From Computer History Wiki
Jump to: navigation, search

The ARPANET was the first packet-based wide-area network; so named because it was initiated, and initially funded, by ARPA. It was intermediate between a virtual circuit network and a datagram network, but closer to the former. There was no 'call setup'; but it had what amounted to connections inside the network, data was delivered reliably and in order, and it had flow control in the network.

It consisted of minicomputers called Interface Message Processors ('IMPs', for short) running specialized packet switching code, joined together with synchronous serial lines; host computers were initially connected to the IMPs by special bit-serial 1822 interfaces.

(The 1822 'Local Host' and 'Distant Host' interface variants had maximum lengths of about 100 feet and 2,000 feet respectively; to allow connection of hosts at much greater distances, a wholly new interface, the 'Very Distant Host', was eventually added. It used a synchronous seral line, at the hardware level; on the IMP end, the existing serial line interfaces were used.)

Later on, customized IMP variants called Terminal Interface Processors ('TIPs', for short) were added to the ARPANET; these provided groups of asynchronous serial lines to which could be attached terminals, which allowed users at the terminals access to the hosts attached to the ARPANET.


The protocol suite used on the early ARPANET included:

These protocols formed the Network Control Protocol, which all the application protocols ran over.

Those included:

See also

Further reading

External links