rfc8836xml2.original.xml   rfc8836.xml 
<?xml version="1.0" encoding="US-ASCII"?> <?xml version='1.0' encoding='utf-8'?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd"> <!DOCTYPE rfc SYSTEM "rfc2629-xhtml.ent">
<?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-ietf-rmcat-cc-requirements-09"
ipr="trust200902">
<front>
<title abbrev="RTP Media Congestion Control Requirements ">Congestion
Control Requirements for Interactive Real-Time Media</title>
<rfc xmlns:xi="http://www.w3.org/2001/XInclude" number="8836" category="info"
submissionType="IETF" consensus="true" ipr="trust200902" obsoletes=""
updates="" xml:lang="en" tocInclude="true" symRefs="true" sortRefs="true"
version="3" docName="draft-ietf-rmcat-cc-requirements-09">
<!-- xml2rfc v2v3 conversion 2.34.0 -->
<front>
<title abbrev="RTP Media Congestion Control
Requirements">Congestion Control Requirements for
Interactive Real-Time Media</title>
<seriesInfo name="RFC" value="8836"/>
<author fullname="Randell Jesup" initials="R." surname="Jesup"> <author fullname="Randell Jesup" initials="R." surname="Jesup">
<organization>Mozilla</organization> <organization>Mozilla</organization>
<address> <address>
<postal> <postal>
<street></street> <street/>
<country>USA</country> <country>USA</country>
</postal> </postal>
<email>randell-ietf@jesup.org</email> <email>randell-ietf@jesup.org</email>
</address> </address>
</author> </author>
<author fullname="Zaheduzzaman Sarker" initials="Z." role="editor" surname="
<author fullname="Zaheduzzaman Sarker" initials="Z." role="editor" Sarker">
surname="Sarker">
<organization>Ericsson</organization> <organization>Ericsson</organization>
<address> <address>
<postal> <postal>
<street></street> <street/>
<city/>
<city></city> <region/>
<code/>
<region></region>
<code></code>
<country>Sweden</country> <country>Sweden</country>
</postal> </postal>
<phone/>
<phone></phone>
<facsimile></facsimile>
<email>zaheduzzaman.sarker@ericsson.com</email> <email>zaheduzzaman.sarker@ericsson.com</email>
<uri/>
<uri></uri>
</address> </address>
</author> </author>
<date month="June" year="2020"/>
<date /> <!-- [rfced] Please insert any keywords (beyond those that appear in
the title) for use on https://www.rfc-editor.org/search.
-->
<keyword>example</keyword>
<abstract> <abstract>
<t>Congestion control is needed for all data transported across the <t>Congestion control is needed for all data transported across the
Internet, in order to promote fair usage and prevent congestion Internet, in order to promote fair usage and prevent congestion
collapse. The requirements for interactive, point-to-point real-time collapse. The requirements for interactive, point-to-point real-time
multimedia, which needs low-delay, semi-reliable data delivery, are multimedia, which needs low-delay, semi-reliable data delivery, are
different from the requirements for bulk transfer like FTP or bursty different from the requirements for bulk transfer like FTP or bursty
transfers like Web pages. Due to an increasing amount of RTP-based transfers like web pages. Due to an increasing amount of RTP-based
real-time media traffic on the Internet (e.g. with the introduction of real-time media traffic on the Internet (e.g., with the introduction of
the Web Real-Time Communication (WebRTC)), it is especially important to the Web Real-Time Communication (WebRTC)), it is especially important to
ensure that this kind of traffic is congestion controlled.</t> ensure that this kind of traffic is congestion controlled.</t>
<t>This document describes a set of requirements that can be used to <t>This document describes a set of requirements that can be used to
evaluate other congestion control mechanisms in order to figure out evaluate other congestion control mechanisms in order to figure out
their fitness for this purpose, and in particular to provide a set of their fitness for this purpose, and in particular to provide a set of
possible requirements for real-time media congestion avoidance possible requirements for a real-time media congestion avoidance
technique.</t> technique.</t>
</abstract> </abstract>
<note title="Requirements Language"> <!-- [rfced] Note that the "Requirements Language" has been removed from the
<t>The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", document because the capitalized keywords are not used.
