Internet Engineering Task Force (IETF) J. Lennox Request for Comments: 8108 Vidyo Updates: 3550, 4585 M. Westerlund Category: Standards Track Ericsson ISSN: 2070-1721 Q. Wu Huawei C. Perkins University of Glasgow March 2017 Sending Multiple RTP Streams in a Single RTP Session Abstract This memo expands and clarifies the behavior of Real-time Transport Protocol (RTP) endpoints that use multiple synchronization sources (SSRCs). This occurs, for example, when an endpoint sends multiple RTP streams in a single RTP session. This memo updates RFC 3550 with regard to handling multiple SSRCs per endpoint in RTP sessions, with a particular focus on RTP Control Protocol (RTCP) behavior. It also updates RFC 4585 to change and clarify the calculation of the timeout of SSRCs and the inclusion of feedback messages. 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 http://www.rfc-editor.org/info/rfc8108. Lennox, et al. Standards Track [Page 1] RFC 8108 Multiple Media Streams in an RTP Session March 2017 Copyright Notice Copyright (c) 2017 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. Lennox, et al. Standards Track [Page 2] RFC 8108 Multiple Media Streams in an RTP Session March 2017 Table of Contents 1. Introduction ....................................................4 2. Terminology .....................................................4 3. Use Cases for Multi-Stream Endpoints ............................4 3.1. Endpoints with Multiple Capture Devices ....................4 3.2. Multiple Media Types in a Single RTP Session ...............5 3.3. Multiple Stream Mixers .....................................5 3.4. Multiple SSRCs for a Single Media Source ...................5 4. Use of RTP by Endpoints That Send Multiple Media Streams ........6 5. Use of RTCP by Endpoints That Send Multiple Media Streams .......6 5.1. RTCP Reporting Requirement .................................7 5.2. Initial Reporting Interval .................................7 5.3. Aggregation of Reports into Compound RTCP Packets ..........8 5.3.1. Maintaining AVG_RTCP_SIZE ...........................9 5.3.2. Scheduling RTCP when Aggregating Multiple SSRCs ....10 5.4. Use of RTP/AVPF or RTP/SAVPF Feedback .....................13 5.4.1. Choice of SSRC for Feedback Packets ................13 5.4.2. Scheduling an RTCP Feedback Packet .................14 6. Adding and Removing SSRCs ......................................15 6.1. Adding RTP Streams ........................................16 6.2. Removing RTP Streams ......................................16 7. RTCP Considerations for Streams with Disparate Rates ...........17 7.1. Timing Out SSRCs ..........................................19 7.1.1. Problems with the RTP/AVPF T_rr_interval Parameter ..........................................19 7.1.2. Avoiding Premature Timeout .........................20 7.1.3. Interoperability between RTP/AVP and RTP/AVPF ......21 7.1.4. Updated SSRC Timeout Rules .........................22 7.2. Tuning RTCP Transmissions .................................22 7.2.1. RTP/AVP and RTP/SAVP ...............................22 7.2.2. RTP/AVPF and RTP/SAVPF .............................24 8. Security Considerations ........................................25 9. References .....................................................26 9.1. Normative References ......................................26 9.2. Informative References ....................................26 Acknowledgments ...................................................29 Authors' Addresses ................................................29 Lennox, et al. Standards Track [Page 3] RFC 8108 Multiple Media Streams in an RTP Session March 2017 1. Introduction At the time the Real-Time Transport Protocol (RTP) [RFC3550] was originally designed, and for quite some time after, endpoints in RTP sessions typically only transmitted a single media source and, thus, used a single RTP stream and synchronization source (SSRC) per RTP session, where separate RTP sessions were typically used for each distinct media type. Recently, however, a number of scenarios have emerged in which endpoints wish to send multiple RTP streams, distinguished by distinct RTP synchronization source (SSRC) identifiers, in a single RTP session. These are outlined in Section 3. Although the initial design of RTP did consider such scenarios, the specification was not consistently written with such use cases in mind; thus, the specification is somewhat unclear in places. This memo updates [RFC3550] to clarify behavior in use cases where endpoints use multiple SSRCs. It also updates [RFC4585] to resolve problems with regard to timeout of inactive SSRCs and to clarify behavior around inclusion of feedback messages. 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 RFC 2119 [RFC2119] and indicate requirement levels for compliant implementations. 3. Use Cases for Multi-Stream Endpoints This section discusses several use cases that have motivated the development of endpoints that sends RTP data using multiple SSRCs in a single RTP session. 3.1. Endpoints with Multiple Capture Devices The most straightforward motivation for an endpoint to send multiple simultaneous RTP streams in a single RTP session is when an endpoint has multiple capture devices and, hence, can generate multiple media sources, of the same media type and characteristics. For example, telepresence systems of the type described by the CLUE Telepresence Framework [CLUE-FRAME] often have multiple cameras or microphones covering various areas of a room and, hence, send several RTP streams of each type within a single RTP session. Lennox, et al. Standards Track [Page 4] RFC 8108 Multiple Media Streams in an RTP Session March 2017 3.2. Multiple Media Types in a Single RTP Session Recent work has updated RTP [MULTI-RTP] and Session Description Protocol (SDP) [SDP-BUNDLE] to remove the historical assumption in RTP that media sources of different media types would always be sent on different RTP sessions. In this work, a single endpoint's audio and video RTP streams (for example) are instead sent in a single RTP session to reduce the number of transport-layer flows used. 3.3. Multiple Stream Mixers There are several RTP topologies that can involve a central device that itself generates multiple RTP streams in a session. An example is a mixer providing centralized compositing for a multi-capture scenario like that described in Section 3.1. In this case, the centralized node is behaving much like a multi-capturer endpoint, generating several similar and related sources. A more complex example is the selective forwarding middlebox, described in Section 3.7 of [RFC7667]. This is a middlebox that receives RTP streams from several endpoints and then selectively forwards modified versions of some RTP streams toward the other endpoints to which it is connected. For each connected endpoint, a separate media source appears in the session for every other source connected to the middlebox, "projected" from the original streams, but at any given time many of them can appear to be inactive (and thus are receivers, not senders, in RTP). This sort of device is closer to being an RTP mixer than an RTP translator: it terminates RTCP reporting about the mixed streams; it can rewrite SSRCs, timestamps, and sequence numbers, as well as the contents of the RTP payloads; and it can turn sources on and off at will without appearing to generate packet loss. Each projected stream will typically preserve its original RTCP source description (SDES) information. 