One document matched: draft-westerlund-avtcore-multiplex-architecture-03.xml
<?xml version="1.0" encoding="US-ASCII"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [
<!ENTITY rfc2119 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml">
<!ENTITY rfc3550 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3550.xml">
<!ENTITY rfc2198 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2198.xml">
<!ENTITY rfc2205 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2205.xml">
<!ENTITY rfc2326 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2326.xml">
<!ENTITY rfc2474 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2474.xml">
<!ENTITY rfc2974 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2974.xml">
<!ENTITY rfc3261 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3261.xml">
<!ENTITY rfc3264 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3264.xml">
<!ENTITY rfc3389 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3389.xml">
<!ENTITY rfc3551 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3551.xml">
<!ENTITY rfc3711 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3711.xml">
<!ENTITY rfc3830 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3830.xml">
<!ENTITY rfc4103 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4103.xml">
<!ENTITY rfc4566 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4566.xml">
<!ENTITY rfc4568 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4568.xml">
<!ENTITY rfc4588 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4588.xml">
<!ENTITY rfc4607 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4607.xml">
<!ENTITY rfc5104 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5104.xml">
<!ENTITY rfc5117 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5117.xml">
<!ENTITY rfc5576 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5576.xml">
<!ENTITY rfc5583 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5583.xml">
<!ENTITY rfc5760 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5760.xml">
<!ENTITY rfc5761 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5761.xml">
<!ENTITY rfc5764 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5764.xml">
<!ENTITY rfc5888 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5888.xml">
<!ENTITY rfc6190 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6190.xml">
<!ENTITY rfc6222 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6222.xml">
<!ENTITY rfc6285 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6285.xml">
<!ENTITY rfc6465 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6465.xml">
<!ENTITY draft-ietf-avtext-multiple-clock-rates SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-avtext-multiple-clock-rates.xml">
<!ENTITY draft-ietf-payload-rtp-howto SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-payload-rtp-howto.xml">
<!ENTITY draft-ietf-avt-srtp-ekt SYSTEM "http://xml.resource.org/public/rfc/bibxml3/reference.I-D.ietf-avt-srtp-ekt.xml">
]>
<?rfc toc="yes"?>
<?rfc tocompact="yes"?>
<?rfc tocdepth="3"?>
<?rfc tocindent="yes"?>
<?rfc symrefs="yes"?>
<?rfc sortrefs="yes"?>
<?rfc comments="yes"?>
<?rfc inline="yes"?>
<?rfc compact="yes"?>
<?rfc subcompact="no"?>
<rfc category="info"
docName="draft-westerlund-avtcore-multiplex-architecture-03"
ipr="trust200902">
<front>
<title abbrev="Guidelines for Multiplexing in RTP">Guidelines for using
the Multiplexing Features of RTP</title>
<author fullname="Magnus Westerlund" initials="M." surname="Westerlund">
<organization>Ericsson</organization>
<address>
<postal>
<street>Farogatan 6</street>
<city>SE-164 80 Kista</city>
<country>Sweden</country>
</postal>
<phone>+46 10 714 82 87</phone>
<email>magnus.westerlund@ericsson.com</email>
</address>
</author>
<author fullname="Bo Burman" initials="B." surname="Burman">
<organization>Ericsson</organization>
<address>
<postal>
<street>Farogatan 6</street>
<city>SE-164 80 Kista</city>
<country>Sweden</country>
</postal>
<phone>+46 10 714 13 11</phone>
<email>bo.burman@ericsson.com</email>
</address>
</author>
<author fullname="Colin Perkins" initials="C. " surname="Perkins">
<organization>University of Glasgow</organization>
<address>
<postal>
<street>School of Computing Science</street>
<city>Glasgow</city>
<code>G12 8QQ</code>
<country>United Kingdom</country>
</postal>
<email>csp@csperkins.org</email>
</address>
</author>
<author fullname="Harald Tveit Alvestrand" initials="H."
surname="Alvestrand">
<organization>Google</organization>
<address>
<postal>
<street>Kungsbron 2</street>
<city>Stockholm</city>
<region/>
<code>11122</code>
<country>Sweden</country>
</postal>
<phone/>
<facsimile/>
<email>harald@alvestrand.no</email>
<uri/>
</address>
</author>
<date day="25" month="February" year="2013"/>
<abstract>
<t>Real-time Transport Protocol (RTP) is a flexible protocol possible to
use in a wide range of applications and network and system topologies.
This flexibility and the implications of different choices should be
understood by any application developer using RTP. To facilitate that
understanding, this document contains an in-depth discussion of the
usage of RTP's multiplexing points; the RTP session and the
Synchronisation Source Identifier (SSRC). The document tries to give
guidance and source material for an analysis on the most suitable
choices for the application being designed.</t>
</abstract>
</front>
<middle>
<section title="Introduction">
<t><xref target="RFC3550">Real-time Transport Protocol (RTP)</xref> is a
commonly used protocol for real-time media transport. It is a protocol
that provides great flexibility and can support a large set of different
applications. RTP has several multiplexing points designed for different
purposes. These enable support of multiple media streams and switching
between different encoding or packetization of the media. By using
multiple RTP sessions, sets of media streams can be structured for
efficient processing or identification. Thus the question for any RTP
application designer is how to best use the RTP session, the SSRC and
the payload type to meet the application's needs.</t>
<t>The purpose of this document is to provide clear information about
the possibilities of RTP when it comes to multiplexing. The RTP
application designer should understand the implications that come from a
particular usage of the RTP multiplexing points. The document will
recommend against some usages as being unsuitable, in general or for
particular purposes.</t>
<t>RTP was from the beginning designed for multiple participants in a
communication session. This is not restricted to multicast, as some may
believe, but also provides functionality over unicast, using either
multiple transport flows below RTP or a network node that re-distributes
the RTP packets. The re-distributing node can for example be a transport
translator (relay) that forwards the packets unchanged, a translator
performing media or protocol translation in addition to forwarding, or
an RTP mixer that creates new conceptual sources from the received
streams. In addition, multiple streams may occur when a single endpoint
have multiple media sources, like multiple cameras or microphones that
need to be sent simultaneously.</t>
<t>This document has been written due to increased interest in more
advanced usage of RTP, resulting in questions regarding the most
appropriate RTP usage. The limitations in some implementations, RTP/RTCP
extensions, and signalling has also been exposed. It is expected that
some limitations will be addressed by updates or new extensions
resolving the shortcomings. The authors also hope that clarification on
the usefulness of some functionalities in RTP will result in more
complete implementations in the future.</t>
<t>The document starts with some definitions and then goes into the
existing RTP functionalities around multiplexing. Both the desired
behaviour and the implications of a particular behaviour depend on which
topologies are used, which requires some consideration. This is followed
by a discussion of some choices in multiplexing behaviour and their
impacts. Some arch-types of RTP usage are discussed. Finally, some
recommendations and examples are provided.</t>
<t>This document is currently an individual contribution, but it is the
intention of the authors that this should become a WG document that
objectively describes and provides suitable recommendations for which
there is WG consensus. Currently this document only represents the views
of the authors. The authors gladly accept any feedback on the document
and will be happy to discuss suitable recommendations.</t>
</section>
<section title="Definitions">
<t/>
<section title="Terminology">
<t>The following terms and abbreviations are used in this
document:</t>
<t><list style="hanging">
<t hangText="Endpoint:">A single entity sending or receiving RTP
packets. It may be decomposed into several functional blocks, but
as long as it behaves a single RTP stack entity it is classified
as a single endpoint.</t>
<t hangText="Multiparty:">A communication situation including
multiple end-points. In this document it will be used to refer to
situations where more than two end-points communicate.</t>
<t hangText="Media Source:">The source of a stream of data of one
Media Type, It can either be a single media capturing device such
as a video camera, a microphone, or a specific output of a media
production function, such as an audio mixer, or some video editing
function. Sending data from a Media Source may cause multiple RTP
sources to send multiple Media Streams.</t>
<t hangText="Media Stream:">A sequence of RTP packets using a
single SSRC that together carries part or all of the content of a
specific Media Type from a specific sender source within a given
RTP session.</t>
<t hangText="RTP Source:">The originator or source of a particular
Media Stream. Identified using an SSRC in a particular RTP
session. An RTP source is the source of a single media stream, and
is associated with a single endpoint and a single Media Source. An
RTP Source is just called a Source in RFC 3550.</t>
<t hangText="Media Sink:">A recipient of a Media Stream. The
endpoint sinking media are Identified using one or more SSRCs.
