Network Working Group H. Lundin
Internet-Draft S. Holmer
Intended status: Informational H. T. Alvestrand, Ed.
Expires: May 01, 2012 Google
October 29, 2011

A Google Congestion Control Algorithm for Real-Time Communication on the World Wide Web


This document describes two methods of congestion control when using real-time communications on the World Wide Web (RTCWEB); one sender-based and one receiver-based.

It is published to aid the discussion on mandatory-to-implement flow control for RTCWEB applications; initial discussion is expected in the RTCWEB WG's mailing list.

Requirements Language

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119].

Status of this Memo

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

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

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

This Internet-Draft will expire on May 01, 2012.

Copyright Notice

Copyright (c) 2011 IETF Trust and the persons identified as the document authors. All rights reserved.

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

Table of Contents

1. Introduction

Congestion control is a requirement for all applications that wish to share the Internet [RFC2914].

The problem of doing congestion control for real-time media is made difficult for a number of reasons:

This memo describes two congestion control algorithms that together are seen to give reasonable performance and reasonable (not perfect) bandwidth sharing with other conferences and with TCP-using applications that share the same links.

The signalling used consists of standard RTP timestamps [RFC3550], standard RTCP feedback reports and Temporary Maximum Media Stream Bit Rate Requests (TMMBR) as defined in [RFC5104] section 3.5.4.

1.1. Mathemathical notation conventions

The mathematics of this document have been transcribed from a more formula-friendly format.

The following notational conventions are used:

The variable X, where X is a vector - conventionally marked by a bar on top of the variable name.
An estimate of the true value of variable X - conventionally marked by a circumflex accent on top of the variable name.
The "i"th value of X - conventionally marked by a subscript i.
[x y z]
A row vector consisting of elements x, y and z.
The transpose of vector X_bar.
The expected value of the stochastic variable X

2. System model

The following elements are in the system:

Together, sender side control and receiver side control implement the congestion control algorithm.

3. Receiver side control

The receive-side algorithm can be further decomposed into three parts: an arrival-time filter, an over-use detector, and a remote rate-control.

3.1. Arrival-time model

This section describes an adaptive filter that continuously updates estimates of network parameters based on the timing of the received frames.

At the receiving side we are observing groups of incoming video packets, where each group of packets corresponding to the same frame having timestamp T(i).

Each frame is assigned a receive time t(i), which corresponds to the time at which the whole frame has been received (ignoring any packet losses). A frame is delayed relative to its predecessor if t(i)-t(i-1)>T(i)-T(i-1), i.e., if the arrival time difference is larger than the timestamp difference.

We define the (relative) inter-arrival time, d(i) as

  d(i) = t(i)-t(i-1)-(T(i)-T(i-1))

Since the time ts to send a frame of size L over a path with a capacity of C is

  ts = L/C

we can model the inter-arrival time as

  d(i) = -------------- + w(i) = dL(i)/C+w(i)

Here, w(i) is a sample from a stochastic process W, which is a function of the capacity C, the current cross traffic X(i), and the current send bit rate R(i). We model W as a white Gaussian process. If we are over-using the channel we expect w(i) to increase, and if a queue on the network path is being emptied, w(i) will decrease; otherwise the mean of w(i) will be zero.

Breaking out the mean of w(i) to make the process zero mean, we get

Equation 5

  d(i) = dL(i)/C + m(i) + v(i)

This is our fundamental model, where we take into account that a large frame needs more time to traverse the link than a small frame, thus arriving with higher relative delay. The noise term represents network jitter and other delay effects not captured by the model.

When graphing the values for d(i) versus dL(i) on a scatterplot, we find that most samples cluster around the center, and the outliers are clustered along a line with average slope 1/C and zero offset.

When using a regular video codec, most frames are roughly the same size after encoding (the central “cloud”); the exceptions are I-frames (or key frames) which are typically much larger than the average causing positive outliers (the I-frame itself) and negative outliers (the frame after an I-frame) on the dL axis.

