Network Working Group                                         C. Perkins
Internet-Draft                                     University of Glasgow
Intended status: Standards Track                           M. Westerlund
Expires: September 13, December 6, 2012                                       Ericsson
                                                                  J. Ott
                                                        Aalto University
                                                          March 12,
                                                            June 4, 2012

  Web Real-Time Communication (WebRTC): Media Transport and Use of RTP
                     draft-ietf-rtcweb-rtp-usage-02
                     draft-ietf-rtcweb-rtp-usage-03

Abstract

   The Web Real-Time Communication (WebRTC) framework provides support
   for direct interactive rich communication using audio, video, text,
   collaboration, games, etc. between two peers' web-browsers.  This
   memo describes the media transport aspects of the WebRTC framework.
   It specifies how the Real-time Transport Protocol (RTP) is used in
   the WebRTC context, and gives requirements for which RTP features,
   profiles, and extensions need to be supported.

Status of this Memo

   This Internet-Draft is submitted in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF).  Note that other groups may also distribute
   working documents as Internet-Drafts.  The list of current Internet-
   Drafts is at http://datatracker.ietf.org/drafts/current/.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   This Internet-Draft will expire on September 13, December 6, 2012.

Copyright Notice

   Copyright (c) 2012 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents

   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  4
   2.  Terminology  . . . . . . . . . . . . . . . . . . . . . . . . .  4
   3.  Discussion and  Rationale  . . . . . . . . . . . . . . . . . . .  4
     3.1.  Supported RTP Topologies . . . . . . . . .  4
   3.  Terminology  . . . . . . . .  4
     3.2.  Signalling Requirements . . . . . . . . . . . . . . . . .  7  5
   4.  WebRTC Use of RTP: Core Protocols  . . . . . . . . . . . . . .  8  6
     4.1.  RTP and RTCP . . . . . . . . . . . . . . . . . . . . . . .  8  6
     4.2.  Choice of RTP Profile  . . . . . . . . . . . . . . . . . .  8  7
     4.3.  Choice of RTP Payload Formats  . . . . . . . . . . . . . .  9  7
     4.4.  RTP Session Multiplexing . . . . . . . . . . . . . . . . . 10  8
     4.5.  RTP and RTCP Multiplexing  . . . . . . . . . . . . . . . . 10  8
     4.6.  Reduced Size RTCP  . . . . . . . . . . . . . . . . . . . . 10  9
     4.7.  Symmetric RTP/RTCP . . . . . . . . . . . . . . . . . . . . 11  9
     4.8.  Generation of the RTCP Canonical Name (CNAME)  . . . . . . 11 10
   5.  WebRTC Use of RTP: Extensions  . . . . . . . . . . . . . . . . 12 10
     5.1.  Conferencing Extensions  . . . . . . . . . . . . . . . . . 12 10
       5.1.1.  Full Intra Request . . . . . . . . . . . . . . . . . . 13 11
       5.1.2.  Picture Loss Indication  . . . . . . . . . . . . . . . 13 11
       5.1.3.  Slice Loss Indication  . . . . . . . . . . . . . . . . 13 11
       5.1.4.  Reference Picture Selection Indication . . . . . . . . 13 12
       5.1.5.  Temporary Maximum Media Stream Bit Rate Request  . . . 13 12
     5.2.  Header Extensions  . . . . . . . . . . . . . . . . . . . . 14 12
       5.2.1.  Rapid Synchronisation  . . . . . . . . . . . . . . . . 14 12
       5.2.2.  Client to Mixer Audio Level  . . . . . . . . . . . . . 14 13
       5.2.3.  Mixer to Client Audio Level  . . . . . . . . . . . . . 15 13
   6.  WebRTC Use of RTP: Improving Transport Robustness  . . . . . . 15 13
     6.1.  Retransmission . . . . . . . . . . . . . . . . . . . . . . 15 14
     6.2.  Forward Error Correction (FEC) . . . . . . . . . . . . . . 16 15
       6.2.1.  Basic Redundancy . . . . . . . . . . . . . . . . . . . 17 15
       6.2.2.  Block Based FEC  . . . . . . . . . . . . . . . . . . . 18 16
       6.2.3.  Recommendations for FEC  . . . . . . . . . . . . . . . 19 17
   7.  WebRTC Use of RTP: Rate Control and Media Adaptation . . . . . 19 17
     7.1.  Congestion Control Requirements  . . . . . . . . . . . . . 20 19
     7.2.  Rate Control Boundary Conditions . . . . . . . . . . . . . 19
     7.3.  RTCP Limiations  . . . . . . . . . . . . . . . . . . . . . 20
     7.3. 19
     7.4.  Legacy Interop Limitations . . . . . . . . . . . . . . . . 21 20
   8.  WebRTC Use of RTP: Performance Monitoring  . . . . . . . . . . 22 21
   9.  IANA Considerations  WebRTC Use of RTP: Future Extensions . . . . . . . . . . . . . 21
   10. Signalling Considerations  . . . . . . . . 22
   10. Security Considerations . . . . . . . . . . 21
   11. WebRTC API Considerations  . . . . . . . . . 22
   11. Acknowledgements . . . . . . . . . 23
     11.1. API MediaStream to RTP Mapping . . . . . . . . . . . . . . 23

   12. References RTP Implementation Considerations  . . . . . . . . . . . . . . 23
     12.1. RTP Sessions and PeerConnection  . . . . . . . . . . . . 23
     12.1. Normative References . 24
     12.2. Multiple Sources . . . . . . . . . . . . . . . . . . 23
     12.2. Informative References . . . 25
     12.3. Multiparty . . . . . . . . . . . . . . . . . 26
   Authors' Addresses . . . . . . . 25
     12.4. SSRC Collision Detection . . . . . . . . . . . . . . . . . 27

1.  Introduction

   The Real-time Transport Protocol (RTP) [RFC3550] provides a framework
   for delivery of audio and video teleconferencing data and other real-
   time media applications.  Previous work has defined the
     12.5. Contributing Sources . . . . . . . . . . . . . . . . . . . 28
     12.6. Media Synchronization  . . . . . . . . . . . . . . . . . . 29
     12.7. Multiple RTP protocol,
   plus numerous profiles, payload formats, and other extensions.  When
   combined with appropriate signalling, these form the basis for many
   teleconferencing systems.

   The Web Real-Time communication (WebRTC) framework is a new activity
   that provides support for direct, interactive, and rich communication
   using audio, video, collaboration, games, etc., between two peers'
   web-browsers.  This memo describes how the RTP framework is used in
   the WebRTC context.  It proposes a baseline set End-points  . . . . . . . . . . . . . . . . . 29
     12.8. Simulcast  . . . . . . . . . . . . . . . . . . . . . . . . 30
     12.9. Differentiated Treatment of RTP features that
   should be implemented by all WebRTC-aware browsers, along with some
   suggested extensions for enhanced functionality.

   The WebRTC overview [I-D.ietf-rtcweb-overview] outlines the complete
   WebRTC framework, of which this memo is a part.

2.  Terminology

   This memo is structured into different topics.  For each topic, one
   or several recommendations from the authors are given.  When it comes
   to the importance of extensions, or the need for implementation
   support, we use three requirement levels to indicate the importance
   of the feature to the WebRTC specification:

   REQUIRED:  Functionality that is absolutely needed to make the WebRTC
      solution work well, or functionality of low complexity that
      provides high value.

   RECOMMENDED:  Should be included as its brings significant benefit,
      but the solution can potentially work without it.

   OPTIONAL:  Something that is useful in some cases, but not always a
      benefit.

3.  Discussion and Rationale

3.1.  Supported Flows  . . . . . . . . . . . . 30
   13. IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 31
   14. Security Considerations  . . . . . . . . . . . . . . . . . . . 32
   15. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 32
   16. References . . . . . . . . . . . . . . . . . . . . . . . . . . 32
     16.1. Normative References . . . . . . . . . . . . . . . . . . . 32
     16.2. Informative References . . . . . . . . . . . . . . . . . . 35
   Appendix A.  Supported RTP Topologies

   RTP supports both unicast and group communication, with participants
   being connected using wide range of transport-layer topologies.  Some
   of these topologies involve only the end-points, while others use RTP
   translators and mixers to provide in-network processing.  Properties
   of some RTP topologies are discussed in [RFC5117], and we further
   describe those expected to be useful for WebRTC in the following.

                            +---+         +---+
                            | A |<------->| B |
                            +---+         +---+

                         Figure 1:  . . . . . . . . . . . . . . 37
     A.1.  Point to Point

   The point to point topology (Figure 1) is going to be very common in
   any single user to single user applications.

                              +---+      +---+
                              | A |<---->| B |
                              +---+      +---+
                                ^         ^
                                 \       /
                                  \     /
                                   v   v
                                   +---+
                                   | C |
                                   +---+

                          Figure 2: Multi-unicast

   For small multiparty sessions it is practical enough to create RTP
   sessions by letting every participant send individual unicast RTP/UDP
   flows to each of the other participants.  This is called multi-
   unicast (Figure 2), and is unfortunately not discussed in the RTP
   Topologies [RFC5117].  This topology has the benefit of not requiring
   central nodes.  The downside is that it increases the used bandwidth
   at each sender by requiring one copy of the media streams for each
   participant that are part of the same session beyond the sender
   itself.  Thus this is limited to scenarios with few end-points unless
   the media is very low bandwidth.

   This topology may be implemented as a single RTP session, spanning
   multiple peer to peer transport layer connections, or as several
   pairwise RTP sessions, one between each pair of peers.  The later
   approach simplifies rate adaptation, but reduces the effectiveness of
   RTCP for debugging purposes, by limiting the ability to send third-
   party RTCP reports.

                    +---+      +------------+      +---+
                    | A |<---->|            |<---->| B |
                    +---+      |            |      +---+
                               |   Mixer    |
                    +---+      |            |      +---+
                    | C |<---->|            |<---->| D |
                    +---+      +------------+      +---+

                Figure 3: RTP . . . . . . . . . . . . . . . . . . . . . . 37
     A.2.  Multi-Unicast (Mesh) . . . . . . . . . . . . . . . . . . . 40
     A.3.  Mixer with Only Unicast Paths

   An RTP mixer (Figure 3) is a centralised point that selects or mixes
   content in a conference to optimise the RTP session so that each end-
   point only needs connect to one entity, the mixer.  The mixer also
   reduces the bit-rate needs as the media sent from the mixer to the
   end-point can be optimised in different ways.  These optimisations
   include methods like only choosing media from the currently most
   active speaker or mixing together audio so that only one audio stream
   is required in stead of 3 in the depicted scenario.  The downside of
   the mixer is that someone is required to provide the actual mixer.

                    +---+      +------------+      +---+
                    | A |<---->|            |<---->| B |
                    +---+      |            |      +---+
                               | Based  . . . . . . . . . . . . . . . . . . . . . . . 43
       A.3.1.  Media Mixing . . . . . . . . . . . . . . . . . . . . . 43
       A.3.2.  Media Switching  . . . . . . . . . . . . . . . . . . . 46
       A.3.3.  Media Projecting . . . . . . . . . . . . . . . . . . . 49
     A.4.  Translator |
                    +---+      |            |      +---+
                    | C |<---->|            |<---->| D |
                    +---+      +------------+      +---+

         Figure 4: RTP Based . . . . . . . . . . . . . . . . . . . . . 52
       A.4.1.  Transcoder . . . . . . . . . . . . . . . . . . . . . . 52
       A.4.2.  Gateway / Protocol Translator (Relay) with Only Unicast Paths

   If one wants a less complex central node it is possible to use an
   relay (called an Transport Translator) (Figure 4) that takes on the
   role  . . . . . . . . . . . . 53
       A.4.3.  Relay  . . . . . . . . . . . . . . . . . . . . . . . . 55
     A.5.  End-point Forwarding . . . . . . . . . . . . . . . . . . . 59
     A.6.  Simulcast  . . . . . . . . . . . . . . . . . . . . . . . . 60
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 61

1.  Introduction

   The Real-time Transport Protocol (RTP) [RFC3550] provides a framework
   for delivery of forwarding the media to the audio and video teleconferencing data and other end-points but doesn't
   perform any real-
   time media processing.  It simply forwards applications.  Previous work has defined the media from all RTP protocol,
   along with numerous profiles, payload formats, and other to all extensions.
   When combined with appropriate signalling, these form the other.  Thus one endpoint A will only need to send basis for
   many teleconferencing systems.

   The Web Real-Time communication (WebRTC) framework is a
   media once to new protocol
   framework that provides support for direct, interactive, real-time
   communication using audio, video, collaboration, games, etc., between
   two peers' web-browsers.  This memo describes how the relay, but it will still receive 3 RTP streams with framework
   is to be used in the media if B, C and D WebRTC context.  It proposes a baseline set of
   RTP features that must be implemented by all currently transmits.

                               +------------+
                               |            |
                    +---+      |            |      +---+
                    | A |<---->| Translator |<---->| B |
                    +---+      |            |      +---+
                               |            |
                               +------------+

               Figure 5: Translator towards Legacy end-point

   To support legacy end-point (B) that don't fulfil WebRTC-aware browsers,
   along with suggested extensions for enhanced functionality.

   The WebRTC overview [I-D.ietf-rtcweb-overview] outlines the requirements of complete
   WebRTC it framework, of which this memo is possible to insert a Translator (Figure 5) that takes on
   the role to ensure that from A's perspective B looks like a fully
   compliant end-point.  Thus it is the combination of the Translator
   and B that looks like the end-point B. part.

   The intention is that the
   presence structure of the translator is transparent to A, however it is not
   certain that is possible.  Thus this case memo is include so that it can
   be discussed if any mechanism specified to be used for WebRTC results as follows.  Section 2 outlines our
   rationale in such issues preparing this memo and how to handle them.

3.2.  Signalling choosing these RTP features.
   Section 3 defines requirement terminology.  Requirements for core RTP is built with the assumption of an external signalling channel
   protocols are described in Section 4 and recommended RTP extensions
   are described in Section 5.  Section 6 outlines mechanisms that can be used
   increase robustness to configure network problems, while Section 7 describes
   the RTP sessions required congestion control and their features. rate adaptation mechanisms.  The basic configuration of an RTP session consists
   discussion of the following
   parameters: required RTP Profile:  The name mechanisms concludes in Section 8 with a
   review of the RTP profile to performance monitoring and network management tools that
   can be used in session.  The
      RTP/AVP [RFC3551] the WebRTC context.  Section 9 gives some guidelines
   for future incorporation of other RTP and RTP/AVPF [RFC4585] profiles can interoperate RTP Control Protocol (RTCP)
   extensions into this framework.  Section 10 describes requirements
   placed on basic level, as can their secure variants RTP/SAVP [RFC3711]
      and RTP/SAVPF [RFC5124].  The secure variants of the profiles do
      not directly interoperate with the non-secure variants, due to signalling channel.  Section 11 discusses the
      presence of additional header fields in addition to any
      cryptographic transformation
   relationship between features of the packet content.

   Transport Information:  Source and destination address(s) and ports
      for RTP and RTCP MUST be signalled for each RTP session.  If RTP framework and RTCP multiplexing [RFC5761] is to be used, such that a single
      port is used for RTP the WebRTC
   application programming interface (API), and RTCP flows, this MUST be signalled (see Section 4.5).  If several 12 discusses RTP sessions are to be multiplexed onto
      a single transport layer flow, this MUST also be signalled (see
      Section 4.4).
   implementation considerations.  This memo concludes with an appendix
   discussing several different RTP Payload Types, media formats, Topologies, and media format
   parameters:  The mapping between media type names (and hence how they affect the
   RTP
      payload formats to be used) session(s) and various implementation details of possible
   realization of central nodes.

2.  Rationale

   The RTP framework comprises the RTP payload type numbers must
      be signalled.  Each media type may also have a number of media
      type parameters that must also be signalled to configure data transfer protocol, the codec RTP
   control protocol, and numerous RTP payload format (the "a=fmtp:" line from SDP).

   RTP Extensions:  The formats, profiles, and
   extensions.  This range of add-ons has allowed RTP extensions one intends to use need meet various
   needs that were not envisaged by the original protocol designers, and
   to support many new media encodings, but raises the question of what
   features should be
      agreed upon, including any parameters for each respective
      extension.  At the very least, this will help avoiding using
      bandwidth for features that supported by new implementations?  The development
   of the other end-point will ignore.  But
      for certain mechanisms there is requirement for this to happen as
      interoperability failure otherwise happens.

   RTCP Bandwidth:  Support WebRTC framework provides an opportunity for exchanging RTCP Bandwidth values us to review the
      end-points will be necessary, as described in "Session Description
      Protocol (SDP) Bandwidth Modifiers for
   available RTP Control Protocol (RTCP)
      Bandwidth" [RFC3556], or something semantically equivalent.  This
      also ensures that the end-points have features and extensions, and to define a common view
   baseline feature set for all WebRTC implementations of RTP.  This
   builds on the RTCP
      bandwidth, this is important as too different view past 15 years development of the
      bandwidths may lead to failure to interoperate.

   These parameters are often expressed in SDP messages conveyed within
   an offer/answer exchange. RTP does not depend on SDP or on the
   offer/answer model, but does require all the necessary parameters to
   be agreed somehow, and provided to mandate the RTP implementation.  We note use of
   extensions that in RTCWEB context it will depend on have shown widespread utility, while still remaining
   compatible with the signalling model and API
   how these parameters need to be configured but they will be need to
   either wide installed base of RTP implementations where
   possible.

   While the baseline set of RTP features and extensions defined in this
   memo is targetted at the API or explicitly signalled between requirements of the peers.

4. WebRTC Use framework, it is
   expected to be broadly useful for other conferencing-related uses of
   RTP.  In particular, it is likely that this set of RTP: Core Protocols RTP features and RTCP are flexible and extensible protocols that allow, on the
   one hand, choosing from a variety of building blocks and combining
   those to meet application needs, but on the other hand, offer the
   ability to create extensions where existing mechanisms are not
   sufficient.  This memo requires a number of RTP and RTCP extensions
   that have been shown to be provide important functionality in the
   WebRTC context be implemented.  It is possible that future
   extensions will be needed: several documents provide guidelines apppropriate for the other desktop or mobile video
   conferencing systems, or for room-based high-quality telepresence
   applications.

3.  Terminology

   This memo specifies various requirements levels for implementation or
   use and
   extension of RTP features and RTCP, including Guidelines for Writers extensions.  When we describe the importance
   of RTP
   Payload Format Specifications [RFC2736] and Guidelines extensions, or the need for Extending implementation support, we use the RTP Control Protocol [RFC5968], and should be consulted before
   extending this memo.

   The
   following describe core features requirement levels to specify the importance of RTP, and core extensions,
   that must be supported the feature
   in all WebRTC implementations.

4.1.  RTP and RTCP

   The Real-time Transport Protocol (RTP) [RFC3550] is REQUIRED to be
   implemented as the media transport protocol for WebRTC.  RTP itself
   comprises two parts: WebRTC framework:

   MUST:  This word, or the RTP data transfer protocol, and terms "REQUIRED" or "SHALL", mean that the RTP
   control protocol (RTCP).  RTCP
      definition is a fundamental and integral part an absolute requirement of the RTP protocol, specification.

   SHOULD:  This word, or the adjective "RECOMMENDED", mean that there
      may exist valid reasons in particular circumstances to ignore a
      particular item, but the full implications must be understood and
      carefully weighed before choosing a different course.

   MAY:  This word, or the adjective "OPTIONAL", mean that an item is REQUIRED
      truly optional.  One vendor may choose to be implemented.

4.2.  Choice of RTP Profile

   The complete specification of RTP for include the item because
      a particular application domain marketplace requires it or because the choice of an RTP Profile.  For WebRTC use, vendor feels
      that it enhances the "Extended
   Secure RTP Profile for Real-time Transport Control Protocol (RTCP)-
   Based Feedback (RTP/SAVPF)" [RFC5124] is REQUIRED to be implemented.
   This builds on product while another vendor may omit the basic RTP/AVP profile [RFC3551],
      same item.  An implementation which does not include a particular
      option MUST be prepared to interoperate with another
      implementation which does include the RTP/AVPF
   feedback profile [RFC4585], and option, though perhaps with
      reduced functionality.  In the secure RTP/SAVP profile
   [RFC3711].

   The RTP/AVPF part of RTP/SAVPF is required same vein an implementation which
      does include a particular option MUST be prepared to get interoperate
      with another implementation which does not include the improved RTCP
   timer model, that allows more flexible transmission option
      (except, of RTCP packets
   in response to events, rather than strictly according to bandwidth.
   This also saves RTCP bandwidth and will commonly only use course, for the full
   amount when there is feature the option provides.)

   These key words are used in a lot of events on which to send feedback.  This
   functionality is needed to make use manner consistent with their definition
   in [RFC2119].

4.  WebRTC Use of RTP: Core Protocols

   The following sections describe the core features of RTP conferencing
   extensions discussed in Section 5.1.  The improved and RTCP timer model
   defined by RTP/AVPF is backwards compatible
   that MUST be implemented, along with legacy systems the mandated RTP profiles and
   payload formats.  Also described are the core extensions providing
   essential features that all WebRTC implementations MUST implement only the RTP/AVP profile given some constraints to
   function effectively on
   parameter configuration such as RTCP banwidth today's networks.

4.1.  RTP and "trr-int". RTCP

   The RTP/SAVP part of RTP/SAVPF Real-time Transport Protocol (RTP) [RFC3550] is REQUIRED to be
   implemented as the media transport protocol for support for Secure WebRTC.  RTP (SRTP)
   [RFC3711].  This provides media encryption, integrity protection,
   replay protection itself
   comprises two parts: the RTP data transfer protocol, and the RTP
   control protocol (RTCP).  RTCP is a limited form fundamental and integral part of source authentication.

   (tbd: There is ongoing discussion on what keying mechanism is to be
   required, what are the mandated cryptographic transforms,
   RTP, and whether
   fallback to plain RTP should be supported.  This section needs to MUST be
   updated based on the results of that discussion.)

4.3.  Choice of implemented in all WebRTC applications.

   The following RTP Payload Formats

   (tbd: say something about the choice and RTCP features are sometimes omitted in limited
   functionality implementations of RTP Payload Format RTP, but are REQUIRED in all WebRTC
   implementations:

   o  Support for
   WebRTC.  If there is use of multiple simultaneous SSRC values in a mandatory to implement set single
      RTP session, including support for RTP end-points that send many
      SSRC values simultaneously.

   o  Random choice of codecs, this
   should reference them.  In any case, it should reference SSRC on joining a discussion session; collision detection
      and resolution for SSRC values.

   o  Support reception of signalling RTP data packets containing CSRC lists, as
      generated by RTP mixers.

   o  Support for sending correct synchronization information in the choice of codec, once that discussion reaches
   closure.)

   Endpoints may signal
      RTCP Sender Reports, with RECOMMENDED support for the rapid RTP
      synchronisation extensions (see Section 5.2.1).

   o  Support for standard RTCP packet types, include SR, RR, SDES, and
      BYE packets.

   o  Support for multiple media formats, or multiple
   configurations of end-points in a single format, provided each uses a different RTP
   payload type number.  An endpoint that has signalled it's support session, and for
   multiple formats is REQUIRED
      scaling the RTCP transmission interval according to accept data the number of
      participants in any the session; support randomised RTCP transmission
      intervals to avoid synchronisation of RTCP reports.

   It is known that a significant number of legacy RTP implementations,
   especially those formats
   at any time, unless it has previously signalled limitations on it's
   decoding capability (this is modified if several RTP sessions are to
   be multiplexed onto one transport layer connection, such that an
   endpoint must be prepared targetted for a source to switch between formats purely VoIP systems, do not support
   all of the same media type at any time; see above features.

   Other implementation considerations are discussed in Section 4.4).  To support rapid
   rate adaptation, 12.

4.2.  Choice of RTP does not require advance signalling for changes
   between payload formats that were signalled during session setup.

4.4. Profile

   The complete specification of RTP Session Multiplexing

   An association amongst for a set particular application domain
   requires the choice of participants communicating with RTP
   is known as an RTP session.  A participant may be involved in
   multiple RTP sessions at the same time.  In a multimedia session,
   each medium has typically been carried in a separate RTP session with
   its own RTCP packets (i.e., one RTP session for Profile.  For WebRTC use, the audio, with a
   separate "Extended
   Secure RTP session running on a different transport connection Profile for
   the video; if SDP Real-time Transport Control Protocol (RTCP)-
   Based Feedback (RTP/SAVPF)" [RFC5124] is used, this corresponds REQUIRED to one RTP session for
   each "m=" line in be implemented.
   This builds on the basic RTP/AVP profile [RFC3551], the SDP).  WebRTC implementations of RTP are
   REQUIRED to implement support of multimedia sessions in this way, profile
   for
   compatibility with legacy systems.

   In today's networks, however, with RTCP-based feedback (RTP/AVPF) [RFC4585], and the prolific use secure RTP
   profile (RTP/SAVP) [RFC3711].

   The RTP/AVPF part of Network
   Address Translators (NAT) and Firewalls (FW), there RTP/SAVPF is a desire required to
   reduce get the number improved RTCP
   timer model, that allows more flexible transmission of transport layer ports used by real-time media
   applications using RTP by combining multimedia traffic RTCP packets
   in a single
   RTP session.  (Details of how this is to be done are tbd, but see
   [I-D.lennox-rtcweb-rtp-media-type-mux],
   [I-D.holmberg-mmusic-sdp-bundle-negotiation] and
   [I-D.westerlund-avtcore-multiplex-architecture].)  WebRTC
   implementations of RTP are REQUIRED response to support multiplexing of
   multimedia sessions onto a single RTP session events, rather than strictly according to (tbd).  If
   such RTP session multiplexing bandwidth.
   This is vital for being able to be used, this MUST be negotiated
   during the signalling phase.

4.5.  RTP and report congestion events.  The RTP/
   AVPF profile also saves RTCP Multiplexing

   Historically, RTP bandwidth, and RTCP have been run on separate UDP ports.  With
   the increased will commonly only use of Network Address/Port Translation (NAPT) this has
   become problematic, since maintaining multiple NAT bindings can be
   costly.  It
   the full RTCP bandwidth allocation when there are many events that
   require feedback.  The RTP/AVPF functionality is also complicates firewall administration, since multiple
   ports must be opened needed to allow RTP traffic.  To reduce these costs and
   session setup times, support for multiplexing make
   use of the RTP data packets and conferencing extensions discussed in Section 5.1.

      Note: The enhanced RTCP control packets on a single port [RFC5761] timer model defined in the RTP/AVPF
      profile is REQUIRED.

   Note backwards compatible with legacy systems that implement
      only the use of RTP and RTCP multiplexed base RTP/AVP profile, given some constraints on parameter
      configuration such as the RTCP bandwidth value and "trr-int" (the
      most important factor for interworking with RTP/AVP end-points via
      a single port
   ensures that there is occasional traffic sent on that port, even if
   there gateway is no active media traffic.  This may be useful to keep-alive
   NAT bindings and set the trr-int parameter to a value representing
      4 seconds).

   The RTP/SAVP part of the RTP/SAVPF profile is recommend method for application level keep-
   alives of support for Secure
   RTP sessions [RFC6263].

4.6.  Reduced Size RTCP

   RTCP packets are usually sent as compound RTCP packets, (SRTP) [RFC3711].  This provides media encryption, integrity
   protection, replay protection and RFC 3550
   requires that those compound a limited form of source
   authentication.

   WebRTC implementation MUST NOT send packets start with an SR or RR packet.
   When using frequent feedback messages, these general statistics are
   not needed in every packet and unnecessarily increase the mean RTCP
   packet size.  This can limit the frequency at which RTCP packets can
   be sent within RTP/AVP profile
   or the RTCP bandwidth share.

   To avoid this problem, [RFC5506] specifies how to reduce RTP/AVPF profile; they MUST use the mean
   RTCP message and allow RTP/SAVPF profile.  WebRTC
   implementations MUST support DTLS-SRTP [RFC5764] for more frequent feedback.  Frequent
   feedback, in turn, key-management.

   (tbd: There is essential to make real-time application quickly
   aware of changing network conditions and allow them to adapt their
   transmission and encoding behaviour.  Support for RFC5506 ongoing discussion on what additional keying mechanism
   is
   REQUIRED.

4.7.  Symmetric RTP/RTCP

   RTP entities choose to be required, what are the RTP and RTCP transport addresses, i.e., IP
   addresses and port numbers, mandated cryptographic transforms.
   This section needs to receive packets be updated based on and bind their
   respective sockets to those.  When sending the results of that
   discussion.)

4.3.  Choice of RTP packets, however, they
   may use a different IP address or port number Payload Formats

   (tbd: say something about the choice of RTP Payload Format for RTP, RTCP, or both;
   e.g., when using
   WebRTC.  If there is a different socket instance for sending and for
   receiving.  Symmetric RTP/RTCP requires that the IP address and port
   number for sending and receiving RTP/RTCP packets are identical.

   The reasons for using symmetric RTP is primarily mandatory to avoid issues with
   NAT and Firewalls by ensuring that the flow is actually bi-
   directional and thus kept alive and registered as flow the intended
   recipient actually wants. implement set of codecs, this
   should reference them.  In addition any case, it saves resources in the form should reference a discussion
   of ports at the end-points, but also in signalling for the network as NAT mappings choice of codec, once that discussion reaches
   closure.)
   Endpoints may signal support for multiple media formats, or firewall state is not unnecessary bloated.  Also the number multiple
   configurations of QoS
   state are reduced.

   Using Symmetric a single format, provided each uses a different RTP and RTCP [RFC4961]
   payload type number.  An endpoint that has signalled it's support for
   multiple formats is REQUIRED.

4.8.  Generation REQUIRED to accept data in any of those formats
   at any time, unless it has previously signalled limitations on it's
   decoding capability.  This is modified if several media types are
   sent in the RTCP Canonical Name (CNAME)

   The RTCP Canonical Name (CNAME) provides a persistent transport-level
   identifier for an RTP endpoint.  While the Synchronisation Source
   (SSRC) identifier for an same RTP endpoint may change if session, in that case a collision source (SSRC) is
   detected, or when the
   restricted to switch between any RTP application is restarted, it's RTCP CNAME payload format established for
   the media type that is meant to stay unchanged, so being sent by that source; see Section 4.4.
   To support rapid rate adaptation, RTP endpoints can be uniquely
   identified and associated with their does not require signalling in
   advance for changes between payload formats that were signalled
   during session setup.

4.4.  RTP media streams.  For proper
   functionality, RTCP CNAMEs should be unique among the participants Session Multiplexing

   An association amongst a set of participants communicating with RTP
   is known as an RTP session.

   The RTP specification [RFC3550] includes guidelines for choosing a
   unique RTP CNAME, but these are not sufficient  A participant may be involved in
   multiple RTP sessions at the presence of NAT
   devices. same time.  In addition, some may find long-term persistent identifiers
   problematic from a privacy viewpoint.  Accordingly, support for
   generating multimedia session,
   each medium has typically been carried in a short-term persistent separate RTP session with
   its own RTCP CNAMEs following method (b)
   as specified in Section 4.2 of "Guidelines packets (i.e., one RTP session for Choosing the audio, with a
   separate RTP Control
   Protocol (RTCP) Canonical Names (CNAMEs)" [RFC6222] session running on a different transport connection for
   the video; if SDP is REQUIRED,
   since used, this addresses both concerns.

5. corresponds to one RTP session for
   each "m=" line in the SDP).  WebRTC Use implementations of RTP: Extensions

   There RTP are a
   REQUIRED to implement support for multimedia sessions in this way,
   for compatibility with legacy systems.

   In today's networks, however, with the widespread use of Network
   Address/Port Translators (NAT/NAPT) and Firewalls (FW), it is
   desirable to reduce the number of transport layer ports used by real-
   time media applications using RTP extensions that could be very useful by combining multimedia traffic in the
   WebRTC context.  One set
   a single RTP session.  (Details of how this is related to conferencing, others be done are more
   generic in nature.  None of these are mandatory to implement, tbd,
   but
   they are expected to be generally useful.

5.1.  Conferencing Extensions

   RTP is inherently a group communication protocol.  Groups can be
   implemented using a centralised server, multi-unicast, or IP
   multicast.  While IP multicast was popular in early deployments, in
   today's practice, overlay-based conferencing dominates, typically
   using one or more central servers to connect endpoints in a star or
   flat tree topology.  These central servers can be implemented in a
   number of ways [RFC5117], of which the following are the most common:

   1.  RTP Translator (Relay) with Only Unicast Paths ([RFC5117],
       section 3.3)

   2.  RTP Mixer with Only Unicast Paths ([RFC5117], section 3.4)

   3.  Point to Multipoint Using a Video Switching MCU ([RFC5117],
       section 3.5)

   4.  Point to Multipoint see [I-D.lennox-rtcweb-rtp-media-type-mux],
   [I-D.holmberg-mmusic-sdp-bundle-negotiation] and
   [I-D.westerlund-avtcore-multiplex-architecture].)  Using Content Modifying MCUs ([RFC5117],
       section 3.6)

   As discussed in [RFC5117] section 3.5, the use of a video switching
   MCU makes single RTP
   session also effects the use of RTCP possibility for congestion control, or any type differentiated treament of
   quality reports, very problematic.  Also, as
   media flows.  This is further discussed in [RFC5117]
   section 3.6, the use Section 12.9.

   WebRTC implementations of RTP are REQUIRED to support multiplexing of
   a content modifying MCU with RTCP termination
   breaks multimedia session onto a single RTP loop detection and removes the ability for receivers session according to
   identify active senders.  According only the first two options are
   recommended. (tbd).
   If such RTP protocol extensions session multiplexing is to be used with conferencing are included
   because they are important in used, this MUST be
   negotiated during the context of centralised
   conferencing, where one signalling phase.  Support for multiple RTP Mixer (Conference Focus) receives
   sessions over a
   participants media streams single UDP flow as defined by
   [I-D.westerlund-avtcore-transport-multiplexing] is RECOMMENDED.

4.5.  RTP and distribute them to the other
   participants.  These messages are defined in the Extended RTCP Multiplexing

   Historically, RTP Profile
   for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/
   AVPF) [RFC4585] and the "Codec Control Messages in the RTCP have been run on separate transport-layer
   ports (e.g., two UDP ports for each RTP session, one port for RTP Audio-
   Visual Profile with Feedback (AVPF)" (CCM) [RFC5104] and are fully
   usable by
   one port for RTCP).  With the Secure variant increased use of Network Address/Port
   Translation (NAPT) this profile (RTP/SAVPF) [RFC5124].

5.1.1.  Full Intra Request

   The Full Intra Request is defined in Sections 3.5.1 and 4.3.1 of CCM
   [RFC5104]. has become problematic, since maintaining
   multiple NAT bindings can be costly.  It is used also complicates firewall
   administration, since multiple ports must be opened to have the mixer request from a allow RTP
   traffic.  To reduce these costs and session
   participants setup times, support for
   multiplexing RTP data packets and RTCP control packets on a new Intra picture.  This single
   port [RFC5761] for each RTP session is used when switching
   between sources REQUIRED.

   (tbd: Are WebRTC implementations required to ensure that support the receivers can decode case where
   the video or
   other predicted media encoding RTP and RTCP are run on separate UDP ports, for interoperability
   with long prediction chains.  It is
   RECOMMENDED legacy systems?)

   Note that this feedback message is supported.

5.1.2.  Picture Loss Indication

   The Picture Loss Indication is defined in Section 6.3.1 of the RTP/
   AVPF profile [RFC4585].  It is used by use of RTP and RTCP multiplexed onto a receiver to tell the encoder single transport
   port ensures that it lost the decoder context and would like to have it repaired
   somehow.  This there is semantically different from the Full Intra Request
   above.  It is RECOMMENDED occasional traffic sent on that this feedback message is supported as
   a loss tolerance mechanism.

5.1.3.  Slice Loss Indication

   The Slice Loss Indicator is defined in Section 6.3.2 of the RTP/AVPF
   profile [RFC4585].  It port, even
   if there is used by a receiver no active media traffic.  This may be useful to tell the encoder that
   it has detected the loss or corruption of one or more consecutive
   macroblocks, keep-
   alive NAT bindings, and would like to have these repaired somehow.  The use
   of this feedback message is OPTIONAL as a loss tolerance mechanism.

5.1.4.  Reference Picture Selection Indication

   Reference Picture Selection Indication (RPSI) is defined in Section
   6.3.3 of the RTP/AVPF profile [RFC4585].  Some video coding standards
   allow the use of older reference pictures than the most recent one recommend method for predictive coding.  If such a codec is application level
   keep-alives of RTP sessions [RFC6263].

4.6.  Reduced Size RTCP

   RTCP packets are usually sent as compound RTCP packets, and [RFC3550]
   requires that those compound packets start with an Sender Report (SR)
   or Receiver Report (RR) packet.  When using frequent RTCP feedback
   messages, these general statistics are not needed in used, every packet and if
   unnecessarily increase the
   encoder has learned about a loss of encoder-decoder synchronicity, a
   known-as-correct reference picture mean RTCP packet size.  This can limit the
   frequency at which RTCP packets can be used for future coding.
   The RPSI message allows sent within the RTCP bandwidth
   share.

   To avoid this problem, [RFC5506] specifies how to be signalled.  The use of this reduce the mean
   RTCP
   feedback message is OPTIONAL as a loss tolerance mechanism.

5.1.5.  Temporary Maximum Media Stream Bit Rate Request

   This feedback message is defined in Section 3.5.4 and 4.2.1 allow for more frequent feedback.  Frequent
   feedback, in CCM
   [RFC5104].  This message and its notification message turn, is used by a
   media receiver, essential to inform the sending party that there make real-time application quickly
   aware of changing network conditions and allow them to adapt their
   transmission and encoding behaviour.  Support for RFC5506 is a current
   limitation on the amount
   REQUIRED.

4.7.  Symmetric RTP/RTCP

   To ease traversal of bandwidth available NAT and firewall devices, implementations are
   REQUIRED to this receiver. implement Symmetric RTP [RFC4961].  This can be requires that
   the IP address and port used for various reasons, sending and can receiving RTP and RTCP
   packets are identical.  The reasons for example be used by an using symmetric RTP mixer is
   primarily to limit the media sender being forwarded avoid issues with NAT and Firewalls by ensuring that the mixer
   (without doing media transcoding) to fit
   flow is actually bi-directional and thus kept alive and registered as
   flow the bottlenecks existing
   towards intended recipient actually wants.  In addition it saves
   resources in the other session participants.  It form of ports at the end-points, but also in the
   network as NAT mappings or firewall state is RECOMMENDED that this
   feedback message not unnecessary bloated.
   Also the amount of QoS state is supported.

5.2.  Header Extensions reduced.

4.8.  Generation of the RTCP Canonical Name (CNAME)

   The RTP specification [RFC3550] RTCP Canonical Name (CNAME) provides a capability to extend persistent transport-level
   identifier for an RTP endpoint.  While the Synchronisation Source
   (SSRC) identifier for an RTP header with in-band data, but endpoint may change if a collision is
   detected, or when the format and semantics of the
   extensions are poorly specified.  The use of header extensions RTP application is
   OPTIONAL, but if are used, it restarted, it's RTCP CNAME
   is REQUIRED meant to stay unchanged, so that they RTP endpoints can be formatted uniquely
   identified and
   signalled according to the general mechanism of associated with their RTP header extensions
   defined in [RFC5285].

   As noted in [RFC5285], the requirement from the media streams.  For proper
   functionality, each RTP endpoint needs to have a unique RTCP CNAME
   value.

   The RTP specification
   that header extensions are "designed so that the header extension may
   be ignored" [RFC3550] stands.  To be specific, header extensions must
   only be used includes guidelines for data that can safely be ignored by the recipient
   without affecting interoperability, and must choosing a
   unique RTP CNAME, but these are not be used when sufficient in the presence of the extension has changed the form or nature NAT
   devices.  In addition, some may find long-term persistent identifiers
   problematic from a privacy viewpoint.  Accordingly, support for
   generating a short-term persistent RTCP CNAMEs following method (b)
   specified in Section 4.2 of the rest "Guidelines for Choosing RTP Control
   Protocol (RTCP) Canonical Names (CNAMEs)" [RFC6222] is REQUIRED,
   since this addresses both concerns.

5.  WebRTC Use of the packet in RTP: Extensions

   There are a way number of RTP extensions that is not compatible with the way are either required to
   obtain full functionality, or extremely useful to improve on the stream
   is signalled (e.g., as defined by
   baseline performance, in the payload type).  Valid examples
   might include metadata that WebRTC application context.  One set of
   these extensions is additional related to conferencing, while others are more
   generic in nature.  The following subsections describe the usual various
   RTP
   information.

   The extensions mandated or strongly recommended within WebRTC.

5.1.  Conferencing Extensions

   RTP rapid synchronisation header extension [RFC6051] is
   recommended, inherently a group communication protocol.  Groups can be
   implemented using a centralised server, multi-unicast, or using IP
   multicast.  While IP multicast was popular in early deployments, in
   today's practice, overlay-based conferencing dominates, typically
   using one or more central servers to connect endpoints in a star or
   flat tree topology.  These central servers can be implemented in a
   number of ways as discussed in Section 5.2.1 we also recommend the
   client to mixer audio level [RFC6464], Appendix A, and consider in the mixer to
   client audio level [RFC6465] as optional feature.

   If memo on RTP
   Topologies [RFC5117].

   As discussed in Section 3.5 of [RFC5117], the use of a video
   switching MCU makes the client-to-mixer use of RTCP for congestion control, or mixer-to-client audio level indications are any
   type of quality reports, very problematic.  Also, as discussed in
   section 3.6 of [RFC5117], the use in SRTP encrypted sessions, it is REQUIRED that of a content modifying MCU with
   RTCP termination breaks RTP loop detection and removes the extensions
   are encrypted according ability
   for receivers to
   [I-D.ietf-avtcore-srtp-encrypted-header-ext] since the information
   contained in these header extensions may be considered sensitive.

