]> Nederlandse Publieke Omroep

james.ietf@gmail.com

Tencent America LLC

spencerdawkins.ietf@gmail.com

applications MOQ Mailing List Internet-Draft QUIC This document describes use cases and requirements that guide the specification of a simple, low-latency media delivery solution for ingest and distribution, using either the QUIC protocol or WebTransport. Note to Readers RFC Editor: please remove this section before publication Source code and issues for this draft can be found at https://github.com/fiestajetsam/draft-gruessing-moq-requirements. Discussion of this draft should take place on the IETF Media Over QUIC (MoQ) mailing list, at https://www.ietf.org/mailman/listinfo/moq.

Introduction This document describes use cases and requirements that guide the specification of a simple, low-latency media delivery solution for ingest and distribution , using either the QUIC protocol or WebTransport as transport protocols.

Note for MOQ Working Group participants When adopted, this document is intended to capture use cases that are in scope for work on the MOQ protocol , and requirements that arise from these use cases. As of this writing, the authors have not planned to request publication on this document, based on our understanding of the IESG's statement on "Support Documents in IETF Working Groups" , which says (among other things):

• While writing down such things as requirements and use cases help to get a common understanding (and often common language) between participants in the working group, support documentation doesn’t always have a high archival value. Under most circumstances, the IESG encourages the community to consider alternate mechanisms for publishing this content, such as on a working group wiki, in an informational appendix of a solution document, or simply as an expired draft.

It seems reasonable for the working group to improve this document, and then consider whether the result justifies publication as a part of the RFC archival document series.

Terminology The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 when, and only when, they appear in all capitals, as shown here.

Distinguishing between Interactive and Live Streaming Use Cases The MOQ charter lists three use cases as being in scope of the MOQ protocol

• use cases including live streaming, gaming, and media conferencing

but does not include (directly or by reference) a definition of "live streaming" or "interactive" (a term that has been used to describe gaming and media conferencing, as distinct from "live streaming"). It seems useful to describe these two terms, as classes of use cases, before we describe individual use cases in more detail. MOQ participants have discussed making this distinction based on quantitative measures such as latency, but since MOQ use cases can include an arbitrary number of relays, we offer a distinction that is based on how users experience that distinction. If two users are able to interact in the way that seems interactive, as described in the proposed definitions, the use case is interactive; if two users are unable to interact in that way, the use case is live streaming. We propose these definitions:

Interactive:

a use case with coupled bidirectional media flows

Interactive use cases have bidirectional media flows sufficiently coupled with each other, that media from one sender can cause the receiver to reply by sending its own media back to the original sender. For instance, a speaker in a conferencing application might make a statement, and then ask, "but what do you folks think?" If one of the listeners is able to answer in a timeframe that seems natural, without without waiting for the current speaker to explicitly "hand over" control of the conversation, this would qualify as "Interactive".

Live Streaming:

a use case with unidirectional media flows, or uncoupled bidirectional flows

Live Streaming use cases allow consumers of media to "watch together", without having a sense that one consumer is experiencing the media before another consumer. This does not require the delivery of live media to be strictly synchronized between media consumers, but only that from the viewpoint of individual consumers, media delivery appears to be synchronized. It is common for live streaming use cases to send media in one direction, and "something else" in the other direction - for instance, a video receiver might be returning requests that the sender change the media encoding or media rate in use, or reorgient a camera. This type of feedback doesn't qualify as "bidirectional media". If two sender/receivers are each sending media to the other, but what's being carried in one direction has no relationship with what's being carried in the other direction, this would not qualify as "Interactive". Note: these descriptions are a starting point. Feedback and pushback are both welcomed.

Use Cases Informing This Proposal Our goal in this section is to understand the range of use cases that are in scope for "Media Over QUIC" . For each use case in this section, we also describe

• the number of senders or receiver in a given session transmitting distinct streams,
• whether a session has bi-directional flows of media from senders and receivers, which may also include timely non-media such as haptics or timed events.

It is likely that we should add other characteristics, as we come to understand them.

Interactive Media The use cases described in this section have one particular attribute in common - the target the lowest possible latency as can be achieved at the trade off of data loss and complexity. For example,

• It may make sense to use FEC and codec-level packet loss concealment , rather than selectively retransmitting only lost packets. These mechanisms use more bytes, but do not require multiple round trips in order to recover from packet loss.
• It's generally infeasible to use congestion control schemes like BBR in many deployments, since BBR has probing mechanisms that rely on temporarily inducing delay, but these mechanisms can then amortize the consequences of induced delay over multiple RTTs.

