CHAPTER 1 BEGINNING

2012.3.29/4.10/4.18/7.3/7.5/7.7/8.3/8.24/10.1/10.13/10.18/10.24

Computer network research and development projects started in USA and UK among others in 1950s~1960s.[Waldrop 2001, Leiner 2003, Kleinrock 2010, Kleinrock 2012, Davies 2010, Computer 2012] In Asia, the computer network research and development started in Australia, India, Japan, Korea and New Zealand among others in 1970s with influence of computer network research efforts in USA and Europe.[PCCS 1985] International collaboration on the computer network research started in 1972. ARPANET demonstration was held during the first International Conference on Computer Communications (ICCC).[Leiner 2003]. During the conference it was decided to form International Network Working Group (INWG).[McKenzie 2011]

Many mainframe computer makers developed the computer network products in 1970s including SNA and VNET at IBM. The VNET protocols were used for BITNET later in 1970s. Many minicomputer makers also developed the computer network products in 1970s including DECNET at DEC, which was used for HEPnets around the world.

The computer network protocol called UUCP was also developed for Unix computers in USA in late 1970s. USENET, news group which runs on top of UUCP was developed in 1980. Various PC communication systems were also developed from late 1970s to 1980s.

1.1 North America

Computer network research and development started in USA and UK in 1960s.[Kleinrock 2010, Kleinrock 2012, Davies 2010, Baran 2002] JCR Licklider at ARPA proposed Galactic Network in early 1960s. [Waldrop 2001] He and his successors at ARPA funded the computer network research and development projects in 1960s, and ARPANET was born with four nodes at UCLA, SRI, UCSB and University of Utah in 1969. This was the beginning of the Internet as we know today. In 1972, a large, very successful demonstration of ARPANET was organized at the International Conference on Computer Communications (ICCC).[ICCC 1972] This was the first public demonstration of ARPANET to the public.

Paul Baran at RAND Corporation which is 10 km away from UCLA carried out the computer network research around the same time, but did not developed a computer network.[Baran 2002] Xerox PARC made the major developments on the computer network including Ethernet and PARC Universal Packet (PIP). Ethernet is perhaps the world’s most pervasive network technology today.[Kleinrock 2010]

1.2 Europe

Donald Davies at NPL in UK led the research team to carry out the computer network research and development in 1960s, and he came up with the term, packet switching.[Davies 2010, Kirstein 2009] The computer network, called Mark I was developed to serve NPL campus in 1970, but could not develop the wide area network in UK since his project proposal was not approved by UK Government. Later, the network protocol suit called Coloured Book was developed in 1970s, and its implementation was realized later as SERCnet and JANET in UK.[Kirstein 2010, Davies 2010]

Louis Pouzin led the research team in France to carry on the computer network research and development in 1972, and demonstrated the packet switching network based on datagram in 1973 with one packet network switch and three computers, called CYCLADES.[CYCLADES 2002] The network supported the concept of the end-to-end connectivity based on the datagram.

1.3 Asia

Computer network research and development started among several Asian countries in 1970s including Australia, India, Japan, Korea and New Zealand. Many of them were inspired by the research and development efforts of ARPANET and other projects in North America and Europe. These research and development efforts in Asia had not contacted each other until early 1980s. Some of the efforts are introduced next.

Australia

The University of Melbourne and Wollongong Group exchanged files between two Unix-based computers using a dial-up line. Bob Kummerfeld and Piers Lauder of the University of Sydney started to develop Australian Computer Science network (ACSnet) with Robert Elz of the University of Melbourne and others. The network became operational in 1980.

India

[by Subramani Ramani]

A team at the Tata Institute of Fundamental Research (TIFR) headed by Srinivasan Ramani was active in computer network research and development from the 1970s. Srinivasan Ramani had early exposure to ARPANET facilities at the Carnegie-Mellon University, and started looking at networking technology from a developing country point of view. The TIFR team developed communication software to create a remote station on a minicomputer made in India, and operationalized it by 1977.

The following three institutions had worked together to run the Computer Network Experiment (COMNEX) over the APPLE satellite built by India.

