hash
stringlengths
32
32
doc_id
stringlengths
5
12
section
stringlengths
5
1.47k
content
stringlengths
0
6.67M
35c11477334a60f3fc6bc1c5bbec28f7
23.801-02
4.8.1.2 Applicability to other aspects
This Key Issue is also relevant to the WT1.2 common exposure frameworks for 6G. This is because the 6G application enablement layer will utilize and potentially re-expose capabilities, which aligns with requirements from other 3GPP working groups. Editor’s note: The applicability of other aspects to this KI is FFS
35c11477334a60f3fc6bc1c5bbec28f7
23.801-02
4.8.1.3 Open issues
Based on the above analysis, the following open issues need to be studied: - How to represent the application enablement layer as part of 3GPP 6G system? - How does the application enablement layer in the 3GPP 6G architecture support serving as a common entry point for consumers to access various capabilities such as...
35c11477334a60f3fc6bc1c5bbec28f7
23.801-02
5 Architecture
This section identifies potential architectural requirements and provides architectural considerations. The detailed structure of this clause will be developed once sufficient progress has been achieved in specifying the key issues.
35c11477334a60f3fc6bc1c5bbec28f7
23.801-02
6 Solutions
This section lists the solutions. The detailed structure of this clause will be developed once after sufficient progress has been achieved on specifying the key issues.
35c11477334a60f3fc6bc1c5bbec28f7
23.801-02
7 Overall evaluation
This section evaluates architecture options and the proposed solutions. The detailed structure of this clause will be developed once after sufficient progress has been achieved on specifying the key issues and solutions.
35c11477334a60f3fc6bc1c5bbec28f7
23.801-02
8 Conclusions
This section provides conclusion and suggestion for normative work for the architecture options and solutions studied in this TR. The detailed structure of this clause will be developed once after sufficient progress has been achieved on Annex A: Use cases and requirement considerations The intent of this annex is t...
7cbc07a015b3b7a0604f125955f91968
26.823
1 Scope
Editors NOTE: scope will be provided
7cbc07a015b3b7a0604f125955f91968
26.823
2 References
The following documents contain provisions which, through reference in this text, constitute provisions of the present document. - References are either specific (identified by date of publication, edition number, version number, etc.) or non‑specific. - For a specific reference, subsequent revisions do not apply. -...
7cbc07a015b3b7a0604f125955f91968
26.823
3 Definitions of terms, symbols and abbreviations
7cbc07a015b3b7a0604f125955f91968
26.823
3.1 Terms
Void
7cbc07a015b3b7a0604f125955f91968
26.823
3.2 Symbols
Void.
7cbc07a015b3b7a0604f125955f91968
26.823
3.3 Abbreviations
For the purposes of the present document, the abbreviations given in TR 21.905 [1] and the following apply. An abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in TR 21.905 [1]. AI Artificial Intelligence AL-FEC Application-Layer Forward Error Correct...
7cbc07a015b3b7a0604f125955f91968
26.823
4 Introduction
Void
7cbc07a015b3b7a0604f125955f91968
26.823
5 Media Applications and Service Scenarios
Editors NOTE: this clause will present the use cases and scenarios considered for the evaluation with simple description example end-to-end procedures and example QoE metrics and 3GPP QoS usage.
7cbc07a015b3b7a0604f125955f91968
26.823
5.1 Introduction
7cbc07a015b3b7a0604f125955f91968
26.823
5.2 Real-Time Communication for Conversational XR
7cbc07a015b3b7a0604f125955f91968
26.823
5.2.1 Description
Real-time communication between two or more users may be augmented by an eXtended Reality (XR) scene shared by all the participants. A user may be represented by his (2D or 3D) avatar. A user may have several XR devices (e.g., XR glasses/headset, immersive audio headset, haptics devices) for a multi-modal immersive exp...
7cbc07a015b3b7a0604f125955f91968
26.823
5.2.2 Typical implementation and end-to-end procedures
The following sample scenario and end-to-end procedure for Real-Time Communication for Conversational XR is provided, derived from the description of section 5.2.1. Both local and remote rendering cases are considered. A group of users are willing to participate to an interactive collaborative AR experience with the su...
7cbc07a015b3b7a0604f125955f91968
26.823
5.2.3 Typical QoE criteria
Real-time services require low end-to-end latency to ensure a suitable user’s QoE for interactive experiences. A pose-to-render-to-photon delay QoE metric, derived from the motion-to-photon latency, for which the rendering is performed remotely, is defined in the TR 26.928 [11]. A threshold value of less than 20-50ms ...
