Generic Compression Algorithm for Cellular Traffic

S

Generic Compression Algorithm for Cellular Traffic

Idea: Udi Goori, IL-Hod HaSharon

The price for leasing or installing ATM (Asynchronous Transfer Mode) lines is high. Therefore, it is desirable to compress data between the BTS (Base Transceiver Station) and the BSC (Base Station Controller) in a way that fewer ATM lines will be used.

The idea described by the following works with ATM lines but also works for other types of high speed communication media. The mechanisms proposed are also applicable for communications of ASUB (Interface) lines between the BSC and the TRAU (Transcoder and Rate Adaption Unit).

To compress the ABIS (Interface) traffic, it is necessary to understand a variety of packets which are being transferred. However, the ABIS structure is very complicated, and includes many types of channels (e.g. 8K, 16K, 32K, 64K); many types of frames (e.g. FR (Full Rate), EFR (Enhanced Full Rate), AMR (Adaptive Multi-Rate), Silence, Idle, GPRS (General Packet Radio Service) Data, Edge data, Signaling); and many vendors have different accomplishments of the GSM (Global System for Mobile Communications) standard.

The technical problem is to find an easy and flexible compression algorithm that will compress a major portion of the traffic without the need to identify a plethora of traffic types of different manufacturers with many variants of the GSM protocol. It is not possible to detect the channel types for a wide variety of equipment and to also achieve good compression.

The proposed solution differentiates between voice channels and other channels. Voice channels can easily be detected by state-of-the art technology.

In a second step, voice channels are compressed by detecting silent periods, which usually occupy more than 50% of the traffic and compresses them accordingly. Al...