3.4. Multiple SSRCs for a Single Media Source There are also several cases where multiple SSRCs can be used to send data from a single media source within a single RTP session. These include, but are not limited to, transport robustness tools, such as the RTP retransmission payload format [RFC4588], that require one SSRC to be used for the media data and another SSRC for the repair data. Similarly, some layered media encoding schemes, for example, H.264 Scalable Video Coding (SVC) [RFC6190], can be used in a configuration where each layer is sent using a different SSRC within a single RTP session. Lennox, et al. Standards Track [Page 5] RFC 8108 Multiple Media Streams in an RTP Session March 2017 4. Use of RTP by Endpoints That Send Multiple Media Streams RTP is inherently a group communication protocol. Each endpoint in an RTP session will use one or more SSRCs, as will some types of RTP- level middlebox. Accordingly, unless restrictions on the number of SSRCs have been signaled, RTP endpoints can expect to receive RTP data packets sent using a number of different SSRCs, within a single RTP session. This can occur irrespective of whether the RTP session is running over a point-to-point connection or a multicast group, since middleboxes can be used to connect multiple transport connections together into a single RTP session (the RTP session is defined by the shared SSRC space, not by the transport connections). Furthermore, if RTP mixers are used, some SSRCs might only be visible in the contributing source (CSRC) list of an RTP packet and in RTCP, and might not appear directly as the SSRC of an RTP data packet. Every RTP endpoint will have an allocated share of the available session bandwidth, as determined by signaling and congestion control. The endpoint needs to keep its total media sending rate within this share. However, endpoints that send multiple RTP streams do not necessarily need to subdivide their share of the available bandwidth independently or uniformly to each RTP stream and its SSRCs. In particular, an endpoint can vary the bandwidth allocation to different streams depending on their needs, and it can dynamically change the bandwidth allocated to different SSRCs (for example, by using a variable-rate codec), provided the total sending rate does not exceed its allocated share. This includes enabling or disabling RTP streams, or their redundancy streams, as more or less bandwidth becomes available. 5. Use of RTCP by Endpoints That Send Multiple Media Streams RTCP is defined in Section 6 of [RFC3550]. The description of the protocol is phrased in terms of the behavior of "participants" in an RTP session, under the assumption that each endpoint is a participant with a single SSRC. However, for correct operation in cases where endpoints have multiple SSRC values, implementations MUST treat each SSRC as a separate participant in the RTP session, so that an endpoint that has multiple SSRCs counts as multiple participants. Lennox, et al. Standards Track [Page 6] RFC 8108 Multiple Media Streams in an RTP Session March 2017 5.1. RTCP Reporting Requirement An RTP endpoint that has multiple SSRCs MUST treat each SSRC as a separate participant in the RTP session. Each SSRC will maintain its own RTCP-related state information and, hence, will have its own RTCP reporting interval that determines when it sends RTCP reports. If the mechanism in [MULTI-STREAM-OPT] is not used, then each SSRC will send RTCP reports for all other SSRCs, including those co-located at the same endpoint. If the endpoint has some SSRCs that are sending data and some that are only receivers, then they will receive different shares of the RTCP bandwidth and calculate different base RTCP reporting intervals. Otherwise, all SSRCs at an endpoint will calculate the same base RTCP reporting interval. The actual reporting intervals for each SSRC are randomized in the usual way, but reports can be aggregated as described in Section 5.3. 5.2. Initial Reporting Interval When a participant joins a unicast session, the following text from Section 6.2 of [RFC3550] is relevant: "For unicast sessions... the delay before sending the initial compound RTCP packet MAY be zero." The basic assumption is that this also ought to apply in the case of multiple SSRCs. Caution has to be exercised, however, when an endpoint (or middlebox) with a large number of SSRCs joins a unicast session, since immediate transmission of many RTCP reports can create a significant burst of traffic, leading to transient congestion and packet loss due to queue overflows. To ensure that the initial burst of traffic generated by an RTP endpoint is no larger than would be generated by a TCP connection, an RTP endpoint MUST NOT send more than four compound RTCP packets with zero initial delay when it joins an RTP session, independent of the number of SSRCs used by the endpoint. Each of those initial compound RTCP packets MAY include aggregated reports from multiple SSRCs, provided the total compound RTCP packet size does not exceed the MTU, and the avg_rtcp_size is maintained as in Section 5.3.1. Aggregating reports from several SSRCs in the initial compound RTCP packets allows a substantial number of SSRCs to report immediately. Endpoints SHOULD prioritize reports on SSRCs that are likely to be most immediately useful, e.g., for SSRCs that are initially senders. An endpoint that needs to report on more SSRCs than will fit into the four compound RTCP reports that can be sent immediately MUST send the other reports later, following the usual RTCP timing rules including timer reconsideration. Those reports MAY be aggregated as described in Section 5.3. Lennox, et al. Standards Track [Page 7] RFC 8108 Multiple Media Streams in an RTP Session March 2017 Note: The above is chosen to match the TCP maximum initial window of four packets [RFC3390], not the larger TCP initial windows for which there is an ongoing experiment [RFC6928]. The reason for this is a desire to be conservative, since an RTP endpoint will also in many cases start sending RTP data packets at the same time as these initial RTCP packets are sent. 5.3. Aggregation of Reports into Compound RTCP Packets As outlined in Section 5.1, an endpoint with multiple SSRCs has to treat each SSRC as a separate participant when it comes to sending RTCP reports. This will lead to each SSRC sending a compound RTCP packet in each reporting interval. Since these packets are coming from the same endpoint, it might reasonably be expected that they can be aggregated to reduce overheads. Indeed, Section 6.1 of [RFC3550] allows RTP translators and mixers to aggregate packets in similar circumstances: It is RECOMMENDED that translators and mixers combine individual RTCP packets from the multiple sources they are forwarding into one compound packet whenever feasible in order to amortize the packet overhead (see Section 7). An example RTCP compound packet as might be produced by a mixer is shown in Fig. 1. If the overall length of a compound packet would exceed the MTU of the network path, it SHOULD be segmented into multiple shorter compound packets to be transmitted in separate packets of the underlying protocol. This does not impair the RTCP bandwidth estimation because each compound packet represents at least one distinct participant. Note that each of the compound packets MUST begin with an SR or RR packet. This allows RTP translators and mixers to generate compound RTCP packets that contain multiple Sender Report (SR) or Receiver Report (RR) packets from different SSRCs, as well as any of the other packet types. There are no restrictions on the order in which the RTCP packets can occur within the compound packet, except the regular rule that the compound RTCP packet starts with an SR or RR packet. Due to this rule, correctly implemented RTP endpoints will be able to handle compound RTCP packets that contain RTCP packets relating to multiple SSRCs. Accordingly, endpoints that use multiple SSRCs can aggregate the RTCP packets sent by their different SSRCs into compound RTCP packets, provided 1) the resulting compound RTCP packets begin with an SR or RR packet, 2) they maintain the average RTCP packet size as described in Section 5.3.1, and 3) they schedule packet transmission and manage aggregation as described in Section 5.3.2. Lennox, et al. Standards Track [Page 8] RFC 8108 Multiple Media Streams in an RTP Session March 2017 5.3.1. Maintaining AVG_RTCP_SIZE The RTCP scheduling algorithm in [RFC3550] works on a per-SSRC basis. Each SSRC sends a single compound RTCP packet in each RTCP reporting interval. When an endpoint uses multiple SSRCs, it is desirable to aggregate the