Filter
Top Products
User Manual - Configuration Guide (Volume 3)
Versatile Routing Platform
Chapter 5
IPHC Configuration
5-1
Chapter 5 IPHC Configuration
5.1 Overview of IPHC
IPHC (IP Header Compression) is the method to classify and compress packet such as
IP, TCP and UDP, according to a series of header compression algorithms specified in
relevant RFC documents, so as to enhance the transmission rate of voice, video and
large packet over low-speed network. Currently, IPHC is mainly applied to the serial
links running such protocols as PPP, FR and HDLC and carrying large amount of voice
information.
As described in RFC2507, the packets need to be compressed can be classified into
the following categories:
1) IP header + TCP header
2) IP header + UDP header
3) IP header + UDP header + RTP header
Among them, (RTP) Realtime Transmit Protocol is an application layer protocol lies
above TCP/UDP, which is mainly used to transmit audio, video and simulated data
information.
The implementation of IPHC depends on the various existing compression algorithms,
whose main concepts are to compress the header since many fields of the headers do
not change or change regularly in the course of one-time connection. It divides header
into TCP and NON_TCP type, and further divides them into smaller categories.
Extracting the unchanged fields or the fields that change regularly from the header of
the same packet type, it does not transmit or only transmit the changed values of those
fields, thus achieving the purpose of compressing the length of the whole packet.
IPHC defines the two ends of the link as compressor and non-compressor end
respectively. The packet processing procedure is described as below:
Compressor: According to the classification method in RFC2507 document, after it
classify and compresses the packets, the compressor notifies the decompressor
through different packet protocol numbers and makes the decompressor perform
different compression in accordance with the corresponding protocol number. If the
compressor receives fault-reporting packet CONTEXT_STATE sent from the
decompressor, it will mark some location digits in the context storage table according to
the information in the packet to identify the faulty packet flow and will performs faulty
processing accordingly. For example, when the next packet arrives, it will transmit
packets to the decompressor in the complete format to help the latter update context
storage table and to implement re-synchronization, etc.
Decompressor: It performs decompression according to the different types of packets
transmitted from the compressor, and restores them into complete packets. If it finds
that there is fault in packet decompression process, it will either discard the faulty
packet or generate CONTEXT_STATE fault-reporting packet according to the actual
conditions.
By using IP header compression on low-speed serial link, it can implement:
z
Enhance the packet Interactive Response Speed
z Reduce the transmission cost of header, enabling small packet to achieve high
link transmission quality and saving bandwidth resource
z
Reduce the packet discard rate on lossy links