SIPPING WG R. Even Internet-Draft Polycom Expires: August 24, 2003 O. Levin RADVISION N. Ismail Cisco Systems, Inc. February 23, 2003 Conferencing media policy requirements draft-even-sipping-media-policy-requirements- 00.txt Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http:// www.ietf.org/ietf/1id-abstracts.txt. The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. This Internet-Draft will expire on August 24, 2003. Copyright Notice Copyright (C) The Internet Society (2003). All Rights Reserved. Abstract This document defines the data model and the requirements for Media Policy, i.e. a set of rules associated with the media distribution of the conference. This document also presents the requirements for the media manipulations that can be done using these rules by conference participants or third parties using any kind of media/ conference policy control protocol. This document does not address the interface between the focus and the media policy. Even, et al. Expires August 24, 2003 [Page 1] Internet-Draft media policy February 2003 Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. Rational . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 4. High Level Architecture . . . . . . . . . . . . . . . . . . . 3 5. Media Policy Data Model . . . . . . . . . . . . . . . . . . . 5 5.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5.2 Mixer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 5.3 Mixer block examples . . . . . . . . . . . . . . . . . . . . . 7 5.4 Video layout . . . . . . . . . . . . . . . . . . . . . . . . . 9 6. Media Policy Requirements . . . . . . . . . . . . . . . . . . 10 6.1 Genera Media Policy Requirements . . . . . . . . . . . . . . . 10 6.2 Video specific requirements . . . . . . . . . . . . . . . . . 11 7. Security Considerations . . . . . . . . . . . . . . . . . . . 11 8. Appendix I - Media Policy Operations . . . . . . . . . . . . . 12 8.1 Stream selection operations . . . . . . . . . . . . . . . . . 12 8.2 Mixing Operation . . . . . . . . . . . . . . . . . . . . . . . 13 8.3 Well known streams . . . . . . . . . . . . . . . . . . . . . . 13 References . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . 14 Intellectual Property and Copyright Statements . . . . . . . . 15 Even, et al. Expires August 24, 2003 [Page 2] Internet-Draft media policy February 2003 1. Introduction The Conferencing Framework [2] presents an overall framework and defines the terminology for SIP [1] tightly coupled conferencing. The conferencing framework architecture includes the media policy. This is a set of rules that describes the media distribution of a conference. This document presents the requirements for the media policy data model and for the manipulations on these rules by conference participants or third parties using any kind of media/ conference policy control protocol. This document does not address the interface between the focus and the media policy and between the focus and the media mixer. 2. Rational The media policy enables a conference participant or an application server to define and manipulate the content of the media streams going to the conference participants. This will enable applications like sidebars, announcement to specific participants, call centers and panel conferences. 3. Terminology The draft relies on the terminology defined in the conferencing framework document[2]. 4. High Level Architecture The basic conferencing architecture used in this document is defined in the Conferencing architecture framework [2]. This document focuses on the media policy component and the requirements to manipulate the media policy by authorized entities. An authorized entity can manipulate the media policy using a supplied application. Examples for such applications include a web application, an interactive voice response application, an interactive Instant Messaging (IM) base application, or an application that uses the media policy control protocol. The Conference policy control protocol (CPCP) provides a standard way for an automated authorized entity to manipulate the media policy. The requirements and definition of the CPCP protocol are out of scope of this document. The media policy is a set of rules that describes the media mixing or switching required for each participant in the conference. This includes the set of sources to be mixed or switched and the rules for their mixing or switching. The focus uses the media policy to Even, et al. Expires August 24, 2003 [Page 3] Internet-Draft media policy February 2003 determine the proper configuration of the mixers. Authorized entities will be notified of changes to the media policy by subscribing to the conference event package. The information about the current contributing sources to the mixed streams can be learned by the information in the RTP header or by the conference event package [4]. The data structures that include the contributing sources of the current streams is in the focus or the mixer and is not in the scope of the work. The initial state of the media policy data structure is defined at the conference creation time. It can be either provisioned or created by using a conference policy control protocol or/and other protocols being used to create the conference. Typically, a focus has access to the media policy and is responsible for translating the media policy data into the actions towards the physical entities ("mixers"). Figure 1 describes an instance of media policy of a conference. The figure shows a single mixer and a single type of stream for ease of drawing but the model does not have such a restriction. Even, et al. Expires August 24, 2003 [Page 4] Internet-Draft media policy February 2003 Conference . Policy . +-----------+ //-----\\ . Control . | | || || . Protocol . | Conference| \\-----// . +------------->| Policy | | | . | . | Server |----> |Conference . | . | | | | . | . +-----------+ | & | . | . | | . | . | Media | . +------------+ . +-----------+ | Policy| . | +----------++ . | | \ // . | | || . | | \-----/ . |Participants||<-------->| Focus | | . | | || SIP . | | | . | | || Dialog . | |<-----------+ . +-+----------++ . +-----------+ . +-----------+ . | . ^ | . | . | | Contributing . | . | | Streams . +-----------+ . | +------------->. | | . | Distributed . | Mixer | . | Streams . | | . +-------------------. +-----------+ . ..................................... Conference Functions Figure 