3.2. Arrival-time filter

The parameters d(i) and dL(i) are readily available for each frame i, and we want to estimate C and m(i) and use those estimates to detect whether or not we are over-using the bandwidth currently available. These parameters are easily estimated by any adaptive filter – we are using the Kalman filter.


  theta_bar(i) = [1/C(i)  m(i)]^T

and call it the state of time i. We model the state evolution from time i to time i+1 as

  theta_bar(i+1) = theta_bar(i) + u_bar(i)

where u_bar(i) is the zero mean white Gaussian process noise with covariance

Equation 7

  Q(i) = E{u_bar(i) u_bar(i)^T}

Given equation 5 we get

Equation 8

  d(i) = h_bar(i)^T theta_bar(i) + v(i)

  h_bar(i) = [dL(i)  1]^T

where v(i) is zero mean white Gaussian measurement noise with variance var_v = sigma(v,i)^2

The Kalman filter recursively updates our estimate

  theta_hat(i) = [1/C_hat(i) m_hat(i)]^T


  z(i) = d(i) - h_bar(i)^T * theta_hat(i-1)

  theta_hat(i) = theta_hat(i-1) + z(i) * k_bar(i)

                           E(i-1) * h_bar(i)
  k_bar(i) = --------------------------------------------
               var_v_hat + h_bar(i)^T * E(i-1) * h_bar(i)

  E(i) = (I - K_bar(i) * h_bar(i)^T) * E(i-1) + Q(i)

I is the 2-by-2 identity matrix.

The variance var_v = sigma(v,i)^2 is estimated using an exponential averaging filter, modified for variable sampling rate

  var_v_hat = beta*sigma(v,i-1)^2 + (1-beta)*z(i)^2

  beta = (1-alpha)^(30/(1000 * f_max))

where f_max = max {1/(T(j) - T(j-1))} for j in i-K+1...i is the highest rate at which frames have been captured by the camera the last K frames and alpha is a filter coefficient typically chosen as a number in the interval [0.1, 0.001]. Since our assumption that v(i) should be zero mean WGN is less accurate in some cases, we have introduced an additional outlier filter around the updates of var_v_hat. If z(i) > 3 var_v_hat the filter is updated with 3 sqrt(var_v_hat) rather than z(i). In a similar way, Q(i) is chosen as a diagonal matrix with main diagonal elements given by

  diag(Q(i)) = 30/(1000 * f_max)[10^-10 10^-2]^T

It is necessary to scale these filter parameters with the frame rate to make the detector respond as quickly at low frame rates as at high frame rates.

3.3. Over-use detector

The offset estimate m(i) is compared with a threshold gamma_1. An estimate above the threshold is considered as an indication of over-use. Such an indication is not enough for the detector to signal over-use to the rate control subsystem. Not until over-use has been detected for at least gamma_2 milliseconds and at least gamma_3 frames, a definitive over-use will be signaled. However, if the offset estimate m(i) was decreased in the last update, over-use will not be signaled even if all the above conditions are met. Similarly, the opposite state, under-use, is detected when m(i) < -gamma_1. If neither over-use nor under-use is detected, the detector will be in the normal state.

3.4. Rate control

The rate control at the receiving side is designed to increase the available bandwidth estimate A_hat as long as the detected state is normal. Doing that assures that we, sooner or later, will reach the available bandwidth of the channel and detect an over-use.

As soon as over-use has been detected the available bandwidth estimate is decreased. In this way we get a recursive and adaptive estimate of the available bandwidth.

In this document we make the assumption that the rate control subsystem is executed periodically and that this period is constant.

The rate control subsystem has 3 states: Increase, Decrease and Hold. "Increase" is the state when no congestion is detected; "Decrease" is the state where congestion is detected, and "Hold" is a state that waits until built-up queues have drained before going to "increase" state.

The state transitions (with blank fields meaning "remain in state") are:

State ---->  | Hold      |Increase    |Decrease
  v          |           |            |
Over-use     | Decrease  |Decrease    |
Normal       | Increase  |            |Hold
Under-use    |           |Hold        |Hold

The subsystem starts in the increase state, where it will stay until over-use or under-use has been detected by the detector subsystem. On every update the available bandwidth is increased with a factor which is a function of the global system response time and the estimated measurement noise variance var_v_hat. The global system response time is the time from an increase that causes over-use until that over-use can be detected by the over-use detector. The variance var_v_hat affects how responsive the Kalman filter is, and is thus used as an indicator of the delay inflicted by the Kalman filter.

