Description
MABR (Multicast Adaptive Bit Rate) is a Rel-19 3GPP work item that brings adaptive bit rate (ABR) streaming to multicast/broadcast services in 5G. Traditionally, multicast in cellular networks (e.g., eMBMS) used a fixed bit rate for all users, which could lead to poor quality for users with bad channel conditions or wasted resources for those with good conditions. MABR introduces the ability to adapt the media bit rate dynamically, either per user or per group, by leveraging unicast and multicast switching, scalable video coding, or multiple multicast streams at different bit rates. This allows the network to deliver the best possible quality to each user while efficiently using radio resources.
The specifications 26.802 and 26.804 define the media streaming and multicast architecture for MABR. 26.802 covers the dynamic adaptive streaming over HTTP (DASH) and its extensions for multicast, while 26.804 covers the multicast and broadcast service architecture. MABR integrates with the 5G Media Streaming (5GMS) architecture, where the application function (AF) and media server coordinate with the 5G core network to manage multicast delivery. The core network's multicast/broadcast user plane function (MB-UPF) and multicast/broadcast session management function (MB-SMF) handle the multicast transport, while the AF provides service information and adaptation policies.
Key components of MABR include the media server that hosts multiple representations of the content at different bit rates, the client that requests and switches between representations based on network conditions, and the network functions that manage multicast delivery. The client uses standard ABR algorithms (e.g., throughput-based, buffer-based) to select the appropriate bit rate. In multicast mode, the network may deliver multiple multicast streams at different bit rates, and the client joins the stream that matches its capabilities and conditions. Alternatively, the network can use scalable video coding (SVC) where a base layer and enhancement layers are multicast, and clients decode the layers they can receive.
In operation, when a user requests a multicast stream, the AF sets up a multicast session with the 5G core. The media server provides the content in multiple bit rates. The network transmits the multicast streams, and the client monitors its reception quality and switches between streams or layers as needed. The switching can be seamless, with the client buffering to avoid interruptions. The network can also use unicast for users who cannot receive the multicast reliably, ensuring service continuity. MABR thus combines the efficiency of multicast with the adaptability of unicast ABR.
MABR is particularly relevant for live sports, news, and other popular content where many users watch the same stream simultaneously. It enables operators to deliver high-quality video to large audiences without overwhelming the network. By adapting to network conditions, it improves user experience and reduces the need for over-provisioning. Rel-19 MABR builds on previous 3GPP work on eMBMS and 5G multicast, adding the adaptive bit rate capability to meet the demands of modern streaming services.
Purpose & Motivation
MABR was created to solve the challenge of delivering high-quality video to many users simultaneously over cellular networks. Traditional multicast services used a single bit rate, which meant that users with poor channel conditions experienced stuttering or low quality, while those with good conditions received no benefit. Unicast adaptive bit rate streaming solves this by adapting to each user, but it does not scale well for popular content because it requires separate resources for each user. MABR combines the scalability of multicast with the adaptability of unicast, providing an efficient and high-quality solution for mass media delivery.
The motivation for MABR comes from the growing consumption of video on mobile devices and the desire of operators to offer compelling multicast services. Previous multicast technologies like eMBMS did not support adaptive bit rate, limiting their effectiveness for live streaming. With the advent of 5G and its improved multicast capabilities, 3GPP saw an opportunity to standardize MABR, enabling interoperability and widespread adoption. The feature addresses the need for efficient spectrum usage, especially during peak events, and improves the quality of experience for users.
Historically, 3GPP has worked on multicast and broadcast since 3G, with eMBMS in LTE and 5G MBS in Rel-17. MABR in Rel-19 represents the next step, adding adaptive bit rate to multicast. It leverages existing ABR standards like DASH and HLS, and integrates with the 5G media streaming architecture. The development of MABR reflects the ongoing evolution of media delivery in cellular networks, aiming to make multicast a viable alternative to unicast for popular content.
What changed in its specifications
The latest release-to-release comparison of specifications that define MABR. Each summary describes the whole specification, not only this term.
Rel-19 adds CMCD reporting, multi-source media delivery, content steering, QoS enhancements, and QUIC-based streaming.
Full comparison in TS 26.804 →Evolution Across Releases
An overview written from the defining specifications, not a verified change history. For the exact text, open the specifications listed below.
Initial introduction of MABR, defining the architecture and procedures for adaptive bit rate multicast in 5G. It specifies the integration with 5GMS, the use of DASH and SVC, and the interaction with 5G core network functions for multicast delivery. This release establishes the foundation for efficient, high-quality multicast streaming services.
Defining Specifications
3GPP specifications that define or reference MABR, with the latest known release. Sourced from the 3GPP document catalog — see methodology.
| Specification | Title | Release |
|---|---|---|
| TS 26.802 vk00 | 5G Media Streaming Multicast Enhancements | Rel-20 read |
| TS 26.804 vk10 | 5G Media Delivery Architecture Extensions | Rel-20 read |