The Soviet Union's launch of
Sputnik spurred the United States to create the
Advanced Research Projects Agency (ARPA, later DARPA) in February
1958 to regain a technological lead.
[4][5] ARPA created the
Information Processing Technology Office (IPTO) to further the research of the
Semi Automatic Ground Environment (SAGE) program, which had networked country-wide
radar systems together for the first time. The IPTO's purpose was to find ways to address the US military's concern about survivability of their communications networks, and as a first step interconnect their computers at the
Pentagon,
Cheyenne Mountain, and
Strategic Air Command headquarters (SAC).
J. C. R. Licklider, a promoter of universal networking, was selected to head the IPTO. Licklider moved from the Psycho-Acoustic Laboratory at
Harvard University to
MIT in
1950, after becoming interested in
information technology. At MIT, he served on a committee that established
Lincoln Laboratory and worked on the SAGE project. In
1957 he became a Vice President at
BBN, where he bought the first production
PDP-1 computer and conducted the first public demonstration of
time-sharing.
At the IPTO, Licklider's successor
Ivan Sutherland in 1965 got
Lawrence Roberts to start a project to make a network, and Roberts based the technology on the work of
Paul Baran,
[6] who had written an exhaustive study for the
United States Air Force that recommended
packet switching (opposed to
circuit switching) to achieve better network robustness and disaster survivability. Roberts had worked at the
MIT Lincoln Laboratory originally established to work on the design of the SAGE system. UCLA professor
Leonard Kleinrock had provided the theoretical foundations for packet networks in 1962, and later, in the 1970s, for
hierarchical routing, concepts which have been the underpinning of the development towards today's Internet.
Sutherland's successor
Robert Taylor convinced Roberts to build on his early packet switching successes and come and be the IPTO Chief Scientist. Once there, Roberts prepared a report called
Resource Sharing Computer Networks which was approved by Taylor in June 1968 and laid the foundation for the launch of the working ARPANET the following year.
After much work, the first two nodes of what would become the
ARPANET were interconnected between Kleinrock's Network Measurement Center at the
UCLA's School of Engineering and Applied Science and
Douglas Engelbart's NLS system at
SRI International (SRI) in
Menlo Park, California, on 29 October 1969. The third site on the ARPANET was the Culler-Fried Interactive Mathematics center at the
University of California at Santa Barbara, and the fourth was the
University of Utah Graphics Department. In an early sign of future growth, there were already fifteen sites connected to the young ARPANET by the end of 1971.
In an independent development,
Donald Davies at the
UK National Physical Laboratory developed the concept of packet switching in the early 1960s, first giving a talk on the subject in 1965, after which the teams in the new field from two sides of the Atlantic ocean first became acquainted. It was actually Davies' coinage of the wording
packet and
packet switching that was adopted as the standard terminology. Davies also built a packet-switched network in the UK, called the Mark I in 1970.
[7] Bolt, Beranek & Newman (BBN), the private contractors for ARPANET, set out to create a separate commercial version after establishing "value added carriers" was legalized in the U.S.
[8] The network they established was called
Telenet and began operation in 1975, installing free public
dial-up access in cities throughout the U.S. Telenet was the first packet-switching network open to the general public.
[9]
Following the demonstration that packet switching worked on the ARPANET, the
British Post Office,
Telenet,
DATAPAC and TRANSPAC collaborated to create the first international packet-switched network service. In the UK, this was referred to as the
International Packet Switched Service (IPSS), in
1978. The collection of
X.25-based networks grew from Europe and the US to cover Canada, Hong Kong and Australia by 1981. The X.25 packet switching standard was developed in the CCITT (now called
ITU-T) around 1976. X.25 was independent of the TCP/IP protocols that arose from the experimental work of DARPA on the ARPANET, Packet Radio Net, and Packet Satellite Net during the same time period.
