| AVT | J. Lennox, Ed. |
| Internet-Draft | Vidyo |
| Intended status: Standards Track | E. Ivov |
| Expires: December 04, 2011 | Jitsi |
| E. Marocco | |
| Telecom Italia | |
| June 02, 2011 |
A Real-Time Transport Protocol (RTP) Header Extension for Client-to-Mixer Audio Level Indication
draft-ietf-avtext-client-to-mixer-audio-level-02
This document defines a mechanism by which packets of Real-Time Transport Protocol (RTP) audio streams can indicate, in an RTP header extension, the audio level of the audio sample carried in the RTP packet. In large conferences, this can reduce the load on an audio mixer or other middlebox which wants to forward only a few of the loudest audio streams, without requiring it to decode and measure every stream that is received.
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 http://datatracker.ietf.org/drafts/current/.
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 December 04, 2011.
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 (http://trustee.ietf.org/license-info) 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.
This document may contain material from IETF Documents or IETF Contributions published or made publicly available before November 10, 2008. The person(s) controlling the copyright in some of this material may not have granted the IETF Trust the right to allow modifications of such material outside the IETF Standards Process. Without obtaining an adequate license from the person(s) controlling the copyright in such materials, this document may not be modified outside the IETF Standards Process, and derivative works of it may not be created outside the IETF Standards Process, except to format it for publication as an RFC or to translate it into languages other than English.
In a centralized Real-Time Transport Protocol (RTP) [RFC3550] audio conference, an audio mixer or forwarder receives audio streams from many or all of the conference participants. It then selectively forwards some of them to other participants in the conference. In large conferences, it is possible that such a server might be receiving a large number of streams, of which only a few should be forwarded to the other conference participants.
In such a scenario, in order to pick the audio streams to forward, a centralized server needs to decode, measure audio levels, and possibly perform voice activity detection on audio data from a large number of streams. The need for such processing limits the size or number of conferences such a server can support.
As an alternative, this document defines an RTP header extension [RFC5285] through which senders of audio packets can indicate the audio level of the packets' payload, reducing the processing load for a server.
The header extension in this draft is different than, but complementary with, the one defined in [I-D.ietf-avtext-mixer-to-client-audio-level], which defines a mechanism by which audio mixers can indicate to clients the levels of the contributing sources that made up the mixed audio.
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] and indicate requirement levels for compliant implementations.
The audio level header extension element carries the level of the audio in the RTP payload of the packet it is associated with, and also an indication as to whether voice activity has been detected in the packet. This information is carried in an RTP header extension element as defined by [RFC5285].
The payload of the audio level header extension element is as follows:
0 1
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ID | len=0 |V| level |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The length field takes the value 0 to indicate that 1 byte follows.
The two-byte header defined in RFC 5285 [RFC5285] may also be used.
The magnitude of the audio level is packed into the seven least significant bits of the single byte of the header extension, shown in Figure 1. The least significant bit of the audio level magnitude is packed into the least significant bit of the byte. The most significant bit of the byte is used as a separate flag bit "V", defined below.
The audio level is expressed in -dBov, with values from 0 to 127 representing 0 to -127 dBov. dBov is the level, in decibels, relative to the overload point of the system, i.e. the maximum-amplitude signal that can be handled by the system without clipping. (Note: Representation relative to the overload point of a system is particularly useful for digital implementations, since one does not need to know the relative calibration of the analog circuitry.) For example, in the case of u-law (audio/pcmu) audio [ITU.G711.1988], the 0 dBov reference would be a square wave with values +/- 8031. (This translates to 6.18 dBm0, relative to u-law's dBm0 definition in Table 6 of G.711.)
The audio level for digital silence, for example for a muted audio source, MAY be represented as 127 (-127 dBov), regardless of the dynamic range of the encoded audio format.
Implementations MAY choose to measure audio levels prior to encoding them in the payload carried in the RTP payload, e.g. on raw linear PCM input.
The audio level header extension only carries the level of the audio in the RTP payload of the packet it is associated with, with no long-term averaging or smoothing applied.
To simplify implementation of the encoding procedures described here, the reference implementation section in [I-D.ietf-avtext-mixer-to-client-audio-level] provides a sample Java implementation of an audio level calculator that helps obtain such values from raw linear PCM audio samples.