7cbc07a015b3b7a0604f125955f91968
26.823
5.2.4 Example QoS usage in the 3GPP Network
void
7cbc07a015b3b7a0604f125955f91968
26.823
5.3 HTTP Video on demand streaming of segmented media
7cbc07a015b3b7a0604f125955f91968
26.823
5.3.1 Description
Video-on-demand Streaming is a popular way of consuming content. Current popular video protocols for on demand streaming include HTTP Live Streaming (HLS) [14] and Dynamic Adaptive Streaming over HTTP (DASH) [13]. Common Media Application Format (CMAF) [15] is a popular format used by these streaming protocols as it c...
7cbc07a015b3b7a0604f125955f91968
26.823
5.3.2 Typical implementation and end-to-end procedures
The typical procedure here as example focusses on on-demand streaming using DASH [13]. Streaming in the 5G System is addressed in TS 26.501, with typical procedures for DASH streaming in clause 5.7.4. In this clause we provide a simplified procedure to give a global overview of how video on demand streaming can work i...
7cbc07a015b3b7a0604f125955f91968
26.823
5.3.3 Typical QoE criteria
Typical QoS and QoE metrics for Live Streaming described as follows: Quality of Experience (QoE) Metrics documented in 3GPP in TS 26.247 [17] include: - Representation switch events - Average throughput - Initial Playout delay - Media start-up delay
7cbc07a015b3b7a0604f125955f91968
26.823
5.3.4 Example QoS usage in the 3GPP network
In the 5GS QoS model this type of traffic in TS 23.501 [3] this type of traffic is characterized by different 5QI's given as examples that could meet the QoS needs of such a service: - 5QI 4: with GBR QoS Flow, 300ms maximum packet delay budget and packet error rate 10-6 and averaging window is 2000 ms. - 5QI ...
7cbc07a015b3b7a0604f125955f91968
26.823
5.4 HTPP Live streaming of segmented media
7cbc07a015b3b7a0604f125955f91968
26.823
5.4.1 Description
HTTP Live Streaming of segmented media is a form of streaming where content is made gradually available, resembling to some extend television broadcasts. With time more media is becoming available. The content may be created live or may pre-recorded. HTTP Live streaming of segmented media is popular way of distributi...
7cbc07a015b3b7a0604f125955f91968
26.823
5.4.2 Typical implementation and end-to-end procedures
The end-to-end procedure example in the clause focuses on streaming using DASH [13]. Streaming in the 5G System is addressed in TS 26.501, with typical procedures for DASH streaming in clause 5.7.4. In this clause we provide a simplified procedure to give a global overview of how live streaming can operate in pract...
7cbc07a015b3b7a0604f125955f91968
26.823
5.4.3 Typical QoE criteria
Typical QoS and QoE metrics for Live Streaming described as follows: Quality of Experience (QoE) Metrics documented in 3GPP in TS 26.247 include: - Representation switch events - Average throughput - Initial Playout delay - Media start-up delay
7cbc07a015b3b7a0604f125955f91968
26.823
5.4.4 Example QoS usage in the 3GPP network
In 5GS QoS model this type of traffic is as example in TS 23.501 associated to 2 different 5QI: - 5QI 6, non-GBR QoS Flow, 300ms maximum packet delay and packet error rate 10-6, see clause 5.7.4 of TS 23.501. - 5QI 4, GBR QoS Flow, 300ms maximum packet delay and packet error rate 10-6, see clause 5.7.4 of TS 23.50...
7cbc07a015b3b7a0604f125955f91968
26.823
5.5 Short Form Video Download
7cbc07a015b3b7a0604f125955f91968
26.823
5.5.1 Description
Short form video download is a popular form of media consumption used in popular social media platforms. A key difference of this form of media consumption compared to live or on demand video streaming is that videos are shorter (typically up to 120 seconds) and that users interact by “swipes” resulting in a quick sw...
7cbc07a015b3b7a0604f125955f91968
26.823
5.5.2 Typical implementation and end-to-end procedures
The example and generalized procedure for short form video download/streaming is shown in Figure 5.4.2-1. In this example a PDU Session with the default QoS is used to connect to the data network, and this PDU session is used to establish a connection to the Application Server (AS) (in this baseline it assumed the defa...
