Packet (computing): structure, role and history of network data units

Typical structure

Packets are usually divided into several logical parts. A common breakdown is:

• Header: control information such as source and destination addresses, protocol identifiers, sequence numbers and quality-of-service markers.
• Payload (data): the actual content being transported — part of a file, a message, or an application datagram.
• Trailer or footer: optional integrity checks (for example, a cyclic redundancy check) or end-of-packet markers.

Different protocol layers attach their own headers as a packet descends the stack (for example, an application payload inside a transport header inside a network header). Typical link technologies impose a maximum packet size (MTU); when data exceed that size they are split into fragments and reassembled at the destination.

Packet switching and history

Packets are central to packet switching, a method that routes independent chunks of data over shared links. The idea emerged in the 1960s and was implemented in early research networks such as ARPANET. Pioneers in the field demonstrated that dividing messages into small blocks improved resilience and utilization compared with dedicating a whole circuit to one conversation. Packet switching underlies the design of many protocols and devices used in modern information technology systems.

Protocols, routing and transport

At the network layer, protocols like IP define how packets are addressed and forwarded. Transport-layer protocols such as TCP and UDP provide end-to-end services: TCP offers ordered, reliable delivery using sequence numbers and acknowledgements; UDP provides a simpler, lower-overhead datagram service. Routers examine packet headers to make forwarding decisions, and switches operate on link-layer frames to move packets within local segments.

Uses, advantages and notable facts

Packets enable multiplexing (many conversations sharing the same links), fault tolerance (alternate paths), and flexible bandwidth use. They also introduce behaviours that application designers must handle: variable delay (jitter), possible loss, reordering of fragments, and the need for retransmission or error correction. Important operational concepts include maximum transmission unit (MTU), fragmentation and reassembly, sequence numbers, and quality-of-service markings used to prioritize traffic.

Distinctions and examples

Not all networked communication is packet-based: traditional telephone circuits use a dedicated channel, while some real-time systems stream data continuously. Common packet-based examples include web browsing, email, file transfer, and real-time applications like VoIP and online gaming, which often use different transport strategies to balance latency and reliability. For further general background see network concepts and resources on computer networking and the Internet.