rfc9330.original.xml   rfc9330.xml 
<?xml version='1.0' encoding='utf-8'?> <?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE rfc [ <!DOCTYPE rfc [
<!ENTITY nbsp "&#160;"> <!ENTITY nbsp "&#160;">
<!ENTITY zwsp "&#8203;"> <!ENTITY zwsp "&#8203;">
<!ENTITY nbhy "&#8209;"> <!ENTITY nbhy "&#8209;">
<!ENTITY wj "&#8288;"> <!ENTITY wj "&#8288;">
]> ]>
<!-- This template is for creating an Internet Draft using xml2rfc,
which is available here: http://xml.resource.org. -->
<?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>
<!-- used by XSLT processors -->
<!-- For a complete list and description of processing instructions (PIs),
please see http://xml.resource.org/authoring/README.html. -->
<!-- Below are generally applicable Processing Instructions (PIs) that most I-Ds
might want to use.
(Here they are set differently than their defaults in xml2rfc v1.32) -->
<?rfc strict="yes" ?>
<!-- give errors regarding ID-nits and DTD validation -->
<!-- control the table of contents (ToC) -->
<?rfc toc="yes"?>
<!-- generate a ToC -->
<?rfc tocdepth="4"?>
<!-- the number of levels of subsections in ToC. default: 3 -->
<!-- control references -->
<?rfc symrefs="yes"?>
<!-- use symbolic references tags, i.e, [RFC2119] instead of [1] -->
<?rfc sortrefs="yes" ?>
<!-- sort the reference entries alphabetically -->
<!-- control vertical white space
(using these PIs as follows is recommended by the RFC Editor) -->
<?rfc compact="yes" ?>
<!-- do not start each main section on a new page -->
<?rfc subcompact="no" ?>
<!-- keep one blank line between list items -->
<!-- end of list of popular I-D processing instructions -->
<rfc xmlns:xi="http://www.w3.org/2001/XInclude" category="info" docName="draft-i
etf-tsvwg-l4s-arch-20" ipr="trust200902" obsoletes="" updates="" submissionType=
"IETF" xml:lang="en" tocInclude="true" tocDepth="4" symRefs="true" sortRefs="tru
e" version="3">
<!-- xml2rfc v2v3 conversion 3.14.1 -->
<!-- category values: std, bcp, info, exp, and historic
ipr values: trust200902, noModificationTrust200902, noDerivativesTrust200902
,
or pre5378Trust200902
you can add the attributes updates="NNNN" and obsoletes="NNNN"
they will automatically be output with "(if approved)" -->
<!-- ***** FRONT MATTER ***** --> <rfc xmlns:xi="http://www.w3.org/2001/XInclude" submissionType="IETF" category=" info" consensus="true" docName="draft-ietf-tsvwg-l4s-arch-20" number="9330" ipr= "trust200902" obsoletes="" updates="" xml:lang="en" tocInclude="true" tocDepth=" 4" symRefs="true" sortRefs="true" version="3">
<front> <front>
<!-- The abbreviated title is used in the page header - it is only necessary <title abbrev="L4S Architecture">Low Latency, Low Loss, and Scalable
if the Throughput (L4S) Internet Service: Architecture</title>
full title is longer than 39 characters --> <seriesInfo name="RFC" value="9330"/>
<author fullname="Bob Briscoe" initials="B." surname="Briscoe" role="editor">
<organization>Independent</organization>
<address>
<postal>
<street/>
<country>United Kingdom</country>
</postal>
<email>ietf@bobbriscoe.net</email>
<uri>https://bobbriscoe.net/</uri>
</address>
</author>
<author fullname="Koen De Schepper" initials="K." surname="De Schepper">
<organization>Nokia Bell Labs</organization>
<address>
<postal>
<street/>
<city>Antwerp</city>
<country>Belgium</country>
</postal>
<email>koen.de_schepper@nokia.com</email>
<uri>https://www.bell-labs.com/about/researcher-profiles/koende_schepper/<
/uri>
</address>
</author>
<author fullname="Marcelo Bagnulo" initials="M." surname="Bagnulo">
<organization>Universidad Carlos III de Madrid</organization>
<address>
<postal>
<street>Av. Universidad 30</street>
<city>Madrid</city>
<code>28911</code>
<country>Spain</country>
</postal>
<phone>34 91 6249500</phone>
<email>marcelo@it.uc3m.es</email>
<uri>https://www.it.uc3m.es</uri>
</address>
</author>
<author fullname="Greg White" initials="G." surname="White">
<organization>CableLabs</organization>
<address>
<postal>
<street/>
<country>United States of America</country>
</postal>
<email>G.White@CableLabs.com</email>
</address>
</author>
<date year="2023" month="January"/>
<area>tsv</area>
<workgroup>tsvwg</workgroup>
<keyword>Performance</keyword>
<keyword>Queuing Delay</keyword>
<keyword>One Way Delay</keyword>
<keyword>Round-Trip Time</keyword>
<keyword>RTT</keyword>
<keyword>Jitter</keyword>
<keyword>Congestion Control</keyword>
<keyword>Congestion Avoidance</keyword>
<keyword>Quality of Service</keyword>
<keyword>QoS</keyword>
<keyword>Quality of Experience</keyword>
<keyword>QoE</keyword>
<keyword>Active Queue Management</keyword>
<keyword>AQM</keyword>
<keyword>Explicit Congestion Notification</keyword>
<keyword>ECN</keyword>
<keyword>Pacing</keyword>
<keyword>Burstiness</keyword>
<title abbrev="L4S Architecture">Low Latency, Low Loss, Scalable <abstract>
Throughput (L4S) Internet Service: Architecture</title> <t>This document describes the L4S architecture, which enables Internet
<seriesInfo name="Internet-Draft" value="draft-ietf-tsvwg-l4s-arch-20"/> applications to achieve low queuing latency, low congestion loss, and scalab
<author fullname="Bob Briscoe" initials="B." role="editor" surname="Briscoe" le
> throughput control. L4S is based on the insight that the root cause of
<organization>Independent</organization> queuing delay is in the capacity-seeking congestion controllers of
<address> senders, not in the queue itself. With the L4S architecture, all Internet
<postal> applications could (but do not have to) transition away from congestion
<street/> control algorithms that cause substantial queuing delay and instead adopt a
<country>UK</country> new class
</postal> of congestion controls that can seek capacity with very little queuing.
<email>ietf@bobbriscoe.net</email> These are aided by a modified form of Explicit Congestion Notification
<uri>https://bobbriscoe.net/</uri> (ECN) from the network. With this new architecture, applications can
</address> have both low latency and high throughput.</t>
</author> <t>The architecture primarily concerns incremental deployment. It
<author fullname="Koen De Schepper" initials="K." surname="De Schepper"> defines mechanisms that allow the new class of L4S congestion controls
<organization>Nokia Bell Labs</organization> to coexist with 'Classic' congestion controls in a shared network. The
<address> aim is for L4S latency and throughput to be usually much better (and
<postal> rarely worse) while typically not impacting Classic performance.</t>
<street/> </abstract>
<city>Antwerp</city> </front>
<country>Belgium</country> <middle>
</postal> <section anchor="l4sps_intro" numbered="true" toc="default">
<email>koen.de_schepper@nokia.com</email> <name>Introduction</name>
<uri>https://www.bell-labs.com/about/researcher-profiles/koende_schepper <t>At any one time, it is increasingly common for all of the traffic in
/</uri> a bottleneck link (e.g., a household's Internet access or Wi-Fi) to come fro
</address> m
</author> applications that prefer low delay: interactive web, web services,
<author fullname="Marcelo Bagnulo" initials="M." surname="Bagnulo Braun"> voice, conversational video, interactive video, interactive remote
<organization>Universidad Carlos III de Madrid</organization> presence, instant messaging, online and cloud-rendered gaming, remote deskto
<address> p, cloud-based
<postal> applications, cloud-rendered virtual reality or augmented reality, and video
<street>Av. Universidad 30</street> -assisted remote control of machinery and
<city>Leganes, Madrid 28911</city> industrial processes. In the last decade or so, much has been done to
<country>Spain</country> reduce propagation delay by placing caches or servers closer to users.
</postal> However, queuing remains a major, albeit intermittent, component of
<phone>34 91 6249500</phone> latency. For instance, spikes of hundreds of milliseconds are not
<email>marcelo@it.uc3m.es</email> uncommon, even with state-of-the-art Active Queue Management
<uri>https://www.it.uc3m.es</uri> (AQM) <xref target="COBALT" format="default"/> <xref target="DOCSIS3AQM" for
</address> mat="default"/>. A Classic AQM in an
</author> access network bottleneck is typically configured to buffer the sawteeth of
<author fullname="Greg White" initials="G." surname="White"> lone flows, which can cause peak overall
<organization>CableLabs</organization> network delay to roughly double during a long-running flow, relative to
<address> expected base (unloaded) path delay <xref target="BufferSize" format="defaul
<postal> t"/>.
<street/> Low loss is also important because, for interactive applications, losses
<country>US</country> translate into even longer retransmission delays.</t>
</postal> <t>It has been demonstrated that, once access network bit rates reach
<email>G.White@CableLabs.com</email> levels now common in the developed world, increasing link capacity
</address> offers diminishing returns if latency (delay) is not addressed <xref target=
</author> "Dukkipati06" format="default"/> <xref target="Rajiullah15" format="default"/>.
<date month="" year=""/> Therefore, the
<area>Transport</area> goal is an Internet service with very low queuing latency, very low
<workgroup>Transport Area Working Group</workgroup> loss, and scalable throughput. Very low queuing latency means less
<keyword>Internet-Draft</keyword> than 1 millisecond (ms) on average and less than about 2 ms at
<keyword>I-D</keyword> the 99th percentile. End-to-end delay above 50 ms <xref target="Raaen14" for
<abstract> mat="default"/>, or even above 20 ms <xref target="NASA04" format="default"/>,
<t>This document describes the L4S architecture, which enables Internet starts to feel unnatural for more demanding interactive applications. Theref
applications to achieve Low queuing Latency, Low Loss, and Scalable ore,
throughput (L4S). L4S is based on the insight that the root cause of removing unnecessary delay variability increases the reach of these
queuing delay is in the capacity-seeking congestion controllers of applications (the distance over which they are comfortable to use) and/or
senders, not in the queue itself. With the L4S architecture all Internet provides additional latency budget that can be used for enhanced processing.
applications could (but do not have to) transition away from congestion This
control algorithms that cause substantial queuing delay, to a new class document describes the L4S architecture for achieving these goals.</t>
of congestion controls that can seek capacity with very little queuing. <t>Differentiated services (Diffserv) offers Expedited Forwarding
These are aided by a modified form of explicit congestion notification (EF) <xref target="RFC3246" format="default"/> for some packets at the expen
(ECN) from the network. With this new architecture, applications can se of
have both low latency and high throughput.</t> others, but this makes no difference when all (or most) of the traffic
<t>The architecture primarily concerns incremental deployment. It at a bottleneck at any one time requires low latency. In contrast, L4S
defines mechanisms that allow the new class of L4S congestion controls still works well when all traffic is L4S -- a service that gives without
to coexist with 'Classic' congestion controls in a shared network. The taking needs none of the configuration or management baggage (traffic
aim is for L4S latency and throughput to be usually much better (and policing or traffic contracts) associated with favouring some traffic
rarely worse), while typically not impacting Classic performance.</t> flows over others.</t>
</abstract> <t>Queuing delay degrades performance intermittently <xref target="Hohlfeld1
</front> 4" format="default"/>.
<middle> It occurs i) when a large enough capacity-seeking
<section anchor="l4sps_intro" numbered="true" toc="default"> (e.g., TCP) flow is running alongside the user's traffic in the
<name>Introduction</name> bottleneck link, which is typically in the access network, or ii) when the
<t>At any one time, it is increasingly common for all of the traffic in low latency application is itself a large capacity-seeking or adaptive
a bottleneck link (e.g. a household's Internet access) to come from rate flow (e.g., interactive video).
applications that prefer low delay: interactive Web, Web services, At these times, the performance
voice, conversational video, interactive video, interactive remote improvement from L4S must be sufficient for network operators to be motivate
presence, instant messaging, online gaming, remote desktop, cloud-based d
applications and video-assisted remote control of machinery and to deploy it.</t>
industrial processes. In the last decade or so, much has been done to <t>Active Queue Management (AQM) is part of the solution to queuing
reduce propagation delay by placing caches or servers closer to users. under load. AQM improves performance for all traffic, but there is a
However, queuing remains a major, albeit intermittent, component of limit to how much queuing delay can be reduced by solely changing the
latency. For instance spikes of hundreds of milliseconds are not network without addressing the root of the problem.</t>
uncommon, even with state-of-the-art active queue management <t>The root of the problem is the presence of standard congestion
(AQM) <xref target="COBALT" format="default"/>, <xref target="DOCSIS3AQM" control (Reno <xref target="RFC5681" format="default"/>) or compatible varia
format="default"/>. Queuing nts
in access network bottlenecks is typically configured to cause overall (e.g., CUBIC <xref target="RFC8312" format="default"/>) that are used in TCP
network delay to roughly double during a long-running flow, relative to and
expected base (unloaded) path delay <xref target="BufferSize" format="defa in other transports, such as QUIC <xref target="RFC9000" format="default"/>.
ult"/>. We shall use
Low loss is also important because, for interactive applications, losses the term 'Classic' for these Reno-friendly congestion controls.
translate into even longer retransmission delays.</t> Classic
<t>It has been demonstrated that, once access network bit rates reach congestion controls induce relatively large sawtooth-shaped excursions
levels now common in the developed world, increasing link capacity of queue occupancy. So if a network operator naively
offers diminishing returns if latency (delay) is not addressed <xref targe attempts to reduce queuing delay by configuring an AQM to operate at a
t="Dukkipati06" format="default"/>, <xref target="Rajiullah15" format="default"/ shallower queue, a Classic congestion control will significantly
>. Therefore, the underutilize the link at the bottom of every sawtooth. These sawteeth have
goal is an Internet service with very Low queueing Latency, very Low also been growing in duration as flow rate scales (see <xref target="l4sps_w
Loss and Scalable throughput (L4S). Very low queuing latency means less hy_primary_components" format="default"/>
than 1 millisecond (ms) on average and less than about 2 ms at and <xref target="RFC3649" format="default"/>).</t>
the 99th percentile. End-to-end delay above 50 ms <xref target="Raaen14" f <t>It has been demonstrated that, if the sending host replaces a Classic
ormat="default"/> or even above 20 ms <xref target="NASA04" format="default"/> congestion control with a 'Scalable' alternative, the performance under load
starts to feel unnatural for more demanding interactive applications. So of all the above
removing unnecessary delay variability increases the reach of these interactive applications can be significantly improved once a suitable AQM i
applications (the distance over which they are comfortable to use). This s
document describes the L4S architecture for achieving these goals.</t> deployed in the network.
<t>Differentiated services (Diffserv) offers Expedited Forwarding Taking the example solution cited below that uses Data Center TCP (DCTCP)
(EF <xref target="RFC3246" format="default"/>) for some packets at the exp <xref target="RFC8257" format="default"/> and a Dual-Queue Coupled AQM <xref
ense of target="RFC9332"
others, but this makes no difference when all (or most) of the traffic format="default"/> on a DSL or Ethernet link,
at a bottleneck at any one time requires low latency. In contrast, L4S queuing delay under heavy load is roughly 1-2 ms at
still works well when all traffic is L4S - a service that gives without the 99th percentile without losing link utilization <xref target="L4Seval22"
taking needs none of the configuration or management baggage (traffic format="default"/> <xref target="DualPI2Linux" format="default"/> (for other li
policing, traffic contracts) associated with favouring some traffic nk types,
flows over others.</t> see <xref target="l4sarch_link-specifics" format="default"/>).
<t>Queuing delay degrades performance intermittently <xref target="Hohlfel This compares with
d14" format="default"/>. It occurs when a large enough capacity-seeking 5-20 ms on <em>average</em> with a Classic
(e.g. TCP) flow is running alongside the user's traffic in the congestion control and current state-of-the-art AQMs, such as
bottleneck link, which is typically in the access network. Or when the Flow Queue CoDel <xref target="RFC8290" format="default"/>, Proportional Int
low latency application is itself a large capacity-seeking or adaptive egral controller Enhanced (PIE) <xref target="RFC8033" format="default"/>, or DO
rate (e.g. interactive video) flow. At these times, the performance CSIS PIE <xref target="RFC8034" format="default"/> and about
improvement from L4S must be sufficient that network operators will be 20-30 ms at the 99th percentile <xref target="DualPI2Linux" format="default"
motivated to deploy it.</t> />.</t>
<t>Active Queue Management (AQM) is part of the solution to queuing <t>L4S is designed for incremental deployment. It is possible to deploy
under load. AQM improves performance for all traffic, but there is a the L4S service at a bottleneck link alongside the existing best efforts
limit to how much queuing delay can be reduced by solely changing the service <xref target="DualPI2Linux" format="default"/> so that unmodified
network; without addressing the root of the problem.</t> applications can start using it as soon as the sender's stack is
<t>The root of the problem is the presence of standard congestion updated. Access networks are typically designed with one link as the
control (Reno <xref target="RFC5681" format="default"/>) or compatible var bottleneck for each site (which might be a home, small enterprise, or
iants mobile device), so deployment at either or both ends of this link should
(e.g. CUBIC <xref target="RFC8312" format="default"/>) that are used in TC give nearly all the benefit in the respective direction.
P and With some
in other transports such as QUIC <xref target="RFC9000" format="default"/> transport protocols, namely TCP <xref target="I-D.ietf-tcpm-accurate-ecn" fo
. We shall use rmat="default"/>, the sender has to check that
the term 'Classic' for these Reno-friendly congestion controls. Classic the receiver has been suitably updated to give more accurate feedback,
congestion controls induce relatively large saw-tooth-shaped excursions whereas with more recent transport protocols, such as QUIC <xref target="RFC
up the queue and down again, which have been growing as flow rate 9000" format="default"/> and Datagram Congestion Control Protocol (DCCP) <xref t
scales <xref target="RFC3649" format="default"/>. So if a network operator arget="RFC4340" format="default"/>, all
naively receivers have always been suitable.</t>
attempts to reduce queuing delay by configuring an AQM to operate at a <t>This document presents the L4S architecture. It consists of three
shallower queue, a Classic congestion control will significantly components: network support to isolate L4S traffic from Classic traffic;
underutilize the link at the bottom of every saw-tooth.</t> protocol features that allow network elements to identify L4S traffic;
<t>It has been demonstrated that if the sending host replaces a Classic and host support for L4S congestion controls. The protocol is defined
congestion control with a 'Scalable' alternative, when a suitable AQM is separately in <xref target="RFC9331" format="default"/> as an experimental
deployed in the network the performance under load of all the above change to Explicit Congestion Notification (ECN). This document
interactive applications can be significantly improved. For instance, describes and justifies the component parts and how they interact to
queuing delay under heavy load with the example DCTCP/DualQ solution provide the low latency, low loss, and scalable Internet service. It also
cited below on a DSL or Ethernet link is roughly 1 to 2 milliseconds at details the approach to incremental deployment, as briefly summarized
the 99th percentile without losing link utilization <xref target="DualPI2L above.</t>
inux" format="default"/>, <xref target="DCttH19" format="default"/> (for other l <section numbered="true" toc="default">
ink types, <name>Document Roadmap</name>
see <xref target="l4sarch_link-specifics" format="default"/>). This compar <t>This document describes the L4S architecture in three passes. First,
es with the brief overview in <xref target="l4s-arch_arch_overview" format="defaul
5-20 ms on <em>average</em> with a Classic t"/> gives the very high-level idea and states the main
congestion control and current state-of-the-art AQMs such as components with minimal rationale. This is only intended to give some
FQ-CoDel <xref target="RFC8290" format="default"/>, PIE <xref target="RFC8 context for the terminology definitions that follow in <xref target="l4sps
033" format="default"/> or DOCSIS PIE <xref target="RFC8034" format="default"/> _Terminology" format="default"/> and to explain the structure of the rest
and about of the document. Then, <xref target="l4sps_components" format="default"/>
20-30 ms at the 99th percentile <xref target="DualPI2Linux" format="defaul goes into more
t"/>.</t> detail on each component with some rationale but still mostly stating
<t>L4S is designed for incremental deployment. It is possible to deploy what the architecture is, rather than why. Finally, <xref target="l4sps_ra
the L4S service at a bottleneck link alongside the existing best efforts tionale" format="default"/> justifies why each element of the solution
service <xref target="DualPI2Linux" format="default"/> so that unmodified was chosen (<xref target="l4sps_why_primary_components" format="default"/>
applications can start using it as soon as the sender's stack is ) and why
updated. Access networks are typically designed with one link as the these choices were different from other solutions (<xref target="l4sps_why
bottleneck for each site (which might be a home, small enterprise or -not" format="default"/>).</t>
mobile device), so deployment at either or both ends of this link should <t>After the architecture has been described, <xref target="l4sarch_applic
give nearly all the benefit in the respective direction. With some ability" format="default"/>
transport protocols, namely TCP and SCTP, the sender has to check that clarifies its applicability by describing the applications and use cases
the receiver has been suitably updated to give more accurate feedback, that motivated the design, the challenges applying the architecture to
whereas with more recent transport protocols such as QUIC and DCCP, all various link technologies, and various incremental deployment models
receivers have always been suitable.</t> (including the two main deployment topologies, different sequences for
<t>This document presents the L4S architecture. It consists of three incremental deployment, and various interactions with preexisting
components: network support to isolate L4S traffic from classic traffic; approaches). The document
protocol features that allow network elements to identify L4S traffic; ends with the usual tailpieces, including extensive discussion of
and host support for L4S congestion controls. The protocol is defined traffic policing and other security considerations in <xref target="l4sps_
separately <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/> as Security_Considerations" format="default"/>.</t>
an experimental
change to Explicit Congestion Notification (ECN). This document
describes and justifies the component parts and how they interact to
provide the scalable, low latency, low loss Internet service. It also
details the approach to incremental deployment, as briefly summarized
above.</t>
<section numbered="true" toc="default">
<name>Document Roadmap</name>
<t>This document describes the L4S architecture in three passes. First
this brief overview gives the very high level idea and states the main
components with minimal rationale. This is only intended to give some
context for the terminology definitions that follow in <xref target="l4s
ps_Terminology" format="default"/>, and to explain the structure of the rest
of the document. Then <xref target="l4sps_components" format="default"/>
goes into more
detail on each component with some rationale, but still mostly stating
what the architecture is, rather than why. Finally, <xref target="l4sps_
rationale" format="default"/> justifies why each element of the solution
was chosen (<xref target="l4sps_why_primary_components" format="default"
/>) and why
these choices were different from other solutions (<xref target="l4sps_w
hy-not" format="default"/>).</t>
<t>Having described the architecture, <xref target="l4sarch_applicabilit
y" format="default"/> clarifies its applicability; that is,
the applications and use-cases that motivated the design, the
challenges applying the architecture to various link technologies, and
various incremental deployment models: including the two main
deployment topologies, different sequences for incremental deployment
and various interactions with pre-existing approaches. The document
ends with the usual tailpieces, including extensive discussion of
traffic policing and other security considerations in <xref target="l4sp
s_Security_Considerations" format="default"/>.</t>
</section>
</section>
<section anchor="l4s-arch_arch_overview" numbered="true" toc="default">
<name>L4S Architecture Overview</name>
<t>Below we outline the three main components to the L4S architecture;
1) the scalable congestion control on the sending host; 2) the AQM at
the network bottleneck; and 3) the protocol between them.</t>
<t>But first, the main point to grasp is that low latency is not
provided by the network - low latency results from the careful behaviour
of the scalable congestion controllers used by L4S senders. The network
does have a role - primarily to isolate the low latency of the carefully
behaving L4S traffic from the higher queuing delay needed by traffic
with pre-existing Classic behaviour. The network also alters the way it
signals queue growth to the transport - It uses the Explicit Congestion
Notification (ECN) protocol, but it signals the very start of queue
growth - immediately without the smoothing delay typical of Classic
AQMs. Because ECN support is essential for L4S, senders use the ECN
field as the protocol that allows the network to identify which packets
are L4S and which are Classic.</t>
<dl newline="false" spacing="normal">
<dt>1) Host:</dt>
<dd>
<t>Scalable congestion controls already exist.
They solve the scaling problem with Classic congestion controls,
such as Reno or Cubic. Because flow rate has scaled since TCP
congestion control was first designed in 1988, assuming the flow
lasts long enough, it now takes hundreds of round trips (and
growing) to recover after a congestion signal (whether a loss or an
ECN mark) as shown in the examples in <xref target="l4sps_why_primary_
components" format="default"/> and <xref target="RFC3649" format="default"/>. Th
erefore, control of queuing and utilization
becomes very slack, and the slightest disturbances (e.g. from
new flows starting) prevent a high rate from being attained.</t>
<t>With a scalable congestion control, the average time
from one congestion signal to the next (the recovery time) remains
invariant as the flow rate scales, all other factors being equal.
This maintains the same degree of control over queueing and
utilization whatever the flow rate, as well as ensuring that high
throughput is more robust to disturbances. The scalable control used
most widely (in controlled environments) is Data Center TCP
(DCTCP <xref target="RFC8257" format="default"/>), which has been impl
emented
and deployed in Windows Server Editions (since 2012), in Linux and
in FreeBSD. Although DCTCP as-is functions well over wide-area round
trip times, most implementations lack certain safety features that
would be necessary for use outside controlled environments like data
centres (see <xref target="l4sarch_sec_non-l4s-neck" format="default"/
>). So scalable
congestion control needs to be implemented in TCP and other
transport protocols (QUIC, SCTP, RTP/RTCP, RMCAT, etc.). Indeed,
between the present document being drafted and published, the
following scalable congestion controls were implemented: TCP
Prague <xref target="PragueLinux" format="default"/>, QUIC Prague, an
L4S
variant of the RMCAT SCReAM controller <xref target="SCReAM" format="d
efault"/>
and the L4S ECN part of BBRv2 <xref target="BBRv2" format="default"/>
intended
for TCP and QUIC transports.</t>
</dd>
<dt>2) Network:</dt>
<dd>
<t>L4S traffic needs to be isolated from the
queuing latency of Classic traffic. One queue per application flow
(FQ) is one way to achieve this, e.g. FQ-CoDel <xref target="RFC8290"
format="default"/>. However, using just two queues is sufficient and
does not require inspection of transport layer headers in the
network, which is not always possible (see <xref target="l4sps_why-not
" format="default"/>). With just two queues, it might seem
impossible to know how much capacity to schedule for each queue
without inspecting how many flows at any one time are using each.
And it would be undesirable to arbitrarily divide access network
capacity into two partitions. The Dual Queue Coupled AQM was
developed as a minimal complexity solution to this problem. It acts
like a 'semi-permeable' membrane that partitions latency but not
bandwidth. As such, the two queues are for transition from Classic
to L4S behaviour, not bandwidth prioritization.</t>
<t><xref target="l4sps_components" format="default"/> gives a high lev
el
explanation of how the per-flow-queue (FQ) and DualQ variants of L4S
work, and <xref target="I-D.ietf-tsvwg-aqm-dualq-coupled" format="defa
ult"/> gives a
full explanation of the DualQ Coupled AQM framework. A specific
marking algorithm is not mandated for L4S AQMs. Appendices of <xref ta
rget="I-D.ietf-tsvwg-aqm-dualq-coupled" format="default"/> give non-normative
examples that have been implemented and evaluated, and give
recommended default parameter settings. It is expected that L4S
experiments will improve knowledge of parameter settings and whether
the set of marking algorithms needs to be limited.<!--{ToDo: Add ref t
o Mohit's draft re L4S FQ, once available.}-->
</t>
</dd>
<dt>3) Protocol:</dt>
<dd>A sending host needs to distinguish L4S
and Classic packets with an identifier so that the network can
classify them into their separate treatments. The L4S identifier
spec. <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/> conc
ludes that
all alternatives involve compromises, but the ECT(1) and CE
codepoints of the ECN field represent a workable solution. As
already explained, the network also uses ECN to immediately signal
the very start of queue growth to the transport.</dd>
</dl>
</section> </section>
<section anchor="l4sps_Terminology" numbered="true" toc="default"> </section>
<name>Terminology</name> <section anchor="l4s-arch_arch_overview" numbered="true" toc="default">
<t>[Note to the RFC Editor (to be removed before publication as an RFC): <name>L4S Architecture Overview</name>
The following definitions are copied from the L4S ECN spec <xref target="I <t>Below, we outline the three main components to the L4S architecture:
-D.ietf-tsvwg-ecn-l4s-id" format="default"/> for the reader's convenience. 1) the Scalable congestion control on the sending host; 2) the AQM at
Except, here, Classic CC and Scalable CC are condensed because they the network bottleneck; and 3) the protocol between them.</t>
refer to <xref target="l4sps_why_primary_components" format="default"/> la <t>But first, the main point to grasp is that low latency is not
ter. Also the provided by the network; low latency results from the careful behaviour
definition of Traffic Policing is not needed in <xref target="I-D.ietf-tsv of the Scalable congestion controllers used by L4S senders. The network
wg-ecn-l4s-id" format="default"/>.]</t> does have a role, primarily to isolate the low latency of the carefully
<dl newline="false" spacing="normal"> behaving L4S traffic from the higher queuing delay needed by traffic
<dt>Classic Congestion Control:</dt> with preexisting Classic behaviour. The network also alters the way it
<dd>A congestion control signals queue growth to the transport. It uses the Explicit Congestion
behaviour that can co-exist with standard Reno <xref target="RFC5681" Notification (ECN) protocol, but it signals the very start of queue
format="default"/> without causing significantly negative impact on growth immediately, without the smoothing delay typical of Classic
its flow rate <xref target="RFC5033" format="default"/>. The scaling p AQMs. Because ECN support is essential for L4S, senders use the ECN
roblem field as the protocol that allows the network to identify which packets
with Classic congestion control is explained, with examples, in are L4S and which are Classic.</t>
<xref target="l4sps_why_primary_components" format="default"/> and in <ol spacing="normal" type="%d)">
<xref target="RFC3649" format="default"/>.</dd> <li><t>Host:</t>
<dt>Scalable Congestion Control:</dt> <t>Scalable congestion controls already exist. They solve the scaling
<dd>A congestion control problem with Classic congestion controls, such as Reno or
where the average time from one congestion signal to the next (the CUBIC. Because flow rate has scaled since TCP congestion control was
recovery time) remains invariant as the flow rate scales, all other first designed in 1988, assuming the flow lasts long enough, it now
factors being equal. For instance, DCTCP averages 2 congestion takes hundreds of round trips (and growing) to recover after a
signals per round-trip whatever the flow rate, as do other recently congestion signal (whether a loss or an ECN mark), as shown in the
developed scalable congestion controls, e.g. Relentless examples in <xref target="l4sps_why_primary_components"
TCP <xref target="Mathis09" format="default"/>, TCP Prague <xref targe format="default"/> and <xref target="RFC3649"
t="I-D.briscoe-iccrg-prague-congestion-control" format="default"/>, <xref target format="default"/>. Therefore, control of queuing and utilization
="PragueLinux" format="default"/>, BBRv2 <xref target="BBRv2" format="default"/> becomes very slack, and the slightest disturbances (e.g., from new
, <xref target="I-D.cardwell-iccrg-bbr-congestion-control" format="default"/> an flows starting) prevent a high rate from being attained.</t>
d the L4S <t>With a Scalable congestion control, the average time from one
variant of SCReAM for real-time media <xref target="SCReAM" format="de congestion signal to the next (the recovery time) remains invariant as
fault"/>, <xref target="RFC8298" format="default"/>). See Section 4.3 flow rate scales, all other factors being equal. This maintains
of <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/> for mor the same degree of control over queuing and utilization, whatever the
e flow rate, as well as ensuring that high throughput is more robust to
explanation.</dd> disturbances. The Scalable control used most widely (in controlled
<dt>Classic service:</dt> environments) is DCTCP <xref target="RFC8257"
<dd>The Classic service is intended for format="default"/>, which has been implemented and deployed in
all the congestion control behaviours that co-exist with Windows Server Editions (since 2012), in Linux, and in
Reno <xref target="RFC5681" format="default"/> (e.g. Reno itself, FreeBSD.
Cubic <xref target="RFC8312" format="default"/>, Compound <xref target Although DCTCP as-is functions well over wide-area round-trip
="I-D.sridharan-tcpm-ctcp" format="default"/>, TFRC <xref target="RFC5348" forma times (RTTs), most implementations lack certain safety features that would
t="default"/>). The term 'Classic queue' means a queue be
necessary for use outside controlled environments, like data centres
(see <xref target="l4sarch_sec_non-l4s-neck" format="default"/>). Therefor
e,
Scalable congestion control needs to be implemented in TCP and other
transport protocols (QUIC, Stream Control Transmission Protocol (SCTP), RT
P/RTCP, RTP Media Congestion Avoidance Techniques (RMCAT), etc.).
Indeed,
between the present document being drafted and published, the
following Scalable congestion controls were implemented: Prague over TCP a
nd QUIC
<xref target="I-D.briscoe-iccrg-prague-congestion-control" format="default
"/> <xref target="PragueLinux" format="default"/>, an L4S
variant of the RMCAT SCReAM controller <xref target="SCReAM-L4S"
format="default"/>, and the L4S ECN part of Bottleneck Bandwidth and Round
-trip propagation time (BBRv2) <xref target="BBRv2"
format="default"/> intended for TCP and QUIC transports.</t>
</li>
<li><t>Network:</t>
<t>L4S traffic needs to be isolated from the queuing latency of
Classic traffic. One queue per application flow (FQ) is one way to
achieve this, e.g., FQ-CoDel <xref target="RFC8290"
format="default"/>. However, using just two queues is sufficient and
does not require inspection of transport layer headers in the network,
which is not always possible (see <xref target="l4sps_why-not"
format="default"/>). With just two queues, it might seem impossible to
know how much capacity to schedule for each queue without inspecting
how many flows at any one time are using each. And it would be
undesirable to arbitrarily divide access network capacity into two
partitions. The Dual-Queue Coupled AQM was developed as a minimal
complexity solution to this problem. It acts like a 'semi-permeable'
membrane that partitions latency but not bandwidth. As such, the two
queues are for transitioning from Classic to L4S behaviour, not bandwidth
prioritization.</t>
<t><xref target="l4sps_components" format="default"/> gives a high-level
explanation of how the per-flow queue (FQ) and DualQ variants of
L4S work, and <xref target="RFC9332"
format="default"/> gives a full explanation of the DualQ Coupled AQM
framework. A specific marking algorithm is not mandated for L4S
AQMs. Appendices of <xref target="RFC9332"
format="default"/> give non-normative examples that have been
implemented and evaluated and give recommended default parameter
settings. It is expected that L4S experiments will improve knowledge
of parameter settings and whether the set of marking algorithms needs
to be limited.