- National Centre for Software Development and Computing Techniques(NCSDCT) at the Tata Institute of Fundamental Research(TIFR),

- Space Application Centre(SAC), and

- Telecom Research Centre(TRC)

COMNEX used a packet switching protocol and connected three cities – Ahmedabad, Bombay and Delhi, and was demonstrated in 1980. It used satellite communication terminals designed by SAC and built by the participating institutions.

The networks team at NCSDCT, headed by Srinivasan Ramani partnered with a government owned company, CMC Ltd., headed by Prem P Gupta then. This partnership triggered a number of projects including the joint development of a message switch which was commercialized later. The partnership also led to the holding of an IFIP TC-6 international conference named Networks-80 in February 1980, in Bombay. [Ramani 1981]

V S Rao working in the database team, headed by P Sadanandan, had designed and implemented an email system at NCSDCT in 1980. The database team, working with the networks team, also developed a proof-of-concept demo of a Railway Passenger Reservation System. CMC launched a project later, to design and develop a nation-wide Passenger Reservation System, the first application of networking technology to impact the lives of millions of Indians.

Japan

[From Haruhisa Ishida, Current Status of the N1 (Japanese inter-university) network with access to supercomputers, PCCS, 1985.]

In 1973, a planning was initiated for an inter-university computer network at the University of Tokyo Computer Centre. Following a preliminary study, a network development project was organized by the members of the University of Tokyo Computer Centre, Kyoto University Computer Center and NTT (Nippon Telegraph '& Telephone Co.). Three Japanese main-framers, Hitachi, Fujitsu, NEC and Tohoku University Computer Center were invited to join the project, thus making it a development project by 7 parties.

Somewhat patterned after the ARPANET in USA, a decision was made to use packet switching service being planned by NTT then. Table 1 summarizes the development of the Nl network resulting from the project. The DDX (Digital Data eXchange) is a digital packet switching network operated by NTT (also line-switching service is available on DDX) since 1980. TLl and TL2 were the early experimental versions of DDX.

By the summer of 1981 when NTT started packet switching service on the DDX as a commercial service, nearly all of 7 large-scale university computer centers joined the Nl network. The features of the Nl network are as follows.

(1) It is a heterogeneous network based on a proprietary protocol called "Nl protocol" and links main-frame computers from Hitachi, Fujitsu, NEC and Mitsubishi. The protocol specification is open to universities.

(2) An all-digital packet-switching network (DDX) is used to connect 33 computers. Thus the communication cost is proportional to the amount of information transmitted (the number of packets). However, since DDX is economical only for long-distance and infrequent use, 18 university computers in Tokyo area are linked to the Tokyo center via leased lines.

(3) The transmission speed is high and typically is 48 kilo-bits/sec. Data is compressed for transmission.

(4) Packet switching allows the multiplexed use of communication lines. A number of TSS terminals and RJE facilities can share the same line.

(5) Users can access Tokyo center from a telephone terminal via PAD (packet assemble/ disassembly) service on DDX and through Venus-P. The latter is an overseas packet switching subnetwork operated by KDD.

(6) The network reliability is high.

Table 1 History of the Nl network (Inter-university network in Japan)

1973 Tokyo Univ., Kyoto Univ., and NTT started development of an experimental network.

1975 Tokyo University started TSS service through telephones

1976 Tokyo and Kyoto · connected through RJE on Nl/TLl

1977 Tokyo and Kyoto connected thr·ough TSS on Nl/TLl

1978 Tokyo and Kyoto connected through TSS on Nl/TL2

1980 NTT starts a commercial digital line switching service on DDX (Digital Data Exchange)

1981 7 university centers join Nl network using packet switching on DDX

1981-1983 Connected Nl network with NTT's DCNA via protocol conversion (Tokyo, Kyoto, Tohoku & NTT)

1984 Tokyo University tested PAD for DDX through 300/1200 bps telephone couplers/modems.

1984-1985 Extended Nl-NVT protocol for Kanji data exchange. Mail facility introduced into Nl network.

