One document matched: draft-ietf-avt-tfrc-profile-10.txt
Differences from draft-ietf-avt-tfrc-profile-09.txt
Internet Engineering Task Force AVT WG
INTERNET-DRAFT Ladan Gharai
Intended status: Standards Track 22 July 2007
draft-ietf-avt-tfrc-profile-10.txt
Expires: January 2008
RTP with TCP Friendly Rate Control
Status of this Memo
By submitting this Internet-Draft, each author represents that any
applicable patent or other IPR claims of which he or she is aware
have been or will be disclosed, and any of which he or she becomes
aware will be disclosed, in accordance with Section 6 of BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that other
groups may also distribute working documents as Internet-Drafts.
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."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html.
Copyright Notice
Copyright (C) The IETF Trust (2007).
Abstract
This memo specifies how the TCP Friendly Rate Control (TFRC) of RTP
flows can be supported using the RTP/AVPF profile and the general RTP
header extension mechanism. AVPF feedback packets and RTP header
extensions are defined to support the exchange of control information
between RTP TFRC senders and receivers. TFRC is an equation-based
congestion control scheme for unicast flows operating in a best
effort Internet environment.
Gharai [Page 1]
INTERNET-DRAFT Expires: January 2008 July 2007
1. Introduction
[Note to RFC Editor: All references to RFC XXXX are to be replaced
with the RFC number of this memo, when published]
This memo specifies how the TCP Friendly Rate Control (TFRC) of RTP
flows can be supported using the RTP/AVPF [RFC3550][RFC4585] profile
and RTP header extensions, by defining a new header extension and
AVPF feedback packet, and related parameters. Any of the AVPF based
RTP profiles, such as SAVPF, can be used to support TFRC RTP flows.
TFRC is an equation-based congestion control scheme for unicast flows
operating in a best effort Internet environment and competing with
TCP traffic. TFRC computes a TCP-friendly data rate based on current
network conditions, as represented by the latest round trip time and
packet loss calculations. The complete TFRC mechanism is described in
detail in [RFC3448bis].
To calculate a TCP-friendly data rate and keep track of round trip
times and packet losses, TFRC senders and receivers rely on
exchanging specific information between each other, i.e: the sender
provides the receiver with the latest updates to round trip time
calculations, while the receiver provides feedback needed to compute
round trip times and on packet losses. This memo defines how this
information can be exchanged between TFRC senders and receiver with
RTP header extensions and an AVPF feedback packet.
2. Terminology
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].
3. Relation to the Datagram Congestion Control Protocol
The TFRC congestion control mechanism is one of a set of congestion
control methods provided by the Datagram Congestion Control Protocol
(DCCP) [RFC4340] In this section we detail the pros and cons of using
TFRC with RTP versus DCCP.
DCCP is a minimal general purpose transport-layer protocol with
unreliable yet congestion controlled packet delivery semantics and
reliable connection setup and teardown [RFC4336]. DCCP currently
supports both TFRC [RFC4342] and TCP-like [RFC4341] congestion
control, and the protocol is structured to support new congestion
control mechanisms defined in the future. A DCCP mapping for RTP has
been standardized for media applications [RFCxRTP]. In addition DCCP
supports a host of other features, such as: use of Explicit
Gharai [Page 2]
INTERNET-DRAFT Expires: January 2008 July 2007
Congestion Notification (ECN) and the ECN Nonce, flexible options
processing, reliable option negotiation, DTLS [ID.DTLS], Path Maximum
Transfer Unit (PMTU) and Service Codes. Naturally an application
using RTP/DCCP as its transport protocol will benefit from the
protocol features supported by DCCP.