7cbc07a015b3b7a0604f125955f91968
26.823
5.5.3 Typical QoE criteria
Typical QoE metrics for Short form video download can be described as follows (according to 3GP DASH in TS 26.247 [17] clause 10): Quality of Experience (QoE): (Average) Start-up delay (i.e. the time a video start playing after a swipe) Average throughput / achieved bit-rate Initial playout delay Switch time (i....
7cbc07a015b3b7a0604f125955f91968
26.823
5.5.4 Example 3GPP QoS usage
Quality of Service (QoS) related characteristics for this service are given as examples in TS 23.501 as follows: - 5QI 4 GBR 300 ms packet delay budget, 10-6 packet error rate - 5QI 6 non-GBR 300 ms packet delay budget, 10-6 packet error rate
7cbc07a015b3b7a0604f125955f91968
26.823
5.6 Media upstream transmission for AI inferencing
7cbc07a015b3b7a0604f125955f91968
26.823
5.6.1 Description
5.6.1.1 Discriminative AI AI inferencing can target various AI tasks such as discriminative and generative tasks. Examples of discriminative AI tasks includes object or facial recognition, detection, or image classification. These AI tasks typically rely on a Convolutional Neural Network (CNN) model. Scene unde...
7cbc07a015b3b7a0604f125955f91968
26.823
5.6.2 Typical implementation and end-to-end procedures
The following sample scenario and end-to-end procedure for AI inferencing is provided derived from the description of section 5.6.1, addressing periodic and aperiodic upstream traffic. - A user launches an AI enabled XR application recognizing, labelling and counting particular objects continuously in his environment ...
7cbc07a015b3b7a0604f125955f91968
26.823
5.6.3 Typical QoE criteria
Several QoE metrics have been defined in the section 4.4 of TR 26.847 [26]. Some of these QoE metrics, mostly related to AI inference quality, depend on the type of task performed by the AI model, e.g., mean Average Precision (mAP) for object tracking, or Word Error Rate (WER) for language translation. Other metrics ...
7cbc07a015b3b7a0604f125955f91968
26.823
5.6.4 Example QoS in 3GPP Network
TR 26.847 [26] does not define QoS criteria and metrics for uplink data traffic. Uplink latency: it can be expected that the input data need to be transmitted within a delay-bound (on the uplink) to enable AI inference and related output(s) to meet the expected round-trip end-to-end latency metric (corresponding to th...
7cbc07a015b3b7a0604f125955f91968
26.823
5.7 High Quality Real-Time Conversational communication
7cbc07a015b3b7a0604f125955f91968
26.823
5.7.1 Description
Immersive communication was extensively studied in 5G resulting in reports such as TR 38.838 [36] and TR 26.928 [11]. Normative work was also completed resulting in updated 5G Core and NG-RAN support for high quality real-time immersive and conversational services. Some of these technologies have been adopted for conve...
7cbc07a015b3b7a0604f125955f91968
26.823
5.7.2 Assumed implementation and end-to-end procedures
This end-end example procedure is based on TS 26.506 and webRTC which is can be used for RTC communication similar to what is supported for example in modern web browsers today. Figure 5.7.2-1 possible procedure for Network operator supported RTC Session based on [31] The diagram showing typical procedures for ...
7cbc07a015b3b7a0604f125955f91968
26.823
5.7.3 Typical QoE criteria
Typical QoE metrics for Real time conversational include based on TR 26.244 [24]: • Session setup delay (service setup time) • Average Video Bitrate, • Bitrate Stability/Switches • Frame freezes • re-buffering frequency • Intra frame and inter frame video quality - Playback freezes ...
7cbc07a015b3b7a0604f125955f91968
26.823
5.7.4 Example QoS usage in 3GPP
Typical QoS support examples in TS 23.501 [3] clause 5.7.4 for 5QI include: - 5QI 7, non-GBR, 100 ms, packet delay budget, 10-3 error loss rate
7cbc07a015b3b7a0604f125955f91968
26.823
5.8 Media upstream/downstream AI inferencing
7cbc07a015b3b7a0604f125955f91968
26.823
5.8.1 Description
An AI/ML operation/model may be split into multiple parts according to the current task and environment. The intention is to offload the computation-intensive, energy-intensive parts to network endpoints, whereas leaving the privacy-sensitive and delay-sensitive parts at the end UE. In that case, the UE executes the op...
7cbc07a015b3b7a0604f125955f91968
26.823
5.8.2 Example End-End procedures
Figure 5.6.2 from TR 26.927 [25] shows a basic workflow for split inference between the network and UE. Steps for the procedures shown are described below. The related formats for transmission are discussed in clause 6 of TR 26.927 [25]. Figure 5.8.2-1: Basic workflow for split inference between the network and ...