This may help to explain why interactive use cases have typically relied on protocols such as RTP , which provide low-level control of packetization and transmission, with addtional support for retransmission as an optional extension.

Gaming

Attribute	Value
Senders/Receivers	One to One
Bi-directional	Yes

In this use case the computation for running a video game (single or multiplayer) is performed externally on a hosted service, with user inputs from input devices sent to the server, and media, usually video and audio of gameplay returned. This may also include the client receiving other types of signaling, such as triggers for haptic feedback, as well as the client sending media such as microphone audio for in-game chat with other players. Latency may be considerably important in this use case as updates to video occur in response user input, with certain genres of games having high requirements in responsiveness and/or a high frequency of user input.

Remote Desktop

Attribute	Value
Senders/Receivers	One to Many
Bi-directional	Yes

Similar to the gaming use case in many requirements, but where a user wishes to observe or control the graphical user interface of another computer through local user interfaces. Latency requirements with this use case are marginally different than the gaming use case as greater input latency may be more tolerated by users. This use case may also include a need to support signalling and/or transmitting of files or devices connected to the user's computer.

Video Conferencing/Telephony

Attribute	Value
Senders/Receivers	Many to Many
Bi-directional	Yes

Where media is both sent and received; This may include audio from both microphone(s) and/or cameras, or may include "screen sharing" or inclusion of other content such as slide, document, or video presentation. This may be done as client/server, or peer to peer with a many to many relationship of both senders and receivers. The target for latency may be as large as 200ms or more for some media types such as audio, but other media types in this use case have much more stringent latency targets.

Hybrid Interactive and Live Media For the video conferencing/telephony use case, there can be additional scenarios where the audience greatly outnumbers the concurrent active participants, but any member of the audience could participate. As this has a much larger total number of participants - as many as Live Media Streaming , but with the bi-directionality of conferencing, this should be considered a "hybrid". There can be additional functionality as well that overlap between the two, such as "live rewind", or recording abilities. Another consideration is the limits of "human bandwidth" - as the number of sources are included into a given session increase, the amount of media that can usefully understood by a single person diminishes. To put it more simply - too many people talking at once is much more difficult to understand than one person speaking at a time, and this varies on the audience and circumstance. Subsequently this will define some limitations in the number of potential concurrent or semi-concurrent, bidirectional communications that occur.

Live Media The use cases in this section like those in do set some expectations to minimise high and/or highly variable latency, however their key difference is that are seldom bi-directional as their basis is on mass-consumption of media or the contribution of it into a platform to syndicate, or distribute. Latency is less noticeable over loss, and may be more accepting of having slightly more latency to increase guarantee of delivery.

Live Media Ingest

Attribute	Value
Senders/Receivers	One to One
Bi-directional	No

Where media is received from a source for onwards handling into a distribution platform. The media may comprise of multiple audio and/or video sources. Bitrates may either be static or set dynamically by signaling of connection information (bandwidth, latency) based on data sent by the receiver, and the media may go through additional steps of transcoding or transformation before being distributed.

Live Media Syndication

Attribute	Value
Senders/Receivers	One to One
Bi-directional	No

Where media is sent onwards to another platform for further distribution and not directly used for presentation to an audience, however may be monitored by operational systems and/or people. The media may be compressed down to a bitrate lower than source, but larger than final distribution output. Streams may be redundant with failover mechanisms in place.

Live Media Streaming

Attribute	Value
Senders/Receivers	One to Many
Bi-directional	No

Where media is received from a live broadcast or stream either as a broadcast with fixed duration or as ongoing 24/7 output. The number of receivers may vary depending on the type of content; breaking news events may see sharp, sudden spikes, whereas sporting and entertainment events may see a more gradual ramp up with a higher sustained peak with some changes based on match breaks or interludes. Such broadcasts may comprise of multiple audio or video outputs with different codecs or bitrates, and may also include other types of media essence such as subtitles or timing signalling information (e.g. markers to indicate change of behaviour in client such as advertisement breaks). The use of "live rewind" where a window of media between the live edge and trailing edge can be made available for clients to playback, either because the local player falls behind edge or because the viewer wishes to play back from a point in the past.

