The IP Protocol

Recall earlier: the TCP (transport layer) protocol is implemented in the two end systems involved in communications.

TCP relies on the Internet Protocol to deliver packets of data from one end system to another. The packets which IP delivers are usually called datagrams.

Datagrams navigate zero or more (sometimes many) routers, interconnecting the individual networks which together make up the Internet. The word "Internet" is just a contraction of the phrase "Interconnected Networks".

A router is essentially a special-purpose computer whose sole function is to transfer datagrams between networks. In earlier times, they were called "IP Gateways". The general structure of the Internet can thus be envisaged:

IP Addressing

The concept of a global Internet addressing scheme is fundamental to IP.

Every Internet-connected system has a unique Internet host address. This is a 32 bit, or 4 byte, binary number.

Internet addresses are written as a dotted sequence of the form:

aa.bb.cc.dd

where aa, bb, etc, are the decimal values (ranging from 0 to 255) of the 4 bytes which make up the internet address, for example:

149.144.21.60

This is called the "Common Internet Address Notation". Internet addresses are commonly referred to as "IP addresses".

Note that 149.144.21.60 is the IP address of "ironbark", or to use its full name, ironbark.bendigo.latrobe.edu.au. We shall see later how the name of a computer (ironbark, redgum, bindi, etc) is mapped to its IP address.

Note that the traditional Internet literature refers to internet-connected computers as hosts, which reflects the historical nature of computers as multi-user timeshared systems.

"Two Part" Addresses

The 4 byte internet address consists of 2 parts, the network part and the host part.

• The network part must be the same for all hosts connected to the same network.
• The host part must be unique within each specific network. It is used to deliver the datagram once it has been routed to the appropriate network.

The structure of an IP address differs depending on the class of the local network to which the host is connected:

Class A Network

the first (or leftmost) field, or byte, of the address specifies the network number and class, and can take values from 1 to 126.

Class B network

the leftmost two fields (bytes) specify the network number and class. The first field is in the range from 128 to 191, and the second from 1 to 254.

Class C network

the leftmost three fields specify network number and class. The first field can be from 192 to 223, the second from 0 to 255, and the third from 1 to 254.

IP Address Classes

The Internet address format can be described graphically:

 

Example 1:

in a class A network, only values from 00000000 to 01111111 are possible. Of these 128 combinations, 126 are permitted (networks 0 and 127 are reserved for other purposes). Hence there can only ever be 126 class A networks, each of which can have a huge number of hosts.

Example 2:

ironbark's IP address is 149.144.21.60 This is a class B network (149.144), and ironbark's host ID is 21.60

Example 3:

203.17.3.5 is a class C address. The network number is 203.17.3, and the host number is 5.

Network numbers above 223 are also reserved for special purposes, outside the scope of this subject.

IP Address Miscellany

The broadcast address is the address used to send a datagram to all hosts on the local network simultaneously. This is the address given by the network number followed by all binary 1's in the host part of the address.

For example: A broadcast address for a host on an example Class-C network is:

195.96.97.255

Where Class-B networks (and, nowadays, class C also) are in use, the host part of the address is commonly split, for management purposes, to allow subnetting - typically this involves defining 254 subnets each of 254 hosts. Each of these subnets behaves, for all intents, like an independent class-C network: thus:

Exactly which bits of the subnetted address are used for the network/subnet part and the host part is defined using an address mask, or netmask thus:

255.255.255.0

IP Datagrams

IP transmits data in packets called datagrams. A datagram contains a header and some data, thus:

Notes:

• The total length of a datagram can be up to 64 KB, but is usually less than 1500 bytes
• The header contains the source and destination IP addresses as well as various other administrative data, a minimum of 20 bytes. In fact, virtually all datagrams have a 20 byte header.

IP: Connectionless Datagram Delivery

IP data transfer across an internet is based on three fundamental principles:

Unreliable delivery

delivery of data is not guaranteed. A packet of data may be lost in the network, may be duplicated (ie: delivered twice) or may be delivered out of order. The IP service will not detect such conditions, nor will it notify the sender or receiver if they occur.

Connectionless delivery

each packet is treated entirely indpendently of all others. No information is kept as to which packets have been forwarded, and packets may travel over different routes to the same destination.

Best-Effort delivery

the packet delivery mechanism is engineered to always deliver packets if possible. It will not gratuitously drop packets: unreliability should only occur when underlying resources (eg buffer space) are exhausted.

These specifications allow the IP service to concentrate on its job: delivering packets. As we have seen, higher level protocols (usually TCP) transform the IP service into a reliable, sequenced interprocess communications mechanism.

IP Routing

There are two levels of complexity involved in IP routing:

Local delivery

when the IP software is presented with a datagram for delivery, it first checks the network part of the destination address to see if matches its own network number. If it does, then the datagram can be delivered locally, and is simply handed to the physical network delivery system (typically an Ethernet device driver) for direct encapsulation and delivery.

Internet delivery

if the network numbers are different, then the datagram must be forwarded to a router (or IP gateway). In this case, the datagram is forwarded (using direct delivery as above) to the appropriate router for on-delivery

• Datagrams cross gateways (routers) from network to network until they reach a network where they can be locally delivered.
• The IP software must keep a routing table to know the IP address of an appropriate router. This is commonly done by defining a default route for each network machine.
• Note that routers must have (at least) two IP addresses!

The Structure of the Internet

The Internet consists of networks (or, more commonly nowadays - subnets) connected by routers. The traceroute utility shows the path a datagram takes - the following traceroute is from ironbark to the Victorian Regional Network (VRN).

 1  r-busbgo.bendigo.latrobe.edu.au (149.144.21.254)  1 ms  2 ms  1 ms
 2  r-bgoatm34.bendigo.latrobe.edu.au (149.144.10.250)  1 ms  1 ms  1 ms
 3  r-sctech-atm.latrobe.edu.au (131.172.239.3)  5 ms  4 ms  3 ms
 4  cisco-ltu-fddi.latrobe.edu.au (131.172.20.12)  5 ms  4 ms  4 ms
 5  vic-gw.vrn.EDU.AU (203.21.130.129)  4 ms  5 ms  4 ms
 

The Transport Layer Revisited: UDP

The User Datagram Protocol provides a connectionless alternative transport service to TCP for applications where reliable stream service is not needed. UDP datagrams can be droppped, duplicated or delivered out of order, exactly as for IP.

The UDP transport service add to IP the ability to deliver a datagram to a specified destination process using a port abstraction, in an analogous way to that used by TCP.

Examples of applications where UDP is used include:

• Any application where loss of a datagram is not critical because later datagrams will imply the missing information. Examples include:
  • A "timekeeper" host sends the current "wallclock" time to a slave host.
  • Routers broadcast copies of their routing tables every 30 seconds.
• An application process which performs its own error correction is used.
• The reduced overhead of connectionless operation suits some time-critical applications where (occasional) loss of data may be unimportant. An example is voice communications over the Internet.