|< Day Day Up >|| |
Internet Protocol breaks data transmissions into smaller chunks called packets. Communication takes place by forwarding these packets from one network to another. Every time it creates a data packet, Internet Protocol attaches a header with the IP address of the source and the destination. Thus, IP addresses are completely virtual, software-based identifiers. To the network, they are a string of thirty-two 1s and 0s, but people commonly represent them as four numbers from 0 to 255 separated by dots; for example, 22.214.171.124. With 32 bits available, the IP address space yields a theoretical maximum of about 4.3 billion unique numbers.
The classical IP address has two basic parts. The first encodes a network to which a computer is attached; the second identifies a specific device attached to that network. The first part is often referred to as the network prefix, the second part as the host ID, where host refers to a computer or some other connected device. As in Ethernet addressing, the two-part structure helps to distribute the responsibility for address assignment. The central authority for IP addresses only needs to hand out network prefixes; the recipients then perform the task of assigning host IDs on their own networks. The original IP addressing structure defined three primary classes of address assignments, with the classes based on the number of bits used by the network prefix. [1 ]
[1 ]Class A assignments used only the first 8 bits to identify the class and the network, providing for a relatively small number (126) of very large networks accommodating up to 16.7 million hosts. Class B assignments used the first 16 bits to identify the class and the network prefix, providing for 16,384 intermediatesized networks accommodating up to 65,536 hosts. Class C assignments provided for a very large number of small networks with only 256 unique identifiers for hosts. There were also two additional classes, D and E, one for multicasting and the other for experimental uses.
|< Day Day Up >|| |