Leonard Kleinrock of MIT and later UCLA developed the theory of packet switching. they proposed that apart from circuit switching (where a new connection has to be made whenever the communication takes place) communication can also take place by transferring data in the form of packets containing their unique address. This would allow electronic signals to be sent and received between different machines without a need for a separate circuit connection every time the communication takes place. This theory of packet switching formed the basis of internet connections as we know today.

The first long-distance communication link was set up between two computers, one in Massachusetts and the other in California by an MIT professor Lawrence Roberts. The link was established with dial-up telephone lines, which worked on the basis of circuit-switching. There were two significant outcomes of this event. First, it proved that wide area networking was indeed a possibility and secondly the technique of circuit switching was insufficient for such a network.

In 1966 DARPA initiated a project called ARPANET, with the help of Lawrence Roberts and his team. Then in 1969, the team was able to link the computers of four universities i.e. UTAH University, UCSB, UCLA, and Stanford University. This was the first time packet switching was used instead of circuit switching. The ARPANET was the first proper form of the modern-day internet.

Soon the ARPANET began to grow, more and more universities and research institutes began to join the network including NASA/Ames, MIT, and Harvard, etc. by 1971 the ARPANET covered almost all the major universities in North America.

Initially, ARPANET was about “Time Sharing” and not about establishing communication links between people. In the words of Larry Roberts “sending messages between people was not an important motivation for a network of scientific computers” it was built to facilitate time sharing between different computing machines. Time-sharing would have allowed the processors to draw additional computing power from idle computers, to perform complex&nbsp.scientific and business intelligence calculations.