  A_hat(i) = eta*A_hat(i-1)
  eta(RTT, var_v_hat) = ------------------------------------------
                           1+e^(b(d*RTT - (c1 * var_v_hat + c2)))

Here, B, b, d, c1 and c2 are design parameters.

Since the system depends on over-using the channel to verify the current available bandwidth estimate, we must make sure that our estimate doesn’t diverge from the rate at which the sender is actually sending. Thus, if the sender is unable to produce a bit stream with the bit rate the receiver is asking for, the available bandwidth estimate must stay within a given bound. Therefore we introduce a threshold

  A_hat(i) < 1.5 * R_hat(i)

where R_hat(i) is the incoming bit rate measured over a T seconds window:

  R_hat(i) = 1/T * sum(L(j)) for j from 1 to N(i)

N(i) is the number of frames received the past T seconds and L(j) is the payload size of frame j.

When an over-use is detected the system transitions to the decrease state, where the available bandwidth estimate is decreased to a factor times the currently incoming bit rate.

  A_hat(i) = alpha*R_hat(i)

alpha is typically chosen to be in the interval [0.8, 0.95].

When the detector signals under-use to the rate control subsystem, we know that queues in the network path are being emptied, indicating that our available bandwidth estimate is lower than the actual available bandwidth. Upon that signal the rate control subsystem will enter the hold state, where the available bandwidth estimate will be held constant while waiting for the queues to stabilize at a lower level – a way of keeping the delay as low as possible. This decrease of delay is wanted, and expected, immediately after the estimate has been reduced due to over-use, but can also happen if the cross traffic over some links is reduced. In either case we want to measure the highest incoming rate during the under-use interval:

  R_max = max{R_hat(i)} for i in 1..K

where K is the number of frames of under-use before returning to the normal state. R_max is a measure of the actual bandwidth available and is a good guess of what bit rate the sender should be able to transmit at. Therefore the available bandwidth will be set to Rmax when we transition from the hold state to the increase state.

One design decision is when to send rate control messages. The time from a change in congestion to the sending of the feedback message is a limitation on how fast the sender can react. Sending too many messages giving no new information is a waste of bandwidth - but in the case of severe congestion, feedback messages can be lost, resulting in a failure to react in a timely manner.

The conclusion is that feedback messages should be sent on a "heartbeat" schedule, allowing the sender side control to react to missing feedback messages by reducing its send rate, but they should also be sent whenever the estimated bandwidth value has changed significantly, without waiting for the heartbeat time, up to some limiting upper bound on the send rate.

The minimum interval is named t_min_fb_interval.

The maximum interval is named t_max_fb_interval.

The permissible values of these intervals will be bounded by the RTP session's RTCP bandwidht and its rtcp_frr setting.

[TODO: Get some example values for these timers]

4. Sender side control

An additional congestion controller resides at the sending side. It bases its decisions on the round-trip time, packet loss and available bandwidth estimates transmitted from the receiving side.

The available bandwidth estimates produced by the receiving side are only reliable when the size of the queues along the channel are large enough. If the queues are very short, over-use will only be visible through packet losses, which aren't used by the receiving side algorithm.

This algorithm is run every time a receive report arrives at the sender, which will happen no more often than t_min_fb_interval, and no less often than t_max_fb_interval. If no receive report is recieved within 2x t_max_fb_interval (indicating at least 2 lost feedback reports), the algorithm will take action as if all packets in the interval have been lost, resulting in a halving of the send rate.

The new send-side estimate is limited by the TCP Friendly Rate Control formula [RFC3448] and the receive-side estimate of the available bandwidth A(i):

                               8 s
As(i) >= ----------------------------------------------------------
         R*sqrt(2*b*p/3) + (t_RTO*(3*sqrt(3*b*p/8) * p * (1+32*p^2)))

As(i) <= A(i)

where b is the number of packets acknowledged by a single TCP acknowledgement (set to 1 per TFRC recommendations), t_RTO is the TCP retransmission timeout value in seconds (set to 4*R) and s is the average packet size in bytes. R is the round-trip time in seconds.