Requirements for Protocol Work Our goal in this section is to understand the requirements that result from the use cases described in .

Notes to the Reader

• Note: the intention for the requirements in this document is that they are useful for MOQ working group participants, to recognize constraints, and useful for readers outside the MOQ working group to understand the high-level functionality of the MOQ protocol, as they consider implementation and deployment of systems that rely on the MOQ protocol.

Specific Protocol Considerations In order to support the various topologies and patterns of media flows with the protocol, the protocol MUST support both sending and receiving of media streams, as separate actions or concurrently in a given connection.

Delivery Assurance vs. Delay Different use cases have varying requirements with respect to the tradeoffs associated in having guarantee of delivery vs delay - in some (such as telephony) it may be acceptable to drop some or all of the media as a result of changes in network connectivity, throughput, or congestion whereas in other scenarios all media must arrive at the receiving end even if delayed. There SHOULD be support for some means for a connection to signal which media may be abandoned, and behaviours of both senders receivers defined when delay or loss occurs. Where multiple variants of media are sent, this SHOULD be done so in a way that provides pipelining so each media stream may be processed in parallel.

Support Webtransport/Raw QUIC as media transport There should be a degree of decoupling from the underlying transport protocols and MoQ itself despite the "Q" in the name, in particular to provide future agility and prevent any potential ossification being tied to specific version(s) of dependant protocols. Many of the use cases will be deployed in contexts where web browsers are the common application runtime; thus the use of existing protocols and APIs is desireable for implementations. Support for WebTransport will be defined, although implementations or deployments running outside browsers will not need to use WebTransport, thus support for the protocol running directly atop QUIC should be provided. Considerations should be made clear with respect to modes where WebTransport "falls back" to using HTTP/2 or other future non-QUIC based protocol.

Media Negotiation & Agility All entities which directly process media will have support for a variety of media codecs, both codecs which exist now and codecs that will be defined in the future. Consequently the protocol will provide the capability for sender and receiver to negotiate which media codecs will be used in a given session. The protocol SHOULD remain codec agnostic as much as possible, and should allow for new media formats and codecs to be supported without change in specification. The working group should consider if a minimum, suggestive set of codecs should be supported for the purposes of interop, however this SHOULD avoid being strict to simplify use cases and deployments that don't require certain capability e.g. telephony which may not require video codecs.

Media Data Model As the protocol will handle many different types of media, classifications, and variations when all entities describe the media a model should be defined which represents this, with a clear addressing scheme. This should factor in at least, but not limited to allow future types:

Media Types

Video, audio, subtitles, ancillary data

Classifications

Codec, language, layers

Variations

For each stream, the resolution(s), bitrate(s). Each variant should be uniquely identifiable and addressable.

Considerations should be made to addressing of individual audio/video frames as opposed to groups, in addition to how the model incorporates signalling of prioritisation, media dependency, and cacheability to all entities.

Publishing Media Many of the use cases have bi-directional flows of media, with clients both sending and receiving media concurrently, thus the protocol should have a unified approach in connection negotiation and signalling to send and received media both at the start and ongoing in the lifetime of a session including describing when flow of media is unsupported (e.g. a live media server signalling it does not support receiving from a given client). In the initiation of a session both client and server must perform negotiation in order to agree upon a variety of details before media can move in any direction:

• Is the client authenticated and subsequently authorised to initiate a connection?
• What media is available, and for each what are the parameters such as codec, bitrate, and resolution etc?
• Can media move bi-directionally, or is it unidirectional only?

Naming and Addressing Media Resources As multiple streams of media may be available for concurrent sending such as multiple camera views or audio tracks, a means of both identifying the technical properties of each resource (codec, bitrate, etc) as well as a useful identification for playback should be part of the protocol. A base level of optional metadata e.g. the known language of an audio track or name of participant's camera should be supported, but further extended metadata of the contents of the media or its ontology should not be supported.

Scoped to an Origin/Domain, Application specific.

Allows subscribing or requesting for the data matching the name by the consumers

Packaging Media Packaging of media describes how raw media will be encapsulated. There are at a high level two approaches to this:

• Within the protocol itself, where the protocol defines the ancillary data required to decode each media type the protocol supports.
• A common encapsulation format such as there are advantages to using an existing generic media packaging format (such as CMAF or other ISOBMFF subsets) which define a generic method for all media and handles ancillary decode information.

