RFC 6828

Internet Engineering Task Force (IETF)                            J. Xia
Request for Comments: 6828                                        Huawei
Category: Informational                                     January 2013
ISSN: 2070-1721

                   Content Splicing for RTP Sessions


   Content splicing is a process that replaces the content of a main
   multimedia stream with other multimedia content and delivers the
   substitutive multimedia content to the receivers for a period of
   time.  Splicing is commonly used for insertion of local
   advertisements by cable operators, whereby national advertisement
   content is replaced with a local advertisement.

   This memo describes some use cases for content splicing and a set of
   requirements for splicing content delivered by RTP.  It provides
   concrete guidelines for how an RTP mixer can be used to handle
   content splicing.

Status of This Memo

   This document is not an Internet Standards Track specification; it is
   published for informational purposes.

   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).  Not all documents
   approved by the IESG are a candidate for any level of Internet
   Standard; see Section 2 of RFC 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at

Xia                           Informational                     [Page 1]

RFC 6828                      RTP Splicing                  January 2013

Copyright Notice

   Copyright (c) 2013 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.

Table of Contents

   1. Introduction ....................................................2
   2. System Model and Terminology ....................................3
   3. Requirements for RTP Splicing ...................................6
   4. Content Splicing for RTP Sessions ...............................7
      4.1. RTP Processing in RTP Mixer ................................7
      4.2. RTCP Processing in RTP Mixer ...............................8
      4.3. Considerations for Handling Media Clipping at the
           RTP Layer .................................................10
      4.4. Congestion Control Considerations .........................11
      4.5. Considerations for Implementing Undetectable Splicing .....13
   5. Implementation Considerations ..................................13
   6. Security Considerations ........................................14
   7. Acknowledgments ................................................15
   8. References .....................................................15
      8.1. Normative References ......................................15
      8.2. Informative References ....................................15
   Appendix A. Why Mixer Is Chosen ...................................17

1.  Introduction

   This document outlines how content splicing can be used in RTP
   sessions.  Splicing, in general, is a process where part of a
   multimedia content is replaced with other multimedia content and
   delivered to the receivers for a period of time.  The substitutive
   content can be provided, for example, via another stream or via local
   media file storage.  One representative use case for splicing is
   local advertisement insertion.  This allows content providers to
   replace national advertising content with their own regional
   advertising content prior to delivering the regional advertising
   content to the receivers.  Besides the advertisement insertion use
   case, there are other use cases in which the splicing technology can

Xia                           Informational                     [Page 2]

RFC 6828                      RTP Splicing                  January 2013

   be applied, for example, splicing a recorded video into a video
   conferencing session or implementing a playlist server that stitches
   pieces of video together.

   Content splicing is a well-defined operation in MPEG-based cable TV
   systems.  Indeed, the Society for Cable Telecommunications Engineers
   (SCTE) has created two standards, [SCTE30] and [SCTE35], to
   standardize MPEG2-TS splicing procedures.  SCTE 30 creates a
   standardized method for communication between advertisement server
   and splicer, and SCTE 35 supports splicing of MPEG2 transport

   When using multimedia splicing into the Internet, the media may be
   transported by RTP.  In this case, the original media content and
   substitutive media content will use the same time period but may
   contain different numbers of RTP packets due to different media
   codecs and entropy coding.  This mismatch may require some
   adjustments of the RTP header sequence number to maintain
   consistency.  [RFC3550] provides the tools to enable seamless content
   splicing in RTP sessions, but to date there have been no clear
   guidelines on how to use these tools.

   This memo outlines the requirements for content splicing in RTP
   sessions and describes how an RTP mixer can be used to meet these

2.  System Model and Terminology

   In this document, the splicer, an intermediary network element,
   handles RTP splicing.  The splicer can receive main content and
   substitutive content simultaneously but will send one of them at one
   point of time.

   When RTP splicing begins, the splicer sends the substitutive content
   to the RTP receiver instead of the main content for a period of time.
   When RTP splicing ends, the splicer switches back to sending the main
   content to the RTP receiver.

   A simplified RTP splicing diagram is depicted in Figure 1, in which
   only one main content flow and one substitutive content flow are
   given.  Actually, the splicer can handle multiple splicing for
   multiple RTP sessions simultaneously.  RTP splicing may happen more
   than once in multiple time slots during the lifetime of the main RTP
   stream.  The methods by which the splicer learns when to start and
   end the splicing are out of scope for this document.