7cbc07a015b3b7a0604f125955f91968
26.823
5.8.3 Quality of experience metrics
Typical Quality of Experience metrics include time to result (response time) and accuracy of the result. It may depend on the use case (i.e. real-time industrial inference versus non-real time inferencing).
7cbc07a015b3b7a0604f125955f91968
26.823
5.8.4 Quality of experience metrics
In the 5GS QoS model this type of traffic for real time inferencing is associated different 5QI values (i.e. see clause 5.7.4 of TS 23.501): 5QI 85: delay critical GBR QoS flow with 5 millisecond packet delay budget 10-5 packet error rate and 255 maximum burst volume5QI 88: delay critical GBR QoS flow with 10 millisec...
7cbc07a015b3b7a0604f125955f91968
26.823
6 Experimental Evaluation
Editors Note: the clause will describe the related tests the tests setup, and the evaluation of the results
7cbc07a015b3b7a0604f125955f91968
26.823
6.1 General Experimental Approach and Test Setup
Void
7cbc07a015b3b7a0604f125955f91968
26.823
6.2 Evaluation #1: Real-Time Communication for Conversational XR
6.2.1 Description 6.2.1.1 Experimental approach This experimental approach relies on a standalone custom client/server XR platform for a simplified emulation of real-time communication for conversational XR as detailed in Figure 5.2.2-2. An XR scene is shared between the client and the server. The custom XR applic...
7cbc07a015b3b7a0604f125955f91968
26.823
6.3 Experimental Evaluation #2: Video on demand segmented HTTP streaming
6.3.1 Description 6.3.1.1 Experimental approach The on-demand streaming use case will be tested using an experimental approach in real and emulated networks. The communication will be between the client and streaming application server (AS). Different implementations for the client and application server will be eva...
7cbc07a015b3b7a0604f125955f91968
26.823
6.4 Experimental Evaluation #3: HTTP Live streaming of segmented media
6.4.1 Description 6.4.1.1 Experimental approach The HTTP live streaming of segmented media use case will be tested using an experimental approach in real and emulated networks. The communication will be between the client and Streaming application server (AS). Different implementations for the client and applicatio...
7cbc07a015b3b7a0604f125955f91968
26.823
6.5 Experimental Evaluation #4: Short Form form video download
6.5.1 Description 6.5.1.1 Experimental approach The short form video download use case will be tested using an experimental approach in real and emulated networks. The communication will be between the client and short form video server (AS). Different implementations for the client and short form video server will...
7cbc07a015b3b7a0604f125955f91968
26.823
6.6 Experimental Evaluation #5: Media upstream transmission for AI inferencing
6.6.1 Description void 6.6.2 Evaluation void
7cbc07a015b3b7a0604f125955f91968
26.823
6.7 Evaluation #6: High Quality Real-Time Conversational Communication
6.7.1 Description 6.7.1.1 Experimental approach High-quality real time communication for conversational use cases will be tested using an experimental approach in real and emulated networks. The communication will be a (two-way) between the client and the RTC application server. Different implementations for the c...
7cbc07a015b3b7a0604f125955f91968
26.823
7 Dynamic Traffic Characteristics and Enhanced QoS Support for media applications and services
Editors NOTE: this clause will document proposed solutions for characterizing dynamic traffic characteristics and enhanced QoS support and potentially new identified QoE metrics
7cbc07a015b3b7a0604f125955f91968
26.823
7.1 Dynamic Traffic Characteristics
Editors NOTE: present for each case corresponding traffic characteristics
7cbc07a015b3b7a0604f125955f91968
26.823
7.2 Enhanced QoS Usage
Editors NOTE: present for each case corresponding case suggested and solutions for enhanced QoS usage
7cbc07a015b3b7a0604f125955f91968
26.823
7.3 Enhanced QoE Metrics
Editors NOTE: present enhanced QoE metrics if any
7cbc07a015b3b7a0604f125955f91968
26.823
8 Analysis and recommendations
Editors NOTE: this clause will analysis and suggested recommendations for normative work
7cbc07a015b3b7a0604f125955f91968
26.823
9 Conclusion
Editors NOTE: this clause will provide the conclusion Annex A: Change history Change history Date Meeting TDoc CR Rev Cat Subject/Comment New version 2025-11 SA4#134 S4-220498 Agreed Skeleton 0.1.0 2026-02 SA4#135 S4-260312 Agreements after SA4#134 (: S4-252126 TR skeleton) S4-252054: U...
