Computer engineer Paul Baran of BBN (Bolt,Beranek,Newman) writes a paper, “On Distributed Communication Networks”, describing what later becomes known as packet switching, in which digital data are sent over a distributed network in small units and reassembled into a whole message at the receiving end. Packet switching will be an integral part of the ARPANET a few years later.
Baran’s paper introduced two radical ideas at that time. The first proposed changing the structure of the existing communication networks. Instead of a common decentralized network that used several interconnected main centers linked to nearby locations, he proposed a distributed network, where each point was only connected to it’s nearest neighbors. This way, messages had multiple pathways to their destinations.
The second idea introduced a theory that messages could be chopped up into smaller chunks of data, and then reassembled at the receiving end. This allowed each one to travel by a different route, because the traditional method of streaming data was inefficient. These ideas prove to be so unconventional none of the telco giants were interested, and even Baran moved on to other projects. It wasn’t until later that Larry Roberts, then head of ARPA, adopted it for their new computer networking program.