Xia                           Informational                     [Page 3]

RFC 6828                      RTP Splicing                  January 2013

         |               | Main Content +-----------+
         |   Main RTP    |------------->|           | Output Content
         |   Content     |              |  Splicer  |--------------->
         +---------------+   ---------->|           |
                            |           +-----------+
                            | Substitutive Content
                  |   Substitutive RTP    |
                  |       Content         |
                  |          or           |
                  |   Local File Storage  |

                    Figure 1: RTP Splicing Architecture

   This document uses the following terminologies.

   Output RTP Stream

      The RTP stream that the RTP receiver is currently receiving.  The
      content of the output of the RTP stream can be either main content
      or substitutive content.

   Main Content

      The multimedia content that is conveyed in the main RTP stream.
      Main content will be replaced by the substitutive content during

   Main RTP Stream

      The RTP stream that the splicer is receiving.  The content of the
      main RTP stream can be replaced by substitutive content for a
      period of time.

   Main RTP Sender

      The sender of RTP packets carrying the main RTP stream.

Xia                           Informational                     [Page 4]

RFC 6828                      RTP Splicing                  January 2013

   Substitutive Content

      The multimedia content that replaces the main content during
      splicing.  The substitutive content can, for example, be contained
      in an RTP stream from a media sender or fetched from local media
      file storage.

   Substitutive RTP Stream

      An RTP stream with new content that will replace the content in
      the main RTP stream.  The substitutive RTP stream and main RTP
      stream are two separate streams.  If the substitutive content is
      provided via a substitutive RTP stream, the substitutive RTP
      stream must pass through the splicer before the substitutive
      content is delivered to the receiver.

   Substitutive RTP Sender

      The sender of RTP packets carrying the substitutive RTP stream.

   Splicing-In Point

      A virtual point in the RTP stream, suitable for substitutive
      content entry, typically in the boundary between two independently
      decodable frames.

   Splicing-Out Point

      A virtual point in the RTP stream, suitable for substitutive
      content exit, typically in the boundary between two independently
      decodable frames.


      An intermediary node that inserts substitutive content into a main
      RTP stream.  The splicer sends substitutive content to the RTP
      receiver instead of main content during splicing.  It is also
      responsible for processing RTP Control Protocol (RTCP) traffic
      between the RTP sender and the RTP receiver.

Xia                           Informational                     [Page 5]

RFC 6828                      RTP Splicing                  January 2013

3.  Requirements for RTP Splicing

   In order to allow seamless content splicing at the RTP layer, the
   following requirements must be met.  Meeting these will also allow,
   but not require, seamless content splicing at layers above RTP.


      The splicer should be agnostic about the network and
      transport-layer protocols used to deliver the RTP streams.


      The splicing operation at the RTP layer must allow splicing at any
      point required by the media content and must not constrain when
      splicing-in or splicing-out operations can take place.


      Splicing of RTP content must be backward compatible with the
      RTP/RTCP protocol, associated profiles, payload formats, and


      The splicer will modify the content of RTP packets and thus break
      the end-to-end security, at a minimum, breaking the data integrity
      and source authentication.  If the splicer is designated to insert
      substitutive content, it must be trusted, i.e., be in the security
      context(s) with the main RTP sender, the substitutive RTP sender,
      and the receivers.  If encryption is employed, the splicer
      commonly must decrypt the inbound RTP packets and re-encrypt the
      outbound RTP packets after splicing.


      The splicer should rewrite as necessary and forward RTCP messages
      (e.g., including packet loss, jitter, etc.) sent from a downstream
      receiver to the main RTP sender or the substitutive RTP sender,
      and thus allow the main RTP sender or substitutive RTP sender to
      learn the performance of the downstream receiver when its content
      is being passed to an RTP receiver.  In addition, the splicer
      should rewrite RTCP messages from the main RTP sender or
      substitutive RTP sender to the receiver.

Xia                           Informational                     [Page 6]

RFC 6828                      RTP Splicing                  January 2013


      The splicer must not affect other RTP sessions running between the
      RTP sender and the RTP receiver and must be transparent for the
      RTP sessions it does not splice.