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this -->
document are to be interpreted as described in <xref
target="RFC2119">RFC 2119</xref>. The terms are presented in many cases
using lowercase for readability.</t>
</note>
</front>
</front>
<middle> <middle>
<section title="Introduction"> <section numbered="true" toc="default">
<name>Introduction</name>
<t>Most of today's TCP congestion control schemes were developed with a <t>Most of today's TCP congestion control schemes were developed with a
focus on an use of the Internet for reliable bulk transfer of focus on a use of the Internet for reliable bulk transfer of
non-time-critical data, such as transfer of large files. They have also non-time-critical data, such as transfer of large files. They have also
been used successfully to govern the reliable transfer of smaller chunks been used successfully to govern the reliable transfer of smaller chunks
of data in as short a time as possible, such as when fetching Web of data in as short a time as possible, such as when fetching web
pages.</t> pages.</t>
<t>These algorithms have also been used for transfer of media streams <t>These algorithms have also been used for transfer of media streams
that are viewed in a non-interactive manner, such as "streaming" video, that are viewed in a non-interactive manner, such as "streaming" video,
where having the data ready when the viewer wants it is important, but where having the data ready when the viewer wants it is important, but
the exact timing of the delivery is not.</t> the exact timing of the delivery is not.</t>
<t>When doing real-time interactive media, the requirements are <t>When handling real-time interactive media, the requirements are
different; one needs to provide the data continuously, within a very different. One needs to provide the data continuously, within a very
limited time window (no more than 100s of milliseconds end-to-end limited time window (no more delay than hundreds of milliseconds
delay), the sources of data may be able to adapt the amount of data that end-to-end). In addition, the sources of data may be able to adapt the
needs sending within fairly wide margins but can be rate limited by the amount of data that needs sending within fairly wide margins, but they can
application- even not always have data to send, and may tolerate some be rate limited by the
amount of packet loss, but since the data is generated in real-time, application -- even not always having data to send. They may tolerate some
amount of packet loss, but since the data is generated in real time,
sending "future" data is impossible, and since it's consumed in sending "future" data is impossible, and since it's consumed in
real-time, data delivered late is commonly useless.</t> real time, data delivered late is commonly useless.</t>
<t>While the requirements for real-time interactive media differ from <t>While the requirements for real-time interactive media differ from
the requirements for the other flow types, these other flow types will the requirements for the other flow types, these other flow types will
be present in the network. The congestion control algorithm for be present in the network. The congestion control algorithm for
real-time interactive media must work properly when these other flow real-time interactive media must work properly when these other flow
types are present as cross traffic on the network.</t> types are present as cross traffic on the network.</t>
<t>One particular protocol portfolio being developed for this use case <t>One particular protocol portfolio being developed for this use case
is WebRTC <xref target="I-D.ietf-rtcweb-overview"></xref>, where one is WebRTC <xref target="RFC8825" format="default"/>, where one
envisions sending multiple flows using the Real-time Transport Protocol envisions sending multiple flows using the Real-time Transport Protocol
(RTP) <xref target="RFC3550"></xref> between two peers, in conjunction (RTP) <xref target="RFC3550" format="default"/> between two peers, in conj unction
with data flows, all at the same time, without having special with data flows, all at the same time, without having special
arrangements with the intervening service providers. As RTP does not arrangements with the intervening service providers. As RTP does not
provide any congestion control mechanism; a set of circuit breakers, provide any congestion control mechanism, a set of circuit breakers,
such as <xref target="I-D.ietf-avtcore-rtp-circuit-breakers"></xref>, such as those described in <xref target="RFC8083" format="default"/>,
are required to protect the network from excessive congestion caused by are required to protect the network from excessive congestion caused by
the non-congestion controlled flows. When the real-time interactive non-congestion-controlled flows. When the real-time interactive
media is congestion controlled, it is recommended that the congestion media is congestion controlled, it is recommended that the
control mechanism operates within the constraints defined by these congestion control mechanism operate within the constraints defined by
circuit breakers when circuit breaker is present and that it should not these
cause congestion collapse when circuit breaker is not implemented.</t> circuit breakers when a circuit breaker is present and that it should not
cause congestion collapse when a circuit breaker is not implemented.</t>
<t>Given that this use case is the focus of this document, use cases <t>Given that this use case is the focus of this document, use cases
involving non-interactive media such as video streaming, and use cases involving non-interactive media such as video streaming and those
using multicast/broadcast-type technologies, are out of scope.</t> using multicast/broadcast-type technologies, are out of scope.</t>
<t>The terminology defined in <xref target="RFC8825" format="default"/>
<t>The terminology defined in <xref is used in this memo.</t>
target="I-D.ietf-rtcweb-overview"></xref> is used in this memo.</t>
</section> </section>
<section title="Requirements"> <section numbered="true" toc="default">
<t><list style="numbers"> <name>Requirements</name>
<ol spacing="normal" type="1">
<li>
<t>The congestion control algorithm must attempt to provide <t>The congestion control algorithm must attempt to provide
as-low-as-possible-delay transit for interactive real-time traffic as-low-as-possible-delay transit for interactive real-time traffic
while still providing a useful amount of bandwidth. There may be while still providing a useful amount of bandwidth. There may be
lower limits on the amount of bandwidth that is useful, but this is lower limits on the amount of bandwidth that is useful, but this is
largely application-specific and the application may be able to largely application specific, and the application may be able to
modify or remove flows in order allow some useful flows to get modify or remove flows in order to allow some useful flows to get
enough bandwidth. (Example: not enough bandwidth for low-latency enough bandwidth.
video+audio, but enough for audio-only.) <list style="letters"> <!-- [rfced] We have updated the text as shown below. Please let us know if
<t>Jitter (variation in the bitrate over short time scales) also any corrections are needed.
is relevant, though moderate amounts of jitter will be absorbed
Original:
(Example: not enough bandwidth for
low-latency video+audio, but enough for audio-only.)
Current:
For example, although there might not be enough bandwidth
for low-latency video+audio, there could be enough for audio only.
-->
For example, although there might not be enough bandwidth
for low-latency video+audio, there could be enough for audio only.
</t>
<ol spacing="normal" type="a">
<li>Jitter (variation in the bitrate over short timescales) is also
relevant, though moderate amounts of jitter will be absorbed
by jitter buffers. Transit delay should be considered to track by jitter buffers. Transit delay should be considered to track
the short-term maximums of delay including jitter.</t>
<t>It should provide this as-low-as-possible-delay transit and the short-term maximums of delay, including jitter.</li>
<li>The algorithm should provide this as-low-as-possible-delay trans
it and
minimize self-induced latency even when faced with intermediate minimize self-induced latency even when faced with intermediate
bottlenecks and competing flows. Competing flows may limit bottlenecks and competing flows. Competing flows may limit
what's possible to achieve.</t> what's possible to achieve.</li>
<li>The algorithm should be resilient to the effects of events, such
<t>It should be resilience to the effects of the events, such as as
routing changes, which may alter or remove bottlenecks or change routing changes, which may alter or remove bottlenecks or change
the bandwidth available especially if there is a reduction in the bandwidth available, especially if there is a reduction in
available bandwidth or increase in observed delay. It is available bandwidth or increase in observed delay. It is
expected that the mechanism reacts quickly to the such events to expected that the mechanism reacts quickly to such events to
avoid delay buildup. In the context of this memo, a 'quick' avoid delay buildup. In the context of this memo, a "quick"
reaction is on the order of a few RTTs, subject to the reaction is on the order of a few RTTs, subject to the
constraints of the media codec, but is likely within a second. constraints of the media codec, but is likely within a second.