However there are a number of benefits to be gained by the
development and standardization of the use RTP with TFRC:
o Media applications lacking congestion control can incorporate
congestion controlled transport without delay by using RTP with
TFRC. Widespread deployment of the DCCP protocol is not currently
in place.
o Use of RTP with TFRC is not contingent on any OS level changes
and can be quickly deployed, because RTP is implemented at the
application layer.
o RTP/UDP flows face the same restrictions in firewall traversal
as do UDP flows and do not require NATs and firewall
modifications. DCCP flows, on the other hand, do require NAT
and firewall modifications, however once these modifications are
in place, they can result in easier NAT and firewall traversal
for RTP/DCCP flows in the future.
o Use of RTP with TFRC with various media applications will give
researchers, implementors and developers a better understanding
of the intricate relationship between media quality and
equation-based congestion control. Hopefully this experience
with congestion control and TFRC will ease the migration of media
applications to DCCP once DCCP is deployed.
Using the AVPF/RTP profile and header extension to support TFRC
provides an immediate means for congestion control in media streams,
in the time until DCCP is deployed.
Additionally, there are also a number of technical differences as to
how (and which) congestion control information is exchanged between
DCCP with CCID3 and RTP:
o Using header extensions the RTP TFRC sender transmits a
send timestamp to the RTP TFRC receiver with every data packet.
In addition to congestion control the send timestamp can be
used by the receiver for jitter calculations.
In contrast DCCP with CCID3 transmits a quad round trip
counter to the receiver.
Gharai [Page 3]
INTERNET-DRAFT Expires: January 2008 July 2007
o The RTP TFRC receiver only provides the RTP TFRC sender
with the loss event rate as computed by the receiver.
In contrast DCCP with CCID3, provides 2 other options for the
transport of loss event rate. A sender may choose to receive
loss intervals or an Ack Vector. These two options provide the
sender with the necessary information to compute the loss event
rate.
o Sequence number: DCCP supports a 48 bit and a 24 bit sequence
number, whereas RTP only supports a 16 bit sequence number. While
this makes RTP susceptible to data injection attacks, it can be
avoided by using the SRTP [RFC3711] profile.
4. The TFRC Information Exchange Loop
TFRC depends on the exchange of congestion control information
between a sender and receiver. In this section we reiterate which
items are exchanged between a TFRC sender and receiver as discussed
in [RFC3448bis]. We note how RTP can accommodates these exchanges.
4.1. Data Packets
As stated in [RFC3448bis] a TFRC sender transmits the following
information in each data packet to the receiver:
o A sequence number, incremented by one for each data packet
transmitted.
o A timestamp indicating the packet send time and the sender's
current estimate of the round-trip time, RTT. This information
is then used by the receiver to compute the TFRC loss intervals.
- or -
A course-grained timestamp incrementing every quarter of a
round trip time, which is then used to determine the TFRC loss
intervals.
The standard RTP sequence number suffices for the functionality
provided by TFRC. A RTP header extension [hdrtxt] is used to
transmit the send timestamp and RTT. This extension is defined in
Section 5.
4.2. Feedback Packets
As stated in [RFC3448bis] a TFRC receiver provides the following
feedback to the sender at least once per RTT or per data packet
Gharai [Page 4]
INTERNET-DRAFT Expires: January 2008 July 2007
received (which ever time interval is larger):
o The send timestamp of the last data packet received, t_i.
o The amount of time elapsed between the receipt of the last
data packet at the receiver, and the generation of this feedback
report, t_delay. This is used by the sender for RTT computations.
o The rate at which the receiver estimates that data was received
since the last feedback report was sent, x_recv.
o The receiver's current estimate of the loss event rate, p, a real
value between 0 and 1.0.
To accommodate the feedback of these values a new AVPF transport
layer feedback message is defined, as detailed in Section 6. The
timing interval between the feedback packets is discussed in Section
7.
5. The Header Extension
The form of the extension block is depicted in Figure 1. The length
field for the extension takes the value 6 to indicate that the
payload is 7 bytes. Two header extension fields are defined and used
as follows:
Send timestamp (t_i): 32 bits
The timestamp indicating when the packet is sent. This timestamp
is measured in microseconds and is used for round trip time
calculations.