There may be more than one Media Sink for one RTP source.</t>
<t hangText="CNAME:">"Canonical name" - identifier associated with
one or more RTP sources from a single endpoint. Defined in <xref
target="RFC3550">the RTP specification</xref>. A CNAME identifies
a synchronisation context. A CNAME is associated with a single
endpoint, although some RTP nodes will use an end-points CNAME on
that end-points behalf. An endpoint may use multiple CNAMEs. A
CNAME is intended to be globally unique and stable for the full
duration of a communication session. <xref target="RFC6222"/><xref
target="I-D.ietf-avtcore-6222bis"/> gives updated guidelines for
choosing CNAMEs.</t>
<t hangText="Media Type:">Audio, video, text or data whose form
and meaning are defined by a specific real-time application.</t>
<t hangText="Multiplex:">The operation of taking multiple entities
as input, aggregating them onto some common resource while keeping
the individual entities addressable such that they can later be
fully and unambiguously separated (de-multiplexed) again.</t>
<t hangText="RTP Session:">As defined by <xref target="RFC3550"/>,
the endpoints belonging to the same RTP Session are those that
share a single SSRC space. That is, those endpoints can see an
SSRC identifier transmitted by any one of the other endpoints. An
endpoint can receive an SSRC either as SSRC or as CSRC in RTP and
RTCP packets. Thus, the RTP Session scope is decided by the
endpoints' network interconnection topology, in combination with
RTP and RTCP forwarding strategies deployed by endpoints and any
interconnecting middle nodes.</t>
<t hangText="RTP Session Group:">One or more RTP sessions that are
used together to perform some function. Examples are multiple RTP
sessions used to carry different layers of a layered encoding. In
an RTP Session Group, CNAMEs are assumed to be valid across all
RTP sessions, and designate synchronisation contexts that can
cross RTP sessions.</t>
<t hangText="Source:">Term that should not be used alone. An RTP
Source, as identified by its SSRC, is the source of a single Media
Stream; a Media Source can be the source of mutiple Media
Streams.</t>
<t hangText="SSRC:">An RTP 32-bit unsigned integer used as
identifier for a RTP Source.</t>
<t hangText="CSRC:">Contributing Source, A SSRC identifier used in
a context, like the RTP headers CSRC list, where it is clear that
the Media Source is not the source of the media stream, instead
only a contributor to the Media Stream.</t>
<t hangText="Signalling:">The process of configuring endpoints to
participate in one or more RTP sessions.</t>
<t/>
</list></t>
</section>
<section title="Subjects Out of Scope">
<t>This document is focused on issues that affect RTP. Thus, issues
that involve signalling protocols, such as whether SIP, Jingle or some
other protocol is in use for session configuration, the particular
syntaxes used to define RTP session properties, or the constraints
imposed by particular choices in the signalling protocols, are
mentioned only as examples in order to describe the RTP issues more
precisely.</t>
<t>This document assumes the applications will use RTCP. While there
are such applications that don't send RTCP, they do not conform to the
RTP specification, and thus should be regarded as reusing the RTP
packet format, not as implementing the RTP protocol.</t>
</section>
</section>
<section anchor="sec-mux-points" title="RTP Concepts">
<t>This section describes the existing RTP tools that are particularly
important when discussing multiplexing of different media streams.</t>
<section title="Session">
<t>The RTP Session is the highest semantic level in RTP and contains
all of the RTP functionality. RTP itself has no normative statements
about the relationship between different RTP sessions.<list
style="hanging">
<t hangText="Identifier:">RTP in itself does not contain any
Session identifier, but relies either on the underlying transport
or on the used signalling protocol, depending on in which context
the identifier is used (e.g. transport or signalling). Due to
this, a single RTP Session may have multiple associated
identifiers belonging to different contexts.<list style="hanging">
<t hangText="Position:">Depending on underlying transport and
signalling protocol. For example, when running RTP on top of
UDP, an RTP endpoint can identify and delimit an RTP Session
from other RTP Sessions through the UDP source and destination
transport address, consisting of network address and port
number(s). Commonly, RTP and RTCP use separate ports and the
destination transport address is in fact an address pair, but
in the case of <xref target="RFC5761">RTP/RTCP
multiplex</xref> there is only a single port. Another example
is <xref target="RFC4566">SDP signalling</xref>, where the
<xref target="RFC5888">grouping framework</xref> uses an
identifier per "m="-line. If there is a one-to-one mapping
between "m="-line and RTP Session, that grouping framework
identifier can identify a single RTP Session.</t>
<t hangText="Usage:">Identify separate RTP Sessions.</t>
<t hangText="Uniqueness:">Globally unique, but identity can
only be detected by the general communication context for the
specific endpoint.</t>
<t hangText="Inter-relation:">Depending on the underlying
transport and signalling protocol.</t>
</list></t>
<t hangText="Special Restrictions:">None.</t>
</list></t>
<t>A RTP source in an RTP session that changes its source transport
address during a session must also choose a new SSRC identifier to
avoid being interpreted as a looped source.</t>
<t>The set of participants considered part of the same RTP Session is
defined by <xref target="RFC3550">the RTP specification</xref> as
those that share a single SSRC space. That is, those participants that
can see an SSRC identifier transmitted by any one of the other
participants. A participant can receive an SSRC either as SSRC or CSRC
in RTP and RTCP packets. Thus, the RTP Session scope is decided by the
participants' network interconnection topology, in combination with
RTP and RTCP forwarding strategies deployed by endpoints and any
interconnecting middle nodes.</t>
</section>
<section title="SSRC">
<t>An SSRC identifies a RTP source or a media sink. For end-points
that both source and sink media streams its SSRCs are used in both
roles. At any given time, a RTP source has one and only one SSRC -
although that may change over the lifetime of the RTP source or sink.
An RTP Session serves one or more RTP sources.<!----><list
style="hanging">
<t hangText="Identifier:">Synchronisation Source (SSRC), 32-bit
unsigned number.<list style="hanging">
<t hangText="Position:">In every RTP and RTCP packet header.
May be present in RTCP payload. May be present in SDP
signalling.</t>
<t hangText="Usage:">Identify individual RTP sources and media
sinks within an RTP Session. Refer to individual RTP sources
and media sinks in RTCP messages and SDP signalling.</t>
<t hangText="Uniqueness:">Randomly chosen, intended to be
globally unique within an RTP Session and not dependent on
network address. SSRC value collisions may occur and must be
handled as specified in <xref target="RFC3550">RTP</xref>.</t>
<t hangText="Inter-relation:">SSRC belonging to the same
synchronisation context (originating from the same endpoint),
within or between RTP Sessions, are indicated through use of
identical SDES CNAME items in RTCP compound packets with those
SSRC as originating source. SDP signalling can provide
explicit <xref target="RFC5576">SSRC grouping </xref>. When
CNAME is inappropriate or insufficient, there exist a few
other methods to relate different SSRC. One such case is
session-based <xref target="RFC4588">RTP
retransmission</xref>. In some cases, the same SSRC Identifier
value is used to relate streams in two different RTP Sessions,
such as in Multi-Session Transmission of <xref
target="RFC6190">scalable video</xref>.</t>
</list></t>
<t hangText="Special Restrictions:">All RTP implementations must
be prepared to use procedures for SSRC collision handling, which
results in an SSRC number change. A RTP source that changes its
RTP Session identifier (e.g. source transport address) during a
session must also choose a new SSRC identifier to avoid being
interpreted as looped source.</t>
<t hangText="">Note that RTP sequence number and RTP timestamp are
scoped by SSRC and thus independent between different SSRCs.</t>
</list></t>
<t>An SSRC identifier is used by different type of sources as well as
sinks:<list style="hanging">
<t hangText="Real Media Source:">Connected to a "physical" media
source, for example a camera or microphone.</t>
<t hangText="Conceptual Media Source:">A source with some
attributed property generated by some network node, for example a
filtering function in an RTP mixer that provides the most active
speaker based on some criteria, or a mix representing a set of
other sources.</t>
<t hangText="Media Sink:">A source that does not generate any RTP
media stream in itself (e.g. an endpoint or middlebox only
receiving in an RTP session), but anyway need a sender SSRC for
use as source in RTCP reports.</t>
</list></t>
<t>Note that a endpoint that generates more than one media type, e.g.
a conference participant sending both audio and video, need not (and
commonly should not) use the same SSRC value across RTP sessions. RTCP
Compound packets containing the CNAME SDES item is the designated
method to bind an SSRC to a CNAME, effectively cross-correlating SSRCs
within and between RTP Sessions as coming from the same endpoint. The
main property attributed to SSRCs associated with the same CNAME is
that they are from a particular synchronisation context and may be
synchronised at playback.</t>
<t>An RTP receiver receiving a previously unseen SSRC value must
interpret it as a new source. It may in fact be a previously existing
source that had to change SSRC number due to an SSRC conflict.
However, the originator of the previous SSRC should have ended the
conflicting source by sending an RTCP BYE for it prior to starting to
send with the new SSRC, so the new SSRC is anyway effectively a new
source.</t>
</section>
<section title="CSRC">
<t>The Contributing Source (CSRC) is not a separate identifier, but an
usage of the SSRC identifier. It is optionally included in the RTP
header as list of up to 15 contributing RTP sources. CSRC shares the
SSRC number space and specifies which set of SSRCs that has
contributed to the RTP payload. However, even though each RTP packet
and SSRC can be tagged with the contained CSRCs, the media
representation of an individual CSRC is in general not possible to
extract from the RTP payload since it is typically the result of a
media mixing (merge) operation (by an RTP mixer) on the individual
media streams corresponding to the CSRC identifiers. The exception is
the case when only a single CSRC is indicated as this represent
forwarding of a media stream, possibly modified. The RTP header
extension for <xref target="RFC6465">Mixer-to-Client Audio Level
Indication</xref> expands on the receivers information about a packet
with a CSRC list. Due to these restrictions, CSRC will not be
considered a fully qualified multiplex point and will be disregarded
in the rest of this document.</t>
</section>
<section title="Payload Type">
<t>Each Media Stream utilises one or more encoding formats, identified
by the Payload Type.</t>
<t>The Payload Type is not a multiplexing point. <xref
target="sec-pt-mux"/> gives some of the many reasons why attempting to
use it as a multiplexing point will have bad results.</t>
<t><list style="hanging">
<t hangText="Identifier:">Payload Type number.<list
style="hanging">
<t hangText="Position:">In every RTP header and in
signalling.</t>
<t hangText="Usage:">Identify a specific Media Stream encoding
format. The format definition may be taken from <xref
target="RFC3551"/> for statically allocated Payload Types, but
should be explicitly defined in signalling, such as SDP, both
for static and dynamic Payload Types. The term "format" here
includes whatever can be described by out-of-band signalling
means. In SDP, the term "format" includes media type, RTP
timestamp sampling rate, codec, codec configuration, payload
format configurations, and various robustness mechanisms such
as <xref target="RFC2198">redundant encodings</xref>.</t>
<t hangText="Uniqueness:">Scoped by sending endpoint within an
RTP Session. To avoid any potential for ambiguity, it is
desirable that payload types are unique across all sending
endpoints within an RTP session, but this is often not true in
practice. All SSRC in an RTP session sent from an single
endpoint share the same Payload Types definitions. The RTP
Payload Type is designed such that only a single Payload Type
is valid at any time instant in the SSRC's RTP timestamp time
line, effectively time-multiplexing different Payload Types if
any change occurs. Used Payload Type may change on a
per-packet basis for an SSRC, for example a speech codec
making use of <xref target="RFC3389">generic Comfort
Noise</xref>.</t>
<t hangText="Inter-relation:">There are some uses where
Payload Type numbers need to be unique across RTP Sessions.
This is for example the case in <xref target="RFC5583">Media
Decoding Dependency</xref> where Payload Types are used to
describe media dependency across RTP Sessions. Another example
is session-based <xref target="RFC4588">RTP
retransmission</xref>.</t>
</list></t>
<t hangText="Special Restrictions:">Using different RTP timestamp
clock rates for the RTP Payload Types in use in the same RTP
Session have issues such as potential for loss of synchronisation.