<!-- [rfced] Please clarify "but equally". Does this mean, "but at the same
time"?
Original:
Reaction on the next RTT is
explicitly not required, since many codecs cannot adapt
their sending rate that quickly, but equally response cannot
be arbitrarily delayed.
-->
Reaction on the next RTT is explicitly not required, since many Reaction on the next RTT is explicitly not required, since many
codecs cannot adapt their sending rate that quickly, but equally codecs cannot adapt their sending rate that quickly, but equally
response cannot be arbitrarily delayed.</t> a response cannot be arbitrarily delayed.</li>
<li>The algorithm should react quickly to handle both local and remo
<t>It should react quickly to handle both local and remote te
interface changes (WLAN to 3G data, etc) which may radically interface changes (e.g., WLAN to 3G data) that may radically
change the bandwidth available or bottlenecks, especially if change the bandwidth available or bottlenecks, especially if
there is a reduction in available bandwidth or increase in there is a reduction in available bandwidth or an increase in
bottleneck delay. It is assumed that an interface change can bottleneck delay. It is assumed that an interface change can
generate a notification to the algorithm.</t> generate a notification to the algorithm.</li>
<li>The real-time interactive media applications can be rate
<t>The real-time interactive media applications can be rate
limited. This means the offered loads can be less than the limited. This means the offered loads can be less than the
available bandwidth at any given moment, and may vary available bandwidth at any given moment and may vary
dramatically over time, including dropping to no load and then dramatically over time, including dropping to no load and then
resuming a high load, such as in a mute/unmute operation. Hence, resuming a high load, such as in a mute/unmute operation. Hence,
the algorithm must be designed to handle such behavior from the algorithm must be designed to handle such behavior from
media source or application. Note that the reaction time between a media source or application. Note that the reaction time between
a change in the bandwidth available from the algorithm and a a change in the bandwidth available from the algorithm and a
change in the offered load is variable, and may be different change in the offered load is variable, and it may be different
when increasing versus decreasing.</t> when increasing versus decreasing.</li>
<li>The algorithm is required to avoid building up queues when
<t>The algorithm requires to avoid building up queues when
competing with short-term bursts of traffic (for example, competing with short-term bursts of traffic (for example,
traffic generated by web-browsing) which can quickly saturate a traffic generated by web browsing), which can quickly saturate a
local-bottleneck router or link, but also clear quickly. The local-bottleneck router or link but clear quickly. The
algorithm should also react quickly to regain its previous share algorithm should also react quickly to regain its previous share
of the bandwidth when the local-bottleneck or link is of the bandwidth when the local bottleneck or link is
cleared.</t> cleared.</li>
<li>Similarly, periodic bursty flows such as MPEG DASH <xref
<t>Similarly periodic bursty flows such as MPEG DASH <xref target="MPEG_DASH" format="default"/> or proprietary media
target="MPEG_DASH"></xref> or proprietary media streaming streaming
algorithms may compete in bursts with the algorithm, and may not algorithms may compete in bursts with the algorithm and may not
be adaptive within a burst. They are often layered on top of TCP be adaptive within a burst. They are often layered on top of TCP
but use TCP in a bursty manner that can interact poorly with but use TCP in a bursty manner that can interact poorly with
competing flows during the bursts. The algorithm must not competing flows during the bursts. The algorithm must not
increase the already existing delay buildup during those bursts. increase the already existing delay buildup during those bursts.
Note that this competing traffic may be on a shared access link, Note that this competing traffic may be on a shared access link,
or the traffic burst may cause a shift in the location of the or the traffic burst may cause a shift in the location of the
bottleneck for the duration of the burst.</t> bottleneck for the duration of the burst.</li>
</list></t> </ol>
</li>
<li>
<t>The algorithm must be fair to other flows, both real-time flows <t>The algorithm must be fair to other flows, both real-time flows
(such as other instances of itself), and TCP flows, both long-lived (such as other instances of itself) and TCP flows, both long-lived flo
and bursts such as the traffic generated by a typical web browsing ws
session. Note that 'fair' is a rather hard-to-define term. It should and bursts such as the traffic generated by a typical web-browsing
be fair with itself, giving fair share of the bandwidth to multiple session. Note that "fair" is a rather hard-to-define term. It should
be fair with itself, giving a fair share of the bandwidth to multiple
flows with similar RTTs, and if possible to multiple flows with flows with similar RTTs, and if possible to multiple flows with
different RTTs.<list style="letters"> different RTTs.
<t>Existing flows at a bottleneck must also be fair to new flows </t>
to that bottleneck, and must allow new flows to ramp up to a <ol spacing="normal" type="a">
<li>Existing flows at a bottleneck must also be fair to new flows
to that bottleneck and must allow new flows to ramp up to a
useful share of the bottleneck bandwidth as quickly as possible. useful share of the bottleneck bandwidth as quickly as possible.
