How Data Travels as Packets

Data is not sent as one giant stream. It is split into small chunks called packets, so many conversations share the network and lost bits are easy to resend.

A packet has two parts: a header with addresses and a payload with the real data. The header is like the address on an envelope.

Every packet carries a source and destination address, so devices along the way know exactly where to forward it. No address, no delivery.

Packets from one message can take different paths to the same place. Each travels on its own, letting the network route around traffic and failures.

Since packets can arrive out of order, the receiver uses sequence numbers to reassemble them into the original message. Missing one? It just asks again.

Packet switching lets many users share the same links at once, taking turns. That is how one connection carries email, video, and browsing together.

Packets sometimes get lost or damaged. Protocols spot the gap and just resend that one piece, not the whole file. That is why big downloads still finish.

Packets have a max size, the MTU (often 1500 bytes on Ethernet). Bigger messages get split to fit, which avoids costly chopping up mid-journey.

Each layer renames the chunk: a packet at the network layer, a frame at the link layer, a segment at transport. Same idea, different stage.

Splitting data into packets gives networks flexibility and resilience: they share links fairly, route around trouble, and recover from loss easily.

Picture mailing a book as numbered pages in separate envelopes. They may arrive out of order, but the numbers rebuild it — exactly how packets work.

Why split data into packets?

Data splits into packets with a header and payload. They travel independently and get reassembled in order — making networks fast and resilient.