      The RTP receiver should not be able to detect any splicing points
      in the RTP stream produced by the splicer on the RTP protocol
      level.  For the advertisement insertion use case, it is important
      to make it difficult for the RTP receiver to detect where an
      advertisement insertion is starting or ending from the RTP
      packets, and thus avoiding the RTP receiver from filtering out the
      advertisement content.  This memo only focuses on making the
      splicing undetectable at the RTP layer.  The corresponding
      processing is depicted in Section 4.5.

4.  Content Splicing for RTP Sessions

   The RTP specification [RFC3550] defines two types of middleboxes: RTP
   translators and RTP mixers.  Splicing is best viewed as a mixing
   operation.  The splicer generates a new RTP stream that is a mix of
   the main RTP stream and the substitutive RTP stream.  An RTP mixer is
   therefore an appropriate model for a content splicer.  In the next
   four subsections (from Section 4.1 to Section 4.4), the document
   analyzes how the mixer handles RTP splicing and how it satisfies the
   general requirements listed in Section 3.  In Section 4.5, the
   document looks at REQ-7 in order to hide the fact that splicing takes

4.1.  RTP Processing in RTP Mixer

   A splicer could be implemented as a mixer that receives the main RTP
   stream and the substitutive content (possibly via a substitutive RTP
   stream), and sends a single output RTP stream to the receiver(s).
   That output RTP stream will contain either the main content or the
   substitutive content.  The output RTP stream will come from the mixer
   and will have the synchronization source (SSRC) of the mixer rather
   than the main RTP sender or the substitutive RTP sender.

   The mixer uses its own SSRC, sequence number space, and timing model
   when generating the output stream.  Moreover, the mixer may insert
   the SSRC of the main RTP stream into the contributing source (CSRC)
   list in the output media stream.

Xia                           Informational                     [Page 7]

RFC 6828                      RTP Splicing                  January 2013

   At the splicing-in point, when the substitutive content becomes
   active, the mixer chooses the substitutive RTP stream as the input
   stream and extracts the payload data (i.e., substitutive content).
   If the substitutive content comes from local media file storage, the
   mixer directly fetches the substitutive content.  After that, the
   mixer encapsulates substitutive content instead of main content as
   the payload of the output media stream and then sends the output RTP
   media stream to the receiver.  The mixer may insert the SSRC of the
   substitutive RTP stream into the CSRC list in the output media
   stream.  If the substitutive content comes from local media file
   storage, the mixer should leave the CSRC list blank.

   At the splicing-out point, when the substitutive content ends, the
   mixer retrieves the main RTP stream as the input stream and extracts
   the payload data (i.e., main content).  After that, the mixer
   encapsulates main content instead of substitutive content as the
   payload of the output media stream and then sends the output media
   stream to the receivers.  Moreover, the mixer may insert the SSRC of
   the main RTP stream into the CSRC list in the output media stream as

   Note that if the content is too large to fit into RTP packets sent to
   the RTP receiver, the mixer needs to transcode or perform
   application-layer fragmentation.  Usually the mixer is deployed as
   part of a managed system and MTU will be carefully managed by this
   system.  This document does not raise any new MTU related issues
   compared to a standard mixer described in [RFC3550].

   Splicing may occur more than once during the lifetime of the main RTP
   stream.  This means the mixer needs to send main content and
   substitutive content in turn with its own SSRC identifier.  From
   receiver point of view, the only source of the output stream is the
   mixer regardless of where the content is coming from.

4.2.  RTCP Processing in RTP Mixer

   By monitoring available bandwidth and buffer levels and by computing
   network metrics such as packet loss, network jitter, and delay, an
   RTP receiver can learn the network performance and communicate this
   to the RTP sender via RTCP reception reports.

   According to the description in Section 7.3 of [RFC3550], the mixer
   splits the RTCP flow between the sender and receiver into two
   separate RTCP loops; the RTP sender has no idea about the situation
   on the receiver.  But splicing is a process where the mixer selects
   one media stream from multiple streams rather than mixing them, so
   the mixer can leave the SSRC identifier in the RTCP report intact

Xia                           Informational                     [Page 8]

RFC 6828                      RTP Splicing                  January 2013

   (i.e., the SSRC of the downstream receiver).  This enables the main
   RTP sender or the substitutive RTP sender to learn the situation on
   the receiver.

   If the RTCP report corresponds to a time interval that is entirely
   main content or entirely substitutive content, the number of output
   RTP packets containing substitutive content is equal to the number of
   input substitutive RTP packets (from the substitutive RTP stream)
   during splicing.  In the same manner, the number of output RTP
   packets containing main content is equal to the number of input main
   RTP packets (from the main RTP stream) during non-splicing unless the
   mixer fragments the input RTP packets.  This means that the mixer
   does not need to modify the loss packet fields in reception report
   blocks in RTCP reports.  But, if the mixer fragments the input RTP
   packets, it may need to modify the loss packet fields to compensate
   for the fragmentation.  Whether the input RTP packets are fragmented
   or not, the mixer still needs to change the SSRC field in the report
   block to the SSRC identifier of the main RTP sender or the
   substitutive RTP sender and rewrite the extended highest sequence
   number field to the corresponding original extended highest sequence
   number before forwarding the RTCP report to the main RTP sender or
   the substitutive RTP sender.

   If the RTCP report spans the splicing-in point or the splicing-out
   point, it reflects the characteristics of the combination of main RTP
   packets and substitutive RTP packets.  In this case, the mixer needs
   to divide the RTCP report into two separate RTCP reports and send
   them to their original RTP senders, respectively.  For each RTCP
   report, the mixer also needs to make the corresponding changes to the
   packet loss fields in the report block besides the SSRC field and the
   extended highest sequence number field.

   If the mixer receives an RTCP extended report (XR) block, it should
   rewrite the XR report block in a similar way to the reception report
   block in the RTCP report.

   Besides forwarding the RTCP reports sent from the RTP receiver, the
   mixer can also generate its own RTCP reports to inform the main RTP
   sender, or the substitutive RTP sender, of the reception quality of
   content not sent to the RTP receiver when it reaches the mixer.
   These RTCP reports use the SSRC of the mixer.  If the substitutive
   content comes from local media file storage, the mixer does not need
   to generate RTCP reports for the substitutive stream.

Xia                           Informational                     [Page 9]

RFC 6828                      RTP Splicing                  January 2013

   Based on the above RTCP operating mechanism, the RTP sender whose
   content is being passed to a receiver will see the reception quality
   of its stream as received by the mixer and the reception quality of
   the spliced stream as received by the receiver.  The RTP sender whose
   content is not being passed to a receiver will only see the reception
   quality of its stream as received by the mixer.

   The mixer must forward RTCP source description (SDES) and BYE packets
   from the receiver to the sender and may forward them in inverse
   direction as defined in Section 7.3 of [RFC3550].

   Once the mixer receives an RTP/Audio-Visual Profile with Feedback
   (AVPF) [RFC4585] transport-layer feedback packet, it must handle it
   carefully, as the feedback packet may contain the information of the
   content that comes from different RTP senders.  In this case, the
   mixer needs to divide the feedback packet into two separate feedback
   packets and process the information in the feedback control
   information (FCI) in the two feedback packets, just as in the RTCP
   report process described above.

   If the substitutive content comes from local media file storage
   (i.e., the mixer can be regarded as the substitutive RTP sender), any
   RTCP packets received from downstream related to the substitutive
   content must be terminated on the mixer without any further

4.3.  Considerations for Handling Media Clipping at the RTP Layer

   This section provides informative guidelines on how to handle media
   substitution at the RTP layer to minimize media impact.  Dealing well
   with the media substitution at the RTP layer is necessary for quality
   implementations.  To perfectly erase any media impact needs more
   considerations at the higher layers.  How the media substitution is
   erased at the higher layers is outside of the scope of this memo.

   If the time duration for any substitutive content mismatches, i.e.,
   shorter or longer than the duration of the main content to be
   replaced, then media degradations may occur at the splicing point and
   thus impact the user's experience.

   If the substitutive content has shorter duration from the main
   content, then there could be a gap in the output RTP stream.  The RTP
   sequence number will be contiguous across this gap, but there will be
   an unexpected jump in the RTP timestamp.  Such a gap would cause the
   receiver to have nothing to play.  This may be unavoidable, unless
   the mixer can adjusts the splice in or splice out point to
   compensate.  This assumes the splicing mixer can send more of the
   main RTP stream in place of the shorter substitutive stream or vary

Xia                           Informational                    [Page 10]

RFC 6828                      RTP Splicing                  January 2013

   the length of the substitutive content.  It is the responsibility of
   the higher-layer protocols and the media providers to ensure that the
   substitutive content is of very similar duration as the main content
   to be replaced.

