End-to-End Operation

The end-to-end principle in computer networking states that certain application-specific features, such as reliability and security, should be implemented in the communicating end nodes of the network rather than intermediary nodes like gateways and routers. This principle was first articulated in 1981 by Saltzer, Reed, and Clark, but its meaning has been continuously reinterpreted since then. The principle is based on the premise that adding these features to the communication subsystem incurs resource penalties, and it is more efficient for the end hosts to implement them.