Packet switching over copper telephone lines with Interface Message Processors (IMPS) connected at each end to interface with a host computer, had been demonstrated and met the criteria of no central routing control and could adapt to down lines. An additional feature directed the recipient of a message to send an acknowledgement to the source. If the acknowledgement was not received in time, the source re-sent the packet. To verify integrity of messages, each IMP used a checksum {a hashing function that uses each bit in a message to yield one sum after transmission for comparison with one another}. In addition, each IMP was programmed to be self-sufficient. If an IMP found its operating system to be damaged, it would request a replacement copy from a neighboring IMP or else shut down to prevent harm to the network.

In the early '70s, with over 15 sites connected, the Arpanet would prove not only efficient, but reasonably reliable too. Soon users would begin regarding it as an operational utility, rather than an experiment. As Bob Kahn, one of the founders noted, the large-scale network project combined "theory, simulation and engineering". Arpanet pioneers had proved the feasibility of packet-switching that underlies much of the networking today.

But in 1971, few people used the Arpanet and Kahn recalled "If you did log on, it was rare that you could do anything useful". So he and Larry Roberts, an ARPA executive, decided to make a splash at the International Conference on Computer Communications to be held at the Washington Hilton on Oct. 24-26, 1972. For the next 15 months, Kahn organized the show, aided by several dozen of the best programmers including Vincent Cerf, Jon Postel and Robert Metcafe who developed programs for useful applications from medical to air traffic control.

Over 1000 people saw the show and marvelled at over 100 pieces of equipment functioning together reliably. From the success, Telnet, now part of Sprint's operating net work group, was created with Roberts himself heading the effort at BBN, the company who had developed the IMP. The show succeeded in legitimizing the notion of computer networking. During the month after the conference, traffic on the Arpanet jumped by 67 percent and has continued this pace even until today.

During the conference, the first computer communication global step was taken when the International Working Group was formed with Vincent Cerf as its chairman, a position he held for 4 years. Members included representatives from Britain's National Physical Laboratory and the French Cyclades packet-switching project.

Kahn moved on too as ARPA's program manager in Arlington, Virginia, a post he held for 13 years. It was a unique position from which to foster the Internet and to answer the need to communicate over multiple dissimilar networks. He undertook the advancement of packet satellite radio networks, a project with military advantages. Digital speech was difficult for an enemy to decipher if encrypted. Packets sent by satellites could reach ships at sea or forces on foreign lands. The radio network had different parameters and Kahn chose to solve their problems in a way that could be applied to connecting ARPA to other dissimilar networks.