   If the substitute content has longer duration than the reserved gap
   duration, there will be an overlap between the substitutive RTP
   stream and the main RTP stream at the splicing-out point.  A
   straightforward approach is that the mixer performs an ungraceful
   action and terminates the splicing and switches back to the main RTP
   stream even if this may cause media stuttering on the receiver.
   Alternatively, the mixer may transcode the substitutive content to
   play at a faster rate than normal, to adjust it to the length of the
   gap in the main content and generate a new RTP stream for the
   transcoded content.  This is a complex operation and very specific to
   the content and media codec used.  Additional approaches exist; these
   types of issues should be taken into account in both mixer
   implementors and media generators to enable smooth substitutions.

4.4.  Congestion Control Considerations

   If the substitutive content has somewhat different characteristics
   from the main content it replaces, or if the substitutive content is
   encoded with a different codec or has different encoding bitrate, it
   might overload the network and might cause network congestion on the
   path between the mixer and the RTP receiver(s) that would not have
   been caused by the main content.

   To be robust to network congestion and packet loss, a mixer that is
   performing splicing must continuously monitor the status of a
   downstream network by monitoring any of the following RTCP reports
   that are used:

   1.  RTCP receiver reports indicate packet loss [RFC3550].

   2.  RTCP NACKs for lost packet recovery [RFC4585].

   3.  RTCP Explicit Congestion Notification (ECN) Feedback information

Xia                           Informational                    [Page 11]

RFC 6828                      RTP Splicing                  January 2013

   Once the mixer detects congestion on its downstream link, it will
   treat these reports as follows:

   1.  If the mixer receives the RTCP receiver reports with packet loss
       indication, it will forward the reports to the substitutive RTP
       sender or the main RTP sender as described in Section 4.2.

   2.  If mixer receives the RTCP NACK packets defined in [RFC4585] from
       the RTP receiver for packet loss recovery, it first identifies
       the content category of lost packets to which the NACK
       corresponds.  Then, the mixer will generate new RTCP NACKs for
       the lost packets with its own SSRC and make corresponding changes
       to their sequence numbers to match original, pre-spliced,
       packets.  If the lost substitutive content comes from local media
       file storage, the mixer acting as the substitutive RTP sender
       will directly fetch the lost substitutive content and retransmit
       it to the RTP receiver.  The mixer may buffer the sent RTP
       packets and do the retransmission.

       It is somewhat complex that the lost packets requested in a
       single RTCP NACK message not only contain the main content but
       also the substitutive content.  To address this, the mixer must
       divide the RTCP NACK packet into two separate RTCP NACK packets:
       one requests for the lost main content, and another requests for
       the lost substitutive content.

   3.  If an ECN-aware mixer receives RTCP ECN feedback (RTCP ECN
       feedback packets or RTCP XR summary reports) defined in [RFC6679]
       from the RTP receiver, it must process them in a similar way to
       the RTP/AVPF feedback packet or RTCP XR process described in
       Section 4.2 of this memo.

   These three methods require the mixer to run a congestion control
   loop and bitrate adaptation between itself and the RTP receiver.  The
   mixer can thin or transcode the main RTP stream or the substitutive
   RTP stream, but such operations are very inefficient and difficult,
   and they also bring undesirable delay.  Fortunately, as noted in this
   memo, the mixer acting as a splicer can rewrite the RTCP packets sent
   from the RTP receiver and forward them to the RTP sender, thus
   letting the RTP sender knows that congestion is being experienced on
   the path between the mixer and the RTP receiver.  Then, the RTP
   sender applies its congestion control algorithm and reduces the media
   bitrate to a value that is in compliance with congestion control
   principles for the slowest link.  The congestion control algorithm
   may be a TCP-friendly bitrate adaptation algorithm specified in
   [RFC5348] or a Datagram Congestion Control Protocol (DCCP) congestion
   control algorithm defined in [RFC5762].

Xia                           Informational                    [Page 12]

RFC 6828                      RTP Splicing                  January 2013

   If the substitutive content comes from local media file storage, the
   mixer must directly reduce the bitrate as if it were the substitutive
   RTP sender.

   From the above analysis, to reduce the risk of congestion and
   maintain the bandwidth consumption stable over time, the substitutive
   RTP stream is recommended to be encoded at an appropriate bitrate to
   match that of the main RTP stream.  If the substitutive RTP stream
   comes from the substitutive RTP sender, this sender should have some
   knowledge about the media encoding bitrate of the main content in
   advance.  Acquiring such knowledge is out of scope in this document.

