FQC

Frame Quality Classification

Interface
Introduced in Rel-8
Frame Quality Classification (FQC) is a mechanism used on the Iu (between RNC and Core Network) and Nb (between MSC Servers) interfaces in 3GPP UMTS networks. It classifies the quality of received speech frames (e.g., as 'Good', 'Bad', or 'Erased') to enable selective frame processing and error concealment, optimizing voice quality across network nodes.

Description

Frame Quality Classification (FQC) is a functional process defined for the Iu and Nb interfaces within the 3GPP UMTS and pre-LTE core network architecture. Its primary role is to assess and tag the quality of individual speech frames as they are transported between network elements, specifically between the Radio Network Controller (RNC) and the Core Network (CN) via the Iu interface, and between Media Gateway (MGW) nodes or MSC Servers via the Nb interface. This classification is crucial because speech frames can be corrupted over the radio interface (Uu) or during internal transport. FQC provides a standardized 'quality indicator' that travels with the frame payload (e.g., in the Iu frame protocol or NbUP protocol), informing downstream nodes about the frame's integrity so they can apply appropriate handling strategies.

On the Iu interface, the RNC performs the initial FQC. It receives frames from the User Equipment (UE) over the air interface and, based on physical layer checks (like CRC results), classifies each frame. A typical classification is 'Good' (frame is intact), 'Bad' (frame is corrupted and unrecoverable), or potentially 'Erased' (frame was expected but not received). This classification is then embedded into the Iu Data Frame sent to the CN's Media Gateway (MGW). The MGW, upon receiving these frames, can use the FQC tag to decide its processing. For instance, a 'Bad' frame might be discarded entirely, or more intelligently, passed to the speech codec's decoder to trigger its internal Frame Erasure Concealment (FEC) or Frame Pattern Substitution (FPS) algorithms, rather than attempting to decode corrupted bits which could produce loud artifacts.

On the Nb interface, which interconnects MGWs in the core network (e.g., in a transcoding or tandem-free operation scenario), FQC tags are preserved and can be forwarded. This allows quality-aware routing and processing across multiple network hops. The mechanism prevents the propagation of corrupted frames and enables consistent application of error concealment at the final point of decoding, which is often at the terminating MGW or the UE. FQC is specified in protocols like Iu User Plane (IuUP) and Nb User Plane (NbUP), detailing the frame structure and quality field encoding. It is a key enabler for maintaining end-to-end voice quality in UMTS by making frame loss and corruption a explicit, managed parameter rather than a hidden impairment.

Purpose & Motivation

FQC was developed to address the challenge of maintaining voice quality in a segmented network where different segments (radio, transport, core) have different error characteristics. In early digital systems, errors were often handled locally at each segment boundary, which could lead to suboptimal overall quality. For UMTS, with its separate RAN and CN domains, there was a need for a coordinated quality management strategy. FQC solves this by providing a common 'language' of frame quality that travels with the user data, allowing each network node to make informed decisions on how to handle each frame.

The historical context is the evolution from GSM's Full Rate/Enhanced Full Rate codecs, where error handling was more tightly bound to the radio link, to the UMTS era with its wider range of codecs (like AMR) and more complex network architecture involving media gateways and potentially multiple transcoding points. Without FQC, a corrupted frame might be blindly forwarded by an MGW, only to cause a severe audio artifact when finally decoded. Alternatively, an MGW might unnecessarily discard a frame that a downstream node could have partially recovered. FQC enables intelligent, end-to-end quality-aware processing, allowing the network to minimize the perceptual impact of errors by applying concealment at the most effective point, typically as close to the listener as possible.

Key Features

  • Classifies speech frames into quality categories (e.g., Good, Bad, Erased)
  • Operates on the Iu interface (RNC to CN) and Nb interface (between MGWs)
  • Embeds quality tags within user plane protocol frames (IuUP/NbUP)
  • Enables downstream nodes (MGWs, UE) to apply appropriate error concealment
  • Prevents propagation and decoding of irrecoverably corrupted frames
  • Facilitates quality-aware processing in tandem-free operation scenarios

Evolution Across Releases

Rel-8 Initial

Specified as part of the Iu User Plane (IuUP) and Nb User Plane (NbUP) protocols for UMTS. Defined the fundamental frame structure with a quality indicator field, enabling the RNC to classify frames based on radio link CRC and forward this classification to the core network for informed media processing.

Defining Specifications

SpecificationTitle
TS 26.102 3GPP TS 26.102
TS 26.202 3GPP TS 26.202
TS 26.453 3GPP TS 26.453
TS 26.454 3GPP TS 26.454
TS 29.163 3GPP TS 29.163