reliable delivery. It does not check if the receiver is
ready when it sends a packet to a host. When it receives a packet, it does not
acknowledge the delivery because it does not send a large number of messages. UDP
required much less bandwidth than TCP. UDP is used by application layer
protocol such as DNS to transmit short messages such as name query and
response. UDP reject corrupted packet as much as TCP did. Streaming media also
use UDP because it is low overhead. The media player will buffered enough of
arriving data to be able to request a new copy of a bad packet from the sender
and receive it before it was needed.
C)
Network
layer
The third layer is Network Layer or
Internet Layer. This layer is accountable to handle addressing, packaging and
routing functions. This layer has many protocols such as Internet Protocol
(IP), Address Resolution Protocol (ARP), Internet Control Message Protocol
(ICMP) and Internet Group Management Protocol (IGMP). IP is a main protocol of the
Internet Layer. This is an unreliable, connectionless protocol for routing
between hosts. This protocol makes its best effort to get the data to receiving
host. If a piece of data does not get through, this protocol does not attempt
to recover that piece of data. It manages the movement of packet between sending
and receiving host. It was responsible for routing packet to their destination.
Dividing them into a size that would fit through the router among their route
and addressing packet. An IP address is a series of number that is assigned to the
physical hardware address of the host. ARP resolves the Media Access Control (MAC)
address which is a unique number given to every network capable device. It
helps to deliver the packets to the receiving host by giving the packets a path
to go through. It looks up physical address that link to an IP address. ICMP is
the protocol that makes sure that all of the packets go through the network. If
a packet is not delivered, or is lost during transmission, this protocol sends
another request to send the packet. It carries error messages and provides the
simple diagnostic capability. IGMP is used by router to track group membership
on the subnet and manages groups of hosts such as computers, routers etc.
D)
Data
Link Layer
The fourth
layer is the Data Link Layer or Network Access layer. This layer handles packet
transmission. This is designed to give the layers that stack on top of this to
be able to function normally across any network medium such as Ethernet or wireless
LAN. The data link layer is divided into two sub layers which is logical link
control (LLC) and media access control (MAC). LLC is a function required to
establish and control of logical links between local devices on a network to
allow different technologies to work seamlessly with the higher layers. MAC is
a procedure to control device access to the network medium. Usually networks sharing
a medium such as a single network cable or a series of cables that are
electrically connected into a single virtual medium. Rules are needed for
managing the medium to avoid conflicts. Example of a MAC is ethernet uses the
CSMA/CD method while Token Ring uses token passing. Data Link Layer also
responsible for Data Framing, Addressing and Error Detection and Handling.
E)
Physical
Layer
The bottom
layer is Physical Layer. The physical layer of the network involves of hardware
elements such as network interface cards, cables and repeaters. Ethernet is the
most common protocol used at the physical layer. Ethernet network such as
10BaseT or 100BaseTX specifies the configuration of the hardware elements such
as type of cables that can be used, the optimal topology such as star topology,
bus topology etc. and the maximum length of cables.
