The Internet relies heavily on two protocols. In the network layer, IP (Internet Protocol) provides an unreliable datagram service and ensures that any host can exchange packets with any other host. Since its creation in the 1970s, IP has seen the addition of several features, including multicast, IPsec (IP security), and QoS (quality of service). The latest revision, IPv6 (IP version 6), supports 16-byte addresses.
The second major protocol is TCP (Transmission Control Protocol), which operates in the transport layer and provides a reliable bytestream service on top of IP. TCP has evolved continuously since the first experiments in research networks.
Still, one of the early design decisions of TCP continues to frustrate many users. TCP and IP are separate protocols, but the separation between the network and transport protocols is not complete. To differentiate the individual data streams among incoming packets, a receiving end host demultiplexes the packets based on the so-called 5-tuple, which includes the IP addresses, port numbers, and protocol identifiers. This implies that a TCP connection is bound to the IP addresses used on the client and the server at connection-establishment time. Despite the growing importance of mobile nodes such as smartphones and tablets, TCP connections cannot move from one IP address to another. When a laptop switches from Ethernet to Wi-Fi it obtains another IP address. All existing TCP connections must be torn down and new connections restarted….
Christoph Paasch and Olivier Bonaventure. ACM Queue. April 2014
Posted in 