Mode-set

Description

Mode-set is a term used in the context of the Adaptive Multi-Rate (AMR) and Adaptive Multi-Rate Wideband (AMR-WB) speech codecs. These codecs are widely used in 3GPP networks for voice services, including VoLTE and VoNR. The mode-set parameter defines which codec modes (bit rates) are permitted for a session. Each AMR mode corresponds to a specific bit rate and robustness level, and the mode-set allows endpoints to agree on a subset of these modes for communication.

In a typical session, the mode-set is negotiated during session setup using protocols such as SIP/SDP (Session Initiation Protocol / Session Description Protocol). The offerer includes a mode-set in the SDP offer, listing the codec modes it supports. The answerer selects a subset of those modes that it also supports and includes the selected mode-set in the SDP answer. This negotiation ensures that both endpoints use compatible modes and can adapt to changing radio conditions by switching between modes within the agreed set.

The mode-set is particularly important for interoperability between different devices and networks. For example, a UE may support all eight AMR modes, while a network may only support a subset. By negotiating a mode-set, the session can proceed with the common modes. The mode-set also enables dynamic adaptation: during a call, the codec can switch between modes within the set based on network conditions, improving voice quality and capacity.

In 3GPP specifications, the mode-set is defined in TS 26.114 (IMS Multimedia Telephony) and TS 26.980 (AMR-WB). The mode-set is signaled as a parameter in the SDP 'a=fmtp' line, with a list of mode indices. For AMR, the modes are numbered 0 to 7 (or 0 to 8 for AMR-WB). The absence of a mode-set implies that all modes are supported, but explicit negotiation is recommended for interoperability.

The mode-set concept has evolved with the introduction of new codecs like EVS (Enhanced Voice Services), which uses a different negotiation mechanism. However, for AMR and AMR-WB, mode-set remains a key parameter for ensuring seamless voice communication across diverse networks and devices.

Purpose & Motivation

The mode-set exists to facilitate interoperability and adaptability in voice sessions using AMR and AMR-WB codecs. AMR and AMR-WB are multi-mode codecs, meaning they can operate at different bit rates (modes) to adapt to varying radio conditions. Without a mode-set, endpoints might assume support for all modes, leading to mismatches if one endpoint cannot handle a particular mode. The mode-set allows endpoints to explicitly agree on the set of modes that can be used, ensuring that both parties can encode and decode speech correctly.

Historically, early voice codecs like GSM FR and EFR were single-mode, so no mode negotiation was needed. The introduction of AMR in 3GPP Release 98/99 brought multi-mode capability, which required a mechanism to negotiate the modes. The mode-set was defined to solve this problem. It also enables efficient use of network resources by allowing the codec to switch to lower bit rates during congestion, within the agreed set.

The mode-set is crucial for VoLTE and VoNR deployments, where UEs from different vendors must interoperate. It ensures that a call can be established even if the endpoints have different capabilities. Furthermore, the mode-set supports dynamic adaptation during a call, which is essential for maintaining voice quality in challenging radio conditions. As networks evolve to 5G, the mode-set remains relevant for AMR-WB and is complemented by EVS for higher quality.