paketswitching

Paketswitching, commonly referred to as packet switching, is a method of transmitting digital communications by dividing data into smaller units called packets. Each packet contains a portion of the user data together with addressing, sequencing, and error-detection information. Packets from multiple sources may be interleaved and routed independently through a shared network.

In operation, packets are independently forwarded by switches and routers. They may take different paths and arrive out of order. Networks often use store-and-forward forwarding, buffering packets at intermediate nodes until they can be transmitted onward. At the destination, packets are reassembled into the original message using sequence numbers and checksums, while lost packets are detected and recovered by higher-layer protocols.

Historically, paketswitching emerged in the 1960s and 1970s as a scalable alternative to circuit switching, the traditional telephone approach in which a dedicated path is reserved for a conversation. It became the foundation of ARPANET and the modern Internet, enabling efficient sharing of network resources and dynamic routing.

Advantages include efficient bandwidth utilization, resilience to link failures, and the ability to multiplex many communications over the same network infrastructure. Drawbacks can include variable latency and jitter, potential packet loss, and the need for end-to-end protocols to ensure reliability and ordering.

Variants include datagram packet switching, which is connectionless, and virtual-circuit switching, which establishes a logical path through the network. In IP networks, packets carry destination addresses and are routed by routers; common transport protocols such as TCP and UDP operate over packet-switched networks to provide reliability or low-latency delivery. The model supports diverse applications, from email and web traffic to streaming and real-time communication.