1: Media Policy in a Conference 5. Media Policy Data Model 5.1 General The fundamental conferencing functionality is being able to combine (i.e. "to mix") in a media specific manner participants' streams that belong to a logical sub-function within a conference (such as participant's video, left audio stream, right audio stream, video streaming presentation, slide presentation) and are of the same media type (such as video, audio, etc.). In the case of using centralized-mixing the resultant stream(s) will be sent back to the participants. In the case of end-point mixing, the original streams, needed to produce the mixed media, will be distributed to the participants that will perform the actual mixing. Even, et al. Expires August 24, 2003 [Page 5] Internet-Draft media policy February 2003 Typically, the maximum number of different mixers in a conference is preconfigured as part of the media conference policy. Mixers MAY be dynamically created and destroyed during the conference lifetime. Conferencing mixing scenarios include simple conferences that have well known mixing characteristics (e.g. mix audio of N loudest speakers) as well as conferences that enable the users of the media policy to create custom mixes (e.g. Mix the audio of John and Mary) An example of an audio mixer is shown in Figure 2 (Wide lines specify a set of a streams). The audio mixer in the example is for a simple audio conference that mixes the audio of three loudest speakers. The speakers will not hear themselves. The mixer will choose the three loudest speakers from all the conference participants and will mix them. The output from the mixer will be four mixed audio streams. The first one is the sum of all three loudest speakers, the other three streams are the streams that will be sent to the speakers and each of them will include the mix of the other two loudest participants. Main Audio | | | | v v v v .................................... . +----------------------+ . . | | . Mixer | Loudest(3) | . . | | . . | | . . +----------------------+ . . ||| . . ||| Loudout . . v . . +----------------------+ . . | | . . | MixMinus(3) | . . | | . . | | . . +----------------------+ . .................................... ||| ||| MiXAudio v Even, et al. Expires August 24, 2003 [Page 6] Internet-Draft media policy February 2003 Figure 2: simple audio conference 5.2 Mixer The mixers are composed of logical blocks "mixer block" each mixer block has input streams, output streams and is characterized by an operation that describes the transformation from the input streams to the output streams. The operations can instruct the mixer block to perform a selection operation on the incoming streams, a mixing function or a transformation on a stream. The number of mixer blocks inside a mixer is dynamic. The output from one block may serve as input to another block by using the output stream descriptor as input to a mixer block. The solution will try to reduce the number of blocks needed to define a conference mix. This can be achieved by having rules that will enable a focus to control a mixer based on minimal number of rules. Each mixer has a set of input streams from the participants (called contributing streams) and a set of output streams to the participants (called distributed streams). For this purpose, a participant can be a SIP UA , a media server (such as an announcement server), or a recording system. Each participant incoming (to the focus) stream MUST NOT be associated with more than a single contributing stream of a single mixer. The number and the type of the contributing and distributed streams is usually dynamic. During the conference lifetime, the input and the output media stream attributes (such as bandwidth, rate, CODEC, resolution, etc.) are being defined dynamically such as by means of the SDP m- lines as a part of the Offer-Answer[3] mechanism. 5.3 Mixer block examples The media policy enables the creation of different conferencing scenarios. An example of a main audio conference and a sidebar is shown in figure 3. Even, et al. Expires August 24, 2003 [Page 7] Internet-Draft media policy February 2003 MainAudioIn SidebarIn | | | | | | v v v v v v +----------------------+ +--------+ +-----------+ | | +----> | In*0.4 | | | | Loudest(3) | | | | |MixMinus(2)| | | | | | | | | | | +--------+ +-----------+ +----------------------+ | | ||| ||| | SideBout | SideBout ||| ||| Loudout | V V v | +----------------------+ +----------------------+ | | | | | |All | Add | | MixMinus(3) | | | | | | | +----------------------+ | | | ||| +----------------------+ | SideBmix ||| MainMiXAudio||| ||| | ||| ||| \\\--- +-----------+ V ||| ----->|SelectMain | v ----- +-----------+ Figure 3: Audio Conference with sidebar The main conference is the same as in figure 2. The sidebar conference includes two participants and the main conference audio at a lower volume. The output from the sidebar includes two mixes each with one of the participants and the main audio with the reduced volume and there will be a stream that will include a mix of all three streams. The blocks of the main conference are showing the operations on the conference incoming streams. The Loudest(3) block gets as input all the contributing audio streams of the conference and selects the three that represent the speakers with the highest volume. Those streams are coming out as Loudout. The Loudout set of streams is an input to the MixMinus(3) block. This block will be used by the focus to tell the media mixer to create 4 output streams, one with all three sources and three that are a mix of each two of the participants. These are the distributed streams and the focus knows who are the receivers of each of the streams. The SelectMain block is used to select the main mix of the conference which include the mix of all the three loudest participants, this stream will be an input to the sidebar. The sidebar has three blocks. The In*0.4 takes the input stream and Even, et al. Expires August 24, 2003 [Page 8] Internet-Draft media policy February 2003 attenuate it to 0.4 of its volume. The MixMinus block is the same as the one in the main mixer but it mix only the two loudest speakers and create three output streams. The Add block get as input all streams called SideBout (from the previous two blocks) and mixes them. Each output stream from the In block is mixed with each output stream of the MixMinus block. This will create three mixed streams that the focus can use to send to the sidebar participants. The main conference can be describe as: MainAudio ------Loudest(3)-----> Loudout Loudout-------MixMinus(3)------>MainMixAudio MainMixAudio------SelectMain------->All The sidebar conference can be described as: SidebarIn-----MixMinus(2)---->SideBout All----------------0.4In------>SideBout SideBout-----------Add------->SideBmix The description include predefined operations which are: Loudest(N), MixMinus(N),Main, In, Mix. Those operation and others will be specified as part of the media rules definition. The CPCP will carry those media policy rules to the conference policy server where they will be stored as media policy. 