Payload Type clock rate switching requires some special
consideration that is described in the <xref
target="I-D.ietf-avtext-multiple-clock-rates">multiple clock rates
specification</xref>.</t>
</list></t>
<t>If there is a true need to send multiple Payload Types for the same
SSRC that are valid for the same RTP Timestamps, then <xref
target="RFC2198">redundant encodings</xref> can be used. Several
additional constraints than the ones mentioned above need to be met to
enable this use, one of which is that the combined payload sizes of
the different Payload Types must not exceed the transport MTU.</t>
<t>Other aspects of RTP payload format use are described in <xref
target="I-D.ietf-payload-rtp-howto">RTP Payload HowTo </xref>.</t>
</section>
</section>
<section title="Multiple Streams Alternatives">
<t>The reasons why an endpoint may choose to send multiple media streams
are widespread. In the below discussion, please keep in mind that the
reasons for having multiple media streams vary and include but are not
limited to the following:<list style="symbols">
<t>Multiple Media Sources</t>
<t>Multiple Media Streams may be needed to represent one Media
Source (for instance when using layered encodings)</t>
<t>A Retransmission stream may repeat the content of another Media
Stream</t>
<t>An FEC stream may provide material that can be used to repair
another Media Stream</t>
<t>Alternative Encodings, for instance different codecs for the same
audio stream</t>
<t>Alternative formats, for instance multiple resolutions of the
same video stream</t>
</list></t>
<t>Thus the choice made due to one reason may not be the choice suitable
for another reason. In the above list, the different items have
different levels of maturity in the discussion on how to solve them. The
clearest understanding is associated with multiple media sources of the
same media type. However, all warrant discussion and clarification on
how to deal with them.</t>
<t>This section reviews the alternatives to enable multi-stream
handling. Let's start with describing mechanisms that could enable
multiple media streams, independent of the purpose for having multiple
streams.</t>
<t><list style="hanging">
<t hangText="Additional SSRC:">Each additional Media Stream gets its
own SSRC within a RTP Session.</t>
<t hangText="Multiple RTP Sessions:">Using additional RTP Sessions
to handle additional Media Streams.</t>
</list>As the below discussion will show, in reality we cannot choose
a single one of the two solutions. To utilise RTP well and as
efficiently as possible, both are needed. The real issue is finding the
right guidance on when to create RTP sessions and when additional SSRCs
in an RTP session is the right choice.</t>
</section>
<section anchor="sec-topologies" title="RTP Topologies and Issues">
<t>The impact of how RTP Multiplex is performed will in general vary
with how the RTP Session participants are interconnected, described by
<xref target="RFC5117">RTP Topology</xref> and its <xref
target="I-D.westerlund-avtcore-rtp-topologies-update">intended
successor</xref>.</t>
<!--BoB: Only the remaining sub-sections of section 5 contain any text in addition to
rtp-topologies-update. Does that mean that those are the only ones that has any specific
usage considerations, or is it rather that it is unclear what "issues" should be described
in this draft? Seriously consider to focus on recommendations, specific to topologies.
Another option would be to just guide and recommend, without using per-topology division.
-->
<section title="Point to Point">
<t>Even the most basic use case, denoted Topo-Point-to-Point in <xref
target="I-D.westerlund-avtcore-rtp-topologies-update"/>, raises a
number of considerations that are discussed in detail <xref
target="sec-discussion">below</xref>. They range over such aspects
as:<list style="symbols">
<t>Does my communication peer support RTP as defined with multiple
SSRCs?</t>
<t>Do I need network differentiation in form of QoS?</t>
<t>Can the application more easily process and handle the media
streams if they are in different RTP sessions?</t>
<t>Do I need to use additional media streams for RTP
retransmission or FEC.</t>
<t>etc.</t>
</list></t>
<t>The application designer will have to make choices here. The point
to point topology can contain one to many RTP sessions with one to
many media sources per session, resulting in one or more RTP source
(SSRC) per media source. </t>
</section>
<section title="Translators & Gateways">
<t>A point to point communication can end up in a situation when the
peer it is communicating with is not compatible with the other peer
for various reasons:<list style="symbols">
<t>No common media codec for a media type thus requiring
transcoding</t>
<t>Different support for multiple RTP sources and RTP sessions</t>
<t>Usage of different media transport protocols, i.e RTP or
other.</t>
<t>Usage of different transport protocols, e.g. UDP, DCCP, TCP</t>
<t>Different security solutions, e.g. IPsec, TLS, DTLS, SRTP with
different keying mechanisms.</t>
</list></t>
<t>This is in many situations resolved by the inclusion of a
translator in-between the two peers, as described by
Topo-PtP-Translator in <xref
target="I-D.westerlund-avtcore-rtp-topologies-update"/>. The
translator's main purpose is to make the peer look to the other peer
like something it is compatible with. There may also be other reasons
than compatibility to insert a translator in the form of a middlebox
or gateway, for example a need to monitor the media streams. If the
stream transport characteristics are changed by the translator,
appropriate media handling can require thorough understanding of the
application logic, specifically any congestion control or media
adaptation.</t>
</section>
<section title="Point to Multipoint Using Multicast">
<t>This section discusses the Point to Multi-point using Multicast to
interconnect the session participants. This includes both Topo-ASM and
Topo-SSM in <xref
target="I-D.westerlund-avtcore-rtp-topologies-update"/>.</t>
<t>Special considerations must be made as multicast is a one to many
distribution system. For example, the only practical method for
adapting the bit-rate sent towards a given receiver for large groups
is to use a set of multicast groups, where each multicast group
represents a particular bit-rate. Otherwise the whole group gets media
adapted to the participant with the worst conditions. The media
encoding is either scalable, where multiple layers can be combined, or
simulcast where a single version is selected. By either selecting or
combing multicast groups, the receiver can control the bit-rate sent
on the path to itself. It is also common that streams that improve
transport robustness are sent in their own multicast group to allow
for interworking with legacy or to support different levels of
protection.</t>
<t>The result of this is some common behaviours for RTP
multicast:<list style="numbers">
<t>Multicast applications use a group of RTP sessions, not one.
Each endpoint will need to be a member of a number of RTP sessions
in order to perform well.</t>
<t>Within each RTP session, the number of media sinks is likely to
be much larger than the number of RTP sources.</t>
<t>Multicast applications need signalling functions to identify
the relationships between RTP sessions.</t>
<t>Multicast applications need signalling functions to identify
the relationships between SSRCs in different RTP sessions.</t>
</list></t>
<t>All multicast configurations share a signalling requirement; all of
the participants will need to have the same RTP and payload type
configuration. Otherwise, A could for example be using payload type 97
as the video codec H.264 while B thinks it is MPEG-2. It should be
noted that <xref target="RFC3264">SDP offer/answer</xref> has issues
with ensuring this property. The signalling aspects of multicast are
not explored further in this memo.</t>
<t>Security solutions for this type of group communications are also
challenging. First of all the key-management and the security protocol
must support group communication. Source authentication becomes more
difficult and requires special solutions. For more discussion on this
please review <xref
target="I-D.ietf-avtcore-rtp-security-options">Options for Securing
RTP Sessions</xref>.</t>
</section>
<section anchor="sec-translator"
title="Point to Multipoint Using an RTP Transport Translator">
<t>This mode is described as Topo-Translator in <xref
target="I-D.westerlund-avtcore-rtp-topologies-update"/>.</t>
<t>Transport Translators (Relays) result in an RTP session situation
that is very similar to how an ASM group RTP session would behave.</t>
<t>One of the most important aspects with the simple relay is that it
is only rewriting transport headers, no RTP modifications nor media
transcoding occur. The most obvious downside of this basic relaying is
that the translator has no control over how many streams need to be
delivered to a receiver. Nor can it simply select to deliver only
certain streams, as this creates session inconsistencies: If the
translator temporarily stops a stream, this prevents some receivers
from reporting on it. From the sender's perspective it will look like
a transport failure. Applications having needs to stop or switch
streams in the central node should consider using an RTP mixer to
avoid this issue.</t>
<t>The Transport Translator has the same signalling requirement as
multicast: All participants must have the same payload type
configuration. Most of the ASM security issues also arise here. Some
alternative when it comes to solution do exist as there after all
exist a central node to communicate with. One that also can enforce
some security policies depending on the level of trust placed in the
node.</t>
</section>
<section anchor="sec-mixer"
title="Point to Multipoint Using an RTP Mixer">
<t>A mixer, described by Topo-Mixer in <xref
target="I-D.westerlund-avtcore-rtp-topologies-update"/>, is a
centralised node that selects or mixes content in a conference to
optimise the RTP session so that each endpoint only needs connect to
one entity, the mixer. 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.</t>
<t>Mixers have some downsides, the first is that the mixer must be a
trusted node as they either perform media operations or at least
repacketize the media. When using SRTP, both media operations and
repacketization requires that the mixer verifies integrity, decrypts
the content, performs the operation and forms new RTP packets,
encrypts and integrity-protects them. This applies to all types of
mixers. The second downside is that all these operations and
optimisations of the session requires processing. How much depends on
the implementation, as will become evident below.</t>
<t>A mixer, unlike a pure transport translator, is always application
specific: the application logic for stream mixing or stream selection
has to be embedded within the mixer, and controlled using application
specific signalling. The implementation of a mixer can take several
different forms and we will discuss the main themes available that
doesn't break RTP.</t>
<t>Please note that a Mixer could also contain translator
functionalities, like a media transcoder to adjust the media bit-rate
or codec used for a particular RTP media stream.</t>
</section>
</section>
<section anchor="sec-discussion" title="Multiple Streams Discussion">
<section title="Introduction">
<t>Using multiple media streams is a well supported feature of RTP.