The working group must agree on which approach should be taken to the packaging of media, taking into consideration the various technical trade offs that each approach provides.

• If the working group decides to describe media encapsulation as part of the MOQ protocol, this will require a new version of the MOQ protocol in order to signal the receiver that a new media encapsulation format may be present.
• If the working group decides to use a common encapsulation format, the mechanisms within the protocol SHOULD allow for new encapsulation formats to be used. Without encapsulation agility, adding or changing the way media is encapsulated will also require a new version of the MOQ protocol, to signal the receiver that a new media encapsulation format may be present.

MOQ protocol specifications will provide details on the supported media encapsulation(s).

Media Consumption Receivers SHOULD be able to as part of negotiation of a session specify which media to receive, not just with respect to the media format and codec, but also the varient thereof such as resolution or bitrate.

Relays, Caches, and other MOQ Network Elements

Pull & Push To enable use cases where receivers may wish to address a particular time of media in addition to having the most recently produced media available, both "pull" and "push" of media SHOULD be supported, with consideration that producers and intermediates SHOULD also signal what media is available (commonly referred to as a "DVR window"). Behaviours around cache durations for each MoQ entity should be defined.

Security

Authentication & Authorisation Whilst QUIC and conversely TLS supports the ability for mutual authentication through client and server presenting certificates and performing validation, this is infeasible in many use cases where provisioning of client TLS certificates is unsupported or infeasible. Thus, support for a primitive method of authentication between MoQ entities SHOULD be included to authenticate entities between one another, noting that implementations and deployments should determine which authorisation model if any is applicable.

Media Encryption End-to-end security describes the use of encryption of the media stream(s) to provide confidentiality in the presence of unauthorized intermediates or observers and prevent or restrict ability to decrypt the media without authorization. Generally, there are three aspects of end-to-end media security:

• Digital Rights Management, which refers to the authorization of receivers to decode a media stream.
• Sender-to-Receiver Media Security, which refers to the ability of media senders and receivers to transfer media while protected from authorized intermediates and observers, and
• Node-to-node Media Security, which refers to security when authorized intermediaries are needed to transform media into a form acceptable to authorized receivers. For example, this might refer to a video transcoder between the media sender and receiver.

**Note: "Node-to-node" refers to a path segment connecting two MOQ nodes, that makes up part of the end-to-end path between the MOQ sender and ultimate MOQ receiver. Support for encrypted media SHOULD be available in the protocol to support the above use cases, with key exchange and decryption authorisation handled externally. The protocol SHOULD provide metadata for entities which process media to perform key exchange and decrypt.

IANA Considerations This document makes no requests of IANA.

Security Considerations As this document is intended to guide discussion and consensus, it introduces no security considerations of its own.