Round trip time (RTT): 24 bits
The round trip time as measured by the RTP TFRC sender in
microseconds.
Gharai [Page 5]
INTERNET-DRAFT Expires: January 2008 July 2007
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 | 0xDE | length=2 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ID | len=6 | RTT |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| send timestamp (t_i) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 1: RTP Sender Header Extension Block
6. TFRC-FB: A New AVPF Transport Layer Feedback Message
A new transport layer AVPF feedback message is defined to support
feedback from the receivers: TFRC-FB. Figure 2 depicts the both the
common packet format (the first 12 octets) and the feedback control
information (FCI) for the TFRC-FB packet.
We note that the TFRC related feedback, is specific to one media
stream sender, therefore all messages in the compound RTCP packet
MUST share the same media source SSRC. In the case where a sender is
sending multiple media streams to a receiver, each media flow will be
allocated its own AVPF feedback flow.
We define four FCI fields for the TFRC-FB message as follows:
Send timestamp (t_i): 32 bits
The send timestamp of the last data packet received by the
RTP TFRC receiver, t_i, in microseconds.
Delay (t_delay): 32 bits
The amount of time elapsed between the receipt of the last data
packet at the RTP TFRC receiver, and the generation of this
feedback report in microseconds. This is used by the RTP TFRC
sender for RTT computations.
Data rate (X_recv): 32 bits
The rate at which the receiver estimates that data was received
since the last feedback report was sent in bytes per second. X_recv
is computed per [RFC3448bis].
Loss event rate (p): 32 bits
The receiver's current estimate of the loss event rate, p,
expressed as a fixed point number with the binary point at the
left edge of the field. (That is equivalent to taking the integer
part after multiplying the loss event rate by 2^32.) The value of
Gharai [Page 6]
INTERNET-DRAFT Expires: January 2008 July 2007
the loss event rate is computed per [RFC3448bis] Section 5.
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=2 | PT=RTPFB | length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| SSRC of packet sender |
+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
| SSRC of media source |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| send timestamp (t_i) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| delay (t_delay) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| data rate at the receiver (x_recv) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| loss event rate (p) |
+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
Figure 2: The AVPF TFRC-FB RTCP Transport Layer Feedback Message
7. RTCP Transmission Intervals and Bandwidth Requirements
When using TFRC rate controlled RTP, the RTCP transmission intervals
must be set according to the requirements of the TFRC algorithm. TFRC
requires a receiver to generate a feedback ack packet at least once
per RTT or per packet received (based on the larger time interval).
These requirements are to ensure timely reaction to congestion.
The TFRC requirements of receiving feedback once per RTT can at times
conflict with the AVP RTCP bandwidth constraints, particularly at
small RTTs of 20 ms or less. Assuming only one TFRC-FB report per
RTCP compound packet, Table 1 lists the RTCP bandwidths at RTTs of 2,
5, 10 and 20 ms and the minimum corresponding RTP data rates, where
RTCP(X) <= (0.05)*RTP(X) is true. For example, according to Table
1, a TFRC RTP flow of less than 3.2 Mbps and a RTT of 5 ms, can not
comply with the 5% RTCP bandwidth constraints (Table 1 assumes each
RTCP packet is 100 bytes). RTP flows facing such circumstances
should take into account the additional RTCP bandwidth needed when
signaling their bandwidth information in SDP [RFC4566].
Based on initial assumptions on round trip time if more than the
recommended 5% is needed for RTCP bandwidth, the applications SHOULD
use the SDP bandwidth modifiers RS and RR [RFC3556] to signal the
amount of RTCP bandwidth needed. If the round trip time assumptions
change after the RTP flows start running, the application MAY
Gharai [Page 7]
INTERNET-DRAFT Expires: January 2008 July 2007
recalculate the amount of RTCP bandwidth needed and re-signal this
new value using its signaling protocol of choice.