However, it can be unclear for most implementers or people writing
RTP/RTCP applications or extensions attempting to apply multiple
streams when it is most appropriate to add an additional SSRC in an
existing RTP session and when it is better to use multiple RTP
sessions. This section tries to discuss the various considerations
needed. The next section then concludes with some guidelines.</t>
</section>
<section title="RTP/RTCP Aspects">
<t>This section discusses RTP and RTCP aspects worth considering when
selecting between using an additional SSRC and Multiple RTP
sessions.</t>
<section anchor="sec-rtp-spec" title="The RTP Specification">
<t>RFC 3550 contains some recommendations and a bullet list with 5
arguments for different aspects of RTP multiplexing. Let's review
Section 5.2 of <xref target="RFC3550"/>, reproduced below:</t>
<t>"For efficient protocol processing, the number of multiplexing
points should be minimised, as described in the <xref
target="ALF">integrated layer processing design principle</xref>. In
RTP, multiplexing is provided by the destination transport address
(network address and port number) which is different for each RTP
session. For example, in a teleconference composed of audio and
video media encoded separately, each medium SHOULD be carried in a
separate RTP session with its own destination transport address.</t>
<t>Separate audio and video streams SHOULD NOT be carried in a
single RTP session and demultiplexed based on the payload type or
SSRC fields. Interleaving packets with different RTP media types but
using the same SSRC would introduce several problems: <list
style="numbers">
<t>If, say, two audio streams shared the same RTP session and
the same SSRC value, and one were to change encodings and thus
acquire a different RTP payload type, there would be no general
way of identifying which stream had changed encodings.</t>
<t>An SSRC is defined to identify a single timing and sequence
number space. Interleaving multiple payload types would require
different timing spaces if the media clock rates differ and
would require different sequence number spaces to tell which
payload type suffered packet loss.</t>
<t>The RTCP sender and receiver reports (see Section 6.4) can
only describe one timing and sequence number space per SSRC and
do not carry a payload type field.</t>
<t>An RTP mixer would not be able to combine interleaved streams
of incompatible media into one stream.</t>
<t>Carrying multiple media in one RTP session precludes: the use
of different network paths or network resource allocations if
appropriate; reception of a subset of the media if desired, for
example just audio if video would exceed the available
bandwidth; and receiver implementations that use separate
processes for the different media, whereas using separate RTP
sessions permits either single- or multiple-process
implementations.</t>
</list></t>
<t>Using a different SSRC for each medium but sending them in the
same RTP session would avoid the first three problems but not the
last two.</t>
<t>On the other hand, multiplexing multiple related sources of the
same medium in one RTP session using different SSRC values is the
norm for multicast sessions. The problems listed above don't apply:
an RTP mixer can combine multiple audio sources, for example, and
the same treatment is applicable for all of them. It may also be
appropriate to multiplex streams of the same medium using different
SSRC values in other scenarios where the last two problems do not
apply."</t>
<t>Let's consider one argument at a time. The first is an argument
for using different SSRC for each individual media stream, which is
very applicable.</t>
<t>The second argument is advocating against using payload type
multiplexing, which still stands as can been seen by the extensive
list of issues found in <xref target="sec-pt-mux"/>.</t>
<t>The third argument is yet another argument against payload type
multiplexing.</t>
<t>The fourth is an argument against multiplexing media streams that
require different handling into the same session. As we saw in the
discussion of RTP mixers, the RTP mixer has to embed application
logic in order to handle streams anyway; the separation of streams
according to stream type is just another piece of application logic,
which may or may not be appropriate for a particular application. A
type of application that can mix different media sources "blindly"
is the audio only "telephone" bridge; most other type of application
needs application-specific logic to perform the mix correctly.</t>
<t>The fifth argument discusses network aspects that we will discuss
more below in <xref target="sec-network-aspects"/>. It also goes
into aspects of implementation, like decomposed endpoints where
different processes or inter-connected devices handle different
aspects of the whole multi-media session.</t>
<t>A summary of RFC 3550's view on multiplexing is to use unique
SSRCs for anything that is its own media/packet stream, and to use
different RTP sessions for media streams that don't share media
type. The first this document support as very valid. The later is
one thing which is further discussed in this document as something
the application developer needs to make a conscious choice for.</t>
<section anchor="sec-multi-media-rec"
title="Different Media Types Recommendations">
<t>The above quote from <xref target="RFC3550">RTP</xref> includes
a strong recommendation:<list style="empty">
<t>"For example, in a teleconference composed of audio and
video media encoded separately, each medium SHOULD be carried
in a separate RTP session with its own destination transport
address."</t>
</list></t>
<t>It was identified in <xref
target="I-D.alvestrand-rtp-sess-neutral">"Why RTP Sessions Should
Be Content Neutral"</xref> that the above statement is poorly
supported by any of the motivations provided in the RTP
specification. This has resulted in the creation of a
specification <xref
target="I-D.ietf-avtcore-multi-media-rtp-session">Multiple Media
Types in an RTP Session specification</xref> which intend to
update this recommendation. That document has a detailed analysis
of the potential issues in having multiple media types in the same
RTP session. This document tries to provide an more over arching
consideration regarding the usage of RTP session and considers
multiple media types in one RTP session as possible choice for the
RTP application designer.</t>
</section>
</section>
<section anchor="sec-self-reporting"
title="Multiple SSRCs in a Session">
<t>Using multiple SSRCs in an RTP session at one endpoint has some
unclarities in the RTP specification. These could potentially lead
to some interoperability issues as well as some potential
significant inefficencies. These are further discussed in <xref
target="I-D.lennox-avtcore-rtp-multi-stream">"RTP Considerations for
Endpoints Sending Multiple Media Streams"</xref>. A application
designer may need to consider these issues and the impact
availability or lack of the optimization in the endpoints has on
their application.</t>
<t>If an application will become affected by the issues described,
using Multiple RTP sessions can mitigate these issues.</t>
</section>
<section title="Handling Varying Sets of Senders">
<t>In some applications, the set of simultaneously active sources
varies within a larger set of session members. A receiver can then
possibly try to use a set of decoding chains that is smaller than
the number of senders, switching the decoding chains between
different senders. As each media decoding chain may contain state,
either the receiver must either be able to save the state of
swapped-out senders, or the sender must be able to send data that
permits the receiver to reinitialise when it resumes activity.</t>
<t>This behaviour will cause similar issues independent of
Additional SSRC or Multiple RTP session.</t>
</section>
<section title="Cross Session RTCP Requests">
<t>There currently exists no functionality to make truly
synchronised and atomic RTCP messages with some type of request
semantics across multiple RTP Sessions. Instead, separate RTCP
messages will have to be sent in each session. This gives streams in
the same RTP session a slight advantage as RTCP messages for
different streams in the same session can be sent in a compound RTCP
packet. Thus providing an atomic operation if different
modifications of different streams are requested at the same
time.</t>
<t>When using multiple RTP sessions, the RTCP timing rules in the
sessions and the transport aspects, such as packet loss and jitter,
prevents a receiver from relying on atomic operations, forcing it to
use more robust and forgiving mechanisms.</t>
</section>
<section anchor="sec-binding-related" title="Binding Related Sources">
<t>A common problem in a number of various RTP extensions has been
how to bind related RTP sources and their media streams together.
This issue is common to both using additional SSRCs and Multiple RTP
sessions.</t>
<t>The solutions can be divided into some groups, RTP/RTCP based,
Signalling based (SDP), grouping related RTP sessions, and grouping
SSRCs within an RTP session. Most solutions are explicit, but some
implicit methods have also been applied to the problem.</t>
<t>The SDP-based signalling solutions are:<list style="hanging">
<t hangText="SDP Media Description Grouping:">The <xref
target="RFC5888">SDP Grouping Framework</xref> uses various
semantics to group any number of media descriptions. These has
previously been considered primarily as grouping RTP sessions,
but this may change.</t>
<t hangText="SDP SSRC grouping:"><xref
target="RFC5576">Source-Specific Media Attributes in SDP</xref>
includes a solution for grouping SSRCs the same way as the
Grouping framework groupes Media Descriptions.</t>
</list></t>
<t>This supports a lot of use cases. Both solutions have
shortcomings in cases where the session's dynamic properties are
such that it is difficult or resource consuming to keep the list of
related SSRCs up to date. As they are two related but still
separated solutions it is not well specified to group SSRCs across
multiple RTP sessions and SDP media descriptions.</t>
<t>Within RTP/RTCP based solutions when binding to a endpoint or
synchronization context, i.e. the CNAME has not be sufficient and
one has multiple RTP sessions has been to using the same SSRC value
across all the RTP sessions. <xref target="RFC4588">RTP
Retransmission</xref> is multiple RTP session mode, <xref
target="RFC5109">Generic FEC</xref>, as well as the <xref
target="RFC6190">RTP payload format for Scalable Video Coding</xref>
in Multi Session Transmission (MST) mode uses this method. This
method clearly works but might have some downside in RTP sessions
with many participating SSRCs. The birthday paradox ensures that if
you populate a single session with 9292 SSRCs at random, the chances
are approximately 1% that at least one collision will occur. When a
collision occur this will force one to change SSRC in all RTP
sessions and thus resynchronizing all of them instead of only the
single media stream having the collision.</t>
<t>It can be noted that Section 8.3 of the <xref
target="RFC3550">RTP Specification</xref> recommends using a single
SSRC space across all RTP sessions for layered coding.</t>
<t>Another solution that has been applied to binding SSRCs have been
an implicit method used by <xref target="RFC4588">RTP
Retransmission</xref> when doing retransmissions in the same RTP
session as the source RTP media stream. This issues an RTP
retransmission request, and then await a new SSRC carrying the RTP
retransmission payload and where that SSRC is from the same CNAME.
This limits a requestor to having only one outstanding request on
any new source SSRCs per endpoint.</t>
<t>There exist no RTP/RTCP based mechanism capable of supporting
explicit association accross multiple RTP sessions as well within an
RTP session. A proposed solution for handling this issue is <xref
target="I-D.westerlund-avtext-rtcp-sdes-srcname"/>. This can
potentially be part of an SDP based solution also by reusing the
same identifiers and name space. </t>
<!---->
</section>
<section title="Forward Error Correction">
<t>There exist a number of Forward Error Correction (FEC) based
schemes for how to reduce the packet loss of the original streams.