7cbc07a015b3b7a0604f125955f91968
26.823
6.5 Experimental Evaluation #4: Short form video download
6.5.1 Description 6.5.1.1 Experimental approach The short form video download use case will be tested using an experimental approach in real and emulated networks. The communication will be between the client and short form video server (AS). Different implementations for the client and short form video server will...
c87765bd88910b968f2a46607f719dbf
26.836
1 Scope
This clause shall start on a new page. The present document …
c87765bd88910b968f2a46607f719dbf
26.836
2 References
The following documents contain provisions which, through reference in this text, constitute provisions of the present document. - References are either specific (identified by date of publication, edition number, version number, etc.) or non‑specific. - For a specific reference, subsequent revisions do not apply. -...
c87765bd88910b968f2a46607f719dbf
26.836
3 Definitions of terms, symbols and abbreviations
This clause and its three (sub) clauses are mandatory. The contents shall be shown as "void" if the TS/TR does not define any terms, symbols, or abbreviations.
c87765bd88910b968f2a46607f719dbf
26.836
3.1 Terms
For the purposes of the present document, the terms given in TR 21.905 [1] and the following apply. A term defined in the present document takes precedence over the definition of the same term, if any, in TR 21.905 [1]. Definition format (Normal) <defined term>: <definition>. example: text used to clarify abstract r...
c87765bd88910b968f2a46607f719dbf
26.836
3.2 Symbols
For the purposes of the present document, the following symbols apply: Symbol format (EW) <symbol> <Explanation>
c87765bd88910b968f2a46607f719dbf
26.836
3.3 Abbreviations
For the purposes of the present document, the abbreviations given in TR 21.905 [1] and the following apply. An abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in TR 21.905 [1]. Abbreviation format (EW) <ABBREVIATION> <Expansion>
c87765bd88910b968f2a46607f719dbf
26.836
4 QUIC-based media delivery protocols
c87765bd88910b968f2a46607f719dbf
26.836
4.1 General
IETF has been working on the standardization of several QUIC-based application protocols. This clause documents the ones that are considered relevant to real‑time and interactive communication.
c87765bd88910b968f2a46607f719dbf
26.836
4.2 Considered protocols
Editor’s note: Each subsequent clause describes an individual protocol.
c87765bd88910b968f2a46607f719dbf
26.836
4.2.1 WebTransport
c87765bd88910b968f2a46607f719dbf
26.836
4.2.1.1 Introduction
Media over QUIC Transport (MOQT) [18] is a publish/subscribe-based binary data transport protocol that is under development by the IETF MOQ Working Group [30] since 2023, designed to run either directly over QUIC [9] or via WebTransport [20]. Although the protocol was originally created for media applications as its na...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.1.2 Features
c87765bd88910b968f2a46607f719dbf
26.836
4.2.1.3 Benefits and limitations
Benefits: - Leverage the features of QUIC for real-time media delivery (e.g., multiple streams, prioritization) and provides means for integration to a CDN infrastructure. - Convergence to a single media delivery protocol suitable from ingest to distribution simplifies workflows for service providers and enables a un...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.1.4 Current applications
MOQT can be layered on top of WebTransport (see clause 4.2.3). WebTransport implementations over QUIC (HTTP/3) are available both for clients and servers: - Google Chrome and Microsoft Edge browsers implement client-side W3C WebTransport API [33] over QUIC. - Many open-source server-side implementations exist. Some ...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.2 RTP over QUIC (RoQ)
c87765bd88910b968f2a46607f719dbf
26.836
4.2.2.1 Introduction
RTP over QUIC (RoQ) [19] is a protocol that has been under development since 2022 by the IETF AVTCORE (Audio/Video Transport Core Maintenance) Working Group  [31] since 2022. The RoQ draft defines a minimal and flexible mapping that allows existing RTP-based real-time media applications to operate over QUIC instead of ...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.2.2 Features
c87765bd88910b968f2a46607f719dbf
26.836
4.2.2.2.1 Security and encapsulation
QUIC includes built-in encryption (TLS v1.3 [6]) for all traffic, so RTP media packets sent over a RoQ session benefit from confidentiality and integrity without requiring a separate DTLS layer as in the case offor SRTP [4]. RoQ [19] supports QUIC streams and QUIC datagrams [14] as primary encapsulation models for carr...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.2.2.2 Multiplexing
RoQ allows multiplexing multiple RTP and RTCP streamsflows to be multiplexed over a single QUIC connection using flow identifiers, simplifying NAT/firewall traversal and reducing port usage. Instead of using separate UDP ports per flow, an application-level flow identifier is inserted as part of the QUIC payload for bo...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.2.2.3 RTCP considerations
