From Ray Litman Photo Graphics' Web Services; Members Lounge;
www.rlpg.com/webservices/members/

Background and History of the Internet

The Internet Society hosts a monograph called called "A Brief History of the Internet." (See http://www.isoc.org/internet/history/) The authors include some of the designers of the essential components of how the Internet works today: Barry M. Leiner, Vinton G. Cerf, David D. Clark, Robert E. Kahn, Leonard Kleinrock, Daniel C. Lynch, Jon Postel, Larry G. Roberts, and Stephen Wolff. The paper notes these key milestones in Internet history:

• 1961: Leonard Kleinrock writes the first paper on packet switched networks.
• 1962: J.C.R. Licklider of MIT writes a paper describing a globally connected "Galactic Network" of computers.
• 1966: Larry Roberts proposes the ARPANET to the Defense Department's Advanced Research Projects Agency (ARPA).
• 1968: ARPA issues Request for Quotations for the Interface Message Processors (IMPs), which became the first routers.
• 1969: First IMP is installed at UCLA.
• Early 1970s: Universities and defense agencies and contractors begin to connect to ARPANET.
• 1973: Bob Kahn and Vint Cerf begin research into what eventually becomes IP - the Internet Protocol and its companion, TCP - the Transmission Control Protocol.
• 1973: Bob Metcalfe develops Ethernet, which had been the subject of his PhD thesis, while working at Xerox.
• Early 1980s: The Personal Computer revolution begins.
• Mid 1980s: Local Area Networks (LANs) begin to flourish in business and university environments. Campus area networks soon follow.
• January 1, 1983: All "old-style" traffic on the ARPANET ceases, as TCP/IP becomes the only protocol used. [Arguably, this is the date of the birth of the Internet as a functioning, practical, production network.]
• 1985: Dennis Jennings chooses TCP/IP as the protocol for the planned National Science Foundation Network (NSFnet).
• 1988: NSF sponsors a series of workshops at Harvard on the commercialization and privatization of the Internet.
• 1988: Kahn et al. write a paper "Towards a National Research Network.

"The Brief History" by Cerf et al. details the key milestones in the development of the Internet infrastructure that were essential for the Internet to evolve into what we know and use today. They cite the conscious decision to transition the Internet from a primarily defense, research, and education network into a national network of networks incorporating private as well as commercial traffic.

But is this story the whole story of the "Beginnings" of the internet.

"When, why and how did the internet begin?"

Writing Contributions By: Ray Litman, Karl Berger, Brian Allen

Forty years ago, the RAND Corporation, one of first and the United States' foremost Cold War not-for-profit think-tanks, was faced with a unique strategic problem. How would the United States Government and U.S. Military be able to successfully communicate in a post-nuclear holocaust world?

Any centralized government authority, any centralized communications network center, would be a primary target for the enemy warheads. The area where communications center was located would be the very first to be hit. After the strike, command and control would be lost or diminished to a point of ineffectiveness?

No matter how thoroughly existing communication systems were hardened or protected, their central offices and telephone units would always be vulnerable to the EMF effect of the nuclear explosions. The dreaded nuclear war would bring any conceivable network to its knees. The post-nuclear United States was going to need a command-and-control network, that would provide a redundant or even primary communication link(s) from military base to base, city to state police, and local to federal government.

Paul Baran, born in Poland in 1926, a former Hughes Aircraft Minuteman Missile designer, went to RAND Corporation (stands for Research AND Development) in 1959. Working in typical Military secrecy, Baran formulated a network that would have "no central authority". Baran's network was designed to work in small isolated sections that could communicate to other small separate computers or pieces of networks. The work continued from 1960 to 1962 and in 1964, "On Distributed Communications, a survivable network of unmanned digital switches implementing a self-learning policy at each node, without need for a central and possibly vulnerable control point, so that overall traffic is effectively routed in a changing environment.", was made public.

The 39 word title described an idea that was deceptively simple. The communications network would be understood to be unreliable at all times. All the nodes (individual computers or terminals) in the network would have the same status as all other nodes. The exchange of information would be divided into packets, each individual packet was separately tagged or addressed. Each packet would begin at some specified origin point, and end at some other specified destination node. Nodes had their own authority to originate, send, and receive messages. Each individual information packet would route & reroute its way through this new concept in networks on an singular basis.

The actual route that the individual packet traveled was inconsequential. Only the fact that the packet(s) made it to their destination and we reassembled properly was important. Each packet was moved randomly from node to node to node, approximately in the direction of its target, until it arrives at the proper destination. The completeness of the network was unimportant. Each packet would stay active in the rest of the network, moving through the active portions of whatever nodes happened to be working. If placed in the context of a system operating after a nuclear first strike, the design elements become abundantly clear. While this delivery system might appear to be "inefficient" in the usual sense it was built to be remarkably rugged. This new concept was a completely different than the direct dial or hard wired connections from the phone company or computer network of the day.

Baran's concept of a redundant, decentralized, nuclear war survivable, packet-switching network was worked on at UCLA, MIT and RAND. However, the first test network based on Baran's work was by the National Physical Laboratory in Great Britain in 1968. Following closely, the Pentagon's Advanced Research Projects Agency funded a more ambitious project. This network was made up of the fastest computers of the day. There weren't more than a handful of these rare and extremely expensive computers. This was the first of the computer networking projects that we take for granted today.

UCLA saw the first of these "node" installations in fall of 1969. the first four nodes on this network were operational by December 1969 the network, ARPANET, was named, after its sponsor, the Pentagon. Baran specified all the critical functions of the Internet: packets with headers for addresses and fields for error detection and packet ordering. He described in detail the autonomous adaptive nodes found in ARPANET.

The four computers transferred data over dedicated high-speed transmission lines. For the first time the computers could even be programmed remotely from terminals located at other facilities located hundreds of miles away. Using the ARPANET, researchers and scientists could make use of one another's computer facilities. Computer-time was in short supply and very expensive in the early'70s. The ARPANET was an idea arriving at the right place and time. Fifteen nodes were operational by 1971. ARPANET had grown to thirty-seven nodes by 1972.

Continued on: Page 2

Internet Glossary

Go Back

Questions? ... Comments... Please E-Mail us!

Home  Photography  Graphics  Video  Models
Web Services  Leftovers  Links  Production Services
New  Affiliations  Privacy Policy