Audio-Video Profile with Feedback

Description

The Audio-Video Profile with Feedback (AVPF) is a Real-time Transport Protocol (RTP) profile standardized by 3GPP in TS 26.922, building upon the foundation of the Audio-Video Profile (AVP). AVPF extends the basic RTP/AVP framework by incorporating explicit feedback mechanisms that allow receivers to send quality reports back to media senders. This feedback is transmitted via RTCP (RTP Control Protocol) packets, which operate alongside the main RTP media streams. The profile defines specific RTCP packet types and timing rules for feedback transmission, ensuring that control traffic remains within acceptable bandwidth limits while providing timely information about network conditions and receiver status.

AVPF operates through a structured feedback system where receivers generate reports based on observed packet loss, jitter, delay, and other quality metrics. These reports include Receiver Report (RR) packets with extended statistics, as well as specialized feedback messages like Picture Loss Indication (PLI), Full Intra Request (FIR), and Temporary Maximum Media Stream Bit Rate Request (TMMBR). The feedback messages follow defined timing rules—either immediate feedback for urgent conditions like picture loss, or regular feedback for continuous quality monitoring. This allows senders to adapt their encoding parameters, such as bitrate, frame rate, or error resilience features, in response to real-time network conditions.

Key architectural components of AVPF include the feedback target, which identifies the media sender or middlebox receiving the feedback; feedback messages with specific formats defined in RFC 4585; and timing rules that govern when feedback can be sent to prevent RTCP bandwidth explosion. The profile supports both point-to-point and multipoint scenarios, making it suitable for conferencing applications. In 3GPP networks, AVPF is integrated with multimedia subsystems like IMS (IP Multimedia Subsystem) to enhance services such as VoLTE (Voice over LTE) and ViLTE (Video over LTE), where maintaining consistent quality over variable radio conditions is critical.

AVPF's role in the network is to bridge the gap between application-layer quality requirements and underlying transport conditions. By providing a standardized mechanism for quality feedback, it enables adaptive streaming, error concealment, and congestion control without requiring deep inspection of media packets by network elements. The profile works in conjunction with other 3GPP features like QoS Class Identifiers (QCIs) and dedicated bearers to ensure end-to-end quality for real-time services. Its feedback mechanisms are particularly valuable in mobile environments where radio conditions can change rapidly, allowing media applications to maintain acceptable user experience despite network variability.

Purpose & Motivation

AVPF was created to address the limitations of the original Audio-Video Profile (AVP), which provided basic RTP capabilities but lacked standardized feedback mechanisms for real-time quality adaptation. In early multimedia systems, applications either used proprietary feedback methods or operated without receiver input, leading to suboptimal performance in lossy or congested networks. The need for a standardized feedback profile became apparent with the growth of real-time video conferencing and mobile video services, where network conditions are unpredictable and quality degradation directly impacts user experience.

The primary problem AVPF solves is the inability of media senders to adapt to network conditions without explicit feedback from receivers. Before AVPF, senders could only infer problems through packet loss observation or required out-of-band signaling, which was often too slow or incompatible across different vendors. AVPF provides a standardized, in-band feedback channel that works across interoperable systems, enabling timely adaptation to packet loss, bandwidth changes, and receiver capabilities. This is especially critical in 3GPP networks where radio resource fluctuations are common, and efficient use of bandwidth is essential for service quality and network capacity.

Historical context shows that AVPF emerged alongside the development of advanced video codecs and the expansion of IMS-based services in 3GPP networks. As LTE networks enabled high-quality video calling and streaming, the need for robust quality adaptation mechanisms became paramount. AVPF addressed this by providing the feedback framework necessary for features like adaptive bitrate streaming, error resilience, and congestion control in standardized multimedia services. Its creation was motivated by the industry's move toward all-IP networks where traditional circuit-switched quality guarantees were no longer available, requiring new packet-based mechanisms to maintain service quality.

Key Features

• Standardized RTCP feedback messages for quality reporting
• Support for immediate feedback for urgent conditions like picture loss
• Timing rules to control RTCP bandwidth and prevent congestion
• Compatibility with both point-to-point and multipoint scenarios
• Integration with 3GPP IMS and multimedia services
• Enables adaptive bitrate, error recovery, and congestion control

Evolution Across Releases

Defining Specifications