compound RTCP packets sent by its SSRCs, reducing the overhead by forming a larger compound RTCP packet. This aggregation can be done as described in Section 5.3.2, provided the average RTCP packet size calculation is updated as follows. Participants in an RTP session update their estimate of the average RTCP packet size (avg_rtcp_size) each time they send or receive an RTCP packet (see Section 6.3.3 of [RFC3550]). When a compound RTCP packet that contains RTCP packets from several SSRCs is sent or received, the avg_rtcp_size estimate for each SSRC that is reported upon is updated using div_packet_size rather than the actual packet size: avg_rtcp_size = (1/16) * div_packet_size + (15/16) * avg_rtcp_size where div_packet_size is packet_size divided by the number of SSRCs reporting in that compound packet. The number of SSRCs reporting in a compound packet is determined by counting the number of different SSRCs that are the source of SR or RR RTCP packets within the compound RTCP packet. Non-compound RTCP packets (i.e., RTCP packets that do not contain an SR or RR packet [RFC5506]) are considered to report on a single SSRC. A participant that doesn't follow the above rule, and instead uses the full RTCP compound packet size to calculate avg_rtcp_size, will derive an RTCP reporting interval that is overly large by a factor that is proportional to the number of SSRCs aggregated into compound RTCP packets and the size of set of SSRCs being aggregated relative to the total number of participants. This increased RTCP reporting interval can cause premature timeouts if it is more than five times the interval chosen by the SSRCs that understand compound RTCP that aggregate reports from many SSRCs. A 1500-octet MTU can fit five typical-size reports into a compound RTCP packet, so this is a real concern if endpoints aggregate RTCP reports from multiple SSRCs. The issue raised in the previous paragraph is mitigated by the modification in timeout behavior specified in Section 7.1.2 of this memo. This mitigation is in place in those cases where the RTCP bandwidth is sufficiently high that an endpoint, using avg_rtcp_size calculated without taking into account the number of reporting SSRCs, can transmit more frequently than approximately every 5 seconds. Note, however, that the non-updated endpoint's RTCP reporting is still negatively impacted even if the premature timeouts of its SSRCs Lennox, et al. Standards Track [Page 9] RFC 8108 Multiple Media Streams in an RTP Session March 2017 are avoided. If compatibility with non-updated endpoints is a concern, the number of reports from different SSRCs aggregated into a single compound RTCP packet SHOULD either be limited to two reports or aggregation ought not be used at all. This will limit the non-updated endpoint's RTCP reporting interval to be no larger than twice the RTCP reporting interval that would be chosen by an endpoint following this specification. 5.3.2. Scheduling RTCP when Aggregating Multiple SSRCs This section revises and extends the behavior defined in Section 6.3 of [RFC3550], and in Section 3.5.3 of [RFC4585] if the RTP/AVPF profile or the RTP/SAVPF profile is used, regarding actions to take when scheduling and sending RTCP packets where multiple reporting SSRCs are aggregating their RTCP packets into the same compound RTCP packet. These changes to the RTCP scheduling rules are needed to maintain important RTCP timing properties, including the inter-packet distribution, and the behavior during flash joins and other changes in session membership. The variables tn, tp, tc, T, and Td used in the following are defined in Section 6.3 of [RFC3550]. The variables T_rr_interval and T_rr_last are defined in [RFC4585]. Each endpoint MUST schedule RTCP transmission independently for each of its SSRCs using the regular calculation of tn for the RTP profile being used. Each time the timer tn expires for an SSRC, the endpoint MUST perform RTCP timer reconsideration and, if applicable, suppression based on T_rr_interval. If the result indicates that a compound RTCP packet is to be sent by that SSRC, and the transmission is not an early RTCP packet [RFC4585], then the endpoint SHOULD try to aggregate RTCP packets of additional SSRCs that are scheduled in the future into the compound RTCP packet before it is sent. The reason to limit or not aggregate due to backwards compatibility reasons is discussed in Section 5.3.1. Aggregation proceeds as follows. The endpoint selects the SSRC that has the smallest tn value after the current time, tc, and prepares the RTCP packets that SSRC would send if its timer tn expired at tc. If those RTCP packets will fit into the compound RTCP packet that is being generated, taking into account the path MTU and the previously added RTCP packets, then they are added to the compound RTCP packet; otherwise, they are discarded. This process is repeated for each SSRC, in order of increasing tn, until the compound RTCP packet is full or all SSRCs have been aggregated. At that point, the compound RTCP packet is sent. Lennox, et al. Standards Track [Page 10] RFC 8108 Multiple Media Streams in an RTP Session March 2017 When the compound RTCP packet is sent, the endpoint MUST update tp, tn, and T_rr_last (if applicable) for each SSRC that was included. These variables are updated as follows: a. For the first SSRC that reported in the compound RTCP packet, set the effective transmission time, tt, of that SSRC to tc. b. For each additional SSRC that reported in the compound RTCP packet, calculate the transmission time that SSRC would have had if it had not been aggregated into the compound RTCP packet. This is derived by taking tn for that SSRC, then performing reconsideration and updating tn until tp + T <= tn. Once this is done, set the effective transmission time, tt, for that SSRC to the calculated value of tn. If the RTP/AVPF profile or the RTP/ SAVPF profile is being used, then suppression based on T_rr_interval MUST NOT be used in this calculation. c. Calculate average effective transmission time, tt_avg, for the compound RTCP packet based on the tt values for all SSRCs sent in the compound RTCP packet. Set tp for each of the SSRCs sent in the compound RTCP packet to tt_avg. If the RTP/AVPF profile or the RTP/SAVPF profile is being used, set T_tt_last for each SSRC sent in the compound RTCP packet to tt_avg. d. For each of the SSRCs sent in the compound RTCP packet, calculate new tn values based on the updated parameters and the usual RTCP timing rules and reschedule the timers. When using the RTP/AVPF profile or the RTP/SAVPF profile, the above mechanism only attempts to aggregate RTCP packets when the compound RTCP packet to be sent is not an early RTCP packet, and hence the algorithm in Section 3.5.3 of [RFC4585] will control RTCP scheduling. If T_rr_interval == 0, or if T_rr_interval != 0 and option 1, 2a, or 2b of the algorithm are chosen, then the above mechanism updates the necessary variables. However, if the transmission is suppressed per option 2c of the algorithm, then tp is updated to tc as aggregation has not taken place. Reverse reconsideration MUST be performed following Section 6.3.4 of [RFC3550]. In some cases, this can lead to the value of tp after reverse reconsideration being larger than tc. This is not a problem, and has the desired effect of proportionally pulling the tp value towards tc (as well as tn) as the reporting interval shrinks in direct proportion the reduced group size. The above algorithm has been shown in simulations [Sim88] [Sim92] to maintain the inter-RTCP packet transmission time distribution for each SSRC and to consume the same amount of bandwidth as Lennox, et al. Standards Track [Page 11] RFC 8108 Multiple Media Streams in an RTP Session March 2017 non-aggregated RTCP packets. With this algorithm, the actual transmission interval for an SSRC triggering an RTCP compound packet transmission is following the regular transmission rules. The value tp is set to somewhere in the interval [0, 1.5/1.21828*Td] ahead of tc. The actual value is the average of one instance of tc and the randomized