(The multiplication by 8 comes because TFRC is computing bandwidth in bytes, while this document computes bandwidth in bits.)

In words: The sender-side estimate will never be larger than the receiver-side estimate, and will never be lower than the estimate from the TFRC formula.

We motivate the packet loss thresholds by noting that if the transmission channel has a small amount of packet loss due to over-use, that amount will soon increase if the sender does not adjust his bit rate. Therefore we will soon enough reach above the 10 % threshold and adjust As(i). However if the packet loss rate does not increase, the losses are probably not related to self-induced channel over-use and therefore we should not react on them.

5. Interoperability Considerations

There are three scenarios of interest, and one included for reference

In the case where both parties implement the algorithms, we expect to see most of the congestion control response to slowly varying conditions happen by TMMBR messages from recipient to sender. At most times, the sender will send less than the congestion-inducing bandwidth limit C, and when he sends more, congestion will be detected before packets are lost.

If sudden changes happen, packets will be lost, and the sender side control will trigger, limiting traffic until the congestion becomes low enough that the system switches back to the receiver-controlled state.

In the case where sender only implements, we expect to see somewhat higher loss rates and delays, but the system will still be overall TCP friendly and self-adjusting; the governing term in the calculation will be the TFRC formula.

In the case where recipient implements this algorithm and sender does not, congestion will be avoided for slow changes as long as the sender understands and obeys TMMBR; there will be no backoff for packet-loss-inducing changes in capacity. Given that some kind of congestion control is mandatory for the sender according to the TMMBR spec, this case has to be reevaluated against the specific congestion control implemented by the sender.

6. Implementation Experience

This algorithm has been implemented in the open-source WebRTC project.

7. Further Work

This draft is offered as input to the congestion control discussion.

Work that can be done on this basis includes:

These are matters for further work; since some of them involve extensions that have not yet been standardized, this could take some time.

8. IANA Considerations

This document makes no request of IANA.

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

9. Security Considerations

An attacker with the ability to insert or remove messages on the connection will, of course, have the ability to mess up rate control, causing people to send either too fast or too slow, and causing congestion.

In this case, the control information is carried inside RTP, and can be protected against modification or message insertion using SRTP, just as for the media. Given that timestamps are carried in the RTP header, which is not encrypted, this is not protected against disclosure, but it seems hard to mount an attack based on timing information only.

10. Acknowledgements

Thanks to Randell Jesup, Magnus Westerlund, Varun Singh, Tim Panton, Soo-Hyun Choo, Jim Gettys, Ingemar Johansson and others for providing valuable feedback on earlier versions of this draft.

11. References

11.1. Normative References

[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC3448] Handley, M., Floyd, S., Padhye, J. and J. Widmer, "TCP Friendly Rate Control (TFRC): Protocol Specification", RFC 3448, January 2003.
[RFC3550] Schulzrinne, H., Casner, S., Frederick, R. and V. Jacobson, "RTP: A Transport Protocol for Real-Time Applications", STD 64, RFC 3550, July 2003.
[RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C. and J. Rey, "Extended RTP Profile for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, July 2006.
[RFC5104] Wenger, S., Chandra, U., Westerlund, M. and B. Burman, "Codec Control Messages in the RTP Audio-Visual Profile with Feedback (AVPF)", RFC 5104, February 2008.

11.2. Informative References

[I-D.gharai-avtcore-rtp-tfrc] Gharai, L and C Perkins, "RTP with TCP Friendly Rate Control", Internet-Draft draft-gharai-avtcore-rtp-tfrc-01, September 2011.
[RFC2914] Floyd, S., "Congestion Control Principles", BCP 41, RFC 2914, September 2000.

Appendix A. New proposed functionality

This section proposes two new functionalities: An RTP header extension for signalling the time of packet emission, and an RTCP feedback message signalling the requested total bandwidth for a section.

If these two functions are available, it is possible to implement the algorithm in this document, or other algorithms that take the same input, in a fashion that is likely to be more precise than the one that depends on RTP timestamps, and can cover multiple flows instead of just one.

This section is intended to be pulled out in a later separate Internet-Draft and be proposed for standardization.