In addition, a flag bit (labeled V) indicates whether the encoder believes the audio packet contains voice activity (1) or does not (0). The voice activity detection algorithm is unspecified and left implementation-specific.
When this header extension is used with RTP data sent using the RTP Payload for Redundant Audio Data [RFC2198], the header's data describes the contents of the primary encoding.
Note: This audio level is defined in the same manner as is audio noise level in the RTP Payload Comfort Noise specification [RFC3389]. In the comfort noice specification, the overall magnitude of the noise level in comfort noise is encoded into the first byte of the payload, with spectral information about the noise in subsequent bytes. This specification's audio level parameter is defined so as to be identical to the comfort noise payload's noise-level byte.
The URI for declaring this header extension in an extmap attribute is "urn:ietf:params:rtp-hdrext:ssrc-audio-level". There is no additional setup information needed for this extension (i.e. no extensionattributes).
Mixers and forwarders generally should not base audio forwarding decisions directly on packet-by-packet audio level information, but rather should apply some analysis of the audio levels and trends. This general rule applies whether audio levels are provided by endpoints (as defined in this document), or are calculated at a server, as would be done in the absence of this information. This section discusses several issues that mixers and forwarders may wish to take into account. (Note that this section provides design guidance only, and is not normative.)
First of all, audio levels should generally be measured over longer intervals than that of a single audio packet. In order to avoid false-positives for short bursts of sound (such as a cough or a dropped microphone), it is often useful to require that a participant's audio level be maintained for some period of time before considering it to be "real", i.e. some type of low-pass filter should be applied to the audio levels. Note, though, that such filtering must be balanced with the need to avoid clipping of the beginning of a speaker's speech.
Additionally, different participants may have their audio input set differently. It may be useful to apply some sort of automatic gain control to the audio levels. There are a number of possible approaches to acheiving this, e.g. by measuring peak audio levels, by average audio levels during speech, or by measuring background audio levels (average audio level levels during non-speech).
A malicious endpoint could choose to set the values in this header extension falsely, so as to falsely claim that audio or voice is or is not present. It is not clear what could be gained by falsely claiming that audio is not present, but an endpoint falsely claiming that audio is present could perform a denial-of-service attack on an audio conference, so as to send silence to suppress other conference members' audio. Thus, a device relying on audio level data from untrusted endpoints SHOULD periodically audit the level information transmitted, taking appropriate corrective action if endpoints appear to be sending incorrect data. (Note that as it is valid for an endpoint to choose to measure audio levels prior to encoding, some degree of discrepancy SHOULD be tolerated.)
In the Secure Real-Time Transport Protocol (SRTP) [RFC3711], RTP header extensions are authenticated but not encrypted. When this header extension is used, audio levels are therefore visible on a packet-by-packet basis to an attacker passively observing the audio stream. As discussed in [I-D.perkins-avt-srtp-vbr-audio], such an attacker might be able to infer information about the conversation, possibly with phoneme-level resolution. In scenarios where this is a concern, additional mechanisms SHOULD be used to protect the confidentiality of the header extension. One solution is header extension encryption [I-D.lennox-avtcore-srtp-encrypted-header-ext].
This document defines a new extension URI to the RTP Compact Header Extensions subregistry of the Real-Time Transport Protocol (RTP) Parameters registry, according to the following data:
Note to RFC Editor: please replace "RFC XXXX" with the number of this RFC.
| [RFC5285] | Singer, D. and H. Desineni, "A General Mechanism for RTP Header Extensions", RFC 5285, July 2008. |
| [RFC2198] | Perkins, C., Kouvelas, I., Hodson, O., Hardman, V., Handley, M., Bolot, J., Vega-Garcia, A. and S. Fosse-Parisis, "RTP Payload for Redundant Audio Data", RFC 2198, September 1997. |
| [RFC2119] | Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. |
| [RFC3550] | Schulzrinne, H., Casner, S., Frederick, R. and V. Jacobson, "RTP: A Transport Protocol for Real-Time Applications", STD 64, RFC 3550, July 2003. |
Note to the RFC-Editor: please remove this section prior to publication as an RFC.