transmission times of the additional SSRCs; thus, the lower range of the interval is more probable. This compensates for the bias that is otherwise introduced by picking the shortest tn value out of the N SSRCs included in aggregate. The algorithm also handles the cases where the number of SSRCs that can be included in an aggregated packet varies. An SSRC that previously was aggregated and fails to fit in a packet still has its own transmission scheduled according to normal rules. Thus, it will trigger a transmission in due time, or the SSRC will be included in another aggregate. The algorithm's behavior under SSRC group size changes is as follows: RTP sessions where the number of SSRCs is growing: When the group size is growing, Td grows in proportion to the number of new SSRCs in the group. When reconsideration is performed due to expiry of the tn timer, that SSRC will reconsider the transmission and with a certain probability reschedule the tn timer. This part of the reconsideration algorithm is only impacted by the above algorithm having tp values that were in the future instead of set to the time of the actual last transmission at the time of updating tp. RTP sessions where the number of SSRCs is shrinking: When the group shrinks, reverse reconsideration moves the tp and tn values towards tc proportionally to the number of SSRCs that leave the session compared to the total number of participants when they left. The setting of the tp value forward in time related to the tc could be believed to have negative effect. However, the reason for this setting is to compensate for bias caused by picking the shortest tn out of the N aggregated. This bias remains over a reduction in the number of SSRCs. The reverse reconsideration compensates the reduction independently of whether or not aggregation is being used. The negative effect that can occur on removing an SSRC is that the most favorable tn belonged to the removed SSRC. The impact of this is limited to delaying the transmission, in the worst case, one reporting interval. In conclusion, the investigations performed have found no significant negative impact on the scheduling algorithm. Lennox, et al. Standards Track [Page 12] RFC 8108 Multiple Media Streams in an RTP Session March 2017 5.4. Use of RTP/AVPF or RTP/SAVPF Feedback This section discusses the transmission of RTP/AVPF feedback packets when the transmitting endpoint has multiple SSRCs. The guidelines in this section also apply to endpoints using the RTP/SAVPF profile. 5.4.1. Choice of SSRC for Feedback Packets When an RTP/AVPF endpoint has multiple SSRCs, it can choose what SSRC to use as the source for the RTCP feedback packets it sends. Several factors can affect that choice: o RTCP feedback packets relating to a particular media type SHOULD be sent by an SSRC that receives that media type. For example, when audio and video are multiplexed onto a single RTP session, endpoints will use their audio SSRC to send feedback on the audio received from other participants. o RTCP feedback packets and RTCP codec control messages that are notifications or indications regarding RTP data processed by an endpoint MUST be sent from the SSRC used for that RTP data. This includes notifications that relate to a previously received request or command [RFC4585][RFC5104]. o If separate SSRCs are used to send and receive media, then the corresponding SSRC SHOULD be used for feedback, since they have differing RTCP bandwidth fractions. This can also affect the consideration of whether or not the SSRC can be used in immediate mode. o Some RTCP feedback packet types require consistency in the SSRC used. For example, if a Temporary Maximum Media Stream Bit Rate Request (TMMBR) limitation [RFC5104] is set by an SSRC, the same SSRC needs to be used to remove the limitation. o If several SSRCs are suitable for sending feedback, it might be desirable to use an SSRC that allows the sending of feedback as an early RTCP packet. When an RTCP feedback packet is sent as part of a compound RTCP packet that aggregates reports from multiple SSRCs, there is no requirement that the compound packet contain an SR or RR packet generated by the sender of the RTCP feedback packet. For reduced- size RTCP packets, aggregation of RTCP feedback packets from multiple sources is not limited further than Section 4.2.2 of [RFC5506]. Lennox, et al. Standards Track [Page 13] RFC 8108 Multiple Media Streams in an RTP Session March 2017 5.4.2. Scheduling an RTCP Feedback Packet When an SSRC has a need to transmit a feedback packet in early mode, it MUST schedule that packet following the algorithm in Section 3.5 of [RFC4585] modified as follows: o To determine whether an RTP session is considered to be a point- to-point session or a multiparty session, an endpoint MUST count the number of distinct RTCP SDES CNAME values used by the SSRCs listed in the SSRC field of RTP data packets it receives and in the "SSRC of sender" field of RTCP SR, RR, RTPFB, or PSFB packets it receives. An RTP session is considered to be a multiparty session if more than one CNAME is used by those SSRCs, unless signaling indicates that the session is to be handled as point to point or RTCP reporting groups [MULTI-STREAM-OPT] are used. If RTCP reporting groups are used, an RTP session is considered to be a point-to-point session if the endpoint receives only a single reporting group and is considered to be a multiparty session if multiple reporting groups are received or a combination of reporting groups and SSRCs that are not part of a reporting group are received. Endpoints MUST NOT determine whether an RTP session is multiparty or point to point based on the type of connection (unicast or multicast) used, or on the number of SSRCs received. o When checking if there is already a scheduled compound RTCP packet containing feedback messages (Step 2 in Section 3.5.2 of [RFC4585]), that check MUST be done considering all local SSRCs. o If an SSRC is not allowed to send an early RTCP packet, then the feedback message MAY be queued for transmission as part of any early or regular scheduled transmission that can occur within the maximum useful lifetime of the feedback message (T_max_fb_delay). This modifies the behavior in item 4a in Section 3.5.2 of [RFC4585]. The first bullet point above specifies a rule to determine if an RTP session is to be considered a point-to-point session or a multiparty session. This rule is straightforward to implement, but is known to incorrectly classify some sessions as multiparty sessions. The known problems are as follows: Endpoint with multiple synchronization contexts: An endpoint that is part of a point-to-point session can have multiple synchronization contexts, for example, due to forwarding an external media source into an interactive real-time conversation. In this case, the classification will consider the peer as two endpoints, while the actual RTP/RTCP transmission will be under the control of one endpoint. Lennox, et al. Standards Track [Page 14] RFC 8108 Multiple Media Streams in an RTP Session March 2017 Selective Forwarding Middlebox: The Selective Forwarding Middlebox (SFM) as defined in Section 3.7 of [RFC7667] has control over the transmission and configurations between itself and each peer endpoint individually. It also fully controls the RTCP packets being forwarded between the individual legs. Thus, this type of middlebox can be compared to the RTP mixer, which uses its own SSRCs to mix or select the media it forwards, that will be classified as a point-to-point RTP session by the above rule. In the above cases, it is very reasonable to use RTCP reporting groups [MULTI-STREAM-OPT]. If that extension is used, an endpoint can indicate that the multitude of CNAMEs are in fact under a single endpoint or middlebox control by using only a single reporting group. The above rules will also classify some sessions where the endpoint is connected to an RTP mixer as being point to point. For example, the mixer could act as gateway to an RTP session based on Any Source Multicast for the discussed endpoint. However, this will, in most cases, be okay, as the RTP mixer provides separation between the two parts of the session. The responsibility falls on the mixer to act accordingly in each domain. Finally, we note that signaling mechanisms could be defined to override the rules when they would result in the wrong classification. 