Appendix A.1. Send timestamp

The send timestamp serves to record the last time at which the packet was available for modification to the RTP sender - that is, as close as possible to the time at which the packet was actually queued for sending on the wire.

     0                   1                   2                   3
     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
    |      0xBE     |      0xBE     |            length=1           |
    |  ID   | len=2 |     send timestamp  (t_i)                     |


Send timestamp (t_i)
24 bits The timestamp indicating when the packet is sent. This timestamp is measured in microseconds and is used for bandwidth estimation.

The absolute value of the send timestamp does not matter. The value MUST be consistent between packets sent from the same sender.

Appendix A.2. Receiver Estimated Max Bitrate (REMB)

Appendix A.2.1. Semantics

This feedback message is used to notify a sender of multiple media streams over the same RTP session of the total estimated available bit rate on the path to the receiving side of this RTP session.

Within the common packet header for feedback messages (as defined in section 6.1 of [RFC4585]), the "SSRC of packet sender" field indicates the source of the notification. The "SSRC of media source" is not used and SHALL be set to 0. RFC 5104 section

The reception of a REMB message SHALL result in that the total bit rate sent on the RTP session this message applies to is equal to or lower than the bit rate in this message. The new bit rate constraint should be applied as fast as resonable. The sender is free to apply additional bandwidth restrictions based on its own restrictions and estimates.

Appendix A.2.2. Message format

This document describes a message using the application specific paylaod type. This is suitable for experimentation; upon standardization, a specific type can be assigned for the purpose.

RTCP message with payload type 206. Reference RFC 3550, 4585 and 5104.

 0                   1                   2                   3               
 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
|V=2|P| FMT=15  |   PT=206      |             length            |                               
|                  SSRC of packet sender                        |
|                  SSRC of media source                         |
|  Unique identifier 'R' 'E' 'M' 'B'                            |
|  Num SSRC     | BR Exp      |  BR Mantissa                    |
|   SSRC feedback                                               |
|  ...                                                          |

The fields V, P, SSRC, and length are defined in the RTP specification [2], the respective meaning being summarized below:

version (V): 2 bits
This field identifies the RTP version. The current version is 2.
padding (P): 1 bit
If set, the padding bit indicates that the packet contains additional padding octets at the end that are not part of the control information but are included in the length field. Always 0
Feedback message type (FMT): 5 bits
This field identifies the type of the FB message and is interpreted relative to the type (transport layer, payload- specific, or application layer feedback). The values for each of the three feedback types are defined in the respective sections below. Always 15, application layer feedback message. RFC 4585 section 6.4
Payload type (PT): 8 bits
This is the RTCP packet type that identifies the packet as being an RTCP FB message. Always PSFB | 206 | Payload-specific FB message. RFC 4585 section 6.4.
Length: (16 bits)
The length of this packet in 32-bit words minus one, including the header and any padding. This is in line with the definition of the length field used in RTCP sender and receiver reports [3]. RFC 4585 section 6.4.
SSRC of packet sender: (32 bits)
The synchronization source identifier for the originator of this packet. RFC 4585 section 6.4.
SSRC of media source:(32 bits)
Always 0.
Unique identifier (32 bits)
Always ‘R’ ‘E’ ‘M’ ‘B’
Num SSRC (8 bits):
Number of SSRCs in this message
BR Exp (6 bits):
The exponential scaling of the mantissa for the maximum total media bit rate value, ignoring all packet overhead. The value is an unsigned integer [0..63]. RFC 5104 section
BR Mantissa (18 bits):
The mantissa of the maximum total media bit rate (ignoring all packet overhead) that the sender of the REMB estimates. The BR is the estimate of the traveled path for the SSRCs reported in this message. The value is an unsigned integer in number of bits per second
SSRC feedback (32 bits)
Consists of one or more SSRC entries which this feedback message applies to.

Appendix B. Change log

Appendix B.1. Version -00 to -01

Authors' Addresses

Henrik Lundin Google Kungsbron 2 Stockholm, 11122 Sweden
Stefan Holmer Google Kungsbron 2 Stockholm, 11122 Sweden EMail:
Harald Alvestrand editor Google Kungsbron 2 Stockholm, 11122 Sweden EMail: