TCP/IP
The interconnection of networks is known as internetworking (or an internet). Each part of an internet is a subnetwork (or stubnet), and Transmission Control Protocol (TCP) and Internet Protocol (IP) are a pair of protocols that allow one stubnet to communicate with another.
A protocol is a set of rules that allows the orderly exchange of information. The IP part corresponds to the network layer of the OSI model and the TCP part to the transport layer. Their operation is transparent to the physical and data link layers and can thus be used on Ethernet, FDDI or Token Ring networks.
The address of the Data Link layer corresponds to the physical address of the node, such as the MAC address (in Ethernet and Token Ring) or the telephone number (for a modem connection). The IP address is assigned to each node on the internet, and is used to identify the location of the network and any subnets.
TCP/IP was originally developed by the US Defense Advanced Research Projects Agency (DARPA), and its objective was to connect a number of universities and other research establishments to DARPA. The resultant internet is now known as the
Internet and uses TCP/IP to transfer data, (each node on the Internet is assigned a unique network
address, called an IP address). Any organization can have its own internets to connect to the Internet the addresses must conform to the Internet addressing format. Common applications that use TCP/IP communications are remote login and file transfer.
File transfer protocol (ftp) for file transfer and telnet allows remote login into another
computer using TCP/IP.
ISO - IP is the TCP/IP standard relating to the network and transport layer of
the OSI adopted by ISO relating to the network and transport layers of the OSI
model. Most currently available systems conform to the IP addressing standard.
TCP/IP and OSI Model
top
TCP/IP Layers
As with all other communications protocol, TCP/IP is composed of layers:
- IP - This layer is responsible for moving packet of data from node to node.
Each packet are based on a four byte destination address (the IP
number).
- TCP -This layer is responsible for verifying the correct delivery of data
from client to server.
- Sockets -This layer is a name given to the package of subroutines that
provide access to TCP/IP on most systems.
TCP/IP hosts (nodes) communicate over interconnected networks using TCP/IP communications.
A node connects one type of network to another. A hardware provides the physical link between the different networks and the hardware and software to convert frames from one network to the other.
Point-to-point link is when a router connects a network of a similar type to another of the same kind through. The difference between a gateway, a router, and a bridge is that
the bridge uses the 48-bit MAC address to route frames, whereas the gateway and router use the IP network address.
comparing the routing with random assigned telephone numbers e.g. the MAC address would be equivalent to a randomly assigned telephone number, whereas the IP address would contain the information on where the telephone is logically located, such as which country, area
code etc.
Internet Gateway layers
The figure above shows how a gateway (or router) routes information. It reads the data frame from the computer on network A, and reads the IP address contained in the frame and makes a decision whether it is routed out of network A to network B. If it does then it relays the frame to network B.
top
The main functions of the IP protocol are to:
- Route IP data packets - internet datagrams. Data passes from node to gateway through the internet.
As the IP protocol program running on each node knows the location of the
gateway on the network which must be able to locate the interconnected
network.
- Fragment the data into smaller units, if it is greater than a given amount
e.g. 64 kB.
- Report errors. If an error occurs when a datagram is being routed or is being reassembled
the node that detects the error reports back to the sourcenode. The
datagrams travel through the network for more than a set time it is deleted,
for each datagram an error message is sent to the source node to inform it that the Internet routing could not find a route for the datagram or that the destination node, or network, does not exist.
top
Many of the existing networking standards were developed before the OSI model was
developed, so no networking technology fits into the model. Also the OSI model was meant to be an abstract method of viewing a network from its physical connection,
through its hardware/software interface, right up to the application program. The
network protocol is the key element that allows computers
over the world to intercommunicate, no matter their operating system, their hardware, their network connection, or their application program.
TCP/IP is the most common protocol for worldwide communications.
TCP/IP does not quite fit into the OSI model, as illustrated in the figure below. The OSI
model uses seven layers where the TCP/IP model uses four layers, which are:
- Network access layer. Specifies the procedures for transmitting data across the network, including how to access the physical medium, such as Ethernet and
FDDI.
-
Internet layer. Responsible for data addressing, transmission, and packet fragmentation and reassembly (IP protocol).
-
Transport layer. Manages all aspects of data routing and delivery including session initiation, error control and sequence checking (TCP/UDP protocols). This includes part of the session layer of the OSI model.
- Application layer. Responsible for everything else. Applications must be responsible for all the presentation and part of the session layer.
OSI and TCP/IP model
top
|
Related Topics
Functions of the transport layer
TCP/IP model
Transmission
control protocol
UDP
TCP specification more
Topics
|