Most of the FEC schemes will protect a single source flow. The
protection is achieved by transmitting a certain amount of redundant
information that is encoded such that it can repair one or more
packet loss over the set of packets they protect. This sequence of
redundant information also needs to be transmitted as its own media
stream, or in some cases instead of the original media stream. Thus
many of these schemes create a need for binding the related flows as
discussed above. They also create additional flows that need to be
transported. Looking at the history of these schemes, there is both
schemes using multiple SSRCs and multiple RTP sessions, and some
schemes that support both modes of operation.</t>
<t>Using multiple RTP sessions supports the case where some set of
receivers may not be able to utilise the FEC information. By placing
it in a separate RTP session, it can easily be ignored.</t>
<t>In usages involving multicast, having the FEC information on its
own multicast group, and therefore in its own RTP session, allows
for flexibility, for example when using <xref target="RFC6285">Rapid
Acquisition of Multicast Groups (RAMS)</xref>. During the RAMS burst
where data is received over unicast and where it is possible to
combine with unicast based <xref
target="RFC4588">retransmission</xref>, there is no need to burst
the FEC data related to the burst of the source media streams needed
to catch up with the multicast group. This saves bandwidth to the
receiver during the burst, enabling quicker catch up. When the
receiver has caught up and joins the multicast group(s) for the
source, it can at the same time join the multicast group with the
FEC information. Having the source stream and the FEC in separate
groups allows for easy separation in the Burst/Retransmission Source
(BRS) without having to individually classify packets.</t>
</section>
<section title="Transport Translator Sessions">
<t>A basic Transport Translator relays any incoming RTP and RTCP
packets to the other participants. The main difference between
Additional SSRCs and Multiple RTP Sessions resulting from this use
case is that with Additional SSRCs it is not possible for a
particular session participant to decide to receive a subset of
media streams. When using separate RTP sessions for the different
sets of media streams, a single participant can choose to leave one
of the sessions but not the other.</t>
</section>
</section>
<section title="Interworking">
<t>There are several different kinds of interworking, and this section
discusses two related ones. The interworking between different
applications and the implications of potentially different choices of
usage of RTP's multiplexing points. The second topic relates to what
limitations may have to be considered working with some legacy
applications.</t>
<section title="Types of Interworking">
<t>It is not uncommon that applications or services of similar
usage, especially the ones intended for interactive communication,
ends up in a situation where one want to interconnect two or more of
these applications.</t>
<t>In these cases one ends up in a situation where one might use a
gateway to interconnect applications. This gateway then needs to
change the multiplexing structure or adhere to limitations in each
application.</t>
<t>There are two fundamental approaches to gatewaying: RTP bridging,
where the gateway acts as an RTP Translator, and the two
applications are members of the same RTP session, and RTP
termination, where there are independent RTP sessions running from
each interconnected application to the gateway.</t>
</section>
<section title="RTP Translator Interworking">
<t>From an RTP perspective the RTP Translator approach could work if
all the applications are using the same codecs with the same payload
types, have made the same multiplexing choices, have the same
capabilities in number of simultaneous media streams combined with
the same set of RTP/RTCP extensions being supported. Unfortunately
this may not always be true.</t>
<t>When one is gatewaying via an RTP Translator, a natural
requirement is that the two applications being interconnected must
use the same approach to multiplexing. Furthermore, if one of the
applications is capable of working in several modes (such as being
able to use Additional SSRCs or Multiple RTP sessions at will), and
the other one is not, successful interconnection depends on locking
the more flexible application into the operating mode where
interconnection can be successful, even if no participants using the
less flexible application are present when the RTP sessions are
being created.</t>
</section>
<section title="Gateway Interworking">
<t>When one terminates RTP sessions at the gateway, there are
certain tasks that the gateway must carry out:</t>
<t><list style="symbols">
<t>Generating appropriate RTCP reports for all media streams
(possibly based on incoming RTCP reports), originating from
SSRCs controlled by the gateway.</t>
<t>Handling SSRC collision resolution in each application's RTP
sessions.</t>
<t>Signalling, choosing and policing appropriate bit-rates for
each session.</t>
</list>If either of the applications has any security applied,
e.g. in the form of SRTP, the gateway must be able to decrypt
incoming packets and re-encrypt them in the other application's
security context. This is necessary even if all that's required is a
simple remapping of SSRC numbers. If this is done, the gateway also
needs to be a member of the security contexts of both sides, of
course.</t>
<t>Other tasks a gateway may need to apply include transcoding (for
incompatible codec types), rescaling (for incompatible video size
requirements), suppression of content that is known not to be
handled in the destination application, or the addition or removal
of redundancy coding or scalability layers to fit the need of the
destination domain.</t>
<t>From the above, we can see that the gateway needs to have an
intimate knowledge of the application requirements; a gateway is by
its nature application specific, not a commodity product.</t>
<t>This fact reveals the potential for these gateways to block
evolution of the applications by blocking unknown RTP and RTCP
extensions that the regular application has been extended with.</t>
<t>If one uses security functions, like SRTP, they can as seen above
incur both additional risk due to the gateway needing to be in
security association between the endpoints, unless the gateway is on
the transport level, and additional complexities in form of the
decrypt-encrypt cycles needed for each forwarded packet. SRTP, due
to its keying structure, also requires that each RTP session must
have different master keys, as use of the same key in two RTP
sessions can result in two-time pads that completely breaks the
confidentiality of the packets.</t>
</section>
<section title="Multiple SSRC Legacy Considerations">
<t>Historically, the most common RTP use cases have been point to
point Voice over IP (VoIP) or streaming applications, commonly with
no more than one media source per endpoint and media type (typically
audio and video). Even in conferencing applications, especially
voice only, the conference focus or bridge has provided a single
stream with a mix of the other participants to each participant. It
is also common to have individual RTP sessions between each endpoint
and the RTP mixer, meaning that the mixer functions as an
RTP-terminating gateway.</t>
<t>When establishing RTP sessions that may contain endpoints that
aren't updated to handle multiple streams following these
recommendations, a particular application can have issues with
multiple SSRCs within a single session. These issues include:</t>
<t><list style="numbers">
<t>Need to handle more than one stream simultaneously rather
than replacing an already existing stream with a new one.</t>
<t>Be capable of decoding multiple streams simultaneously.</t>
<t>Be capable of rendering multiple streams simultaneously.</t>
</list></t>
<t>This indicates that gateways attempting to interconnect to this
class of devices must make sure that only one media stream of each
type gets delivered to the endpoint if it's expecting only one, and
that the multiplexing format is what the device expects. It is
highly unlikely that RTP translator-based interworking can be made
to function successfully in such a context.</t>
</section>
</section>
<section anchor="sec-network-aspects" title="Network Aspects">
<t>The multiplexing choice has impact on network level mechanisms that
need to be considered by the implementor.</t>
<section title="Quality of Service">
<t>When it comes to Quality of Service mechanisms, they are either
flow based or marking based. <xref target="RFC2205">RSVP</xref> is
an example of a flow based mechanism, while <xref
target="RFC2474">Diff-Serv</xref> is an example of a Marking based
one. For a marking based scheme, the method of multiplexing will not
affect the possibility to use QoS.</t>
<t>However, for a flow based scheme there is a clear difference
between the methods. Additional SSRC will result in all media
streams being part of the same 5-tuple (protocol, source address,
destination address, source port, destination port) which is the
most common selector for flow based QoS. Thus, separation of the
level of QoS between media streams is not possible. That is however
possible when using multiple RTP sessions, where each media stream
for which a separate QoS handling is desired can be in a different
RTP session that can be sent over different 5-tuples.</t>
</section>
<section title="NAT and Firewall Traversal">
<t>In today's network there exist a large number of middleboxes. The
ones that normally have most impact on RTP are Network Address
Translators (NAT) and Firewalls (FW).</t>
<t>Below we analyze and comment on the impact of requiring more
underlying transport flows in the presence of NATs and
Firewalls:</t>
<t><list style="hanging">
<t hangText="End-Point Port Consumption:">A given IP address
only has 65536 available local ports per transport protocol for
all consumers of ports that exist on the machine. This is
normally never an issue for an end-user machine. It can become
an issue for servers that handle large number of simultaneous
streams. However, if the application uses ICE to authenticate
STUN requests, a server can serve multiple endpoints from the
same local port, and use the whole 5-tuple (source and
destination address, source and destination port, protocol) as
identifier of flows after having securely bound them to the
remote endpoint address using the STUN request. In theory the
minimum number of media server ports needed are the maximum
number of simultaneous RTP Sessions a single endpoint may use.
In practice, implementation will probably benefit from using
more server ports to simplify implementation or avoid
performance bottlenecks.</t>
<t hangText="NAT State:">If an endpoint sits behind a NAT, each
flow it generates to an external address will result in a state
that has to be kept in the NAT. That state is a limited
resource. In home or Small Office/Home Office (SOHO) NATs,
memory or processing are usually the most limited resources. For
large scale NATs serving many internal endpoints, available
external ports are likely the scarce resource. Port limitations
is primarily a problem for larger centralised NATs where
endpoint independent mapping requires each flow to use one port
for the external IP address. This affects the maximum number of
internal users per external IP address. However, it is worth
pointing out that a real-time video conference session with
audio and video is likely using less than 10 UDP flows, compared
to certain web applications that can use 100+ TCP flows to
various servers from a single browser instance.</t>
<t hangText="NAT Traversal Excess Time:">Making the NAT/FW
traversal takes a certain amount of time for each flow. It also
takes time in a phase of communication between accepting to
communicate and the media path being established which is fairly
critical. The best case scenario for how much extra time it
takes after finding the first valid candidate pair following the
specified ICE procedures are: 1.5*RTT + Ta*(Additional_Flows-1),
where Ta is the pacing timer, which ICE specifies to be no
smaller than 20 ms. That assumes a message in one direction, and
then an immediate triggered check back. The reason it isn't
more, is that ICE first finds one candidate pair that works
prior to attempting to establish multiple flows. Thus, there is
no extra time until one has found a working candidate pair.
Based on that working pair the needed extra time is to in
parallel establish the, in most cases 2-3, additional flows.
However, packet loss causes extra delays, at least 100 ms, which
is the minimal retransmission timer for ICE.</t>
<t hangText="NAT Traversal Failure Rate:">Due to the need to
establish more than a single flow through the NAT, there is some
risk that establishing the first flow succeeds but that one or
more of the additional flows fail. The risk that this happens is
hard to quantify, but it should be fairly low as one flow from
the same interfaces has just been successfully established. Thus
only rare events such as NAT resource overload, or selecting
particular port numbers that are filtered etc, should be reasons
for failure.</t>
<t
hangText="Deep Packet Inspection and Multiple Streams:">Firewalls
differ in how deeply they inspect packets. There exist some
potential that deeply inspecting firewalls will have similar
legacy issues with multiple SSRCs as some stack
implementations.</t>
</list></t>
<t>Additional SSRC keeps the additional media streams within one RTP
Session and transport flow and does not introduce any additional NAT
traversal complexities per media stream. This can be compared with
normally one or two additional transport flows per RTP session when
using multiple RTP sessions. Additional lower layer transport flows
will be required, unless an explicit de-multiplexing layer is added
between RTP and the transport protocol. A proposal for how to
multiplex multiple RTP sessions over the same single lower layer
transport exist in <xref
target="I-D.westerlund-avtcore-transport-multiplexing"/>.</t>
</section>
<section title="Multicast">
<t>Multicast groups provides a powerful semantics for a number of
real-time applications, especially the ones that desire
broadcast-like behaviours with one endpoint transmitting to a large
number of receivers, like in IPTV. But that same semantics do result
in a certain number of limitations.</t>
<t>One limitation is that for any group, sender side adaptation to
the actual receiver properties causes degradation for all
participants to what is supported by the receiver with the worst
conditions among the group participants. In most cases this is not
acceptable. Instead various receiver based solutions are employed to
ensure that the receivers achieve best possible performance. By
using scalable encoding and placing each scalability layer in a
different multicast group, the receiver can control the amount of
traffic it receives. To have each scalability layer on a different
multicast group, one RTP session per multicast group is used.</t>
<t>RTP can't function correctly if media streams sent over different
multicast groups where considered part of the same RTP session.
First of all the different layers needs different SSRCs or the
sequence number space seen for a receiver of any sub set of the
layers would have sender side holes. Thus triggering packet loss
reactions. Also any RTCP reporting of such a session would be non
consistent and making it difficult for the sender to determine the
sessions actual state.</t>
<t>Thus it appears easiest and most straightforward to use multiple
RTP sessions. In addition, the transport flow considerations in
multicast are a bit different from unicast. First of all there is no
shortage of port space, as each multicast group has its own port
space.</t>
</section>
<section title="Multiplexing multiple RTP Session on a Single Transport">
<t>For applications that doesn't need flow based QoS and like to
save ports and NAT/FW traversal costs and where usage of multiple
media types in one RTP session is not suitable, there is a proposal
for how to achieve <xref
target="I-D.westerlund-avtcore-transport-multiplexing">multiplexing
of multiple RTP sessions over the same lower layer transport</xref>.
