Hacker Culture A to Z

Chapter 16. P

Packet switching

All data transmitted over computer networks, whether through the internet or much smaller networks, is sent in the form of packets. It’s a way of managing the bandwidth in our vast global telecommunications infrastructure. Paul Baran at the RAND Corporation for ARPA (now DARPA) and Donald Davies of the National Physical Laboratory in England invented packet switching separately in the early and mid-1960s, and the concept of network packets soon followed.

ARPAnet, the precursor to the modern internet, launched in 1969. The various academic institutions, government entities, and big tech companies that joined it used a wide range of packet-switching standards, many of which were technologically incompatible. To address this problem, Vint Cerf and Bob Kahn invented TCP/IP in the mid-1970s.

TCP/IP packets have components that interface with four of the seven layers of the open systems interconnection (OSI) model—application layer, transport layer, network layer, and data link layer. There are a wide range of TCP/IP ports which use different-looking packets. Most TCP/IP packets contain a header, payload, and trailer. The header contains the sender and recipient’s IP addresses, an indicator of the protocol (such as HTTPS, SMTP, or FTP), and the packet’s number in its sequence. The payload is the content itself, such as the audio and video data in the YouTube videos you watch, or the digital file of an ebook. Finally, the trailer declares the end of the ...