A useful share will depend on the media types involved, total A useful share will depend on the media types involved, total
bandwidth available and the user experience requirements of a bandwidth available, and the user-experience requirements of a
particular service. Note that relative RTTs may affect the rate particular service. Note that relative RTTs may affect the rate
new flows can ramp up to a reasonable share.</t> at which new flows can ramp up to a reasonable share.</li>
</list></t> </ol>
</li>
<t>The algorithm should not starve competing TCP flows, and should <li>
as best as possible avoid starvation by TCP flows.<list <t>The algorithm should not starve competing TCP flows and should,
style="letters"> as best as possible, avoid starvation by TCP flows.</t>
<t>The congestion control should prioritise achieving a useful <ol spacing="normal" type="a">
<li>The congestion control should prioritize achieving a useful
share of the bandwidth depending on the media types and total share of the bandwidth depending on the media types and total
available bandwidth over achieving as low as possible transit available bandwidth over achieving as-low-as-possible transit
delay, when these two requirements are in conflict.</t> delay, when these two requirements are in conflict.</li>
</list></t> </ol>
</li>
<t>The algorithm should as quickly as possible adapt to initial <li>
network conditions at the start of a flow. This should occur both if <t>The algorithm should adapt as quickly as possible to initial
network conditions at the start of a flow. This should occur whether
the initial bandwidth is above or below the bottleneck bandwidth. the initial bandwidth is above or below the bottleneck bandwidth.
<list style="letters"> </t>
<t>The algorithm should allow different modes of adaptation for <ol spacing="normal" type="a">
example,the startup adaptation may be faster than adaptation <li>The algorithm should allow different modes of adaptation; for
example, the startup adaptation may be faster than adaptation
later in a flow. It should allow for both slow-start operation later in a flow. It should allow for both slow-start operation
(adapt up) and history-based startup (start at a point expected (adapt up) and history-based startup (start at a point expected
to be at or below channel bandwidth from historical information, to be at or below channel bandwidth from historical information,
which may need to adapt down quickly if the initial guess is which may need to adapt down quickly if the initial guess is
wrong). Starting too low and/or adapting up too slowly can cause wrong). Starting too low and/or adapting up too slowly can cause
a critical point in a personal communication to be poor a critical point in a personal communication to be poor
("Hello!"). Starting over-bandwidth causes other problems for ("Hello!").
<!-- [rfced] Does "over-bandwidth" mean "excess bandwith"?
Original:
Starting
over-bandwidth causes other problems for user experience, so
there's a tension here.
Perhaps:
Starting with too much bandwith causes other problems for user experience,
so there's a tension here.
-->
Starting over-bandwidth causes other problems for
user experience, so there's a tension here. Alternative methods user experience, so there's a tension here. Alternative methods
to help startup like probing during setup with dummy data may be to help startup, such as probing during setup with dummy data, may
useful in some applications; in some cases there will be a be
useful in some applications; in some cases, there will be a
considerable gap in time between flow creation and the initial considerable gap in time between flow creation and the initial
flow of data. Again, A flow may need to change adaptation rates flow of data. Again, a flow may need to change adaptation rates
due to network conditions or changes in the provided flows (such due to network conditions or changes in the provided flows (such
as un-muting or sending data after a gap).</t> as unmuting or sending data after a gap).</li>
</list></t> </ol>
</li>
<li>
<t>The algorithm should be stable if the RTP streams are halted or <t>The algorithm should be stable if the RTP streams are halted or
discontinuous (for example - Voice Activity Detection). <list discontinuous (for example, when using Voice Activity Detection). </t>
style="letters"> <ol spacing="normal" type="a">
<t>After stream resumption, the algorithm should attempt to <li>After stream resumption, the algorithm should attempt to
rapidly regain its previous share of the bandwidth; the rapidly regain its previous share of the bandwidth; the
aggressiveness with which this is done will decay with the aggressiveness with which this is done will decay with the
length of the pause.</t> length of the pause.</li>
</list></t> </ol>
</li>
<t>The algorithm should where possible merge information across <li>
multiple RTP streams sent between two endpoints, when those RTP <t>Where possible, the algorithm should merge information across
multiple RTP streams sent between two endpoints when those RTP
streams share a common bottleneck, whether or not those streams are streams share a common bottleneck, whether or not those streams are
multiplexed onto the same ports, in order to allow congestion multiplexed onto the same ports. This will allow congestion
control of the set of streams together instead of as multiple control of the set of streams together instead of as multiple
independent streams. This allows better overall bandwidth independent streams. It will also allow better overall bandwidth
management, faster response to changing conditions, and fairer management, faster response to changing conditions, and fairer
sharing of bandwidth with other network users.<list style="letters"> sharing of bandwidth with other network users.</t>
<t>The algorithm should also share information and adaptation <ol spacing="normal" type="a">
<li>The algorithm should also share information and adaptation
with other non-RTP flows between the same endpoints, such as a with other non-RTP flows between the same endpoints, such as a
WebRTC DataChannel <xref WebRTC data channel <xref target="RFC8831" format="default"/>, whe
target="I-D.ietf-rtcweb-data-channel"></xref>, when n
possible.</t> possible.</li>
<li>When there are multiple streams across the same 5-tuple
<t>When there are multiple streams across the same 5-tuple
coordinating their bandwidth use and congestion control, the coordinating their bandwidth use and congestion control, the
algorithm should allow the application to control the relative algorithm should allow the application to control the relative
split of available bandwidth. The most correlated bandwidth split of available bandwidth. The most correlated bandwidth
usage would be with other flows on the same 5-tuple, but there usage would be with other flows on the same 5-tuple, but there
may be use in coordinating measurement and control of the local may be use in coordinating measurement and control of the local
link(s). Use of information about previous flows, especially on link(s). Use of information about previous flows, especially on
the same 5-tuple, may be useful input to the algorithm, the same 5-tuple, may be useful input to the algorithm,
especially to startup performance of a new flow.</t> especially regarding startup performance of a new flow.</li>
</list></t> </ol>
</li>
<li>
<!--[rfced] We're having difficulty parsing the sentence below. Is it the
network elements or the algorithm that convey cogestion-related information?