Using such a solution would allow Multiple RTP session without most
of the perceived downsides of Multiple RTP sessions creating a need
for additional transport flows.</t>
</section>
</section>
<section anchor="sec-security-aspects" title="Security Aspects">
<t>When dealing with point-to-point, 2-member RTP sessions only, there
are few security issues that are relevant to the choice of having one
RTP session or multiple RTP sessions. However, there are a few aspects
of multiparty sessions that might warrant consideration. For general
information of possible methods of securing RTP, please review <xref
target="I-D.ietf-avtcore-rtp-security-options">RTP Security
Options</xref>.</t>
<section title="Security Context Scope">
<t>When using <xref target="RFC3711">SRTP</xref> the security
context scope is important and can be a necessary differentiation in
some applications. As SRTP's crypto suites (so far) is built around
symmetric keys, the receiver will need to have the same key as the
sender. This results in that no one in a multi-party session can be
certain that a received packet really was sent by the claimed sender
or by another party having access to the key. In most cases this is
a sufficient security property, but there are a few cases where this
does create situations.</t>
<t>The first case is when someone leaves a multi-party session and
one wants to ensure that the party that left can no longer access
the media streams. This requires that everyone re-keys without
disclosing the keys to the excluded party.</t>
<t>A second case is when using security as an enforcing mechanism
for differentiation. Take for example a scalable layer or a high
quality simulcast version which only premium users are allowed to
access. The mechanism preventing a receiver from getting the high
quality stream can be based on the stream being encrypted with a key
that user can't access without paying premium, having the
key-management limit access to the key.</t>
<t><xref target="RFC3711">SRTP</xref> has not special functions for
dealing with different sets of master keys for different SSRCs. The
key-management functions has different capabilities to establish
different set of keys, normally on a per end-point basis. <xref
target="RFC5764">DTLS-SRTP</xref> and <xref
target="RFC4568">Security Descriptions</xref> for example establish
different keys for outgoing and incoming traffic from an end-point.
This key usage must be written into the cryptographic context,
possibly associated with different SSRCs.</t>
</section>
<section title="Key Management for Multi-party session">
<t>Performing key-management for multi-party session can be a
challenge. This section considers some of the issues.</t>
<t>Multi-party sessions, such as transport translator based sessions
and multicast sessions, cannot use <xref target="RFC4568">Security
Description</xref> nor <xref target="RFC5764">DTLS-SRTP</xref>
without an extension as each endpoint provides its set of keys. In
centralised conference, the signalling counterpart is a conference
server and the media plane unicast counterpart (to which DTLS
messages would be sent) is the transport translator. Thus an
extension like <xref target="I-D.ietf-avt-srtp-ekt">Encrypted Key
Transport</xref> is needed or a <xref target="RFC3830">MIKEY</xref>
based solution that allows for keying all session participants with
the same master key.</t>
</section>
<section title="Complexity Implications">
<t>The usage of security functions can surface complexity
implications of the choice of multiplexing and topology. This
becomes especially evident in RTP topologies having any type of
middlebox that processes or modifies RTP/RTCP packets. Where there
is very small overhead for an RTP translator or mixer to rewrite an
SSRC value in the RTP packet of an unencrypted session, the cost of
doing it when using cryptographic security functions is higher. For
example if using <xref target="RFC3711">SRTP</xref>, the actual
security context and exact crypto key are determined by the SSRC
field value. If one changes it, the encryption and authentication
tag must be performed using another key. Thus changing the SSRC
value implies a decryption using the old SSRC and its security
context followed by an encryption using the new one.</t>
</section>
</section>
</section>
<section title="Arch-Types">
<t>This section discusses some arch-types of how RTP multiplexing can be
used in applications to achieve certain goals and a summary of their
implications. For each arch-type there is discussion of benefits and
downsides.</t>
<section title="Single SSRC per Session">
<t>In this arch-type each endpoint in a point-to-point session has
only a single SSRC, thus the RTP session contains only two SSRCs, one
local and one remote. This session can be used both unidirectional,
i.e. only a single media stream or bi-directional, i.e. both endpoints
have one media stream each. If the application needs additional media
flows between the endpoints, they will have to establish additional
RTP sessions.</t>
<t>The Pros:<list style="numbers">
<t>This arch-type has great legacy interoperability potential as
it will not tax any RTP stack implementations.</t>
<t>The signalling has good possibilities to negotiate and describe
the exact formats and bit-rates for each media stream, especially
using today's tools in SDP.</t>
<t>It does not matter if usage or purpose of the media stream is
signalled on media stream level or session level as there is no
difference.</t>
<t>It is possible to control security association per RTP session
with current key-management.</t>
</list></t>
<t>The Cons:<list style="letters">
<t>The number of required RTP sessions grows directly in
proportion with the number of media streams, which has the
implications:<list style="symbols">
<t>Linear growth of the amount of NAT/FW state with number of
media streams.</t>
<t>Increased delay and resource consumption from NAT/FW
traversal.</t>
<t>Likely larger signalling message and signalling processing
requirement due to the amount of session related
information.</t>
<t>Higher potential for a single media stream to fail during
transport between the endpoints.</t>
</list></t>
<t>When the number of RTP sessions grows, the amount of explicit
state for relating media stream also grows, linearly or possibly
exponentially, depending on how the application needs to relate
media streams.</t>
<t>The port consumption may become a problem for centralised
services, where the central node's port consumption grows rapidly
with the number of sessions.</t>
<t>For applications where the media streams are highly dynamic in
their usage, i.e. entering and leaving, the amount of signalling
can grow high. Issues arising from the timely establishment of
additional RTP sessions can also arise.</t>
<t>Cross session RTCP requests needs is likely to exist and may
cause issues.</t>
<t>If the same SSRC value is reused in multiple RTP sessions
rather than being randomly chosen, interworking with applications
that uses another multiplexing structure than this application
will have issues and require SSRC translation.</t>
<t>Cannot be used with Any Source Multicast (ASM) as one cannot
guarantee that only two endpoints participate as packet senders.
Using SSM, it is possible to restrict to these requirements if no
RTCP feedback is injected back into the SSM group.</t>
<t>For most security mechanisms, each RTP session or transport
flow requires individual key-management and security association
establishment thus increasing the overhead.</t>
</list></t>
<t>RTP applications that need to inter-work with legacy RTP
applications, like VoIP and video conferencing, can potentially
benefit from this structure. However, a large number of media
descriptions in SDP can also run into issues with existing
implementations. For any application needing a larger number of media
flows, the overhead can become very significant. This structure is
also not suitable for multi-party sessions, as any given media stream
from each participant, although having same usage in the application,
must have its own RTP session. In addition, the dynamic behaviour that
can arise in multi-party applications can tax the signalling system
and make timely media establishment more difficult.</t>
</section>
<section anchor="sec-multiple-ssrc-single-session"
title="Multiple SSRCs of the Same Media Type">
<t>In this arch-type, each RTP session serves only a single media
type. The RTP session can contain multiple media streams, either from
a single endpoint or due to multiple endpoints. This commonly creates
a low number of RTP sessions, typically only two one for audio and one
for video with a corresponding need for two listening ports when using
RTP and RTCP multiplexing.</t>
<t>The Pros:<list style="numbers">
<t>Low number of RTP sessions needed compared to single SSRC case.
This implies:<list style="symbols">
<t>Reduced NAT/FW state</t>
<t>Lower NAT/FW Traversal Cost in both processing and
delay.</t>
</list></t>
<t>Allows for early de-multiplexing in the processing chain in RTP
applications where all media streams of the same type have the
same usage in the application.</t>
<t>Works well with media type de-composite endpoints.</t>
<t>Enables Flow-based QoS with different prioritisation between
media types.</t>
<t>For applications with dynamic usage of media streams, i.e. they
come and go frequently, having much of the state associated with
the RTP session rather than an individual SSRC can avoid the need
for in-session signalling of meta-information about each SSRC.</t>
<t>Low overhead for security association establishment.</t>
</list></t>
<t>The Cons:<list style="letters">
<t>May have some need for cross session RTCP requests for things
that affect both media types in an asynchronous way.</t>
<t>Some potential for concern with legacy implementations that
does not support the RTP specification fully when it comes to
handling multiple SSRC per endpoint.</t>
<t>Will not be able to control security association for sets of
media streams within the same media type with today's
key-management mechanisms, only between SDP media
descriptions.</t>
</list></t>
<t>For RTP applications where all media streams of the same media type
share same usage, this structure provides efficiency gains in amount
of network state used and provides more faith sharing with other media
flows of the same type. At the same time, it is still maintaining
almost all functionalities when it comes to negotiation in the
signalling of the properties for the individual media type and also
enabling flow based QoS prioritisation between media types. It handles
multi-party session well, independently of multicast or centralised
transport distribution, as additional sources can dynamically enter
and leave the session.</t>
</section>
<section title="Multiple Sessions for one Media type">
<t>In this arch-type one goes one step further than in the <xref
target="sec-multiple-ssrc-single-session">above</xref> by using
multiple RTP sessions also for a single media type. The main reason
for going in this direction is that the RTP application needs
separation of the media streams due to their usage. Some typical
reasons for going to this arch-type are scalability over multicast,
simulcast, need for extended QoS prioritisation of media streams due
to their usage in the application, or the need for fine granular
signalling using today's tools.</t>
<t>The Pros:<list style="numbers">
<t>More suitable for Multicast usage where receivers can
individually select which RTP sessions they want to participate
in, assuming each RTP session has its own multicast group.</t>
<t>Detailed indication of the application's usage of the media
stream, where multiple different usages exist.</t>
<t>Less need for SSRC specific explicit signalling for each media
stream and thus reduced need for explicit and timely
signalling.</t>
<t>Enables detailed QoS prioritisation for flow based
mechanisms.</t>
<t>Works well with de-composite endpoints.</t>
<t>Handles dynamic usage of media streams well.</t>
<t>For transport translator based multi-party sessions, this
structure allows for improved control of which type of media
streams an endpoint receives.</t>
<t>The scope for who is included in a security association can be
structured around the different RTP sessions, thus enabling such
functionality with existing key-management.</t>
</list></t>
<t>The Cons:<list style="letters">
<t>Increases the amount of RTP sessions compared to Multiple SSRCs
of the Same Media Type.</t>
<t>Increased amount of session configuration state.</t>
<t>May need synchronised cross-session RTCP requests and require
some consideration due to this.</t>
<t>For media streams that are part of scalability, simulcast or
transport robustness it will be needed to bind sources, which must
support multiple RTP sessions.</t>
<t>Some potential for concern with legacy implementations that
does not support the RTP specification fully when it comes to
handling multiple SSRC per endpoint.</t>
<t>Higher overhead for security association establishment.</t>
<t>If the applications need finer control than on media type level
over which session participants that are included in different
sets of security associations, most of today's key-management will
have difficulties establishing such a session.</t>
</list></t>
<t>For more complex RTP applications that have several different
usages for media streams of the same media type and / or uses
scalability or simulcast, this solution can enable those functions at
the cost of increased overhead associated with the additional
sessions. This type of structure is suitable for more advanced
applications as well as multicast based applications requiring
differentiation to different participants.</t>
</section>
<section title="Multiple Media Types in one Session">
<t>This arch-type is to use a single RTP session for multiple
different media types, like audio and video, and possibly also
transport robustness mechanisms like FEC or Retransmission. Each media
stream will use its own SSRC and a given SSRC value from a particular
endpoint will never use the SSRC for more than a single media
type.</t>
<t>The Pros:<list style="numbers">
<t>Single RTP session which implies:<list style="symbols">
<t>Minimal NAT/FW state.</t>
<t>Minimal NAT/FW Traversal Cost.</t>
<t>Fate-sharing for all media flows.</t>
</list></t>
<t>Enables separation of the different media types based on the
payload types so media type specific endpoint or central
processing can still be supported despite single session.</t>
<t>Can handle dynamic allocations of media streams well on an RTP
level. Depends on the application's needs for explicit indication
of the stream usage and how timely that can be signalled.</t>
<t>Minimal overhead for security association establishment.</t>
</list></t>
<t>The Cons:<list style="letters">
<t>Less suitable for interworking with other applications that
uses individual RTP sessions per media type or multiple sessions
for a single media type, due to need of SSRC translation.</t>
<t>Negotiation of bandwidth for the different media types is
currently not possible in SDP. This requires SDP extensions to
enable payload or source specific bandwidth. Likely to be a
problem due to media type asymmetry in required bandwidth.</t>
<t>Not suitable for de-composite end-points as it requires higher
bandwidth and processing.</t>
<t>Flow based QoS cannot provide separate treatment to some media
streams compared to other in the single RTP session.</t>
<t>If there is significant asymmetry between the media streams
RTCP reporting needs, there are some challenges in configuration
and usage to avoid wasting RTCP reporting on the media stream that
does not need that frequent reporting.</t>
<t>Not suitable for applications where some receivers like to
receive only a subset of the media streams, especially if
multicast or transport translator is being used.</t>
<t>Additional concern with legacy implementations that does not
support the RTP specification fully when it comes to handling
multiple SSRC per endpoint, as also multiple simultaneous media
types needs to be handled.</t>
<t>If the applications need finer control over which session
participants that are included in different sets of security
associations, most key-management will have difficulties
establishing such a session.</t>
</list></t>
<t/>
<!--MW: Add new summary for this arch-type.
