Status of This Memo

This is an Internet Standards Track document.¶

This document is a product of the Internet Engineering Task Force (IETF). It represents the consensus of the IETF community. It has received public review and has been approved for publication by the Internet Engineering Steering Group (IESG). Further information on Internet Standards is available in Section 2 of RFC 7841.¶

Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at https://www.rfc-editor.org/info/rfc8849.¶

Copyright Notice

Copyright (c) 2021 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 (https://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.¶

Table of Contents

1. Introduction

Telepresence systems can send and receive multiple media streams. The CLUE Framework [RFC8845] defines Media Captures (MCs) as a source of Media, from one or more Capture Devices. A Media Capture may also be constructed from other Media streams. A middlebox can express conceptual Media Captures that it constructs from Media streams it receives. A Multiple Content Capture (MCC) is a special Media Capture composed of multiple Media Captures.¶

SIP Offer/Answer [RFC3264] uses SDP [RFC4566] to describe the RTP media streams [RFC3550]. Each RTP stream has a unique Synchronization Source (SSRC) within its RTP session. The content of the RTP stream is created by an encoder in the endpoint. This may be an original content from a camera or a content created by an intermediary device like a Multipoint Control Unit (MCU).¶

This document makes recommendations for the CLUE architecture about how RTP and RTP Control Protocol (RTCP) streams should be encoded and transmitted and how their relation to CLUE Media Captures should be communicated. The proposed solution supports multiple RTP topologies [RFC7667].¶

With regards to the media (audio, video, and timed text), systems that support CLUE use RTP for the media, SDP for codec and media transport negotiation (CLUE individual encodings), and the CLUE protocol for Media Capture description and selection. In order to associate the media in the different protocols, there are three mappings that need to be specified:¶

1. CLUE individual encodings to SDP¶
2. RTP streams to SDP (this is not a CLUE-specific mapping)¶
3. RTP streams to MC to map the received RTP stream to the current MC in the MCC.¶

2. Terminology

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶

Definitions from the CLUE Framework (see Section 3 of [RFC8845]) are used by this document as well.¶

3. RTP Topologies for CLUE

The typical RTP topologies used by CLUE telepresence systems specify different behaviors for RTP and RTCP distribution. A number of RTP topologies are described in [RFC7667]. For CLUE telepresence, the relevant topologies include Point-to-Point, as well as Media-Mixing Mixers, Media-Switching Mixers, and Selective Forwarding Middleboxes.¶

In the Point-to-Point topology, one peer communicates directly with a single peer over unicast. There can be one or more RTP sessions, each sent on a separate 5-tuple, that have a separate SSRC space, with each RTP session carrying multiple RTP streams identified by their SSRC. All SSRCs are recognized by the peers based on the information in the RTCP Source description (SDES) report that includes the Canonical Name (CNAME) and SSRC of the sent RTP streams. There are different Point-to-Point use cases as specified in the CLUE use case [RFC7205]. In some cases, a CLUE session that, at a high level, is Point-to-Point may nonetheless have an RTP stream that is best described by one of the mixer topologies. For example, a CLUE endpoint can produce composite or switched captures for use by a receiving system with fewer displays than the sender has cameras. The Media Capture may be described using an MCC.¶

For the media mixer topology [RFC7667], the peers communicate only with the mixer. The mixer provides mixed or composited media streams, using its own SSRC for the sent streams. If needed by the CLUE endpoint, the conference roster information including conference participants, endpoints, media, and media-id (SSRC) can be determined using the conference event package [RFC4575] element.¶

Media-Switching Mixers and Selective Forwarding Middleboxes behave as described in [RFC7667].¶

4. Mapping CLUE Capture Encodings to RTP Streams

The different topologies described in Section 3 create different SSRC distribution models and RTP stream multiplexing points.¶

Most video conferencing systems today can separate multiple RTP sources by placing them into RTP sessions using the SDP description; the video conferencing application can also have some knowledge about the purpose of each RTP session. For example, video conferencing applications that have a primary video source and a slides video source can send each media source in a separate RTP session with a content attribute [RFC4796], enabling different application behavior for each received RTP media source. Demultiplexing is straightforward because each Media Capture is sent as a single RTP stream, with each RTP stream being sent in a separate RTP session, on a distinct UDP 5-tuple. This will also be true for mapping the RTP streams to Capture Encodings, if each Capture Encoding uses a separate RTP session and the consumer can identify it based on the receiving RTP port. In this case, SDP only needs to label the RTP session with an identifier that can be used to identify the Media Capture in the CLUE description. The SDP label attribute serves as this identifier.¶

Each Capture Encoding MUST be sent as a separate RTP stream. CLUE endpoints MUST support sending each such RTP stream in a separate RTP session signaled by an SDP "m=" line. They MAY also support sending some or all of the RTP streams in a single RTP session, using the mechanism described in [RFC8843] to relate RTP streams to SDP "m=" lines.¶