Original:
7. The algorithm should not require any special support from
network elements to convey congestion related information to be
functional.
Perhaps 1:
7. The algorithm should not require that network elements convey
congestion-related information for the algorithm to
be functional.
Perhaps 2:
7. The algorithm should not require any special support from network
elements to be able to convey congestion-related information.
-->
<t>The algorithm should not require any special support from network <t>The algorithm should not require any special support from network
elements to convey congestion related information to be functional. elements to convey congestion-related information to be functional.
As much as possible, it should leverage available information about As much as possible, it should leverage available information about
the incoming flow to provide feedback to the sender. Examples of the incoming flow to provide feedback to the sender. Examples of
this information are the packet arrival times, acknowledgements and this information are the packet arrival times, acknowledgements and
feedback, packet timestamps, and packet losses, Explicit Congestion feedback, packet timestamps, packet losses, and Explicit Congestion
Notification (ECN) <xref target="RFC3168"></xref>; all of these can Notification (ECN) <xref target="RFC3168" format="default"/>; all of t
hese can
provide information about the state of the path and any bottlenecks. provide information about the state of the path and any bottlenecks.
However, the use of available information is algorithm However, the use of available information is algorithm
dependent.<list style="letters"> dependent.</t>
<t>Extra information could be added to the packets to provide <ol spacing="normal" type="a">
<li>Extra information could be added to the packets to provide
more detailed information on actual send times (as opposed to more detailed information on actual send times (as opposed to
sampling times), but should not be required.</t> sampling times), but such information should not be required.</li>
</list></t> </ol>
</li>
<li>
<t>Since the assumption here is a set of RTP streams, the <t>Since the assumption here is a set of RTP streams, the
backchannel typically should be done via RTCP<xref backchannel typically should be done via the RTP Control Protocol
target="RFC3550"></xref>; one alternative would be to include it (RTCP) <xref target="RFC3550" format="default"/>; instead, one alternat
instead in a reverse RTP channel using header extensions.<list ive
style="letters"> would be to include it
<t>In order to react sufficiently quickly when using RTCP for a in a reverse-RTP channel using header extensions.</t>
backchannel, an RTP profile such as RTP/AVPF <xref <ol spacing="normal" type="a">
target="RFC4585"></xref> or RTP/SAVPF <xref <li>In order to react sufficiently quickly when using RTCP for a
target="RFC5124"></xref> that allows sufficiently frequent backchannel, an RTP profile such as RTP/AVPF <xref target="RFC4585
" format="default"/> or RTP/SAVPF <xref target="RFC5124" format="default"/> that
allows sufficiently frequent
feedback must be used. Note that in some cases, backchannel feedback must be used. Note that in some cases, backchannel
messages may be delayed until the RTCP channel can be allocated messages may be delayed until the RTCP channel can be allocated
enough bandwidth, even under AVPF rules. This may also imply enough bandwidth, even under AVPF rules. This may also imply
negotiating a higher maximum percentage for RTCP data or negotiating a higher maximum percentage for RTCP data or
allowing solutions to violate or modify the rules specified for allowing solutions to violate or modify the rules specified for
AVPF.</t> AVPF.</li>
<li>Bandwidth for the feedback messages should be minimized
<t>Bandwidth for the feedback messages should be minimized (such using techniques such as those in <xref target="RFC5506"
as via RFC 5506 <xref target="RFC5506"></xref>to allow RTCP format="default"/>, to allow RTCP
without Sender Reports/Receiver Reports)</t> without Sender/Receiver Reports.</li>
<li>Backchannel data should be minimized to avoid taking too much
<t>Backchannel data should be minimized to avoid taking too much
reverse-channel bandwidth (since this will often be used in a reverse-channel bandwidth (since this will often be used in a
bidirectional set of flows). In areas of stability, backchannel bidirectional set of flows). In areas of stability, backchannel
data may be sent more infrequently so long as algorithm data may be sent more infrequently so long as algorithm
stability and fairness are maintained. When the channel is stability and fairness are maintained. When the channel is
unstable or has not yet reached equilibrium after a change, unstable or has not yet reached equilibrium after a change,
backchannel feedback may be more frequent and use more backchannel feedback may be more frequent and use more
reverse-channel bandwidth. This is an area with considerable reverse-channel bandwidth. This is an area with considerable
flexibility of design, and different approaches to backchannel flexibility of design, and different approaches to backchannel
messages and frequency are expected to be evaluated.</t> messages and frequency are expected to be evaluated.</li>
</list></t> </ol>
</li>
<li>
<!-- [rfced] Please review the update below and let us know if any corrections
are needed.
<t>Flows managed by this algorithm and flows competing against at a Original:
bottleneck may have different DSCP<xref target="RFC5865"></xref> 9. Flows managed by this algorithm and flows competing against at a
markings depending on the type of traffic, or may be subject to bottleneck may have different DSCP[RFC5865] markings depending
on the type of traffic, or may be subject to flow-based QoS.
Current:
9. Flows managed by this algorithm and flows competing against each
other at a bottleneck may have different Differentiated Services
Code Point (DSCP) [RFC5865] markings depending on the type of
traffic or may be subject to flow-based QoS.