</t>
</li>
<li><t>Protocol:</t>
<t>A sending host needs to distinguish L4S and Classic packets with an
identifier so that the network can classify them into their separate
treatments. The L4S identifier spec <xref
target="RFC9331" format="default"/> concludes that
all alternatives involve compromises, but the ECT(1) and Congestion Experi
enced (CE) codepoints
of the ECN field represent a workable solution. As already explained,
the network also uses ECN to immediately signal the very start of
queue growth to the transport.</t>
</li>
</ol>
</section>
<section anchor="l4sps_Terminology" numbered="true" toc="default">
<name>Terminology</name>
<dl newline="false" spacing="normal">
<dt>Classic Congestion Control:</dt>
<dd>A congestion control
behaviour that can coexist with standard Reno <xref target="RFC5681" forma
t="default"/> without causing significantly negative impact on
its flow rate <xref target="RFC5033" format="default"/>. The scaling probl
em
with Classic congestion control is explained, with examples, in
<xref target="l4sps_why_primary_components" format="default"/> and in <xre
f target="RFC3649" format="default"/>.</dd>
<dt>Scalable Congestion Control:</dt>
<dd>A congestion control
where the average time from one congestion signal to the next (the
recovery time) remains invariant as flow rate scales, all other
factors being equal.
For instance, DCTCP averages 2 congestion
signals per round trip, whatever the flow rate, as do other recently
developed Scalable congestion controls, e.g., Relentless
TCP <xref target="I-D.mathis-iccrg-relentless-tcp" format="default"/>, Pra
gue for TCP and QUIC <xref target="I-D.briscoe-iccrg-prague-congestion-control"
format="default"/> <xref target="PragueLinux" format="default"/>, BBRv2 <xref ta
rget="BBRv2" format="default"/> <xref target="I-D.cardwell-iccrg-bbr-congestion-
control" format="default"/>, and the L4S
variant of SCReAM for real-time media <xref target="SCReAM-L4S" format="de
fault"/> <xref target="RFC8298" format="default"/>. See
<xref target="RFC9331" format="default" sectionFormat="of" section="4.3"/>
for more
explanation.</dd>
<dt>Classic Service:</dt>
<dd>The Classic service is intended for
all the congestion control behaviours that coexist with
Reno <xref target="RFC5681" format="default"/> (e.g., Reno itself,
CUBIC <xref target="RFC8312" format="default"/>, Compound <xref target="I-
D.sridharan-tcpm-ctcp" format="default"/>, and TFRC <xref target="RFC5348" forma
t="default"/>). The term 'Classic queue' means a queue
providing the Classic service.</dd> providing the Classic service.</dd>
<dt>Low-Latency, Low-Loss Scalable throughput (L4S) service:</dt> <dt>Low Latency, Low Loss, and Scalable throughput (L4S) service:</dt>
<dd> <dd>
<t>The <t>The
'L4S' service is intended for traffic from scalable congestion 'L4S' service is intended for traffic from Scalable congestion
control algorithms, such as the Prague congestion control <xref target control algorithms, such as the Prague congestion control <xref target
="I-D.briscoe-iccrg-prague-congestion-control" format="default"/>, which was ="I-D.briscoe-iccrg-prague-congestion-control" format="default"/>, which was
derived from DCTCP  <xref target="RFC8257" format="default"/>. The L4S derived from DCTCP <xref target="RFC8257" format="default"/>. The L4S
service service
is for more general traffic than just Prague -- it allows the is for more general traffic than just Prague -- it allows the
set of congestion controls with similar scaling properties to Prague set of congestion controls with similar scaling properties to Prague
to evolve, such as the examples listed above (Relentless, SCReAM). to evolve, such as the examples listed above (Relentless, SCReAM, etc. ).
The term 'L4S queue' means a queue providing the L4S service.</t> The term 'L4S queue' means a queue providing the L4S service.</t>
<t>The terms Classic or L4S can also qualify other <t>The terms Classic or L4S can also qualify other
nouns, such as 'queue', 'codepoint', 'identifier', 'classification', nouns, such as 'queue', 'codepoint', 'identifier', 'classification',
'packet', 'flow'. For example: an L4S packet means a packet with an 'packet', and 'flow'. For example, an L4S packet means a packet with a n
L4S identifier sent from an L4S congestion control.</t> L4S identifier sent from an L4S congestion control.</t>
<t>Both Classic and L4S services can cope with a <t>Both Classic and L4S services can cope with a
proportion of unresponsive or less-responsive traffic as well, but proportion of unresponsive or less-responsive traffic as well but,
in the L4S case its rate has to be smooth enough or low enough to in the L4S case, its rate has to be smooth enough or low enough to
not build a queue (e.g. DNS, VoIP, game sync datagrams, not build a queue (e.g., DNS, Voice over IP (VoIP), game sync datagram
s,
etc.).</t> etc.).</t>
</dd> </dd>
<dt>Reno-friendly:</dt> <dt>Reno-friendly:</dt>
<dd>The subset of Classic traffic that is <dd>The subset of Classic traffic that is
friendly to the standard Reno congestion control defined for TCP in friendly to the standard Reno congestion control defined for TCP in
<xref target="RFC5681" format="default"/>. The TFRC spec. <xref target ="RFC5348" format="default"/> indirectly implies that 'friendly' is defined as <xref target="RFC5681" format="default"/>. The TFRC spec <xref target= "RFC5348" format="default"/> indirectly implies that 'friendly' is defined as
"generally within a factor of two of the sending rate of a TCP flow "generally within a factor of two of the sending rate of a TCP flow
under the same conditions". Reno-friendly is used here in place of under the same conditions". Reno-friendly is used here in place of
'TCP-friendly', given the latter has become imprecise, because the 'TCP-friendly', given the latter has become imprecise, because the
TCP protocol is now used with so many different congestion control TCP protocol is now used with so many different congestion control
behaviours, and Reno is used in non-TCP transports such as behaviours, and Reno is used in non-TCP transports, such as
QUIC <xref target="RFC9000" format="default"/>.</dd> QUIC <xref target="RFC9000" format="default"/>.</dd>
<dt>Classic ECN:</dt> <dt>Classic ECN:</dt>
<dd> <dd>
<t>The original Explicit Congestion <t>The original Explicit Congestion
Notification (ECN) protocol <xref target="RFC3168" format="default"/>, which Notification (ECN) protocol <xref target="RFC3168" format="default"/> that
requires ECN signals to be treated as equivalent to drops, both when requires ECN signals to be treated as equivalent to drops, both when
generated in the network and when responded to by the sender.</t> generated in the network and when responded to by the sender.</t>
<t>L4S uses the ECN field as an identifier <xref target="I-D.ietf-tsvw <t>For L4S, the names used for the four codepoints of the 2-bit
g-ecn-l4s-id" format="default"/> with the names for the four IP-ECN field are unchanged from those defined in the ECN spec
codepoints of the 2-bit IP-ECN field unchanged from those defined in <xref target="RFC3168" format="default"/>, i.e., Not-ECT, ECT(0),
the ECN spec <xref target="RFC3168" format="default"/>: Not ECT, ECT(0 ECT(1), and CE, where ECT stands for ECN-Capable Transport and CE
), ECT(1) stands for Congestion Experienced. A packet marked with the CE
and CE, where ECT stands for ECN-Capable Transport and CE stands for codepoint is termed 'ECN-marked' or sometimes just 'marked' where
Congestion Experienced. A packet marked with the CE codepoint is the context makes ECN obvious.</t>
termed 'ECN-marked' or sometimes just 'marked' where the context
makes ECN obvious.</t>
</dd> </dd>
<dt>Site:</dt> <dt>Site:</dt>
<dd>A home, mobile device, small enterprise or <dd>A home, mobile device, small enterprise, or
campus, where the network bottleneck is typically the access link to campus where the network bottleneck is typically the access link to
the site. Not all network arrangements fit this model but it is a the site. Not all network arrangements fit this model, but it is a
useful, widely applicable generalization.</dd> useful, widely applicable generalization.</dd>
<dt>Traffic policing:</dt> <dt>Traffic Policing:</dt>
<dd>Limiting traffic by dropping packets <dd>Limiting traffic by dropping packets
or shifting them to lower service class (as opposed to introducing or shifting them to a lower service class (as opposed to introducing
delay, which is termed traffic shaping). Policing can involve delay, which is termed 'traffic shaping'). Policing can involve
limiting average rate and/or burst size. Policing focused on limiting the average rate and/or burst size. Policing focused on
limiting queuing but not average flow rate is termed congestion limiting queuing but not the average flow rate is termed 'congestion
policing, latency policing, burst policing or queue protection in policing', 'latency policing', 'burst policing', or 'queue protection'
in
this document. Otherwise, the term rate policing is used.</dd> this document. Otherwise, the term rate policing is used.</dd>
</dl> </dl>
</section> </section>
<section anchor="l4sps_components" numbered="true" toc="default"> <section anchor="l4sps_components" numbered="true" toc="default">
<name>L4S Architecture Components</name> <name>L4S Architecture Components</name>
<t>The L4S architecture is composed of the elements in the following <t>The L4S architecture is composed of the elements in the following
three subsections.</t> three subsections.</t>
<section anchor="l4sps_protocol_components" numbered="true" toc="default"> <section anchor="l4sps_protocol_components" numbered="true" toc="default">
<name>Protocol Mechanisms</name> <name>Protocol Mechanisms</name>
<t>The L4S architecture involves: a) unassignment of the previous use <t>The L4S architecture involves: a) unassignment of the previous use
of the identifier; b) reassignment of the same identifier; and c) of the identifier; b) reassignment of the same identifier; and c)
optional further identifiers:</t> optional further identifiers:</t>
<ol spacing="normal" type="a"><li> <ol spacing="normal" type="a"><li>
<t>An essential aspect of a scalable congestion control is the use <t>An essential aspect of a Scalable congestion control is the use
of explicit congestion signals. 'Classic' ECN <xref target="RFC3168" of explicit congestion signals. Classic ECN <xref target="RFC3168" f
format="default"/> requires an ECN signal to be treated as ormat="default"/> requires an ECN signal to be treated as
equivalent to drop, both when it is generated in the network and equivalent to drop, both when it is generated in the network and
when it is responded to by hosts. L4S needs networks and hosts to when it is responded to by hosts. L4S needs networks and hosts to
support a more fine-grained meaning for each ECN signal that is support a more fine-grained meaning for each ECN signal that is
less severe than a drop, so that the L4S signals:</t> less severe than a drop, so that the L4S signals:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>can be much more frequent;</li> <li>can be much more frequent and</li>
<li>can be signalled immediately, without the significant delay <li>can be signalled immediately, without the significant delay
required to smooth out fluctuations in the queue.</li> required to smooth out fluctuations in the queue.</li>
</ul> </ul>
<t>To enable L4S, the standards track Classic ECN <t>To enable L4S, the Standards Track Classic ECN
spec. <xref target="RFC3168" format="default"/> has had to be update spec <xref target="RFC3168" format="default"/> has had to be updated
d to allow to allow
L4S packets to depart from the 'equivalent to drop' constraint. L4S packets to depart from the 'equivalent-to-drop' constraint.
<xref target="RFC8311" format="default"/> is a standards track updat <xref target="RFC8311" format="default"/> is a Standards Track updat
e to relax e to
specific requirements in RFC 3168 (and certain other standards relax specific requirements in <xref target="RFC3168" format="default
track RFCs), which clears the way for the experimental changes "/>
(and certain other Standards
Track RFCs), which clears the way for the experimental changes
proposed for L4S. Also, the ECT(1) codepoint was previously proposed for L4S. Also, the ECT(1) codepoint was previously
assigned as the experimental ECN nonce <xref target="RFC3540" format ="default"/>, which RFC 8311 recategorizes as historic to assigned as the experimental ECN nonce <xref target="RFC3540" format ="default"/>, which <xref target="RFC8311" format="default"/> recategorizes as h istoric to
make the codepoint available again.</t> make the codepoint available again.</t>
</li> </li>
<li> <li>
<t><xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/> speci fies that <t><xref target="RFC9331" format="default"/> specifies that
ECT(1) is used as the identifier to classify L4S packets into a ECT(1) is used as the identifier to classify L4S packets into a
separate treatment from Classic packets. This satisfies the separate treatment from Classic packets. This satisfies the
requirement for identifying an alternative ECN treatment in <xref ta rget="RFC4774" format="default"/>.</t> requirement for identifying an alternative ECN treatment in <xref ta rget="RFC4774" format="default"/>.</t>
<t>The CE codepoint is <t>The CE codepoint is
used to indicate Congestion Experienced by both L4S and Classic used to indicate Congestion Experienced by both L4S and Classic
treatments. This raises the concern that a Classic AQM earlier on treatments. This raises the concern that a Classic AQM earlier on
the path might have marked some ECT(0) packets as CE. Then these the path might have marked some ECT(0) packets as CE. Then, these
packets will be erroneously classified into the L4S queue. packets will be erroneously classified into the L4S queue.
Appendix B of the L4S ECN spec <xref target="I-D.ietf-tsvwg-ecn-l4s- id" format="default"/> explains why five unlikely <xref target="RFC9331" format="default" section="B" sectionFormat="o f"/> explains why five unlikely
eventualities all have to coincide for this to have any eventualities all have to coincide for this to have any
detrimental effect, which even then would only involve a detrimental effect, which even then would only involve a
vanishingly small likelihood of a spurious retransmission.</t> vanishingly small likelihood of a spurious retransmission.</t>
</li> </li>
<li>A network operator might wish to include certain unresponsive, <li>A network operator might wish to include certain unresponsive,
non-L4S traffic in the L4S queue if it is deemed to be smoothly non-L4S traffic in the L4S queue if it is deemed to be paced smoothl
enough paced and low enough rate not to build a queue. For y
enough and at a low enough rate not to build a queue, for
instance, VoIP, low rate datagrams to sync online games, instance, VoIP, low rate datagrams to sync online games,
relatively low rate application-limited traffic, DNS, LDAP, etc. relatively low rate application-limited traffic, DNS, Lightweight Di rectory Access Protocol (LDAP), etc.
This traffic would need to be tagged with specific identifiers, This traffic would need to be tagged with specific identifiers,
e.g. a low latency Diffserv Codepoint such as Expedited e.g., a low-latency Diffserv codepoint such as Expedited
Forwarding (EF <xref target="RFC3246" format="default"/>), Non-Queue Forwarding (EF) <xref target="RFC3246" format="default"/>, Non-Queue
-Building -Building
(NQB <xref target="I-D.ietf-tsvwg-nqb" format="default"/>), or (NQB) <xref target="I-D.ietf-tsvwg-nqb" format="default"/>, or
operator-specific identifiers.</li> operator-specific identifiers.</li>
</ol> </ol>
</section> </section>
<section anchor="l4sps_network_components" numbered="true" toc="default"> <section anchor="l4sps_network_components" numbered="true" toc="default">
<name>Network Components</name> <name>Network Components</name>
<t>The L4S architecture aims to provide low latency without the <em>need </em> for per-flow operations in network <t>The L4S architecture aims to provide low latency without the <em>need </em> for per-flow operations in network
components. Nonetheless, the architecture does not preclude per-flow components. Nonetheless, the architecture does not preclude per-flow
solutions. The following bullets describe the known arrangements: a) solutions. The following bullets describe the known arrangements: a)
the DualQ Coupled AQM with an L4S AQM in one queue coupled from a the DualQ Coupled AQM with an L4S AQM in one queue coupled from a
Classic AQM in the other; b) Per-Flow Queues with an instance of a Classic AQM in the other; b) per-flow queues with an instance of a
Classic and an L4S AQM in each queue; c) Dual queues with per-flow Classic and an L4S AQM in each queue; and c) Dual queues with per-flow
AQMs, but no per-flow queues:</t> AQMs but no per-flow queues:</t>
<ol spacing="normal" type="a"><li> <ol spacing="normal" type="a"><li>
<t>The Dual Queue Coupled AQM (illustrated in <xref target="l4sps_fi g_components" format="default"/>) achieves the 'semi-permeable' <t>The Dual-Queue Coupled AQM (illustrated in <xref target="l4sps_fi g_components" format="default"/>) achieves the 'semi-permeable'
membrane property mentioned earlier as follows:</t> membrane property mentioned earlier as follows:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>Latency isolation: Two separate queues are used to isolate <li>Latency isolation: Two separate queues are used to isolate
L4S queuing delay from the larger queue that Classic traffic L4S queuing delay from the larger queue that Classic traffic
needs to maintain full utilization. <!--Each has its own AQM wit needs to maintain full utilization.</li>
h the L4S AQM configured for a very shallow (sub-millisecond) target delay, and
the Classic AQM for
5-15 ms, which is needed to absorb the larger saw-toothing pattern of Classic co
ngestion controls (otherwise they under-utilize
the link). -->
</li>
<li>Bandwidth pooling: The two queues act as if they are a <li>Bandwidth pooling: The two queues act as if they are a
single pool of bandwidth in which flows of either type get single pool of bandwidth in which flows of either type get
roughly equal throughput without the scheduler needing to roughly equal throughput without the scheduler needing to
identify any flows. This is achieved by having an AQM in each identify any flows. This is achieved by having an AQM in each
queue, but the Classic AQM provides a congestion signal to queue, but the Classic AQM provides a congestion signal to
both queues in a manner that ensures a consistent response both queues in a manner that ensures a consistent response
from the two classes of congestion control. Specifically, the from the two classes of congestion control. Specifically, the
Classic AQM generates a drop/mark probability based on Classic AQM generates a drop/mark probability based on
congestion in its own queue, which it uses both to drop/mark congestion in its own queue, which it uses both to drop/mark
packets in its own queue and to affect the marking probability packets in its own queue and to affect the marking probability
in the L4S queue. The strength of the coupling of the in the L4S queue. The strength of the coupling of the
congestion signalling between the two queues is enough to make congestion signalling between the two queues is enough to make
the L4S flows slow down to leave the right amount of capacity the L4S flows slow down to leave the right amount of capacity
for the Classic flows (as they would if they were the same for the Classic flows (as they would if they were the same
type of traffic sharing the same queue).</li> type of traffic sharing the same queue).</li>
</ul> </ul>
<t>Then the scheduler can serve the L4S queue with priority <t>Then, the scheduler can serve the L4S queue with priority
(denoted by the '1' on the higher priority input), because the L4S (denoted by the '1' on the higher priority input), because the L4S
traffic isn't offering up enough traffic to use all the priority traffic isn't offering up enough traffic to use all the priority
that it is given. Therefore:</t> that it is given. Therefore:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>for latency isolation on short time-scales (sub-round-trip) <li>for latency isolation on short timescales (sub-round-trip),
the prioritization of the L4S queue protects its low latency the prioritization of the L4S queue protects its low latency
by allowing bursts to dissipate quickly;</li> by allowing bursts to dissipate quickly;</li>
<li>but for bandwidth pooling on longer time-scales (round-trip <li>but for bandwidth pooling on longer timescales (round-trip
and longer) the Classic queue creates an equal and opposite and longer), the Classic queue creates an equal and opposite
pressure against the L4S traffic to ensure that neither has pressure against the L4S traffic to ensure that neither has
priority when it comes to bandwidth - the tension between priority when it comes to bandwidth -- the tension between
prioritizing L4S and coupling the marking from the Classic AQM prioritizing L4S and coupling the marking from the Classic AQM
results in approximate per-flow fairness.</li> results in approximate per-flow fairness.</li>
</ul> </ul>
<t>To protect against unresponsive traffic taking advantage <t>To protect against the prioritization of persistent L4S traffic
of the prioritization of the L4S queue and starving the Classic deadlocking the Classic queue for a while in some implementations,
queue, it is advisable for the priority to be conditional, not it is advisable for the priority to be conditional, not
strict (see Appendix A of the DualQ spec <xref target="I-D.ietf-tsvw strict (see <xref target="RFC9332" format="default" section="A" sect
g-aqm-dualq-coupled" format="default"/>). </t> ionFormat="of">the DualQ spec</xref>). </t>
<t>When there is no Classic traffic, the L4S <t>When there is no Classic traffic, the L4S
queue's own AQM comes into play. It starts congestion queue's own AQM comes into play. It starts congestion
marking with a very shallow queue, so L4S traffic maintains very marking with a very shallow queue, so L4S traffic maintains very
low queuing delay.</t> low queuing delay.</t>
<t>If either queue becomes <t>If either queue becomes persistently overloaded, drop of some
persistently overloaded, drop of ECN-capable packets is ECN-capable packets is introduced, as recommended in <xref
introduced, as recommended in Section 7 of the ECN spec <xref target target="RFC3168" sectionFormat="of" section="7">the ECN
="RFC3168" format="default"/> and Section 4.2.1 of the AQM spec</xref> and <xref target="RFC7567" sectionFormat="of"
recommendations <xref target="RFC7567" format="default"/>. Then both section="4.2.1">the AQM recommendations</xref>. The trade-offs with
queues different approaches
introduce the same level of drop (not shown in the figure).</t> are discussed in <xref
<t>The Dual Queue Coupled AQM has been specified as target="RFC9332" sectionFormat="of" section="4.2.3">the DualQ
generically as possible <xref target="I-D.ietf-tsvwg-aqm-dualq-coupl spec</xref> (not shown in the figure here).</t>
ed" format="default"/> without specifying the <t>The Dual-Queue Coupled AQM has been specified as
generically as possible <xref target="RFC9332" format="default"/> wi
thout specifying the
particular AQMs to use in the two queues so that designers are particular AQMs to use in the two queues so that designers are
free to implement diverse ideas. Informational appendices in that free to implement diverse ideas. Informational appendices in that
draft give pseudocode examples of two different specific AQM document give pseudocode examples of two different specific AQM
approaches: one called DualPI2 (pronounced Dual PI approaches: one called DualPI2 (pronounced Dual PI
Squared) <xref target="DualPI2Linux" format="default"/> that uses th Squared) <xref target="DualPI2Linux" format="default"/> that uses th
e PI2 e PI2
variant of PIE, and a zero-config variant of RED called Curvy RED. variant of PIE and a zero-config variant of Random Early Detection (
RED) called Curvy RED.
A DualQ Coupled AQM based on PIE has also been specified and A DualQ Coupled AQM based on PIE has also been specified and
implemented for Low Latency DOCSIS <xref target="DOCSIS3.1" format=" default"/>.</t> implemented for Low Latency DOCSIS <xref target="DOCSIS3.1" format=" default"/>.</t>
<figure anchor="l4sps_fig_components"> <figure anchor="l4sps_fig_components">
<name>Components of an L4S DualQ Coupled AQM Solution: 1) Scalable <name>Components of an L4S DualQ Coupled AQM Solution</name>
Sending Host; 2) Isolation in separate network queues; and 3) Packet Identifica <artwork align="center" name="" type="" alt=""><![CDATA[
tion Protocol</name> (3) (2)
<artwork align="center" name="" type="" alt=""><![CDATA[
(3) (2)
.-------^------..------------^------------------. .-------^------..------------^------------------.
,-(1)-----. _____ ,-(1)-----. _____
; ________ : L4S -------. | | ; ________ : L4S -------. | |
:|Scalable| : _\ ||__\_|mark | :|Scalable| : _\ ||__\_|mark |
:| sender | : __________ / / || / |_____|\ _________ :| sender | : __________ / / || / |_____|\ _________
:|________|\; | |/ -------' ^ \1|condit'nl| :|________|\; | |/ -------' ^ \1|condit'nl|
`---------'\_| IP-ECN | Coupling : \|priority |_\ `---------'\_| IP-ECN | Coupling : \|priority |_\
________ / |Classifier| : /|scheduler| / ________ / |Classifier| : /|scheduler| /
|Classic |/ |__________|\ -------. __:__ / |_________| |Classic |/ |__________|\ -------. __:__ / |_________|
| sender | \_\ || | ||__\_|mark/|/ | sender | \_\ || | ||__\_|mark/|/
|________| / || | || / |drop | |________| / || | || / |drop |
Classic -------' |_____| Classic -------' |_____|
(1) Scalable sending host
(2) Isolation in separate network queues
(3) Packet identification protocol
]]></artwork> ]]></artwork>
</figure> </figure>
</li> </li>
<li>Per-Flow Queues and AQMs: A scheduler with per-flow queues such <li>Per-Flow Queues and AQMs: A scheduler with per-flow queues, such
as FQ-CoDel or FQ-PIE can be used for L4S. For instance within as FQ-CoDel or FQ-PIE, can be used for L4S. For instance, within
each queue of an FQ-CoDel system, as well as a CoDel AQM, there is each queue of an FQ-CoDel system, as well as a CoDel AQM, there is
typically also the option of ECN marking at an immediate typically also the option of ECN marking at an immediate
(unsmoothed) shallow threshold to support use in data centres (see (unsmoothed) shallow threshold to support use in data centres (see
Sec.5.2.7 of the FQ-CoDel spec <xref target="RFC8290" format="defaul t"/>). In <xref target="RFC8290" sectionFormat="of" section="5.2.7">the FQ-CoD el spec</xref>). In
Linux, this has been modified so that the shallow threshold can be Linux, this has been modified so that the shallow threshold can be
solely applied to ECT(1) packets <xref target="FQ_CoDel_Thresh" form solely applied to ECT(1) packets <xref target="FQ_CoDel_Thresh" form
at="default"/>. Then, if there is a flow of non-ECN or at="default"/>. Then, if there is a flow of Not-ECT or
ECT(0) packets in the per-flow-queue, the Classic AQM ECT(0) packets in the per-flow queue, the Classic AQM
(e.g. CoDel) is applied; while if there is a flow of ECT(1) (e.g., CoDel) is applied; whereas, if there is a flow of ECT(1)
packets in the queue, the shallower (typically sub-millisecond) packets in the queue, the shallower (typically sub-millisecond)
threshold is applied. In addition, ECT(0) and not-ECT packets threshold is applied.
could potentially be classified into a separate flow-queue from In addition, ECT(0) and Not-ECT packets
could potentially be classified into a separate flow queue from
ECT(1) and CE packets to avoid them mixing if they share a common ECT(1) and CE packets to avoid them mixing if they share a common
flow-identifier (e.g. in a VPN).</li> flow identifier (e.g., in a VPN).</li>
<li> <li>
<t>Dual-queues, but per-flow AQMs: It should also be possible to <t>Dual queues but per-flow AQMs: It should also be possible to
use dual queues for isolation, but with per-flow marking to use dual queues for isolation but with per-flow marking to
control flow-rates (instead of the coupled per-queue marking of control flow rates (instead of the coupled per-queue marking of
the Dual Queue Coupled AQM). One of the two queues would be for the Dual-Queue Coupled AQM). One of the two queues would be for
isolating L4S packets, which would be classified by the ECN isolating L4S packets, which would be classified by the ECN
codepoint. Flow rates could be controlled by flow-specific codepoint. Flow rates could be controlled by flow-specific
marking. The policy goal of the marking could be to differentiate marking. The policy goal of the marking could be to differentiate
flow rates (e.g. <xref target="Nadas20" format="default"/>, which re flow rates (e.g., <xref target="Nadas20" format="default"/>, which r
quires equires
additional signalling of a per-flow 'value'), or to equalize additional signalling of a per-flow 'value') or to equalize
flow-rates (perhaps in a similar way to Approx Fair flow rates (perhaps in a similar way to Approx Fair
CoDel <xref target="AFCD" format="default"/>, <xref target="I-D.mort CoDel <xref target="AFCD" format="default"/> <xref target="I-D.morto
on-tsvwg-codel-approx-fair" format="default"/>, but with two queues n-tsvwg-codel-approx-fair" format="default"/> but with two queues
not one).</t> not one).</t>
<t>Note that whenever the term <t>Note that, whenever the term
'DualQ' is used loosely without saying whether marking is 'DualQ' is used loosely without saying whether marking is
per-queue or per-flow, it means a dual queue AQM with per-queue per queue or per flow, it means a dual-queue AQM with per-queue
marking.</t> marking.</t>
</li> </li>
</ol> </ol>
</section> </section>
<section anchor="l4sps_host_components" numbered="true" toc="default"> <section anchor="l4sps_host_components" numbered="true" toc="default">
<name>Host Mechanisms</name> <name>Host Mechanisms</name>
<t>The L4S architecture includes two main mechanisms in the end host <t>The L4S architecture includes two main mechanisms in the end host
that we enumerate next:</t> that we enumerate next:</t>
<ol spacing="normal" type="a"><li> <ol spacing="normal" type="a"><li>
<t>Scalable Congestion Control at the sender: <xref target="l4s-arch _arch_overview" format="default"/> defines a scalable congestion <t>Scalable congestion control at the sender: <xref target="l4s-arch _arch_overview" format="default"/> defines a Scalable congestion
control as one where the average time from one congestion signal control as one where the average time from one congestion signal
to the next (the recovery time) remains invariant as the flow rate to the next (the recovery time) remains invariant as flow rate
scales, all other factors being equal. Data Center TCP is the most scales, all other factors being equal. DCTCP is the most
widely used example. It has been documented as an informational widely used example. It has been documented as an informational
record of the protocol currently in use in controlled record of the protocol currently in use in controlled
environments <xref target="RFC8257" format="default"/>. A draft list environments <xref target="RFC8257" format="default"/>. A list of sa
of safety fety
and performance improvements for a scalable congestion control to and performance improvements for a Scalable congestion control to
be usable on the public Internet has been drawn up (the so-called be usable on the public Internet has been drawn up (see the so-calle
'Prague L4S requirements' in Appendix A of <xref target="I-D.ietf-ts d
vwg-ecn-l4s-id" format="default"/>). The subset that involve 'Prague L4S requirements' in <xref target="RFC9331" format="default"
sectionFormat="of" section="A"/>).
The subset that involve
risk of harm to others have been captured as normative risk of harm to others have been captured as normative
requirements in Section 4 of <xref target="I-D.ietf-tsvwg-ecn-l4s-id " format="default"/>. TCP Prague <xref target="I-D.briscoe-iccrg-prague-congesti on-control" format="default"/> has been requirements in <xref target="RFC9331" format="default" sectionForma t="of" section="4"/>. TCP Prague <xref target="I-D.briscoe-iccrg-prague-congesti on-control" format="default"/> has been
implemented in Linux as a reference implementation to address implemented in Linux as a reference implementation to address
these requirements <xref target="PragueLinux" format="default"/>.</t > these requirements <xref target="PragueLinux" format="default"/>.</t >
<t>Transport protocols other than TCP use various <t>Transport protocols other than TCP use various
congestion controls that are designed to be friendly with Reno. congestion controls that are designed to be friendly with Reno.
Before they can use the L4S service, they will need to be updated Before they can use the L4S service, they will need to be updated
to implement a scalable congestion response, which they will have to implement a Scalable congestion response, which they will have
to indicate by using the ECT(1) codepoint. Scalable variants are to indicate by using the ECT(1) codepoint. Scalable variants are
under consideration for more recent transport protocols, under consideration for more recent transport protocols
e.g. QUIC, and the L4S ECN part of BBRv2 <xref target="BBRv2" format (e.g., QUIC), and the L4S ECN part of BBRv2 <xref target="BBRv2" for
="default"/>, <xref target="I-D.cardwell-iccrg-bbr-congestion-control" format="d mat="default"/> <xref target="I-D.cardwell-iccrg-bbr-congestion-control" format=
efault"/> is a scalable "default"/> is a Scalable
congestion control intended for the TCP and QUIC transports, congestion control intended for the TCP and QUIC transports,
amongst others. Also, an L4S variant of the RMCAT SCReAM amongst others. Also, an L4S variant of the RMCAT SCReAM
controller <xref target="RFC8298" format="default"/> has been controller <xref target="RFC8298" format="default"/> has been
implemented <xref target="SCReAM" format="default"/> for media trans implemented <xref target="SCReAM-L4S" format="default"/> for media t
ported ransported
over RTP.</t> over RTP.</t>
<t>Section 4.3 of the L4S ECN <t> <xref target="RFC9331" format="default" sectionFormat="of" secti
spec <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/> def on="4.3">the L4S ECN spec</xref> defines
ines Scalable congestion control in more detail and specifies the
scalable congestion control in more detail, and specifies the requirements that an L4S Scalable congestion control has to comply
requirements that an L4S scalable congestion control has to comply
with.</t> with.</t>
</li> </li>
<li> <li>
<t>The ECN feedback in some transport protocols is already <t>The ECN feedback in some transport protocols is already
sufficiently fine-grained for L4S (specifically DCCP <xref target="R sufficiently fine-grained for L4S (specifically DCCP <xref target="R
FC4340" format="default"/> and QUIC <xref target="RFC9000" format="default"/>). FC4340" format="default"/> and QUIC <xref target="RFC9000" format="default"/>).
But But
others either require update or are in the process of being others either require updates or are in the process of being
updated:</t> updated:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>For the case of TCP, the feedback protocol for ECN embeds <li>For the case of TCP, the feedback protocol for ECN embeds
the assumption from Classic ECN <xref target="RFC3168" format="d efault"/> the assumption from Classic ECN <xref target="RFC3168" format="d efault"/>
that an ECN mark is equivalent to a drop, making it unusable that an ECN mark is equivalent to a drop, making it unusable
for a scalable TCP. Therefore, the implementation of TCP for a Scalable TCP. Therefore, the implementation of TCP
receivers will have to be upgraded <xref target="RFC7560" format receivers will have to be upgraded <xref target="RFC7560" format
="default"/>. Work to standardize and implement more ="default"/>.