6. Adding and Removing SSRCs The set of SSRCs present in a single RTP session can vary over time due to changes in the number of endpoints in the session or due to changes in the number or type of RTP streams being sent. Every endpoint in an RTP session will have at least one SSRC that it uses for RTCP reporting, and for sending media if desired. It can also have additional SSRCs, for sending extra media sources or for additional RTCP reporting. If the set of media sources being sent changes, then the set of SSRCs being sent will change. Changes in the media format or clock rate might also require changes in the set of SSRCs used. An endpoint can also have more SSRCs than it has active RTP streams, and send RTCP relating to SSRCs that are not currently sending RTP data packets so that its peers are aware of the SSRCs, and have the associated context (e.g., clock synchronization and an SDES CNAME) in place to be able to play out media as soon as they becomes active. In the following, we describe some considerations around adding and removing RTP streams and their associated SSRCs. Lennox, et al. Standards Track [Page 15] RFC 8108 Multiple Media Streams in an RTP Session March 2017 6.1. Adding RTP Streams When an endpoint joins an RTP session, it can have zero, one, or more RTP streams it will send, or that it is prepared to send. If it has no RTP stream it plans to send, it still needs an SSRC that will be used to send RTCP feedback. If it will send one or more RTP streams, it will need the corresponding number of SSRC values. The SSRCs used by an endpoint are made known to other endpoints in the RTP session by sending RTP and RTCP packets. SSRCs can also be signaled using non-RTP means (e.g., [RFC5576]). Unless restricted by signaling, an endpoint can, at any time, send an additional RTP stream, identified by a new SSRC (this might be associated with a signaling event, but that is outside the scope of this memo). This makes the new SSRC visible to the other endpoints in the session, since they share the single SSRC space inherent in the definition of an RTP session. An endpoint that has never sent an RTP stream will have an SSRC that it uses for RTCP reporting. If that endpoint wants to start sending an RTP stream, it is RECOMMENDED that it use its existing SSRC for that stream, since otherwise the participant count in the RTP session will be unnecessarily increased, leading to a longer RTCP reporting interval and larger RTCP reports due to cross reporting. If the endpoint wants to start sending more than one RTP stream, it will need to generate a new SSRC for the second and any subsequent RTP streams. An endpoint that has previously stopped sending an RTP stream, and that wants to start sending a new RTP stream, cannot generally reuse the existing SSRC, and often needs to generate a new SSRC, because an SSRC cannot change media type (e.g., audio to video) or RTP timestamp clock rate [RFC7160] and because the SSRC might be associated with a particular semantic by the application (note: an RTP stream can pause and restart using the same SSRC, provided RTCP is sent for that SSRC during the pause; these rules only apply to new RTP streams reusing an existing SSRC). 6.2. Removing RTP Streams An SSRC is removed from an RTP session in one of two ways. When an endpoint stops sending RTP and RTCP packets using an SSRC, then that SSRC will eventually time out as described in Section 6.3.5 of [RFC3550]. Alternatively, an SSRC can be explicitly removed from use by sending an RTCP BYE packet as described in Section 6.3.7 of [RFC3550]. It is RECOMMENDED that SSRCs be removed from use by sending an RTCP BYE packet. Note that [RFC3550] requires that the RTCP BYE SHOULD be the last RTP/RTCP packet sent in the RTP session Lennox, et al. Standards Track [Page 16] RFC 8108 Multiple Media Streams in an RTP Session March 2017 for an SSRC. If an endpoint needs to restart an RTP stream after sending an RTCP BYE for its SSRC, it needs to generate a new SSRC value for that stream. The finality of sending RTCP BYE means that endpoints need to consider if the ceasing of transmission of an RTP stream is temporary or permanent. Temporary suspension of media transmission using a particular RTP stream (SSRC) needs to maintain that SSRC as an active participant, by continuing RTCP transmission for it. That way the media sending can be resumed immediately, knowing that the context is in place. When permanently halting transmission, a participant needs to send an RTCP BYE to allow the other participants to use the RTCP bandwidth resources and clean up their state databases. An endpoint that ceases transmission of all its RTP streams but remains in the RTP session MUST maintain at least one SSRC that is to be used for RTCP reporting and feedback (i.e., it cannot send a BYE for all SSRCs, but needs to retain at least one active SSRC). As some Feedback packets can be bound to media type, there might be a need to maintain one SSRC per media type within an RTP session. An alternative can be to create a new SSRC to use for RTCP reporting and feedback. However, to avoid the perception that an endpoint drops completely out of an RTP session, such a new SSRC ought to be established first -- before terminating all the existing SSRCs. 7. RTCP Considerations for Streams with Disparate Rates An RTP session has a single set of parameters that configure the session bandwidth. These are the RTCP sender and receiver fractions (e.g., the SDP "b=RR:" and "b=RS:" lines [RFC3556]) and the parameters of the RTP/AVPF profile [RFC4585] (e.g., trr-int) if that profile (or its secure extension, RTP/SAVPF [RFC5124]) is used. As a consequence, the base RTCP reporting interval, before randomization, will be the same for every sending SSRC in an RTP session. Similarly, every receiving SSRC in an RTP session will have the same base reporting interval, although this can differ from the reporting interval chosen by sending SSRCs. This uniform RTCP reporting interval for all SSRCs can result in RTCP reports being sent more often, or too seldom, than is considered desirable for an RTP stream. For example, consider a scenario in which an audio flow sending at tens of kilobits per second is multiplexed into an RTP session with a multi-megabit high-quality video flow. If the session bandwidth is configured based on the video sending rate, and the default RTCP bandwidth fraction of 5% of the session bandwidth is used, it is likely that the RTCP bandwidth will exceed the audio sending rate. If the reduced minimum RTCP interval described in Section 6.2 of [RFC3550] is then used in the session, as appropriate for video where Lennox, et al. Standards Track [Page 17] RFC 8108 Multiple Media Streams in an RTP Session March 2017 rapid feedback on damaged I-frames is wanted, the uniform reporting interval for all senders could mean that audio sources are expected to send RTCP packets more often than they send audio data packets. This bandwidth mismatch can be reduced by careful tuning of the RTCP parameters, especially trr_int when the RTP/AVPF profile is used, but cannot be avoided entirely as it is inherent in the design of the RTCP timing rules, and affects all RTP sessions that contain flows with greatly mismatched bandwidth. Different media rates or desired RTCP behaviors can also occur with SSRCs carrying the same media type. A common case in multiparty conferencing is when a small number of video streams are shown in high resolution, while the others are shown as low-resolution thumbnails, with the choice of which is shown in high resolution being voice-activity controlled. Here the differences are both in actual media rate and in choices for what feedback messages might be needed. Other examples of differences that can exist are due to the intended usage of a media