5.2.1.  Rapid Synchronisation

   Many identify active senders.  Accordingly, only RTP sessions require synchronisation between audio, video,
   Transport Translators (relays), RTP Mixers, and
   other content.  This synchronisation is performed by receivers, using
   information contained end-point based
   forwarding topologies are supported in RTCP SR packets, as described WebRTC.  These RECOMMENDED
   topologies are expected to be supported by all WebRTC end-points
   (these three topologies require no special support in the RTP
   specification [RFC3550].  This basic mechanism can be slow, however,
   so it is RECOMMENDED that end-point,
   if the rapid RTP synchronisation extensions
   described features mandated in [RFC6051] be implemented.  The rapid synchronisation
   extensions use the general RTP header extension mechanism [RFC5285],
   which requires signalling, but this memo are otherwise backwards compatible.

5.2.2.  Client to Mixer Audio Level implemented).

   The Client RTP protocol extensions to Mixer Audio Level [RFC6464] is be used with conferencing, described
   below, are not required for correctness; an RTP header extension
   used by a client to inform a mixer about endpoint that does
   not implement these extensions will work correctly, but offer poor
   performance.  Support for the level listed extensions will greatly improve
   the quality of audio activity experience, however, in the packet the header is attached to.  This enables context of centralised
   conferencing, where one RTP Mixer (Conference Focus) receives a central node
   participants media streams and distribute them to make mixing or selection decisions without decoding or detailed
   inspection of the payload.  Thus reducing the needed complexity other
   participants.  These messages are defined in
   some types of central the Extended RTP nodes.

   The Client to Mixer Audio Level [RFC6464] is RECOMMENDED to be
   implemented.

5.2.3.  Mixer to Client Audio Level

   The Mixer to Client Audio Level header extension [RFC6465] provides Profile
   for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/
   AVPF) [RFC4585] and the client "Codec Control Messages in the RTP Audio-
   Visual Profile with Feedback (AVPF)" (CCM) [RFC5104] and are fully
   usable by the audio level Secure variant of this profile (RTP/SAVPF) [RFC5124].

5.1.1.  Full Intra Request

   The Full Intra Request is defined in Sections 3.5.1 and 4.3.1 of the different sources mixed into a
   common mix from the RTP mixer.  Thus enabling a user interface
   Codec Control Messages [RFC5104].  This message is used to
   indicate have the relative activity level of
   mixer request a session participant, rather
   than just being included or not based on new Intra picture from a participant in the CSRC field. session.
   This is a
   pure optimisations of non critical functions and thus optional
   functionality.

   The Mixer to Client Audio Level [RFC6465] is OPTIONAL used when switching between sources to implement.

6.  WebRTC Use of RTP: Improving Transport Robustness

   There are some tools ensure that the
   receivers can make RTP flows robust against Packet
   loss and reduce decode the impact on video or other predicted media quality.  However they all add
   extra bits compared to encoding with
   long prediction chains.  It is REQUIRED that this feedback message is
   supported by RTP senders in WebRTC, since it greatly improves the
   user experience when using centralised mixers-based conferencing.

5.1.2.  Picture Loss Indication

   The Picture Loss Indication is defined in Section 6.3.1 of the RTP/
   AVPF profile [RFC4585].  It is used by a non-robust stream.  These extra bits needs receiver to be considered and tell the aggregate bit-rate needs sending
   encoder that it lost the decoder context and would like to be rate
   controlled.  Thus improving robustness might require a lower base
   encoding quality but has have it
   repaired somehow.  This is semantically different from the potential Full Intra
   Request above as there can exist multiple methods to give fulfil the
   request.  It is RECOMMENDED that quality with
   fewer errors in it.

6.1.  Retransmission

   Support for RTP retransmission as defined by "RTP Retransmission
   Payload Format" [RFC4588] this feedback message is RECOMMENDED. supported
   as a loss tolerance mechanism.

5.1.3.  Slice Loss Indication

   The retransmission scheme Slice Loss Indicator is defined in RTP allows flexible application Section 6.3.2 of
   retransmissions.  Only selected missing packets can be requested by the receiver.  It also allows for the sender to prioritise between
   missing packets based on senders knowledge about their content.
   Compared to TCP, RTP retransmission also allows one to give up on a
   packet that despite retransmission(s) still has not been received
   within a time window.

   "WebRTC Media Transport Requirements" [I-D.cbran-rtcweb-data] raises
   two issues that they think makes RTP Retransmission unsuitable for
   RTCWEB.  We here consider these issues and explain why they are in
   fact not RTP/AVPF
   profile [RFC4585].  It is used by a reason receiver to exclude RTP retransmission from tell the tool box
   available to RTCWEB media sessions.

   The additional latency added by [RFC4588] will exceed encoder that
   it has detected the latency
   threshold for interactive voice and video:  RTP Retransmission will
      require at least loss or corruption of one round trip time for a retransmission request or more consecutive
   macroblocks, and repair packet would like to arrive.  Thus the general suitability have these repaired somehow.  The use
   of this feedback message is OPTIONAL as a loss tolerance mechanism.

5.1.4.  Reference Picture Selection Indication

   Reference Picture Selection Indication (RPSI) is defined in Section
   6.3.3 of
      using retransmissions will depend on the actual network path
      latency between RTP/AVPF profile [RFC4585].  Some video coding standards
   allow the end-points.  In many use of older reference pictures than the actual usages most recent one
   for predictive coding.  If such a codec is in used, and if the
      latency between two end-points will
   encoder has learned about a loss of encoder-decoder synchronicity, a
   known-as-correct reference picture can be low enough used for RTP
      retransmission future coding.
   The RPSI message allows this to be effective.  Interactive communication with
      end-to-end delays of 400 ms still provide a fair quality.  Even
      removing half signalled.  The use of that this RTCP
   feedback message is OPTIONAL as a loss tolerance mechanism.

5.1.5.  Temporary Maximum Media Stream Bit Rate Request

   This feedback message is defined in end-point delays allows functional
      retransmission between end-points on the same continent.  In
      addition, some applications may accept temporary delay spikes to
      allow for retransmission Sections 3.5.4 and 4.2.1 of crucial codec information such an
      parameter sets, intra picture etc, rather than getting no the
   Codec Control Messages [RFC5104].  This message and its notification
   message is used by a media at
      all.

   The undesirable increase in packet transmission at receiver, to inform the point when
   congestion occurs:  Congestion loss will impact sending party that
   there is a current limitation on the rate controls
      view amount of bandwidth available bit-rate for transmission.  When using
      retransmission one will have to prioritise between performing
      retransmissions
   this receiver.  This can be for various reasons, and the quality one can achieve with ones
      adaptable codecs.  In many use cases one prefer error free or low
      rates of error with reduced base quality over high degrees of
      error at a higher base quality.

   The RTCWEB end-point implementations will need to both select when to
   enable for example
   be used by an RTP retransmissions based on API settings and measurements of mixer to limit the actual round trip time.  In addition for each NACK request that a media sender receives it will need to make a prioritisation based on
   the importance of being forwarded by
   the requested media, mixer (without doing media transcoding) to fit the probability that bottlenecks
   existing towards the
   packet will reach other session participants.  It is REQUIRED that
   this feedback message is supported.

5.2.  Header Extensions

   The RTP specification [RFC3550] provides the receiver in time for being usable, capability to include
   RTP header extensions containing in-band data, but the
   consumption of available bit-rate format and
   semantics of the impact extensions are poorly specified.  The use of header
   extensions is OPTIONAL in the media quality WebRTC context, but if they are used,
   they MUST be formatted and signalled following the general mechanism
   for new encodings.

   To conclude, RTP header extensions defined in [RFC5285], since this gives
   well-defined semantics to RTP header extensions.

   As noted in [RFC5285], the issues raised are implementation concerns requirement from the RTP specification
   that an
   implementation needs to take into consideration, they header extensions are not
   arguments against including a highly versatile "designed so that the header extension may
   be ignored" [RFC3550] stands.  To be specific, header extensions MUST
   only be used for data that can safely be ignored by the recipient
   without affecting interoperability, and efficient packet
   loss repair mechanism.

6.2.  Forward Error Correction (FEC)

   Support MUST NOT be used when the
   presence of some type the extension has changed the form or nature of FEC to combat the effects rest
   of packet loss is
   beneficial, but is heavily application dependent.  However, some FEC
   mechanisms are encumbered.

   The main benefit from FEC is the relatively low additional delay
   needed to protect against packet losses.  The transmission of any
   repair packets should preferably be done with in a time delay way that is
   just larger than any loss events normally encountered.  That way the
   repair packet isn't also lost in not compatible with the same event as the source data.

   The amount of repair packets needed varies depending on the amount
   and pattern of packet loss to be recovered, and on the mechanism used
   to derive repair data.  The later choice also effects way the stream
   is signalled (e.g., as defined by the payload type).  Valid examples
   might include metadata that is additional delay required to both encode the repair packets usual RTP
   information.

5.2.1.  Rapid Synchronisation

   Many RTP sessions require synchronisation between audio, video, and
   other content.  This synchronisation is performed by receivers, using
   information contained in RTCP SR packets, as described in the receiver to RTP
   specification [RFC3550].  This basic mechanism can be able to recover slow, however,
   so it is RECOMMENDED that the lost packet(s).

6.2.1.  Basic Redundancy rapid RTP synchronisation extensions
   described in [RFC6051] be implemented.  The method for providing basic redundancy is rapid synchronisation
   extensions use the general RTP header extension mechanism [RFC5285],
   which requires signalling, but are otherwise backwards compatible.

5.2.2.  Client to simply retransmit a
   some time earlier sent packet.  This Mixer Audio Level

   The Client to Mixer Audio Level [RFC6464] is relatively simple in theory,
   i.e. one saves any outgoing source (original) packet in an RTP header extension
   used by a buffer
   marked with client to inform a timestamp mixer about the level of actual transmission, some X ms later one
   transmit this audio activity
   in the packet again.  Where X the header is selected attached to.  This enables a central node
   to be longer than make mixing or selection decisions without decoding or detailed
   inspection of the common loss events. payload.  Thus any loss events shorter than X can be
   recovered assuming that one doesn't get an another loss event before
   all reducing the packets lost needed complexity in the first event
   some types of central RTP nodes.  It can also be used to save
   decoding resources in a WebRTC receiver in a mesh topology, which if
   it has been received. limited decoding resources, may select to decode only the most
   relevant media streams based on audio activity levels.

   The downside of basic redundancy Client-to-Mixer Audio Level [RFC6464] extension is RECOMMENDED to
   be implemented.  If it is implemented, it is REQUIRED that the overhead.  To provide each
   packet with once chance of recovery, then header
   extensions are encrypted according to
   [I-D.ietf-avtcore-srtp-encrypted-header-ext] since the transmission rate
   increases with 100% as one needs information
   contained in these header extensions can be considered sensitive.

5.2.3.  Mixer to send each packet twice.  It is
   possible Client Audio Level

   The Mixer to only redundantly send really important packets thus
   reducing Client Audio Level header extension [RFC6465] provides
   the overhead below 100% for some other trade-off is
   overhead.

   In addition client with the basic retransmission audio level of the same packet using the
   same SSRC in the same different sources mixed into a
   common mix by a RTP mixer.  This enables a user interface to indicate
   the relative activity level of each session is participant, rather than
   just being included or not possible in RTP context.
   The reason based on the CSRC field.  This is a pure
   optimisations of non critical functions, and is hence OPTIONAL to
   implement.  If it is implemented, it is REQUIRED that one would then destroy the RTCP reporting if one
   sends the same packet twice with header
   extensions are encrypted according to
   [I-D.ietf-avtcore-srtp-encrypted-header-ext] since the same sequence number.  Thus one
   needs more elaborate mechanisms.

   RTP Payload Format Support:  Some RTP payload format do support basic
      redundancy within the RTP paylaod format itself.  Examples information
   contained in these header extensions can be considered sensitive.

6.  WebRTC Use of RTP: Improving Transport Robustness

   There are
      AMR-WB [RFC4867] and G.719 [RFC5404]. some tools that can make RTP Payload for Redundant Audio Data:  This audio flows robust against Packet
   loss and text redundancy
      format defined in [RFC2198] allows for multiple levels of
      redundancy with different delay in their transmissions, as long as reduce the source plus payload parts impact on media quality.  However they all add
   extra bits compared to a non-robust stream.  These extra bits need to
   be redundantly transmitted
      together fits into one MTU.  This should work fine for most
      interactive audio considered, and text use cases as both the codec bit-rates
      and aggregate bit-rate must be rate controlled.
   Thus improving robustness might require a lower base encoding
   quality, but has the framing intervals normally allow for this requirement potential to
      hold.  This payload format also don't increase give that quality with fewer
   errors.  The mechanisms described in the following sub-sections can
   be used to improve tolerance to packet rate, as
      original data and redundant data are sent together.  This format
      does not allow perfect recovery, only recovery of information
      deemed necessary for audio, for example the sequence number of the
      original data is lost.

   RTP loss.

6.1.  Retransmission Format:  The

   Support for RTP retransmission as defined by "RTP Retransmission
   Payload format Format" [RFC4588] is RECOMMENDED.

   The retransmission scheme in RTP allows flexible application of
   retransmissions.  Only selected missing packets can be used requested by
   the receiver.  It also allows for the sender to pro-actively send redundant packets using
      either SSRC or session multiplexing.  By using different SSRCs or
      a different session for the redundant packets the RTCP receiver
      reports will be correct.  The retransmission payload format is
      used to recover the prioritise between
   missing packets original data thus enabling a perfect
      recovery.

   Duplication Grouping Semantics in the Session Description Protocol:
      This [I-D.begen-mmusic-redundancy-grouping] is proposal for new
      SDP signalling based on senders knowledge about their content.
   Compared to indicate media stream duplication using
      different TCP, RTP sessions, or different SSRCs retransmission also allows one to separate the source
      and the redundant copy of the stream.

6.2.2.  Block Based FEC

   Block based redundancy collects give up on a number of source packets into
   packet that despite retransmission(s) still has not been received
   within a
   data block time window.

   "WebRTC Media Transport Requirements" [I-D.cbran-rtcweb-data] raises
   two issues that they think makes RTP Retransmission unsuitable for processing.  The processing results
   WebRTC.  We here consider these issues and explain why they are in some number of
   repair packets that is then transmitted
   fact not a reason to exclude RTP retransmission from the other end allowing the
   receiver to attempt tool box
   available to recover some number of lost packets in the
   block. WebRTC media sessions.

   The benefit of block based approaches is additional latency added by [RFC4588] will exceed the overhead which
   can be lower than 100% latency
   threshold for interactive voice and still recover video:  RTP Retransmission will
      require at least one or more lost source
   packet from the block.  The optimal block codes allows round trip time for each
   received a retransmission request
      and repair packet to repair a single loss within the block. arrive.  Thus 3 repair packets that are received should allow for any set of 3
   packets within the block to be recovered.  In reality one commonly
   don't reach this level of performance for any block sizes and number general suitability of repair packets, and taking
      using retransmissions will depend on the computational complexity into
   account there are even more trade-offs to make among actual network path
      latency between the codes.

   One result end-points.  In many of the block based approach is actual usages the extra delay, as one
   needs to collect
      latency between two end-points will be low enough data together before being able for RTP
      retransmission to calculate
   the repair packets.  In addition sufficient amount be effective.  Interactive communication with
      end-to-end delays of 400 ms still provide a fair quality.  Even
      removing half of that in end-point delays allows functional
      retransmission between end-points on the block needs
   to be received prior to recovery.  Thus additional same continent.  In
      addition, some applications may accept temporary delay are added on
   both sending and receiving side to ensure possibility spikes to recover any
   packet within the block.
      allow for retransmission of crucial codec information such an
      parameter sets, intra picture etc, rather than getting no media at
      all.

   The redundancy overhead and the undesirable increase in packet transmission pattern of source and
   repair data can be altered from block to block, thus allowing a
   adaptive process adjusting to meet at the actual amount of point when
   congestion occurs:  Congestion loss seen on will impact the network path and reported in RTCP.

   The alternatives that exist rate controls
      view of available bit-rate for block based FEC with RTP are transmission.  When using
      retransmission one will have to prioritise between performing
      retransmissions and the
   following:

   RTP Payload Format for Generic Forward Error Correction:  This quality one can achieve with ones
      adaptable codecs.  In many use cases one prefer error free or low
      rates of error with reduced base quality over high degrees of
      error at a higher base quality.

   The WebRTC end-point implementations will need to both select when to
   enable RTP
      payload format [RFC5109] defines an XOR retransmissions based recovery packet.
      This is on API settings and measurements of
   the simplest processing wise that an block based FEC
      scheme can be.  It also results in some limited properties, as actual round trip time.  In addition for each repair packet can only repair NACK request that a single loss.  To handle
      multiple close losses
   media sender receives it will need to make a scheme prioritisation based on
   the importance of hierarchical encodings are need.
      Thus increasing the overhead significantly.

   Forward Error Correction (FEC) Framework:  This framework
      [I-D.ietf-fecframe-framework] defines how not only RTP packets but
      how arbitrary requested media, the probability that the
   packet flows can be protected.  Some solutions
      produced or under development will reach the receiver in FECFRAME WG are RTP specific.
      There exist alternatives supporting block codes such as Reed-
      Salomon and Raptor.

6.2.3.  Recommendations time for FEC

   (tbd)

7.  WebRTC Use being usable, the
   consumption of RTP: Rate Control available bit-rate and Media Adaptation

   WebRTC will be used in very varied network environment with a
   hetrogenous set of link technologies, including wired and wireless,
   interconnecting peers at different topological locations resulting in
   network paths with widely varying one way delays, bit-rate capacity,
   load levels and traffic mixes.  In addition individual end-points
   will open one or more WebRTC sessions between one or more peers.
   Each of these session may contain different mixes the impact of the media quality
   for new encodings.

   To conclude, the issues raised are implementation concerns that an
   implementation needs to take into consideration, they are not
   arguments against including a highly versatile and data
   flows.  Assymetric usage efficient packet
   loss repair mechanism.

6.2.  Forward Error Correction (FEC)

   Support of media bit-rates and number some type of media
   streams is also to be expected.  A single end-point may receive zero FEC to many simultanous media streams while itself transmitting one or
   more streams.

   The WebRTC application combat the effects of packet loss is very dependent
   beneficial, but is heavily application dependent.  However, some FEC
   mechanisms are encumbered.

   The main benefit from a quality perspective
   on FEC is the media adapation working well so relatively low additional delay
   needed to protect against packet losses.  The transmission of any
   repair packets should preferably be done with a time delay that an end-point doesn't
   transmit significantly more is
   just larger than any loss events normally encountered.  That way the path is capable
   repair packet isn't also lost in the same event as the source data.

   The amount of handling.  If
   it would, repair packets needed varies depending on the result would be high levels amount
   and pattern of packet loss or to be recovered, and on the mechanism used
   to derive repair data.  The later choice also effects the the
   additional delay
   spikes causing media degradations.

   WebRTC applications using more than a single media stream of any
   media type or data flows has an additional concern.  In this case the
   different flows should try required to avoid affecting each other negatively.
   In addition both encode the repair packets and in case there is a resource limiation,
   the available
   resources needs receiver to be shared.  How able to share them recover the lost packet(s).

6.2.1.  Basic Redundancy

   The method for providing basic redundancy is something to simply retransmit a
   some time earlier sent packet.  This is relatively simple in theory,
   i.e. one saves any outgoing source (original) packet in a buffer
   marked with a timestamp of actual transmission, some X ms later one
   transmit this packet again.  Where X is selected to be longer than
   the
   application should prioritize so common loss events.  Thus any loss events shorter than X can be
   recovered assuming that one doesn't get an another loss event before
   all the limiation packets lost in quality or
   capabilities are the ones that provide first event has been received.

   The downside of basic redundancy is the least affect on overhead.  To provide each
   packet with once chance of recovery, then the
   application.

   This hetrogenous situation results in a requirement transmission rate
   increases with 100% as one needs to have
   functionality that adapts send each packet twice.  It is
   possible to only redundantly send really important packets thus
   reducing the available capacity and that competes
   fairly with other network flows.  If it would not compete fairly
   enough WebRTC could be used as an attack method overhead below 100% for starving out some other traffic on specific links as long as the attacker trade-off is able to
   create traffic across a specific link.  This is not far-fetched for a
   web-service capable of attracting large number of end-points and use
   overhead.

   In addition the service, combined with BGP routing state a server could pick
   client pairs to drive traffic to specific paths.

   The above estalish a clear need based on several reasons why there
   need to be a well working media adaptation mechanism.  This mechanism
   also have a number basic retransmission of requirements on what services it should provide
   and what performance it needs to provide. the same packet using the
   same SSRC in the same RTP session is not possible in RTP context.
   The biggest issue reason is that there one would then destroy the RTCP reporting if one
   sends the same packet twice with the same sequence number.  Thus one
   needs more elaborate mechanisms.