MCCs bring another mapping issue, in that an MCC represents multiple Media Captures that can be sent as part of the MCC if configured by the consumer. When receiving an RTP stream that is mapped to the MCC, the consumer needs to know which original MC it is in order to get the MC parameters from the advertisement. If a consumer requested a MCC, the original MC does not have a Capture Encoding, so it cannot be associated with an "m=" line using a label as described in "CLUE Signaling" [RFC8848]. It is important, for example, to get correct scaling information for the original MC, which may be different for the various MCs that are contributing to the MCC.¶

5. MCC Constituent CaptureID Definition

For an MCC that can represent multiple switched MCs, there is a need to know which MC is represented in the current RTP stream at any given time. This requires a mapping from the SSRC of the RTP stream conveying a particular MCC to the constituent MC. In order to address this mapping, this document defines an RTP header extension and SDES item that includes the captureID of the original MC, allowing the consumer to use the MC's original source attributes like the spatial information.¶

This mapping temporarily associates the SSRC of the RTP stream conveying a particular MCC with the captureID of the single original MC that is currently switched into the MCC. This mapping cannot be used for a composed case where more than one original MC is composed into the MCC simultaneously.¶

If there is only one MC in the MCC, then the media provider MUST send the captureID of the current constituent MC in the RTP header extension and as an RTCP CaptureID SDES item. When the media provider switches the MC it sends within an MCC, it MUST send the captureID value for the MC that just switched into the MCC in an RTP header extension and as an RTCP CaptureID SDES item as specified in [RFC7941].¶

If there is more than one MC composed into the MCC, then the media provider MUST NOT send any of the MCs' captureIDs using this mechanism. However, if an MCC is sending Contributing Source (CSRC) information in the RTP header for a composed capture, it MAY send the captureID values in the RTCP SDES packets giving source information for the SSRC values sent as CSRCs.¶

If the media provider sends the captureID of a single MC switched into an MCC, then later sends one composed stream of multiple MCs in the same MCC, it MUST send the special value "-", a single-dash character, as the captureID RTP header extension and RTCP CaptureID SDES item. The single-dash character indicates there is no applicable value for the MCC constituent CaptureID. The media consumer interprets this as meaning that any previous CaptureID value associated with this SSRC no longer applies. As [RFC8846] defines the captureID syntax as "xs:ID", the single-dash character is not a legal captureID value, so there is no possibility of confusing it with an actual captureID.¶

 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|   CaptId=14   |     length    | CaptureID                     |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|   ....        |
+-+-+-+-+-+-+-+-+

6. Examples

In this partial advertisement, the media provider advertises a composed capture VC7 made of a big picture representing the current speaker (VC3) and two picture-in-picture boxes representing the previous speakers (the previous one -- VC5 -- and the oldest one -- VC6).¶

  <ns2:mediaCapture
     xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
     xsi:type="ns2:videoCaptureType" captureID="VC7"
       mediaType="video">
         <ns2:captureSceneIDREF>CS1</ns2:captureSceneIDREF>
         <ns2:nonSpatiallyDefinable>true</ns2:nonSpatiallyDefinable>
         <ns2:content>
               <ns2:captureIDREF>VC3</ns2:captureIDREF>
               <ns2:captureIDREF>VC5</ns2:captureIDREF>
               <ns2:captureIDREF>VC6</ns2:captureIDREF>
         </ns2:content>
                 <ns2:maxCaptures>3</ns2:maxCaptures>
           <ns2:allowSubsetChoice>false</ns2:allowSubsetChoice>
         <ns2:description lang="en">big picture of the current
           speaker pips about previous speakers</ns2:description>
           <ns2:priority>1</ns2:priority>
           <ns2:lang>it</ns2:lang>
           <ns2:mobility>static</ns2:mobility>
           <ns2:view>individual</ns2:view>
       </ns2:mediaCapture>

In this case, the media provider will send capture IDs VC3, VC5, or VC6 as an RTP header extension and RTCP SDES message for the RTP stream associated with the MC.¶

Note that this is part of the full advertisement message example from the CLUE data model example [RFC8846] and is not a valid XML document.¶

7. Communication Security

CLUE endpoints MUST support RTP/SAVPF profiles and the Secure Real-time Transport Protocol (SRTP) [RFC3711]. CLUE endpoints MUST support DTLS [RFC6347] and DTLS-SRTP [RFC5763] [RFC5764] for SRTP keying.¶

All media channels SHOULD be secure via SRTP and the RTP/SAVPF profile unless the RTP media and its associated RTCP are secure by other means (see [RFC7201] and [RFC7202]).¶

All CLUE implementations MUST support DTLS 1.2 with the TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256 cipher suite and the P-256 curve [FIPS186]. The DTLS-SRTP protection profile SRTP_AES128_CM_HMAC_SHA1_80 MUST be supported for SRTP. Implementations MUST favor cipher suites that support Perfect Forward Secrecy (PFS) over non-PFS cipher suites and SHOULD favor Authenticated Encryption with Associated Data (AEAD) over non-AEAD cipher suites. Encrypted SRTP Header extensions [RFC6904] MUST be supported.¶

Implementations SHOULD implement DTLS 1.2 with the TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256 cipher suite. Implementations MUST favor cipher suites that support Perfect Forward Secrecy (PFS) over non- PFS cipher suites and SHOULD favor Authenticated Encryption with Associated Data (AEAD) over non-AEAD cipher suites.¶