source. A media source carrying the video of the speaker in a conference is different from a document camera. Basic parameters that can differ in this case are frame-rate, acceptable end-to-end delay, and the Signal-to-Noise Ratio (SNR) fidelity of the image. These differences affect not only the needed bitrates, but also possible transmission behaviors, usable repair mechanisms, what feedback messages the control and repair requires, the transmission requirements on those feedback messages, and monitoring of the RTP stream delivery. Other similar scenarios can also exist. Sending multiple media types in a single RTP session causes that session to contain more SSRCs than if each media type was sent in a separate RTP session. For example, if two participants each send an audio and a video RTP stream in a single RTP session, that session will comprise four SSRCs; but if separate RTP sessions had been used for audio and video, each of those two RTP sessions would comprise only two SSRCs. Hence, sending multiple RTP streams in an RTP session increases the amount of cross reporting between the SSRCs, as each SSRC reports on all other SSRCs in the session. This increases the size of the RTCP reports, causing them to be sent less often than would be the case if separate RTP sessions where used for a given RTCP bandwidth. Finally, when an RTP session contains multiple media types, it is important to note that the RTCP reception quality reports, feedback messages, and extended report blocks used might not be applicable to all media types. Endpoints will need to consider the media type of each SSRC, and only send or process reports and feedback that apply to that particular SSRC and its media type. Signaling solutions Lennox, et al. Standards Track [Page 18] RFC 8108 Multiple Media Streams in an RTP Session March 2017 might have shortcomings when it comes to indicating that a particular set of RTCP reports or feedback messages only apply to a particular media type within an RTP session. From an RTCP perspective, therefore, it can be seen that there are advantages to using separate RTP sessions for each media source, rather than sending multiple media sources in a single RTP session. However, these are frequently offset by the need to reduce port use, to ease NAT/firewall traversal, achieved by combining media sources into a single RTP session. The following sections consider some of the issues with using RTCP in sessions with multiple media sources in more detail. 7.1. Timing Out SSRCs Various issues have been identified with timing out SSRC values when sending multiple RTP streams in an RTP session. 7.1.1. Problems with the RTP/AVPF T_rr_interval Parameter The RTP/AVPF profile includes a method to prevent regular RTCP reports from being sent too often. This mechanism is described in Section 3.5.3 of [RFC4585]; it is controlled by the T_rr_interval parameter. It works as follows. When a regular RTCP report is sent, a new random value, T_rr_current_interval, is generated, drawn evenly in the range 0.5 to 1.5 times T_rr_interval. If a regular RTCP packet is to be sent earlier than T_rr_current_interval seconds after the previous regular RTCP packet, and there are no feedback messages to be sent, then that regular RTCP packet is suppressed and the next regular RTCP packet is scheduled. The T_rr_current_interval is recalculated each time a regular RTCP packet is sent. The benefit of suppression is that it avoids wasting bandwidth when there is nothing requiring frequent RTCP transmissions, but still allows utilization of the configured bandwidth when feedback is needed. Unfortunately, this suppression mechanism skews the distribution of the RTCP sending intervals compared to the regular RTCP reporting intervals. The standard RTCP timing rules, including reconsideration and the compensation factor, result in the intervals between sending RTCP packets having a distribution that is skewed towards the upper end of the range [0.5/1.21828, 1.5/1.21828]*Td, where Td is the deterministic calculated RTCP reporting interval. With Td = 5 s, this distribution covers the range [2.052 s, 6.156 s]. In comparison, the RTP/AVPF suppression rules act in an interval that is 0.5 to 1.5 times T_rr_interval; for T_rr_interval = 5s, this is [2.5 s, 7.5 s]. Lennox, et al. Standards Track [Page 19] RFC 8108 Multiple Media Streams in an RTP Session March 2017 The effect of this is that the time between consecutive RTCP packets when using T_rr_interval suppression can become large. The maximum time interval between sending one regular RTCP packet and the next, when T_rr_interval is being used, occurs when T_rr_current_interval takes its maximum value and a regular RTCP packet is suppressed at the end of the suppression period, then the next regular RTCP packet is scheduled after its largest possible reporting interval. Taking the worst case of the two intervals gives a maximum time between two RTCP reports of 1.5*T_rr_interval + 1.5/1.21828*Td. This behavior can be surprising when Td and T_rr_interval have the same value. That is, when T_rr_interval is configured to match the regular RTCP reporting interval. In this case, one might expect that regular RTCP packets are sent according to their usual schedule, but feedback packets can be sent early. However, the above-mentioned issue results in the RTCP packets actually being sent in the range [0.5*Td, 2.731*Td] with a highly non-uniform distribution, rather than the range [0.41*Td, 1.23*Td]. This is perhaps unexpected, but is not a problem in itself. However, when coupled with packet loss, it raises the issue of premature timeout. 7.1.2. Avoiding Premature Timeout In RTP/AVP [RFC3550] the timeout behavior is simple; it is 5 times Td, where Td is calculated with a Tmin value of 5 seconds. In other words, if the configured RTCP bandwidth allows for an average RTCP reporting interval shorter than 5 seconds, the timeout is 25 seconds of no activity from the SSRC (RTP or RTCP); otherwise, the timeout is 5 average reporting intervals. RTP/AVPF [RFC4585] introduces different timeout behaviors depending on the value of T_rr_interval. When T_rr_interval is 0, it uses the same timeout calculation as RTP/AVP. However, when T_rr_interval is non-zero, it replaces Tmin in the timeout calculation, most likely to speed up detection of timed out SSRCs. However, using a non-zero T_rr_interval has two consequences for RTP behavior. First, due to suppression, the number of RTP and RTCP packets sent by an SSRC that is not an active RTP sender can become very low, because of the issue discussed in Section 7.1.1. As the RTCP packet interval can be as long as 2.73*Td, during a 5*Td time period, an endpoint might in fact transmit only a single RTCP packet. The long intervals result in fewer RTCP packets, to a point where a single RTCP packet loss can sometimes result in timing out an SSRC. Second, the RTP/AVPF changes to the timeout rules reduce robustness to misconfiguration. It is common to use RTP/AVPF configured such that RTCP packets can be sent frequently to allow rapid feedback; Lennox, et al. Standards Track [Page 20] RFC 8108 Multiple Media Streams in an RTP Session March 2017 however, this makes timeouts very sensitive to T_rr_interval. For example, if two SSRCs are configured, one with T_rr_interval = 0.1 s and the other with T_rr_interval = 0.6 s, then this small difference will result in the SSRC with the shorter T_rr_interval timing out the other if it stops sending RTP packets, since the other RTCP reporting interval is more than five times its own. When RTP/AVP is used, or RTP/AVPF with T_rr_interval = 0, this is a non-issue, as the timeout period will be 25 s, and differences between configured RTCP bandwidth can only cause premature timeouts when the reporting intervals are greater than 5 s and differ by a factor of five. To limit the scope for such problematic misconfiguration, we define an update to the RTP/AVPF timeout rules in Section 7.1.4. 