   RTP Payload Format Support:  Some RTP payload format do support basic
      redundancy within the RTP paylaod format itself.  Examples are no standardised
      AMR-WB [RFC4867] and ready to use
   mechanism that can simply be included G.719 [RFC5404].

   RTP Payload for Redundant Audio Data:  This audio and text redundancy
      format defined in WebRTC Thus there will be
   need [RFC2198] allows for multiple levels of
      redundancy with different delay in their transmissions, as long as
      the IETF to produce such a specification.  Therefore the
   suggested way forward is source plus payload parts to specify requirements on any solution be redundantly transmitted
      together fits into one MTU.  This should work fine for most
      interactive audio and text use cases as both the media adaptation.  These requirements is codec bit-rates
      and the framing intervals normally allow for now proposed to be
   documented in this specification.  In addition a proposed detailed
   solution will be developed, but is expected to take longer time to
   finalize than this document.

7.1.  Congestion Control Requirements

   Requirements for congestion control of WebRTC sessions are discussed
   in [I-D.jesup-rtp-congestion-reqs].

   Implementations are REQUIRED requirement to implement the RTP circuit breakers
   described in [I-D.perkins-avtcore-rtp-circuit-breakers].

7.2.  RTCP Limiations

   Experience with
      hold.  This payload format also don't increase the congestion control algorithms of TCP [RFC5681],
   TFRC [RFC5348], and DCCP [RFC4341], [RFC4342], [RFC4828], has shown
   that feedback on packet arrivals needs to be rate, as
      original data and redundant data are sent roughly once per
   round trip time.  We note that the capabilities together.  This format
      does not allow perfect recovery, only recovery of real-time media
   traffic to adapt to changing path conditions may be less rapid than information
      deemed necessary for audio, for example the elastic applications TCP was designed for, but frequent
   feedback sequence number of the
      original data is still required lost.

   RTP Retransmission Format:  The RTP Retransmission Payload format
      [RFC4588] can be used to allow pro-actively send redundant packets using
      either SSRC or session multiplexing.  By using different SSRCs or
      a different session for the congestion control algorithm
   to track redundant packets the path dynamics.

   The total RTCP bandwidth receiver
      reports will be correct.  The retransmission payload format is limited in its transmission rate
      used to a
   fraction of recover the RTP traffic (by default 5%).  RTCP packets are larger
   than, e.g., TCP ACKs (even when non-compound RTCP packets are used).
   The media stream bit rate original data thus limits the maximum feedback rate as enabling a
   function of the mean RTCP packet size.

   Interactive communication may not be able to afford waiting perfect
      recovery.

   Duplication Grouping Semantics in the Session Description Protocol:
      This [I-D.begen-mmusic-redundancy-grouping] is proposal for
   packet losses to occur new
      SDP signalling to indicate congestion, because an increase in
   playout delay due media stream duplication using
      different RTP sessions, or different SSRCs to queuing (most prominent separate the source
      and the redundant copy of the stream.

6.2.2.  Block Based FEC

   Block based redundancy collects a number of source packets into a
   data block for processing.  The processing results in wireless networks)
   may easily lead to some number of
   repair packets being dropped due that is then transmitted to late arrival at the
   receiver.  Therefore, more sophisticated cues may need other end allowing the
   receiver to be reported
   -- attempt to be defined in a suitable congestion control framework as noted
   above -- which, recover some number of lost packets in turn, increase the report size again.  For
   example, different RTCP XR report blocks (jointly) provide
   block.  The benefit of block based approaches is the
   necessary details overhead which
   can be lower than 100% and still recover one or more lost source
   packet from the block.  The optimal block codes allows for each
   received repair packet to implement repair a variety of congestion control
   algorithms, but the (compound) report size grows quickly.

   In group communication, single loss within the share block.
   Thus 3 repair packets that are received should allow for any set of RTCP bandwidth needs 3
   packets within the block to be
   shared by all group members, reducing the capacity and thus the
   reporting frequency per node.

   Example: assuming 512 kbit/s video yields 3200 bytes/s RTCP
   bandwidth, split across two entities in a point-to-point session.  An
   endpoint could thus send a report recovered.  In reality one commonly
   don't reach this level of 100 bytes about every 70ms or performance for every other frame in a 30 fps video.

7.3.  Legacy Interop Limitations

   Congestion control interoperability with most type any block sizes and number
   of legacy devices,
   even using an translator could be difficult.  There repair packets, and taking the computational complexity into
   account there are numerous
   reasons for this:

   No RTCP Support:  There exist legacy implementations that does not even implement RTCP at all.  Thus no feedback at all is provided.

   RTP/AVP Minimal RTCP Interval of 5s:  RTP [RFC3550] under more trade-offs to make among the RTP/AVP
      profile specifies a recommended minimal fixed interval codes.

   One result of 5
      seconds.  Sending RTCP report blocks as seldom the block based approach is the extra delay, as 5 seconds makes
      it very difficult for a sender one
   needs to use these reports and react collect enough data together before being able to
      any congestion event.

   RTP/AVP Scaled Minimal Interval:  If a legacy device uses the scaled
      minimal RTCP compound interval, calculate
   the "RECOMMENDED value for repair packets.  In addition sufficient amount of the
      reduced minimum in seconds is 360 divided by block needs
   to be received prior to recovery.  Thus additional delay are added on
   both sending and receiving side to ensure possibility to recover any
   packet within the session bandwidth
      in kilobits/second" ([RFC3550], section 6.2). block.

   The minimal
      interval drops below a second, still several times redundancy overhead and the RTT in
      almost all paths in the Internet, when the session bandwidht
      becomes 360 kbps.  A session bandwidth transmission pattern of 1 Mbps still has source and
   repair data can be altered from block to block, thus allowing a
      minimal interval of 360 ms.  Thus, with the exception for rather
      high bandwidth sessions, getting frequent enough RTCP Report
      Blocks
   adaptive process adjusting to report on the order of the RTT is very difficult as long
      as the legacy device uses meet the RTP/AVP profile.

   RTP/AVPF Supporting Legacy Device:  If a legacy device supports RTP/
      AVPF, then that enables negotation actual amount of important parameters for
      frequent reporting, such as loss seen on
   the "trr-int" parameter, network path and the
      possibility reported in RTCP.

   The alternatives that exist for block based FEC with RTP are the end-point supports some useful feedback
      format
   following:

   RTP Payload Format for congestion control purpose such as TMMBR [RFC5104].

   It has been suggested on Generic Forward Error Correction:  This RTP
      payload format [RFC5109] defines an XOR based recovery packet.
      This is the RTCWEB mailing list simplest processing wise that if
   interoperating with really limited legacy devices an WebRTC end-point
   may not send more than 64 kbps of media streams, to avoid it causing
   massive congestion on most paths block based FEC
      scheme can be.  It also results in the Internet when communicating
   with a legacy node not providing sufficient feedback for effective
   congestion control.  This warrants further discussion some limited properties, as there is
   clearly
      each repair packet can only repair a number of link layers that don't even provide that amount
   of bit-rate consistently, and that assumes no competing traffic.

8.  WebRTC Use of RTP: Performance Monitoring

   RTCP does contains a basic set single loss.  To handle
      multiple close losses a scheme of hierarchical encodings are need.
      Thus increasing the overhead significantly.

   Forward Error Correction (FEC) Framework:  This framework
      [I-D.ietf-fecframe-framework] defines how not only RTP flow monitoring points like packets but
      how arbitrary packet loss and jitter. flows can be protected.  Some solutions
      produced or under development in FECFRAME WG are RTP specific.
      There exist a number alternatives supporting block codes such as Reed-
      Salomon and Raptor.

6.2.3.  Recommendations for FEC

   Open Issue: Decision of extensions that
   could be included in the set need for FEC and if to be supported.  However, included in most cases
   which RTP monitoring that is needed depends on the application,
   recommendation which
   makes it difficult FEC scheme to select which be supported needs to include when the set be
   documented.

7.  WebRTC Use of
   applications is RTP: Rate Control and Media Adaptation

   WebRTC will be used in very large.

   Exposing some metrics varied network environment with a
   hetrogenous set of link technologies, including wired and wireless,
   interconnecting peers at different topological locations resulting in the
   network paths with widely varying one way delays, bit-rate capacity,
   load levels and traffic mixes.  In addition individual end-points
   will open one or more WebRTC API should be considered allowing
   the application to gather the measurements sessions between one or more peers.
   Each of interest.  However,
   security implications for the these session may contain different mixes of media and data sets exposed will need
   to be considered in this.

9.  IANA Considerations

   This memo makes no request
   flows.  Assymetric usage of IANA.

   Note media bit-rates and number of media
   streams is also to RFC Editor: this section may be removed expected.  A single end-point may receive zero
   to many simultanous media streams while itself transmitting one or
   more streams.

   The WebRTC application is very dependent from a quality perspective
   on publication as the media adapation working well so that an
   RFC.

10.  Security Considerations

   RTP and its various extensions each have their own security
   considerations.  These should be taken into account when considering end-point doesn't
   transmit significantly more than the security properties path is capable of handling.  If
   it would, the complete suite.  We currently don't
   think this suite creates result would be high levels of packet loss or delay
   spikes causing media degradations.

   WebRTC applications using more than a single media stream of any
   media type or data flows has an additional security issues concern.  In this case the
   different flows should try to avoid affecting each other negatively.
   In addition in case there is a resource limiation, the available
   resources needs to be shared.  How to share them is something the
   application should prioritize so that the limiation in quality or
   properties.  The use of SRTP [RFC3711] will
   capabilities are the ones that provide protection or
   mitigation against all the fundamental issues by offering
   confidentiality, integrity least affect on the
   application.

   This hetrogenous situation results in a requirement to have
   functionality that adapts to the available capacity and that competes
   fairly with other network flows.  If it would not compete fairly
   enough WebRTC could be used as an attack method for starving out
   other traffic on specific links as long as the attacker is able to
   create traffic across a specific link.  This is not far-fetched for a
   web-service capable of attracting large number of end-points and use
   the service, combined with BGP routing state a server could pick
   client pairs to drive traffic to specific paths.

   The above estalish a clear need based on several reasons why there
   need to be a well working media adaptation mechanism.  This mechanism
   also have a number of requirements on what services it should provide
   and what performance it needs to provide.

   The biggest issue is that there are no standardised and ready to use
   mechanism that can simply be included in WebRTC.  Thus there will be
   need for the IETF to produce such a specification.  Therefore the
   suggested way forward is to specify requirements on any solution for
   the media adaptation.  These requirements is for now proposed to be
   documented in this specification.  In addition a proposed detailed
   solution will be developed, but is expected to take longer time to
   finalize than this document.

7.1.  Congestion Control Requirements

   Requirements for congestion control of WebRTC sessions are discussed
   in [I-D.jesup-rtp-congestion-reqs].

   Implementations are REQUIRED to implement the RTP circuit breakers
   described in [I-D.perkins-avtcore-rtp-circuit-breakers].

7.2.  Rate Control Boundary Conditions

   The session establishment signalling will establish certain boundary
   that the media bit-rate adaptation can act within.  First of all the
   set of media codecs provide practical limitations in the supported
   bit-rate span where it can provide useful quality, which
   packetization choices that exist.  Next the signalling can establish
   maximum media bit-rate boundaries using SDP b=AS or b=CT.

7.3.  RTCP Limiations

   Experience with the congestion control algorithms of TCP [RFC5681],
   TFRC [RFC5348], and DCCP [RFC4341], [RFC4342], [RFC4828], has shown
   that feedback on packet arrivals needs to be sent roughly once per
   round trip time.  We note that the capabilities of real-time media
   traffic to adapt to changing path conditions may be less rapid than
   for the elastic applications TCP was designed for, but frequent
   feedback is still required to allow the congestion control algorithm
   to track the path dynamics.

   The total RTCP bandwidth is limited in its transmission rate to a
   fraction of the RTP traffic (by default 5%).  RTCP packets are larger
   than, e.g., TCP ACKs (even when non-compound RTCP packets are used).
   The media stream bit rate thus limits the maximum feedback rate as a
   function of the mean RTCP packet size.

   Interactive communication may not be able to afford waiting for
   packet losses to occur to indicate congestion, because an increase in
   playout delay due to queuing (most prominent in wireless networks)
   may easily lead to packets being dropped due to late arrival at the
   receiver.  Therefore, more sophisticated cues may need to be reported
   -- to be defined in a suitable congestion control framework as noted
   above -- which, in turn, increase the report size again.  For
   example, different RTCP XR report blocks (jointly) provide the
   necessary details to implement a variety of congestion control
   algorithms, but the (compound) report size grows quickly.

   In group communication, the share of RTCP bandwidth needs to be
   shared by all group members, reducing the capacity and thus the
   reporting frequency per node.

   Example: assuming 512 kbit/s video yields 3200 bytes/s RTCP
   bandwidth, split across two entities in a point-to-point session.  An
   endpoint could thus send a report of 100 bytes about every 70ms or
   for every other frame in a 30 fps video.

7.4.  Legacy Interop Limitations

   Congestion control interoperability with most type of legacy devices,
   even using an translator could be difficult.  There are numerous
   reasons for this:

   No RTCP Support:  There exist legacy implementations that does not
      even implement RTCP at all.  Thus no feedback at all is provided.

   RTP/AVP Minimal RTCP Interval of 5s:  RTP [RFC3550] under the RTP/AVP
      profile specifies a recommended minimal fixed interval of 5
      seconds.  Sending RTCP report blocks as seldom as 5 seconds makes
      it very difficult for a sender to use these reports and react to
      any congestion event.

   RTP/AVP Scaled Minimal Interval:  If a legacy device uses the scaled
      minimal RTCP compound interval, the "RECOMMENDED value for the
      reduced minimum in seconds is 360 divided by the session bandwidth
      in kilobits/second" ([RFC3550], section 6.2).  The minimal
      interval drops below a second, still several times the RTT in
      almost all paths in the Internet, when the session bandwidht
      becomes 360 kbps.  A session bandwidth of 1 Mbps still has a
      minimal interval of 360 ms.  Thus, with the exception for rather
      high bandwidth sessions, getting frequent enough RTCP Report
      Blocks to report on the order of the RTT is very difficult as long
      as the legacy device uses the RTP/AVP profile.

   RTP/AVPF Supporting Legacy Device:  If a legacy device supports RTP/
      AVPF, then that enables negotation of important parameters for
      frequent reporting, such as the "trr-int" parameter, and the
      possibility that the end-point supports some useful feedback
      format for congestion control purpose such as TMMBR [RFC5104].

   It has been suggested on the WebRTC mailing list that if
   interoperating with really limited legacy devices an WebRTC end-point
   may not send more than 64 kbps of media streams, to avoid it causing
   massive congestion on most paths in the Internet when communicating
   with a legacy node not providing sufficient feedback for effective
   congestion control.  This warrants further discussion as there is
   clearly a number of link layers that don't even provide that amount
   of bit-rate consistently, and that assumes no competing traffic.

8.  WebRTC Use of RTP: Performance Monitoring

   RTCP does contains a basic set of RTP flow monitoring points like
   packet loss and jitter.  There exist a number of extensions that
   could be included in the set to be supported.  However, in most cases
   which RTP monitoring that is needed depends on the application, which
   makes it difficult to select which to include when the set of
   applications is very large.

   Exposing some metrics in the WebRTC API should be considered allowing
   the application to gather the measurements of interest.  However,
   security implications for the different data sets exposed will need
   to be considered in this.

9.  WebRTC Use of RTP: Future Extensions

   It is possible that the core set of RTP protocols and RTP extensions
   specified in this memo will prove insufficient for the future needs
   of WebRTC applications.  In this case, future updates to this memo
   MUST be made following the Guidelines for Writers of RTP Payload
   Format Specifications [RFC2736] and Guidelines for Extending the RTP
   Control Protocol [RFC5968], and SHOULD take into account any future
   guidelines for extending RTP and related protocols that have been
   developed.

   Authors of future extensions are urged to consider the wide range of
   environments in which RTP is used when recommending extensions, since
   extensions that are applicable in some scenarios can be problematic
   in others.  Where possible, the WebRTC framework should adopt RTP
   extensions that are of general utility, to enable easy gatewaying to
   other applications using RTP, rather than adopt mechanisms that are
   narrowly targetted at specific WebRTC use cases.

10.  Signalling Considerations

   RTP is built with the assumption of an external signalling channel
   that can be used to configure the RTP sessions and their features.
   The basic configuration of an RTP session consists of the following
   parameters:

   RTP Profile:  The name of the RTP profile to be used in session.  The
      RTP/AVP [RFC3551] and RTP/AVPF [RFC4585] profiles can interoperate
      on basic level, as can their secure variants RTP/SAVP [RFC3711]
      and RTP/SAVPF [RFC5124].  The secure variants of the profiles do
      not directly interoperate with the non-secure variants, due to the
      presence of additional header fields in addition to any
      cryptographic transformation of the packet content.  As WebRTC
      requires the usage of the SAVPF profile only a single profile will
      need to be signalled.  Interworking functions may transform this
      into SAVP for a legacy use case by indicating to the WebRTC end-
      point a SAVPF end-point and limiting the usage of the a=rtcp
      attribute to indicate a trr-int value of 4 seconds.

   Transport Information:  Source and destination address(s) and ports
      for RTP and RTCP MUST be signalled for each RTP session.  In
      WebRTC these end-points will be provided by ICE that signalls
      candidates and arrive at nominated candidate pairs.  If RTP and
      RTCP multiplexing [RFC5761] is to be used, such that a single port
      is used for RTP and RTCP flows, this MUST be signalled (see
      Section 4.5).  If several RTP sessions are to be multiplexed onto
      a single transport layer flow, this MUST also be signalled (see
      Section 4.4).

   RTP Payload Types, media formats, and media format
   parameters:  The mapping between media type names (and hence the RTP
      payload formats to be used) and the RTP payload type numbers must
      be signalled.  Each media type may also have a number of media
      type parameters that must also be signalled to configure the codec
      and RTP payload format (the "a=fmtp:" line from SDP).

   RTP Extensions:  The RTP extensions one intends to use need to be
      agreed upon, including any parameters for each respective
      extension.  At the very least, this will help avoiding using
      bandwidth for features that the other end-point will ignore.  But
      for certain mechanisms there is requirement for this to happen as
      interoperability failure otherwise happens.

   RTCP Bandwidth:  Support for exchanging RTCP Bandwidth values to the
      end-points will be necessary, as described in "Session Description
      Protocol (SDP) Bandwidth Modifiers for RTP Control Protocol (RTCP)
      Bandwidth" [RFC3556], or something semantically equivalent.  This
      also ensures that the end-points have a common view of the RTCP
      bandwidth, this is important as too different view of the
      bandwidths may lead to failure to interoperate.

   These parameters are often expressed in SDP messages conveyed within
   an offer/answer exchange.  RTP does not depend on SDP or on the
   offer/answer model, but does require all the necessary parameters to
   be agreed somehow, and provided to the RTP implementation.  We note
   that in the WebRTC context it will depend on the signalling model and
   API how these parameters need to be configured but they will be need
   to either set in the API or explicitly signalled between the peers.

11.  WebRTC API Considerations

   The following sections describe how the WebRTC API features map onto
   the RTP mechanisms described in this memo.

11.1.  API MediaStream to RTP Mapping

   The WebRTC API and its media function have the concept of a
   MediaStream that consists of zero or more tracks.  Where a track is
   an individual stream of media from any type of media source like a
   microphone or a camera, but also coneptual sources, like a audio mix
   or a video composition.  The tracks within a MediaStream are expected
   to be synchronized.

   A track correspondes to the media received with one particular SSRC.
   There might be additional SSRCs associated with that SSRC, like for
   RTP retransmission or Forward Error Correction.  However, one SSRC
   will identify a media stream and its timing.

   Thus a MediaStream is a collection of SSRCs carrying the different
   media included in the synchornized aggregate.  Thus also the
   synchronization state associated with the included SSRCs are part of
   concept.  One important thing to consider is that there can be
   multiple different MediaStreams containing a given Track (SSRC).
   Thus to avoid unnecessary duplication of media at transport level one
   need to do the binding of which MediaStreams a given SSRC is
   associated with at signalling level.

   A proposal for how the binding between MediaStreams and SSRC can be
   done exist in "Cross Session Stream Identification in the Session
   Description Protocol" [I-D.alvestrand-rtcweb-msid].

12.  RTP Implementation Considerations

   The following provide some guidance on the implementation of the RTP
   features described in this memo.

   This section discusses RTP functionality that is part of the RTP
   standard, required by decisions made, or to enable use cases raised
   and their motivations.  This discussion is done from an WebRTC end-
   point perspective.  It will occassional go into central nodes, but as
   the specification is for an end-point that is where the focus lies.
   For more discussion on the central nodes and details about RTP
   topologies please reveiw Appendix A.

   The section will touch on the relation with certain RTP/RTCP
   extensions, but will focus on the RTP core functionality.  The
   definition of what functionalities and the level of requirement on
   implementing it is defined in Section 2.