-->
<t>Flows managed by this algorithm and flows competing against each
other at a
bottleneck may have different Differentiated Services Code Point
(DSCP) <xref target="RFC5865" format="default"/>
markings depending on the type of traffic or may be subject to
flow-based QoS. A particular bottleneck or section of the network flow-based QoS. A particular bottleneck or section of the network
path may or may not honor DSCP markings. The algorithm should path may or may not honor DSCP markings. The algorithm should
attempt to leverage DSCP markings when they're available.<list attempt to leverage DSCP markings when they're available.</t>
style="letters"> <ol spacing="normal" type="a">
<t>In WebRTC, a division of packets into 4 classes is envisioned <!-- [rfced] The text indicates there are 4 classes, but 5 seem to be listed.
Please clarify.
A. In WebRTC, a division of packets into 4 classes is
envisioned in order of priority: faster-than-audio, audio,
video, best-effort, and bulk-transfer.
-->
<li>In WebRTC, a division of packets into 4 classes is envisioned
in order of priority: faster-than-audio, audio, video, in order of priority: faster-than-audio, audio, video,
best-effort, and bulk-transfer. Typically the flows managed by best-effort, and bulk-transfer. Typically, the flows managed by
this algorithm would be audio or video in that hierarchy, and this algorithm would be audio or video in that hierarchy, and
feedback flows would be faster-than-audio.</t> feedback flows would be faster-than-audio.</li>
</list></t> </ol>
</li>
<t>The algorithm should sense the unexpected lack of backchannel <li>The algorithm should sense the unexpected lack of backchannel
information as a possible indication of a channel overuse problem information as a possible indication of a channel-overuse problem
and react accordingly to avoid burst events causing a congestion and react accordingly to avoid burst events causing a congestion
collapse.</t> collapse.</li>
<li>The algorithm should be stable and maintain low delay when faced
<t>The algorithm should be stable and maintain low-delay when faced
with Active Queue Management (AQM) algorithms. Also note that these with Active Queue Management (AQM) algorithms. Also note that these
algorithms may apply across multiple queues in the bottleneck, or to algorithms may apply across multiple queues in the bottleneck or to
a single queue</t> a single queue.</li>
</list></t> </ol>
</section> </section>
<section numbered="true" toc="default">
<section title="Deficiencies of existing mechanisms "> <name>Deficiencies of Existing Mechanisms</name>
<t>Among the existing congestion control mechanisms TCP Friendly Rate <t>Among the existing congestion control mechanisms, TCP Friendly Rate
Control (TFRC) <xref target="RFC5348"></xref> is the one which claims to Control (TFRC) <xref target="RFC5348" format="default"/> is the one that c
be suitable for real-time interactive media. TFRC is, an equation based, laims to
congestion control mechanism which provides reasonably fair share of the be suitable for real-time interactive media. TFRC is an equation-based
congestion control mechanism that provides a reasonably fair share of
bandwidth when competing with TCP flows and offers much lower throughput bandwidth when competing with TCP flows and offers much lower throughput
variations than TCP. This is achieved by a slower response to the variations than TCP. This is achieved by a slower response to the
available bandwidth change than TCP. TFRC is designed to perform best available bandwidth change than TCP. TFRC is designed to perform best
with applications which has fixed packet size and does not have fixed with applications that have a fixed packet size and do not have a fixed
period between sending packets.</t> period between sending packets.</t>
<t>TFRC detects loss events and reacts to congestion-caused loss by
reducing its sending rate. It allows applications to
increase the sending rate until loss is observed in the flows. As
noted in IAB/IRTF report <xref target="RFC7295" format="default"/>, large
buffers
are available in the network elements, which introduce additional delay
in the communication. It becomes important to take all possible
congestion indications into consideration.
<!-- [rfced] We are having trouble parsing this sentence. Please consider
whether the suggested update correctly conveys the intended meaning.
<t>TFRC operates on detecting loss events and reacts to loss caused by Original:
congestion by reducing its sending rate. It allows applications to Looking at the current Internet
increase the sending rate until loss is observed in the flows. As it is deployment, TFRC's only consideration of loss events as congestion
noted in IAB/IRTF report <xref target="RFC7295"></xref> large buffers indication can be considered as biggest lacking.
are available in the network elements which introduces additional delay
in the communication, it becomes important to take all possible
congestion indications into considerations. Looking at the current
Internet deployment, TFRC's only consideration of loss events as
congestion indication can be considered as biggest lacking.</t>
<t>A typical real-time interactive communication includes live encoded Suggested:
audio and video flow(s). In such a communication scenario an audio Looking at the current
source typically needs fixed interval between packets, needs to vary Internet deployment, TFRC's biggest deficiency is that it only considers
loss events as a congestion indication.
-->
Looking at the current Internet
deployment, TFRC's only consideration of loss events as congestion
indication can be considered as biggest lacking.
</t>
<t>A typical real-time interactive communication includes live-encoded
<!--[rfced] We're having difficulty parsing the following sentence. What does
"their" refer to? In other words, what is the the thing whose segment size
needs to vary? What sends smaller packets?
Original:
In such a communication scenario an audio
source typically needs fixed interval between packets, needs to vary
their segment size instead of their packet rate in response to
congestion and sends smaller packets, ...
Perhaps:
In such a communication scenario, an audio
source typically needs a fixed interval between packets and needs to
vary the segment size of the packets instead of the packet rate in response
to congestion; therefore, it sends smaller packets.