7.1.3. Interoperability between RTP/AVP and RTP/AVPF If endpoints implementing the RTP/AVP and RTP/AVPF profiles (or their secure variants) are combined within a single RTP session, and the RTP/AVPF endpoints use a non-zero T_rr_interval that is significantly below 5 seconds, there is a risk that the RTP/AVPF endpoints will prematurely time out the SSRCs of the RTP/AVP endpoints, due to their different RTCP timeout rules. Conversely, if the RTP/AVPF endpoints use a T_rr_interval that is significantly larger than 5 seconds, there is a risk that the RTP/AVP endpoints will time out the SSRCs of the RTP/AVPF endpoints. Mixing endpoints using two different RTP profiles within a single RTP session is NOT RECOMMENDED. However, if mixed RTP profiles are used, and the RTP/AVPF endpoints are not updated to follow Section 7.1.4 of this memo, then the RTP/AVPF session SHOULD be configured to use T_rr_interval = 4 seconds to avoid premature timeouts. The choice of T_rr_interval = 4 seconds for interoperability might appear strange. Intuitively, this value ought to be 5 seconds, to make both the RTP/AVP and RTP/AVPF use the same timeout period. However, the behavior outlined in Section 7.1.1 shows that actual RTP/AVPF reporting intervals can be longer than expected. Setting T_rr_interval = 4 seconds gives actual RTCP intervals near to those expected by RTP/AVP, ensuring interoperability. Lennox, et al. Standards Track [Page 21] RFC 8108 Multiple Media Streams in an RTP Session March 2017 7.1.4. Updated SSRC Timeout Rules To ensure interoperability and avoid premature timeouts, all SSRCs in an RTP session MUST use the same timeout behavior. However, previous specifications are inconsistent in this regard. To avoid interoperability issues, this memo updates the timeout rules as follows: o For the RTP/AVP, RTP/SAVP, RTP/AVPF, and RTP/SAVPF profiles, the timeout interval SHALL be calculated using a multiplier of five times the deterministic RTCP reporting interval. That is, the timeout interval SHALL be 5*Td. o For the RTP/AVP, RTP/SAVP, RTP/AVPF, and RTP/SAVPF profiles, calculation of Td, for the purpose of calculating the participant timeout only, SHALL be done using a Tmin value of 5 seconds and not the reduced minimal interval, even if the reduced minimum interval is used to calculate RTCP packet transmission intervals. This changes the behavior for the RTP/AVPF or RTP/SAVPF profiles when T_rr_interval != 0. Specifically, the first paragraph of Section 3.5.4 of [RFC4585] is updated to use Tmin instead of T_rr_interval in the timeout calculation for RTP/AVPF entities. 7.2. Tuning RTCP Transmissions This subsection discusses what tuning can be done to reduce the downsides of the shared RTCP packet intervals. First, what possibilities exist for the RTP/AVP [RFC3551] profile are listed followed by what additional tools are provided by RTP/AVPF [RFC4585]. 7.2.1. RTP/AVP and RTP/SAVP When using the RTP/AVP or RTP/SAVP profiles, the options for tuning the RTCP reporting intervals are limited to the RTCP sender and receiver bandwidth, and whether the minimum RTCP interval is scaled according to the bandwidth. As the scheduling algorithm includes both randomization and reconsideration, one cannot simply calculate the expected average transmission interval using the formula for Td given in Section 6.3.1 of [RFC3550]. However, by considering the inputs to that expression, and the randomization and reconsideration rules, we can begin to understand the behavior of the RTCP transmission interval. Lennox, et al. Standards Track [Page 22] RFC 8108 Multiple Media Streams in an RTP Session March 2017 Let's start with some basic observations: a. Unless the scaled minimum RTCP interval is used, Td prior to randomization and reconsideration can never be less than Tmin. The default value of Tmin is 5 seconds. b. If the scaled minimum RTCP interval is used, Td can become as low as 360 divided by RTP Session bandwidth in kilobits per second. In SDP, the RTP session bandwidth is signaled using a "b=AS" line. An RTP Session bandwidth of 72 kbps results in Tmin being 5 seconds. An RTP session bandwidth of 360 kbps of course gives a Tmin of 1 second, and to achieve a Tmin equal to once every frame for a 25 frame-per-second video stream requires an RTP session bandwidth of 9 Mbps. Use of the RTP/AVPF or RTP/SAVPF profile allows more frequent RTCP reports for the same bandwidth, as discussed below. c. The value of Td scales with the number of SSRCs and the average size of the RTCP reports to keep the overall RTCP bandwidth constant. d. The actual transmission interval for a Td value is in the range [0.5*Td/1.21828, 1.5*Td/1.21828], and the distribution is skewed, due to reconsideration, with the majority of the probability mass being above Td. This means, for example, that for Td = 5 s, the actual transmission interval will be distributed in the range [2.052 s, 6.156 s], and tending towards the upper half of the interval. Note that Tmin parameter limits the value of Td before randomization and reconsideration are applied, so the actual transmission interval will cover a range extending below Tmin. Given the above, we can calculate the number of SSRCs, n, that an RTP session with 5% of the session bandwidth assigned to RTCP can support while maintaining Td equal to Tmin. This will tell us how many RTP streams we can report on, keeping the RTCP overhead within acceptable bounds. We make two assumptions that simplify the calculation: that all SSRCs are senders, and that they all send compound RTCP packets comprising an SR packet with n-1 report blocks, followed by an SDES packet containing a 16 octet CNAME value [RFC7022] (such RTCP packets will vary in size between 54 and 798 octets depending on n, up to the maximum of 31 report blocks that can be included in an SR packet). If we put this packet size, and a 5% RTCP bandwidth fraction into the RTCP interval calculation in Section 6.3.1 of [RFC3550], and calculate the value of n needed to give Td = Tmin for the scaled minimum interval, we find n=9 SSRCs can be supported (irrespective of the interval, due to the way the reporting interval scales with the session bandwidth). We see that to support more SSRCs without changing the scaled minimum interval, we need to increase the RTCP Lennox, et al. Standards Track [Page 23] RFC 8108 Multiple Media Streams in an RTP Session March 2017 bandwidth fraction from 5%; changing the session bandwidth to a higher value would reduce the Tmin. However, if using the default 5% allocation of RTCP bandwidth, an increase will result in more SSRCs being supported given a fixed Td target. Based on the above, when using the RTP/AVP profile or the RTP/SAVP profile, the key limitation for rapid RTCP reporting in small unicast sessions is going to be the Tmin value. The RTP session bandwidth configured in RTCP has to be sufficiently high to reach the reporting goals the application has following the rules for the scaled minimal RTCP interval. 7.2.2. RTP/AVPF and RTP/SAVPF When using RTP/AVPF or RTP/SAVPF, we have a powerful additional tool for tuning RTCP transmissions: the T_rr_interval parameter. Use of this parameter allows short RTCP reporting intervals; alternatively it gives the ability to sent frequent RTCP feedback without sending frequent regular RTCP reports. The use of the RTP/AVPF or RTP/SAVPF profile with T_rr_interval set to a value greater than zero but smaller than Tmin allows more frequent RTCP feedback than the RTP/AVP or RTP/SAVP profiles, for a given RTCP bandwidth. This happens because Tmin is set to zero after the transmission of the initial RTCP report, causing the reporting interval for later packet to be determined by the usual RTCP bandwidth-based calculation, with Tmin=0, and the T_rr_interval. This has the effect that we are no longer restricted by the minimal interval (whether the default 5-second minimum or the reduced minimum interval). Rather, the RTCP bandwidth and the T_rr_interval are the governing factors, allowing faster feedback. Applications that care about rapid regular RTCP feedback ought to consider using the RTP/ AVPF or RTP/SAVPF profile, even if they don't use the feedback features of that profile. The use of the RTP/AVPF or RTP/SAVPF profile allows RTCP