The early ARPANET ran on the
Network Control Program (NCP), implementing the host-to-host connectivity and switching layers of the protocol stack, designed and first implemented in December 1970 by a team called the Network Working Group (NWG) led by
Steve Crocker. To respond to the network's rapid growth as more and more locations connected,
Vinton Cerf and
Robert Kahn developed the first description of the now widely used TCP protocols during 1973 and published a paper on the subject in May 1974. Use of the term "Internet" to describe a single global
TCP/IP network originated in December 1974 with the publication of
RFC 675, the first full specification of TCP that was written by Vinton Cerf, Yogen Dalal and Carl Sunshine, then at
Stanford University. During the next nine years, work proceeded to refine the protocols and to implement them on a wide range of operating systems. The first
TCP/IP-based wide-area network was operational by 1 January 1983 when all hosts on the ARPANET were switched over from the older NCP protocols.
T3 NSFNET Backbone, c. 1992
In 1985, the United States'
National Science Foundation (NSF) commissioned the construction of the
NSFNET, a university 56
kilobit/second network backbone using computers called "
fuzzballs" by their inventor,
David L. Mills. The following year, NSF sponsored the conversion to a higher-speed 1.5
megabit/second network that became operational in 1988. A key decision to use the DARPA
TCP/IP protocols was made by Dennis Jennings, then in charge of the Supercomputer program at NSF. The NSFNET backbone was upgraded to 45 Mbit/s in 1991 and decommissioned in 1995 when it was replaced by new backbone networks operated by commercial
Internet Service Providers.
The opening of the NSFNET to other networks began in 1988.
[10] The US Federal Networking Council approved the interconnection of the NSFNET to the commercial
MCI Mail system in that year and the link was made in the summer of 1989. Other commercial electronic mail services were soon connected, including OnTyme, Telemail and Compuserve. In that same year, three commercial Internet service providers (ISPs) began operations:
UUNET,
PSINet, and CERFNET. Important, separate networks that offered gateways into, then later merged with, the Internet include
Usenet and
BITNET. Various other commercial and educational networks, such as
Telenet (by that time renamed to Sprintnet),
Tymnet,
Compuserve and
JANET were interconnected with the growing Internet in the 1980s as the TCP/IP protocol became increasingly popular. The adaptability of TCP/IP to existing communication networks allowed for rapid growth. The open availability of the specifications and reference code permitted commercial vendors to build interoperable network components, such as routers, making standardized network gear available from many companies. This aided in the rapid growth of the Internet and the proliferation of local-area networking. It seeded the widespread implementation and rigorous standardization of TCP/IP on
UNIX and virtually every other common operating system.
Although the basic applications and guidelines that make the Internet possible had existed for almost two decades, the network did not gain a public face until the 1990s. On 6 August 1991,
CERN, a pan-European organization for particle research, publicized the new
World Wide Web project. The Web was invented by British scientist
Tim Berners-Lee in 1989. An early popular
web browser was
ViolaWWW, patterned after
HyperCard and built using the
X Window System. It was eventually replaced in popularity by the
Mosaic web browser. In 1993, the
National Center for Supercomputing Applications at the
University of Illinois released version 1.0 of Mosaic, and by late 1994 there was growing public interest in the previously academic, technical Internet. By 1996 usage of the word
Internet had become commonplace, and consequently, so had its use as a
synecdoche in reference to the World Wide Web.
Meanwhile, over the course of the decade, the Internet successfully accommodated the majority of previously existing public computer networks (although some networks, such as
FidoNet, have remained separate). During the late 1990s, it was estimated that traffic on the public Internet grew by 100 percent per year, while the mean annual growth in the number of Internet users was thought to be between 20% and 50%.
[11] This growth is often attributed to the lack of central administration, which allows organic growth of the network, as well as the non-proprietary open nature of the Internet protocols, which encourages vendor interoperability and prevents any one company from exerting too much control over the network.
[12] As of 31 March 2011, the estimated total number of
Internet users was 2.095 billion (30.2% of world population)