RoQ aims to minimize RTCP traffic by utilizing data already accessible in the QUIC layer. QUIC’s transport-level feedback can be used to complement or partially replace traditional RTCP features, potentially reducing control overhead. - QUIC acknowledgments can be used to compute the lost packet statistics, which are...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.2.3 Benefits and limitations
Benefits: - Reuse established RTP payload formats, media semantics and timing. This supports, for example, audioA/Vvideo lip-sync across multiple streams, etc. - Built-in authentication and encryption via QUIC’s use of /TLS v1.3 transport security (no need for separate DTLS-SRTP [4]). - Multiple media, control, and ...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.2.4 Current applications
Open-source implementations exist: - Go implementation by TUM: https://github.com/mengelbart/roq - Gstreamer plugin by BBC: https://github.com/bbc/gst-roq - Meetecho C library imquic implements RoQ in addition to MOQT: https://github.com/meetecho/imquic/ No commercial deployments have been identified, further explo...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3 Media over QUIC Transport (MOQT)
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.1 Introduction
WebTransport is a modern web API and protocol framework that enables secure, low‑latency, bidirectional communication between browsers or web apps and servers. WebTransport is designed to cover use cases where the WebSocket API is too limited (single, ordered, reliable byte stream over TCP) and where the WebRTC Data Ch...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.2 Features
WebTransport provides a session-based communication model supporting multiple independent unidirectional and bidirectional streams, ensuring reliable and ordered delivery of byte streams. It also allows for unreliable delivery using QUIC datagrams. The API exposes readable/writable streams and datagrams to developers,...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.2.1 Introduction
Below is a non-exhaustive summary of MOQT's key features. For further details, refer to draft-ietf-moq-transport [18].
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.2.2 Object-based data model
MOQT’s data model represents each MOQT schedulable unit of content as an Object within a named Track, organized into Groups and Subgroups. Objects are identified by an Object ID within a Group/Subgroup and carry metadata including Track Alias, Group ID, Object IDs, Publisher Priority, and optional extension headers. A...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.2.3 Publish/Subscribe workflow
MOQT’s publish/subscribe workflow enables publishers to make media Tracks available within designated namespaces, allowing subscribers to select only the namespaces and Tracks they need. Publishers are endpoints that handle subscriptions by sending requested Objects from the requested Track; the initial publisher of a...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.2.4 Data transport over streams and datagrams
Objects are transmitted either on QUIC streams (reliable, ordered) as defined in RFC 9000 [9] or as QUIC datagrams (unreliable, unordered) as defined in RFC 9221 [14]. Grouping mutually dependent Objects together in a QUIC stream can provide operational advantages, such as improved prioritization. This is achieved by ...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.2.5 Relay behaviour and scalability
MOQT Relays support both fan-in and fan-out: they can ingest tracks from multiple publishers (fan-in), acquire Tracks once and serve many subscribers (fan-out), thereby facilitating scalable distribution in a manner analogous to Content Delivery Networks (CDNs). Furthermore, Relays function as policy enforcement points...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.3 Benefits and limitations
Benefits: - Leverage benefits of QUIC within the standard web security model (origin-based access control, secure contexts) providing browser support. The W3C WebTransport API [33] exposes the capabilities of QUIC through a high-level, stream-oriented JavaScript interface aligned with the Web Streams paradigm, offeri...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.4 Related Internet Drafts in the IETF MOQ WG
The IETF Media Over QUIC Working Group is also developing container formats that specify encapsulation of MOQT data and media streaming formats operating over MOQT that specify media packaging as well as signalling mechanisms for negotiation between MOQT endpoints. These formats include: - Low Overhead Media Container...
c87765bd88910b968f2a46607f719dbf
26.836
4.2.3.5 Current applications
- Several open-source implementations of the IETF draft exist. A non-exhaustive list is given below: - Google’s production-ready implementation supports MOQT: https://github.com/google/quiche - Meta provides an experimental MOQT Relay and live encoder/player designed to work with it: - Relay: https://github.com/face...
c87765bd88910b968f2a46607f719dbf
26.836
4.3 Summary
Editor’s note: Provide a summary and comparison of the protocols described in the previous clause.
c87765bd88910b968f2a46607f719dbf
26.836
5 Evaluation of QUIC-based media delivery protocols for RTC