-->
audio and video flow(s). In such a communication scenario, an audio
source typically needs a fixed interval between packets and needs to vary
their segment size instead of their packet rate in response to their segment size instead of their packet rate in response to
congestion and sends smaller packets, a variant of TFRC , Small-Packet congestion and sends smaller packets. A variant of TFRC, Small-Packet
TFRC (TFRC-SP) <xref target="RFC4828"></xref> addresses the issues TFRC (TFRC-SP) <xref target="RFC4828" format="default"/>, addresses the is
related to such kind of sources ; a video source generally varies video sues
frame sizes, can produce large frames which need to be further related to such kind of sources. A video source generally varies video
frame sizes, can produce large frames that need to be further
fragmented to fit into path Maximum Transmission Unit (MTU) size, and fragmented to fit into path Maximum Transmission Unit (MTU) size, and
have almost fixed interval between producing frames under a certain has an almost fixed interval between producing frames under a certain
frame rate, TFRC is known to be less optimal when using with such video frame rate. TFRC is known to be less optimal when using such video
sources.</t> sources.</t>
<t>There are also some mismatches between TFRC's design assumptions and <t>There are also some mismatches between TFRC's design assumptions and
how the media sources in a typical real-time interactive application how the media sources in a typical real-time interactive application
works. TFRC is design to maintain smooth sending rate however media work. TFRC is designed to maintain a smooth sending rate; however, media
sources can change rates in steps for both rate increase and rate sources can change rates in steps for both rate increase and rate
decrease. TFRC can operate in two modes - i) Bytes per second and ii) decrease. TFRC can operate in two modes: i) bytes per second and ii)
packets per second, where typical real-time interactive media sources packets per second, where typical real-time interactive media sources
operates on bit per second. There are also limitations on how quickly operate on bit per second. There are also limitations on how quickly
the media sources can adapt to specific sending rates. The modern video the media sources can adapt to specific sending rates. Modern video
encoders can operate on a mode where they can vary the output bitrate a encoders can operate in a mode in which they can vary the output bitrate a
lot depending on the way there are configured, the current scene it is lot depending on the way they are configured, the current scene they are
encoding and more. Therefore, it is possible that the video source does encoding, and more. Therefore, it is possible that the video source will
not always output at a bitrate they are allowed to. TFRC tries to raise not always output at an allowable bitrate. TFRC tries to increase
its sending rate when transmitting at maximum allowed rate and increases its sending rate when transmitting at the maximum allowed rate, and it inc
only twice the current transmission rate hence it may create issues when reases
the video source vary their bitrates.</t> only twice the current transmission rate; hence, it may create issues when
the video sources vary their bitrates.</t>
<t>Moreover, there are number of studies on TFRC which shows it's <t>Moreover, there are a number of studies on TFRC that show its
limitations which includes TFRC's unfairness on low statistically limitations, including TFRC's unfairness to low statistically
multiplexed links, oscillatory behavior, performance issue in highly multiplexed links, oscillatory behavior, performance issues in highly
dynamic loss rates conditions and more <xref target="CH09"></xref>.</t> dynamic loss-rate conditions, and more <xref target="CH09" format="default
"/>.</t>
<t>Looking at all these deficiencies it can be concluded that the <t>Looking at all these deficiencies, it can be concluded that the
requirements of congestion control mechanism for real-time interactive requirements for a congestion control mechanism for real-time interactive
media cannot be met by TFRC as defined in the standard.</t> media cannot be met by TFRC as defined in the standard.</t>
</section> </section>
<section anchor="IANA" numbered="true" toc="default">
<section anchor="IANA" title="IANA Considerations"> <name>IANA Considerations</name>
<t>This document makes no request of IANA.</t> <t>This document has no IANA actions.</t>
<t>Note to RFC Editor: this section may be removed on publication as an
RFC.</t>
</section> </section>
<section anchor="Security" numbered="true" toc="default">
<section anchor="Security" title="Security Considerations"> <name>Security Considerations</name>
<t>An attacker with the ability to delete, delay or insert messages in <t>An attacker with the ability to delete, delay, or insert messages into
the flow can fake congestion signals, unless they are passed on a the flow can fake congestion signals, unless they are passed on a
tamper-proof path. Since some possible algorithms depend on the timing tamper-proof path. Since some possible algorithms depend on the timing
of packet arrival, even a traditional protected channel does not fully of packet arrival, even a traditional, protected channel does not fully
mitigate such attacks.</t> mitigate such attacks.</t>
<t>An attack that reduces bandwidth is not necessarily significant, <t>An attack that reduces bandwidth is not necessarily significant,
since an on-path attacker could break the connection by discarding all since an on-path attacker could break the connection by discarding all
packets. Attacks that increase the perceived available bandwidth are packets. Attacks that increase the perceived available bandwidth are
conceivable, and need to be evaluated. Such attacks could result in conceivable and need to be evaluated. Such attacks could result in
starvation of competing flows and permit amplification attacks.</t> starvation of competing flows and permit amplification attacks.</t>