RTT RTCP(X) RTP(X)
+------------------------------+
| 20 ms | 40 kbps | 0.8 Mbps |
| 10 ms | 80 kbps | 1.6 Mbps |
| 5 ms | 160 kbps | 3.2 Mbps |
| 2 ms | 400 kbps | 8.0 Mbps |
+------------------------------+
Table 1: RTCP bandwidth for TFRC flows with corresponding
RTTs of 20, 10, 5 and 2 ms. Assuming, 100 byte RTCP packets
and one RTCP packet per RTT.
Additionally, to support the transmission of a feedback packet once per
RTT, the AVPF T_rr_interval variable MUST NOT be set to a value larger
than the current round trip time, RTT, as this would prevent generating
feedback packets at least once per RTT (see RFC 4585, Section 3.4,m).
8. SDP Usage
RTP flows using TFRC congestion control MUST signal their use of the
AVPF profile and RTCP feedback packets, the round trip time (RTT) and
send timestamp extension, and MAY also signal an initial RTCP
bandwidth usage:
v=0
o=alice 2890844526 2890844526 IN IP4 host.example.com
s=congestion control with TFRC
c=IN IP4 host.example.com
m=video 5400 RTP/AVPF 112
a=rtpmap:112 H261/90000
a=extmap:4 urn:ietf:params:rtp-hdtext:rtt-sendts
a=rtcp-fb * tfrc
b=AS:400
b=RS:800
b=RR:4000
8.1. Usage with the SDP Offer/Answer Model
TBC
Gharai [Page 8]
INTERNET-DRAFT Expires: January 2008 July 2007
9. IANA Considerations
In this section we detail IANA registry values that need to be
registered. Two new values must be reistered for the AVPF profile:
The new RTP/AVPF feedback packet, TFRC-FB. The following format
(FMT) values must be registerd in the FMT sub-registry of the RTPFB
payload type:
Value name: TFRC-FB
Long name: TFRC feedback
Value: 5
Reference: RFC XXXX
The new rtcp-fd-id "tfrc" must be registered with the "rtcp-fb"
attribute registry:
Value name: tfrc
Long name: TFRC Feedback
Reference: RFC XXXX
For the new header extension, the name rtt-sendts must be registered
into the rtp-hdrext section of the urn:ietf: namespace, referring to
RFC XXXX.
10. Security Considerations
This memo defines how to use the RTP AVPF profile and the general RTP
header extensions to support TFRC congestion control. Therefore RTP
packets using these mechanisms are subject to the security
considerations discussed in the RTP specification [RFC3550], the
RTP/AVPF profile specification [RFC4585] and the general header
extensions mechanism [hdrtxt]. Combining these mechanisms does not
pose any additional security implications. Applications requiring
authentication and integrity protection, or applictions operating in
environments that must strictly adhere to the TFRC send rate (and
fear manipulation of the feedback messsages) can use the SAVPF
[SAVPF] profile.
11. Acknowledgments
This memo is based upon work supported by the U.S. National Science
Foundation (NSF) under Grant No. 0334182. Any opinions, findings and
conclusions or recommendations expressed in this material are those
of the authors and do not necessarily reflect the views of NSF.
Gharai [Page 9]
INTERNET-DRAFT Expires: January 2008 July 2007
12. Author's Address
Ladan Gharai <ladan@gharai.org>
Normative References
[RFC2119] S. Bradner, "Key words for use in RFCs to Indicate
Requirement Levels", Internet Engineering Task Force,
RFC 2119, March 1997.
[RFC3550] H. Schulzrinne, S. Casner, R. Frederick and V. Jacobson,
"RTP: A Transport Protocol for Real-Time Applications",
Internet Engineering Task Force, RFC 3550 (STD0064), July
2003.
[RFC3556] S. Casner, "Session Description Protocol (SDP) Bandwidth
Modifiers for RTP Control Protocol (RTCP) Bandwidth", RFC
3556, July 2003.
[RFC3711] M. Baugher, D. McGrew, M. Naslund, E. Carrara, K. Norrman,
"The Secure Real-time Transport Protocol", RFC 3711, March
2004.
