Header compression mechanisms

Compression of IP packet headers in data networks began to appear in the 1990s, when the Internet began to be democratized and individuals wanted to connect from home using modems. Various working groups within the IETF (Internet Engineering Task Force) began to tackle the problem, and set about improving Internet performance, particularly for IP over low-speed serial links. Header compression is one of the new features designed to improve the performance of low-speed serial links.

The first compression mechanisms were targeted solely at TCP/IPv4 flows, which, along with Telnet, FTP and then the Web, accounted for the majority of traffic. Since then, new applications have appeared and multimedia services have developed, leading to an increase in traffic using the UDP protocol, putting the previously defined header compression mechanisms to the test. The increase in the number of devices connected to the Internet also raised the issue of address scarcity. The IETF therefore set about defining a new version of the IP protocol (called IPv6 for IP version 6) to overcome the addressing problems of the previous IPv4 version. An IPv6 header is 40 bytes long, whereas an IPv4 header is 20 bytes long. All these developments have led to the redefinition of header compression mechanisms with CTCP and CRTP.

The development of new mechanisms is always necessary, as cellular telephony aims to integrate Internet services, mainly multimedia services, and also offer users connectivity to the Internet. Releases 99, 4 and 5 of the 3rd generation UMTS (Universal Mobile Telecommunication System) cellular telephony standard add header compression to their protocol stack to take account of the high transmission error rate on the radio link. This is known as ROHC robust header compression.