References Normative References In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings. Informative References This memorandum describes RTP, the real-time transport protocol. RTP provides end-to-end network transport functions suitable for applications transmitting real-time data, such as audio, video or simulation data, over multicast or unicast network services. RTP does not address resource reservation and does not guarantee quality-of- service for real-time services. The data transport is augmented by a control protocol (RTCP) to allow monitoring of the data delivery in a manner scalable to large multicast networks, and to provide minimal control and identification functionality. RTP and RTCP are designed to be independent of the underlying transport and network layers. The protocol supports the use of RTP-level translators and mixers. Most of the text in this memorandum is identical to RFC 1889 which it obsoletes. There are no changes in the packet formats on the wire, only changes to the rules and algorithms governing how the protocol is used. The biggest change is an enhancement to the scalable timer algorithm for calculating when to send RTCP packets in order to minimize transmission in excess of the intended rate when many participants join a session simultaneously. [STANDARDS-TRACK] This document describes a framework for using Forward Error Correction (FEC) codes with applications in public and private IP networks to provide protection against packet loss. The framework supports applying FEC to arbitrary packet flows over unreliable transport and is primarily intended for real-time, or streaming, media. This framework can be used to define Content Delivery Protocols that provide FEC for streaming media delivery or other packet flows. Content Delivery Protocols defined using this framework can support any FEC scheme (and associated FEC codes) that is compliant with various requirements defined in this document. Thus, Content Delivery Protocols can be defined that are not specific to a particular FEC scheme, and FEC schemes can be defined that are not specific to a particular Content Delivery Protocol. [STANDARDS-TRACK] This document defines the Opus interactive speech and audio codec. Opus is designed to handle a wide range of interactive audio applications, including Voice over IP, videoconferencing, in-game chat, and even live, distributed music performances. It scales from low bitrate narrowband speech at 6 kbit/s to very high quality stereo music at 510 kbit/s. Opus uses both Linear Prediction (LP) and the Modified Discrete Cosine Transform (MDCT) to achieve good compression of both speech and music. [STANDARDS-TRACK] This document defines the core of the QUIC transport protocol. QUIC provides applications with flow-controlled streams for structured communication, low-latency connection establishment, and network path migration. QUIC includes security measures that ensure confidentiality, integrity, and availability in a range of deployment circumstances. Accompanying documents describe the integration of TLS for key negotiation, loss detection, and an exemplary congestion control algorithm. Google Google Google Google Google This document specifies the BBR congestion control algorithm. BBR ("Bottleneck Bandwidth and Round-trip propagation time") uses recent measurements of a transport connection's delivery rate, round-trip time, and packet loss rate to build an explicit model of the network path. BBR then uses this model to control both how fast it sends data and the maximum volume of data it allows in flight in the network at any time. Relative to loss-based congestion control algorithms such as Reno [RFC5681] or CUBIC [RFC8312], BBR offers substantially higher throughput for bottlenecks with shallow buffers or random losses, and substantially lower queueing delays for bottlenecks with deep buffers (avoiding "bufferbloat"). BBR can be implemented in any transport protocol that supports packet-delivery acknowledgment. Thus far, open source implementations are available for TCP [RFC793] and QUIC [RFC9000]. This document specifies version 2 of the BBR algorithm, also sometimes referred to as BBRv2 or bbr2. Facebook Facebook Facebook Facebook RUSH is an application-level protocol for ingesting live video. This document describes the protocol and how it maps onto QUIC. Discussion Venues This note is to be removed before publishing as an RFC. Discussion of this document takes place on the mailing list (), which is archived at . Source for this draft and an issue tracker can be found at https://github.com/afrind/draft-rush. Twitch Meta Cisco Google This document defines the core behavior for Warp, a live media transport protocol over QUIC. Media is split into objects based on the underlying media encoding and transmitted independently over QUIC streams. QUIC streams are prioritized based on the delivery order, allowing less important objects to be starved or dropped during congestion. cisco Cisco This specification outlines the design for a media delivery protocol over QUIC. It aims at supporting multiple application classes with varying latency requirements including ultra low latency applications such as interactive communication and gaming. It is based on a publish/subscribe metaphor where entities publish and subscribe to data that is sent through, and received from, relays in the cloud. The information subscribed to is named such that this forms an overlay information centric network. The relays allow for efficient large scale deployments. cisco Cisco Private Octopus Inc. Recently new use cases have emerged requiring higher scalability of media delivery for interactive realtime applications and much lower latency for streaming applications and a combination thereof. draft-jennings-moq-arch specifies architectural aspects of QuicR, a media delivery protocol based on publish/subscribe metaphor and Relay based delivery tree, that enables a wide range of realtime applications with different resiliency and latency needs. This specification defines the protocol aspects of the QuicR media delivery architecture. Google The WebTransport Protocol Framework enables clients constrained by the Web security model to communicate with a remote server using a secure multiplexed transport. It consists of a set of individual protocols that are safe to expose to untrusted applications, combined with a model that allows them to be used interchangeably. This document defines the overall requirements on the protocols used in WebTransport, as well as the common features of the protocols, support for some of which may be optional.

Acknowledgements The authors would like to thank several authors of individual drafts that fed into the "Media Over QUIC" charter process:

• Kirill Pugin, Alan Frindell, Jordi Cenzano, and Jake Weissman (,
• Luke Curley (), and
• Cullen Jennings and Suhas Nandakumar (), together with Christian Huitema ().

We would also like to thank Suhas Nandakumar for his presentation, "Progressing MOQ" , at the October 2022 MOQ virtual interim meeting. We used his outline as a starting point for the Requirements section (). We would also like to thank Cullen Jennings for suggesting that we distinguish between interactive and live streaming use cases based on the users' perception, rather than quantitative measurements. In addition we would also like to thank Lucas Pardue, Alan Frindell, and Bernard Aboba for their reviews of the document. James Gruessing would also like to thank Francesco Illy and Nicholas Book for their part in providing the needed motivation.