4.5.  Considerations for Implementing Undetectable Splicing

   If it is desirable to prevent receivers from detecting that splicing
   is occurring at the RTP layer, the mixer must not include a CSRC list
   in outgoing RTP packets and must not forward RTCP messages from the
   main RTP sender or from the substitutive RTP sender.  Due to the
   absence of a CSRC list in the output RTP stream, the RTP receiver
   only initiates SDES, BYE, and Application-specific functions (APP)
   packets to the mixer without any knowledge of the main RTP sender and
   the substitutive RTP sender.

   The CSRC list identifies the contributing sources; these SSRC
   identifiers of contributing sources are kept globally unique for each
   RTP session.  The uniqueness of the SSRC identifier is used to
   resolve collisions and to detect RTP-level forwarding loops as
   defined in Section 8.2 of [RFC3550].  A danger that loops involving
   those contributing sources will not be detected will be created by
   the absence of a CSRC list in this case.  The loops could occur if
   either the mixer is misconfigured to form a loop or a second
   mixer/translator is added, causing packets to loop back to upstream
   of the original mixer.  An undetected RTP packet loop is a serious
   denial-of-service threat, which can consume all available bandwidth
   or mixer processing resources until the looped packets are dropped as
   a result of congestion.  So, non-RTP means must be used to detect and
   resolve loops if the mixer does not add a CSRC list.

5.  Implementation Considerations

   When the mixer is used to handle RTP splicing, the RTP receiver does
   not need any RTP/RTCP extension for splicing.  As a trade-off,
   additional overhead could be induced on the mixer, which uses its own
   sequence number space and timing model.  So the mixer will rewrite
   the RTP sequence number and timestamp, whatever splicing is active or
   not, and generate RTCP flows for both sides.  In case the mixer
   serves multiple main RTP streams simultaneously, this may lead to
   more overhead on the mixer.

Xia                           Informational                    [Page 13]

RFC 6828                      RTP Splicing                  January 2013

   If an undetectable splicing requirement is required, the CSRC list is
   not included in the outgoing RTP packet; this brings a potential
   issue with loop detection as briefly described in Section 4.5.

6.  Security Considerations

   The splicing application is subject to the general security
   considerations of the RTP specification [RFC3550].

   The mixer acting as splicer replaces some content with other content
   in RTP packets, thus breaking any RTP-level end-to-end security, such
   as integrity protection and source authentication.  Thus, any
   RTP-level or outside security mechanism, such as IPsec [RFC4301] or
   Datagram Transport Layer Security [RFC6347], will use a security
   association between the splicer and the receiver.  When using the
   Secure Real-Time Transport Protocol (SRTP) [RFC3711], the splicer
   could be provisioned with the same security association as the main
   RTP sender.  Using a limitation in the SRTP security services
   regarding source authentication, the splicer can modify and
   re-protect the RTP packets without enabling the receiver to detect if
   the data comes from the original source or from the splicer.

   Security goals to have source authentication all the way from the RTP
   main sender to the receiver through the splicer is not possible with
   splicing and any existing solutions.  A new solution can
   theoretically be developed that enables identifying the participating
   entities and what each provides, i.e., the different media sources,
   main and substituting, and the splicer providing the RTP-level
   integration of the media payloads in a common timeline and
   synchronization context.  Such a solution would obviously not meet
   REQ-7 and will be detectable on the RTP level.

   The nature of this RTP service offered by a network operator
   employing a content splicer is that the RTP-layer security
   relationship is between the receiver and the splicer, and between the
   sender and the splicer, but is not end-to-end between the receiver
   and the sender.  This appears to invalidate the undetectability goal,
   but in the common case, the receiver will consider the splicer as the
   main media source.

   Some RTP deployments use RTP payload security mechanisms (e.g.,
   ISMACryp [ISMACryp]).  If any payload internal security mechanisms
   are used, only the RTP sender and the RTP receiver establish that
   security context, in which case any middlebox (e.g., splicer) between
   the RTP sender and the RTP receiver will not get such keying
   material.  This may impact the splicer's ability to perform splicing
   if it is dependent on RTP payload-level hints for finding the splice
   in and out points.  However, other potential solutions exist to

Xia                           Informational                    [Page 14]

RFC 6828                      RTP Splicing                  January 2013

   specify or mark where the splicing points exist in the media streams.
   When using RTP payload security mechanisms, SRTP or other security
   mechanisms at RTP or lower layers can be used to provide integrity
   and source authentication between the splicer and the RTP receiver.