12.1.  RTP Sessions and PeerConnection

   An RTP session is an association among RTP nodes, which have one
   common SSRC space.  An RTP session can include any number of end-
   points and nodes sourcing, sinking, manipulating or reporting on the
   media streams being sent within the RTP session.  A PeerConnection
   being a point to point association between an end-point and another
   node.  That peer node may be both an end-point or centralized
   processing node of some type, thus the RTP session may terminate
   immediately on the far end of the PeerConnection, but it may also
   continue as further discused below in Multiparty (Section 12.3) and
   Multiple RTP End-points (Section 12.7).

   A PeerConnection can contain one or more RTP session depending on how
   it is setup and how many UDP flows it uses.  A common usage has been
   to have one RTP session per media type, e.g. one for audio and one
   for Video, each sent over different UDP flows.  However, the default
   usage in WebRTC will be to use one RTP session for all media types.
   This usage then uses only one UDP flow, as also RTP and RTCP
   multiplexing is mandated (Section 4.5).  However, for legacy
   interworking and network prioritization (Section 12.9) based on flows
   a WebRTC end-point needs to support a mode of operation where one RTP
   session per media type is used.  Currently each RTP session must use
   its own UDP flow.  Discussion are ongoing if a solution enabling
   multiple RTP sessions over a single UDP flow, see Section 4.4.

   The multi-unicast or mesh based multi-party topology (Figure 1) is
   best to raise in this section as it concers the relation between RTP
   sessions and PeerConnections.  In this topology, each participant
   sends individual unicast RTP/UDP/IP flows to each of the other
   participants using independent PeerConnections in a full mesh.  This
   topology has the benefit of not requiring central nodes.  The
   downside is that it increases the used bandwidth at each sender by
   requiring one copy of the media streams for each participant that are
   part of the same session beyond the sender itself.  Hence, this
   topology is limited to scenarios with few participants unless the
   media is very low bandwidth.

                              +---+      +---+
                              | A |<---->| B |
                              +---+      +---+
                                ^         ^
                                 \       /
                                  \     /
                                   v   v
                                   +---+
                                   | C |
                                   +---+

                          Figure 1: Multi-unicast

   The multi-unicast topology could be implemented as a single RTP
   session, spanning multiple peer-to-peer transport layer connections,
   or as several pairwise RTP sessions, one between each pair of peers.
   To maintain a coherent mapping between the relation between RTP
   sessions and PeerConnections we recommend that one implements this as
   individual RTP sessions.  The only downside is that end-point A will
   not learn of the quality of any transmission happening between B and
   C based on RTCP.  This has not been seen as a significant downside as
   no one has yet seen a clear need for why A would need to know about
   the B's and C's communication.  An advantage of using separate RTP
   sessions is that it enables using different media bit-rates to the
   differnt peers, thus not forcing B to endure the same quality
   reductions if there are limiations in the transport from A to C as C
   will.

12.2.  Multiple Sources

   A WebRTC end-point may have multiple cameras, microphones or audio
   inputs thus a single end-point can source multiple media streams
   concurrently of the same media type.  In addition the above discussed
   criteria to support multiple media types in one single RTP session
   results that also an end-point that has one audio and one video
   source still need two transmit using two SSRCs concurrently.  As
   multi-party conferences are supported, as discussed below in
   Section 12.3, a WebRTC end-point will need to be capable of
   receiving, decoding and playout multiple media streams of the same
   type concurrently.

   Open Issue:Are any mechanism needed to signal limiations in the
   number of SSRC that an end-point can handle?

12.3.  Multiparty

   There exist numerous situations and clear use cases for WebRTC
   supporting sessions supoprting multi-party.  This can be realized in
   a number of ways using a number of different implementations
   strategies.  This focus on the different set of WebRTC end-point
   requirements that arise from different sets of multi-party
   topologies.

   The multi-unicast mesh (Figure 1) based multi-party topoology
   discussed above provides a non-centralized solution but can easily
   tax the end-points outgoing paths.  It may also consume large amount
   of encoding resources if each outgoing stream is specifically
   encoded.  If an encoding is transmitted to multiple parties, either
   as in the mesh case or when using relaying central nodes (see below)
   a requirement on the end-point becomes to be able to create media
   streams suitable to multiple destinations requirements.  These
   requirements may both be dependent on transport path and the
   different end-points preferences related to playout of the media.

                    +---+      +------------+      +---+
                    | A |<---->|            |<---->| B |
                    +---+      |            |      +---+
                               |   Mixer    |
                    +---+      |            |      +---+
                    | C |<---->|            |<---->| D |
                    +---+      +------------+      +---+

                Figure 2: RTP Mixer with Only Unicast Paths

   A Mixer (Figure 2) is an RTP end-point that optimizes the
   transmission of media streams from certain perspectives, either by
   only sending some of the received media stream to any given receiver
   or by providing a combined media stream out of a set of contributing
   streams.  There exist various methods of implementation as discussed
   in Appendix A.3.  A common aspect is that these central nodes a
   number of tools to control the media encoding provided by a WebRTC
   end-point.  This includes functions like requesting breaking the
   encoding chain and have the encoder produce a so called Intra frame.
   Another is limiting the bit-rate of a given stream to better suit the
   mixer view of the multiple down-streams.  Others are controling the
   most suitable frame-rate, picture resultion, the trade-off between
   frame-rate and spatial quality.

   A mixer gets a significant responsibility to correctly perform
   congestion control, identity management, manage synchronization while
   providing a for the application suitable media optimization.

   Mixers also need to be a trusted node when it comes to security as it
   manipulates either RTP or the media itself before sending it on
   towards the end-point(s) thus must be able to decrypt and then
   encrypt it before sending it out.  There exist one type of central
   node, the relay that one doesn't need to trust with the keys to the
   media.  The relay operates only on the IP/UDP level of the transport.
   It is configured so that it would forward any RTP/RTCP packets from A
   to the other participants B-D.

                 +---+                               +---+
                 |   |         +-----------+         |   |
                 | A |<------->| DTLS-SRTP |<------->| C |
                 |   |<--   -->|   HOST    |<--   -->|   |
                 +---+   \ /   +-----------+   \ /   +---+
                          X                     X
                 +---+   / \   +-----------+   / \   +---+
                 |   |<--   -->|    RTP    |<--   -->|   |
                 | B |<------->|   RELAY   |<------->| D |
                 |   |         +-----------+         |   |
                 +---+                               +---+

             Figure 3: DTLS-SRTP host and RTP Relay Separated

   To accomplish the security properties discussed above using a relay
   one need to have a separate key handling server and also support for
   distribute the different keys such as Encrypted Key Transport
   [I-D.ietf-avt-srtp-ekt].  The relay also creates a situation where
   there is multiple end-points visible in the RTCP reporting and any
   feedback events.  Thus becoming yet another situation in addition to
   Mesh where the end-point will have to have logic for merging
   different requirements and preferences.  This is more detail
   discussed in Section 12.7.

                          +---+    +---+    +---+
                          | A |--->| B |--->| C |
                          +---+    +---+    +---+

                     Figure 4: MediaStream Forwarding

   The above Figure 4 depicts a possible scenario where an WebRTC end-
   point (A) sends a media stream to B. B decides to forward the media
   stream to C. This can either be realized in B (WebRTC end-point)
   using a simple relay functionality creating similar consideration and
   implementation requirements.  Another implmentation strategy in B
   could be to select to transcode the media from A to C, thus breaking
   most of the dependecies between A and C. In that case A is not
   required to be aware of B forwarding the media to C.

12.4.  SSRC Collision Detection

   The RTP standard [RFC3550] requires any RTP implementation to have
   support for detecting and handling SSRC collisions, i.e. when two
   different end-points uses the same SSRC value.  This requirement
   applies also to WebRTC end-points.  There exist several scenarios
   where SSRC collisions may occur.

   In a point to point session where each SSRC are associated with
   either of the two end-points and where the main media carrying SSRC
   identifier will be announced in the signalling there is less likely
   to occur due to the information about used SSRCs provided by Source-
   Specific SDP Attributes [RFC5576].  Still if both end-points starts
   uses an new SSRC identifier prior to having signalled it to the peer
   and received acknowledgement on the signalling message there can be
   collisions.  The Source-Specific SDP Attributes [RFC5576] contains no
   mechanism to resolve SSRC collisions or reject a end-points usage of
   an SSRC.

   There could also appear unsignalled SSRCs, this may be considered a
   bug.  This is more likely than it appears as certain RTP
   functionalities need extra SSRCs to provide functionality related to
   another SSRC, for example SSRC multiplexed RTP retransmission
   [RFC4588].  In those cases an end-point can create a new SSRC which
   strictly don't need to be announced over the signalling channel to
   function correctly on both RTP and PeerConnection level.

   The more likely cases for SSRC collision is that multiple end-points
   in an multiparty creates new soruces and signalls those towards the
   central server.  In cases where the SSRC/CSRC are propogated between
   the different end-points from the central node collisions can occur.

   Another scenario is when the central node manage to connect an end-
   points PeerConnection to another PeerConnectio the end-point it has.
   Thus forming a loop where the end-point will receive its own traffic.
   This must be considered a bug, but still if it occurs it is important
   that the end-point can handle the situation.

12.5.  Contributing Sources

   Contributing Sources (CSRC) is a functionality in RTP header that
   enables a RTP node combing multiple sources into one to identify the
   sources that has gone into the combination.  For WebRTC end-point the
   support of contributing sources are trivial.  The set of CSRC are
   provided for a given RTP packet.  This information can then be
   exposed towards the applications using some form of API, most likely
   a mapping back into MediaStream identities to avoid having to expose
   two namespaces and the handling of SSRC collision handling to the
   JavaScript.

   There are also at least one extension that is dependent on the CRSRC
   list being used, that is the Mixer to client audio level [RFC6465],
   that enhances the information provided by the CSRC to actual energy
   levels for audio for each contributing source.

12.6.  Media Synchronization

   When an end-point has more than one media source being sent one need
   to consider if these media source are to be synchronized.  In RTP/
   RTCP synchronziation is provided by having a set of media streams be
   indicated as comming from the same synchroniztion context and logical
   end-point by using the same CNAME identifier.

   The next provision is that all media sources internal clock, i.e.
   what drives the RTP timestamp can be correlated with a system clock
   that is provided in RTCP Sender Reports encoded in an NTP format.  By
   having the RTP timestamp to system clock being provided for all
   sources the relation of the different media stream, also across
   multiple RTP sessions can if chosen to be synchronized.  The
   requirement is for the media sender to provide the information, the
   receiver can chose to use it or not.

12.7.  Multiple RTP End-points

   A number of usages of RTP discussed here results in that an WebRTC
   end-point sending media in an RTP session out over an PeerConnection
   will receive receiver reports from multiple RTP receiving nodes.
   Note that receiving multiple receiver reports are expected due to
   that any RTP node that has multiple SSRCs are required to report on
   the media sender.  The difference here is that they are multiple
   nodes, and thus will have different path characteristics.

   The topologies relevant to WebRTC when this can occur are centralized
   relay and a end-point forwarding a media stream.  Mixers are expected
   to not forward media stream reports across itself due to the
   difference in the media stream provided to different end-points which
   the original media source lacks information about the mixers
   manipulation.

   Having multiple RTP nodes receive ones RTP flow and send reports and
   feedback about it has several impacts.  As previously discussed
   (Section 12.3) any codec control and rate control needs to be capable
   of merging the requirements and preferences to provide a single best
   according to the situation media stream.  Specifically when it comes
   to congestion control it needs to be capable of identifying the
   different end-points to form independent congestion state information
   for each different path.

   Providing source authentication in multi-party is a challange.  In
   the mixer based topologies an end-points source authentication is
   based on verifying that media comes from the mixer by cryptographic
   verification and secondly trust the mixer to correctly identify any
   source towards the end-point.  In RTP sessions where multiple end-
   points are directly visible to an end-point all end-points have
   knowledge about each others master keys, and can thus inject packets
   claimed to come from another end-point in the session.  Any node
   performing relay can perform non-cryptographic mitigation by
   preventing forwarding of packets that has SSRC fields that has
   previously come from other end-points.  For cryptographic
   verification of the source SRTP will require additional security
   mechanisms, like TESLA for SRTP [RFC4383].

12.8.  Simulcast

   This section discusses simulcast in the meaning of providing a node,
   for example a Mixer, with multiple different encoded version of the
   same media source.  In the WebRTC context that appears to be most
   easily accomplished by establishing mutliple PeerConnection all being
   feed the same set of MediaStreams.  Each PeerConnection is then
   configured to deliver a particular media quality and thus media bit-
   rate.  This will work well as long as the end-point implements
   indepdentent media encoding for each PeerConnection and not share the
   encoder.  Simulcast will fail if the end-point uses a common encoder
   instance to multiple PeerConnections.

   Thus it should be considered to explicitly signal which of the two
   implementation strategies that are desired and which will be done.
   At least making the application and possible the central node
   interested in receiving simulcast of an end-points media streams to
   be aware if it will function or not.

12.9.  Differentiated Treatment of Flows

   There exist use cases for differentiated treatment of media streams.
   Such differentiation can happen at several places in the system.
   First of all is the prioritization within the end-point for which
   media streams that should be sent, there allocation of bit-rate out
   of the current available aggregate as determined by the congestion
   control.

   Secondly, the transport can prioritize a media streams.  This is done
   according to three methods;

   Diffserv:  The end-point could mark the packet with a diffserv code
      point to indicate to the network how the WebRTC application and
      browser would like this particular packet treated.

   Flow based:  Prioritization of all packets belonging to a particular
      media flow or RTP session by keeping them in separated UDP flows.
      Thus enabling either end-point initiated or network initiated
      prioritization of the flow.

   Deep Packet Inspection:  A network classifier (DPI) inspects the
      packet and tries to determine if the packet represents a
      particular application and type that is to be prioritized.

   With the exception of diffserv both flow based and DPI have issues
   with running multiple media types and flows on a single UDP flow,
   especially when combined with data transport (SCTP/DTLS).  DPI has
   issues due to that multiple different type of flows are aggregated
   and thus becomes more difficult to apply analysis on.  The flow based
   differentiation will provide the same treatment to all packets within
   the flow.  Thus relative prioritization is not possible.  In addition
   if the resources are limited it may not be possible to provide
   differential treatment compared to best-effort for all the flows in a
   WebRTC application.

   When flow based differentiation is available the WebRTC application
   needs to know about so that it can provide the separation of the
   media streams onto different UDP flows to enable a more granular
   usage of flow based differentiation.

   Diffserv is based on that either the end-point or a classifier can
   mark the packets with an appropriate DSCP so the packets is treated
   according to that marking.  If the end-point is to mark the traffic
   there exist two requirements in the WebRTC context.  The first is
   that the WebRTC application or browser knows which DSCP to use and
   that it can use them on some set of media streams.  Secondly the
   information needs to be propagated to the operating system when
   transmitting the packet.

   Open Issue: How will the WebRTC application and/or browser know that
   differentiated treatment is desired and available and ensure that it
   gets the information required to correctly configure the WebRTC
   multimedia conference.

13.  IANA Considerations

   This memo makes no request of IANA.

   Note to RFC Editor: this section may be removed on publication as an
   RFC.

14.  Security Considerations

   RTP and its various extensions each have their own security
   considerations.  These should be taken into account when considering
   the security properties of the complete suite.  We currently don't
   think this suite creates any additional security issues or
   properties.  The use of SRTP [RFC3711] will provide protection or
   mitigation against all the fundamental issues by offering
   confidentiality, integrity and partial source authentication.  A
   mandatory to implement media security solution will be required to be
   picked.  We currently don't discuss the key-management aspect of SRTP
   in this memo, that needs to be done taking the WebRTC communication
   model into account.

   The guidelines in [I-D.ietf-avtcore-srtp-vbr-audio] apply when using
   variable bit rate (VBR) audio codecs, for example Opus or the Mixer
   audio level header extensions.

   Security considerations for the WebRTC work are discussed in
   [I-D.ietf-rtcweb-security].

15.  Acknowledgements

   The authors would like to thank Harald Alvestrand, Cary Bran, Charles
   Eckel and Cullen Jennings for valuable feedback.

16.  References

16.1.  Normative References

   [I-D.holmberg-mmusic-sdp-bundle-negotiation]
              Holmberg, C. and H. Alvestrand, "Multiplexing Negotiation
              Using Session Description Protocol (SDP) Port Numbers",
              draft-holmberg-mmusic-sdp-bundle-negotiation-00 (work in
              progress), October 2011.

   [I-D.ietf-avtcore-srtp-encrypted-header-ext]
              Lennox, J., "Encryption of Header Extensions in the Secure
              Real-Time Transport Protocol (SRTP)",
              draft-ietf-avtcore-srtp-encrypted-header-ext-01 (work in
              progress), October 2011.

   [I-D.ietf-avtcore-srtp-vbr-audio]
              Perkins, C. and J. Valin, "Guidelines for the use of
              Variable Bit Rate Audio with Secure RTP",
              draft-ietf-avtcore-srtp-vbr-audio-04 (work in progress),
              December 2011.

   [I-D.ietf-rtcweb-overview]
              Alvestrand, H., "Overview: Real Time Protocols for Brower-
              based Applications", draft-ietf-rtcweb-overview-03 (work
              in progress), March 2012.

   [I-D.ietf-rtcweb-security]
              Rescorla, E., "Security Considerations for RTC-Web",
              draft-ietf-rtcweb-security-02 (work in progress),
              March 2012.

   [I-D.jesup-rtp-congestion-reqs]
              Jesup, R. and H. Alvestrand, "Congestion Control
              Requirements For Real Time Media",
              draft-jesup-rtp-congestion-reqs-00 (work in progress),
              March 2012.

   [I-D.lennox-rtcweb-rtp-media-type-mux]
              Rosenberg, J. and J. Lennox, "Multiplexing Multiple Media
              Types In a Single Real-Time Transport Protocol (RTP)
              Session", draft-lennox-rtcweb-rtp-media-type-mux-00 (work
              in progress), October 2011.

   [I-D.perkins-avtcore-rtp-circuit-breakers]
              Perkins, C. and V. Singh, "RTP Congestion Control: Circuit
              Breakers for Unicast Sessions",
              draft-perkins-avtcore-rtp-circuit-breakers-00 (work in
              progress), March 2012.

   [I-D.westerlund-avtcore-multiplex-architecture]
              Westerlund, M., Burman, B., and C. Perkins, "RTP
              Multiplexing Architecture",
              draft-westerlund-avtcore-multiplex-architecture-01 (work
              in progress), March 2012.

   [I-D.westerlund-avtcore-transport-multiplexing]
              Westerlund, M. and C. Perkins, "Multiple RTP Sessions on a
              Single Lower-Layer Transport",
              draft-westerlund-avtcore-transport-multiplexing-02 (work
              in progress), March 2012.

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

   [RFC2736]  Handley, M. and C. Perkins, "Guidelines for Writers of RTP
              Payload Format Specifications", BCP 36, RFC 2736,
              December 1999.

   [RFC3550]  Schulzrinne, H., Casner, S., Frederick, R., and V.
              Jacobson, "RTP: A Transport Protocol for Real-Time
              Applications", STD 64, RFC 3550, July 2003.

   [RFC3551]  Schulzrinne, H. and S. Casner, "RTP Profile for Audio and
              Video Conferences with Minimal Control", STD 65, RFC 3551,
              July 2003.

   [RFC3556]  Casner, S., "Session Description Protocol (SDP) Bandwidth
              Modifiers for RTP Control Protocol (RTCP) Bandwidth",
              RFC 3556, July 2003.

   [RFC3711]  Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K.
              Norrman, "The Secure Real-time Transport Protocol (SRTP)",
              RFC 3711, March 2004.

   [RFC4585]  Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey,
              "Extended RTP Profile for Real-time Transport Control
              Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585,
              July 2006.

   [RFC4588]  Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R.
              Hakenberg, "RTP Retransmission Payload Format", RFC 4588,
              July 2006.

   [RFC4961]  Wing, D., "Symmetric RTP / RTP Control Protocol (RTCP)",
              BCP 131, RFC 4961, July 2007.

   [RFC5104]  Wenger, S., Chandra, U., Westerlund, M., and B. Burman,
              "Codec Control Messages in the RTP Audio-Visual Profile
              with Feedback (AVPF)", RFC 5104, February 2008.

   [RFC5109]  Li, A., "RTP Payload Format for Generic Forward Error
              Correction", RFC 5109, December 2007.

   [RFC5124]  Ott, J. and E. Carrara, "Extended Secure RTP Profile for
              Real-time Transport Control Protocol (RTCP)-Based Feedback
              (RTP/SAVPF)", RFC 5124, February 2008.

   [RFC5285]  Singer, D. and H. Desineni, "A General Mechanism for RTP
              Header Extensions", RFC 5285, July 2008.

   [RFC5506]  Johansson, I. and M. Westerlund, "Support for Reduced-Size
              Real-Time Transport Control Protocol (RTCP): Opportunities
              and Consequences", RFC 5506, April 2009.

   [RFC5761]  Perkins, C. and M. Westerlund, "Multiplexing RTP Data and
              Control Packets on a Single Port", RFC 5761, April 2010.