1985 NTT started commercial service for PAD/DDX

1985 Universities began international mail exchange through KDD's Venus-P and USENET.

Korea

Korea’s TCP/IP network began with a small computer network research group headed by Kilnam Chon in 1979. The group was formed at Korea Institute of Electronics Technology (KIET), the national laboratory for research and development on computer and semiconductor founded in the late 1970s, and a part of Electronics and Telecommunications Research Institute (ETRI) now. The group started research and development on the computer networks. Concurrently, a graduate course on computer networking was created at Seoul National University (SNU).

In 1980, a research proposal to develop the first computer network in Korea was made to Ministry of Commerce and Industry of Korean Government, but it was rejected. A revised proposal was made to the same ministry as part of National Project on Computer Research and Development in 1981, and it was approved. An implementation plan was developed for this project in 1981, and the network based on TCP/IP was named Software Development Network (SDN).[Chon 1985c] Later, it was renamed as System Development Network.

New Zealand

[by Brain Carpenter, 2012]

Brian Carpenter’s first involvement in a wide-area networking project was in the period 1974-76, at Massey University, Palmerston North, New Zealand. At that time, all the NZ universities had installed the same type of mainframe computer, the Burroughs B6700. There was naturally much interest in data communications. By 1975, the ARPANET was quite well known, and had grown to include 99 computers in the USA, Norway, and the UK. Partly inspired by this, Massey University got together with Victoria University in Wellington in a project that we called Kiwinet. An early debate was the choice of network protocols. There turned out to be two contenders, unless we invented something ourselves: EPSS and, of course, ARPANET. The EPSS was the Experimental Packet Switching System of the British Post Office, a precursor of X.25. In those days the New Zealand Post Office usually followed the lead of the British, so the only way for Kiwinet to get official support was to pick X.25. Unfortunately the project met many practical difficulties; New Zealand finally connected to the Internet in 1989, thirteen years after Brian Carpenter left.

Reference

[Baran 2002] Paul Baran, “The Beginnings of Packet Switching: Some Underlying Concepts,” IEEE Communications Magazine, vol.40, no.7, July 2002, pp.42-48.

[Chon 1985c] Kilnam Chon, et al., "SDN: A Computer Network for Korean Research Community," Proceedings of Pacific Computer Communications Symposium, Seoul, 1985, pp.567-570.

[Computer 2012] Computer History Museum, www.ComputerHistory.org, 2012.

[CYCLADES 2002] A History of CYCLADES, THINK, U. of Texeas, 2002.

[Davies 2010] Howard Davies and Beatrice Bressan, History of International Research Networking, Wiley-Blackwell, May 2010.

[ICCC 1972] ICCC, Washington, D.C., 1972.

[Kirstein 2009] Peter Kirstein, "The early history of packet switching in the UK," IEEE Communications Magazine, Feb. 2009, pp.18-26]

[Kleinrock 2010] Leonard Kleinrock, "An Early History of the Internet," IEEE Communications Magazine, vol.48, no.8, Aug. 2010, pp.26-36.

[Kleinrock 2012] Leonard Kleinrock, "A Brief History of the Internet," Global INET, Geneva, April 2012. [video and ppf]

[Leiner 2003] Bob Leiner, et al., A Brief History of the Internet, Internet Society, 2003 (revised). http://www.internetsociety.org/internet/internet-51/history-internet/brief-history-internet

[McKenzie 2011] Alex McKenzie, "INWG and the concept of the Internet: An Eyewitness Account," IEEE Annals of the History of Computing, vol.33, no.1, Jan. 2011, pp.66-71.

[PCCS 1985] PCCS (Pacific Computer Communications Symposium), Seoul, Korea, 1985. (Also available from North Holland in 1986)

[Ramani 1981] Srinivasan Ramani, Proceedings of IFIP Conference on Networks 80, Bombay, North Holland, 1981.

[Waldrop 2001] Mitchell Waldrop, The Dream Machine: JCR Licklider and theRevolution that Made Computing Personal, Viking Penguin, 2001.