A core design philosophy of the Internet and communication systems, the end-to-end principle states that application-specific functions, like reliability and error checking, should reside in the end hosts of a network rather than in the intermediary nodes. The network itself should be kept simple, focusing only on delivering packets. This places intelligence at the edges and creates a ‘dumb’ network core.
The end-to-end principle was a radical departure from the design of earlier telecommunication networks, like the telephone system, which placed significant intelligence and state management within the network core. The argument is that intermediary nodes (routers) cannot fully and correctly implement functions for all possible applications. For example, ensuring a file transfer is error-free requires acknowledgment from the final recipient, something a router partway through the path cannot guarantee. By implementing this logic in the end hosts (e.g., using TCP), the network core can be simplified to just forwarding packets. This design has profound implications: it fosters 创新, as new applications can be deployed without changing the network infrastructure. It also enhances scalability and robustness, as the failure of an intermediary node does not corrupt the state of an ongoing 沟通, which is managed by the endpoints. However, the principle is not absolute and has been debated, with network functions like firewalls and Network Address Translation (NAT) violating a strict interpretation.
迎接新挑战
机械工程师、项目、工艺工程师或研发经理
可在短时间内接受新的挑战。
通过 LinkedIn 联系我
塑料金属电子集成、成本设计、GMP、人体工程学、中高容量设备和耗材、精益制造、受监管行业、CE 和 FDA、CAD、Solidworks、精益西格玛黑带、医疗 ISO 13485
(如果日期不详或不相关，例如 "流体力学"，则对其显著出现的时间作了四舍五入的估计）。
相关发明、创新和技术原理
{{标题}}
{%，如果摘录 %}{{ 摘录 | truncatewords：55 }}
{% endif %}