   [RFC5764]  McGrew, D. and E. Rescorla, "Datagram Transport Layer
              Security (DTLS) Extension to Establish Keys for the Secure
              Real-time Transport Protocol (SRTP)", RFC 5764, May 2010.

   [RFC6051]  Perkins, C. and T. Schierl, "Rapid Synchronisation of RTP
              Flows", RFC 6051, November 2010.

   [RFC6222]  Begen, A., Perkins, C., and D. Wing, "Guidelines for
              Choosing RTP Control Protocol (RTCP) Canonical Names
              (CNAMEs)", RFC 6222, April 2011.

   [RFC6464]  Lennox, J., Ivov, E., and E. Marocco, "A Real-time
              Transport Protocol (RTP) Header Extension for Client-to-
              Mixer Audio Level Indication", RFC 6464, December 2011.

   [RFC6465]  Ivov, E., Marocco, E., and J. Lennox, "A Real-time
              Transport Protocol (RTP) Header Extension for Mixer-to-
              Client Audio Level Indication", RFC 6465, December 2011.

16.2.  Informative References

   [I-D.alvestrand-rtcweb-msid]
              Alvestrand, H., "Cross Session Stream Identification in
              the Session Description Protocol",
              draft-alvestrand-rtcweb-msid-02 (work in progress),
              May 2012.

   [I-D.begen-mmusic-redundancy-grouping]
              Begen, A., Cai, Y., and H. Ou, "Duplication Grouping
              Semantics in the Session Description Protocol",
              draft-begen-mmusic-redundancy-grouping-03 (work in
              progress), March 2012.

   [I-D.cbran-rtcweb-data]
              Bran, C. and C. Jennings, "RTC-Web Non-Media Data
              Transport Requirements", draft-cbran-rtcweb-data-00 (work
              in progress), July 2011.

   [I-D.ietf-avt-srtp-ekt]
              Wing, D., McGrew, D., and K. Fischer, "Encrypted Key
              Transport for Secure RTP", draft-ietf-avt-srtp-ekt-03
              (work in progress), October 2011.

   [I-D.ietf-fecframe-framework]
              Watson, M., Begen, A., and V. Roca, "Forward Error
              Correction (FEC) Framework",
              draft-ietf-fecframe-framework-15 (work in progress),
              June 2011.

   [RFC2198]  Perkins, C., Kouvelas, I., Hodson, O., Hardman, V.,
              Handley, M., Bolot, J., Vega-Garcia, A., and S. Fosse-
              Parisis, "RTP Payload for Redundant Audio Data", RFC 2198,
              September 1997.

   [RFC4341]  Floyd, S. and E. Kohler, "Profile for Datagram Congestion
              Control Protocol (DCCP) Congestion Control ID 2: TCP-like
              Congestion Control", RFC 4341, March 2006.

   [RFC4342]  Floyd, S., Kohler, E., and J. Padhye, "Profile for
              Datagram Congestion Control Protocol (DCCP) Congestion
              Control ID 3: TCP-Friendly Rate Control (TFRC)", RFC 4342,
              March 2006.

   [RFC4383]  Baugher, M. and E. Carrara, "The Use of Timed Efficient
              Stream Loss-Tolerant Authentication (TESLA) in the Secure
              Real-time Transport Protocol (SRTP)", RFC 4383,
              February 2006.

   [RFC4828]  Floyd, S. and E. Kohler, "TCP Friendly Rate Control
              (TFRC): The Small-Packet (SP) Variant", RFC 4828,
              April 2007.

   [RFC4867]  Sjoberg, J., Westerlund, M., Lakaniemi, A., and Q. Xie,
              "RTP Payload Format and File Storage Format for the
              Adaptive Multi-Rate (AMR) and Adaptive Multi-Rate Wideband
              (AMR-WB) Audio Codecs", RFC 4867, April 2007.

   [RFC5117]  Westerlund, M. and S. Wenger, "RTP Topologies", RFC 5117,
              January 2008.

   [RFC5348]  Floyd, S., Handley, M., Padhye, J., and J. Widmer, "TCP
              Friendly Rate Control (TFRC): Protocol Specification",
              RFC 5348, September 2008.

   [RFC5404]  Westerlund, M. and I. Johansson, "RTP Payload Format for
              G.719", RFC 5404, January 2009.

   [RFC5576]  Lennox, J., Ott, J., and T. Schierl, "Source-Specific
              Media Attributes in the Session Description Protocol
              (SDP)", RFC 5576, June 2009.

   [RFC5681]  Allman, M., Paxson, V., and E. Blanton, "TCP Congestion
              Control", RFC 5681, September 2009.

   [RFC5968]  Ott, J. and C. Perkins, "Guidelines for Extending the RTP
              Control Protocol (RTCP)", RFC 5968, September 2010.

   [RFC6263]  Marjou, X. and A. Sollaud, "Application Mechanism for
              Keeping Alive the NAT Mappings Associated with RTP / RTP
              Control Protocol (RTCP) Flows", RFC 6263, June 2011.

Appendix A.  Supported RTP Topologies

   RTP supports both unicast and group communication, with participants
   being connected using wide range of transport-layer topologies.  Some
   of these topologies involve only the end-points, while others use RTP
   translators and mixers to provide in-network processing.  Properties
   of some RTP topologies are discussed in [RFC5117], and we further
   describe those expected to be useful for WebRTC in the following.  We
   also goes into important RTP session aspects that the topology or
   implementation variant can place on a WebRTC end-point.

A.1.  Point to Point

   The point-to-point RTP topology (Figure 5) is the simplest scenario
   for WebRTC applications.  This is going to be very common for user to
   user calls.

                            +---+         +---+
                            | A |<------->| B |
                            +---+         +---+

                         Figure 5: Point to Point

   This being the basic one lets use the topology to high-light a couple
   of details that are common for all RTP usage in the WebRTC context.
   First is the intention to multiplex RTP and RTCP over the same UDP-
   flow.  Secondly is the question of using only a single RTP session or
   one per media type for legacy interoperability.  Thirdly is the
   question of using multiple sender sources (SSRCs) per end-point.

   Historically, RTP and RTCP have been run on separate UDP ports.  With
   the increased use of Network Address/Port Translation (NAPT) this has
   become problematic, since maintaining multiple NAT bindings can be
   costly.  It also complicates firewall administration, since multiple
   ports must be opened to allow RTP traffic.  To reduce these costs and
   session setup times, support for multiplexing RTP data packets and
   RTCP control packets on a single port [RFC5761] will be supported.

   In cases where there is only one type of media (e.g., a voice-only
   call) this topology will be implemented as a single RTP session, with
   bidirectional flows of RTP and RTCP packets, all then multiplexed
   onto a single 5-tuple.  If multiple types of media are to be used
   (e.g., audio and video), then each type media can be sent as a
   separate RTP session using a different 5-tuple, allowing for separate
   transport level treatment of each type of media.  Alternatively, all
   types of media can be multiplexed onto a single 5-tuple as a single
   RTP session, or as several RTP sessions if using a demultiplexing
   shim.  Multiplexing different types of media onto a single 5-tuple
   places some limitations on how RTP is used, as described in "RTP
   Multiplexing Architecture"
   [I-D.westerlund-avtcore-multiplex-architecture].  It is not expected
   that these limitations will significantly affect the scenarios
   targetted by WebRTC, but they may impact interoperability with legacy
   systems.

   An RTP session have good support for simultanously transport multiple
   media sources.  Each media source uses an unique SSRC identifier and
   each SSRC has independent RTP sequence number and timestamp spaces.
   This is being utilized in WebRTC for several cases.  One is to enable
   multiple media sources of the same type, an end-point that has two
   video cameras can potentially transmitt video from both to its
   peer(s).  Another usage is when a single RTP session is being used
   for both multiple media types, thus an end-point can transmit both
   audio and video to the peer(s).  Thirdly to support multi-party cases
   as will be discussed below support for multiple SSRC of the same
   media type are required.

   Thus we can introduce a couple of different notiations in the below
   two alternate figures of a single peer connection in a a point to
   point setup.  The first depicting a setup where the peer connection
   established has two different RTP sessions, one for audio and one for
   video.  The second one using a single RTP session.  In both cases A
   has two video streams to send and one audio stream.  B has only one
   audio and video stream.  These are used to illustrate the relation
   between a peerConnection, the UDP flow(s), the RTP session(s) and the
   SSRCs that will be used in the later cases also.  In the below
   figures RTCP flows are not included.  They will flow bi-directionally
   between any RTP session instances in the different nodes.

            +-A-------------+                 +-B-------------+
            | +-PeerC1------|                 |-PeerC1------+ |
            | | +-UDP1------|                 |-UDP1------+ | |
            | | | +-RTP1----|                 |-RTP1----+ | | |
            | | | | +-Audio-|                 |-Audio-+ | | | |
            | | | | |    AA1|---------------->|       | | | | |
            | | | | |       |<----------------|BA1    | | | | |
            | | | | +-------|                 |-------+ | | | |
            | | | +---------|                 |---------+ | | |
            | | +-----------|                 |-----------+ | |
            | |             |                 |             | |
            | | +-UDP2------|                 |-UDP2------+ | |
            | | | +-RTP2----|                 |-RTP1----+ | | |
            | | | | +-Video-|                 |-Video-+ | | | |
            | | | | |    AV1|---------------->|       | | | | |
            | | | | |    AV2|---------------->|       | | | | |
            | | | | |       |<----------------|BV1    | | | | |
            | | | | +-------|                 |-------+ | | | |
            | | | +---------|                 |---------+ | | |
            | | +-----------|                 |-----------+ | |
            | +-------------|                 |-------------+ |
            +---------------+                 +---------------+

              Figure 6: Point to Point: Multiple RTP sessions

   As can be seen above in the Point to Point: Multiple RTP sessions
   (Figure 6) the single Peer Connection contains two RTP sessions over
   different UDP flows UDP 1 and UDP 2, i.e. their 5-tuples will be
   different, normally on source and destination ports.  The first RTP
   session (RTP1) carries audio, one stream in each direction AA1 and
   BA1.  The second RTP session contains two video streams from A (AV1
   and AV2) and one from B to A (BV1).

            +-A-------------+                 +-B-------------+
            | +-PeerC1------|                 |-PeerC1------+ |
            | | +-UDP1------|                 |-UDP1------+ | |
            | | | +-RTP1----|                 |-RTP1----+ | | |
            | | | | +-Audio-|                 |-Audio-+ | | | |
            | | | | |    AA1|---------------->|       | | | | |
            | | | | |       |<----------------|BA1    | | | | |
            | | | | +-------|                 |-------+ | | | |
            | | | |         |                 |         | | | |
            | | | | +-Video-|                 |-Video-+ | | | |
            | | | | |    AV1|---------------->|       | | | | |
            | | | | |    AV2|---------------->|       | | | | |
            | | | | |       |<----------------|BV1    | | | | |
            | | | | +-------|                 |-------+ | | | |
            | | | +---------|                 |---------+ | | |
            | | +-----------|                 |-----------+ | |
            | +-------------|                 |-------------+ |
            +---------------+                 +---------------+

               Figure 7: Point to Point: Single RTP session.

   In (Figure 7) there is only a single UDP flow and RTP session (RTP1).
   This RTP session carries a total of five (5) media streams (SSRCs).
   From A to B there is Audio (AA1) and two video (AV1 and AV2).  From B
   to A there is Audio (BA1) and Video (BV1).

A.2.  Multi-Unicast (Mesh)

   For small multiparty calls, it is practical to set up a multi-unicast
   topology (Figure 8); unfortunately not discussed in the RTP
   Topologies RFC [RFC5117].  In this topology, each participant sends
   individual unicast RTP/UDP/IP flows to each of the other participants
   using independent PeerConnections in a full mesh.

                              +---+      +---+
                              | A |<---->| B |
                              +---+      +---+
                                ^         ^
                                 \       /
                                  \     /
                                   v   v
                                   +---+
                                   | C |
                                   +---+

                          Figure 8: Multi-unicast

   This topology has the benefit of not requiring central nodes.  The
   downside is that it increases the used bandwidth at each sender by
   requiring one copy of the media streams for each participant that are
   part of the same session beyond the sender itself.  Hence, this
   topology is limited to scenarios with few participants unless the
   media is very low bandwidth.  The multi-unicast topology could be
   implemented as a single RTP session, spanning multiple peer-to-peer
   transport layer connections, or as several pairwise RTP sessions, one
   between each pair of peers.  To maintain a coherent mapping between
   the relation between RTP sessions and PeerConnections we recommend
   that one implements this as individual RTP sessions.  The only
   downside is that end-point A will not learn of the quality of any
   transmission happening between B and C based on RTCP.  This has not
   been seen as a significant downside as now one has yet seen a need
   for why A would need to know about the B's and C's communication.  An
   advantage of using separate RTP sessions is that it enables using
   different media bit-rates to the differnt peers, thus not forcing B
   to endure the same quality reductions if there are limiations in the
   transport from A to C as C will.

        +-A------------------------+              +-B-------------+
        |+---+       +-PeerC1------|              |-PeerC1------+ |
        ||MIC|       | +-UDP1------|              |-UDP1------+ | |
        |+---+       | | +-RTP1----|              |-RTP1----+ | | |
        | |  +----+  | | | +-Audio-|              |-Audio-+ | | | |
        | +->|ENC1|--+-+-+-+--->AA1|------------->|       | | | | |
        | |  +----+  | | | |       |<-------------|BA1    | | | | |
        | |          | | | +-------|              |-------+ | | | |
        | |          | | +---------|              |---------+ | | |
        | |          | +-----------|              |-----------+ | |
        | |          +-------------|              |-------------+ |
        | |                        |              |---------------+
        | |                        |
        | |                        |              +-C-------------+
        | |          +-PeerC2------|              |-PeerC2------+ |
        | |          | +-UDP2------|              |-UDP2------+ | |
        | |          | | +-RTP2----|              |-RTP2----+ | | |
        | |  +----+  | | | +-Audio-|              |-Audio-+ | | | |
        | +->|ENC2|--+-+-+-+--->AA2|------------->|       | | | | |
        |    +----+  | | | |       |<-------------|CA1    | | | | |
        |            | | | +-------|              |-------+ | | | |
        |            | | +---------|              |---------+ | | |
        |            | +-----------|              |-----------+ | |
        |            +-------------|              |-------------+ |
        +--------------------------+              +---------------+

            Figure 9: Session strcuture for Multi-Unicast Setup

   Lets review how the RTP sessions looks from A's perspective by
   considering both how the media is a handled and what PeerConnections
   and RTP sessions that are setup in Figure 9.  A's microphone is
   captured and the digital audio can then be feed into two different
   encoder instances each beeing associated with two different
   PeerConnections (PeerC1 and PeerC2) each containing independent RTP
   sessions (RTP1 and RTP2).  The SSRCs in each RTP session will be
   completely independent and the media bit-rate produced by the encoder
   can also be tuned to address any congestion control requirements
   between A and B differently then for the path A to C.

   For media encodings which are more resource consuming, like video,
   one could expect that it will be common that end-points that are
   resource costrained will use a different implementation strategy
   where the encoder is shared between the different PeerConnections as
   shown below Figure 10.
        +-A----------------------+                 +-B-------------+
        |+---+                   |                 |               |
        ||CAM|     +-PeerC1------|                 |-PeerC1------+ |
        |+---+     | +-UDP1------|                 |-UDP1------+ | |
        |  |       | | +-RTP1----|                 |-RTP1----+ | | |
        |  V       | | | +-Video-|                 |-Video-+ | | | |
        |+----+    | | | |       |<----------------|BV1    | | | | |
        ||ENC |----+-+-+-+--->AV1|---------------->|       | | | | |
        |+----+    | | | +-------|                 |-------+ | | | |
        |  |       | | +---------|                 |---------+ | | |
        |  |       | +-----------|                 |-----------+ | |
        |  |       +-------------|                 |-------------+ |
        |  |                     |                 |---------------+
        |  |                     |
        |  |                     |                 +-C-------------+
        |  |       +-PeerC2------|                 |-PeerC2------+ |
        |  |       | +-UDP2------|                 |-UDP2------+ | |
        |  |       | | +-RTP2----|                 |-RTP2----+ | | |
        |  |       | | | +-Video-|                 |-Video-+ | | | |
        |  +-------+-+-+-+--->AV2|---------------->|       | | | | |
        |          | | | |       |<----------------|CV1    | | | | |
        |          | | | +-------|                 |-------+ | | | |
        |          | | +---------|                 |---------+ | | |
        |          | +-----------|                 |-----------+ | |
        |          +-------------|                 |-------------+ |
        +------------------------+                 +---------------+

               Figure 10: Single Encoder Multi-Unicast Setup

   This will clearly save resources consumed by encoding but does
   introduce the need for the end-point A to make decisions on how it
   encodes the media so it suites delivery to both B and C. This is not
   limited to congestion control, also prefered resolution to receive
   based on dispaly area available is another aspect requiring
   consideration.  The need for this type of descion logic does arise in
   several different topologies and implementation.

A.3.  Mixer Based

   An mixer (Figure 11) is a centralised point that selects or mixes
   content in a conference to optimise the RTP session so that each end-
   point only needs connect to one entity, the mixer.  The mixer can
   also reduce the bit-rate needed from the mixer down to a conference
   participants as the media sent from the mixer to the end-point can be
   optimised in different ways.  These optimisations include methods
   like only choosing media from the currently most active speaker or
   mixing together audio so that only one audio stream is required in
   stead of 3 in the depicted scenario (Figure 11).

                    +---+      +------------+      +---+
                    | A |<---->|            |<---->| B |
                    +---+      |            |      +---+
                               |   Mixer    |
                    +---+      |            |      +---+
                    | C |<---->|            |<---->| D |
                    +---+      +------------+      +---+

               Figure 11: RTP Mixer with Only Unicast Paths

   Mixers has two downsides, the first is that the mixer must be a
   trusted node as they either performs media operations or at least
   repacketize the media.  Both type of operations requires when using
   SRTP that the mixer verifies integrity, decrypts the content, perform
   its operation and form new RTP packets, encrypts and integegrity
   protect them.  This applies to all types of mixers described below.

   The second downside is that all these operations and optimization of
   the session requires processing.  How much depends on the
   implementation as will become evident below.

   The implementation of an mixer can take several different forms and
   we will discuss the main themes available that doesn't break RTP.

   Please note that a Mixer could also contain translator
   functionalities, like a media transcoder to adjust the media bit-rate
   or codec used on a particular media stream.

A.3.1.  Media Mixing

   This type of mixer is one which clearly can be called RTP mixer is
   likely the one that most thinks of when they hear the term mixer.

   Its basic patter of operation is that it will receive the different
   participants media stream.  Select which that are to be included in a
   media domain mix of the incomming media streams.  Then create a
   single outgoing stream from this mix.

   Audio mixing is straight forward and commonly possible to do for a
   number of participants.  Lets assume that you want to mix N number of
   streams from different participants.  Then the mixer need to perform
   N decodings.  Then it needs to produce N or N+1 mixes, the reasons
   that different mixes are needed are so that each contributing source
   get a mix which don't contain themselves, as this would result in an
   echo.  When N is lower than the number of all participants one may
   produce a Mix of all N streams for the group that are curently not
   included in the mix, thus N+1 mixes.  These audio streams are then
   encoded again, RTP packetized and sent out.

   Video can't really be "mixed" and produce something particular useful
   for the users, however creating an composition out of the contributed
   video streams can be done.  In fact it can be done in a number of
   ways, tiling the different streams creating a chessboard, selecting
   someone as more important and showing them large and a number of
   other sources as smaller is another.  Also here one commonly need to
   produce a number of different compositions so that the contributing
   part doesn't need to see themselves.  Then the mixer re-encodes the
   created video stream, RTP packetize it and send it out

   The problem with media mixing is that it both consume large amount of
   media processing and encoding resources.  The second is the quality
   degradation created by decoding and re-encoding the media stream.
   Its advantage is that it is quite simplistic for the clients to
   handle as they don't need to handle local mixing and composition.