MW: One arch-type that we should consider adding is the one with multiple RTP sessions over
a single transport.-->
</section>
<section title="Summary">
<t>There are some clear relations between these arch-types. Both the
"single SSRC per RTP session" and the "multiple media types in one
session" are cases which require full explicit signalling of the media
stream relations. However, they operate on two different levels where
the first primarily enables session level binding, and the second
needs to do it all on SSRC level. From another perspective, the two
solutions are the two extreme points when it comes to number of RTP
sessions required.</t>
<t>The two other arch-types "Multiple SSRCs of the Same Media Type"
and "Multiple Sessions for one Media Type" are examples of two other
cases that first of all allows for some implicit mapping of the role
or usage of the media streams based on which RTP session they appear
in. It thus potentially allows for less signalling and in particular
reduced need for real-time signalling in dynamic sessions. They also
represent points in between the first two when it comes to amount of
RTP sessions established, i.e. representing an attempt to reduce the
amount of sessions as much as possible without compromising the
functionality the session provides both on network level and on
signalling level.</t>
</section>
</section>
<section title="Summary considerations and guidelines">
<t/>
<section title="Guidelines">
<t>This section contains a number of recommendations for implementors
or specification writers when it comes to handling multi-stream.<list
style="hanging">
<t hangText="Do not Require the same SSRC across Sessions:">As
discussed in <xref target="sec-binding-related"/> there exist
drawbacks in using the same SSRC in multiple RTP sessions as a
mechanism to bind related media streams together. It is instead
recommended that a mechanism to explicitly signal the relation is
used, either in RTP/RTCP or in the used signalling mechanism that
establishes the RTP session(s).</t>
<t hangText="Use additional SSRCs additional Media Sources:">In
the cases an RTP endpoint needs to transmit additional media
streams of the same media type in the application, with the same
processing requirements at the network and RTP layers, it is
recommended to send them as additional SSRCs in the same RTP
session. For example a telepresence room where there are three
cameras, and each camera captures 2 persons sitting at the table,
sending each camera as its own SSRC within a single RTP session is
recommended.</t>
<t
hangText="Use additional RTP sessions for streams with different requirements:">When
media streams have different processing requirements from the
network or the RTP layer at the endpoints, it is recommended that
the different types of streams are put in different RTP
sessions.<vspace blankLines="0"/>This includes the case where
different participants want different subsets of the set of RTP
streams.</t>
<t hangText="When using multiple RTP Sessions use grouping:">When
using Multiple RTP session solutions, it is recommended to be
explicitly group the involved RTP sessions when needed using the
signalling mechanism, for example <xref target="RFC5888">The
Session Description Protocol (SDP) Grouping Framework.</xref>,
using some appropriate grouping semantics.</t>
<t
hangText="RTP/RTCP Extensions May Support Additional SSRCs as well as Multiple RTP sessions:">When
defining an RTP or RTCP extension, the creator needs to consider
if this extension is applicable to usage with additional SSRCs and
Multiple RTP sessions. Any extension intended to be generic is
recommended to support both. Applications that are not as
generally applicable will have to consider if interoperability is
better served by defining a single solution or providing both
options.</t>
<t hangText="Transport Support Extensions:">When defining new
RTP/RTCP extensions intended for transport support, like the
retransmission or FEC mechanisms, they are recommended to include
support for both additional SSRCs and multiple RTP sessions so
that application developers can choose freely from the set of
mechanisms without concerning themselves with which of the
multiplexing choices a particular solution supports.</t>
</list></t>
<t/>
</section>
</section>
<section anchor="IANA" title="IANA Considerations">
<t>This document makes no request of IANA.</t>
<t>Note to RFC Editor: this section may be removed on publication as an
RFC.</t>
</section>
<section anchor="Security" title="Security Considerations">
<t>There is discussion of the security implications of choosing SSRC vs
Multiple RTP session in <xref target="sec-security-aspects"/>.</t>
</section>
</middle>
<back>
<references title="Normative References">
&rfc3550;
</references>
<references title="Informative References">
&rfc2198;
&rfc2205;
&rfc2326;
&rfc2474;
&rfc2974;
&rfc3261;
&rfc3264;
&rfc3389;
&rfc3551;
&rfc3711;
&rfc3830;
&rfc4103;
&rfc4566;
&rfc4568;
&rfc4588;
&rfc4607;
&rfc5104;
&rfc5117;
&rfc5583;
&rfc5576;
&rfc5760;
&rfc5761;
&rfc5764;
&rfc5888;
&rfc6190;
&rfc6285;
&rfc6465;
&rfc6222;
&draft-ietf-avtext-multiple-clock-rates;
&draft-ietf-payload-rtp-howto;
&draft-ietf-avt-srtp-ekt;
<reference anchor="ALF">
<front>
<title>Architectural Considerations for a New Generation of
Protocols</title>
<author initials="D." surname="Clark">
<organization>IEEE Computer Communications Review, Vol.
20(4)</organization>
</author>
<author initials="D." surname="Tennenhouse">
<organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<facsimile/>
<email/>
<uri/>
</address>
</author>
<date month="September" year="1990"/>
</front>
<seriesInfo name="SIGCOMM Symposium on Communications Architectures and Protocols"
value="(Philadelphia, Pennsylvania), pp. 200--208, IEEE Computer Communications Review, Vol. 20(4)"/>
</reference>
<?rfc include='reference.I-D.ietf-mmusic-sdp-bundle-negotiation'?>
<?rfc include='reference.I-D.alvestrand-rtp-sess-neutral'?>
<?rfc include='reference.I-D.lennox-mmusic-sdp-source-selection'?>
<?rfc include='reference.I-D.westerlund-avtext-rtcp-sdes-srcname'?>
<?rfc include='reference.I-D.westerlund-avtcore-max-ssrc'?>
<?rfc include='reference.I-D.westerlund-avtcore-transport-multiplexing'?>
<?rfc include='reference.I-D.ietf-avtcore-rtp-security-options'?>
<?rfc include='reference.I-D.ietf-avtcore-multi-media-rtp-session'?>
<?rfc include='reference.I-D.lennox-avtcore-rtp-multi-stream'?>
<?rfc include='reference.I-D.westerlund-avtcore-rtp-topologies-update'?>
<?rfc include='reference.I-D.ietf-avtcore-6222bis'?>
<?rfc include='reference.RFC.5109'?>
</references>
<section anchor="sec-pt-mux" title="Dismissing Payload Type Multiplexing">
<t>This section documents a number of reasons why using the payload type
as a multiplexing point for most things related to multiple streams is
unsuitable. If one attempts to use Payload type multiplexing beyond it's
defined usage, that has well known negative effects on RTP. To use
Payload type as the single discriminator for multiple streams implies
that all the different media streams are being sent with the same SSRC,
thus using the same timestamp and sequence number space. This has many
effects:</t>
<t><list style="numbers">
<t>Putting restraint on RTP timestamp rate for the multiplexed
media. For example, media streams that use different RTP timestamp
rates cannot be combined, as the timestamp values need to be
consistent across all multiplexed media frames. Thus streams are
forced to use the same rate. When this is not possible, Payload Type
multiplexing cannot be used.</t>
<t>Many RTP payload formats may fragment a media object over
multiple packets, like parts of a video frame. These payload formats
need to determine the order of the fragments to correctly decode
them. Thus it is important to ensure that all fragments related to a
frame or a similar media object are transmitted in sequence and
without interruptions within the object. This can relatively simple
be solved on the sender side by ensuring that the fragments of each
media stream are sent in sequence.</t>
<t>Some media formats require uninterrupted sequence number space
between media parts. These are media formats where any missing RTP
sequence number will result in decoding failure or invoking of a
repair mechanism within a single media context. The <xref
target="RFC4103">text/T140 payload format</xref> is an example of
such a format. These formats will need a sequence numbering
abstraction function between RTP and the individual media stream
before being used with Payload Type multiplexing.</t>
<t>Sending multiple streams in the same sequence number space makes
it impossible to determine which Payload Type and thus which stream
a packet loss relates to.</t>
<t>If <xref target="RFC4588">RTP Retransmission</xref> is used and
there is a loss, it is possible to ask for the missing packet(s) by
SSRC and sequence number, not by Payload Type. If only some of the
Payload Type multiplexed streams are of interest, there is no way of
telling which missing packet(s) belong to the interesting stream(s)
and all lost packets must be requested, wasting bandwidth.</t>
<t>The current RTCP feedback mechanisms are built around providing
feedback on media streams based on stream ID (SSRC), packet
(sequence numbers) and time interval (RTP Timestamps). There is
almost never a field to indicate which Payload Type is reported, so
sending feedback for a specific media stream is difficult without
extending existing RTCP reporting.</t>
<t>The current <xref target="RFC5104">RTCP media control
messages</xref> specification is oriented around controlling
particular media flows, i.e. requests are done addressing a
particular SSRC. Such mechanisms would need to be redefined to
support Payload Type multiplexing.</t>
<t>The number of payload types are inherently limited. Accordingly,
using Payload Type multiplexing limits the number of streams that
can be multiplexed and does not scale. This limitation is
exacerbated if one uses solutions like <xref target="RFC5761">RTP
and RTCP multiplexing</xref> where a number of payload types are
blocked due to the overlap between RTP and RTCP.</t>
<t>At times, there is a need to group multiplexed streams and this
is currently possible for RTP Sessions and for SSRC, but there is no
defined way to group Payload Types.</t>
<t>It is currently not possible to signal bandwidth requirements per
media stream when using Payload Type Multiplexing.</t>
<t>Most existing SDP media level attributes cannot be applied on a
per Payload Type level and would require re-definition in that
context.</t>
<t>A legacy endpoint that doesn't understand the indication that
different RTP payload types are different media streams may be
slightly confused by the large amount of possibly overlapping or
identically defined RTP Payload Types.</t>
</list></t>
</section>
<section title="Proposals for Future Work">
<t>The above discussion and guidelines indicates that a small set of
extension mechanisms could greatly improve the situation when it comes
