The DARPA Internet Protocol Suite is more
commonly known as the "TCP/IP Protocol".
It is most widely used protocol on the internet.
Developed by ARPA (DARPA) in late 60's and 70's for
the ARPANET, for research.
In 1984, ARPANET splits off MILNET.
There are two distinct layers and several protocols in the suite.
We describe the three most important ones: IP, TCP and UDP.
The IP (= Internet Protocol)
corresponds to the network layer (layer 3).
Built on top of IP are two protocols, TCP and UDP.
These two protocols correspond to the transport layer (layer 4).
Users usually do not interact with IP but with TCP or UDP.
IP DATAGRAMS
IP provides connectionless and unreliable communication.
Each IP datagram has a source and destination Internet Address.
The IP protocol performs a checksum on the datagrams, and if
checksum fails, the datagram is simply discarded. It is up to
the higher levels to detect this and re-transmit.
However, IP does perform fragmentation service.
TCP (= Transmission Control Protocol)
TCP is a connection-oriented, reliable, full-duplex,
byte-stream service.
UDP (= User Datagram Protocol)
UDP is a connection-less, unreliable datagram service.
Some standard applications based on this suite:
FTP
file transfer protocol
TCP
TFTP
Trivial file transfer protocol
UDP
Telnet
Remote login service
TCP
SMTP
Simple Mail Transfer Protocol
TCP
Other protocol suites:
SNA (from IBM), XNS (from Xerox), OSI.
INTERNET ADDRESSES
This is assigned by a central
authority, the Network Information Center (NIC)
which is located at SRI International.
It is a unique 32 bit number representing the network ID and
host computer ID (the latter is unique only within the network).
This is usually written as 4 decimal digits, corresponding to the
four bytes in the 32 bit address. E.g., 128.3.0.5.
There are 4 classes of addresses (A, B, C, D) which can be
distinguished by the first four bits of the 32-bit address:
0, 10, 110, 1110.
E.g., 128.3.0.5 (class B), 192.43.235.6 (class C).