      +-A-------------+             +-MIXER--------------------------+
      | +-PeerC1------|             |-PeerC1--------+                |
      | | +-UDP1------|             |-UDP1--------+ |                |
      | | | +-RTP1----|             |-RTP1------+ | |        +-----+ |
      | | | | +-Audio-|             |-Audio---+ | | | +---+  |     | |
      | | | | |    AA1|------------>|---------+-+-+-+-|DEC|->|     | |
      | | | | |       |<------------|MA1 <----+ | | | +---+  |     | |
      | | | | |       |             |(BA1+CA1)|\| | | +---+  |     | |
      | | | | +-------|             |---------+ +-+-+-|ENC|<-| B+C | |
      | | | +---------|             |-----------+ | | +---+  |     | |
      | | +-----------|             |-------------+ |        |  M  | |
      | +-------------|             |---------------+        |  E  | |
      +---------------+             |                        |  D  | |
                                    |                        |  I  | |
      +-B-------------+             |                        |  A  | |
      | +-PeerC2------|             |-PeerC2--------+        |     | |
      | | +-UDP2------|             |-UDP2--------+ |        |  M  | |
      | | | +-RTP2----|             |-RTP2------+ | |        |  I  | |
      | | | | +-Audio-|             |-Audio---+ | | | +---+  |  X  | |
      | | | | |    BA1|------------>|---------+-+-+-+-|DEC|->|  E  | |
      | | | | |       |<------------|MA2 <----+ | | | +---+  |  R  | |
      | | | | +-------|             |(BA1+CA1)|\| | | +---+  |     | |
      | | | +---------|             |---------+ +-+-+-|ENC|<-| A+C | |
      | | +-----------|             |-----------+ | | +---+  |     | |
      | +-------------|             |-------------+ |        |     | |
      +---------------+             |---------------+        |     | |
                                    |                        |     | |
      +-C-------------+             |                        |     | |
      | +-PeerC3------|             |-PeerC3--------+        |     | |
      | | +-UDP3------|             |-UDP3--------+ |        |     | |
      | | | +-RTP3----|             |-RTP3------+ | |        |     | |
      | | | | +-Audio-|             |-Audio---+ | | | +---+  |     | |
      | | | | |    CA1|------------>|---------+-+-+-+-|DEC|->|     | |
      | | | | |       |<------------|MA3 <----+ | | | +---+  |     | |
      | | | | +-------|             |(BA1+CA1)|\| | | +---+  |     | |
      | | | +---------|             |---------+ +-+-+-|ENC|<-| A+B | |
      | | +-----------|             |-----------+ | | +---+  |     | |
      | +-------------|             |-------------+ |        +-----+ |
      +---------------+             |---------------+                |
                                    +--------------------------------+

            Figure 12: Session and SSRC details for Media Mixer

   From an RTP perspective media mixing can be very straight forward as
   can be seen in Figure 12.  The mixer present one SSRC towards the
   peer client, e.g.  MA1 to Peer A, which is the media mix of the other
   particpants.  As each peer receives a different version produced by
   the mixer there are no actual relation between the different RTP
   sessions in the actual media or the transport level information.
   There is however one connection between RTP1-RTP3 in this figure.  It
   has to do with the SSRC space and the identity information.  When A
   receives the MA1 stream which is a combination of BA1 and CA1 streams
   in the other PeerConnections RTP could enable the mixer to include
   CSRC information in the MA1 stream to identify the contributing
   source BA1 and CA1.

   The CSRC has in its turn utility in RTP extensions, like the in
   Section 5.2.3 discussed Mixer to Client audio levels RTP header
   extension [RFC6465].  If the SSRC from one PeerConnection are used as
   CSRC in another PeerConnection then RTP1, RTP2 and RTP3 becomes one
   joint session as they have a common SSRC space.  At this stage one
   also need to consider which RTCP information one need to expose in
   the different legs.  For the above situation commonly nothing more
   than the Source Description (SDES) information and RTCP BYE for CSRC
   need to be exposed.  The main goal would be to enable the correct
   binding against the application logic and other information sources.
   This also enables loop detection in the RTP session.

A.3.1.1.  RTP Session Termination

   There exist an possible implementation choice to have the RTP
   sessions being separated between the different legs in the multi-
   party communication session and only generate media streams in each
   without carrying on RTP/RTCP level any identity information about the
   contributing sources.  This removes both the functionaltiy that CSRC
   can provide and the possibility to use any extensions that build on
   CSRC and the loop detection.  It may appear a simplification if SSRC
   collision would occur between two different end-points as they can be
   avoide to be resolved and instead remapped between the independent
   sessions if at all exposed.  However, SSRC/CSRC remapping
   requiresthat SSRC/CSRC are never exposed to the WebRTC javascript
   client to use as reference.  This as they only have local importance
   if they are used on a multi-party session scope the result would be
   missreferencing.  Also SSRC collision handling will still be needed
   as it may occur between the mixer and the end-point.

   Session termination may appear to resolve some issues, it however
   creates other issues that needs resolving, like loop detection,
   identification of contributing sources and the need to handle mapped
   identities and ensure that the right one is used towards the right
   identities and never used directly between multiple end-points.

A.3.2.  Media Switching

   An RTP Mixer based on media switching avoids the media decoding and
   encoding cycle in the mixer, but not the decryption and re-encryption
   cycle as one rewrites RTP headers.  This both reduces the amount of
   computational resources needed in the mixer and increases the media
   quality per transmitted bit.  This is achieve by letting the mixer
   have a number of SSRCs that represents conceptual or functional
   streams the mixer produces.  These streams are created by selecting
   media from one of the by the mixer received media streams and forward
   the media using the mixers own SSRCs.  The mixer can then switch
   between available sources if that is required by the concept for the
   source, like currently active speaker.

   To achieve a coherent RTP media stream from the mixer's SSRC the
   mixer is forced to rewrite the incoming RTP packet's header.  First
   the SSRC field must be set to the value of the Mixer's SSRC.
   Secondly, the sequence number must be the next in the sequence of
   outgoing packets it sent.  Thirdly the RTP timestamp value needs to
   be adjusted using an offset that changes each time one switch media
   source.  Finally depending on the negotiation the RTP payload type
   value representing this particular RTP payload configuration may have
   to be changed if the different PeerConnections have not arrived on
   the same numbering for a given configuration.  This also requires
   that the different end-points do support a common set of codecs,
   otherwise media transcoding for codec compatibility is still
   required.

   Lets consider the operation of media switching mixer that supports a
   video conference with six participants (A-F) where the two latest
   speakers in the conference are shown to each participants.  Thus the
   mixer has two SSRCs sending video to each peer.

      +-A-------------+             +-MIXER--------------------------+
      | +-PeerC1------|             |-PeerC1--------+                |
      | | +-UDP1------|             |-UDP1--------+ |                |
      | | | +-RTP1----|             |-RTP1------+ | |        +-----+ |
      | | | | +-Video-|             |-Video---+ | | |        |     | |
      | | | | |    AV1|------------>|---------+-+-+-+------->|     | |
      | | | | |       |<------------|MV1 <----+-+-+-+-BV1----|     | |
      | | | | |       |<------------|MV2 <----+-+-+-+-EV1----|     | |
      | | | | +-------|             |---------+ | | |        |     | |
      | | | +---------|             |-----------+ | |        |     | |
      | | +-----------|             |-------------+ |        |  S  | |
      | +-------------|             |---------------+        |  W  | |
      +---------------+             |                        |  I  | |
                                    |                        |  T  | |
      +-B-------------+             |                        |  C  | |
      | +-PeerC2------|             |-PeerC2--------+        |  H  | |
      | | +-UDP2------|             |-UDP2--------+ |        |     | |
      | | | +-RTP2----|             |-RTP2------+ | |        |  M  | |
      | | | | +-Video-|             |-Video---+ | | |        |  A  | |
      | | | | |    BV1|------------>|---------+-+-+-+------->|  T  | |
      | | | | |       |<------------|MV3 <----+-+-+-+-AV1----|  R  | |
      | | | | |       |<------------|MV4 <----+-+-+-+-EV1----|  I  | |
      | | | | +-------|             |---------+ | | |        |  X  | |
      | | | +---------|             |-----------+ | |        |     | |
      | | +-----------|             |-------------+ |        |     | |
      | +-------------|             |---------------+        |     | |
      +---------------+             |                        |     | |
                                    :                        :     : :
                                    :                        :     : :
      +-F-------------+             |                        |     | |
      | +-PeerC6------|             |-PeerC6--------+        |     | |
      | | +-UDP6------|             |-UDP6--------+ |        |     | |
      | | | +-RTP6----|             |-RTP6------+ | |        |     | |
      | | | | +-Video-|             |-Video---+ | | |        |     | |
      | | | | |    CV1|------------>|---------+-+-+-+------->|     | |
      | | | | |       |<------------|MV11 <---+-+-+-+-AV1----|     | |
      | | | | |       |<------------|MV12 <---+-+-+-+-EV1----|     | |
      | | | | +-------|             |---------+ | | |        |     | |
      | | | +---------|             |-----------+ | |        |     | |
      | | +-----------|             |-------------+ |        +-----+ |
      | +-------------|             |---------------+                |
      +---------------+             +--------------------------------+

                   Figure 13: Media Switching RTP Mixer

   The Media Switching RTP mixer can similar to the Media Mixing one
   reduce the bit-rate needed towards the different peers by selecting
   and switching in a sub-set of media streams out of the ones it
   receives from the conference participations.

   To ensure that a media receiver can correctly decode the media stream
   after a switch, it becomes necessary to ensure for state saving
   codecs that they start from default state at the point of switching.
   Thus one common tool for video is to request that the encoding
   creates an intra picture, something that isn't dependent on earlier
   state.  This can be done using Full Intra Request RTCP codec control
   message as discussed in Section 5.1.1.

   Also in this type of mixer one could consider to terminate the RTP
   sessions fully between the different PeerConnection.  The same
   arguments and conisderations as discussed in Appendix A.3.1.1 applies
   here.

A.3.3.  Media Projecting

   Another method for handling media in the RTP mixer is to project all
   potential sources (SSRCs) into a per end-point independent RTP
   session.  The mixer can then select which of the potential sources
   that are currently actively transmitting media, despite that the
   mixer in another RTP session recieves media from that end-point.
   This is similar to the media switching Mixer but have some important
   differences in RTP details.

      +-A-------------+             +-MIXER--------------------------+
      | +-PeerC1------|             |-PeerC1--------+                |
      | | +-UDP1------|             |-UDP1--------+ |                |
      | | | +-RTP1----|             |-RTP1------+ | |        +-----+ |
      | | | | +-Video-|             |-Video---+ | | |        |     | |
      | | | | |    AV1|------------>|---------+-+-+-+------->|     | |
      | | | | |       |<------------|BV1 <----+-+-+-+--------|     | |
      | | | | |       |<------------|CV1 <----+-+-+-+--------|     | |
      | | | | |       |<------------|DV1 <----+-+-+-+--------|     | |
      | | | | |       |<------------|EV1 <----+-+-+-+--------|     | |
      | | | | |       |<------------|FV1 <----+-+-+-+--------|     | |
      | | | | +-------|             |---------+ | | |        |     | |
      | | | +---------|             |-----------+ | |        |     | |
      | | +-----------|             |-------------+ |        |  S  | |
      | +-------------|             |---------------+        |  W  | |
      +---------------+             |                        |  I  | |
                                    |                        |  T  | |
      +-B-------------+             |                        |  C  | |
      | +-PeerC2------|             |-PeerC2--------+        |  H  | |
      | | +-UDP2------|             |-UDP2--------+ |        |     | |
      | | | +-RTP2----|             |-RTP2------+ | |        |  M  | |
      | | | | +-Video-|             |-Video---+ | | |        |  A  | |
      | | | | |    BV1|------------>|---------+-+-+-+------->|  T  | |
      | | | | |       |<------------|AV1 <----+-+-+-+--------|  R  | |
      | | | | |       |<------------|CV1 <----+-+-+-+--------|  I  | |
      | | | | |       | :    :    : |: :  : : : : : :  :  : :|  X  | |
      | | | | |       |<------------|FV1 <----+-+-+-+--------|     | |
      | | | | +-------|             |---------+ | | |        |     | |
      | | | +---------|             |-----------+ | |        |     | |
      | | +-----------|             |-------------+ |        |     | |
      | +-------------|             |---------------+        |     | |
      +---------------+             |                        |     | |
                                    :                        :     : :
                                    :                        :     : :
      +-F-------------+             |                        |     | |
      | +-PeerC6------|             |-PeerC6--------+        |     | |
      | | +-UDP6------|             |-UDP6--------+ |        |     | |
      | | | +-RTP6----|             |-RTP6------+ | |        |     | |
      | | | | +-Video-|             |-Video---+ | | |        |     | |
      | | | | |    CV1|------------>|---------+-+-+-+------->|     | |
      | | | | |       |<------------|AV1 <----+-+-+-+--------|     | |
      | | | | |       | :    :    : |: :  : : : : : :  :  : :|     | |
      | | | | |       |<------------|EV1 <----+-+-+-+--------|     | |
      | | | | +-------|             |---------+ | | |        |     | |
      | | | +---------|             |-----------+ | |        |     | |
      | | +-----------|             |-------------+ |        +-----+ |
      | +-------------|             |---------------+                |
      +---------------+             +--------------------------------+
                     Figure 14: Media Projecting Mixer

   So in this six participant conference depicted above in (Figure 14)
   one can see that end-point A will in this case be aware of 5 incoming
   SSRCs, BV1-FV1.  If this mixer intend to have the same behavior as in
   Appendix A.3.2 where the mixer provides the end-points with the two
   latest speaking end-points, then only two out of these five SSRCs
   will concurrently transmitt media to A. As the mixer selects which
   source in the different RTP sessions that transmit media to the end-
   points each media stream will require some rewriting when being
   projected from one session into another.  The main thing is that the
   sequence number will need to be consequitvely incremented based on
   the packet actually being transmitted in each RTP session.  Thus the
   RTP sequence number offset will change each time a source is turned
   on in RTP session.

   As the RTP sessions are independent the SSRC numbers used can be
   handled indepdentently also thus working around any SSRC collisions
   by having remapping tables between the RTP sessions.  However the
   related MediaStream signalling must be correspondlingly changed to
   ensure consistent MediaStream to SSRC mappings between the different
   PeerConnections and the same comment that higher functions must not
   use SSRC as references to media streams applies also here.

   The mixer will also be responsible to act on any RTCP codec control
   requests comming from an end-point and decide if it can act on it
   locally or needs to translate the request into the RTP session that
   contains the media source.  Both end-points and the mixer will need
   to implement conference related codec control functionalities to
   provide a good experience.  Full Intra Request to request from the
   media source to provide switching points between the sources,
   Temporary Maximum Media Bit-rate Request (TMMBR) to enable the mixer
   to aggregate congestion control response towards the media source and
   have it adjust its bit-rate in case the limitation is not in the
   source to mixer link.

   This version of the mixer also puts different requirements on the
   end-point when it comes to decoder instances and handling of the
   media streams providing media.  As each projected SSRC can at any
   time provide media the end-point either needs to handle having thus
   many allocated decoder instances or have efficient switching of
   decoder contexts in a more limited set of actual decoder instances to
   cope with the switches.  The WebRTC application also gets more
   responsibility to update how the media provides is to be presented to
   the user.

A.4.  Translator Based

   There is also a variety of translators.  The core commonality is that
   they do not need to make themselves visible in the RTP level by
   having an SSRC themselves.  Instead they sit between one or more end-
   point and perform translation at some level.  It can be media
   transcoding, protocol translation or covering missing functionality
   for a legacy device or simply relay packets between transport domains
   or to realize multi-party.  We will go in details below.

A.4.1.  Transcoder

   A transcoder operates on media level and really used for two
   purposes, the first is to allow two end-points that doesn't have a
   common set of media codecs to communicate by translating from one
   codec to another.  The second is to change the bit-rate to a lower
   one.  For WebRTC end-points communicating with each other only the
   first one should at all be relevant.  In certain legacy deployment
   media transcoder will be necessary to ensure both codecs and bit-rate
   falls within the envelope the legacy device supports.

   As transcoding requires access to the media the transcoder must
   within the security context and access any media encryption and
   integrity keys.  On the RTP plane a media transcoder will in practice
   fork the RTP session into two different domains that are highly
   decoupled when it comes to media parameters and reporting, but not
   identities.  To maintain signalling bindings to SSRCs a transcoder is
   likely needing to use the SSRC of one end-point to represent the
   transcoded media stream to the other end-point(s).  The congestion
   control loop can be terminated in the transcoder as the media bit-
   rate being sent by the transcoder can be adjusted independently of
   the incoming bit-rate.  However, for optimizing performance and
   resource consumption the translator needs to consider what signals or
   bit-rate reductions it should send towards the source end-point.  For
   example receving a 2.5 mbps video stream and then send out a 250 kbps
   video stream after transcoding is a vaste of resources.  In most
   cases a 500 kbps video stream from the source in the right resolution
   is likely to provide equal quality after transcoding as the 2.5 mbps
   source stream.  At the same time increasing media bit-rate futher
   than what is needed to represent the incoming quality accurate is
   also wasted resources.

       +-A-------------+             +-Translator------------------+
       | +-PeerC1------|             |-PeerC1--------+             |
       | | +-UDP1------|             |-UDP1--------+ |             |
       | | | +-RTP1----|             |-RTP1------+ | |             |
       | | | | +-Audio-|             |-Audio---+ | | | +---+       |
       | | | | |    AA1|------------>|---------+-+-+-+-|DEC|----+  |
       | | | | |       |<------------|BA1 <----+ | | | +---+    |  |
       | | | | |       |             |         |\| | | +---+    |  |
       | | | | +-------|             |---------+ +-+-+-|ENC|<-+ |  |
       | | | +---------|             |-----------+ | | +---+  | |  |
       | | +-----------|             |-------------+ |        | |  |
       | +-------------|             |---------------+        | |  |
       +---------------+             |                        | |  |
                                     |                        | |  |
       +-B-------------+             |                        | |  |
       | +-PeerC2------|             |-PeerC2--------+        | |  |
       | | +-UDP2------|             |-UDP2--------+ |        | |  |
       | | | +-RTP1----|             |-RTP1------+ | |        | |  |
       | | | | +-Audio-|             |-Audio---+ | | | +---+  | |  |
       | | | | |    BA1|------------>|---------+-+-+-+-|DEC|--+ |  |
       | | | | |       |<------------|AA1 <----+ | | | +---+    |  |
       | | | | |       |             |         |\| | | +---+    |  |
       | | | | +-------|             |---------+ +-+-+-|ENC|<---+  |
       | | | +---------|             |-----------+ | | +---+       |
       | | +-----------|             |-------------+ |             |
       | +-------------|             |---------------+             |
       +---------------+             +-----------------------------+

                        Figure 15: Media Transcoder

   Figure 15 exposes some important details.  First of all you can see
   the SSRC identifiers used by the translator are the corresponding
   end-points.  Secondly, there is a relation between the RTP sessions
   in the two different PeerConnections that are represtented by having
   both parts be identified by the same level and they need to share
   certain contexts.  Also certain type of RTCP messages will need to be
   bridged between the two parts.  Certain RTCP feedback messages are
   likely needed to be soruced by the translator in response to actions
   by the translator and its media encoder.

A.4.2.  Gateway / Protocol Translator

   Gateways are used when some protocol feature that is required is not
   supported by an end-point wants to participate in session.  This RTP
   translator in Figure 16 takes on the role of ensuring that from the
   perspective of participant A, participant B appears as a fully
   compliant WebRTC end-point (that is, it is the combination of the
   Translator and participant B that looks like a WebRTC end point).

                               +------------+
                               |            |
                    +---+      | Translator |      +---+
                    | A |<---->| to legacy  |<---->| B |
                    +---+      | end-point  |      +---+
                    WebRTC     |            |     Legacy
                               +------------+

       Figure 16: Gateway (RTP translator) towards legacy end-point

   For WebRTC there are a number of requirements that could force the
   need for a gateway if a WebRTC end-point is to communicate with a
   legacy end-point, such as support of ICE and DTLS-SRTP for
   keymanagement.  On RTP level the main functions that may be missing
   in a legacy implementation that otherswise support RTP are RTCP in
   general, SRTP implementation, congestion control and feedback
   messages required to make it work.

       +-A-------------+             +-Translator------------------+
       | +-PeerC1------|             |-PeerC1------+               |
       | | +-UDP1------|             |-UDP1------+ |               |
       | | | +-RTP1----|             |-RTP1-----------------------+|
       | | | | +-Audio-|             |-Audio---+                  ||
       | | | | |    AA1|------------>|---------+----------------+ ||
       | | | | |       |<------------|BA1 <----+--------------+ | ||
       | | | | |       |<---RTCP---->|<--------+----------+   | | ||
       | | | | +-------|             |---------+      +---+-+ | | ||
       | | | +---------|             |---------------+| T   | | | ||
       | | +-----------|             |-----------+ | || R   | | | ||
       | +-------------|             |-------------+ || A   | | | ||
       +---------------+             |               || N   | | | ||
                                     |               || S   | | | ||
       +-B-(Legacy)----+             |               || L   | | | ||
       |               |             |               || A   | | | ||
       |   +-UDP2------|             |-UDP2------+   || T   | | | ||
       |   | +-RTP1----|             |-RTP1----------+| E   | | | ||
       |   | | +-Audio-|             |-Audio---+      +---+-+ | | ||
       |   | | |       |<---RTCP---->|<--------+----------+   | | ||
       |   | | |    BA1|------------>|---------+--------------+ | ||
       |   | | |       |<------------|AA1 <----+----------------+ ||
       |   | | +-------|             |---------+                  ||
       |   | +---------|             |----------------------------+|
       |   +-----------|             |-----------+                 |
       |               |             |                             |
       +---------------+             +-----------------------------+

                  Figure 17: RTP/RTCP Protocol Translator

   The legacy gateway may be implemented in several ways and what it
   need to change is higly dependent on what functions it need to proxy
   for the legacy end-point.  One possibility is depicted in Figure 17
   where the RTP media streams are compatible and forward without
   changes.  However, their RTP header values are captured to enable the
   RTCP translator to create RTCP reception information related to the
   leg between the end-point and the translator.  This can then be
   combined with the more basic RTCP reports that the legacy endpoint
   (B) provides to give compatible and expected RTCP reporting to A.
   Thus enabling at least full congestion control on the path between A
   and the translator.  If B has limited possibilities for congestion
   response for the media then the translator may need the capabilities
   to perform media transcoding to address cases where it otherwise
   would need to terminate media transmission.