feedback packets to be sent frequently, without also requiring regular RTCP reports to be sent frequently, since T_rr_interval limits the rate at which regular RTCP packets can be sent, while still permitting RTCP feedback packets to be sent. Applications that can use feedback packets for some RTP streams, e.g., video streams, but don't want frequent regular reporting for other RTP streams, can configure the T_rr_interval to a value so that the regular reporting for both audio and video is at a level that is considered acceptable for the audio. They could then use feedback packets, which will include RTCP SR/RR packets unless reduced size RTCP feedback packets [RFC5506] are used, Lennox, et al. Standards Track [Page 24] RFC 8108 Multiple Media Streams in an RTP Session March 2017 for the video reporting. This allows the available RTCP bandwidth to be devoted on the feedback that provides the most utility for the application. Using T_rr_interval still requires one to determine suitable values for the RTCP bandwidth value. Indeed, it might make this choice even more important, as this is more likely to affect the RTCP behavior and performance than when using the RTP/AVP or RTP/SAVP profile, as there are fewer limitations affecting the RTCP transmission. When T_rr_interval is non-zero, there are configurations that need to be avoided. If the RTCP bandwidth chosen is such that the Td value is smaller than, but close to, T_rr_interval, then the actual regular RTCP packet transmission interval can become very large, as discussed in Section 7.1.1. Therefore, for configuration where one intends to have Td smaller than T_rr_interval, then Td is RECOMMENDED to be targeted at values less than 1/4th of T_rr_interval, which results in the range becoming [0.5*T_rr_interval, 1.81*T_rr_interval]. With the RTP/AVPF or RTP/SAVPF profiles, using T_rr_interval = 0 has utility and results in a behavior where the RTCP transmission is only limited by the bandwidth, i.e., no Tmin limitations at all. This allows more frequent regular RTCP reporting than can be achieved using the RTP/AVP profile. Many configurations of RTCP will not consume all the bandwidth that they have been configured to use, but this configuration will consume what it has been given. Note that the same behavior will be achieved as long as T_rr_interval is smaller than 1/3 of Td as that prevents T_rr_interval from affecting the transmission. There exists no method for using different regular RTCP reporting intervals depending on the media type or individual RTP stream, other than using a separate RTP session for each type or stream. 8. Security Considerations When using the secure RTP protocol (RTP/SAVP) [RFC3711], or the secure variant of the feedback profile (RTP/SAVPF) [RFC5124], the cryptographic context of a compound secure RTCP packet is the SSRC of the sender of the first RTCP (sub-)packet. This could matter in some cases, especially for keying mechanisms such as MIKEY [RFC3830] that allow use of per-SSRC keying. Otherwise, the standard security considerations of RTP apply; sending multiple RTP streams from a single endpoint in a single RTP session does not appear to have different security consequences than sending the same number of RTP streams spread across different RTP sessions. Lennox, et al. Standards Track [Page 25] RFC 8108 Multiple Media Streams in an RTP Session March 2017 9. References 9.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997, . [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. Jacobson, "RTP: A Transport Protocol for Real-Time Applications", STD 64, RFC 3550, DOI 10.17487/RFC3550, July 2003, . [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. Norrman, "The Secure Real-time Transport Protocol (SRTP)", RFC 3711, DOI 10.17487/RFC3711, March 2004, . [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, DOI 10.17487/RFC4585, July 2006, . [RFC5124] Ott, J. and E. Carrara, "Extended Secure RTP Profile for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/SAVPF)", RFC 5124, DOI 10.17487/RFC5124, February 2008, . [RFC5506] Johansson, I. and M. Westerlund, "Support for Reduced-Size Real-Time Transport Control Protocol (RTCP): Opportunities and Consequences", RFC 5506, DOI 10.17487/RFC5506, April 2009, . 9.2. Informative References [CLUE-FRAME] Duckworth, M., Ed., Pepperell, A., and S. Wenger, "Framework for Telepresence Multi-Streams", Work in Progress, draft-ietf-clue-framework-25, January 2016. [MULTI-RTP] Westerlund, M., Perkins, C., and J. Lennox, "Sending Multiple Types of Media in a Single RTP Session", Work in Progress, draft-ietf-avtcore-multi-media-rtp-session-13, December 2015. Lennox, et al. Standards Track [Page 26] RFC 8108 Multiple Media Streams in an RTP Session March 2017 [MULTI-STREAM-OPT] Lennox, J., Westerlund, M., Wu, Q., and C. Perkins, "Sending Multiple Media Streams in a Single RTP Session: Grouping RTCP Reception Statistics and Other Feedback", Work in Progress, draft-ietf-avtcore-rtp-multi- stream-optimisation-12, March 2016. [RFC3390] Allman, M., Floyd, S., and C. Partridge, "Increasing TCP's Initial Window", RFC 3390, DOI 10.17487/RFC3390, October 2002, . [RFC3551] Schulzrinne, H. and S. Casner, "RTP Profile for Audio and Video Conferences with Minimal Control", STD 65, RFC 3551, DOI 10.17487/RFC3551, July 2003, . [RFC3556] Casner, S., "Session Description Protocol (SDP) Bandwidth Modifiers for RTP Control Protocol (RTCP) Bandwidth", RFC 3556, DOI 10.17487/RFC3556, July 2003, . [RFC3830] Arkko, J., Carrara, E., Lindholm, F., Naslund, M., and K. Norrman, "MIKEY: Multimedia Internet KEYing", RFC 3830, DOI 10.17487/RFC3830, August 2004, . [RFC4588] Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R. Hakenberg, "RTP Retransmission Payload Format", RFC 4588, DOI 10.17487/RFC4588, 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, DOI 10.17487/RFC5104, February 2008, . [RFC5576] Lennox, J., Ott, J., and T. Schierl, "Source-Specific Media Attributes in the Session Description Protocol (SDP)", RFC 5576, DOI 10.17487/RFC5576, June 2009, . [RFC6190] Wenger, S., Wang, Y., Schierl, T., and A. Eleftheriadis, "RTP Payload Format for Scalable Video Coding", RFC 6190, DOI 10.17487/RFC6190, May 2011, . Lennox, et al. Standards Track [Page 27] RFC 8108 Multiple Media Streams in an RTP Session March 2017 [RFC6928] Chu, J., Dukkipati, N., Cheng, Y., and M. Mathis, "Increasing TCP's Initial Window", RFC 6928, DOI 10.17487/RFC6928, April 2013, . [RFC7022] Begen, A., Perkins, C., Wing, D., and E. Rescorla, "Guidelines for Choosing RTP Control Protocol (RTCP) Canonical Names (CNAMEs)", RFC 7022, DOI 10.17487/RFC7022, September 2013, . [RFC7160] Petit-Huguenin, M. and G. Zorn, Ed., "Support for Multiple Clock Rates in an RTP Session", RFC 7160, DOI 10.17487/RFC7160, April 2014, . [RFC7667] Westerlund, M. and S. Wenger, "RTP Topologies", RFC 7667, DOI 10.17487/RFC7667, November 2015, . [SDP-BUNDLE] Holmberg, C., Alvestrand, H., and C. Jennings, "Negotiating Media Multiplexing Using the Session Description Protocol (SDP)", Work in Progress, draft-ietf-mmusic-sdp-bundle-negotiation-36, October 2016. [Sim88] Westerlund, M., "SIMULATION RESULTS FOR MULTI-STREAM", IETF 88 Proceedings, November 2013, . [Sim92] Westerlund, M., Lennox, J., Perkins, C., and Q. Wu, "Changes in RTP Multi-stream", IETF 92 Proceedings, March 2015, . Lennox, et al. Standards Track [Page 28] RFC 8108 Multiple Media Streams in an RTP Session March 2017 Acknowledgments The authors like to thank Harald Alvestrand and everyone else who has been involved in the development of this document. Authors' Addresses Jonathan Lennox Vidyo, Inc. 433 Hackensack Avenue Seventh Floor Hackensack, NJ 07601 United States of America Email: jonathan@vidyo.com Magnus Westerlund Ericsson Farogatan 2 SE-164 80 Kista Sweden Phone: +46 10 714 82 87 Email: magnus.westerlund@ericsson.com Qin Wu Huawei 101 Software Avenue, Yuhua District Nanjing, Jiangsu 210012 China Email: bill.wu@huawei.com Colin Perkins University of Glasgow School of Computing Science Glasgow G12 8QQ United Kingdom Email: csp@csperkins.org Lennox, et al. Standards Track [Page 29]