to using multiple streams independently of Multiple RTP session or
Additional SSRC. These extensions are:<list style="hanging">
<t hangText="Media Source Identification:">A Media source
identification that can be used to bind together media streams that
are related to the same media source. A <xref
target="I-D.westerlund-avtext-rtcp-sdes-srcname">proposal</xref>
exist for a new SDES item SRCNAME that also can be used with the
a=ssrc SDP attribute to provide signalling layer binding
information.</t>
<t hangText="SSRC limitations within RTP sessions:">By providing a
signalling solution that allows the signalling peers to explicitly
express both support and limitations on how many simultaneous media
streams an endpoint can handle within a given RTP Session. That
ensures that usage of Additional SSRC occurs when supported and
without overloading an endpoint. This extension is proposed in <xref
target="I-D.westerlund-avtcore-max-ssrc"/>.</t>
</list></t>
</section>
<section anchor="sec-rtp-clarifications"
title="RTP Specification Clarifications">
<t>This section describes a number of clarifications to the RTP
specifications that are likely necessary for aligned behaviour when RTP
sessions contain more SSRCs than one local and one remote.</t>
<t>All of the below proposals are under consideration in <xref
target="I-D.lennox-avtcore-rtp-multi-stream"/>.</t>
<section title="RTCP Reporting from all SSRCs">
<t>When one has multiple SSRC in an RTP node, all these SSRC must send
some RTP or RTCP packet as long as the SSRC exist. It is not
sufficient that only one SSRC in the node sends report blocks on the
incoming RTP streams; any SSRC that intends to remain in the session
must send some packets to avoid timing out according to the rules in
RFC 3550 section 6.3.5.</t>
<t>It has been hypothesised that a third party monitor may be confused
by not necessarily being able to determine that all these SSRC are in
fact co-located and originate from the same stack instance; if this
hypothesis is true, this may argue for having all the sources send
full reception reports, even though they are reporting the same packet
delivery.</t>
<t>The contrary argument is that such double reporting may confuse the
third party monitor even more by making it seem that utilisation of
the last-hop link to the recipient is (number of SSRCs) times higher
than what it actually is.</t>
</section>
<section title="RTCP Self-reporting">
<t>For any RTP node that sends more than one SSRC, there is the
question if SSRC1 needs to report its reception of SSRC2 and vice
versa. The reason that they in fact need to report on all other local
streams as being received is report consistency. The hypothetical
third party monitor that considers the full matrix of media streams
and all known SSRC reports on these media streams would detect a gap
in the reports which could be a transport issue unless identified as
in fact being sources from the same node.</t>
<!--MW: Moved the proposal out of the text. I am not certain this is fly and
don't have time to consider it well enough.
Our proposal is that RFC3550 is updated to clarify that one needs to report
on all SSRCs one knows exist with the sole exception of the local SSRCs
that has the same CNAME as the SSRC providing the report. That way a third
party monitor can use the CNAME data from the various SSRCs to determine
that the gap in reporting is not valid.
Note: There is nothing preventing a node to send an SSRC with the same
CNAME as one or more other SSRCs originating from another node. In fact
an obvious case for this to occur is when the creation of Forward Error
Correction data is performed at a boundary to another transport domain. Thus
any node in this case would need to report on both the actual arrived
stream and send sender reports on the stream it creates.
The result of the above exception is that a 3rd party monitor can't
detect if there is an fault in the transport from the original source and
the secondary node generating the new source with shared CNAME.-->
</section>
<section title="Combined RTCP Packets">
<t>When a node contains multiple SSRCs, it is questionable if an RTCP
compound packet can only contain RTCP packets from a single SSRC or if
multiple SSRCs can include their packets in a joint compound packet.
The high level question is a matter for any receiver processing on
what to expect. In addition to that question there is the issue of how
to use the RTCP timer rules in these cases, as the existing rules are
focused on determining when a single SSRC can send.</t>
</section>
</section>
<section title="Signalling considerations">
<t>Signalling is not an architectural consideration for RTP itself, so
this discussion has been moved to an appendix. However, it is hugely
important for anyone building complete applications, so it is deserving
of discussion.</t>
<t>The issues raised here need to be addressed in the WGs that deal with
signalling; they cannot be addressed by tweaking, extending or profiling
RTP.</t>
<section title="Signalling Aspects">
<t>There exist various signalling solutions for establishing RTP
sessions. Many are <xref target="RFC4566">SDP</xref> based, however
SDP functionality is also dependent on the signalling protocols
carrying the SDP. Where <xref target="RFC2326">RTSP</xref> and <xref
target="RFC2974">SAP</xref> both use SDP in a declarative fashion,
while <xref target="RFC3261">SIP</xref> uses SDP with the additional
definition of <xref target="RFC3264">Offer/Answer</xref>. The impact
on signalling and especially SDP needs to be considered as it can
greatly affect how to deploy a certain multiplexing point choice.</t>
<section title="Session Oriented Properties">
<t>One aspect of the existing signalling is that it is focused
around sessions, or at least in the case of SDP the media
description. There are a number of things that are signalled on a
session level/media description but those are not necessarily
strictly bound to an RTP session and could be of interest to signal
specifically for a particular media stream (SSRC) within the
session. The following properties have been identified as being
potentially useful to signal not only on RTP session level:<list
style="symbols">
<t>Bitrate/Bandwidth exist today only at aggregate or a common
any media stream limit, unless either codec-specific bandwidth
limiting or RTCP signalling using TMMBR is used.</t>
<t>Which SSRC that will use which RTP Payload Types (this will
be visible from the first media packet, but is sometimes useful
to know before packet arrival).</t>
</list></t>
<t>Some of these issues are clearly SDP's problem rather than RTP
limitations. However, if the aim is to deploy an solution using
additional SSRCs that contains several sets of media streams with
different properties (encoding/packetization parameter, bit-rate,
etc), putting each set in a different RTP session would directly
enable negotiation of the parameters for each set. If insisting on
Additional SSRC only, a number of signalling extensions are needed
to clarify that there are multiple sets of media streams with
different properties and that they shall in fact be kept different,
since a single set will not satisfy the application's
requirements.</t>
<t>For some parameters, such as resolution and framerate, a
SSRC-linked mechanism has been proposed: <xref
target="I-D.lennox-mmusic-sdp-source-selection"/>.</t>
</section>
<section title="SDP Prevents Multiple Media Types">
<t>SDP chose to use the m= line both to delineate an RTP session and
to specify the top level of the MIME media type; audio, video, text,
image, application. This media type is used as the top-level media
type for identifying the actual payload format bound to a particular
payload type using the rtpmap attribute. This binding has to be
loosened in order to use SDP to describe RTP sessions containing
multiple MIME top level types.</t>
<t>There is an accepted WG item in the MMUSIC WG to define how <xref
target="I-D.ietf-mmusic-sdp-bundle-negotiation">multiple media lines
describe a single underlying transport</xref> and thus it becomes
possible in SDP to define one RTP session with media types having
different MIME top level types.</t>
</section>
<section title="Signalling Media Stream Usage">
<t>Media streams being transported in RTP has some particular usage
in an RTP application. This usage of the media stream is in many
applications so far implicitly signalled. For example, an
application may choose to take all incoming audio RTP streams, mix
them and play them out. However, in more advanced applications that
use multiple media streams there will be more than a single usage or
purpose among the set of media streams being sent or received. RTP
applications will need to signal this usage somehow. The signalling
used will have to identify the media streams affected by their
RTP-level identifiers, which means that they have to be identified
either by their session or by their SSRC + session.</t>
<t>In some applications, the receiver cannot utilise the media
stream at all before it has received the signalling message
describing the media stream and its usage. In other applications,
there exists a default handling that is appropriate.</t>
<t>If all media streams in an RTP session are to be treated in the
same way, identifying the session is enough. If SSRCs in a session
are to be treated differently, signalling must identify both the
session and the SSRC.</t>
<t>If this signalling affects how any RTP central node, like an RTP
mixer or translator that selects, mixes or processes streams, treats
the streams, the node will also need to receive the same signalling
to know how to treat media streams with different usage in the right
fashion.</t>
</section>
</section>
</section>
<section title="Changes from -01 to -02">
<t><list style="symbols">
<t>Added Harald Alvestrand as co-author.</t>
<t>Removed unused term "Media aggregate".</t>
<t>Added term "RTP session group", noted that CNAMEs are assumed to
bind across the sessions of an RTP session group, and used it when
appropriate (TODO)</t>
<t>Moved discussion of signalling aspects to appendix</t>
<t>Removed all suggestion that PT can be a multiplexing point</t>
<t>Normalised spelling of "endpoint" to follow RFC 3550 and not use
a hyphen.</t>
<t>Added CNAME to definition list.</t>
<t>Added term "Media Sink" for the thing that is identified by a
listen-only SSRC.</t>
<t>Added term "RTP source" for the thing that transmits one media
stream, separating it from "Media Source". [[OUTSTANDING: Whether to
use "RTP Source" or "Media Sender" here]]</t>
<t>Rewrote section on distributed endpoint, noting that this, like
any endpoint that wants a subset of a set of RTP streams, needs
multiple RTP sessions.</t>
<t>Removed all substantive references to the undefined term
"purpose" from the main body of the document when it referred to the
purpose of an RTP stream.</t>
<t>Moved the summary section of section 6 to the guidelines section
that it most closely supports.</t>
<t/>
</list></t>
</section>
</back>
</rfc>
| PAFTECH AB 2003-2026 | 2026-04-23 14:21:25 |