7.  Acknowledgments

   The following individuals have reviewed the earlier versions of this
   specification and provided very valuable comments: Colin Perkins,
   Magnus Westerlund, Roni Even, Tom Van Caenegem, Joerg Ott, David R.
   Oran, Cullen Jennings, Ali C. Begen, Charles Eckel, and Ning Zong.

8.  References

8.1.  Normative References

   [RFC3550]   Schulzrinne, H., Casner, S., Frederick, R., and V.
               Jacobson, "RTP: A Transport Protocol for Real-Time
               Applications", STD 64, RFC 3550, July 2003.

   [RFC4585]   Ott, J., Wenger, S., Sato, N., Burmeister, C., and J.
               Rey, "Extended RTP Profile for Real-time Transport
               Control Protocol (RTCP)-Based Feedback (RTP/AVPF)",
               RFC 4585, July 2006.

   [RFC6679]   Westerlund, M., Johansson, I., Perkins, C., O'Hanlon, P.,
               and K. Carlberg, "Explicit Congestion Notification (ECN)
               for RTP over UDP", RFC 6679, August 2012.

8.2.  Informative References

   [ISMACryp]  Internet Streaming Media Alliance (ISMA), "ISMA
               Encryption and Authentication Specification 2.0",
               November 2007.

   [RFC3711]   Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K.
               Norrman, "The Secure Real-time Transport Protocol
               (SRTP)", RFC 3711, March 2004.

   [RFC4301]   Kent, S. and K. Seo, "Security Architecture for the
               Internet Protocol", RFC 4301, December 2005.

   [RFC5348]   Floyd, S., Handley, M., Padhye, J., and J. Widmer, "TCP
               Friendly Rate Control (TFRC): Protocol Specification",
               RFC 5348, September 2008.

   [RFC5762]   Perkins, C., "RTP and the Datagram Congestion Control
               Protocol (DCCP)", RFC 5762, April 2010.

Xia                           Informational                    [Page 15]

RFC 6828                      RTP Splicing                  January 2013

   [RFC6347]   Rescorla, E. and N. Modadugu, "Datagram Transport Layer
               Security Version 1.2", RFC 6347, January 2012.

   [SCTE30]    Society of Cable Telecommunications Engineers (SCTE),
               "Digital Program Insertion Splicing API", 2009.

   [SCTE35]    Society of Cable Telecommunications Engineers (SCTE),
               "Digital Program Insertion Cueing Message for Cable",

Xia                           Informational                    [Page 16]

RFC 6828                      RTP Splicing                  January 2013

Appendix A.  Why Mixer Is Chosen

   Both a translator and mixer can realize splicing by changing a set of
   RTP parameters.

   A translator has no SSRC; hence it is transparent to the RTP sender
   and receiver.  Therefore, the RTP sender sees the full path to the
   receiver when the translator is passing its content.  When a
   translator inserts the substitutive content, the RTP sender could get
   a report on the path up to the translator itself.  Additionally, if
   splicing does not occur yet, the translator does not need to rewrite
   the RTP header, and the overhead on the translator can be avoided.

   If a mixer is used to do splicing, it can also allow the RTP sender
   to learn the situation of its content on the receiver or on the mixer
   just like the translator does, which is specified in Section 4.2.
   Compared to the translator, the mixer's outstanding benefit is that
   it is pretty straightforward to do with RTCP messages, for example,
   bit-rate adaptation to handle varying network conditions.  But the
   translator needs more considerations, and its implementation is more

   From the above analysis, both the translator and mixer have their own
   advantages: less overhead or less complexity on handling RTCP.  After
   long and sophisticated discussions, the avtext WG members decided
   that they prefer less complexity rather than less overhead and are
   inclined to choose a mixer to do splicing.

   If one chooses a mixer as splicer, the overhead on the mixer must be
   taken into account even if the splicing has not occurred yet.

Author's Address

   Jinwei Xia
   Software No.101
   Nanjing, Yuhuatai District 210012

   Phone: +86-025-86622310
   EMail: xiajinwei@huawei.com

Xia                           Informational                    [Page 17]