<t>Algorithm designers should consider the possibility of malicious <t>Algorithm designers should consider the possibility of malicious
on-path attackers.</t> on-path attackers.</t>
</section> </section>
<section anchor="Acknowledgements" title="Acknowledgements">
<t>This document is the result of discussions in various fora of the
WebRTC effort, in particular on the rtp-congestion@alvestrand.no mailing
list. Many people contributed their thoughts to this.</t>
</section>
</middle> </middle>
<back> <back>
<references title="Normative References">
<?rfc include="reference.RFC.2119"?>
<?rfc include='reference.RFC.3550'?>
<?rfc include='reference.RFC.4585'?> <references>
<name>References</name>
<?rfc include='reference.RFC.5124'?> <references>
<name>Normative References</name>
<?rfc include='reference.I-D.ietf-rtcweb-overview'?> <xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
</references> ence.RFC.3550.xml"/>
<xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
<references title="Informative References"> ence.RFC.4585.xml"/>
<?rfc include='reference.RFC.3168'?> <xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
ence.RFC.5124.xml"/>
<?rfc include='reference.RFC.5506'?>
<?rfc include='reference.RFC.5865'?>
<?rfc include='reference.RFC.5348'?>
<?rfc include='reference.RFC.4828'?> <!-- draft-ietf-rtcweb-overview: RFC 8825 -->
<reference anchor="RFC8825" target="https://www.rfc-editor.org/info/rfc8825">
<front>
<title>Overview: Real-Time Protocols for Browser-Based Applications</title>
<author initials="H." surname="Alvestrand" fullname="Harald T. Alvestrand">
<organization />
</author>
<date month="June" year="2020" />
</front>
<seriesInfo name="RFC" value="8825" />
<seriesInfo name="DOI" value="10.17487/RFC8825"/>
</reference>
<?rfc include='reference.RFC.7295'?> </references>
<?rfc include='reference.I-D.ietf-rtcweb-data-channel'?> <references>
<name>Informative References</name>
<xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
ence.RFC.3168.xml"/>
<xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
ence.RFC.5506.xml"/>
<xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
ence.RFC.5865.xml"/>
<xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
ence.RFC.5348.xml"/>
<xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
ence.RFC.4828.xml"/>
<xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
ence.RFC.7295.xml"/>
<?rfc include='reference.I-D.ietf-avtcore-rtp-circuit-breakers'?> <!-- draft-ietf-rtcweb-data-channel: 8831 -->
<reference anchor="RFC8831" target="https://www.rfc-editor.org/info/rfc8831">
<front>
<title>WebRTC Data Channels</title>
<author initials="R" surname="Jesup" fullname="Randell Jesup">
<organization/>
</author>
<author initials="S" surname="Loreto" fullname="Salvatore Loreto">
<organization/>
</author>
<author initials="M" surname="Tüxen" fullname="Michael Tüxen">
<organization/>
</author>
<date month='June' year='2020'/>
</front>
<seriesInfo name="RFC" value="8831"/>
<seriesInfo name="DOI" value="10.17487/RFC8831"/>
</reference>
<reference anchor="MPEG_DASH"> <!-- I-D.ietf-avtcore-rtp-circuit-breakers; RFC 8083 -->
<front> <xi:include href="https://xml2rfc.tools.ietf.org/public/rfc/bibxml/refer
<title>Dynamic adaptive streaming over HTTP (DASH) -- Part 1: Media ence.RFC.8083.xml"/>
presentation description and segment formats</title>
<author></author> <!--[rfced] FYI, we updated the following reference to point to the latest
version. Let us know if it's important to cite the 2012 version.
<date month="April" year="2012" /> Original:
</front> [MPEG_DASH]
"Dynamic adaptive streaming over HTTP (DASH) - Part 1:
Media presentation description and segment formats", April
2012.
<format target="http://standards.iso.org/ittf/PubliclyAvailableStandards Current:
/c057623_ISO_IEC_23009-1_2012.zip" [MPEG_DASH]
type="TXT" /> ISO, "Information Technology - Dynamic adaptive streaming
</reference> over HTTP (DASH) - Part 1: Media presentation description
and segment formats", ISO/IEC 23009-1:2019, December 2019,
<https://www.iso.org/standard/79329.html>.
-->
<reference anchor="MPEG_DASH" target="https://www.iso.org/standard/79329
.html">
<front>
<title>Information Technology -- Dynamic adaptive streaming over
HTTP (DASH) -- Part 1: Media presentation description and segment
formats</title>
<seriesInfo name="ISO/IEC" value="23009-1:2019"/>
<author>
<organization>ISO</organization>
</author>
<date month="December" year="2019"/>
</front>
</reference>
<reference anchor="CH09"> <reference anchor="CH09">
<front> <front>
<title>Designing TCP-Friendly Window-based Congestion Control for <title>Designing TCP-Friendly Window-based Congestion Control for
Real-time Multimedia Applications</title> Real-time Multimedia Applications</title>
<seriesInfo name="Proceedings of" value="PFLDNeT"/>
<author fullname="Soo-Hyun Choi" initials="S" surname="Choi">
<organization/>
</author>
<author fullname="Mark Handley" initials="M" surname="Handley">
<organization/>
</author>
<date month="May" year="2009"/>
</front>
</reference>
</references>
</references>
<author fullname="Soo-Hyun Choi" initials="S" surname="Choi"> <section anchor="Acknowledgements" numbered="false" toc="default">
<organization></organization> <name>Acknowledgements</name>
</author> <t>This document is the result of discussions in various fora of the
WebRTC effort, in particular on the &lt;rtp-congestion@alvestrand.no&gt; m
<author fullname="Mark Handley" initials="M" surname="Handley"> ailing
<organization></organization> list. Many people contributed their thoughts to this.</t>
</section>
<address>
<postal>
<street></street>
<city></city>
<region></region>
<code></code>
<country></country>
</postal>
<phone></phone>
<facsimile></facsimile>
<email></email>
<uri></uri>
</address>
</author>
<date day="21" month="May" year="2009" />
</front>
<seriesInfo name="PFLDNeT 2009 Workshop" value="" />
<format target="www.hpcc.jp/pfldnet2009/Program_files/1569199301.pdf"
type="PDF" />
</reference>
</references>
</back> </back>
</rfc> </rfc>
 End of changes. 110 change blocks. 
356 lines changed or deleted 477 lines changed or added

This html diff was produced by rfcdiff 1.45. The latest version is available from http://tools.ietf.org/tools/rfcdiff/