Description
Traffic Channel Enhanced Full rate Speech (TCH-EFS) is a specific type of Traffic Channel (TCH) in GSM systems dedicated to carrying speech encoded with the Enhanced Full Rate (EFR) codec. The EFR codec, standardized as ETSI/3GPP codec AMR-EFR (Adaptive Multi-Rate - Enhanced Full Rate), operates at a fixed rate of 12.2 kbps. It is an ACELP (Algebraic Code-Excited Linear Prediction) based codec that provides significantly improved voice quality compared to the original GSM Full Rate (FR) codec (13 kbps) and is comparable to the Half Rate (HR) codec's quality but at a higher bit rate. The TCH-EFS channel encompasses the entire physical and logical chain for transporting this encoded speech.
Technically, the TCH-EFS occupies one full timeslot per TDMA frame on the radio interface, similar to a TCH/F (Full Rate TCH). The speech signal is processed by the EFR codec, producing 244 bits per 20 ms speech frame. These bits are classified into Class 1a (most sensitive, with CRC protection), Class 1b (moderately sensitive), and Class 2 (least sensitive). Channel coding applies a convolutional code with rate 1/2 to Class 1 bits, adding redundancy for error protection, while Class 2 bits are transmitted uncoded. The resulting 456 bits per 20 ms block are then interleaved over 8 consecutive TDMA frames to combat fading and burst errors. After ciphering, the bits are modulated using GMSK and transmitted.
The TCH-EFS is managed by the network's Base Station Controller (BSC) and Transcoding and Rate Adaptation Unit (TRAU). The BSC allocates a TCH-EFS based on codec negotiation during call setup (using signaling like the Codec Negotiation procedure). The TRAU, often located between the BSC and the core network, performs the EFR encoding/decoding and adapts the 64 kbps PCM stream from the core to the 12.2 kbps radio interface rate. TCH-EFS operation is detailed in TS 46.055, which specifies the transcoding and channel coding rules. It ensures interoperability between mobile stations and networks supporting EFR, providing a consistent high-quality voice experience.
Purpose & Motivation
TCH-EFS was developed to address the voice quality limitations of the original GSM Full Rate (FR) speech codec, which, while a breakthrough for digital cellular, exhibited noticeable artifacts and poor performance in noisy environments. The Enhanced Full Rate codec, initially developed for the North American D-AMPS (IS-136) standard as the IS-641 codec, was adopted by GSM to provide 'wireline-like' quality. This was a key competitive feature as mobile networks matured and user expectations for voice clarity increased. The TCH-EFS channel provided the standardized bearer to deploy this improved codec across GSM networks globally.
The primary problem solved was enhancing the perceived voice quality without requiring additional radio spectrum or changing the fundamental TDMA channel structure. The TCH-EFS uses the same radio resources (one timeslot per frame) as TCH/F but employs a more advanced codec algorithm. This allowed network operators to upgrade software in handsets and network elements (TRAU) to support EFR, offering a tangible quality improvement to subscribers. It also improved performance in conditions with background noise and reduced listener fatigue during long calls.
Historically, TCH-EFS was part of the evolution towards more efficient and adaptive speech coding. It preceded the Adaptive Multi-Rate (AMR) codec, which became the ultimate GSM/UMTS voice codec. AMR could dynamically switch between multiple rates (including 12.2 kbps EFR mode) based on channel conditions, optimizing quality versus capacity. However, TCH-EFS remained specified as a fixed-mode option for compatibility and scenarios where AMR negotiation is not used. Its inclusion from Release 8 onwards in 3GPP specs ensures support in modern networks for legacy interoperability and testing, as many GSM networks continued to use EFR as a high-quality fallback or preferred codec.
Key Features
- Carries speech encoded with the Enhanced Full Rate (EFR) codec at 12.2 kbps
- Provides near-toll-quality voice, superior to original GSM Full Rate
- Uses ACELP-based coding with sophisticated error protection classes
- Occupies one full timeslot per TDMA frame (same as TCH/F)
- Includes specific channel coding and interleaving defined in TS 46.055
- Enabled via codec negotiation during call setup
Evolution Across Releases
Initial standardization of the TCH-EFS channel in 3GPP TS 46.055. Defined the complete transcoding and channel coding procedures for the Enhanced Full Rate speech codec on a full-rate traffic channel. Established the mapping of EFR-encoded speech frames (244 bits/20ms) into the GSM channel coding structure (456 bits/block) for transmission over the radio interface.
Defining Specifications
| Specification | Title |
|---|---|
| TS 46.055 | 3GPP TR 46.055 |