[RFC4566] M. Handley, V. Jacobson and C. Perkins, "SDP: Session
Description Protocol", RFC 4566, July 2006.
[RFC4585] J. Ott, S. Wenger, A. Sato, C. Burmeister and J. Ray,
"Extended RTP Profile for RTCP-based Feedback (RTP/AVPF)",
RFC 4585, July 2006.
[RFC3448bis] M. Handley, S. Floyed, J. Padhye and J. widmer,
"TCP Friendly Rate Control (TRFC): Protocol Specification",
draft-ietf-dccp-rfc3448bis-01.txt, March 2007.
[hdrext] D. Singer, "A general mechanism for RTP Header Extensions",
IETF Work in Progress, draft-ietf-avt-rtp-hdrext-12.txt.
[SAVPF] J. Ott, E. Carrara, "Extended Secure RTP Profile for
RTCP-based Feedback (RTP/SAVPF)," IETF Work in Progress,
draft-ietf-avt-profile-savpf-10.txt.
Informative References
[RFC3711] M. Baugher, D. McGrew, M. Naslund, E. Carrara, K. Norrman,
"The Secure Real-time Transport Protocol", RFC 3711, March
2004.
Gharai [Page 10]
INTERNET-DRAFT Expires: January 2008 July 2007
[RFC4336] Floyd, S., Handley, M., and E. Kohler, "Problem Statement
for the Datagram Congestion Control Protocol (DCCP)", RFC
4336, March 2006.
[RFC4340] Kohler, E., Handley, M., and S. Floyd, "Datagram Congestion
Control Protocol (DCCP)", RFC 4340, March 2006.
[RFC4341] Floyd, S. and E. Kohler, "Profile for Datagram Congestion
Control Protocol (DCCP) Congestion Control ID 2: TCP-like
Congestion Control", RFC 4341, March 2006.
[RFC4342] Floyd, S., Kohler, E., and J. Padhye, "Profile for Datagram
Congestion Control Protocol (DCCP) Congestion Control ID 3:
TCP-Friendly Rate Control (TFRC)", RFC 4342, March 2006.
[RFCxRTP] C. Perkins, "RTP and the Datagram Congestion Control
Protocol (DCCP)", IETF Work in Progress, draft-ietf-dccp-
rtp-07.txt.
[ID.DTLS] T.Phelan, "Datagram Transport Layer Security (DTLS)
over the Datagram Congestion Control Protocol (DCCP)", IETF
Work in Progress, draft-ietf-dccp-dtls-00.txt.
13. IPR Notice
The IETF takes no position regarding the validity or scope of any
Intellectual Property Rights or other rights that might be claimed to
pertain to the implementation or use of the technology described in
this document or the extent to which any license under such rights
might or might not be available; nor does it represent that it has
made any independent effort to identify any such rights. Information
on the procedures with respect to rights in RFC documents can be
found in BCP 78 and BCP 79.
Copies of IPR disclosures made to the IETF Secretariat and any
assurances of licenses to be made available, or the result of an
attempt made to obtain a general license or permission for the use of
such proprietary rights by implementers or users of this
specification can be obtained from the IETF on-line IPR repository at
http://www.ietf.org/ipr.
The IETF invites any interested party to bring to its attention any
copyrights, patents or patent applications, or other proprietary
rights that may cover technology that may be required to implement
this standard. Please address the information to the IETF at ietf-
ipr@ietf.org.
Gharai [Page 11]
INTERNET-DRAFT Expires: January 2008 July 2007
14. Full Copyright Statement
Copyright (C) The IETF Trust (2007).
This document is subject to the rights, licenses and restrictions
contained in BCP 78, and except as set forth therein, the authors
retain all their rights.
This document and the information contained herein are provided on an
"AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Gharai [Page 12]
| PAFTECH AB 2003-2026 | 2026-04-24 06:56:26 |