Work to standardize and implement more
accurate ECN feedback for TCP (AccECN) is in accurate ECN feedback for TCP (AccECN) is in
progress <xref target="I-D.ietf-tcpm-accurate-ecn" format="defau lt"/>, progress <xref target="I-D.ietf-tcpm-accurate-ecn" format="defau lt"/>
<xref target="PragueLinux" format="default"/>.</li> <xref target="PragueLinux" format="default"/>.</li>
<li>ECN feedback was only roughly sketched in an appendix of <li>ECN feedback was only roughly sketched in the appendix of
the now obsoleted second specification of SCTP <xref target="RFC 4960" format="default"/>, while a fuller specification was proposed the now obsoleted second specification of SCTP <xref target="RFC 4960" format="default"/>, while a fuller specification was proposed
in a long-expired draft <xref target="I-D.stewart-tsvwg-sctpecn" format="default"/>. A new design would need in a long-expired document <xref target="I-D.stewart-tsvwg-sctpe cn" format="default"/>. A new design would need
to be implemented and deployed before SCTP could support to be implemented and deployed before SCTP could support
L4S.</li> L4S.</li>
<li>For RTP, sufficient ECN feedback was defined in <xref target=" <li>For RTP, sufficient ECN feedback was defined in <xref target="
RFC6679" format="default"/>, but <xref target="RFC8888" format="default"/> defin RFC6679" format="default"/>, but <xref target="RFC8888" format="default"/> defin
es the es the
latest standards track improvements.</li> latest Standards Track improvements.</li>
</ul> </ul>
</li> </li>
</ol> </ol>
</section> </section>
</section> </section>
<section anchor="l4sps_rationale" numbered="true" toc="default"> <section anchor="l4sps_rationale" numbered="true" toc="default">
<name>Rationale</name> <name>Rationale</name>
<t/>
<section anchor="l4sps_why_primary_components" numbered="true" toc="defaul t"> <section anchor="l4sps_why_primary_components" numbered="true" toc="defaul t">
<name>Why These Primary Components?</name> <name>Why These Primary Components?</name>
<dl newline="false" spacing="normal"> <dl newline="false" spacing="normal">
<dt>Explicit congestion signalling (protocol):</dt> <dt>Explicit congestion signalling (protocol):</dt>
<dd> <dd>
<t>Explicit <t>Explicit
congestion signalling is a key part of the L4S approach. In congestion signalling is a key part of the L4S approach. In
contrast, use of drop as a congestion signal creates a tension contrast, use of drop as a congestion signal creates tension
because drop is both an impairment (less would be better) and a because drop is both an impairment (less would be better) and a
useful signal (more would be better):</t> useful signal (more would be better):</t>
<ul spacing="normal"> <ul spacing="normal">
<li>Explicit congestion signals can be used many times per <li>Explicit congestion signals can be used many times per
round trip, to keep tight control, without any impairment. round trip to keep tight control without any impairment.
Under heavy load, even more explicit signals can be applied, Under heavy load, even more explicit signals can be applied
so that the queue can be kept short whatever the load. In so that the queue can be kept short whatever the load. In
contrast, Classic AQMs have to introduce very high packet drop contrast, Classic AQMs have to introduce very high packet drop
at high load to keep the queue short. By using ECN, an L4S at high load to keep the queue short. By using ECN, an L4S
congestion control's sawtooth reduction can be smaller and congestion control's sawtooth reduction can be smaller and
therefore return to the operating point more often, without therefore return to the operating point more often, without
worrying that more sawteeth will cause more signals. The worrying that more sawteeth will cause more signals. The
consequent smaller amplitude sawteeth fit between an empty consequent smaller amplitude sawteeth fit between an empty
queue and a very shallow marking threshold (~1 ms in the queue and a very shallow marking threshold (~1 ms in the
public Internet), so queue delay variation can be very low, public Internet), so queue delay variation can be very low,
without risk of under-utilization.</li> without risk of underutilization.</li>
<li>Explicit congestion signals can be emitted immediately to <li>Explicit congestion signals can be emitted immediately to
track fluctuations of the queue. L4S shifts smoothing from the track fluctuations of the queue. L4S shifts smoothing from the
network to the host. The network doesn't know the round trip network to the host. The network doesn't know the round-trip
times of any of the flows. So if the network is responsible times (RTTs) of any of the flows. So if the network is responsib
le
for smoothing (as in the Classic approach), it has to assume a for smoothing (as in the Classic approach), it has to assume a
worst case RTT, otherwise long RTT flows would become worst case RTT, otherwise long RTT flows would become
unstable. This delays Classic congestion signals by 100-200 unstable. This delays Classic congestion signals by 100-200
ms. In contrast, each host knows its own round trip time. So, ms. In contrast, each host knows its own RTT. So,
in the L4S approach, the host can smooth each flow over its in the L4S approach, the host can smooth each flow over its
own RTT, introducing no more smoothing delay than strictly own RTT, introducing no more smoothing delay than strictly
necessary (usually only a few milliseconds). A host can also necessary (usually only a few milliseconds). A host can also
choose not to introduce any smoothing delay if appropriate, choose not to introduce any smoothing delay if appropriate,
e.g. during flow start-up.</li> e.g., during flow start-up.</li>
</ul> </ul>
<t>Neither of the above are feasible if explicit congestion <t>Neither of the above are feasible if explicit congestion
signalling has to be considered 'equivalent to drop' (as was signalling has to be considered 'equivalent to drop' (as was
required with Classic ECN <xref target="RFC3168" format="default"/>) , because required with Classic ECN <xref target="RFC3168" format="default"/>) , because
drop is an impairment as well as a signal. So drop cannot be drop is an impairment as well as a signal. So drop cannot be
excessively frequent, and drop cannot be immediate, otherwise too excessively frequent, and drop cannot be immediate; otherwise, too
many drops would turn out to have been due to only a transient many drops would turn out to have been due to only a transient
fluctuation in the queue that would not have warranted dropping a fluctuation in the queue that would not have warranted dropping a
packet in hindsight. Therefore, in an L4S AQM, the L4S queue uses packet in hindsight. Therefore, in an L4S AQM, the L4S queue uses
a new L4S variant of ECN that is not equivalent to drop (see a new L4S variant of ECN that is not equivalent to drop (see
section 5.2 of the L4S ECN spec <xref target="I-D.ietf-tsvwg-ecn-l4s <xref target="RFC9331" format="default" sectionFormat="of" section="
-id" format="default"/>), while the Classic queue 5.2">the L4S ECN spec</xref>), while the Classic queue
uses either Classic ECN <xref target="RFC3168" format="default"/> or uses either Classic ECN <xref target="RFC3168" format="default"/> or
drop, drop,
which are equivalent to each other.</t> which are still equivalent to each other.</t>
<t>Before <t>Before
Classic ECN was standardized, there were various proposals to give Classic ECN was standardized, there were various proposals to give
an ECN mark a different meaning from drop. However, there was no an ECN mark a different meaning from drop. However, there was no
particular reason to agree on any one of the alternative meanings, particular reason to agree on any one of the alternative meanings,
so 'equivalent to drop' was the only compromise that could be so 'equivalent to drop' was the only compromise that could be
reached. RFC 3168 contains a statement that:</t> reached. <xref target="RFC3168" format="default"/> contains a statem
<ul empty="true" spacing="normal"> ent that:</t>
<li>"An environment where all end nodes were ECN-Capable could <ul empty="true">
allow new criteria to be developed for setting the CE <li><t indent="1">An environment where all end nodes were
codepoint, and new congestion control mechanisms for end-node ECN-Capable could allow new criteria to be developed for
reaction to CE packets. However, this is a research issue, and setting the CE codepoint, and new congestion control
as such is not addressed in this document."</li> mechanisms for end-node reaction to CE packets. However, this
</ul> is a research issue, and as such is not addressed in this
document.</t></li></ul>
</dd> </dd>
<dt>Latency isolation (network):</dt> <dt>Latency isolation (network):</dt>
<dd>L4S congestion controls <dd>L4S congestion controls
keep queue delay low whereas Classic congestion controls need a keep queue delay low, whereas Classic congestion controls need a
queue of the order of the RTT to avoid under-utilization. One queue of the order of the RTT to avoid underutilization. One
queue cannot have two lengths, therefore L4S traffic needs to be queue cannot have two lengths; therefore, L4S traffic needs to be
isolated in a separate queue (e.g. DualQ) or queues isolated in a separate queue (e.g., DualQ) or queues
(e.g. FQ).</dd> (e.g., FQ).</dd>
<dt>Coupled congestion notification:</dt> <dt>Coupled congestion notification:</dt>
<dd>Coupling the <dd>Coupling the
congestion notification between two queues as in the DualQ Coupled congestion notification between two queues as in the DualQ Coupled
AQM is not necessarily essential, but it is a simple way to allow AQM is not necessarily essential, but it is a simple way to allow
senders to determine their rate, packet by packet, rather than be senders to determine their rate packet by packet, rather than be
overridden by a network scheduler. An alternative is for a network overridden by a network scheduler. An alternative is for a network
scheduler to control the rate of each application flow (see scheduler to control the rate of each application flow (see the
discussion in <xref target="l4sps_why-not" format="default"/>).</dd> discussion in <xref target="l4sps_why-not" format="default"/>).</dd>
<dt>L4S packet identifier (protocol):</dt> <dt>L4S packet identifier (protocol):</dt>
<dd>Once there are at <dd>Once there are at
least two treatments in the network, hosts need an identifier at least two treatments in the network, hosts need an identifier at
the IP layer to distinguish which treatment they intend to the IP layer to distinguish which treatment they intend to
use.</dd> use.</dd>
<dt>Scalable congestion notification:</dt> <dt>Scalable congestion notification:</dt>
<dd>A scalable <dd>A Scalable
congestion control in the host keeps the signalling frequency from congestion control in the host keeps the signalling frequency from
the network high whatever the flow rate, so that queue delay the network high, whatever the flow rate, so that queue delay
variations can be small when conditions are stable, and rate can variations can be small when conditions are stable, and rate can
track variations in available capacity as rapidly as possible track variations in available capacity as rapidly as possible
otherwise.</dd> otherwise.</dd>
<dt>Low loss:</dt> <dt>Low loss:</dt>
<dd>Latency is not the only concern of L4S. <dd>Latency is not the only concern of L4S.
The 'Low Loss' part of the name denotes that L4S generally The 'Low Loss' part of the name denotes that L4S generally
achieves zero congestion loss due to its use of ECN. Otherwise, achieves zero congestion loss due to its use of ECN. Otherwise,
loss would itself cause delay, particularly for short flows, due loss would itself cause delay, particularly for short flows, due
to retransmission delay <xref target="RFC2884" format="default"/>.</ dd> to retransmission delay <xref target="RFC2884" format="default"/>.</ dd>
<dt>Scalable throughput:</dt> <dt>Scalable throughput:</dt>
<dd> <dd>
<t>The "Scalable throughput" part <t>The 'Scalable throughput' part
of the name denotes that the per-flow throughput of scalable of the name denotes that the per-flow throughput of Scalable
congestion controls should scale indefinitely, avoiding the congestion controls should scale indefinitely, avoiding the
imminent scaling problems with Reno-friendly congestion control imminent scaling problems with Reno-friendly congestion control
algorithms <xref target="RFC3649" format="default"/>. It was known w hen TCP algorithms <xref target="RFC3649" format="default"/>. It was known w hen TCP
congestion avoidance was first developed in 1988 that it would not congestion avoidance was first developed in 1988 that it would not
scale to high bandwidth-delay products (see footnote 6 in <xref targ et="TCP-CA" format="default"/>). Today, regular broadband flow rates over WAN scale to high bandwidth-delay products (see footnote 6 in <xref targ et="TCP-CA" format="default"/>). Today, regular broadband flow rates over WAN
distances are already beyond the scaling range of Classic Reno distances are already beyond the scaling range of Classic Reno
congestion control. So `less unscalable' Cubic <xref target="RFC8312 " format="default"/> and Compound <xref target="I-D.sridharan-tcpm-ctcp" format= "default"/> variants of TCP have been congestion control. So 'less unscalable' CUBIC <xref target="RFC8312 " format="default"/> and Compound <xref target="I-D.sridharan-tcpm-ctcp" format= "default"/> variants of TCP have been
successfully deployed. However, these are now approaching their successfully deployed. However, these are now approaching their
scaling limits. </t> scaling limits. </t>
<t>For instance, we will <t>For instance, we will
consider a scenario with a maximum RTT of 30 ms at the peak consider a scenario with a maximum RTT of 30 ms at the peak
of each sawtooth. As Reno packet rate scales 8x from 1,250 to of each sawtooth. As Reno packet rate scales 8 times from 1,250 to
10,000 packet/s (from 15 to 120 Mb/s with 1500 B 10,000 packet/s (from 15 to 120 Mb/s with 1500 B
packets), the time to recover from a congestion event rises packets), the time to recover from a congestion event rises
proportionately by 8x as well, from 422 ms to 3.38 s. It proportionately by 8 times as well, from 422 ms to 3.38 s. It
is clearly problematic for a congestion control to take multiple is clearly problematic for a congestion control to take multiple
seconds to recover from each congestion event. Cubic <xref target="R FC8312" format="default"/> was developed to be less unscalable, but it is seconds to recover from each congestion event. CUBIC <xref target="R FC8312" format="default"/> was developed to be less unscalable, but it is
approaching its scaling limit; with the same max RTT of approaching its scaling limit; with the same max RTT of
30 ms, at 120 Mb/s Cubic is still fully in its 30 ms, at 120 Mb/s, CUBIC is still fully in its
Reno-friendly mode, so it takes about 4.3 s to recover. Reno-friendly mode, so it takes about 4.3 s to recover.
However, once the flow rate scales by 8x again to 960 Mb/s it However, once flow rate scales by 8 times again to 960 Mb/s it
enters true Cubic mode, with a recovery time of 12.2 s. From enters true CUBIC mode, with a recovery time of 12.2 s. From
then on, each further scaling by 8x doubles Cubic's recovery time then on, each further scaling by 8 times doubles CUBIC's recovery ti
(because the cube root of 8 is 2), e.g. at 7.68 Gb/s the me
recovery time is 24.3 s. In contrast, a scalable congestion (because the cube root of 8 is 2), e.g., at 7.68 Gb/s, the
control like DCTCP or TCP Prague induces 2 congestion signals per recovery time is 24.3 s. In contrast, a Scalable congestion
control like DCTCP or Prague induces 2 congestion signals per
round trip on average, which remains invariant for any flow rate, round trip on average, which remains invariant for any flow rate,
keeping dynamic control very tight.</t> keeping dynamic control very tight.</t>
<t>For a <t>For a
feel of where the global average lone-flow download sits on this feel of where the global average lone-flow download sits on this
scale at the time of writing (2021), according to <xref target="BDPd scale at the time of writing (2021), according to <xref target="BDPd
ata" format="default"/> globally averaged fixed access capacity was 103 ata" format="default"/>, the global average fixed access capacity was 103
Mb/s in 2020 and averaged base RTT to a CDN was 25-34ms in 2019. Mb/s in 2020 and the average base RTT to a CDN was 25 to 34 ms in 20
19.
Averaging of per-country data was weighted by Internet user Averaging of per-country data was weighted by Internet user
population (data collected globally is necessarily of variable population (data collected globally is necessarily of variable
quality, but the paper does double-check that the outcome compares quality, but the paper does double-check that the outcome compares
well against a second source). So a lone CUBIC flow would at best well against a second source). So a lone CUBIC flow would at best
take about 200 round trips (5 s) to recover from each of its take about 200 round trips (5 s) to recover from each of its
sawtooth reductions, if the flow even lasted that long. This is sawtooth reductions, if the flow even lasted that long. This is
described as 'at best' because it assumes everyone uses an AQM, described as 'at best' because it assumes everyone uses an AQM,
whereas in reality most users still have a (probably bloated) whereas in reality, most users still have a (probably bloated)
tail-drop buffer. In the tail-drop case, likely average recovery tail-drop buffer.
time would be at least 4x 5 s, if not more, because RTT under load In the tail-drop case, the likely average recovery
would be at least double that of an AQM, and recovery time depends time would be at least 4 times 5 s, if not more, because RTT under l
oad
would be at least double that of an AQM, and the recovery time of Re
no-friendly flows depends
on the square of RTT.</t> on the square of RTT.</t>
<t>Although work on <t>Although work on
scaling congestion controls tends to start with TCP as the scaling congestion controls tends to start with TCP as the
transport, the above is not intended to exclude other transports transport, the above is not intended to exclude other transports
(e.g. SCTP, QUIC) or less elastic algorithms (e.g., SCTP and QUIC) or less elastic algorithms
(e.g. RMCAT), which all tend to adopt the same or similar (e.g., RMCAT), which all tend to adopt the same or similar
developments.</t> developments.</t>
</dd> </dd>
</dl> </dl>
</section> </section>
<section anchor="l4sps_why-not" numbered="true" toc="default"> <section anchor="l4sps_why-not" numbered="true" toc="default">
<name>What L4S adds to Existing Approaches</name> <name>What L4S Adds to Existing Approaches</name>
<t>All the following approaches address some part of the same problem <t>All the following approaches address some part of the same problem
space as L4S. In each case, it is shown that L4S complements them or space as L4S. In each case, it is shown that L4S complements them or
improves on them, rather than being a mutually exclusive improves on them, rather than being a mutually exclusive
alternative:</t> alternative:</t>
<dl newline="false" spacing="normal"> <dl newline="false" spacing="normal">
<dt>Diffserv:</dt> <dt>Diffserv:</dt>
<dd> <dd>
<t>Diffserv addresses the problem of <t>Diffserv addresses the problem of
bandwidth apportionment for important traffic as well as queuing bandwidth apportionment for important traffic as well as queuing
latency for delay-sensitive traffic. Of these, L4S solely latency for delay-sensitive traffic. Of these, L4S solely
addresses the problem of queuing latency. Diffserv will still be addresses the problem of queuing latency. Diffserv will still be
necessary where important traffic requires priority (e.g. for necessary where important traffic requires priority (e.g., for
commercial reasons, or for protection of critical infrastructure commercial reasons or for protection of critical infrastructure
traffic) - see <xref target="I-D.briscoe-tsvwg-l4s-diffserv" format= traffic) -- see <xref target="I-D.briscoe-tsvwg-l4s-diffserv" format
"default"/>. ="default"/>.
Nonetheless, the L4S approach can provide low latency for all Nonetheless, the L4S approach can provide low latency for all
traffic within each Diffserv class (including the case where there traffic within each Diffserv class (including the case where there
is only the one default Diffserv class).</t> is only the one default Diffserv class).</t>
<t>Also, Diffserv can only provide a latency benefit <t>Also, Diffserv can only provide a latency benefit
if a small subset of the traffic on a bottleneck link requests low if a small subset of the traffic on a bottleneck link requests low
latency. As already explained, it has no effect when all the latency. As already explained, it has no effect when all the
applications in use at one time at a single site (home, small applications in use at one time at a single site (e.g., a home, smal
business or mobile device) require low latency. In contrast, l
business, or mobile device) require low latency. In contrast,
because L4S works for all traffic, it needs none of the management because L4S works for all traffic, it needs none of the management
baggage (traffic policing, traffic contracts) associated with baggage (traffic policing or traffic contracts) associated with
favouring some packets over others. This lack of management favouring some packets over others. This lack of management
baggage ought to give L4S a better chance of end-to-end baggage ought to give L4S a better chance of end-to-end
deployment.</t> deployment.</t>
<t>In particular, because networks
tend not to trust end systems to identify which packets should be <t>In particular, if networks do not trust end systems to identify w
favoured over others, where networks assign packets to Diffserv hich
classes they tend to use packet inspection of application flow packets should be favoured, they assign packets to Diffserv classes
identifiers or deeper inspection of application signatures. Thus, themselves. However, the techniques available to such networks, like
nowadays, Diffserv doesn't always sit well with encryption of the inspection of flow identifiers or deeper inspection of application
layers above IP <xref target="RFC8404" format="default"/>. So users signatures, do not always sit well with encryption of the layers abo
have to choose ve
between privacy and QoS.</t> IP <xref target="RFC8404" format="default"/>. In these cases, users
can have either privacy or Quality of Service (QoS), but not both.</
t>
<t>As with Diffserv, <t>As with Diffserv,
the L4S identifier is in the IP header. But, in contrast to the L4S identifier is in the IP header. But, in contrast to
Diffserv, the L4S identifier does not convey a want or a need for Diffserv, the L4S identifier does not convey a want or a need for
a certain level of quality. Rather, it promises a certain a certain level of quality. Rather, it promises a certain
behaviour (scalable congestion response), which networks can behaviour (Scalable congestion response), which networks can
objectively verify if they need to. This is because low delay objectively verify if they need to. This is because low delay
depends on collective host behaviour, whereas bandwidth priority depends on collective host behaviour, whereas bandwidth priority
depends on network behaviour.</t> depends on network behaviour.</t>
</dd> </dd>
<dt>State-of-the-art AQMs:</dt> <dt>State-of-the-art AQMs:</dt>
<dd>AQMs such as PIE and FQ-CoDel <dd>AQMs for Classic traffic, such as PIE and FQ-CoDel,
give a significant reduction in queuing delay relative to no AQM give a significant reduction in queuing delay relative to no AQM
at all. L4S is intended to complement these AQMs, and should not at all. L4S is intended to complement these AQMs and should not
distract from the need to deploy them as widely as possible. distract from the need to deploy them as widely as possible.
Nonetheless, AQMs alone cannot reduce queuing delay too far Nonetheless, AQMs alone cannot reduce queuing delay too far
without significantly reducing link utilization, because the root without significantly reducing link utilization, because the root
cause of the problem is on the host - where Classic congestion cause of the problem is on the host -- where Classic congestion
controls use large saw-toothing rate variations. The L4S approach controls use large sawtoothing rate variations. The L4S approach
resolves this tension between delay and utilization by enabling resolves this tension between delay and utilization by enabling
hosts to minimize the amplitude of their sawteeth. A single-queue hosts to minimize the amplitude of their sawteeth. A single-queue
Classic AQM is not sufficient to allow hosts to use small sawteeth Classic AQM is not sufficient to allow hosts to use small sawteeth
for two reasons: i) smaller sawteeth would not get lower delay in for two reasons: i) smaller sawteeth would not get lower delay in
an AQM designed for larger amplitude Classic sawteeth, because a an AQM designed for larger amplitude Classic sawteeth, because a
queue can only have one length at a time; and ii) much smaller queue can only have one length at a time and ii) much smaller
sawteeth implies much more frequent sawteeth, so L4S flows would sawteeth implies much more frequent sawteeth, so L4S flows would
drive a Classic AQM into a high level of ECN-marking, which would drive a Classic AQM into a high level of ECN-marking, which would
appear as heavy congestion to Classic flows, which in turn would appear as heavy congestion to Classic flows, which in turn would
greatly reduce their rate as a result (see <xref target="l4sarch_sec _classic-ecn-neck" format="default"/>).</dd> greatly reduce their rate as a result (see <xref target="l4sarch_sec _classic-ecn-neck" format="default"/>).</dd>
<dt>Per-flow queuing or marking:</dt> <dt>Per-flow queuing or marking:</dt>
<dd> <dd>
<t>Similarly, per-flow <t>Similarly, per-flow
approaches such as FQ-CoDel or Approx Fair CoDel <xref target="AFCD" approaches, such as FQ-CoDel or Approx Fair CoDel <xref target="AFCD
format="default"/> are not incompatible with the L4S approach. " format="default"/>, are not incompatible with the L4S approach.
However, per-flow queuing alone is not enough - it only isolates However, per-flow queuing alone is not enough -- it only isolates
the queuing of one flow from others; not from itself. Per-flow the queuing of one flow from others, not from itself. Per-flow
implementations need to have support for scalable congestion implementations need to have support for Scalable congestion
control added, which has already been done for FQ-CoDel in Linux control added, which has already been done for FQ-CoDel in Linux
(see Sec.5.2.7 of <xref target="RFC8290" format="default"/> and <xre (see <xref target="RFC8290" sectionFormat="of" section="5.2.7"/> and
f target="FQ_CoDel_Thresh" format="default"/>). Without this simple modification <xref target="FQ_CoDel_Thresh" format="default"/>). Without this simple modific
, ation,
per-flow AQMs like FQ-CoDel would still not be able to support per-flow AQMs, like FQ-CoDel, would still not be able to support
applications that need both very low delay and high bandwidth, applications that need both very low delay and high bandwidth,
e.g. video-based control of remote procedures, or interactive e.g., video-based control of remote procedures or interactive
cloud-based video (see Note <xref format="counter" target="l4sarch_n cloud-based video (see Note <xref format="counter" target="l4sarch_n
ote_app_shuffle"/> below).</t> ote_app_shuffle"/> below).</t>
<t>Although per-flow techniques are not incompatible <t>Although per-flow techniques are not incompatible
with L4S, it is important to have the DualQ alternative. This is with L4S, it is important to have the DualQ alternative. This is
because handling end-to-end (layer 4) flows in the network (layer because handling end-to-end (layer 4) flows in the network (layer
3 or 2) precludes some important end-to-end functions. For 3 or 2) precludes some important end-to-end functions. For
instance:</t> instance:</t>
<ol spacing="normal" type="a"><li> <ol spacing="normal" type="A"><li>
<t>Per-flow forms of L4S like FQ-CoDel are incompatible with <t>Per-flow forms of L4S, like FQ-CoDel, are incompatible with
full end-to-end encryption of transport layer identifiers for full end-to-end encryption of transport layer identifiers for
privacy and confidentiality (e.g. IPSec or encrypted VPN privacy and confidentiality (e.g., IPsec or encrypted VPN
tunnels, as opposed to DTLS over UDP), because they require tunnels, as opposed to DTLS over UDP), because they require
packet inspection to access the end-to-end transport flow packet inspection to access the end-to-end transport flow
identifiers. </t> identifiers. </t>
<t>In contrast, the DualQ <t>In contrast, the DualQ
form of L4S requires no deeper inspection than the IP layer. form of L4S requires no deeper inspection than the IP layer.
So, as long as operators take the DualQ approach, their users So as long as operators take the DualQ approach, their users
can have both very low queuing delay and full end-to-end can have both very low queuing delay and full end-to-end
encryption <xref target="RFC8404" format="default"/>.</t> encryption <xref target="RFC8404" format="default"/>.</t>
</li> </li>
<li> <li>
<t>With per-flow forms of L4S, the network takes over control <t>With per-flow forms of L4S, the network takes over control of
of the relative rates of each application flow. Some see it as the relative rates of each application flow. Some see it as
an advantage that the network will prevent some flows running an advantage that the network will prevent some flows running
faster than others. Others consider it an inherent part of the faster than others. Others consider it an inherent part of the
Internet's appeal that applications can control their rate Internet's appeal that applications can control their rate
while taking account of the needs of others via congestion while taking account of the needs of others via congestion
signals. They maintain that this has allowed applications with signals.
interesting rate behaviours to evolve, for instance, variable They maintain that this has allowed applications with
bit-rate video that varies around an equal share rather than interesting rate behaviours to evolve, for instance: i) a variab
being forced to remain equal at every instant, or e2e le
scavenger behaviours <xref target="RFC6817" format="default"/> t bit-rate video that varies around an equal share, rather than
hat use being forced to remain equal at every instant or ii) end-to-end
less than an equal share of capacity <xref target="LEDBAT_AQM" f scavenger behaviours <xref target="RFC6817" format="default"/> t
ormat="default"/>.</t> hat use
less than an equal share of capacity <xref target="LEDBAT_AQM" f
ormat="default"/>.</t>
<t>The L4S <t>The L4S
architecture does not require the IETF to commit to one architecture does not require the IETF to commit to one
approach over the other, because it supports both, so that the approach over the other, because it supports both so that the
'market' can decide. Nonetheless, in the spirit of 'Do one 'market' can decide. Nonetheless, in the spirit of 'Do one
thing and do it well' <xref target="McIlroy78" format="default"/ >, the thing and do it well' <xref target="McIlroy78" format="default"/ >, the
DualQ option provides low delay without prejudging the issue DualQ option provides low delay without prejudging the issue
of flow-rate control. Then, flow rate policing can be added of flow-rate control. Then, flow rate policing can be added
separately if desired. This allows application control up to a separately if desired. In contrast to scheduling, a policer woul
point, but the network can still choose to set the point at d allow application control up to a
which it intervenes to prevent one flow completely starving point, but the network would still be able to set the point at
which it intervened to prevent one flow completely starving
another.</t> another.</t>
</li> </li>
<!-- <t>fq prevents any one flow from consuming mor
e than 1/N of
the capacity at any instant, where N is the number of flows.
This is fine if all flows are elastic, but it does not sit
well with a variable bit rate real-time multimedia flow, which
requires wriggle room to sometimes take more and other times
less than a 1/N share.<vspace blankLines="1"/>It might seem
that an fq scheduler offers the benefit that it prevents
individual flows from hogging all the bandwidth. However, L4S
has been deliberately designed so that policing of individual
flows can be added as a policy choice, rather than requiring
one specific policy choice as the mechanism itself. {ToDo:
refer to paper on FQ+LEDBAT rather than explain it here - but
we might end up removing this whole bullet} On the other other
end of the spectrum, fq also prevent a flow from using less
than 1/N (otherwise the flow would equally underutilize the
link when N=1). In a shared queue, all flows get equal
congestion signal feedback, which allows
less-than-best-effort-flows to use a lower rate to probability
ratio than Reno-friendly traffic. With fq, the capacity is
split by N equal parts, and congestion feedback is only valid
for the 1/N capacity partition ocupied by the less-than
best-effort flow as if the flow is alone (N=1) and would then
try to fully utilize the available capacity.{/ToDo} A
scheduler (like fq) has to decide packet-by-packet which flow
to schedule without knowing application intent. Whereas a
separate policing function can be configured less strictly, so
that senders can still control the instantaneous rate of each
flow dependent on the needs of each application (e.g. variable
rate video), giving more wriggle-room before a flow is deemed
non-compliant. Also policing of queuing and of flow-rates can
be applied independently.</t>
</ol> </ol>
<t>Note: </t> <t>Note: </t>
<ol spacing="normal" type="1"><li anchor="l4sarch_note_app_shuffle"> <ol spacing="normal" type="1">
It might seem that <li anchor="l4sarch_note_app_shuffle">It might seem that
self-inflicted queuing delay within a per-flow queue should self-inflicted queuing delay within a per-flow queue should
not be counted, because if the delay wasn't in the network it not be counted, because if the delay wasn't in the network, it
would just shift to the sender. However, modern adaptive would just shift to the sender. However, modern adaptive
applications, e.g. HTTP/2 <xref target="RFC9113" format="default "/> applications, e.g., HTTP/2 <xref target="RFC9113" format="defaul t"/>
or some interactive media applications (see <xref target="l4sarc h_apps" format="default"/>), can keep low latency objects at the or some interactive media applications (see <xref target="l4sarc h_apps" format="default"/>), can keep low latency objects at the
front of their local send queue by shuffling priorities of front of their local send queue by shuffling priorities of
other objects dependent on the progress of other transfers other objects dependent on the progress of other transfers
(for example see <xref target="lowat" format="default"/>). They cannot shuffle (for example, see <xref target="lowat" format="default"/>). They cannot shuffle
objects once they have released them into the network.</li> objects once they have released them into the network.</li>
</ol> </ol>
</dd> </dd>
<dt>Alternative Back-off ECN (ABE):</dt> <dt>Alternative Back-off ECN (ABE):</dt>
<dd>Here again, L4S is <dd>Here again, L4S is
not an alternative to ABE but a complement that introduces much not an alternative to ABE but a complement that introduces much
lower queuing delay. ABE <xref target="RFC8511" format="default"/> a lters the lower queuing delay. ABE <xref target="RFC8511" format="default"/> a lters the
host behaviour in response to ECN marking to utilize a link better host behaviour in response to ECN marking to utilize a link better
and give ECN flows faster throughput. It uses ECT(0) and assumes and give ECN flows faster throughput. It uses ECT(0) and assumes
the network still treats ECN and drop the same. Therefore, ABE the network still treats ECN and drop the same. Therefore, ABE
exploits any lower queuing delay that AQMs can provide. But, as exploits any lower queuing delay that AQMs can provide. But, as
explained above, AQMs still cannot reduce queuing delay too far explained above, AQMs still cannot reduce queuing delay too much
without losing link utilization (to allow for other, non-ABE, without losing link utilization (to allow for other, non-ABE,
flows).</dd> flows).</dd>
<dt>BBR:</dt> <dt>BBR:</dt>
<dd> <dd>
<t>Bottleneck Bandwidth and Round-trip propagation <t>Bottleneck Bandwidth and Round-trip propagation
time (BBR <xref target="I-D.cardwell-iccrg-bbr-congestion-control" f ormat="default"/>) controls time (BBR) <xref target="I-D.cardwell-iccrg-bbr-congestion-control" format="default"/> controls
queuing delay end-to-end without needing any special logic in the queuing delay end-to-end without needing any special logic in the
network, such as an AQM. So it works pretty-much on any path. BBR network, such as an AQM. So it works pretty much on any path. BBR
keeps queuing delay reasonably low, but perhaps not quite as low keeps queuing delay reasonably low, but perhaps not quite as low
as with state-of-the-art AQMs such as PIE or FQ-CoDel, and as with state-of-the-art AQMs, such as PIE or FQ-CoDel, and
certainly nowhere near as low as with L4S. Queuing delay is also certainly nowhere near as low as with L4S. Queuing delay is also
not consistently low, due to BBR's regular bandwidth probing not consistently low, due to BBR's regular bandwidth probing
spikes and its aggressive flow start-up phase.</t> spikes and its aggressive flow start-up phase.</t>
<t>L4S complements BBR. Indeed, BBRv2 can use L4S ECN <t>L4S complements BBR. Indeed, BBRv2 can use L4S ECN
where available and a scalable L4S congestion control behaviour in where available and a Scalable L4S congestion control behaviour in
response to any ECN signalling from the path <xref target="BBRv2" fo response to any ECN signalling from the path <xref target="BBRv2" fo
rmat="default"/>. The L4S ECN signal complements the delay based rmat="default"/>. The L4S ECN signal complements the delay-based
congestion control aspects of BBR with an explicit indication that congestion control aspects of BBR with an explicit indication that
hosts can use, both to converge on a fair rate and to keep below a hosts can use, both to converge on a fair rate and to keep below a
shallow queue target set by the network. Without L4S ECN, both shallow queue target set by the network. Without L4S ECN, both
these aspects need to be assumed or estimated.</t> these aspects need to be assumed or estimated.</t>
</dd> </dd>
</dl> </dl>
</section> </section>
</section> </section>
<section anchor="l4sarch_applicability" numbered="true" toc="default"> <section anchor="l4sarch_applicability" numbered="true" toc="default">
<name>Applicability</name> <name>Applicability</name>
<section anchor="l4sarch_apps" numbered="true" toc="default"> <section anchor="l4sarch_apps" numbered="true" toc="default">
<name>Applications</name> <name>Applications</name>
<t>A transport layer that solves the current latency issues will <t>A transport layer that solves the current latency issues will
provide new service, product and application opportunities.</t> provide new service, product, and application opportunities.</t>
<t>With the L4S approach, the following existing applications also <t>With the L4S approach, the following existing applications also
experience significantly better quality of experience under load: experience significantly better quality of experience under load:
</t> </t>
<ul spacing="normal"> <ul spacing="normal">
<li>Gaming, including cloud based gaming;</li> <li>gaming, including cloud-based gaming;</li>
<li>VoIP;</li> <li>VoIP;</li>
<li>Video conferencing;</li> <li>video conferencing;</li>
<li>Web browsing;</li> <li>web browsing;</li>
<li>(Adaptive) video streaming;</li> <li>(adaptive) video streaming; and</li>
<li>Instant messaging.</li> <li>instant messaging.</li>
</ul> </ul>
<t>The significantly lower queuing latency also enables some <t>The significantly lower queuing latency also enables some
interactive application functions to be offloaded to the cloud that interactive application functions to be offloaded to the cloud that
would hardly even be usable today: </t> would hardly even be usable today, including:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>Cloud based interactive video;</li> <li>cloud-based interactive video and</li>
<li>Cloud based virtual and augmented reality.</li> <li>cloud-based virtual and augmented reality.</li>
</ul> </ul>
<t>The above two applications have been successfully demonstrated with <t>The above two applications have been successfully demonstrated with
L4S, both running together over a 40 Mb/s broadband access link L4S, both running together over a 40 Mb/s broadband access link
loaded up with the numerous other latency sensitive applications in loaded up with the numerous other latency-sensitive applications in
the previous list as well as numerous downloads - all sharing the same the previous list, as well as numerous downloads, with all sharing the s
bottleneck queue simultaneously <xref target="L4Sdemo16" format="default ame
"/>. For bottleneck queue simultaneously <xref target="L4Sdemo16" format="default
"/> <xref target="L4Sdemo16-Video" format="default"/>. For
the former, a panoramic video of a football stadium could be swiped the former, a panoramic video of a football stadium could be swiped
and pinched so that, on the fly, a proxy in the cloud could generate a and pinched so that, on the fly, a proxy in the cloud could generate a
sub-window of the match video under the finger-gesture control of each sub-window of the match video under the finger-gesture control of each
user. For the latter, a virtual reality headset displayed a viewport user. For the latter, a virtual reality headset displayed a viewport
taken from a 360-degree camera in a racing car. The user's head taken from a 360-degree camera in a racing car. The user's head
movements controlled the viewport extracted by a cloud-based proxy. In movements controlled the viewport extracted by a cloud-based proxy. In
both cases, with 7 ms end-to-end base delay, the additional both cases, with a 7 ms end-to-end base delay, the additional
queuing delay of roughly 1 ms was so low that it seemed the video queuing delay of roughly 1 ms was so low that it seemed the video
was generated locally. </t> was generated locally. </t>
<t>Using a swiping finger gesture or head movement to pan a video are <t>Using a swiping finger gesture or head movement to pan a video are
extremely latency-demanding actions -- far more demanding than extremely latency-demanding actions -- far more demanding than
VoIP. Because human vision can detect extremely low delays of the VoIP -- because human vision can detect extremely low delays of the
order of single milliseconds when delay is translated into a visual order of single milliseconds when delay is translated into a visual
lag between a video and a reference point (the finger or the lag between a video and a reference point (the finger or the
orientation of the head sensed by the balance system in the inner ear orientation of the head sensed by the balance system in the inner ear,
-- the vestibular system). With an alternative AQM, the video i.e., the vestibular system). With an alternative AQM, the video
noticeably lagged behind the finger gestures and head movements.</t> noticeably lagged behind the finger gestures and head movements.</t>
<t>Without the low queuing delay of L4S, cloud-based applications like <t>Without the low queuing delay of L4S, cloud-based applications like
these would not be credible without significantly more access these would not be credible without significantly more access-network ba
bandwidth (to deliver all possible video that might be viewed) and ndwidth
(to deliver all possible areas of the video that might be viewed) and
more local processing, which would increase the weight and power more local processing, which would increase the weight and power
consumption of head-mounted displays. When all interactive processing consumption of head-mounted displays. When all interactive processing
can be done in the cloud, only the data to be rendered for the end can be done in the cloud, only the data to be rendered for the end
user needs to be sent.</t> user needs to be sent.</t>
<t>Other low latency high bandwidth applications such as:</t> <t>Other low latency high bandwidth applications, such as:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>Interactive remote presence;</li> <li>interactive remote presence and</li>
<li>Video-assisted remote control of machinery or industrial <li>video-assisted remote control of machinery or industrial
processes.</li> processes</li>
</ul> </ul>
<t>are not credible at all without very low queuing delay. No <t>are not credible at all without very low queuing delay. No
amount of extra access bandwidth or local processing can make up for amount of extra access bandwidth or local processing can make up for
lost time.</t> lost time.</t>
</section> </section>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>Use Cases</name> <name>Use Cases</name>
<t>The following use-cases for L4S are being considered by various <t>The following use cases for L4S are being considered by various
interested parties:</t> interested parties:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>Where the bottleneck is one of various types of access network: <li>where the bottleneck is one of various types of access network,
e.g. DSL, Passive Optical Networks (PON), DOCSIS cable, e.g., DSL, Passive Optical Networks (PONs), DOCSIS cable,
mobile, satellite (see <xref target="l4sarch_link-specifics" format= mobile, satellite; or where it's a Wi-Fi link (see <xref target="l4s
"default"/> for arch_link-specifics" format="default"/> for
some technology-specific details)</li> some technology-specific details)</li>
<li> <li>
<t>Private networks of heterogeneous data centres, where there is <t>private networks of heterogeneous data centres, where there is
no single administrator that can arrange for all the simultaneous no single administrator that can arrange for all the simultaneous
changes to senders, receivers and network needed to deploy changes to senders, receivers, and networks needed to deploy
DCTCP:</t> DCTCP:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>a set of private data centres interconnected over a wide <li>a set of private data centres interconnected over a wide
area with separate administrations, but within the same area with separate administrations but within the same
company</li> company</li>
<li>a set of data centres operated by separate companies <li>a set of data centres operated by separate companies
interconnected by a community of interest network interconnected by a community of interest network
(e.g. for the finance sector)</li> (e.g., for the finance sector)</li>
<li>multi-tenant (cloud) data centres where tenants choose <li>multi-tenant (cloud) data centres where tenants choose
their operating system stack (Infrastructure as a Service - their operating system stack (Infrastructure as a Service
IaaS)</li> (IaaS))</li>
</ul> </ul>
</li> </li>
<li> <li>
<t>Different types of transport (or application) congestion <t>different types of transport (or application) congestion
control:</t> control:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>elastic (TCP/SCTP);</li> <li>elastic (TCP/SCTP);</li>
<li>real-time (RTP, RMCAT);</li> <li>real-time (RTP, RMCAT); and</li>
<li>query (DNS/LDAP).</li> <li>query-response (DNS/LDAP).</li>
</ul> </ul>
</li> </li>
<li> <li>
<t>Where low delay quality of service is required, but without <t>where low delay QoS is required but without
inspecting or intervening above the IP layer <xref target="RFC8404" inspecting or intervening above the IP layer <xref target="RFC8404"
format="default"/>:</t> format="default"/>:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>mobile and other networks have tended to inspect higher <li>Mobile and other networks have tended to inspect higher
layers in order to guess application QoS requirements. layers in order to guess application QoS requirements.
However, with growing demand for support of privacy and However, with growing demand for support of privacy and
encryption, L4S offers an alternative. There is no need to encryption, L4S offers an alternative. There is no need to
select which traffic to favour for queuing, when L4S can give select which traffic to favour for queuing when L4S can give
favourable queuing to all traffic.</li> favourable queuing to all traffic.</li>
</ul> </ul>
</li> </li>
<li>If queuing delay is minimized, applications with a fixed delay <li>If queuing delay is minimized, applications with a fixed delay
budget can communicate over longer distances, or via a longer budget can communicate over longer distances or via more circuitous
chain of service functions <xref target="RFC7665" format="default"/> paths, e.g., longer
or onion chains of service functions <xref target="RFC7665" format="default"/
> or of onion
routers.</li> routers.</li>
<li>If delay jitter is minimized, it is possible to reduce the <li>If delay jitter is minimized, it is possible to reduce the
dejitter buffers on the receive end of video streaming, which dejitter buffers on the receiving end of video streaming, which
should improve the interactive experience</li> should improve the interactive experience.</li>
</ul> </ul>
</section> </section>
<section anchor="l4sarch_link-specifics" numbered="true" toc="default"> <section anchor="l4sarch_link-specifics" numbered="true" toc="default">
<name>Applicability with Specific Link Technologies</name> <name>Applicability with Specific Link Technologies</name>
<t>Certain link technologies aggregate data from multiple packets into <t>Certain link technologies aggregate data from multiple packets into
bursts, and buffer incoming packets while building each burst. Wi-Fi, bursts and buffer incoming packets while building each burst. Wi-Fi,
PON and cable all involve such packet aggregation, whereas fixed PON, and cable all involve such packet aggregation, whereas fixed
Ethernet and DSL do not. No sender, whether L4S or not, can do Ethernet and DSL do not. No sender, whether L4S or not, can do
anything to reduce the buffering needed for packet aggregation. So an anything to reduce the buffering needed for packet aggregation. So an
AQM should not count this buffering as part of the queue that it AQM should not count this buffering as part of the queue that it
controls, given no amount of congestion signals will reduce it.</t> controls, given no amount of congestion signals will reduce it.</t>
<t>Certain link technologies also add buffering for other reasons, <t>Certain link technologies also add buffering for other reasons,
specifically:</t> specifically:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>Radio links (cellular, Wi-Fi, satellite) that are distant from <li>Radio links (cellular, Wi-Fi, or satellite) that are distant from
the source are particularly challenging. The radio link capacity the source are particularly challenging. The radio link capacity
can vary rapidly by orders of magnitude, so it is considered can vary rapidly by orders of magnitude, so it is considered
desirable to hold a standing queue that can utilize sudden desirable to hold a standing queue that can utilize sudden
increases of capacity;</li> increases of capacity.</li>
<li>Cellular networks are further complicated by a perceived need <li>Cellular networks are further complicated by a perceived need
to buffer in order to make hand-overs imperceptible;</li> to buffer in order to make hand-overs imperceptible.</li>
</ul> </ul>
<t>L4S cannot remove the need for all these different forms of <t>L4S cannot remove the need for all these different forms of
buffering. However, by removing 'the longest pole in the tent' buffering. However, by removing 'the longest pole in the tent'
(buffering for the large sawteeth of Classic congestion controls), L4S (buffering for the large sawteeth of Classic congestion controls), L4S
exposes all these 'shorter poles' to greater scrutiny.</t> exposes all these 'shorter poles' to greater scrutiny.</t>
<t>Until now, the buffering needed for these additional reasons tended <t>Until now, the buffering needed for these additional reasons tended
to be over-specified - with the excuse that none were 'the longest to be over-specified -- with the excuse that none were 'the longest
pole in the tent'. But having removed the 'longest pole', it becomes pole in the tent'. But having removed the 'longest pole', it becomes
worthwhile to minimize them, for instance reducing packet aggregation worthwhile to minimize them, for instance, reducing packet aggregation
burst sizes and MAC scheduling intervals.</t> burst sizes and MAC scheduling intervals.</t>
<t>Also certain link types, particularly radio-based links, are far <t>Also, certain link types, particularly radio-based links, are far
more prone to transmission losses. <xref target="l4sarch_sec_non-l4s-nec k" format="default"/> explains how an L4S response to more prone to transmission losses. <xref target="l4sarch_sec_non-l4s-nec k" format="default"/> explains how an L4S response to
loss has to be as drastic as a Classic response. Nonetheless, research loss has to be as drastic as a Classic response. Nonetheless, research
referred to in the same section has demonstrated potential for referred to in the same section has demonstrated potential for
considerably more effective loss repair at the link layer, due to the considerably more effective loss repair at the link layer, due to the
relaxed ordering constraints of L4S packets.</t> relaxed ordering constraints of L4S packets.</t>
</section> </section>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>Deployment Considerations</name> <name>Deployment Considerations</name>
<t>L4S AQMs, whether DualQ <xref target="I-D.ietf-tsvwg-aqm-dualq-couple <t>L4S AQMs, whether DualQ <xref target="RFC9332" format="default"/> or
d" format="default"/> or FQ, e.g. <xref target="RFC8290" format="default"/> are, FQ <xref target="RFC8290" format="default"/>, are in themselves an incremental d
in themselves, an incremental deployment eployment
mechanism for L4S - so that L4S traffic can coexist with existing mechanism for L4S -- so that L4S traffic can coexist with existing
Classic (Reno-friendly) traffic. <xref target="l4sarch_deploy_top" forma t="default"/> Classic (Reno-friendly) traffic. <xref target="l4sarch_deploy_top" forma t="default"/>
explains why only deploying an L4S AQM in one node at each end of the explains why only deploying an L4S AQM in one node at each end of the
access link will realize nearly all the benefit of L4S.</t> access link will realize nearly all the benefit of L4S.</t>
<t>L4S involves both end systems and the network, so <xref target="l4s_a <t>L4S involves both the network and end systems, so <xref target="l4s_a
rch_deploy_seq" format="default"/> suggests some typical sequences to rch_deploy_seq" format="default"/> suggests some typical sequences to
deploy each part, and why there will be an immediate and significant deploy each part and why there will be an immediate and significant
benefit after deploying just one part.</t> benefit after deploying just one part.</t>
<t><xref target="l4sarch_sec_non-l4s-neck" format="default"/> and <xref target="l4sarch_sec_classic-ecn-neck" format="default"/> describe the converse <t>Sections <xref target="l4sarch_sec_non-l4s-neck" format="counter"/> a nd <xref target="l4sarch_sec_classic-ecn-neck" format="counter"/> describe the c onverse
incremental deployment case where there is no L4S AQM at the network incremental deployment case where there is no L4S AQM at the network
bottleneck, so any L4S flow traversing this bottleneck has to take bottleneck, so any L4S flow traversing this bottleneck has to take
care in case it is competing with Classic traffic.</t> care in case it is competing with Classic traffic.</t>
<section anchor="l4sarch_deploy_top" numbered="true" toc="default"> <section anchor="l4sarch_deploy_top" numbered="true" toc="default">
<name>Deployment Topology</name> <name>Deployment Topology</name>
<t>L4S AQMs will not have to be deployed throughout the Internet <t>L4S AQMs will not have to be deployed throughout the Internet
before L4S can benefit anyone. Operators of public Internet access before L4S can benefit anyone. Operators of public Internet access
networks typically design their networks so that the bottleneck will networks typically design their networks so that the bottleneck will
nearly always occur at one known (logical) link. This confines the nearly always occur at one known (logical) link. This confines the
cost of queue management technology to one place.</t> cost of queue management technology to one place.</t>
<t>The case of mesh networks is different and will be discussed <t>The case of mesh networks is different and will be discussed
later in this section. But the known bottleneck case is generally later in this section.
However, the known-bottleneck case is generally
true for Internet access to all sorts of different 'sites', where true for Internet access to all sorts of different 'sites', where
the word 'site' includes home networks, small- to medium-sized the word 'site' includes home networks, small- to medium-sized
campus or enterprise networks and even cellular devices (<xref target= campus or enterprise networks and even cellular devices (<xref
"l4sarch_fig_access_topology" format="default"/>). Also, this known-bottleneck target="l4sarch_fig_access_topology" format="default"/>).
case tends to be applicable whatever the access link technology; Also, this known-bottleneck
whether xDSL, cable, PON, cellular, line of sight wireless or case tends to be applicable whatever the access link technology,
whether xDSL, cable, PON, cellular, line of sight wireless, or
satellite.</t> satellite.</t>
<t>Therefore, the full benefit of the L4S service should be <t>Therefore, the full benefit of the L4S service should be
available in the downstream direction when an L4S AQM is deployed at available in the downstream direction when an L4S AQM is deployed at
the ingress to this bottleneck link. And similarly, the full the ingress to this bottleneck link. And similarly, the full
upstream service will be available once an L4S AQM is deployed at upstream service will typically be available once an L4S AQM is deploy
the ingress into the upstream link. (Of course, multi-homed sites ed at
the ingress into the upstream link. (Of course, multihomed sites
would only see the full benefit once all their access links were would only see the full benefit once all their access links were
covered.)</t> covered.)</t>
<figure anchor="l4sarch_fig_access_topology"> <figure anchor="l4sarch_fig_access_topology">
<name>Likely location of DualQ (DQ) Deployments in common access top <name>Likely Location of DualQ (DQ) Deployments in Common Access Top
ologies</name> ologies</name>
<artwork name="" type="" align="left" alt=""><![CDATA[ <artwork name="" type="" align="left" alt=""><![CDATA[
______ ______
( ) ( )
__ __ ( ) __ __ ( )
|DQ\________/DQ|( enterprise ) |DQ\________/DQ|( enterprise )
___ |__/ \__| ( /campus ) ___ |__/ \__| ( /campus )
( ) (______) ( ) (______)
( ) ___||_ ( ) ___||_
+----+ ( ) __ __ / \ +----+ ( ) __ __ / \
| DC |-----( Core )|DQ\_______________/DQ|| home | | DC |-----( Core )|DQ\_______________/DQ|| home |
+----+ ( ) |__/ \__||______| +----+ ( ) |__/ \__||______|
(_____) __ (_____) __
skipping to change at line 1236 skipping to change at line 1213
|__/ \ ____/DQ||| ||mobile |__/ \ ____/DQ||| ||mobile
\/ \__|||_||device \/ \__|||_||device
| o | | o |
`---' `---'
]]></artwork> ]]></artwork>
</figure> </figure>
<t>Deployment in mesh topologies depends on how overbooked the core <t>Deployment in mesh topologies depends on how overbooked the core
is. If the core is non-blocking, or at least generously provisioned is. If the core is non-blocking, or at least generously provisioned
so that the edges are nearly always the bottlenecks, it would only so that the edges are nearly always the bottlenecks, it would only
be necessary to deploy an L4S AQM at the edge bottlenecks. For be necessary to deploy an L4S AQM at the edge bottlenecks.
For
example, some data-centre networks are designed with the bottleneck example, some data-centre networks are designed with the bottleneck
in the hypervisor or host NICs, while others bottleneck at the in the hypervisor or host Network Interface Controllers (NICs), while
others
bottleneck at the
top-of-rack switch (both the output ports facing hosts and those top-of-rack switch (both the output ports facing hosts and those
facing the core).</t> facing the core).</t>
<t>An L4S AQM would often next be needed where the Wi-Fi links in a <t>An L4S AQM would often next be needed where the Wi-Fi links in a
home sometimes become the bottleneck. And an L4S AQM would home sometimes become the bottleneck. Also an L4S AQM would
eventually also need to be deployed at any other persistent eventually need to be deployed at any other persistent
bottlenecks such as network interconnections, e.g. some public bottlenecks, such as network interconnections, e.g., some public
Internet exchange points and the ingress and egress to WAN links Internet exchange points and the ingress and egress to WAN links
interconnecting data-centres.</t> interconnecting data centres.</t>
</section> </section>
<section anchor="l4s_arch_deploy_seq" numbered="true" toc="default"> <section anchor="l4s_arch_deploy_seq" numbered="true" toc="default">
<name>Deployment Sequences</name> <name>Deployment Sequences</name>
<t>For any one L4S flow to provide benefit, it requires three (or <t>For any one L4S flow to provide benefit, it requires three (or
sometimes two) parts to have been deployed: i) the congestion sometimes two) parts to have been deployed: i) the congestion
control at the sender; ii) the AQM at the bottleneck; and iii) older control at the sender; ii) the AQM at the bottleneck; and iii) older
transports (namely TCP) need upgraded receiver feedback too. This transports (namely TCP) need upgraded receiver feedback too. This
was the same deployment problem that ECN faced <xref target="RFC8170" format="default"/> so we have learned from that experience.</t> was the same deployment problem that ECN faced <xref target="RFC8170" format="default"/>, so we have learned from that experience.</t>
<t>Firstly, L4S deployment exploits the fact that DCTCP already <t>Firstly, L4S deployment exploits the fact that DCTCP already
exists on many Internet hosts (Windows, FreeBSD and Linux); both exists on many Internet hosts (e.g., Windows, FreeBSD, and Linux), bot h
servers and clients. Therefore, an L4S AQM can be deployed at a servers and clients. Therefore, an L4S AQM can be deployed at a
network bottleneck to immediately give a working deployment of all network bottleneck to immediately give a working deployment of all
the L4S parts for testing, as long as the ECT(0) codepoint is the L4S parts for testing, as long as the ECT(0) codepoint is
switched to ECT(1). DCTCP needs some safety concerns to be fixed for switched to ECT(1). DCTCP needs some safety concerns to be fixed for
general use over the public Internet (see Section 4.3 of the L4S ECN general use over the public Internet (see <xref target="RFC9331" forma
spec <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/>), but t="default" sectionFormat="of" section="4.3">the L4S ECN spec</xref>), but DCTCP
DCTCP is is
not on by default, so these issues can be managed within controlled not on by default, so these issues can be managed within controlled
deployments or controlled trials.</t> deployments or controlled trials.</t>
<t>Secondly, the performance improvement with L4S is so significant <t>Secondly, the performance improvement with L4S is so significant
that it enables new interactive services and products that were not that it enables new interactive services and products that were not
previously possible. It is much easier for companies to initiate new previously possible. It is much easier for companies to initiate new
work on deployment if there is budget for a new product trial. If, work on deployment if there is budget for a new product trial.
in contrast, there were only an incremental performance improvement In contrast, if there were only an incremental performance improvement
(as with Classic ECN), spending on deployment tends to be much (as with Classic ECN), spending on deployment tends to be much
harder to justify.</t> harder to justify.</t>
<t>Thirdly, the L4S identifier is defined so that initially network <t>Thirdly, the L4S identifier is defined so that network
operators can enable L4S exclusively for certain customers or operators can initially enable L4S exclusively for certain customers o
certain applications. But this is carefully defined so that it does r
certain applications. However, this is carefully defined so that it do
es
not compromise future evolution towards L4S as an Internet-wide not compromise future evolution towards L4S as an Internet-wide
service. This is because the L4S identifier is defined not only as service. This is because the L4S identifier is defined not only as
the end-to-end ECN field, but it can also optionally be combined the end-to-end ECN field, but it can also optionally be combined
with any other packet header or some status of a customer or their with any other packet header or some status of a customer or their
access link (see section 5.4 of <xref target="I-D.ietf-tsvwg-ecn-l4s-i d" format="default"/>). Operators could do this access link (see <xref target="RFC9331" format="default" sectionFormat ="of" section="5.4"/>). Operators could do this
anyway, even if it were not blessed by the IETF. However, it is best anyway, even if it were not blessed by the IETF. However, it is best
for the IETF to specify that, if they use their own local for the IETF to specify that, if they use their own local
identifier, it must be in combination with the IETF's identifier. identifier, it must be in combination with the IETF's identifier, ECT( 1).
Then, if an operator has opted for an exclusive local-use approach, Then, if an operator has opted for an exclusive local-use approach,
later they only have to remove this extra rule to make the service they only have to remove this extra rule later to make the service
work Internet-wide - it will already traverse middleboxes, peerings, work across the Internet -- it will already traverse middleboxes, peer
etc. <!--{K: Review up to here}--> ings,
etc.
</t> </t>
<figure anchor="l4s_arch_fig_deploy_seq"> <figure anchor="l4s_arch_fig_deploy_seq">
<name>Example L4S Deployment Sequence</name> <name>Example L4S Deployment Sequence</name>
<artwork name="" type="" align="left" alt=""><![CDATA[+-+----------- <artwork name="" type="" align="left" alt=""><![CDATA[
---------+----------------------+---------------------+ +-+--------------------+----------------------+---------------------+
| | Servers or proxies | Access link | Clients | | | Servers or proxies | Access link | Clients |
+-+--------------------+----------------------+---------------------+ +-+--------------------+----------------------+---------------------+
|0| DCTCP (existing) | | DCTCP (existing) | |0| DCTCP (existing) | | DCTCP (existing) |
+-+--------------------+----------------------+---------------------+ +-+--------------------+----------------------+---------------------+
|1| |Add L4S AQM downstream| | |1| |Add L4S AQM downstream| |
| | WORKS DOWNSTREAM FOR CONTROLLED DEPLOYMENTS/TRIALS | | | WORKS DOWNSTREAM FOR CONTROLLED DEPLOYMENTS/TRIALS |
+-+--------------------+----------------------+---------------------+ +-+--------------------+----------------------+---------------------+
|2| Upgrade DCTCP to | |Replace DCTCP feedb'k| |2| Upgrade DCTCP to | |Replace DCTCP feedb'k|
| | TCP Prague | | with AccECN | | | TCP Prague | | with AccECN |
| | FULLY WORKS DOWNSTREAM | | | FULLY WORKS DOWNSTREAM |
skipping to change at line 1321 skipping to change at line 1300
sequences in which the parts of L4S might be deployed. It consists sequences in which the parts of L4S might be deployed. It consists
of the following stages, preceded by a presumption that DCTCP is of the following stages, preceded by a presumption that DCTCP is
already installed at both ends:</t> already installed at both ends:</t>
<ol spacing="normal" type="1"><li> <ol spacing="normal" type="1"><li>
<t>DCTCP is not applicable for use over the public Internet, so <t>DCTCP is not applicable for use over the public Internet, so
it is emphasized here that any DCTCP flow has to be completely it is emphasized here that any DCTCP flow has to be completely
contained within a controlled trial environment. </t> contained within a controlled trial environment. </t>
<t>Within this trial environment, once an L4S AQM <t>Within this trial environment, once an L4S AQM
has been deployed, the trial DCTCP flow will experience has been deployed, the trial DCTCP flow will experience
immediate benefit, without any other deployment being needed. In immediate benefit, without any other deployment being needed. In
this example downstream deployment is first, but in other this example, downstream deployment is first, but in other
scenarios the upstream might be deployed first. If no AQM at all scenarios, the upstream might be deployed first. If no AQM at all
was previously deployed for the downstream access, an L4S AQM was previously deployed for the downstream access, an L4S AQM
greatly improves the Classic service (as well as adding the L4S greatly improves the Classic service (as well as adding the L4S
service). If an AQM was already deployed, the Classic service service). If an AQM was already deployed, the Classic service
will be unchanged (and L4S will add an improvement on top).</t> will be unchanged (and L4S will add an improvement on top).</t>
</li> </li>
<li> <li>
<t>In this stage, the name 'TCP Prague' <xref target="I-D.briscoe- iccrg-prague-congestion-control" format="default"/> is used <t>In this stage, the name 'TCP Prague' <xref target="I-D.briscoe- iccrg-prague-congestion-control" format="default"/> is used
to represent a variant of DCTCP that is designed to be used in a to represent a variant of DCTCP that is designed to be used in a
production Internet environment (that is, it has to comply with production Internet environment (that is, it has to comply with
all the requirements in Section 4 of the L4S ECN spec <xref target ="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/>, which then means it can be all the requirements in <xref target="RFC9331" format="default" se ction="4" sectionFormat="of">the L4S ECN spec</xref>, which then means it can be
used over the public Internet). If the application is primarily used over the public Internet). If the application is primarily
unidirectional, 'TCP Prague' at one end will provide all the unidirectional, 'TCP Prague' at the sending end will provide all
benefit needed.</t> the benefit needed, as long as the receiving end supports Accurate
ECN (AccECN)
feedback <xref target="I-D.ietf-tcpm-accurate-ecn" format="default
"/>.</t>
<t>For TCP transports, <t>For TCP transports,
Accurate ECN feedback (AccECN) <xref target="I-D.ietf-tcpm-accurat e-ecn" format="default"/> is needed at the other AccECN feedback is needed at the other
end, but it is a generic ECN feedback facility that is already end, but it is a generic ECN feedback facility that is already
planned to be deployed for other purposes, e.g. DCTCP, BBR. planned to be deployed for other purposes, e.g., DCTCP and BBR.
The two ends can be deployed in either order, because, in TCP, The two ends can be deployed in either order because, in TCP,
an L4S congestion control only enables itself if it has an L4S congestion control only enables itself if it has
negotiated the use of AccECN feedback with the other end during negotiated the use of AccECN feedback with the other end during
the connection handshake. Thus, deployment of TCP Prague on a the connection handshake. Thus, deployment of TCP Prague on a
server enables L4S trials to move to a production service in one server enables L4S trials to move to a production service in one
direction, wherever AccECN is deployed at the other end. This direction, wherever AccECN is deployed at the other end. This
stage might be further motivated by the performance improvements stage might be further motivated by the performance improvements
of TCP Prague relative to DCTCP (see Appendix A.2 of the L4S ECN of TCP Prague relative to DCTCP (see <xref target="RFC9331" forma
spec <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/>). t="default" sectionFormat="of" section="A.2">the L4S ECN spec</xref>).</t>
</t>
<t>Unlike TCP, from the outset, QUIC ECN <t>Unlike TCP, from the outset, QUIC ECN
feedback <xref target="RFC9000" format="default"/> has supported L 4S. feedback <xref target="RFC9000" format="default"/> has supported L 4S.
Therefore, if the transport is QUIC, one-ended deployment of a Therefore, if the transport is QUIC, one-ended deployment of a
Prague congestion control at this stage is simple and Prague congestion control at this stage is simple and
sufficient.</t> sufficient.</t>
<t>For QUIC, if a proxy sits in <t>For QUIC, if a proxy sits in
the path between multiple origin servers and the access the path between multiple origin servers and the access
bottlenecks to multiple clients, then upgrading the proxy with a bottlenecks to multiple clients, then upgrading the proxy with a
Scalable congestion control would provide the benefits of L4S Scalable congestion control would provide the benefits of L4S
over all the clients' downstream bottlenecks in one go --- over all the clients' downstream bottlenecks in one go --
whether or not all the origin servers were upgraded. Conversely, whether or not all the origin servers were upgraded. Conversely,
where a proxy has not been upgraded, the clients served by it where a proxy has not been upgraded, the clients served by it
will not benefit from L4S at all in the downstream, even when will not benefit from L4S at all in the downstream, even when
any origin server behind the proxy has been upgraded to support any origin server behind the proxy has been upgraded to support
L4S.</t> L4S.</t>
<t>For TCP, a proxy upgraded to support <t>For TCP, a proxy upgraded to support
'TCP Prague' would provide the benefits of L4S downstream to all 'TCP Prague' would provide the benefits of L4S downstream to all
clients that support AccECN (whether or not they support L4S as clients that support AccECN (whether or not they support L4S as
well). And in the upstream, the proxy would also support AccECN well). And in the upstream, the proxy would also support AccECN
as a receiver, so that any client deploying its own L4S support as a receiver, so that any client deploying its own L4S support
would benefit in the upstream direction, irrespective of whether would benefit in the upstream direction, irrespective of whether
any origin server beyond the proxy supported AccECN.</t> any origin server beyond the proxy supported AccECN.</t>
</li> </li>
<li>This is a two-move stage to enable L4S upstream. An L4S AQM <li>This is a two-move stage to enable L4S upstream. An L4S AQM
or TCP Prague can be deployed in either order as already or TCP Prague can be deployed in either order as already
explained. To motivate the first of two independent moves, the explained. To motivate the first of two independent moves, the
deferred benefit of enabling new services after the second move deferred benefit of enabling new services after the second move
has to be worth it to cover the first mover's investment risk. has to be worth it to cover the first mover's investment risk.
As explained already, the potential for new interactive services As explained already, the potential for new interactive services
provides this motivation. An L4S AQM also improves the upstream provides this motivation. An L4S AQM also improves the upstream
Classic service - significantly if no other AQM has already been Classic service significantly if no other AQM has already been
deployed.</li> deployed.</li>
</ol> </ol>
<t>Note that other deployment sequences might occur. For <t>Note that other deployment sequences might occur. For
instance: the upstream might be deployed first; a non-TCP protocol instance, the upstream might be deployed first; a non-TCP protocol
might be used end-to-end, e.g. QUIC, RTP; a body such as the might be used end to end, e.g., QUIC and RTP; a body, such as the
3GPP might require L4S to be implemented in 5G user equipment, or 3GPP, might require L4S to be implemented in 5G user equipment; or
other random acts of kindness.</t> other random acts of kindness might arise.</t>
</section> </section>
<section anchor="l4sarch_sec_non-l4s-neck" numbered="true" toc="default" > <section anchor="l4sarch_sec_non-l4s-neck" numbered="true" toc="default" >
<name>L4S Flow but Non-ECN Bottleneck</name> <name>L4S Flow but Non-ECN Bottleneck</name>
<t>If L4S is enabled between two hosts, the L4S sender is required <t>If L4S is enabled between two hosts, the L4S sender is required
to coexist safely with Reno in response to any drop (see Section 4.3 to coexist safely with Reno in response to any drop (see <xref target=
of the L4S ECN spec <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="d "RFC9331" format="default" sectionFormat="of" section="4.3">the L4S ECN spec</xr
efault"/>).</t> ef>).</t>
<t>Unfortunately, as well as protecting Classic traffic, this rule <t>Unfortunately, as well as protecting Classic traffic, this rule
degrades the L4S service whenever there is any loss, even if the degrades the L4S service whenever there is any loss, even if the
cause is not persistent congestion at a bottleneck, e.g.:</t> cause is not persistent congestion at a bottleneck, for example:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>congestion loss at other transient bottlenecks, e.g. due <li>congestion loss at other transient bottlenecks, e.g., due
to bursts in shallower queues;</li> to bursts in shallower queues;</li>
<li>transmission errors, e.g. due to electrical <li>transmission errors, e.g., due to electrical
interference;</li> interference; and</li>
<li>rate policing.</li> <li>rate policing.</li>
</ul> </ul>
<t>Three complementary approaches are in progress to address this <t>Three complementary approaches are in progress to address this
issue, but they are all currently research:</t> issue, but they are all currently research:</t>
<ul spacing="normal"> <ul spacing="normal">
<li>In Prague congestion control, ignore certain losses deemed <li>In Prague congestion control, ignore certain losses deemed
unlikely to be due to congestion (using some ideas from unlikely to be due to congestion (using some ideas from
BBR <xref target="I-D.cardwell-iccrg-bbr-congestion-control" forma t="default"/> regarding BBR <xref target="I-D.cardwell-iccrg-bbr-congestion-control" forma t="default"/> regarding
isolated losses). This could mask any of the above types of loss isolated losses). This could mask any of the above types of loss
while still coexisting with drop-based congestion controls.</li> while still coexisting with drop-based congestion controls.</li>
<li>A combination of RACK, L4S and link retransmission without <li>A combination of Recent Acknowledgement (RACK) <xref target="RFC 8985" format="default"/>, L4S, and link retransmission without
resequencing could repair transmission errors without the head resequencing could repair transmission errors without the head
of line blocking delay usually associated with link-layer of line blocking delay usually associated with link-layer
retransmission <xref target="UnorderedLTE" format="default"/>, <xr ef target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/>;</li> retransmission <xref target="UnorderedLTE" format="default"/> <xre f target="RFC9331" format="default"/>.</li>
<li>Hybrid ECN/drop rate policers (see <xref target="l4s_arch_sec_po licing" format="default"/>).</li> <li>Hybrid ECN/drop rate policers (see <xref target="l4s_arch_sec_po licing" format="default"/>).</li>
</ul> </ul>
<t>L4S deployment scenarios that minimize these issues <t>L4S deployment scenarios that minimize these issues
(e.g. over wireline networks) can proceed in parallel to this (e.g., over wireline networks) can proceed in parallel to this
research, in the expectation that research success could continually research, in the expectation that research success could continually
widen L4S applicability.</t> widen L4S applicability.</t>
</section> </section>
<section anchor="l4sarch_sec_classic-ecn-neck" numbered="true" toc="defa ult"> <section anchor="l4sarch_sec_classic-ecn-neck" numbered="true" toc="defa ult">
<name>L4S Flow but Classic ECN Bottleneck</name> <name>L4S Flow but Classic ECN Bottleneck</name>
<t>Classic ECN support is starting to materialize on the Internet as <t>Classic ECN support is starting to materialize on the Internet as
an increased level of CE marking. It is hard to detect whether this an increased level of CE marking. It is hard to detect whether this
is all due to the addition of support for ECN in implementations of is all due to the addition of support for ECN in implementations of
FQ-CoDel and/or FQ-COBALT, which is not generally problematic, FQ-CoDel and/or FQ-COBALT, which is not generally problematic,
because flow-queue (FQ) scheduling inherently prevents a flow from because flow queue (FQ) scheduling inherently prevents a flow from
exceeding the 'fair' rate irrespective of its aggressiveness. exceeding the 'fair' rate irrespective of its aggressiveness.
However, some of this Classic ECN marking might be due to However, some of this Classic ECN marking might be due to
single-queue ECN deployment. This case is discussed in Section 4.3 single-queue ECN deployment. This case is discussed in
of the L4S ECN spec <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="d <xref target="RFC9331" format="default" sectionFormat="of" section="4.
efault"/>.</t> 3"> the L4S ECN spec</xref>.</t>
</section> </section>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>L4S AQM Deployment within Tunnels</name> <name>L4S AQM Deployment within Tunnels</name>
<t>An L4S AQM uses the ECN field to signal congestion. So, in common <t>An L4S AQM uses the ECN field to signal congestion. So in common
with Classic ECN, if the AQM is within a tunnel or at a lower layer, with Classic ECN, if the AQM is within a tunnel or at a lower layer,
correct functioning of ECN signalling requires correct propagation correct functioning of ECN signalling requires standards-compliant pro
of the ECN field up the layers <xref target="RFC6040" format="default" pagation
/>, <xref target="I-D.ietf-tsvwg-rfc6040update-shim" format="default"/>, <xref t of the ECN field up the layers <xref target="RFC6040" format="default"
arget="I-D.ietf-tsvwg-ecn-encap-guidelines" format="default"/>.</t> /> <xref target="I-D.ietf-tsvwg-rfc6040update-shim" format="default"/> <xref tar
get="I-D.ietf-tsvwg-ecn-encap-guidelines" format="default"/>.</t>
</section> </section>
</section> </section>
</section> </section>
<section anchor="l4sps_IANA" numbered="true" toc="default"> <section anchor="l4sps_IANA" numbered="true" toc="default">
<name>IANA Considerations (to be removed by RFC Editor)</name> <name>IANA Considerations</name>
<t>This specification contains no IANA considerations.</t> <t>This document has no IANA actions.</t>
</section> </section>
<section anchor="l4sps_Security_Considerations" numbered="true" toc="default "> <section anchor="l4sps_Security_Considerations" numbered="true" toc="default ">
<name>Security Considerations</name> <name>Security Considerations</name>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>Traffic Rate (Non-)Policing</name> <name>Traffic Rate (Non-)Policing</name>
<t/>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>(Non-)Policing Rate per Flow</name> <name>(Non-)Policing Rate per Flow</name>
<t>In the current Internet, ISPs usually enforce separation between <t>In the current Internet, ISPs usually enforce separation between
the capacity of shared links assigned to different 'sites' the capacity of shared links assigned to different 'sites'
(e.g. households, businesses or mobile users - see terminology (e.g., households, businesses, or mobile users -- see terminology
in <xref target="l4sps_Terminology" format="default"/>) using some for m of in <xref target="l4sps_Terminology" format="default"/>) using some for m of
scheduler <xref target="RFC0970" format="default"/>. And they use vari scheduler <xref target="RFC0970" format="default"/>. And they use vari
ous ous
techniques like redirection to traffic scrubbing facilities to deal techniques, like redirection to traffic scrubbing facilities, to deal
with flooding attacks. However, there has never been a universal with flooding attacks. However, there has never been a universal
need to police the rate of individual application flows - the need to police the rate of individual application flows -- the
Internet has generally always relied on self-restraint of congestion Internet has generally always relied on self-restraint of congestion
controls at senders for sharing intra-'site' capacity.</t> controls at senders for sharing intra-'site' capacity.</t>
<t>L4S has been designed not to upset this status quo. If a DualQ is <t>L4S has been designed not to upset this status quo. If a DualQ is
used to provide L4S service, section 4.2 of <xref target="I-D.ietf-tsv wg-aqm-dualq-coupled" format="default"/> explains how it is used to provide L4S service, <xref target="RFC9332" format="default" s ectionFormat="of" section="4.2"/> explains how it is
designed to give no more rate advantage to unresponsive flows than a designed to give no more rate advantage to unresponsive flows than a
single-queue AQM would, whether or not there is traffic single-queue AQM would, whether or not there is traffic
overload.</t> overload.</t>
<t>Also, in case per-flow rate policing is ever required, it can be <t>Also, in case per-flow rate policing is ever required, it can be
added because it is orthogonal to the distinction between L4S and added because it is orthogonal to the distinction between L4S and
Classic. As explained in <xref target="l4sps_why-not" format="default" />, the DualQ Classic. As explained in <xref target="l4sps_why-not" format="default" />, the DualQ
variant of L4S provides low delay without prejudging the issue of variant of L4S provides low delay without prejudging the issue of
flow-rate control. So, if flow-rate control is needed, flow-rate control. So if flow-rate control is needed,
per-flow-queuing (FQ) with L4S support can be used instead, or flow per-flow queuing (FQ) with L4S support can be used instead, or flow
rate policing can be added as a modular addition to a DualQ. rate policing can be added as a modular addition to a DualQ.
However, per-flow rate control is not usually deployed as a security However, per-flow rate control is not usually deployed as a security
mechanism, because an active attacker can just shard its traffic mechanism, because an active attacker can just shard its traffic
over more flow IDs if the rate of each is restricted.</t> over more flow identifiers if the rate of each is restricted.</t>
</section> </section>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>(Non-)Policing L4S Service Rate</name> <name>(Non-)Policing L4S Service Rate</name>
<t><xref target="l4sps_why-not" format="default"/> explains how Diffse rv only makes a <t><xref target="l4sps_why-not" format="default"/> explains how Diffse rv only makes a
difference if some packets get less favourable treatment than difference if some packets get less favourable treatment than
others, which typically requires traffic rate policing for a low others, which typically requires traffic rate policing for a low
latency class. In contrast, it should not be necessary to latency class. In contrast, it should not be necessary to
rate-police access to the L4S service to protect the Classic rate-police access to the L4S service to protect the Classic
service, because L4S is designed to reduce delay without harming the service, because L4S is designed to reduce delay without harming the
delay or rate of any Classic traffic. </t> delay or rate of any Classic traffic. </t>
<t>During early deployment (and perhaps always), some networks will <t>During early deployment (and perhaps always), some networks will
not offer the L4S service. In general, these networks should not not offer the L4S service. In general, these networks should not
need to police L4S traffic. They are required (by both the ECN need to police L4S traffic. They are required (by both the ECN
spec <xref target="RFC3168" format="default"/> and the L4S ECN spec <x ref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/>) not to change the L4S spec <xref target="RFC3168" format="default"/> and the L4S ECN spec <x ref target="RFC9331" format="default"/>) not to change the L4S
identifier, which would interfere with end-to-end congestion identifier, which would interfere with end-to-end congestion
control. If they already treat ECN traffic as Not-ECT, they can control. If they already treat ECN traffic as Not-ECT, they can
merely treat L4S traffic as Not-ECT too. At a bottleneck, such merely treat L4S traffic as Not-ECT too. At a bottleneck, such
networks will introduce some queuing and dropping. When a scalable networks will introduce some queuing and dropping. When a Scalable
congestion control detects a drop it will have to respond safely congestion control detects a drop, it will have to respond safely
with respect to Classic congestion controls (as required in Section with respect to Classic congestion controls (as required in
4.3 of <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/>). T <xref target="RFC9331" format="default" sectionFormat="of" section="4.
his will 3"/>). This will
degrade the L4S service to be no better (but never worse) than degrade the L4S service to be no better (but never worse) than
Classic best efforts, whenever a non-ECN bottleneck is encountered Classic best efforts whenever a non-ECN bottleneck is encountered
on a path (see <xref target="l4sarch_sec_non-l4s-neck" format="default "/>).</t> on a path (see <xref target="l4sarch_sec_non-l4s-neck" format="default "/>).</t>
<t>In cases that are expected to be rare, networks that solely <t>In cases that are expected to be rare, networks that solely
support Classic ECN <xref target="RFC3168" format="default"/> in a sin gle queue support Classic ECN <xref target="RFC3168" format="default"/> in a sin gle queue
bottleneck might opt to police L4S traffic so as to protect bottleneck might opt to police L4S traffic so as to protect
competing Classic ECN traffic (for instance, see Section 6.1.3 of competing Classic ECN traffic (for instance, see
the L4S operational guidance <xref target="I-D.ietf-tsvwg-l4sops" form <xref target="I-D.ietf-tsvwg-l4sops" format="default" sectionFormat="o
at="default"/>). However, Section 4.3 of the L4S f"
ECN spec <xref target="I-D.ietf-tsvwg-ecn-l4s-id" format="default"/> r section="6.1.3">the L4S operational guidance</xref>). However, <xref
ecommends target="RFC9331" format="default" sectionFormat="of"
section="4.3"> the L4S ECN spec</xref> recommends
that the sender adapts its congestion response to properly coexist that the sender adapts its congestion response to properly coexist
with Classic ECN flows, i.e. reverting to the self-restraint with Classic ECN flows, i.e., reverting to the self-restraint
approach.</t> approach.</t>
<t>Certain network operators might choose to restrict access to the <t>Certain network operators might choose to restrict access to the
L4S service, perhaps only to selected premium customers as a L4S service, perhaps only to selected premium customers as a
value-added service. Their packet classifier (item 2 in <xref target=" l4sps_fig_components" format="default"/>) could identify such customers value-added service. Their packet classifier (item 2 in <xref target=" l4sps_fig_components" format="default"/>) could identify such customers
against some other field (e.g. source address range) as well as against some other field (e.g., source address range), as well as
classifying on the ECN field. If only the ECN L4S identifier classifying on the ECN field. If only the ECN L4S identifier
matched, but not the source address (say), the classifier could matched, but not (say) the source address, the classifier could
direct these packets (from non-premium customers) into the Classic direct these packets (from non-premium customers) into the Classic
queue. Explaining clearly how operators can use additional local queue. Explaining clearly how operators can use additional local
classifiers (see section 5.4 of the L4S ECN spec <xref target="I-D.iet f-tsvwg-ecn-l4s-id" format="default"/>) is intended to remove any classifiers (see <xref target="RFC9331" section="5.4" sectionFormat="o f"/>) is intended to remove any
motivation to clear the L4S identifier. Then at least the L4S ECN motivation to clear the L4S identifier. Then at least the L4S ECN
identifier will be more likely to survive end-to-end even though the identifier will be more likely to survive end to end, even though the
service may not be supported at every hop. Such local arrangements service may not be supported at every hop.
Such local arrangements
would only require simple registered/not-registered packet would only require simple registered/not-registered packet
classification, rather than the managed, application-specific classification, rather than the managed, application-specific
traffic policing against customer-specific traffic contracts that traffic policing against customer-specific traffic contracts that
Diffserv uses.</t> Diffserv uses.</t>
</section> </section>
</section> </section>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>'Latency Friendliness'</name> <name>'Latency Friendliness'</name>
<t>Like the Classic service, the L4S service relies on self-restraint <t>Like the Classic service, the L4S service relies on self-restraint to
- limiting rate in response to congestion. In addition, the L4S limit the rate in response to congestion. In addition, the L4S
service requires self-restraint in terms of limiting latency service requires self-restraint in terms of limiting latency
(burstiness). It is hoped that self-interest and guidance on dynamic (burstiness). It is hoped that self-interest and guidance on dynamic
behaviour (especially flow start-up, which might need to be behaviour (especially flow start-up, which might need to be
standardized) will be sufficient to prevent transports from sending standardized) will be sufficient to prevent transports from sending
excessive bursts of L4S traffic, given the application's own latency excessive bursts of L4S traffic, given the application's own latency
will suffer most from such behaviour.</t> will suffer most from such behaviour.</t>
<t>Because the L4S service can reduce delay without discernibly <t>Because the L4S service can reduce delay without discernibly
increasing the delay of any Classic traffic, it should not be increasing the delay of any Classic traffic, it should not be
necessary to police L4S traffic to protect the delay of Classic. necessary to police L4S traffic to protect the delay of Classic traffic.
However, whether burst policing becomes necessary to protect other L4S However, whether burst policing becomes necessary to protect other L4S
traffic remains to be seen. Without it, there will be potential for traffic remains to be seen. Without it, there will be potential for
attacks on the low latency of the L4S service.</t> attacks on the low latency of the L4S service.</t>
<t>If needed, various arrangements could be used to address this <t>If needed, various arrangements could be used to address this
concern:</t> concern:</t>
<dl newline="false" spacing="normal"> <dl newline="false" spacing="normal">
<dt>Local bottleneck queue protection:</dt> <dt>Local bottleneck queue protection:</dt>
<dd>A per-flow <dd>A per-flow
(5-tuple) queue protection function <xref target="I-D.briscoe-docsis -q-protection" format="default"/> has been developed for (5-tuple) queue protection function <xref target="I-D.briscoe-docsis -q-protection" format="default"/> has been developed for
the low latency queue in DOCSIS, which has adopted the DualQ L4S the low latency queue in DOCSIS, which has adopted the DualQ L4S
architecture. It protects the low latency service from any architecture. It protects the low latency service from any
queue-building flows that accidentally or maliciously classify queue-building flows that accidentally or maliciously classify
themselves into the low latency queue. It is designed to score themselves into the low latency queue. It is designed to score
flows based solely on their contribution to queuing (not flow rate flows based solely on their contribution to queuing (not flow rate
in itself). Then, if the shared low latency queue is at risk of in itself). Then, if the shared low latency queue is at risk of
exceeding a threshold, the function redirects enough packets of exceeding a threshold, the function redirects enough packets of
the highest scoring flow(s) into the Classic queue to preserve low the highest scoring flow(s) into the Classic queue to preserve low
latency.</dd> latency.</dd>
<dt>Distributed traffic scrubbing:</dt> <dt>Distributed traffic scrubbing:</dt>
<dd>Rather than policing <dd>Rather than policing
locally at each bottleneck, it may only be necessary to address locally at each bottleneck, it may only be necessary to address
problems reactively, e.g. punitively target any deployments problems reactively, e.g., punitively target any deployments
of new bursty malware, in a similar way to how traffic from of new bursty malware, in a similar way to how traffic from
flooding attack sources is rerouted via scrubbing facilities.</dd> flooding attack sources is rerouted via scrubbing facilities.</dd>
<dt>Local bottleneck per-flow scheduling:</dt> <dt>Local bottleneck per-flow scheduling:</dt>
<dd>Per-flow <dd>Per-flow
scheduling should inherently isolate non-bursty flows from bursty scheduling should inherently isolate non-bursty flows from bursty fl ows
(see <xref target="l4sps_why-not" format="default"/> for discussion of the merits (see <xref target="l4sps_why-not" format="default"/> for discussion of the merits
of per-flow scheduling relative to per-flow policing).</dd> of per-flow scheduling relative to per-flow policing).</dd>
<dt>Distributed access subnet queue protection:</dt> <dt>Distributed access subnet queue protection:</dt>
<dd>Per-flow <dd>Per-flow
queue protection could be arranged for a queue structure queue protection could be arranged for a queue structure
distributed across a subnet intercommunicating using lower layer distributed across a subnet intercommunicating using lower layer
control messages (see Section 2.1.4 of <xref target="QDyn" format="d efault"/>). For control messages (see Section 2.1.4 of <xref target="QDyn" format="d efault"/>). For
instance, in a radio access network, user equipment already sends instance, in a radio access network, user equipment already sends
regular buffer status reports to a radio network controller, which regular buffer status reports to a radio network controller, which
could use this information to remotely police individual could use this information to remotely police individual
flows.</dd> flows.</dd>
<dt>Distributed Congestion Exposure to Ingress Policers:</dt> <dt>Distributed Congestion Exposure to ingress policers:</dt>
<dd>The <dd>The
Congestion Exposure (ConEx) architecture <xref target="RFC7713" form Congestion Exposure (ConEx) architecture <xref target="RFC7713" form
at="default"/> uses egress audit to motivate senders to at="default"/> uses an egress audit to motivate senders to
truthfully signal path congestion in-band where it can be used by truthfully signal path congestion in-band, where it can be used by
ingress policers. An edge-to-edge variant of this architecture is ingress policers. An edge-to-edge variant of this architecture is
also possible.</dd> also possible.</dd>
<dt>Distributed Domain-edge traffic conditioning:</dt> <dt>Distributed domain-edge traffic conditioning:</dt>
<dd>An <dd>An
architecture similar to Diffserv <xref target="RFC2475" format="defa ult"/> may architecture similar to Diffserv <xref target="RFC2475" format="defa ult"/> may
be preferred, where traffic is proactively conditioned on entry to be preferred, where traffic is proactively conditioned on entry to
a domain, rather than reactively policed only if it leads to a domain, rather than reactively policed only if it leads to
queuing once combined with other traffic at a bottleneck.</dd> queuing once combined with other traffic at a bottleneck.</dd>
<dt>Distributed core network queue protection:</dt> <dt>Distributed core network queue protection:</dt>
<dd>The <dd>The
policing function could be divided between per-flow mechanisms at policing function could be divided between per-flow mechanisms at
the network ingress that characterize the burstiness of each flow the network ingress that characterize the burstiness of each flow
into a signal carried with the traffic, and per-class mechanisms into a signal carried with the traffic and per-class mechanisms
at bottlenecks that act on these signals if queuing actually at bottlenecks that act on these signals if queuing actually
occurs once the traffic converges. This would be somewhat similar occurs once the traffic converges. This would be somewhat similar
to <xref target="Nadas20" format="default"/>, which is in turn simil ar to the idea to <xref target="Nadas20" format="default"/>, which is in turn simil ar to the idea
behind core stateless fair queuing.</dd> behind core stateless fair queuing.</dd>
</dl> </dl>
<t>No single one of these possible queue protection capabilities is <t>No single one of these possible queue protection capabilities is
considered an essential part of the L4S architecture, which works considered an essential part of the L4S architecture, which works
without any of them under non-attack conditions (much as the Internet without any of them under non-attack conditions (much as the Internet
normally works without per-flow rate policing). Indeed, even where normally works without per-flow rate policing).
latency policers are deployed, under normal circumstances they would Indeed, even where
not intervene, and if operators found they were not necessary they latency policers are deployed, under normal circumstances, they would
not intervene, and if operators found they were not necessary, they
could disable them. Part of the L4S experiment will be to see whether could disable them. Part of the L4S experiment will be to see whether
such a function is necessary, and which arrangements are most such a function is necessary and which arrangements are most
appropriate to the size of the problem.</t> appropriate to the size of the problem.</t>
</section> </section>
<section anchor="l4s_arch_sec_policing" numbered="true" toc="default"> <section anchor="l4s_arch_sec_policing" numbered="true" toc="default">
<name>Interaction between Rate Policing and L4S</name> <name>Interaction between Rate Policing and L4S</name>
<t>As mentioned in <xref target="l4sps_why-not" format="default"/>, L4S should remove <t>As mentioned in <xref target="l4sps_why-not" format="default"/>, L4S should remove
the need for low latency Diffserv classes. However, those Diffserv the need for low latency Diffserv classes. However, those Diffserv
classes that give certain applications or users priority over classes that give certain applications or users priority over
capacity, would still be applicable in certain scenarios capacity would still be applicable in certain scenarios
(e.g. corporate networks). Then, within such Diffserv classes, (e.g., corporate networks). Then, within such Diffserv classes,
L4S would often be applicable to give traffic low latency and low loss L4S would often be applicable to give traffic low latency and low loss
as well. Within such a Diffserv class, the bandwidth available to a as well. Within such a Diffserv class, the bandwidth available to a
user or application is often limited by a rate policer. Similarly, in user or application is often limited by a rate policer. Similarly, in
the default Diffserv class, rate policers are sometimes used to the default Diffserv class, rate policers are sometimes used to
partition shared capacity.</t> partition shared capacity.</t>
<t>A classic rate policer drops any packets exceeding a set rate, <t>A Classic rate policer drops any packets exceeding a set rate,
usually also giving a burst allowance (variants exist where the usually also giving a burst allowance (variants exist where the
policer re-marks non-compliant traffic to a discard-eligible Diffserv policer re-marks noncompliant traffic to a discard-eligible Diffserv
codepoint, so they can be dropped elsewhere during contention). codepoint, so they can be dropped elsewhere during contention).
Whenever L4S traffic encounters one of these rate policers, it will Whenever L4S traffic encounters one of these rate policers, it will
experience drops and the source will have to fall back to a Classic experience drops and the source will have to fall back to a Classic
congestion control, thus losing the benefits of L4S (<xref target="l4sar ch_sec_non-l4s-neck" format="default"/>). So, in networks that already use congestion control, thus losing the benefits of L4S (<xref target="l4sar ch_sec_non-l4s-neck" format="default"/>). So in networks that already use
rate policers and plan to deploy L4S, it will be preferable to rate policers and plan to deploy L4S, it will be preferable to
redesign these rate policers to be more friendly to the L4S redesign these rate policers to be more friendly to the L4S
service.</t> service.</t>
<t>L4S-friendly rate policing is currently a research area (note that <t>L4S-friendly rate policing is currently a research area (note that
this is not the same as latency policing). It might be achieved by this is not the same as latency policing). It might be achieved by
setting a threshold where ECN marking is introduced, such that it is setting a threshold where ECN marking is introduced, such that it is
just under the policed rate or just under the burst allowance where just under the policed rate or just under the burst allowance where
drop is introduced. For instance the two-rate three-colour drop is introduced. For instance, the two-rate, three-colour
marker <xref target="RFC2698" format="default"/> or a PCN threshold and marker <xref target="RFC2698" format="default"/> or a PCN threshold and
excess-rate marker <xref target="RFC5670" format="default"/> could mark excess-rate marker <xref target="RFC5670" format="default"/> could mark
ECN at the ECN at the
lower rate and drop at the higher. Or an existing rate policer could lower rate and drop at the higher. Or an existing rate policer could
have congestion-rate policing added, e.g. using the 'local' have congestion-rate policing added, e.g., using the 'local'
(non-ConEx) variant of the ConEx aggregate congestion (non-ConEx) variant of the ConEx aggregate congestion
policer <xref target="I-D.briscoe-conex-policing" format="default"/>. It policer <xref target="I-D.briscoe-conex-policing" format="default"/>. It
might might
also be possible to design scalable congestion controls to respond also be possible to design Scalable congestion controls to respond
less catastrophically to loss that has not been preceded by a period less catastrophically to loss that has not been preceded by a period
of increasing delay.</t> of increasing delay.</t>
<t>The design of L4S-friendly rate policers will require a separate <t>The design of L4S-friendly rate policers will require a separate,
dedicated document. For further discussion of the interaction between dedicated document. For further discussion of the interaction between
L4S and Diffserv, see <xref target="I-D.briscoe-tsvwg-l4s-diffserv" form at="default"/>.</t> L4S and Diffserv, see <xref target="I-D.briscoe-tsvwg-l4s-diffserv" form at="default"/>.</t>
</section> </section>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>ECN Integrity</name> <name>ECN Integrity</name>
<t>Various ways have been developed to protect the integrity of the <t>Various ways have been developed to protect the integrity of the
congestion feedback loop (whether signalled by loss, Classic ECN or congestion feedback loop (whether signalled by loss, Classic ECN, or
L4S ECN) against misbehaviour by the receiver, sender or network (or L4S ECN) against misbehaviour by the receiver, sender, or network (or
all three). Brief details of each including applicability, pros and all three). Brief details of each, including applicability, pros, and
cons is given in Appendix C.1 of the L4S ECN spec <xref target="I-D.ietf cons, are given in <xref target="RFC9331" format="default" sectionFormat
-tsvwg-ecn-l4s-id" format="default"/>.</t> ="of" section="C.1">the L4S ECN spec</xref>.</t>
</section> </section>
<section numbered="true" toc="default"> <section numbered="true" toc="default">
<name>Privacy Considerations</name> <name>Privacy Considerations</name>
<t>As discussed in <xref target="l4sps_why-not" format="default"/>, the L4S <t>As discussed in <xref target="l4sps_why-not" format="default"/>, the L4S
architecture does not preclude approaches that inspect end-to-end architecture does not preclude approaches that inspect end-to-end
transport layer identifiers. For instance, L4S support has been added transport layer identifiers. For instance, L4S support has been added
to FQ-CoDel, which classifies by application flow ID in the network. to FQ-CoDel, which classifies by application flow identifier in the netw ork.
However, the main innovation of L4S is the DualQ AQM framework that However, the main innovation of L4S is the DualQ AQM framework that
does not need to inspect any deeper than the outermost IP header, does not need to inspect any deeper than the outermost IP header,
because the L4S identifier is in the IP-ECN field.</t> because the L4S identifier is in the IP-ECN field.</t>
<t>Thus, the L4S architecture enables very low queuing delay without <t>Thus, the L4S architecture enables very low queuing delay without
<em>requiring</em> inspection of information above <em>requiring</em> inspection of information above
the IP layer. This means that users who want to encrypt application the IP layer. This means that users who want to encrypt application
flow identifiers, e.g. in IPSec or other encrypted VPN tunnels, flow identifiers, e.g., in IPsec or other encrypted VPN tunnels,
don't have to sacrifice low delay <xref target="RFC8404" format="default don't have to sacrifice low delay <xref target="RFC8404" format="default
"/>.</t> "/>.</t>
<t>Because L4S can provide low delay for a broad set of applications <t>Because L4S can provide low delay for a broad set of applications
that choose to use it, there is no need for individual applications or that choose to use it, there is no need for individual applications or
classes within that broad set to be distinguishable in any way while classes within that broad set to be distinguishable in any way while
traversing networks. This removes much of the ability to correlate traversing networks. This removes much of the ability to correlate
between the delay requirements of traffic and other identifying between the delay requirements of traffic and other identifying
features <xref target="RFC6973" format="default"/>. There may be some ty pes of features <xref target="RFC6973" format="default"/>. There may be some ty pes of
traffic that prefer not to use L4S, but the coarse binary traffic that prefer not to use L4S, but the coarse binary
categorization of traffic reveals very little that could be exploited categorization of traffic reveals very little that could be exploited
to compromise privacy.</t> to compromise privacy.</t>
</section> </section>
</section> </section>
</middle> </middle>
<!-- *****BACK MATTER ***** -->
<back> <back>
<displayreference target="I-D.ietf-tcpm-accurate-ecn" to="ACCECN"/>
<displayreference target="I-D.ietf-tsvwg-nqb" to="NQB-PHB"/>
<displayreference target="I-D.briscoe-conex-policing" to="CONG-POLICING"/>
<displayreference target="I-D.stewart-tsvwg-sctpecn" to="ECN-SCTP"/>
<displayreference target="I-D.sridharan-tcpm-ctcp" to="CTCP"/>
<displayreference target="I-D.ietf-tsvwg-rfc6040update-shim" to="ECN-SHIM"/>
<displayreference target="I-D.ietf-tsvwg-ecn-encap-guidelines" to="ECN-ENCAP"/>
<displayreference target="I-D.ietf-tsvwg-l4sops" to="L4SOPS"/>
<displayreference target="I-D.briscoe-tsvwg-l4s-diffserv" to="L4S-DIFFSERV"/>
<displayreference target="I-D.briscoe-docsis-q-protection" to="DOCSIS-Q-PROT"/>
<displayreference target="I-D.cardwell-iccrg-bbr-congestion-control" to="BBR-CC"
/>
<displayreference target="I-D.briscoe-iccrg-prague-congestion-control" to="PRAGU
E-CC"/>
<displayreference target="I-D.morton-tsvwg-codel-approx-fair" to="CODEL-APPROX-F
AIR"/>
<displayreference target="I-D.mathis-iccrg-relentless-tcp" to="RELENTLESS"/>
<references> <references>
<name>Informative References</name> <name>Informative References</name>
<reference anchor="RFC0970" target="https://www.rfc-editor.org/info/rfc970
" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.0970.xml"> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.0970.
<front> xml"/>
<title>On Packet Switches With Infinite Storage</title> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2475.
<author initials="J." surname="Nagle" fullname="J. Nagle"> xml"/>
<organization/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2698.
</author> xml"/>
<date year="1985" month="December"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2884.
<abstract> xml"/>
<t>The purpose of this RFC is to focus discussion on a particular pr <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3168.
oblem in the ARPA-Internet and possible methods of solution. Most prior work xml"/>
on congestion in datagram systems focuses on buffer management. In this <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.4774.
memo the case of a packet switch with infinite storage is considered. Such a xml"/>
packet switch can never run out of buffers. It can, however, still become co <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6679.
ngested. The meaning of congestion in an infinite-storage system is explored xml"/>
. An unexpected result is found that shows a datagram network with infinite <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3540.
storage, first-in-first-out queuing, at least two packet switches, and a fini xml"/>
te packet lifetime will, under overload, drop all packets. By attacking the <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3246.
problem of congestion for the infinite-storage case, new solutions applicable xml"/>
to switches with finite storage may be found. No proposed solutions this <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3649.
document are intended as standards for the ARPA-Internet at this time.</t> xml"/>
</abstract> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.4340.
</front> xml"/>
<seriesInfo name="RFC" value="970"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.4960.
<seriesInfo name="DOI" value="10.17487/RFC0970"/> xml"/>
</reference> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5033.
<reference anchor="RFC2475" target="https://www.rfc-editor.org/info/rfc247 xml"/>
5" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.2475.xml"> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5348.
<front> xml"/>
<title>An Architecture for Differentiated Services</title> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5670.
<author initials="S." surname="Blake" fullname="S. Blake"> xml"/>
<organization/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5681.
</author> xml"/>
<author initials="D." surname="Black" fullname="D. Black"> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6040.
<organization/> xml"/>
</author> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6817.
<author initials="M." surname="Carlson" fullname="M. Carlson"> xml"/>
<organization/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6973.
</author> xml"/>
<author initials="E." surname="Davies" fullname="E. Davies"> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7560.
<organization/> xml"/>
</author> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7665.
<author initials="Z." surname="Wang" fullname="Z. Wang"> xml"/>
<organization/>
</author> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.ietf-tc
<author initials="W." surname="Weiss" fullname="W. Weiss"> pm-accurate-ecn.xml"/>
<organization/>
</author> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7713.
<date year="1998" month="December"/> xml"/>
<abstract> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7567.
<t>This document defines an architecture for implementing scalable s xml"/>
ervice differentiation in the Internet. This memo provides information for the <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8033.
Internet community.</t> xml"/>
</abstract> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8034.
</front> xml"/>
<seriesInfo name="RFC" value="2475"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8170.
<seriesInfo name="DOI" value="10.17487/RFC2475"/> xml"/>
</reference> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8257.
<reference anchor="RFC2698" target="https://www.rfc-editor.org/info/rfc269 xml"/>
8" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.2698.xml"> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8290.
<front> xml"/>
<title>A Two Rate Three Color Marker</title> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8298.
<author initials="J." surname="Heinanen" fullname="J. Heinanen"> xml"/>
<organization/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8311.
</author> xml"/>
<author initials="R." surname="Guerin" fullname="R. Guerin"> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8312.
<organization/> xml"/>
</author> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8404.
<date year="1999" month="September"/> xml"/>
<abstract> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8511.
<t>This document defines a Two Rate Three Color Marker (trTCM), whic xml"/>
h can be used as a component in a Diffserv traffic conditioner. This memo provi <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8888.
des information for the Internet community.</t> xml"/>
</abstract> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8985.
</front> xml"/>
<seriesInfo name="RFC" value="2698"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.9000.
<seriesInfo name="DOI" value="10.17487/RFC2698"/> xml"/>
</reference> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.9113.
<reference anchor="RFC2884" target="https://www.rfc-editor.org/info/rfc288 xml"/>
4" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.2884.xml">
<front> <reference anchor="RFC9332" target="https://www.rfc-editor.org/info/rfc9332">
<title>Performance Evaluation of Explicit Congestion Notification (ECN <front>
) in IP Networks</title> <title>Dual-Queue Coupled Active Queue Management (AQM) for Low Latency, Low
<author initials="J." surname="Hadi Salim" fullname="J. Hadi Salim"> Loss, and Scalable Throughput (L4S)</title>
<organization/> <author initials="K" surname="De Schepper" fullname="Koen De Schepper">
</author> <organization>Nokia Bell Labs</organization>
<author initials="U." surname="Ahmed" fullname="U. Ahmed"> </author>
<organization/> <author initials="B" surname="Briscoe" fullname="Bob Briscoe" role="editor">
</author> <organization>Independent</organization>
<date year="2000" month="July"/> </author>
<abstract> <author initials="G" surname="White" fullname="Greg White">
<t>This memo presents a performance study of the Explicit Congestion <organization>CableLabs</organization>
Notification (ECN) mechanism in the TCP/IP protocol using our implementation on </author>
the Linux Operating System. This memo provides information for the Internet co <date month="January" year="2023"/>
mmunity.</t> </front>
</abstract> <seriesInfo name="RFC" value="9332"/>
</front> <seriesInfo name="DOI" value="10.17487/RFC9332"/>
<seriesInfo name="RFC" value="2884"/> </reference>
<seriesInfo name="DOI" value="10.17487/RFC2884"/>
</reference> <reference anchor="RFC9331" target="https://www.rfc-editor.org/info/rfc9331">
<reference anchor="RFC3168" target="https://www.rfc-editor.org/info/rfc316 <front>
8" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.3168.xml"> <title>The Explicit Congestion Notification (ECN) Protocol for Low Latency,
<front> Low Loss, and Scalable Throughput (L4S)</title>
<title>The Addition of Explicit Congestion Notification (ECN) to IP</t <author initials="K" surname="De Schepper" fullname="Koen De Schepper">
itle> <organization>Nokia Bell Labs</organization>
<author initials="K." surname="Ramakrishnan" fullname="K. Ramakrishnan </author>
"> <author initials="B" surname="Briscoe" fullname="Bob Briscoe" role="editor">
<organization/> <organization>Independent</organization>
</author> </author>
<author initials="S." surname="Floyd" fullname="S. Floyd"> <date month="January" year="2023"/>
<organization/> </front>
</author> <seriesInfo name="RFC" value="9331"/>
<author initials="D." surname="Black" fullname="D. Black"> <seriesInfo name="DOI" value="10.17487/RFC9331"/>
<organization/> </reference>
</author>
<date year="2001" month="September"/> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.ietf-ts
<abstract> vwg-nqb.xml"/>
<t>This memo specifies the incorporation of ECN (Explicit Congestion <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.briscoe
Notification) to TCP and IP, including ECN's use of two bits in the IP header. -conex-policing.xml"/>
[STANDARDS-TRACK]</t> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.stewart
</abstract> -tsvwg-sctpecn.xml"/>
</front> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.sridhar
<seriesInfo name="RFC" value="3168"/> an-tcpm-ctcp.xml"/>
<seriesInfo name="DOI" value="10.17487/RFC3168"/> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.ietf-ts
</reference> vwg-rfc6040update-shim.xml"/>
<reference anchor="RFC4774" target="https://www.rfc-editor.org/info/rfc477 <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.ietf-ts
4" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.4774.xml"> vwg-ecn-encap-guidelines.xml"/>
<front>
<title>Specifying Alternate Semantics for the Explicit Congestion Noti <reference anchor="I-D.ietf-tsvwg-l4sops" target="https://datatracker.ietf.org/d
fication (ECN) Field</title> oc/html/draft-ietf-tsvwg-l4sops-03">
<author initials="S." surname="Floyd" fullname="S. Floyd"> <front>
<organization/> <title>
</author> Operational Guidance for Deployment of L4S in the Internet
<date year="2006" month="November"/> </title>
<abstract> <author fullname="Greg White" initials="G." surname="White" role="editor">
<t>There have been a number of proposals for alternate semantics for <organization>CableLabs</organization>
the Explicit Congestion Notification (ECN) field in the IP header RFC 3168. Th </author>
is document discusses some of the issues in defining alternate semantics for the <date month="April" day="28" year="2022"/>
ECN field, and specifies requirements for a safe coexistence in an Internet tha </front>
t could include routers that do not understand the defined alternate semantics. <seriesInfo name="Internet-Draft" value="draft-ietf-tsvwg-l4sops-03"/>
This document evolved as a result of discussions with the authors of one recent <format type="TXT" target="https://www.ietf.org/archive/id/draft-ietf-tsvwg-l4so
proposal for such alternate semantics. This document specifies an Internet Bes ps-03.txt"/>
t Current Practices for the Internet Community, and requests discussion and sugg </reference>
estions for improvements.</t>
</abstract> <xi:include
</front> href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.briscoe-tsvwg-l4s-d
<seriesInfo name="BCP" value="124"/> iffserv.xml"/>
<seriesInfo name="RFC" value="4774"/>
<seriesInfo name="DOI" value="10.17487/RFC4774"/> <reference anchor="I-D.briscoe-docsis-q-protection" target="https://datatracker.
</reference> ietf.org/doc/html/draft-briscoe-docsis-q-protection-06">
<reference anchor="RFC6679" target="https://www.rfc-editor.org/info/rfc667 <front>
9" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.6679.xml"> <title>The DOCSIS(R) Queue Protection Algorithm to Preserve Low Latency</tit
<front> le>
<title>Explicit Congestion Notification (ECN) for RTP over UDP</title> <author initials="B" surname="Briscoe" fullname="Bob Briscoe" role="editor">
<author initials="M." surname="Westerlund" fullname="M. Westerlund"> <organization>Independent</organization>
<organization/> </author>
</author> <author initials="G" surname="White" fullname="Greg White">
<author initials="I." surname="Johansson" fullname="I. Johansson"> <organization>CableLabs</organization>
<organization/> </author>
</author> <date day="13" month="May" year="2022"/>
<author initials="C." surname="Perkins" fullname="C. Perkins"> </front>
<organization/> <seriesInfo name="Internet-Draft" value="draft-briscoe-docsis-q-protection-06"
</author> />
<author initials="P." surname="O'Hanlon" fullname="P. O'Hanlon"> </reference>
<organization/>
</author> <xi:include
<author initials="K." surname="Carlberg" fullname="K. Carlberg"> href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.cardwell-iccrg-bbr-
<organization/> congestion-control.xml"/>
</author>
<date year="2012" month="August"/> <reference anchor="I-D.briscoe-iccrg-prague-congestion-control" target="https://
<abstract> datatracker.ietf.org/doc/html/draft-briscoe-iccrg-prague-congestion-control-01">
<t>This memo specifies how Explicit Congestion Notification (ECN) ca <front>
n be used with the Real-time Transport Protocol (RTP) running over UDP, using th <title>Prague Congestion Control</title>
e RTP Control Protocol (RTCP) as a feedback mechanism. It defines a new RTCP Ex <author initials="K" surname="De Schepper">
tended Report (XR) block for periodic ECN feedback, a new RTCP transport feedbac <organization>Nokia Bell Labs</organization>
k message for timely reporting of congestion events, and a Session Traversal Uti </author>
lities for NAT (STUN) extension used in the optional initialisation method using <author fullname="Olivier Tilmans">
Interactive Connectivity Establishment (ICE). Signalling and procedures for ne <organization>Nokia Bell Labs</organization>
gotiation of capabilities and initialisation methods are also defined. [STANDAR </author>
DS-TRACK]</t> <author fullname="Bob Briscoe" role="editor">
</abstract> <organization>Independent</organization>
</front> </author>
<seriesInfo name="RFC" value="6679"/> <date day="11" month="July" year="2022"/>
<seriesInfo name="DOI" value="10.17487/RFC6679"/> </front>
</reference> <seriesInfo name="Internet-Draft" value="draft-briscoe-iccrg-prague-congestion
<reference anchor="RFC3540" target="https://www.rfc-editor.org/info/rfc354 -control-01"/>
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.3540.xml"> </reference>
<front>
<title>Robust Explicit Congestion Notification (ECN) Signaling with No <xi:include
nces</title> href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.morton-tsvwg-codel-
<author initials="N." surname="Spring" fullname="N. Spring"> approx-fair.xml"/>
<organization/>
</author>
<author initials="D." surname="Wetherall" fullname="D. Wetherall">
<organization/>
</author>
<author initials="D." surname="Ely" fullname="D. Ely">
<organization/>
</author>
<date year="2003" month="June"/>
<abstract>
<t>This note describes the Explicit Congestion Notification (ECN)-no
nce, an optional addition to ECN that protects against accidental or malicious c
oncealment of marked packets from the TCP sender. It improves the robustness of
congestion control by preventing receivers from exploiting ECN to gain an unfai
r share of network bandwidth. The ECN-nonce uses the two ECN-Capable Transport
(ECT)codepoints in the ECN field of the IP header, and requires a flag in the TC
P header. It is computationally efficient for both routers and hosts. This mem
o defines an Experimental Protocol for the Internet community.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="3540"/>
<seriesInfo name="DOI" value="10.17487/RFC3540"/>
</reference>
<reference anchor="RFC3246" target="https://www.rfc-editor.org/info/rfc324
6" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.3246.xml">
<front>
<title>An Expedited Forwarding PHB (Per-Hop Behavior)</title>
<author initials="B." surname="Davie" fullname="B. Davie">
<organization/>
</author>
<author initials="A." surname="Charny" fullname="A. Charny">
<organization/>
</author>
<author initials="J.C.R." surname="Bennet" fullname="J.C.R. Bennet">
<organization/>
</author>
<author initials="K." surname="Benson" fullname="K. Benson">
<organization/>
</author>
<author initials="J.Y." surname="Le Boudec" fullname="J.Y. Le Boudec">
<organization/>
</author>
<author initials="W." surname="Courtney" fullname="W. Courtney">
<organization/>
</author>
<author initials="S." surname="Davari" fullname="S. Davari">
<organization/>
</author>
<author initials="V." surname="Firoiu" fullname="V. Firoiu">
<organization/>
</author>
<author initials="D." surname="Stiliadis" fullname="D. Stiliadis">
<organization/>
</author>
<date year="2002" month="March"/>
<abstract>
<t>This document defines a PHB (per-hop behavior) called Expedited F
orwarding (EF). The PHB is a basic building block in the Differentiated Service
s architecture. EF is intended to provide a building block for low delay, low j
itter and low loss services by ensuring that the EF aggregate is served at a cer
tain configured rate. This document obsoletes RFC 2598. [STANDARDS-TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="3246"/>
<seriesInfo name="DOI" value="10.17487/RFC3246"/>
</reference>
<reference anchor="RFC3649" target="https://www.rfc-editor.org/info/rfc364
9" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.3649.xml">
<front>
<title>HighSpeed TCP for Large Congestion Windows</title>
<author initials="S." surname="Floyd" fullname="S. Floyd">
<organization/>
</author>
<date year="2003" month="December"/>
<abstract>
<t>The proposals in this document are experimental. While they may
be deployed in the current Internet, they do not represent a consensus that this
is the best method for high-speed congestion control. In particular, we note t
hat alternative experimental proposals are likely to be forthcoming, and it is n
ot well understood how the proposals in this document will interact with such al
ternative proposals. This document proposes HighSpeed TCP, a modification to TC
P's congestion control mechanism for use with TCP connections with large congest
ion windows. The congestion control mechanisms of the current Standard TCP cons
trains the congestion windows that can be achieved by TCP in realistic environme
nts. For example, for a Standard TCP connection with 1500-byte packets and a 10
0 ms round-trip time, achieving a steady-state throughput of 10 Gbps would requi
re an average congestion window of 83,333 segments, and a packet drop rate of at
most one congestion event every 5,000,000,000 packets (or equivalently, at most
one congestion event every 1 2/3 hours). This is widely acknowledged as an unr
ealistic constraint. To address his limitation of TCP, this document proposes H
ighSpeed TCP, and solicits experimentation and feedback from the wider community
.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="3649"/>
<seriesInfo name="DOI" value="10.17487/RFC3649"/>
</reference>
<reference anchor="RFC4340" target="https://www.rfc-editor.org/info/rfc434
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.4340.xml">
<front>
<title>Datagram Congestion Control Protocol (DCCP)</title>
<author initials="E." surname="Kohler" fullname="E. Kohler">
<organization/>
</author>
<author initials="M." surname="Handley" fullname="M. Handley">
<organization/>
</author>
<author initials="S." surname="Floyd" fullname="S. Floyd">
<organization/>
</author>
<date year="2006" month="March"/>
<abstract>
<t>The Datagram Congestion Control Protocol (DCCP) is a transport pr
otocol that provides bidirectional unicast connections of congestion-controlled
unreliable datagrams. DCCP is suitable for applications that transfer fairly la
rge amounts of data and that can benefit from control over the tradeoff between
timeliness and reliability. [STANDARDS-TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="4340"/>
<seriesInfo name="DOI" value="10.17487/RFC4340"/>
</reference>
<reference anchor="RFC4960" target="https://www.rfc-editor.org/info/rfc496
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.4960.xml">
<front>
<title>Stream Control Transmission Protocol</title>
<author initials="R." surname="Stewart" fullname="R. Stewart" role="ed
itor">
<organization/>
</author>
<date year="2007" month="September"/>
<abstract>
<t>This document obsoletes RFC 2960 and RFC 3309. It describes the
Stream Control Transmission Protocol (SCTP). SCTP is designed to transport Publ
ic Switched Telephone Network (PSTN) signaling messages over IP networks, but is
capable of broader applications.</t>
<t>SCTP is a reliable transport protocol operating on top of a conne
ctionless packet network such as IP. It offers the following services to its us
ers:</t>
<t>-- acknowledged error-free non-duplicated transfer of user data,
</t>
<t>-- data fragmentation to conform to discovered path MTU size,</t
>
<t>-- sequenced delivery of user messages within multiple streams,
with an option for order-of-arrival delivery of individual user messages,</t>
<t>-- optional bundling of multiple user messages into a single SCT
P packet, and</t>
<t>-- network-level fault tolerance through supporting of multi-hom
ing at either or both ends of an association.</t>
<t> The design of SCTP includes appropriate congestion avoidance beh
avior and resistance to flooding and masquerade attacks. [STANDARDS-TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="4960"/>
<seriesInfo name="DOI" value="10.17487/RFC4960"/>
</reference>
<reference anchor="RFC5033" target="https://www.rfc-editor.org/info/rfc503
3" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.5033.xml">
<front>
<title>Specifying New Congestion Control Algorithms</title>
<author initials="S." surname="Floyd" fullname="S. Floyd">
<organization/>
</author>
<author initials="M." surname="Allman" fullname="M. Allman">
<organization/>
</author>
<date year="2007" month="August"/>
<abstract>
<t>The IETF's standard congestion control schemes have been widely s
hown to be inadequate for various environments (e.g., high-speed networks). Rec
ent research has yielded many alternate congestion control schemes that signific
antly differ from the IETF's congestion control principles. Using these new con
gestion control schemes in the global Internet has possible ramifications to bot
h the traffic using the new congestion control and to traffic using the currentl
y standardized congestion control. Therefore, the IETF must proceed with cautio
n when dealing with alternate congestion control proposals. The goal of this do
cument is to provide guidance for considering alternate congestion control algor
ithms within the IETF. This document specifies an Internet Best Current Practic
es for the Internet Community, and requests discussion and suggestions for impro
vements.</t>
</abstract>
</front>
<seriesInfo name="BCP" value="133"/>
<seriesInfo name="RFC" value="5033"/>
<seriesInfo name="DOI" value="10.17487/RFC5033"/>
</reference>
<reference anchor="RFC5348" target="https://www.rfc-editor.org/info/rfc534
8" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.5348.xml">
<front>
<title>TCP Friendly Rate Control (TFRC): Protocol Specification</title
>
<author initials="S." surname="Floyd" fullname="S. Floyd">
<organization/>
</author>
<author initials="M." surname="Handley" fullname="M. Handley">
<organization/>
</author>
<author initials="J." surname="Padhye" fullname="J. Padhye">
<organization/>
</author>
<author initials="J." surname="Widmer" fullname="J. Widmer">
<organization/>
</author>
<date year="2008" month="September"/>
<abstract>
<t>This document specifies TCP Friendly Rate Control (TFRC). TFRC i
s a congestion control mechanism for unicast flows operating in a best-effort In
ternet environment. It is reasonably fair when competing for bandwidth with TCP
flows, but has a much lower variation of throughput over time compared with TCP
, making it more suitable for applications such as streaming media where a relat
ively smooth sending rate is of importance.</t>
<t>This document obsoletes RFC 3448 and updates RFC 4342. [STANDARD
S-TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="5348"/>
<seriesInfo name="DOI" value="10.17487/RFC5348"/>
</reference>
<reference anchor="RFC5670" target="https://www.rfc-editor.org/info/rfc567
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.5670.xml">
<front>
<title>Metering and Marking Behaviour of PCN-Nodes</title>
<author initials="P." surname="Eardley" fullname="P. Eardley" role="ed
itor">
<organization/>
</author>
<date year="2009" month="November"/>
<abstract>
<t>The objective of Pre-Congestion Notification (PCN) is to protect
the quality of service (QoS) of inelastic flows within a Diffserv domain in a si
mple, scalable, and robust fashion. This document defines the two metering and
marking behaviours of PCN-nodes. Threshold-metering and -marking marks all PCN-
packets if the rate of PCN-traffic is greater than a configured rate ("PCN-thres
hold-rate"). Excess- traffic-metering and -marking marks a proportion of PCN-pa
ckets, such that the amount marked equals the rate of PCN-traffic in excess of a
configured rate ("PCN-excess-rate"). The level of marking allows PCN-boundary-
nodes to make decisions about whether to admit or terminate PCN-flows. [STANDAR
DS-TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="5670"/>
<seriesInfo name="DOI" value="10.17487/RFC5670"/>
</reference>
<reference anchor="RFC5681" target="https://www.rfc-editor.org/info/rfc568
1" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.5681.xml">
<front>
<title>TCP Congestion Control</title>
<author initials="M." surname="Allman" fullname="M. Allman">
<organization/>
</author>
<author initials="V." surname="Paxson" fullname="V. Paxson">
<organization/>
</author>
<author initials="E." surname="Blanton" fullname="E. Blanton">
<organization/>
</author>
<date year="2009" month="September"/>
<abstract>
<t>This document defines TCP's four intertwined congestion control a
lgorithms: slow start, congestion avoidance, fast retransmit, and fast recovery.
In addition, the document specifies how TCP should begin transmission after a
relatively long idle period, as well as discussing various acknowledgment genera
tion methods. This document obsoletes RFC 2581. [STANDARDS-TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="5681"/>
<seriesInfo name="DOI" value="10.17487/RFC5681"/>
</reference>
<reference anchor="RFC6040" target="https://www.rfc-editor.org/info/rfc604
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.6040.xml">
<front>
<title>Tunnelling of Explicit Congestion Notification</title>
<author initials="B." surname="Briscoe" fullname="B. Briscoe">
<organization/>
</author>
<date year="2010" month="November"/>
<abstract>
<t>This document redefines how the explicit congestion notification
(ECN) field of the IP header should be constructed on entry to and exit from any
IP-in-IP tunnel. On encapsulation, it updates RFC 3168 to bring all IP-in-IP t
unnels (v4 or v6) into line with RFC 4301 IPsec ECN processing. On decapsulatio
n, it updates both RFC 3168 and RFC 4301 to add new behaviours for previously un
used combinations of inner and outer headers. The new rules ensure the ECN fiel
d is correctly propagated across a tunnel whether it is used to signal one or tw
o severity levels of congestion; whereas before, only one severity level was sup
ported. Tunnel endpoints can be updated in any order without affecting pre-exis
ting uses of the ECN field, thus ensuring backward compatibility. Nonetheless,
operators wanting to support two severity levels (e.g., for pre-congestion notif
ication -- PCN) can require compliance with this new specification. A thorough
analysis of the reasoning for these changes and the implications is included. I
n the unlikely event that the new rules do not meet a specific need, RFC 4774 gi
ves guidance on designing alternate ECN semantics, and this document extends tha
t to include tunnelling issues. [STANDARDS-TRACK]</t>
</abstract>
</front>
<seriesInfo name="RFC" value="6040"/>
<seriesInfo name="DOI" value="10.17487/RFC6040"/>
</reference>
<reference anchor="RFC6817" target="https://www.rfc-editor.org/info/rfc681
7" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.6817.xml">
<front>
<title>Low Extra Delay Background Transport (LEDBAT)</title>
<author initials="S." surname="Shalunov" fullname="S. Shalunov">
<organization/>
</author>
<author initials="G." surname="Hazel" fullname="G. Hazel">
<organization/>
</author>
<author initials="J." surname="Iyengar" fullname="J. Iyengar">
<organization/>
</author>
<author initials="M." surname="Kuehlewind" fullname="M. Kuehlewind">
<organization/>
</author>
<date year="2012" month="December"/>
<abstract>
<t>Low Extra Delay Background Transport (LEDBAT) is an experimental
delay-based congestion control algorithm that seeks to utilize the available ban
dwidth on an end-to-end path while limiting the consequent increase in queueing
delay on that path. LEDBAT uses changes in one-way delay measurements to limit
congestion that the flow itself induces in the network. LEDBAT is designed for
use by background bulk-transfer applications to be no more aggressive than stand
ard TCP congestion control (as specified in RFC 5681) and to yield in the presen
ce of competing flows, thus limiting interference with the network performance o
f competing flows. This document defines an Experimental Protocol for the Inte
rnet community.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="6817"/>
<seriesInfo name="DOI" value="10.17487/RFC6817"/>
</reference>
<reference anchor="RFC6973" target="https://www.rfc-editor.org/info/rfc697
3" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.6973.xml">
<front>
<title>Privacy Considerations for Internet Protocols</title>
<author initials="A." surname="Cooper" fullname="A. Cooper">
<organization/>
</author>
<author initials="H." surname="Tschofenig" fullname="H. Tschofenig">
<organization/>
</author>
<author initials="B." surname="Aboba" fullname="B. Aboba">
<organization/>
</author>
<author initials="J." surname="Peterson" fullname="J. Peterson">
<organization/>
</author>
<author initials="J." surname="Morris" fullname="J. Morris">
<organization/>
</author>
<author initials="M." surname="Hansen" fullname="M. Hansen">
<organization/>
</author>
<author initials="R." surname="Smith" fullname="R. Smith">
<organization/>
</author>
<date year="2013" month="July"/>
<abstract>
<t>This document offers guidance for developing privacy consideratio
ns for inclusion in protocol specifications. It aims to make designers, impleme
nters, and users of Internet protocols aware of privacy-related design choices.
It suggests that whether any individual RFC warrants a specific privacy conside
rations section will depend on the document's content.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="6973"/>
<seriesInfo name="DOI" value="10.17487/RFC6973"/>
</reference>
<reference anchor="RFC7560" target="https://www.rfc-editor.org/info/rfc756
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7560.xml">
<front>
<title>Problem Statement and Requirements for Increased Accuracy in Ex
plicit Congestion Notification (ECN) Feedback</title>
<author initials="M." surname="Kuehlewind" fullname="M. Kuehlewind" ro
le="editor">
<organization/>
</author>
<author initials="R." surname="Scheffenegger" fullname="R. Scheffenegg
er">
<organization/>
</author>
<author initials="B." surname="Briscoe" fullname="B. Briscoe">
<organization/>
</author>
<date year="2015" month="August"/>
<abstract>
<t>Explicit Congestion Notification (ECN) is a mechanism where netwo
rk nodes can mark IP packets, instead of dropping them, to indicate congestion t
o the endpoints. An ECN-capable receiver will feed this information back to the
sender. ECN is specified for TCP in such a way that it can only feed back one
congestion signal per Round-Trip Time (RTT). In contrast, ECN for other transpo
rt protocols, such as RTP/UDP and SCTP, is specified with more accurate ECN feed
back. Recent new TCP mechanisms (like Congestion Exposure (ConEx) or Data Center
TCP (DCTCP)) need more accurate ECN feedback in the case where more than one ma
rking is received in one RTT. This document specifies requirements for an updat
e to the TCP protocol to provide more accurate ECN feedback.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="7560"/>
<seriesInfo name="DOI" value="10.17487/RFC7560"/>
</reference>
<reference anchor="RFC7665" target="https://www.rfc-editor.org/info/rfc766
5" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7665.xml">
<front>
<title>Service Function Chaining (SFC) Architecture</title>
<author initials="J." surname="Halpern" fullname="J. Halpern" role="ed
itor">
<organization/>
</author>
<author initials="C." surname="Pignataro" fullname="C. Pignataro" role
="editor">
<organization/>
</author>
<date year="2015" month="October"/>
<abstract>
<t>This document describes an architecture for the specification, cr
eation, and ongoing maintenance of Service Function Chains (SFCs) in a network.
It includes architectural concepts, principles, and components used in the cons
truction of composite services through deployment of SFCs, with a focus on those
to be standardized in the IETF. This document does not propose solutions, prot
ocols, or extensions to existing protocols.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="7665"/>
<seriesInfo name="DOI" value="10.17487/RFC7665"/>
</reference>
<reference anchor="I-D.ietf-tcpm-accurate-ecn" target="https://datatracker
.ietf.org/api/v1/doc/document/draft-ietf-tcpm-accurate-ecn/" xml:base="https://b
ib.ietf.org/public/rfc/bibxml-ids/reference.I-D.ietf-tcpm-accurate-ecn.xml">
<front>
<title>More Accurate ECN Feedback in TCP</title>
<author fullname="Bob Briscoe"/>
<author fullname="Mirja Kühlewind"/>
<author fullname="Richard Scheffenegger"/>
<date day="25" month="July" year="2022"/>
<abstract>
<t>Explicit Congestion Notification (ECN) is a mechanism where netwo
rk
nodes can mark IP packets instead of dropping them to indicate
incipient congestion to the end-points. Receivers with an ECN-
capable transport protocol feed back this information to the sender.
ECN was originally specified for TCP in such a way that only one
feedback signal can be transmitted per Round-Trip Time (RTT). Recent
new TCP mechanisms like Congestion Exposure (ConEx), Data Center TCP
(DCTCP) or Low Latency Low Loss Scalable Throughput (L4S) need more
accurate ECN feedback information whenever more than one marking is
received in one RTT. This document updates the original ECN
specification to specify a scheme to provide more than one feedback
signal per RTT in the TCP header. Given TCP header space is scarce,
it allocates a reserved header bit previously assigned to the ECN-
Nonce. It also overloads the two existing ECN flags in the TCP
header. The resulting extra space is exploited to feed back the IP-
ECN field received during the 3-way handshake as well. Supplementary
feedback information can optionally be provided in a new TCP option,
which is never used on the TCP SYN. The document also specifies the
treatment of this updated TCP wire protocol by middleboxes.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-tcpm-accurate-ecn-20
"/>
</reference>
<reference anchor="RFC7713" target="https://www.rfc-editor.org/info/rfc771
3" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7713.xml">
<front>
<title>Congestion Exposure (ConEx) Concepts, Abstract Mechanism, and R
equirements</title>
<author initials="M." surname="Mathis" fullname="M. Mathis">
<organization/>
</author>
<author initials="B." surname="Briscoe" fullname="B. Briscoe">
<organization/>
</author>
<date year="2015" month="December"/>
<abstract>
<t>This document describes an abstract mechanism by which senders in
form the network about the congestion recently encountered by packets in the sam
e flow. Today, network elements at any layer may signal congestion to the recei
ver by dropping packets or by Explicit Congestion Notification (ECN) markings, a
nd the receiver passes this information back to the sender in transport-layer fe
edback. The mechanism described here enables the sender to also relay this cong
estion information back into the network in-band at the IP layer, such that the
total amount of congestion from all elements on the path is revealed to all IP e
lements along the path, where it could, for example, be used to provide input to
traffic management. This mechanism is called Congestion Exposure, or ConEx. T
he companion document, "Congestion Exposure (ConEx) Concepts and Use Cases" (RFC
6789), provides the entry point to the set of ConEx documentation.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="7713"/>
<seriesInfo name="DOI" value="10.17487/RFC7713"/>
</reference>
<reference anchor="RFC7567" target="https://www.rfc-editor.org/info/rfc756
7" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7567.xml">
<front>
<title>IETF Recommendations Regarding Active Queue Management</title>
<author initials="F." surname="Baker" fullname="F. Baker" role="editor
">
<organization/>
</author>
<author initials="G." surname="Fairhurst" fullname="G. Fairhurst" role
="editor">
<organization/>
</author>
<date year="2015" month="July"/>
<abstract>
<t>This memo presents recommendations to the Internet community conc
erning measures to improve and preserve Internet performance. It presents a str
ong recommendation for testing, standardization, and widespread deployment of ac
tive queue management (AQM) in network devices to improve the performance of tod
ay's Internet. It also urges a concerted effort of research, measurement, and u
ltimate deployment of AQM mechanisms to protect the Internet from flows that are
not sufficiently responsive to congestion notification.</t>
<t>Based on 15 years of experience and new research, this document r
eplaces the recommendations of RFC 2309.</t>
</abstract>
</front>
<seriesInfo name="BCP" value="197"/>
<seriesInfo name="RFC" value="7567"/>
<seriesInfo name="DOI" value="10.17487/RFC7567"/>
</reference>
<reference anchor="RFC8033" target="https://www.rfc-editor.org/info/rfc803
3" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8033.xml">
<front>
<title>Proportional Integral Controller Enhanced (PIE): A Lightweight
Control Scheme to Address the Bufferbloat Problem</title>
<author initials="R." surname="Pan" fullname="R. Pan">
<organization/>
</author>
<author initials="P." surname="Natarajan" fullname="P. Natarajan">
<organization/>
</author>
<author initials="F." surname="Baker" fullname="F. Baker">
<organization/>
</author>
<author initials="G." surname="White" fullname="G. White">
<organization/>
</author>
<date year="2017" month="February"/>
<abstract>
<t>Bufferbloat is a phenomenon in which excess buffers in the networ
k cause high latency and latency variation. As more and more interactive applic
ations (e.g., voice over IP, real-time video streaming, and financial transactio
ns) run in the Internet, high latency and latency variation degrade application
performance. There is a pressing need to design intelligent queue management sc
hemes that can control latency and latency variation, and hence provide desirabl
e quality of service to users.</t>
<t>This document presents a lightweight active queue management desi
gn called "PIE" (Proportional Integral controller Enhanced) that can effectively
control the average queuing latency to a target value. Simulation results, theo
retical analysis, and Linux testbed results have shown that PIE can ensure low l
atency and achieve high link utilization under various congestion situations. T
he design does not require per-packet timestamps, so it incurs very little overh
ead and is simple enough to implement in both hardware and software.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8033"/>
<seriesInfo name="DOI" value="10.17487/RFC8033"/>
</reference>
<reference anchor="RFC8034" target="https://www.rfc-editor.org/info/rfc803
4" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8034.xml">
<front>
<title>Active Queue Management (AQM) Based on Proportional Integral Co
ntroller Enhanced PIE) for Data-Over-Cable Service Interface Specifications (DOC
SIS) Cable Modems</title>
<author initials="G." surname="White" fullname="G. White">
<organization/>
</author>
<author initials="R." surname="Pan" fullname="R. Pan">
<organization/>
</author>
<date year="2017" month="February"/>
<abstract>
<t>Cable modems based on Data-Over-Cable Service Interface Specifica
tions (DOCSIS) provide broadband Internet access to over one hundred million use
rs worldwide. In some cases, the cable modem connection is the bottleneck (lowe
st speed) link between the customer and the Internet. As a result, the impact o
f buffering and bufferbloat in the cable modem can have a significant effect on
user experience. The CableLabs DOCSIS 3.1 specification introduces requirements
for cable modems to support an Active Queue Management (AQM) algorithm that is
intended to alleviate the impact that buffering has on latency-sensitive traffic
, while preserving bulk throughput performance. In addition, the CableLabs DOCS
IS 3.0 specifications have also been amended to contain similar requirements. T
his document describes the requirements on AQM that apply to DOCSIS equipment, i
ncluding a description of the "DOCSIS-PIE" algorithm that is required on DOCSIS
3.1 cable modems.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8034"/>
<seriesInfo name="DOI" value="10.17487/RFC8034"/>
</reference>
<reference anchor="RFC8170" target="https://www.rfc-editor.org/info/rfc817
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8170.xml">
<front>
<title>Planning for Protocol Adoption and Subsequent Transitions</titl
e>
<author initials="D." surname="Thaler" fullname="D. Thaler" role="edit
or">
<organization/>
</author>
<date year="2017" month="May"/>
<abstract>
<t>Over the many years since the introduction of the Internet Protoc
ol, we have seen a number of transitions throughout the protocol stack, such as
deploying a new protocol, or updating or replacing an existing protocol. Many p
rotocols and technologies were not designed to enable smooth transition to alter
natives or to easily deploy extensions; thus, some transitions, such as the intr
oduction of IPv6, have been difficult. This document attempts to summarize some
basic principles to enable future transitions, and it also summarizes what make
s for a good transition plan.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8170"/>
<seriesInfo name="DOI" value="10.17487/RFC8170"/>
</reference>
<reference anchor="RFC8257" target="https://www.rfc-editor.org/info/rfc825
7" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8257.xml">
<front>
<title>Data Center TCP (DCTCP): TCP Congestion Control for Data Center
s</title>
<author initials="S." surname="Bensley" fullname="S. Bensley">
<organization/>
</author>
<author initials="D." surname="Thaler" fullname="D. Thaler">
<organization/>
</author>
<author initials="P." surname="Balasubramanian" fullname="P. Balasubra
manian">
<organization/>
</author>
<author initials="L." surname="Eggert" fullname="L. Eggert">
<organization/>
</author>
<author initials="G." surname="Judd" fullname="G. Judd">
<organization/>
</author>
<date year="2017" month="October"/>
<abstract>
<t>This Informational RFC describes Data Center TCP (DCTCP): a TCP c
ongestion control scheme for data-center traffic. DCTCP extends the Explicit Co
ngestion Notification (ECN) processing to estimate the fraction of bytes that en
counter congestion rather than simply detecting that some congestion has occurre
d. DCTCP then scales the TCP congestion window based on this estimate. This me
thod achieves high-burst tolerance, low latency, and high throughput with shallo
w- buffered switches. This memo also discusses deployment issues related to the
coexistence of DCTCP and conventional TCP, discusses the lack of a negotiating
mechanism between sender and receiver, and presents some possible mitigations.
This memo documents DCTCP as currently implemented by several major operating sy
stems. DCTCP, as described in this specification, is applicable to deployments
in controlled environments like data centers, but it must not be deployed over t
he public Internet without additional measures.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8257"/>
<seriesInfo name="DOI" value="10.17487/RFC8257"/>
</reference>
<reference anchor="RFC8290" target="https://www.rfc-editor.org/info/rfc829
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8290.xml">
<front>
<title>The Flow Queue CoDel Packet Scheduler and Active Queue Manageme
nt Algorithm</title>
<author initials="T." surname="Hoeiland-Joergensen" fullname="T. Hoeil
and-Joergensen">
<organization/>
</author>
<author initials="P." surname="McKenney" fullname="P. McKenney">
<organization/>
</author>
<author initials="D." surname="Taht" fullname="D. Taht">
<organization/>
</author>
<author initials="J." surname="Gettys" fullname="J. Gettys">
<organization/>
</author>
<author initials="E." surname="Dumazet" fullname="E. Dumazet">
<organization/>
</author>
<date year="2018" month="January"/>
<abstract>
<t>This memo presents the FQ-CoDel hybrid packet scheduler and Activ
e Queue Management (AQM) algorithm, a powerful tool for fighting bufferbloat and
reducing latency.</t>
<t>FQ-CoDel mixes packets from multiple flows and reduces the impact
of head-of-line blocking from bursty traffic. It provides isolation for low-ra
te traffic such as DNS, web, and videoconferencing traffic. It improves utilisa
tion across the networking fabric, especially for bidirectional traffic, by keep
ing queue lengths short, and it can be implemented in a memory- and CPU-efficien
t fashion across a wide range of hardware.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8290"/>
<seriesInfo name="DOI" value="10.17487/RFC8290"/>
</reference>
<reference anchor="RFC8298" target="https://www.rfc-editor.org/info/rfc829
8" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8298.xml">
<front>
<title>Self-Clocked Rate Adaptation for Multimedia</title>
<author initials="I." surname="Johansson" fullname="I. Johansson">
<organization/>
</author>
<author initials="Z." surname="Sarker" fullname="Z. Sarker">
<organization/>
</author>
<date year="2017" month="December"/>
<abstract>
<t>This memo describes a rate adaptation algorithm for conversationa
l media services such as interactive video. The solution conforms to the packet
conservation principle and uses a hybrid loss-and-delay- based congestion contr
ol algorithm. The algorithm is evaluated over both simulated Internet bottlenec
k scenarios as well as in a Long Term Evolution (LTE) system simulator and is sh
own to achieve both low latency and high video throughput in these scenarios.</t
>
</abstract>
</front>
<seriesInfo name="RFC" value="8298"/>
<seriesInfo name="DOI" value="10.17487/RFC8298"/>
</reference>
<reference anchor="RFC8311" target="https://www.rfc-editor.org/info/rfc831
1" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8311.xml">
<front>
<title>Relaxing Restrictions on Explicit Congestion Notification (ECN)
Experimentation</title>
<author initials="D." surname="Black" fullname="D. Black">
<organization/>
</author>
<date year="2018" month="January"/>
<abstract>
<t>This memo updates RFC 3168, which specifies Explicit Congestion N
otification (ECN) as an alternative to packet drops for indicating network conge
stion to endpoints. It relaxes restrictions in RFC 3168 that hinder experimenta
tion towards benefits beyond just removal of loss. This memo summarizes the ant
icipated areas of experimentation and updates RFC 3168 to enable experimentation
in these areas. An Experimental RFC in the IETF document stream is required to
take advantage of any of these enabling updates. In addition, this memo makes
related updates to the ECN specifications for RTP in RFC 6679 and for the Datagr
am Congestion Control Protocol (DCCP) in RFCs 4341, 4342, and 5622. This memo a
lso records the conclusion of the ECN nonce experiment in RFC 3540 and provides
the rationale for reclassification of RFC 3540 from Experimental to Historic; th
is reclassification enables new experimental use of the ECT(1) codepoint.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8311"/>
<seriesInfo name="DOI" value="10.17487/RFC8311"/>
</reference>
<reference anchor="RFC8312" target="https://www.rfc-editor.org/info/rfc831
2" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8312.xml">
<front>
<title>CUBIC for Fast Long-Distance Networks</title>
<author initials="I." surname="Rhee" fullname="I. Rhee">
<organization/>
</author>
<author initials="L." surname="Xu" fullname="L. Xu">
<organization/>
</author>
<author initials="S." surname="Ha" fullname="S. Ha">
<organization/>
</author>
<author initials="A." surname="Zimmermann" fullname="A. Zimmermann">
<organization/>
</author>
<author initials="L." surname="Eggert" fullname="L. Eggert">
<organization/>
</author>
<author initials="R." surname="Scheffenegger" fullname="R. Scheffenegg
er">
<organization/>
</author>
<date year="2018" month="February"/>
<abstract>
<t>CUBIC is an extension to the current TCP standards. It differs f
rom the current TCP standards only in the congestion control algorithm on the se
nder side. In particular, it uses a cubic function instead of a linear window i
ncrease function of the current TCP standards to improve scalability and stabili
ty under fast and long-distance networks. CUBIC and its predecessor algorithm h
ave been adopted as defaults by Linux and have been used for many years. This d
ocument provides a specification of CUBIC to enable third-party implementations
and to solicit community feedback through experimentation on the performance of
CUBIC.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8312"/>
<seriesInfo name="DOI" value="10.17487/RFC8312"/>
</reference>
<reference anchor="RFC8404" target="https://www.rfc-editor.org/info/rfc840
4" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8404.xml">
<front>
<title>Effects of Pervasive Encryption on Operators</title>
<author initials="K." surname="Moriarty" fullname="K. Moriarty" role="
editor">
<organization/>
</author>
<author initials="A." surname="Morton" fullname="A. Morton" role="edit
or">
<organization/>
</author>
<date year="2018" month="July"/>
<abstract>
<t>Pervasive monitoring attacks on the privacy of Internet users are
of serious concern to both user and operator communities. RFC 7258 discusses t
he critical need to protect users' privacy when developing IETF specifications a
nd also recognizes that making networks unmanageable to mitigate pervasive monit
oring is not an acceptable outcome: an appropriate balance is needed. This docu
ment discusses current security and network operations as well as management pra
ctices that may be impacted by the shift to increased use of encryption to help
guide protocol development in support of manageable and secure networks.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8404"/>
<seriesInfo name="DOI" value="10.17487/RFC8404"/>
</reference>
<reference anchor="RFC8511" target="https://www.rfc-editor.org/info/rfc851
1" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8511.xml">
<front>
<title>TCP Alternative Backoff with ECN (ABE)</title>
<author initials="N." surname="Khademi" fullname="N. Khademi">
<organization/>
</author>
<author initials="M." surname="Welzl" fullname="M. Welzl">
<organization/>
</author>
<author initials="G." surname="Armitage" fullname="G. Armitage">
<organization/>
</author>
<author initials="G." surname="Fairhurst" fullname="G. Fairhurst">
<organization/>
</author>
<date year="2018" month="December"/>
<abstract>
<t>Active Queue Management (AQM) mechanisms allow for burst toleranc
e while enforcing short queues to minimise the time that packets spend enqueued
at a bottleneck. This can cause noticeable performance degradation for TCP conn
ections traversing such a bottleneck, especially if there are only a few flows o
r their bandwidth-delay product (BDP) is large. The reception of a Congestion E
xperienced (CE) Explicit Congestion Notification (ECN) mark indicates that an AQ
M mechanism is used at the bottleneck, and the bottleneck network queue is there
fore likely to be short. Feedback of this signal allows the TCP sender-side ECN
reaction in congestion avoidance to reduce the Congestion Window (cwnd) by a sm
aller amount than the congestion control algorithm's reaction to inferred packet
loss. Therefore, this specification defines an experimental change to the TCP r
eaction specified in RFC 3168, as permitted by RFC 8311.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8511"/>
<seriesInfo name="DOI" value="10.17487/RFC8511"/>
</reference>
<reference anchor="RFC8888" target="https://www.rfc-editor.org/info/rfc888
8" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8888.xml">
<front>
<title>RTP Control Protocol (RTCP) Feedback for Congestion Control</ti
tle>
<author initials="Z." surname="Sarker" fullname="Z. Sarker">
<organization/>
</author>
<author initials="C." surname="Perkins" fullname="C. Perkins">
<organization/>
</author>
<author initials="V." surname="Singh" fullname="V. Singh">
<organization/>
</author>
<author initials="M." surname="Ramalho" fullname="M. Ramalho">
<organization/>
</author>
<date year="2021" month="January"/>
<abstract>
<t>An effective RTP congestion control algorithm requires more fine-
grained feedback on packet loss, timing, and Explicit Congestion Notification (E
CN) marks than is provided by the standard RTP Control Protocol (RTCP) Sender Re
port (SR) and Receiver Report (RR) packets. This document describes an RTCP feed
back message intended to enable congestion control for interactive real-time tra
ffic using RTP. The feedback message is designed for use with a sender-based con
gestion control algorithm, in which the receiver of an RTP flow sends back to th
e sender RTCP feedback packets containing the information the sender needs to pe
rform congestion control.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="8888"/>
<seriesInfo name="DOI" value="10.17487/RFC8888"/>
</reference>
<reference anchor="RFC9000" target="https://www.rfc-editor.org/info/rfc900
0" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.9000.xml">
<front>
<title>QUIC: A UDP-Based Multiplexed and Secure Transport</title>
<author initials="J." surname="Iyengar" fullname="J. Iyengar" role="ed
itor">
<organization/>
</author>
<author initials="M." surname="Thomson" fullname="M. Thomson" role="ed
itor">
<organization/>
</author>
<date year="2021" month="May"/>
<abstract>
<t>This document defines the core of the QUIC transport protocol. Q
UIC provides applications with flow-controlled streams for structured communicat
ion, low-latency connection establishment, and network path migration. QUIC incl
udes security measures that ensure confidentiality, integrity, and availability
in a range of deployment circumstances. Accompanying documents describe the int
egration of TLS for key negotiation, loss detection, and an exemplary congestion
control algorithm.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="9000"/>
<seriesInfo name="DOI" value="10.17487/RFC9000"/>
</reference>
<reference anchor="RFC9113" target="https://www.rfc-editor.org/info/rfc911
3" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.9113.xml">
<front>
<title>HTTP/2</title>
<author initials="M." surname="Thomson" fullname="M. Thomson" role="ed
itor">
<organization/>
</author>
<author initials="C." surname="Benfield" fullname="C. Benfield" role="
editor">
<organization/>
</author>
<date year="2022" month="June"/>
<abstract>
<t>This specification describes an optimized expression of the seman
tics of the Hypertext Transfer Protocol (HTTP), referred to as HTTP version 2 (H
TTP/2). HTTP/2 enables a more efficient use of network resources and a reduced l
atency by introducing field compression and allowing multiple concurrent exchang
es on the same connection.</t>
<t>This document obsoletes RFCs 7540 and 8740.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="9113"/>
<seriesInfo name="DOI" value="10.17487/RFC9113"/>
</reference>
<reference anchor="I-D.ietf-tsvwg-aqm-dualq-coupled" target="https://datat
racker.ietf.org/api/v1/doc/document/draft-ietf-tsvwg-aqm-dualq-coupled/" xml:bas
e="https://bib.ietf.org/public/rfc/bibxml-ids/reference.I-D.ietf-tsvwg-aqm-dualq
-coupled.xml">
<front>
<title>DualQ Coupled AQMs for Low Latency, Low Loss and Scalable Throu
ghput (L4S)</title>
<author fullname="Koen De Schepper"/>
<author fullname="Bob Briscoe"/>
<author fullname="Greg White"/>
<date day="7" month="July" year="2022"/>
<abstract>
<t>This specification defines a framework for coupling the Active Qu
eue
Management (AQM) algorithms in two queues intended for flows with
different responses to congestion. This provides a way for the
Internet to transition from the scaling problems of standard TCP
Reno-friendly ('Classic') congestion controls to the family of
'Scalable' congestion controls. These are designed for consistently
very Low queuing Latency, very Low congestion Loss and Scaling of
per-flow throughput (L4S) by using Explicit Congestion Notification
(ECN) in a modified way. Until the Coupled DualQ, these L4S senders
could only be deployed where a clean-slate environment could be
arranged, such as in private data centres. The coupling acts like a
semi-permeable membrane: isolating the sub-millisecond average
queuing delay and zero congestion loss of L4S from Classic latency
and loss; but pooling the capacity between any combination of
Scalable and Classic flows with roughly equivalent throughput per
flow. The DualQ achieves this indirectly, without having to inspect
transport layer flow identifiers and without compromising the
performance of the Classic traffic, relative to a single queue. The
DualQ design has low complexity and requires no configuration for the
public Internet.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-tsvwg-aqm-dualq-coup
led-24"/>
</reference>
<reference anchor="I-D.ietf-tsvwg-ecn-l4s-id" target="https://datatracker.
ietf.org/api/v1/doc/document/draft-ietf-tsvwg-ecn-l4s-id/" xml:base="https://bib
.ietf.org/public/rfc/bibxml-ids/reference.I-D.ietf-tsvwg-ecn-l4s-id.xml">
<front>
<title>Explicit Congestion Notification (ECN) Protocol for Very Low Qu
euing Delay (L4S)</title>
<author fullname="Koen De Schepper"/>
<author fullname="Bob Briscoe"/>
<date day="8" month="August" year="2022"/>
<abstract>
<t>This specification defines the protocol to be used for a new netw
ork
service called low latency, low loss and scalable throughput (L4S).
L4S uses an Explicit Congestion Notification (ECN) scheme at the IP
layer that is similar to the original (or 'Classic') ECN approach,
except as specified within. L4S uses 'scalable' congestion control,
which induces much more frequent control signals from the network and
it responds to them with much more fine-grained adjustments, so that
very low (typically sub-millisecond on average) and consistently low
queuing delay becomes possible for L4S traffic without compromising
link utilization. Thus even capacity-seeking (TCP-like) traffic can
have high bandwidth and very low delay at the same time, even during
periods of high traffic load.</t>
<t>The L4S identifier defined in this document distinguishes L4S fro
m
'Classic' (e.g. TCP-Reno-friendly) traffic. Then, network
bottlenecks can be incrementally modified to distinguish and isolate
existing traffic that still follows the Classic behaviour, to prevent
it degrading the low queuing delay and low loss of L4S traffic. This
experimental track specification defines the rules that L4S
transports and network elements need to follow, with the intention
that L4S flows neither harm each other's performance nor that of
Classic traffic. It also suggests open questions to be investigated
during experimentation. Examples of new active queue management
(AQM) marking algorithms and examples of new transports (whether TCP-
like or real-time) are specified separately.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-tsvwg-ecn-l4s-id-28"
/>
</reference>
<reference anchor="I-D.ietf-tsvwg-nqb" target="https://datatracker.ietf.or
g/api/v1/doc/document/draft-ietf-tsvwg-nqb/" xml:base="https://bib.ietf.org/publ
ic/rfc/bibxml-ids/reference.I-D.ietf-tsvwg-nqb.xml">
<front>
<title>A Non-Queue-Building Per-Hop Behavior (NQB PHB) for Differentia
ted Services</title>
<author fullname="Greg White"/>
<author fullname="Thomas Fossati"/>
<date day="4" month="March" year="2022"/>
<abstract>
<t>This document specifies properties and characteristics of a Non-
Queue-Building Per-Hop Behavior (NQB PHB). The purpose of this NQB
PHB is to provide a separate queue that enables smooth, low-data-
rate, application-limited traffic flows, which would ordinarily share
a queue with bursty and capacity-seeking traffic, to avoid the
latency, latency variation and loss caused by such traffic. This PHB
is implemented without prioritization and without rate policing,
making it suitable for environments where the use of either these
features may be restricted. The NQB PHB has been developed primarily
for use by access network segments, where queuing delays and queuing
loss caused by Queue-Building protocols are manifested, but its use
is not limited to such segments. In particular, applications to
cable broadband links, Wi-Fi links, and mobile network radio and core
segments are discussed. This document recommends a specific
Differentiated Services Code Point (DSCP) to identify Non-Queue-
Building flows.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-tsvwg-nqb-10"/>
</reference>
<reference anchor="I-D.briscoe-conex-policing" target="https://www.ietf.or
g/archive/id/draft-briscoe-conex-policing-01.txt" xml:base="https://bib.ietf.org
/public/rfc/bibxml-ids/reference.I-D.briscoe-conex-policing.xml">
<front>
<title>Network Performance Isolation using Congestion Policing</title>
<author fullname="Bob Briscoe"/>
<date day="14" month="February" year="2014"/>
<abstract>
<t>This document describes why policing using congestion information
can isolate users from network performance degradation due to each other's usag
e, but without losing the multiplexing benefits of a LAN- style network where an
yone can use any amount of any resource. Extensive numerical examples and diagra
ms are given. The document is agnostic to how the congestion information reaches
the policer. The congestion exposure (ConEX) protocol is recommended, but other
tunnel feedback mechanisms have been proposed.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-briscoe-conex-policing-01
"/>
</reference>
<reference anchor="I-D.stewart-tsvwg-sctpecn" target="https://www.ietf.org
/archive/id/draft-stewart-tsvwg-sctpecn-05.txt" xml:base="https://bib.ietf.org/p
ublic/rfc/bibxml-ids/reference.I-D.stewart-tsvwg-sctpecn.xml">
<front>
<title>ECN for Stream Control Transmission Protocol (SCTP)</title>
<author fullname="Randall R. Stewart"/>
<author fullname="Michael Tuexen"/>
<author fullname="Xuesong Dong"/>
<date day="15" month="January" year="2014"/>
<abstract>
<t>This document describes the addition of the ECN to the Stream Con
trol Transmission Protocol (SCTP).</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-stewart-tsvwg-sctpecn-05"
/>
</reference>
<reference anchor="I-D.sridharan-tcpm-ctcp" target="https://datatracker.ie
tf.org/api/v1/doc/document/draft-sridharan-tcpm-ctcp/" xml:base="https://bib.iet
f.org/public/rfc/bibxml-ids/reference.I-D.sridharan-tcpm-ctcp.xml">
<front>
<title>Compound TCP: A New TCP Congestion Control for High-Speed and L
ong Distance Networks</title>
<author fullname="Murali Sridharan"/>
<author fullname="Kun Tan"/>
<author fullname="Deepak Bansal"/>
<author fullname="Dave Thaler"/>
<date day="29" month="October" year="2007"/>
<abstract>
<t>Compound TCP (CTCP) is a modification to TCP's congestion control
&#13;
mechanism for use with TCP connections with large congestion windows. &#13;
This document describes the Compound TCP algorithm in detail, and &#13;
solicits experimentation and feedback from the wider community. The &#13;
key idea behind CTCP is to add a scalable delay-based component to the &#13;
standard TCP's loss-based congestion control. The sending rate of CTCP &#13;
is controlled by both loss and delay components. The delay-based &#13;
component has a scalable window increasing rule that not only &#13;
efficiently uses the link capacity, but on sensing queue build up, &#13;
proactively reduces the sending rate.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-sridharan-tcpm-ctcp-02"/>
</reference>
<reference anchor="I-D.ietf-tsvwg-rfc6040update-shim" target="https://data
tracker.ietf.org/api/v1/doc/document/draft-ietf-tsvwg-rfc6040update-shim/" xml:b
ase="https://bib.ietf.org/public/rfc/bibxml-ids/reference.I-D.ietf-tsvwg-rfc6040
update-shim.xml">
<front>
<title>Propagating Explicit Congestion Notification Across IP Tunnel H
eaders Separated by a Shim</title>
<author fullname="Bob Briscoe"/>
<date day="11" month="July" year="2022"/>
<abstract>
<t>RFC 6040 on "Tunnelling of Explicit Congestion Notification" made
the
rules for propagation of ECN consistent for all forms of IP in IP
tunnel. This specification updates RFC 6040 to clarify that its
scope includes tunnels where two IP headers are separated by at least
one shim header that is not sufficient on its own for wide area
packet forwarding. It surveys widely deployed IP tunnelling
protocols that use such shim header(s) and updates the specifications
of those that do not mention ECN propagation (L2TPv2, L2TPv3, GRE,
Teredo and AMT). This specification also updates RFC 6040 with
configuration requirements needed to make any legacy tunnel ingress
safe.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-tsvwg-rfc6040update-
shim-15"/>
</reference>
<reference anchor="I-D.ietf-tsvwg-ecn-encap-guidelines" target="https://da
tatracker.ietf.org/api/v1/doc/document/draft-ietf-tsvwg-ecn-encap-guidelines/" x
ml:base="https://bib.ietf.org/public/rfc/bibxml-ids/reference.I-D.ietf-tsvwg-ecn
-encap-guidelines.xml">
<front>
<title>Guidelines for Adding Congestion Notification to Protocols that
Encapsulate IP</title>
<author fullname="Bob Briscoe"/>
<author fullname="John Kaippallimalil"/>
<date day="11" month="July" year="2022"/>
<abstract>
<t>The purpose of this document is to guide the design of congestion
notification in any lower layer or tunnelling protocol that
encapsulates IP. The aim is for explicit congestion signals to
propagate consistently from lower layer protocols into IP. Then the
IP internetwork layer can act as a portability layer to carry
congestion notification from non-IP-aware congested nodes up to the
transport layer (L4). Following these guidelines should assure
interworking among IP layer and lower layer congestion notification
mechanisms, whether specified by the IETF or other standards bodies.
This document updates the advice to subnetwork designers about ECN in
RFC 3819.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-tsvwg-ecn-encap-guid
elines-17"/>
</reference>
<reference anchor="I-D.ietf-tsvwg-l4sops" target="https://datatracker.ietf
.org/api/v1/doc/document/draft-ietf-tsvwg-l4sops/" xml:base="https://bib.ietf.or
g/public/rfc/bibxml-ids/reference.I-D.ietf-tsvwg-l4sops.xml">
<front>
<title>Operational Guidance for Deployment of L4S in the Internet</tit
le>
<author fullname="Greg White"/>
<date day="28" month="April" year="2022"/>
<abstract>
<t>This document is intended to provide guidance in order to ensure
successful deployment of Low Latency Low Loss Scalable throughput
(L4S) in the Internet. Other L4S documents provide guidance for
running an L4S experiment, but this document is focused solely on
potential interactions between L4S flows and flows using the original
('Classic') ECN over a Classic ECN bottleneck link. The document
discusses the potential outcomes of these interactions, describes
mechanisms to detect the presence of Classic ECN bottlenecks, and
identifies opportunities to prevent and/or detect and resolve
fairness problems in such networks. This guidance is aimed at
operators of end-systems, operators of networks, and researchers.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-tsvwg-l4sops-03"/>
</reference>
<reference anchor="I-D.briscoe-tsvwg-l4s-diffserv" target="https://datatra
cker.ietf.org/api/v1/doc/document/draft-briscoe-tsvwg-l4s-diffserv/" xml:base="h
ttps://bib.ietf.org/public/rfc/bibxml-ids/reference.I-D.briscoe-tsvwg-l4s-diffse
rv.xml">
<front>
<title>Interactions between Low Latency, Low Loss, Scalable Throughput
(L4S) and Differentiated Services</title>
<author fullname="Bob Briscoe"/>
<date day="2" month="July" year="2018"/>
<abstract>
<t>L4S and Diffserv offer somewhat overlapping services (low latency
and
low loss), but bandwidth allocation is out of scope for L4S.
Therefore there is scope for the two approaches to complement each
other, but also to conflict. This informational document explains
how the two approaches interact, how they can be arranged to
complement each other and in which cases one can stand alone without
needing the other.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-briscoe-tsvwg-l4s-diffser
v-02"/>
</reference>
<reference anchor="I-D.briscoe-docsis-q-protection" target="https://datatr
acker.ietf.org/api/v1/doc/document/draft-briscoe-docsis-q-protection/" xml:base=
"https://bib.ietf.org/public/rfc/bibxml-ids/reference.I-D.briscoe-docsis-q-prote
ction.xml">
<front>
<title>The DOCSIS(r) Queue Protection Algorithm to Preserve Low Latenc
y</title>
<author fullname="Bob Briscoe"/>
<author fullname="Greg White"/>
<date day="13" month="May" year="2022"/>
<abstract>
<t>This informational document explains the specification of the que
ue
protection algorithm used in DOCSIS technology since version 3.1. A
shared low latency queue relies on the non-queue-building behaviour
of every traffic flow using it. However, some flows might not take
such care, either accidentally or maliciously. If a queue is about
to exceed a threshold level of delay, the queue protection algorithm
can rapidly detect the flows most likely to be responsible. It can
then prevent harm to other traffic in the low latency queue by
ejecting selected packets (or all packets) of these flows. The
document is designed for four types of audience: a) congestion
control designers who need to understand how to keep on the 'good'
side of the algorithm; b) implementers of the algorithm who want to
understand it in more depth; c) designers of algorithms with similar
goals, perhaps for non-DOCSIS scenarios; and d) researchers
interested in evaluating the algorithm.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-briscoe-docsis-q-protecti
on-06"/>
</reference>
<reference anchor="I-D.cardwell-iccrg-bbr-congestion-control" target="http
s://datatracker.ietf.org/api/v1/doc/document/draft-cardwell-iccrg-bbr-congestion
-control/" xml:base="https://bib.ietf.org/public/rfc/bibxml-ids/reference.I-D.ca
rdwell-iccrg-bbr-congestion-control.xml">
<front>
<title>BBR Congestion Control</title>
<author fullname="Neal Cardwell"/>
<author fullname="Yuchung Cheng"/>
<author fullname="Soheil Hassas Yeganeh"/>
<author fullname="Ian Swett"/>
<author fullname="Van Jacobson"/>
<date day="7" month="March" year="2022"/>
<abstract>
<t>This document specifies the BBR congestion control algorithm. BB
R
("Bottleneck Bandwidth and Round-trip propagation time") uses recent
measurements of a transport connection's delivery rate, round-trip
time, and packet loss rate to build an explicit model of the network
path. BBR then uses this model to control both how fast it sends
data and the maximum volume of data it allows in flight in the
network at any time. Relative to loss-based congestion control
algorithms such as Reno [RFC5681] or CUBIC [RFC8312], BBR offers
substantially higher throughput for bottlenecks with shallow buffers
or random losses, and substantially lower queueing delays for
bottlenecks with deep buffers (avoiding "bufferbloat"). BBR can be
implemented in any transport protocol that supports packet-delivery
acknowledgment. Thus far, open source implementations are available
for TCP [RFC793] and QUIC [RFC9000]. This document specifies version
2 of the BBR algorithm, also sometimes referred to as BBRv2 or bbr2.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-cardwell-iccrg-bbr-conges
tion-control-02"/>
</reference>
<reference anchor="I-D.briscoe-iccrg-prague-congestion-control" target="ht
tps://datatracker.ietf.org/api/v1/doc/document/draft-briscoe-iccrg-prague-conges
tion-control/" xml:base="https://bib.ietf.org/public/rfc/bibxml-ids/reference.I-
D.briscoe-iccrg-prague-congestion-control.xml">
<front>
<title>Prague Congestion Control</title>
<author fullname="Koen De Schepper"/>
<author fullname="Olivier Tilmans"/>
<author fullname="Bob Briscoe"/>
<date day="11" month="July" year="2022"/>
<abstract>
<t>This specification defines the Prague congestion control scheme,
which is derived from DCTCP and adapted for Internet traffic by
implementing the Prague L4S requirements. Over paths with L4S
support at the bottleneck, it adapts the DCTCP mechanisms to achieve
consistently low latency and full throughput. It is defined
independently of any particular transport protocol or operating
system, but notes are added that highlight issues specific to certain
transports and OSs. It is mainly based on the current default
options of the reference Linux implementation of TCP Prague, but it
includes experience from other implementations where available. It
separately describes non-default and optional parts, as well as
future plans.</t>
<t>The implementation does not satisfy all the Prague requirements (
yet)
and the IETF might decide that certain requirements need to be
relaxed as an outcome of the process of trying to satisfy them all.
In two cases, research code is replaced by placeholders until full
evaluation is complete.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-briscoe-iccrg-prague-cong
estion-control-01"/>
</reference>
<reference anchor="I-D.morton-tsvwg-codel-approx-fair" target="https://www
.ietf.org/archive/id/draft-morton-tsvwg-codel-approx-fair-01.txt" xml:base="http
s://bib.ietf.org/public/rfc/bibxml-ids/reference.I-D.morton-tsvwg-codel-approx-f
air.xml">
<front>
<title>Controlled Delay Approximate Fairness AQM</title>
<author fullname="Jonathan Morton"/>
<author fullname="Peter G. Heist"/>
<date day="9" month="March" year="2020"/>
<abstract>
<t>This note presents CodelAF, or Controlled Delay Approximate Fairn
ess in full, as an alternative to single-queue AQM or Fair Queue implementations
in the low-cost or high-speed network hardware spaces. It builds on the seminal
work in Codel [RFC8289], and guides multiple competing flows towards similar th
roughputs by differential congestion signalling, whilst requiring only a single
FIFO queue. It may also be combined with CNQ [I-D.morton-tsvwg-cheap-nasty-queue
ing] to provide a latency optimisation for sparse flows.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-morton-tsvwg-codel-approx
-fair-01"/>
</reference>
<reference anchor="Hohlfeld14" target="https://doi.acm.org/10.1145/2663716 .2663730"> <reference anchor="Hohlfeld14" target="https://doi.acm.org/10.1145/2663716 .2663730">
<front> <front>
<title>A QoE Perspective on Sizing Network Buffers</title> <title>A QoE Perspective on Sizing Network Buffers</title>
<author fullname="Oliver Hohlfeld" initials="O." surname="Hohlfeld "> <author fullname="Oliver Hohlfeld" initials="O." surname="Hohlfeld ">
<organization/> <organization/>
</author> </author>
<author fullname="Enric Pujol" initials="E." surname="Pujol"> <author fullname="Enric Pujol" initials="E." surname="Pujol">
<organization/> <organization/>
<address> </author>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author>
<author fullname="Florin Ciucu" initials="F." surname="Ciucu"> <author fullname="Florin Ciucu" initials="F." surname="Ciucu">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Anja Feldmann" initials="A." surname="Feldmann"> <author fullname="Anja Feldmann" initials="A." surname="Feldmann">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Paul Barford" initials="P." surname="Barford"> <author fullname="Paul Barford" initials="P." surname="Barford">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<date month="November" year="2014"/> <date month="November" year="2014"/>
</front> </front>
<seriesInfo name="Proc. ACM Internet Measurement Conf (IMC'14)" value="h <seriesInfo name="DOI" value="10.1145/2663716.2663730"/>
mm"/> <refcontent>IMC '14: Proceedings of the 2014 Conference on Internet Measu
</reference> rement, pp. 333-346</refcontent>
<reference anchor="Mathis09" target="https://www.gdt.id.au/~gdt/presentati
ons/2010-07-06-questnet-tcp/reference-materials/papers/mathis-relentless-congest
ion-control.pdf">
<front>
<title>Relentless Congestion Control</title>
<author fullname="Matt Mathis" initials="M." surname="Mathis">
<organization>PSC</organization>
</author>
<date month="May" year="2009"/>
</front>
<seriesInfo name="PFLDNeT'09" value=""/>
</reference> </reference>
<!--{ToDo: DCttH ref will need to be updated, once stable}-->
<reference anchor="DCttH19" target="https://bobbriscoe.net/pubs.html#DCttH <xi:include
_TR"> href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.mathis-iccrg-relent
less-tcp.xml"/>
<reference anchor="L4Seval22" target="https://arxiv.org/abs/2209.01078">
<front> <front>
<title>`Data Centre to the Home': Ultra-Low Latency for All</title> <title>Dual Queue Coupled AQM: Deployable Very Low Queuing Delay for
<author fullname="Koen De Schepper" initials="K." surname="De Schepper All</title>
"> <author fullname="Koen De Schepper" initials="K."
surname="De Schepper">
<organization>Nokia Bell Labs</organization> <organization>Nokia Bell Labs</organization>
</author> </author>
<author fullname="Olga Bondarenko" initials="O." surname="Bondarenko"> <author fullname="Olga Albisser" initials="O." surname="Albisser">
<organization>Simula Research Lab</organization> <organization>Simula Research Lab</organization>
</author> </author>
<author fullname="Olivier" initials="O." surname="Tilmans"> <author fullname="Olivier" initials="O." surname="Tilmans">
<organization>Nokia Bell Labs</organization> <organization>Nokia Bell Labs</organization>
</author> </author>
<author fullname="Bob Briscoe" initials="B." surname="Briscoe"> <author fullname="Bob Briscoe" initials="B." surname="Briscoe">
<organization>Independent (bobbriscoe.net)</organization> <organization>Independent (bobbriscoe.net)</organization>
</author> </author>
<date month="July" year="2019"/> <date month="September" year="2022"/>
</front> </front>
<seriesInfo name="Updated RITE project Technical Report" value=""/> <seriesInfo name="DOI" value="10.48550/arXiv.2209.01078"/>
<format target="https://bobbriscoe.net/projects/latency/dctth_journal_dr <refcontent>TR-BB-2022-001, arXiv:2209.01078 [cs.NI]</refcontent>
aft20190726.pdf" type="PDF"/> <format target="https://arxiv.org/pdf/2209.01078" type="PDF"/>
</reference> </reference>
<reference anchor="L4Sdemo16" target="https://dl.acm.org/citation.cfm?doid
=2910017.2910633 (videos of demos: https://riteproject.eu/dctth/#1511dispatchwg <reference anchor="L4Sdemo16" target="https://dl.acm.org/citation.cfm?doid
)"> =2910017.2910633">
<front> <front>
<title>Ultra-Low Delay for All: Live Experience, Live <title>Ultra-Low Delay for All: Live Experience, Live
Analysis</title> Analysis</title>
<author fullname="Olga Bondarenko" initials="O." surname="Bondarenko"> <author fullname="Olga Bondarenko" initials="O." surname="Bondarenko">
<organization>Simula Research Lab</organization> <organization>Simula Research Lab</organization>
</author> </author>
<author fullname="Koen De Schepper" initials="K." surname="De Schepper "> <author fullname="Koen De Schepper" initials="K." surname="De Schepper ">
<organization>Bell Labs</organization> <organization>Bell Labs</organization>
</author> </author>
<author fullname="Ing-jyh Tsang" initials="I." surname="Tsang"> <author fullname="Ing-jyh Tsang" initials="I." surname="Tsang">
<organization>Bell Labs</organization> <organization>Bell Labs</organization>
</author> </author>
<author fullname="Bob Briscoe" initials="B." surname="Briscoe"> <author fullname="Bob Briscoe" initials="B." surname="Briscoe">
<organization>BT</organization> <organization>BT</organization>
</author> </author>
<author fullname="Andreas Petlund" initials="A." surname="Petlund">
<organization></organization>
</author>
<author fullname="Carstan Griwodz" initials="C." surname="Griwodz">
<organization></organization>
</author>
<date month="May" year="2016"/> <date month="May" year="2016"/>
</front> </front>
<seriesInfo name="Proc. MMSYS'16" value="pp33:1--33:4"/> <seriesInfo name="DOI" value="10.1145/2910017.2910633"/>
<format target="https://dl.acm.org/citation.cfm?doid=2910017.2910633" ty <refcontent>Proceedings of the 7th International Conference on Multimedia
pe="PDF"/> Systems, Article No. 33, pp. 1-4</refcontent>
</reference> </reference>
<reference anchor="L4Sdemo16-Video" target="https://riteproject.eu/dctth/#1511di
spatchwg">
<front>
<title>Videos used in IETF dispatch WG 'Ultra-Low Queuing Delay for Al
l Apps' slot</title>
<author>
</author>
</front>
</reference>
<reference anchor="TCP-CA" target="https://ee.lbl.gov/papers/congavoid.pdf "> <reference anchor="TCP-CA" target="https://ee.lbl.gov/papers/congavoid.pdf ">
<front> <front>
<title>Congestion Avoidance and Control</title> <title>Congestion Avoidance and Control</title>
<author fullname="Van Jacobson" initials="V." surname="Jacobson"> <author fullname="Van Jacobson" initials="V." surname="Jacobson">
<organization/> <organization/>
</author> </author>
<author fullname="Michael J. Karels" initials="M.J." surname="Karels"> <author fullname="Michael J. Karels" initials="M."
surname="Karels">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<date month="November" year="1988"/> <date month="November" year="1988"/>
</front> </front>
<seriesInfo name="Laurence Berkeley Labs Technical Report" value=""/> <seriesInfo name="Laurence Berkeley Labs Technical Report" value=""/>
<format target="https://ee.lbl.gov/papers/congavoid.pdf" type="PDF"/>
</reference> </reference>
<reference anchor="UnorderedLTE"> <reference anchor="UnorderedLTE">
<front> <front>
<title>Implementing immediate forwarding for 4G in a network <title>Implementing immediate forwarding for 4G in a network
simulator</title> simulator</title>
<author fullname="Magnus Vevik Austrheim" initials="M.V." surname="Aus trheim"> <author fullname="Magnus Vevik Austrheim" initials="M." surname="Austr heim">
<organization/> <organization/>
</author> </author>
<date month="June" year="2019"/> <date year="2018"/>
</front> </front>
<seriesInfo name="Master's Thesis, Uni Oslo" value=""/> <refcontent>Master's Thesis, University of Oslo</refcontent>
</reference> </reference>
<reference anchor="PragueLinux" target="https://www.netdevconf.org/0x13/se
ssion.html?talk-tcp-prague-l4s"> <reference anchor="PragueLinux" target="https://www.netdevconf.org/0x13/session.
html?talk-tcp-prague-l4s">
<front> <front>
<title>Implementing the `TCP Prague' Requirements for Low Latency <title>Implementing the 'TCP Prague' Requirements for Low Latency
Low Loss Scalable Throughput (L4S)</title> Low Loss Scalable Throughput (L4S)</title>
<author fullname="Bob Briscoe" initials="B." surname="Briscoe"> <author fullname="Bob Briscoe" initials="B." surname="Briscoe">
<organization>Independent</organization> <organization>Independent</organization>
</author> </author>
<author fullname="Koen De Schepper" initials="K." surname="De Schepper "> <author fullname="Koen De Schepper" initials="K." surname="De Schepper ">
<organization>Nokia Bell Labs</organization> <organization>Nokia Bell Labs</organization>
</author> </author>
<author fullname="Olga Albisser" initials="O." surname="Albisser"> <author fullname="Olga Albisser" initials="O." surname="Albisser">
<organization>Simula Research Lab</organization> <organization>Simula Research Lab</organization>
</author> </author>
skipping to change at line 2962 skipping to change at line 1964
<organization>Nokia Bell Labs</organization> <organization>Nokia Bell Labs</organization>
</author> </author>
<author fullname="Mirja Kühlewind" initials="M." surname="Kühlewind"> <author fullname="Mirja Kühlewind" initials="M." surname="Kühlewind">
<organization>ETH Zurich</organization> <organization>ETH Zurich</organization>
</author> </author>
<author fullname="Asad Sajjad Ahmed" initials="A.S." surname="Ahmed"> <author fullname="Asad Sajjad Ahmed" initials="A.S." surname="Ahmed">
<organization>Simula Research Lab</organization> <organization>Simula Research Lab</organization>
</author> </author>
<date month="March" year="2019"/> <date month="March" year="2019"/>
</front> </front>
<seriesInfo name="Proc. Linux Netdev 0x13" value=""/> <refcontent>Proceedings Linux Netdev 0x13</refcontent>
<format target="https://www.files.netdevconf.org/f/4d6939d5f1fb404fafd1/ ?dl=1" type="PDF"/> <format target="https://www.files.netdevconf.org/f/4d6939d5f1fb404fafd1/ ?dl=1" type="PDF"/>
</reference> </reference>
<reference anchor="DualPI2Linux" target="https://www.netdevconf.org/0x13/s ession.html?talk-DUALPI2-AQM"> <reference anchor="DualPI2Linux" target="https://www.netdevconf.org/0x13/s ession.html?talk-DUALPI2-AQM">
<front> <front>
<title>DUALPI2 - Low Latency, Low Loss and Scalable (L4S) <title>DUALPI2 - Low Latency, Low Loss and Scalable (L4S)
AQM</title> AQM</title>
<author fullname="Olga Albisser" initials="O." surname="Albisser"> <author fullname="Olga Albisser" initials="O." surname="Albisser">
<organization>Simula Research Lab</organization> <organization>Simula Research Lab</organization>
</author> </author>
<author fullname="Koen De Schepper" initials="K." surname="De Schepper "> <author fullname="Koen De Schepper" initials="K." surname="De Schepper ">
<organization>Nokia Bell Labs</organization> <organization>Nokia Bell Labs</organization>
</author> </author>
skipping to change at line 2986 skipping to change at line 1989
<organization>Independent</organization> <organization>Independent</organization>
</author> </author>
<author fullname="Olivier Tilmans" initials="O." surname="Tilmans"> <author fullname="Olivier Tilmans" initials="O." surname="Tilmans">
<organization>Nokia Bell Labs</organization> <organization>Nokia Bell Labs</organization>
</author> </author>
<author fullname="Henrik Steen" initials="H." surname="Steen"> <author fullname="Henrik Steen" initials="H." surname="Steen">
<organization>Simula Research Lab</organization> <organization>Simula Research Lab</organization>
</author> </author>
<date month="March" year="2019"/> <date month="March" year="2019"/>
</front> </front>
<seriesInfo name="Proc. Linux Netdev 0x13" value=""/> <refcontent>Proceedings of Linux Netdev 0x13</refcontent>
<format target="https://www.files.netdevconf.org/f/febbe8c6a05b4ceab641/ ?dl=1" type="PDF"/> <format target="https://www.files.netdevconf.org/f/febbe8c6a05b4ceab641/ ?dl=1" type="PDF"/>
</reference> </reference>
<reference anchor="DOCSIS3.1" target="https://specification-search.cablela
bs.com/CM-SP-MULPIv3.1"> <reference anchor="DOCSIS3.1" target="https://specification-search.cablelabs.com
/CM-SP-MULPIv3.1">
<front> <front>
<title>MAC and Upper Layer Protocols Interface (MULPI) <title>MAC and Upper Layer Protocols Interface (MULPI)
Specification, CM-SP-MULPIv3.1</title> Specification, CM-SP-MULPIv3.1</title>
<author fullname="" surname=""> <author fullname="" surname="">
<organization>CableLabs</organization> <organization>CableLabs</organization>
</author> </author>
<date day="21" month="January" year="2019"/> <date day="21" month="January" year="2019"/>
</front> </front>
<seriesInfo name="Data-Over-Cable Service Interface Specifications DOCSI S® 3.1" value="Version i17 or later"/> <seriesInfo name="Data-Over-Cable Service Interface Specifications DOCSI S 3.1" value="Version i17 or later"/>
</reference> </reference>
<reference anchor="AFCD" target="https://doi.org/10.1016/j.jnca.2016.03.02 1"> <reference anchor="AFCD" target="https://doi.org/10.1016/j.jnca.2016.03.02 1">
<front> <front>
<title>Towards fair and low latency next generation high speed <title>Towards fair and low latency next generation high speed
networks: AFCD queuing</title> networks: AFCD queuing</title>
<author fullname="Lin Xue" initials="L." surname="Xue"> <author fullname="Lin Xue" initials="L." surname="Xue">
<organization/> <organization/>
</author> </author>
<author fullname="Suman Kumar" initials="S." surname="Kumar"> <author fullname="Suman Kumar" initials="S." surname="Kumar">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Cheng Cui" initials="C." surname="Cui"> <author fullname="Cheng Cui" initials="C." surname="Cui">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Praveenkumar Kondikoppa" initials="P." surname="Kond ikoppa"> <author fullname="Praveenkumar Kondikoppa" initials="P." surname="Kond ikoppa">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Chui-Hui Chiu" initials="C-H." surname="Chiu"> <author fullname="Chui-Hui Chiu" initials="C-H." surname="Chiu">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Seung-Jong Park" initials="S-J." surname="Park"> <author fullname="Seung-Jong Park" initials="S-J." surname="Park">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<date month="July" year="2016"/> <date month="July" year="2016"/>
</front> </front>
<seriesInfo name="Journal of Network and Computer Applications" value="7 <seriesInfo name="DOI" value="10.1016/j.jnca.2016.03.021"/>
0:183--193"/> <refcontent>Journal of Network and Computer Applications, Volume 70, pp.
183-193</refcontent>
</reference> </reference>
<reference anchor="Nadas20" target="https://doi.org/10.1145/3404868.340666 9"> <reference anchor="Nadas20" target="https://doi.org/10.1145/3404868.340666 9">
<front> <front>
<title>A Congestion Control Independent L4S Scheduler</title> <title>A Congestion Control Independent L4S Scheduler</title>
<author fullname="Szilveszter Nádas" initials="S." surname="Nádas"> <author fullname="Szilveszter Nádas" initials="S." surname="Nádas">
<organization/> <organization/>
</author> </author>
<author fullname="Gergo Gombos" initials="G." surname="Gombos"> <author fullname="Gergő Gombos" initials="G." surname="Gombos">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Ferenc Fejes" initials="F." surname="Fejes"> <author fullname="Ferenc Fejes" initials="F." surname="Fejes">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Sándor Laki" initials="S." surname="Laki"> <author fullname="Sándor Laki" initials="S." surname="Laki">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<date month="July" year="2020"/> <date month="July" year="2020"/>
</front> </front>
<seriesInfo name="Proc. Applied Networking Research Workshop (ANRW '20)" <seriesInfo name="DOI" value="10.1145/3404868.3406669"/>
value="45--51"/> <refcontent>ANRW '20: Proceedings of the Applied Networking Research Work
shop, pp. 45-51</refcontent>
</reference> </reference>
<reference anchor="QDyn" target="https://arxiv.org/abs/1904.07044"> <reference anchor="QDyn" target="https://arxiv.org/abs/1904.07044">
<front> <front>
<title>Rapid Signalling of Queue Dynamics</title> <title>Rapid Signalling of Queue Dynamics</title>
<author fullname="Bob Briscoe" initials="B." surname="Briscoe"> <author fullname="Bob Briscoe" initials="B." surname="Briscoe">
<organization>bobbriscoe.net Ltd</organization> <organization>bobbriscoe.net Ltd</organization>
</author> </author>
<date month="September" year="2017"/> <date month="April" year="2019"/>
</front> </front>
<seriesInfo name="bobbriscoe.net Technical Report" value="TR-BB-2017-001 <seriesInfo name="DOI" value="10.48550/arXiv.1904.07044"/>
; arXiv:1904.07044 [cs.NI]"/> <refcontent>TR-BB-2017-001, arXiv:1904.07044 [cs.NI]</refcontent>
<format target="https://arxiv.org/pdf/1904.07044" type="PDF"/> <format target="https://arxiv.org/pdf/1904.07044" type="PDF"/>
</reference> </reference>
<reference anchor="McIlroy78" target="https://archive.org/details/bstj57-6 -1899"> <reference anchor="McIlroy78" target="https://archive.org/details/bstj57-6 -1899">
<front> <front>
<title>UNIX Time-Sharing System: Foreword</title> <title>UNIX Time-Sharing System: Foreword</title>
<author fullname="Doug McIlroy" initials="M.D." surname="McIlroy"> <author fullname="Doug McIlroy" initials="M.D." surname="McIlroy">
<organization/> <organization/>
</author> </author>
<author initials="E. N." surname="Pinson"> <author initials="E. N." surname="Pinson">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author initials="B. A." surname="Tague"> <author initials="B. A." surname="Tague">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<date month="July" year="1978"/> <date month="July" year="1978"/>
</front> </front>
<seriesInfo name="The Bell System Technical Journal" value="57:6(1902--1 <seriesInfo name="DOI" value="10.1002/j.1538-7305.1978.tb02135.x"/>
903)"/> <refcontent>The Bell System Technical Journal 57: 6, pp. 1899-1904</refc
ontent>
</reference> </reference>
<reference anchor="LEDBAT_AQM" target="https://ieeexplore.ieee.org/documen t/8109367"> <reference anchor="LEDBAT_AQM" target="https://ieeexplore.ieee.org/documen t/8109367">
<front> <front>
<title>Characterising LEDBAT Performance Through Bottlenecks Using <title>Characterising LEDBAT Performance Through Bottlenecks Using
PIE, FQ-CoDel and FQ-PIE Active Queue Management</title> PIE, FQ-CoDel and FQ-PIE Active Queue Management</title>
<author fullname="Rasool Al-Saadi" initials="R." surname="Al-Saadi"> <author fullname="Rasool Al-Saadi" initials="R." surname="Al-Saadi">
<organization/> <organization/>
</author> </author>
<author fullname="Grenville Armitage" initials="G." surname="Armitage" > <author fullname="Grenville Armitage" initials="G." surname="Armitage" >
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Jason But" initials="J." surname="But"> <author fullname="Jason But" initials="J." surname="But">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<date year="2017"/> <date month="October" year="2017"/>
</front> </front>
<seriesInfo name="Proc. IEEE 42nd Conference on Local Computer Networks <seriesInfo name="DOI" value="10.1109/LCN.2017.22"/>
(LCN)" value="278--285"/> <refcontent>IEEE 42nd Conference on Local Computer Networks (LCN)</refco
ntent>
</reference> </reference>
<reference anchor="BBRv2" target="https://github.com/google/bbr/blob/v2alp ha/README.md"> <reference anchor="BBRv2" target="https://github.com/google/bbr">
<front> <front>
<title>TCP BBR v2 Alpha/Preview Release</title> <title>TCP BBR v2 Alpha/Preview Release</title>
<author fullname="Neal Cardwell" initials="N" surname="Cardwell"> <author>
<organization/> <organization/>
</author> </author>
<date/> <date month="June" year="2022" />
</front> </front>
<seriesInfo name="GitHub repository;" value="Linux congestion control mo <refcontent>commit 17700ca</refcontent>
dule"/>
</reference> </reference>
<reference anchor="SCReAM" target="https://github.com/EricssonResearch/scr
eam/blob/master/README.md"> <reference anchor="SCReAM-L4S" target="https://github.com/EricssonResearch
/scream">
<front> <front>
<title>SCReAM</title> <title>SCReAM</title>
<author fullname="Ingemar Johansson" initials="I" surname="Johansson"> <author>
<organization/> <organization/>
</author> </author>
<date/> <date month="June" year="2022"/>
</front> </front>
<seriesInfo name="GitHub repository;" value=""/> <refcontent>commit fda6c53</refcontent>
<format target="https://github.com/google/bbr/blob/v2alpha/README.md" ty
pe="Source code"/>
</reference> </reference>
<reference anchor="BDPdata" target="https://arxiv.org/abs/2107.01003"> <reference anchor="BDPdata" target="https://arxiv.org/abs/2107.01003">
<front> <front>
<title>PI2 Parameters</title> <title>PI2 Parameters</title>
<author fullname="Bob Briscoe" initials="B." surname="Briscoe"> <author fullname="Bob Briscoe" initials="B." surname="Briscoe">
<organization/> <organization/>
</author> </author>
<date month="July" year="2021"/> <date month="October" year="2021"/>
</front> </front>
<seriesInfo name="Technical Report" value="TR-BB-2021-001 arXiv:2107.010 <seriesInfo name="DOI" value="10.48550/arXiv.2107.01003"/>
03 [cs.NI]"/> <refcontent>TR-BB-2021-001, arXiv:2107.01003 [cs.NI]</refcontent>
<format target="https://arxiv.org/pdf/2107.01003" type="PDF"/>
</reference> </reference>
<reference anchor="BufferSize" target="https://doi.org/10.1145/1015467.101 5499"> <reference anchor="BufferSize" target="https://doi.org/10.1145/1015467.101 5499">
<front> <front>
<title>Sizing Router Buffers</title> <title>Sizing Router Buffers</title>
<author fullname="Guido Appenzeller" initials="G." surname="Appenzelle r"> <author fullname="Guido Appenzeller" initials="G." surname="Appenzelle r">
<organization>Stanford Uni</organization> <organization>Stanford University</organization>
</author> </author>
<author fullname="Isaac Keslassy" initials="I." surname="Keslassy"> <author fullname="Isaac Keslassy" initials="I." surname="Keslassy">
<organization>Stanford Uni</organization> <organization>Stanford University</organization>
</author> </author>
<author fullname="Nick McKeown" initials="N." surname="McKeown"> <author fullname="Nick McKeown" initials="N." surname="McKeown">
<organization>Stanford Uni</organization> <organization>Stanford University</organization>
</author> </author>
<date month="September" year="2004"/> <date month="October" year="2004"/>
</front> </front>
<seriesInfo name="In Proc. SIGCOMM'04" value="34(4):281--292"/> <seriesInfo name="DOI" value="10.1145/1015467.1015499"/>
<format target="http://yuba.stanford.edu/~nickm/papers/sigcomm2004.pdf" <refcontent>SIGCOMM '04: Proceedings of the 2004 conference on Applicati
type="PDF"/> ons, technologies, architectures, and protocols for computer communications, pp.
281-292</refcontent>
</reference> </reference>
<reference anchor="COBALT" target="https://ieeexplore.ieee.org/abstract/do cument/8847054"> <reference anchor="COBALT" target="https://ieeexplore.ieee.org/abstract/do cument/8847054">
<front> <front>
<title>Design and Evaluation of COBALT Queue Discipline</title> <title>Design and Evaluation of COBALT Queue Discipline</title>
<author fullname="Jendaipou Palmei" initials="J." surname="Palmei"> <author fullname="Jendaipou Palmei" initials="J." surname="Palmei">
<organization/> <organization/>
</author> </author>
<author fullname="Shefali Gupta" initials="S." surname="Gupta"> <author fullname="Shefali Gupta" initials="S." surname="Gupta">
<organization/> <organization/>
</author> </author>
<author fullname="Pasquale Imputato" initials="P." surname="Imputato"> <author fullname="Pasquale Imputato" initials="P." surname="Imputato">
skipping to change at line 3308 skipping to change at line 2181
<organization/> <organization/>
</author> </author>
<author fullname="Stefan Avallone" initials="S." surname="Avallone"> <author fullname="Stefan Avallone" initials="S." surname="Avallone">
<organization/> <organization/>
</author> </author>
<author fullname="Dave Täht" initials="D." surname="Täht"> <author fullname="Dave Täht" initials="D." surname="Täht">
<organization/> <organization/>
</author> </author>
<date month="July" year="2019"/> <date month="July" year="2019"/>
</front> </front>
<seriesInfo name="In Proc. IEEE Int'l Symp. Local and Metropolitan Area <seriesInfo name="DOI" value="10.1109/LANMAN.2019.8847054"/>
Networks (LANMAN'19)" value="2019:1-6"/> <refcontent>IEEE International Symposium on Local and Metropolitan Area N
<format target="https://www.researchgate.net/publication/336087375_Desig etworks (LANMAN)</refcontent>
n_and_Evaluation_of_COBALT_Queue_Discipline" type="PDF"/>
</reference> </reference>
<reference anchor="DOCSIS3AQM" target="{https://www.cablelabs.com/wp-conte
nt/uploads/2013/11/Active_Queue_Management_Algorithms_DOCSIS_3_0.pdf"> <reference anchor="DOCSIS3AQM" target="https://www.cablelabs.com/wp-conten
t/uploads/2013/11/Active_Queue_Management_Algorithms_DOCSIS_3_0.pdf">
<front> <front>
<title>Active Queue Management Algorithms for DOCSIS 3.0; A <title>Active Queue Management Algorithms for DOCSIS 3.0: A
Simulation Study of CoDel, SFQ-CoDel and PIE in DOCSIS 3.0 Simulation Study of CoDel, SFQ-CoDel and PIE in DOCSIS 3.0
Networks</title> Networks</title>
<author fullname="Greg White" initials="G." surname="White"> <author fullname="Greg White" initials="G." surname="White">
<organization/> <organization/>
</author> </author>
<date month="April" year="2013"/> <date month="April" year="2013"/>
</front> </front>
<seriesInfo name="CableLabs Technical Report" value=""/> <refcontent>CableLabs Technical Report</refcontent>
</reference> </reference>
<reference anchor="FQ_CoDel_Thresh" target="https://git.kernel.org/pub/scm
/linux/kernel/git/netdev/net-next.git/commit/?id=dfcb63ce1de6b10b"> <reference anchor="FQ_CoDel_Thresh" target="https://git.kernel.org/pub/scm/linux
/kernel/git/netdev/net-next.git/commit/?id=dfcb63ce1de6b10b">
<front> <front>
<title>fq_codel: generalise ce_threshold marking for subset of <title>fq_codel: generalise ce_threshold marking for subset of
traffic</title> traffic</title>
<author fullname="Toke Høiland-Jørgensen" initials="T." surname="Høila nd-Jørgensen"> <author>
<organization/> <organization/>
</author> </author>
<date day="20" month="October" year="2021"/> <date month="October" year="2021"/>
</front> </front>
<seriesInfo name="Linux Patch" value="Commit ID: dfcb63ce1de6b10b"/> <refcontent>commit dfcb63ce1de6b10b</refcontent>
</reference> </reference>
<reference anchor="Dukkipati06" target="https://dl.acm.org/doi/10.1145/111 1322.1111336"> <reference anchor="Dukkipati06" target="https://dl.acm.org/doi/10.1145/111 1322.1111336">
<front> <front>
<title>Why Flow-Completion Time is the Right Metric for Congestion <title>Why Flow-Completion Time is the Right Metric for Congestion
Control</title> Control</title>
<author fullname="Nandita Dukkipati" initials="N." surname="Dukkipati" > <author fullname="Nandita Dukkipati" initials="N." surname="Dukkipati" >
<organization>Stanford Uni</organization> <organization>Stanford University</organization>
</author> </author>
<author fullname="Nick McKeown" initials="N." surname="McKeown"> <author fullname="Nick McKeown" initials="N." surname="McKeown">
<organization>Stanford Uni</organization> <organization>Stanford University</organization>
</author> </author>
<date month="January" year="2006"/> <date month="January" year="2006"/>
</front> </front>
<seriesInfo name="ACM CCR" value="36(1):59--62"/> <seriesInfo name="DOI" value="10.1145/1111322.1111336"/>
<format target="http://yuba.stanford.edu/rcp/flowCompTime-dukkipati.pdf" <refcontent>ACM SIGCOMM Computer Communication Review, Volume 36, Issue
type="PDF"/> 1, pp. 59-62</refcontent>
</reference> </reference>
<reference anchor="Rajiullah15" target="https://www.diva-portal.org/smash/ get/diva2:846109/FULLTEXT01.pdf"> <reference anchor="Rajiullah15" target="https://www.diva-portal.org/smash/ get/diva2:846109/FULLTEXT01.pdf">
<front> <front>
<title>Towards a Low Latency Internet: Understanding and <title>Towards a Low Latency Internet: Understanding and
Solutions</title> Solutions</title>
<author fullname="Mohammad Rajiullah" initials="M." surname="Rajiullah "> <author fullname="Mohammad Rajiullah" initials="M." surname="Rajiullah ">
<organization>Karlstad Uni</organization> <organization>Karlstad University Studies</organization>
</author> </author>
<date year="2015"/> <date year="2015"/>
</front> </front>
<seriesInfo name="Master's Thesis; Karlstad Uni, Dept of Maths &amp; CS " value="2015:41"/> <refcontent>Dissertation, Karlstad University</refcontent>
</reference> </reference>
<reference anchor="lowat" target="https://blog.cloudflare.com/http-2-prior itization-with-nginx/"> <reference anchor="lowat" target="https://blog.cloudflare.com/http-2-prior itization-with-nginx/">
<front> <front>
<title>Optimizing HTTP/2 prioritization with BBR and <title>Optimizing HTTP/2 prioritization with BBR and
tcp_notsent_lowat</title> tcp_notsent_lowat</title>
<author fullname="Patrick Meenan" initials="P." surname="Meenan"> <author fullname="Patrick Meenan" initials="P." surname="Meenan">
<organization>Cloudflare</organization> <organization>Cloudflare</organization>
</author> </author>
<date day="12" month="October" year="2018"/> <date month="October" year="2018"/>
</front> </front>
<seriesInfo name="Cloudflare Blog" value=""/> <refcontent>Cloudflare Blog</refcontent>
</reference> </reference>
<reference anchor="NASA04" target="https://ntrs.nasa.gov/api/citations/201 20009198/downloads/20120009198.pdf?attachment=true"> <reference anchor="NASA04" target="https://ntrs.nasa.gov/api/citations/201 20009198/downloads/20120009198.pdf?attachment=true">
<front> <front>
<title>Latency Requirements for Head-Worn Display S/EVS <title>Latency Requirements for Head-Worn Display S/EVS
Applications</title> Applications</title>
<author fullname="Randall E. Bailey" initials="R.R." surname="Bailey"> <author fullname="Randall E. Bailey" initials="R." surname="Bailey">
<organization>NASA</organization> <organization>NASA Langley Research Center</organization>
</author> </author>
<author fullname="J.J. (Trey) Arthur III" initials="J.J." surname="Tre <author fullname="J.J. (Trey) Arthur III" initials="J." surname="Trey
y Arthur III"> Arthur III">
<organization>NASA</organization> <organization>NASA Langley Research Center</organization>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<author fullname="Steven P. Williams" initials="S.P." surname="William <author fullname="Steven P. Williams" initials="S." surname="Williams"
s"> >
<organization>NASA</organization> <organization>NASA Langley Research Center</organization>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<date month="April" year="2004"/> <date month="April" year="2004"/>
</front> </front>
<seriesInfo name="SPIE Defense and Security Symposium" value="LF99-1955" <seriesInfo name="DOI" value="10.1117/12.554462"/>
/> <refcontent>Proceedings of SPIE 5424</refcontent>
</reference> </reference>
<reference anchor="Raaen14" target="http://ojs.bibsys.no/index.php/NIK/art
icle/view/9/6"> <reference anchor="Raaen14" target="http://ojs.bibsys.no/index.php/NIK/article/v
iew/9/6">
<front> <front>
<title>Latency thresholds for usability in games: A survey</title> <title>Latency Thresholds for Usability in Games: A Survey</title>
<author fullname="Kjetil Raaen" initials="K." surname="Raaen"> <author fullname="Kjetil Raaen" initials="K." surname="Raaen">
<organization/> <organization/>
</author> </author>
<author fullname="Tor-Morten Grønli" initials="T-M." surname="Grønli"> <author fullname="Tor-Morten Grønli" initials="T-M." surname="Grønli">
<organization/> <organization/>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country/>
</postal>
<phone/>
<email/>
<uri/>
</address>
</author> </author>
<date year="2014"/> <date year="2014"/>
</front> </front>
<seriesInfo name="Norsk IKT-konferanse for forskning og utdanning" value <refcontent>Norsk IKT-konferanse for forskning og utdanning (Norwegian
=""/> ICT conference for research and education)</refcontent>
</reference> </reference>
</references> </references>
<!-- <section title="Change Log (to be Deleted before Publication)">
<t>A detailed version history can be accessed at
&lt;http://datatracker.ietf.org/doc/draft-briscoe-aqm-ecn-roadmap/history/
&gt;</t>
<t><list style="hanging">
<t hangText="From briscoe-...-00 to briscoe-...-01:">Technical
changes:<list style="symbols">
<t/>
</list>Editorial changes:<list style="symbols">
<t/>
</list></t>
</list></t>
</section>
<section numbered="false" toc="default"> <section numbered="false" toc="default">
<name>Acknowledgements</name> <name>Acknowledgements</name>
<t>Thanks to Richard Scheffenegger, Wes Eddy, Karen Nielsen, David <t>Thanks to <contact fullname="Richard Scheffenegger"/>, <contact
Black, Jake Holland, Vidhi Goel, Ermin Sakic, Praveen Balasubramanian, fullname="Wes Eddy"/>, <contact fullname="Karen Nielsen"/>, <contact
Gorry Fairhurst, Mirja Kuehlewind, Philip Eardley, Neal Cardwell, Pete fullname="David Black"/>, <contact fullname="Jake Holland"/>, <contact
Heist and Martin Duke for their useful review comments. Thanks also to fullname="Vidhi Goel"/>, <contact fullname="Ermin Sakic"/>, <contact
the area reviewers: Marco Tiloca, Lars Eggert, Roman Danyliw and fullname="Praveen Balasubramanian"/>, <contact fullname="Gorry
Eric Vyncke.</t> Fairhurst"/>, <contact fullname="Mirja Kuehlewind"/>, <contact
<t>Bob Briscoe and Koen De Schepper were part-funded by the European fullname="Philip Eardley"/>, <contact fullname="Neal Cardwell"/>,
Community under its Seventh Framework Programme through the Reducing <contact fullname="Pete Heist"/>, and <contact fullname="Martin Duke"/>
Internet Transport Latency (RITE) project (ICT-317700). The contribution for their useful review comments. Thanks also to the area reviewers:
of Koen De Schepper was also part-funded by the 5Growth and DAEMON EU <contact fullname="Marco Tiloca"/>, <contact fullname="Lars Eggert"/>,
H2020 projects. Bob Briscoe was also part-funded by the Research Council <contact fullname="Roman Danyliw"/>, and <contact fullname="Éric
of Norway through the TimeIn project, partly by CableLabs and partly by Vyncke"/>.</t>
the Comcast Innovation Fund. The views expressed here are solely those <t><contact fullname="Bob Briscoe"/> and <contact fullname="Koen De
of the authors.</t> Schepper"/> were partly funded by the European Community under its Seventh
Framework Programme through the Reducing Internet Transport Latency
(RITE) project (ICT-317700). The contribution of <contact fullname="Koen
De Schepper"/> was also partly funded by the 5Growth and DAEMON EU H2020
projects. <contact fullname="Bob Briscoe"/> was also partly funded by the
Research Council of Norway through the TimeIn project, partly by
CableLabs, and partly by the Comcast Innovation Fund. The views expressed
here are solely those of the authors.</t>
</section> </section>
</back> </back>
</rfc> </rfc>
 End of changes. 351 change blocks. 
2844 lines changed or deleted 1313 lines changed or added

This html diff was produced by rfcdiff 1.48.