   As the translator are generating RTP/RTCP traffic on behalf of B to A
   it will need to be able to correctly protect these packets that it
   translates or generates.  Thus security context information are
   required in this type of translator if it operates on the RTP/RTCP
   packet content or media.  In fact one of the more likley scenario is
   that the translator (gateway) will need to have two different
   security contexts one towards A and one towards B and for each RTP/
   RTCP packet do a authenticity verification, decryption followed by a
   encryption and integirty protection operation to resolve missmatch in
   security systems.

A.4.3.  Relay

   There exist a class of translators that operates on transport level
   below RTP and thus do not effect RTP/RTCP packets directly.  They
   come in two distinct flavors, the one used to bridge between two
   different transport or address domains to more function as a gateway
   and the second one which is to to provide a group communication
   feature as depicted below in Figure 18.

                    +---+      +------------+      +---+
                    | A |<---->|            |<---->| B |
                    +---+      |            |      +---+
                               | Translator |
                    +---+      |            |      +---+
                    | C |<---->|            |<---->| D |
                    +---+      +------------+      +---+

         Figure 18: RTP Translator (Relay) with Only Unicast Paths

   The first kind is straight forward and is likely to exist in WebRTC
   context when an legacy end-point is compatible with the exception for
   ICE, and thus needs a gateway that terminates the ICE and then
   forwards all the RTP/RTCP traffic and keymanagment to the end-point
   only rewriting the IP/UDP to forward the packet to the legacy node.

   The second type is useful if one wants a less complex central node or
   a central node that is outside of the security context and thus do
   not have access to the media.  This relay takes on the role of
   forwarding the media (RTP and RTCP) packets to the other end-points
   but doesn't perform any RTP or media processing.  Such a device
   simply forwards the media from each sender to all of the other
   particpants, and is sometimes called a transport-layer translator.
   In Figure 18, participant A will only need to send a media once to
   the relay, which will redistribute it by sending a copy of the stream
   to participants B, C, and D. Participant A will still receive three
   RTP streams with the media from B, C and D if they transmit
   simultaneously.  This is from an RTP perspective resulting in an RTP
   session that behaves equivalent to one transporter over an IP Any
   Source Multicast (ASM).

   This results in one common RTP session between all participants
   despite that there will be independent PeerConnections created to the
   translator as depicted below Figure 19.

      +-A-------------+             +-RELAY--------------------------+
      | +-PeerC1------|             |-PeerC1--------+                |
      | | +-UDP1------|             |-UDP1--------+ |                |
      | | | +-RTP1----|             |-RTP1-------------------------+ |
      | | | | +-Video-|             |-Video---+                    | |
      | | | | |    AV1|------------>|---------------------------+  | |
      | | | | |       |<------------|BV1 <--------------------+ |  | |
      | | | | |       |<------------|CV1 <------------------+ | |  | |
      | | | | +-------|             |---------+             | | |  | |
      | | | +---------|             |-------------------+   ^ ^ V  | |
      | | +-----------|             |-------------+ |   |   | | |  | |
      | +-------------|             |---------------+   |   | | |  | |
      +---------------+             |                   |   | | |  | |
                                    |                   |   | | |  | |
      +-B-------------+             |                   |   | | |  | |
      | +-PeerC2------|             |-PeerC2--------+   |   | | |  | |
      | | +-UDP2------|             |-UDP2--------+ |   |   | | |  | |
      | | | +-RTP2----|             |-RTP1--------------+   | | |  | |
      | | | | +-Video-|             |-Video---+             | | |  | |
      | | | | |    BV1|------------>|-----------------------+ | |  | |
      | | | | |       |<------------|AV1 <----------------------+  | |
      | | | | |       |<------------|CV1 <--------------------+ |  | |
      | | | | +-------|             |---------+             | | |  | |
      | | | +---------|             |-------------------+   | | |  | |
      | | +-----------|             |-------------+ |   |   V ^ V  | |
      | +-------------|             |---------------+   |   | | |  | |
      +---------------+             |                   |   | | |  | |
                                    :                   |   | | |  | |
                                    :                   |   | | |  | |
      +-C-------------+             |                   |   | | |  | |
      | +-PeerC3------|             |-PeerC3--------+   |   | | |  | |
      | | +-UDP3------|             |-UDP3--------+ |   |   | | |  | |
      | | | +-RTP3----|             |-RTP1--------------+   | | |  | |
      | | | | +-Video-|             |-Video---+             | | |  | |
      | | | | |    CV1|------------>|-------------------------+ |  | |
      | | | | |       |<------------|AV1 <----------------------+  | |
      | | | | |       |<------------|BV1 <------------------+      | |
      | | | | +-------|             |---------+                    | |
      | | | +---------|             |------------------------------+ |
      | | +-----------|             |-------------+ |                |
      | +-------------|             |---------------+                |
      +---------------+             +--------------------------------+

                  Figure 19: Transport Multi-party Relay

   As the Relay RTP and partial source authentication.  A
   mandatory to implement media security solution will be required to be
   picked.  We currently don't discuss the key-management aspect of SRTP
   in this memo, that needs to be done taking RTCP packets between the WebRTC communication
   model into account.

   The guidelines in [I-D.ietf-avtcore-srtp-vbr-audio] apply when using
   variable bit rate (VBR) audio codecs, for example Opus or UDP flows as indicated
   by the Mixer
   audio level header extensions.

   Security considerations arrows for the media flow a given WebRTC work are discussed in
   [I-D.ietf-rtcweb-security].

11.  Acknowledgements

   The authors would end-point, like to thank Harald Alvestrand, Cary Bran, and
   Cullen Jennings for valuable feedback.

12.  References

12.1.  Normative References

   [I-D.holmberg-mmusic-sdp-bundle-negotiation]
              Holmberg, C. and H. Alvestrand, "Multiplexing Negotiation
              Using Session Description Protocol (SDP) Port Numbers",
              draft-holmberg-mmusic-sdp-bundle-negotiation-00 (work in
              progress), October 2011.

   [I-D.ietf-avtcore-srtp-encrypted-header-ext]
              Lennox, J., "Encryption of Header Extensions in the Secure
              Real-Time Transport Protocol (SRTP)",
              draft-ietf-avtcore-srtp-encrypted-header-ext-00 (work in
              progress), June 2011.

   [I-D.ietf-avtcore-srtp-vbr-audio]
              Perkins, C. and J. Valin, "Guidelines for the use of
              Variable Bit Rate Audio with Secure RTP",
              draft-ietf-avtcore-srtp-vbr-audio-03 (work in progress),
              July 2011.

   [I-D.ietf-rtcweb-overview]
              Alvestrand, H., "Overview: Real Time Protocols for Brower-
              based Applications", draft-ietf-rtcweb-overview-02 (work
              in progress), September 2011.

   [I-D.ietf-rtcweb-security]
              Rescorla, E., "Security Considerations for RTC-Web",
              draft-ietf-rtcweb-security-01 (work in progress),
              October 2011.

   [I-D.jesup-rtp-congestion-reqs]
              Jesup, R. and H. Alvestrand, "Congestion Control
              Requirements For Real Time Media",
              draft-jesup-rtp-congestion-reqs-00 (work in progress),
              March 2012.

   [I-D.lennox-rtcweb-rtp-media-type-mux]
              Lennox, J. and J. Rosenberg, "Multiplexing Multiple Media
              Types In a Single Real-Time Transport Protocol (RTP)
              Session", draft-lennox-rtcweb-rtp-media-type-mux-00 (work
              in progress), October 2011.

   [I-D.perkins-avtcore-rtp-circuit-breakers]
              Perkins, C. A
   will see the remote sources BV1 and V. Singh, "RTP Congestion Control: Circuit
              Breakers for Unicast Sessions",
              draft-perkins-avtcore-rtp-circuit-breakers-00 (work in
              progress), March 2012.

   [I-D.westerlund-avtcore-multiplex-architecture]
              Westerlund, M., Burman, B., CV1.  There will be also two
   different network paths between A, and B or C. Perkins, "RTP
              Multiplexing Architecture",
              draft-westerlund-avtcore-multiplex-architecture-00 (work This results in progress), October 2011.

   [RFC2736]  Handley, M. and C. Perkins, "Guidelines for Writers that
   the client A must be capable of RTP
              Payload Format Specifications", BCP 36, RFC 2736,
              December 1999.

   [RFC3550]  Schulzrinne, H., Casner, S., Frederick, R., handlilng that when determining
   congestion state that there might exist multiple destinations on the
   far side of a PeerConnection and V.
              Jacobson, "RTP: A Transport Protocol for Real-Time
              Applications", STD 64, RFC 3550, July 2003.

   [RFC3551]  Schulzrinne, H. that these paths shall be treated
   differently.  It also results in a requirement to combine the
   different congestion states into a decision to transmit a particular
   media stream suitable to all participants.

   It is also important to note that the relay can not perform selective
   relaying of some sources and S. Casner, "RTP Profile for Audio not others.  The reason is that the RTCP
   reporting in that case becomes incosistent and
              Video Conferences with Minimal Control", STD 65, RFC 3551,
              July 2003.

   [RFC3556]  Casner, S., "Session Description Protocol (SDP) Bandwidth
              Modifiers for without explicit
   information about it being blocked must be interpret as severe
   congestion.

   In this usage it is also necessary that the session management has
   configured a common set of RTP Control Protocol (RTCP) Bandwidth",
              RFC 3556, July 2003.

   [RFC3711]  Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K.
              Norrman, "The Secure Real-time Transport Protocol (SRTP)",
              RFC 3711, March 2004.

   [RFC4585]  Ott, J., Wenger, S., Sato, N., Burmeister, C., configuration including RTP payload
   formats as when A sends a packet with pt=97 it will arrive at both B
   and J. Rey,
              "Extended C carrying pt=97 and having the same packetization and encoding,
   no entity will have manipulated the packet.

   When it comes to security there exist some additional requirements to
   ensure that the property that the relay can't read the media traffic
   is enforced.  First of all the key to be used must be agreed such so
   that the relay doesn't get it, e.g. no DTLS-SRTP handshake with the
   relay, instead some other method must be used.  Secondly, the keying
   structure must be capable of handling multiple end-points in the same
   RTP Profile session.

   The second problem can basically be solved in two ways.  Either a
   common master key from which all derive their per source key for Real-time Transport Control
              Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585,
              July 2006.

   [RFC4588]  Rey, J., Leon, D., Miyazaki, A., Varsa, V.,
   SRTP.  The second alternative which might be more practical is that
   each end-point has its own key used to protects all RTP/RTCP packets
   it sends.  Each participants key are then distributed to the other
   participants.  This second method could be implemented using DTLS-
   SRTP to a special key server and R.
              Hakenberg, "RTP Retransmission Payload Format", RFC 4588,
              July 2006.

   [RFC4961]  Wing, D., "Symmetric then use Encrypted Key Transport
   [I-D.ietf-avt-srtp-ekt] to distribute the actual used key to the
   other participants in the RTP session Figure 20.  The first one could
   be achieved using MIKEY messages in SDP.

                 +---+                               +---+
                 |   |         +-----------+         |   |
                 | A |<------->| DTLS-SRTP |<------->| C |
                 |   |<--   -->|   HOST    |<--   -->|   |
                 +---+   \ /   +-----------+   \ /   +---+
                          X                     X
                 +---+   / \   +-----------+   / \   +---+
                 |   |<--   -->|    RTP Control Protocol (RTCP)",
              BCP 131, RFC 4961, July 2007.

   [RFC5104]  Wenger, S., Chandra, U., Westerlund, M.,    |<--   -->|   |
                 | B |<------->|   RELAY   |<------->| D |
                 |   |         +-----------+         |   |
                 +---+                               +---+

             Figure 20: DTLS-SRTP host and RTP Relay Separated

   The relay can still verify that a given SSRC isn't used or spoofed by
   another participant within the multi-party session by binding SSRCs
   on their first usage to a given source address and B. Burman,
              "Codec Control Messages in port pair.
   Packets carrying that source SSRC from other addresses can be
   suppressed to prevent spoofing.  This is possible as long as SRTP is
   used which leaves the SSRC of the packet originator in RTP Audio-Visual Profile
              with Feedback (AVPF)", RFC 5104, February 2008.

   [RFC5109]  Li, A., "RTP Payload Format for Generic Forward Error
              Correction", RFC 5109, December 2007.

   [RFC5124]  Ott, J. and E. Carrara, "Extended Secure RTP Profile RTCP
   packets in the clear.  If such packet level method for
              Real-time Transport Control Protocol (RTCP)-Based Feedback
              (RTP/SAVPF)", RFC 5124, February 2008.

   [RFC5285]  Singer, D. and H. Desineni, "A General Mechanism enforcing
   source authentication within the group, then there exist
   cryptographic methods such as TESLA [RFC4383] that could be used for
   true source authentication.

A.5.  End-point Forwarding

   An WebRTC end-point (B in Figure 21) will receive a MediaStream (set
   of SSRCs) over a PeerConnection (from A).  For the moment is not
   decided if the end-point is allowed or not to in its turn send that
   MediaStream over another PeerConnection to C. This section discusses
   the RTP
              Header Extensions", RFC 5285, July 2008.

   [RFC5506]  Johansson, I. and M. Westerlund, "Support end-point implications of allowing such functionality,
   which on the API level is extremely simplistic to perform.

                          +---+    +---+    +---+
                          | A |--->| B |--->| C |
                          +---+    +---+    +---+

                     Figure 21: MediaStream Forwarding

   There exist two main approaches to how B forwards the media from A to
   C. The first one is to simply relay the media stream.  The second one
   is for Reduced-Size
              Real-Time Transport Control Protocol (RTCP): Opportunities B to act as a transcoder.  Lets consider both approaches.

   A relay approache will result in that the WebRTC end-points will have
   to have the same capabilities as being discussed in Relay
   (Appendix A.4.3).  Thus A will see an RTP session that is extended
   beyond the PeerConnection and Consequences", RFC 5506, April 2009.

   [RFC5761]  Perkins, C. see two different receiving end-points
   with different path characteristics (B and M. Westerlund, "Multiplexing RTP Data C).  Thus A's congestion
   control needs to be capable of handling this.  The security solution
   can either support mechanism that allows A to inform C about the key
   A is using despite B and
              Control Packets C having agreed on a Single Port", RFC 5761, April 2010.

   [RFC6051]  Perkins, C. another set of keys.
   Alternatively B will decrypt and T. Schierl, "Rapid Synchronisation then re-encrypt using a new key.
   The relay based approach has the advantage that B does not need to
   transcode the media thus both maintaining the quality of RTP
              Flows", RFC 6051, November 2010.

   [RFC6222]  Begen, A., Perkins, C., the encoding
   and D. Wing, "Guidelines for
              Choosing RTP Control Protocol (RTCP) Canonical Names
              (CNAMEs)", RFC 6222, April 2011.

   [RFC6464]  Lennox, J., Ivov, E., reducing B's complexity requirements.  If the right security
   solutions are supported then also C will be able to verify the
   authenticity of the media comming from A. As downside A are forced to
   take both B and E. Marocco, "A Real-time
              Transport Protocol (RTP) Header Extension for Client-to- C into consideration when delivering content.

   The media transcoder approach is similar to having B act as Mixer Audio Level Indication", RFC 6464, December 2011.

   [RFC6465]  Ivov, E., Marocco, E., and J. Lennox, "A Real-time
              Transport Protocol (RTP) Header Extension for Mixer-to-
              Client Audio Level Indication", RFC 6465, December 2011.

12.2.  Informative References

   [I-D.begen-mmusic-redundancy-grouping]
              Begen, A., Cai, Y., and H. Ou, "Duplication Grouping
              Semantics in
   terminating the RTP session combined with the Session Description Protocol",
              draft-begen-mmusic-redundancy-grouping-01 (work transcoder as discussed
   in
              progress), June 2011.

   [I-D.cbran-rtcweb-data]
              Bran, C. and C. Jennings, "RTC-Web Non-Media Data
              Transport Requirements", draft-cbran-rtcweb-data-00 (work Appendix A.4.1.  A will only see B as receiver of its media.  B
   will responsible to produce a media stream suitable for the B to C
   PeerConnection.  This may require media transcoding for congestion
   control purpose to produce a suitable bit-rate.  Thus loosing media
   quality in progress), July 2011.

   [I-D.ietf-fecframe-framework]
              Watson, M., Begen, A., the transcoding and V. Roca, "Forward Error
              Correction (FEC) Framework",
              draft-ietf-fecframe-framework-15 (work forcing B to spend the resource on the
   transcoding.  The media transcoding does result in progress),
              June 2011.

   [RFC2198]  Perkins, C., Kouvelas, I., Hodson, O., Hardman, V.,
              Handley, M., Bolot, J., Vega-Garcia, A., and S. Fosse-
              Parisis, "RTP Payload a separation of
   the two different legs removing almost all dependencies.  B could
   choice to implement logic to optimize its media transcoding
   operation, by for Redundant Audio Data", RFC 2198,
              September 1997.

   [RFC4341]  Floyd, S. and E. Kohler, "Profile example requesting media properties that are
   suitable for Datagram Congestion
              Control Protocol (DCCP) Congestion Control ID 2: TCP-like
              Congestion Control", RFC 4341, March 2006.

   [RFC4342]  Floyd, S., Kohler, E., C also, thus trying to avoid it having to transcode the
   content and J. Padhye, "Profile only forward the media payloads between the two sides.
   For that optimization to be practical WebRTC end-points must support
   sufficiently good tools for
              Datagram Congestion Control Protocol (DCCP) Congestion
              Control ID 3: TCP-Friendly Rate Control (TFRC)", RFC 4342,
              March 2006.

   [RFC4828]  Floyd, S. and E. Kohler, "TCP Friendly Rate Control
              (TFRC): The Small-Packet (SP) Variant", RFC 4828,
              April 2007.

   [RFC4867]  Sjoberg, J., Westerlund, M., Lakaniemi, A., and Q. Xie,
              "RTP Payload Format and File Storage Format codec control.

A.6.  Simulcast

   This section discusses simulcast in the meaning of providing a node,
   for example a stream switching Mixer, with multiple different encoded
   version of the
              Adaptive Multi-Rate (AMR) and Adaptive Multi-Rate Wideband
              (AMR-WB) Audio Codecs", RFC 4867, April 2007.

   [RFC5117]  Westerlund, M. and S. Wenger, "RTP Topologies", RFC 5117,
              January 2008.

   [RFC5348]  Floyd, S., Handley, M., Padhye, J., and J. Widmer, "TCP
              Friendly Rate Control (TFRC): Protocol Specification",
              RFC 5348, September 2008.

   [RFC5404]  Westerlund, M. same media source.  In the WebRTC context that appears
   to be most easily accomplished by establishing mutliple
   PeerConnection all being feed the same set of MediaStreams.  Each
   PeerConnection is then configured to deliver a particular media
   quality and I. Johansson, "RTP Payload Format for
              G.719", RFC 5404, January 2009.

   [RFC5681]  Allman, M., Paxson, V., thus media bit-rate.  This will work well as long as the
   end-point implements media encoding according to Figure 9.  Then each
   PeerConnection will receive an independently encoded version and E. Blanton, "TCP Congestion
              Control", RFC 5681, September 2009.

   [RFC5968]  Ott, J. the
   codec parameters can be agreed specifically in the context of this
   PeerConnection.

   For simulcast to work one needs to prevent that the end-point deliver
   content encoded as depicted in Figure 10.  If a single encoder
   instance is feed to multiple PeerConnections the intention of
   performing simulcast will fail.

   Thus it should be considered to explicitly signal which of the two
   implementation strategies that are desired and C. Perkins, "Guidelines for Extending which will be done.
   At least making the RTP
              Control Protocol (RTCP)", RFC 5968, September 2010.

   [RFC6263]  Marjou, X. application and A. Sollaud, "Application Mechanism for
              Keeping Alive possible the NAT Mappings Associated with RTP / RTP
              Control Protocol (RTCP) Flows", RFC 6263, June 2011. central node
   interested in receiving simulcast of an end-points media streams to
   be aware if it will function or not.

Authors' Addresses

   Colin Perkins
   University of Glasgow
   School of Computing Science
   Glasgow  G12 8QQ
   United Kingdom

   Email: csp@csperkins.org

   Magnus Westerlund
   Ericsson
   Farogatan 6
   SE-164 80 Kista
   Sweden

   Phone: +46 10 714 82 87
   Email: magnus.westerlund@ericsson.com

   Joerg Ott
   Aalto University
   School of Electrical Engineering
   Espoo  02150
   Finland

   Email: jorg.ott@aalto.fi