NULL Protection profiles MUST NOT be used for RTP or RTCP.¶

CLUE endpoints MUST generate short-term persistent RTCP CNAMEs, as specified in [RFC7022], and thus can't be used for long-term tracking of the users.¶

8. IANA Considerations

This document defines a new extension URI in the "RTP SDES Compact Header Extensions" subregistry of the "Real-Time Transport Protocol (RTP) Parameters" registry, according to the following data:¶

Extension URI:

urn:ietf:params:rtp-hdrext:sdes:CaptId¶

Description:

CLUE CaptId¶

Contact:

Roni Even <ron.even.tlv@gmail.com>¶

Reference:

RFC 8849¶

The IANA has registered one new RTCP SDES items in the "RTCP SDES Item Types" registry, as follows:¶

9. Security Considerations

The security considerations of the RTP specification, the RTP/SAVPF profile, and the various RTP/RTCP extensions and RTP payload formats that form the complete protocol suite described in this memo apply. It is believed that there are no new security considerations resulting from the combination of these various protocol extensions.¶

The "Extended Secure RTP Profile for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/SAVPF)" document [RFC5124] provides the handling of fundamental issues by offering confidentiality, integrity, and partial source authentication. A mandatory-to-implement and use media security solution is created by combining this secured RTP profile and DTLS-SRTP keying [RFC5764] as defined in the communication security section of this memo (Section 7).¶

RTCP packets convey a CNAME identifier that is used to associate RTP packet streams that need to be synchronized across related RTP sessions. Inappropriate choice of CNAME values can be a privacy concern, since long-term persistent CNAME identifiers can be used to track users across multiple calls. The communication security section of this memo (Section 7) mandates the generation of short- term persistent RTCP CNAMEs, as specified in [RFC7022], so they can't be used for long-term tracking of the users.¶

Some potential denial-of-service attacks exist if the RTCP reporting interval is configured to an inappropriate value. This could be done by configuring the RTCP bandwidth fraction to an excessively large or small value using the SDP "b=RR:" or "b=RS:" lines [RFC3556], or some similar mechanism, or by choosing an excessively large or small value for the RTP/AVPF minimal receiver report interval (if using SDP, this is the "a=rtcp-fb:... trr-int" parameter) [RFC4585]. The risks are as follows:¶

1. The RTCP bandwidth could be configured to make the regular reporting interval so large that effective congestion control cannot be maintained, potentially leading to denial of service due to congestion caused by the media traffic;¶
2. The RTCP interval could be configured to a very small value, causing endpoints to generate high-rate RTCP traffic, which potentially leads to denial of service due to the non-congestion-controlled RTCP traffic; and¶
3. RTCP parameters could be configured differently for each endpoint, with some of the endpoints using a large reporting interval and some using a smaller interval, leading to denial of service due to premature participant timeouts, which are due to mismatched timeout periods that are based on the reporting interval (this is a particular concern if endpoints use a small but non-zero value for the RTP/AVPF minimal receiver report interval (trr-int) [RFC4585], as discussed in [RFC8108]).¶

Premature participant timeout can be avoided by using the fixed (non- reduced) minimum interval when calculating the participant timeout [RFC8108]. To address the other concerns, endpoints SHOULD ignore parameters that configure the RTCP reporting interval to be significantly longer than the default five-second interval specified in [RFC3550] (unless the media data rate is so low that the longer reporting interval roughly corresponds to 5% of the media data rate) or that configure the RTCP reporting interval small enough that the RTCP bandwidth would exceed the media bandwidth.¶

The guidelines in [RFC6562] apply when using variable bit rate (VBR) audio codecs such as Opus.¶

Encryption of the header extensions is RECOMMENDED, unless there are known reasons, like RTP middleboxes performing voice-activity-based source selection or third-party monitoring that will greatly benefit from the information, and this has been expressed using API or signaling. If further evidence is produced to show that information leakage is significant from audio level indications, then the use of encryption needs to be mandated at that time.¶

In multi-party communication scenarios using RTP middleboxes, the middleboxes are REQUIRED, by this protocol, to not weaken the sessions' security. The middlebox SHOULD maintain confidentiality, maintain integrity, and perform source authentication. The middlebox MAY perform checks that prevent any endpoint participating in a conference to impersonate another. Some additional security considerations regarding multi-party topologies can be found in [RFC7667].¶

The CaptureID is created as part of the CLUE protocol. The CaptId SDES item is used to convey the same CaptureID value in the SDES item. When sending the SDES item, the security considerations specified in Section 6 of [RFC7941] and in the communication security section of this memo (see Section 7) are applicable. Note that since the CaptureID is also carried in CLUE protocol messages, it is RECOMMENDED that this SDES item use at least similar protection profiles as the CLUE protocol messages carried in the CLUE data channel.¶

Acknowledgments

The authors would like to thank Allyn Romanow and Paul Witty for contributing text to this work. Magnus Westerlund helped draft the security section.¶

Authors' Addresses