5.4 Video layout Video mixing has two general models. The first one is video switching where the conference bridge sends one of the incoming streams as is to the participants. An example is voice activated video switching where all the participants see the loudest speaker while the speaker sees the previous speaker. The second one is known as continuous presence. In this mode the video mixer builds a composite video that displays each contributing stream in a sub-window. Example is a 2x2 display as in figure 4; each sub-window has a number. Even, et al. Expires August 24, 2003 [Page 9] Internet-Draft media policy February 2003 +-----------------------+------------------------+ | | | | 1 | 2 | | | | +-----------------------+------------------------+ | | | | 3 | 4 | | | | +-----------------------+------------------------+ Figure 4: 2x2 video layout There are well known window layout that can be specified by naming convention to enable simple rules for mapping a video stream to a window. We can add custom window definition support that will serve the creation of more complicated layouts. The numbering of the sub-windows will start from the top-left sub-window and from left to right as in figure 4. 6. Media Policy Requirements All the requirements are based on having a privilege mechanism that authorizes users to access and manipulate the media policy data. 6.1 Genera Media Policy Requirements REQ-GP1: A participant MUST be able to specify its own unique topology. REQ-GP2: It must be possible for a group of users to receive the same mix. This mix may be a conference common mix. REQ-GP3:It MUST be possible to dynamically modify the number of contributing streams associated with a mixer. REQ-GP4: It MUST be possible to define the mixing function for each participant in the conference. REQ-GP6: It SHOULD be possible to send a participant multiple streams from one mixer. This requirement is to enable support for end- point mixing. REQ-GP7: It SHOULD be possible to define relationships between different mixers. The relationships can be time synchronized such as specifying that the audio mixer and video mixer is a pair to establish lip-synchs. REQ-GP8: It SHOULD be possible to define the number of different Even, et al. Expires August 24, 2003 [Page 10] Internet-Draft media policy February 2003 topologies and the number of streams in each of them that will be mixed in a mixer. For example the conference will support only one video topology that will go to all the participants, the video topology will support 2x2 display, or each participant will be able to receive his own audio topology that will include up to 4 contributing sources. 6.2 Video specific requirements Video is a bit different than audio when mixing is concerned. In multipoint video the common mixing modes are: Video switching where one of the contributing sources is sent to all participants, the video source may be forced by the media policy control protocol or may be dynamic by using for example a voice activated video switching mode where the participants will see the loudest speaker. "Continuous presence" or tiled windows display where the topology is composing one video stream that has a layout defining the shape and position of viewing windows that will be displayed to the participants. The layout includes N viewing windows so that in each of the windows there is one contributing stream. Even though the viewing windows can be of any shape we will address in this work only rectangular windows of any size. The windows may overlap. The section defines the specific requirements for media policy and media policy control to enable "Continuous presence" REQ-V1: It should be possible to define rectangular overlapping windows in a video mix. REQ-V2: It should be possible to map a stream to a window based on some mode like having one window display the loudest speaker or the floor holder while for the remaining windows fixed input streams are used. REQ-V3: It should be possible for authorized participants to change the layout of the video topology. REQ-V4: It should be possible for authorized participants to define the mapping of a stream to a window. 7. Security Considerations The media policy control protocol may enables unauthorized users to manipulate the media mixing of conferences, this may enable them to listen to conference or eject unsolicited media streams. The Even, et al. Expires August 24, 2003 [Page 11] Internet-Draft media policy February 2003 protocol should provide authentication of the users. The media policy data may include information about the sources and targets of mixer, if this information will be transferred in the protocol in the clear that may cause a security risk. The protocol should allow for encryption of the media policy transferred in the media policy control protocol. 8. Appendix I - Media Policy Operations The examples in sections 5.2 and 5.3 included the mixer blocks and their description. The appendix will try to list well known operations that can serve to define building blocks. Those operations will be part of the media policy syntax that is part of the media policy document. The operation appear here for information only. Each mixer is built from mixer blocks. Each block defines an operation on a set of streams, the operations can be a selection or a mixing operation and the result is a set of streams (One or more) depending on the operation. By collection a set of such operations we define a mixer in the media policy. The input and output of the operations are streams. The streams can be named by tagging individual m-lines in their SDP with an associated i-line or by specifying an alias in the contents of the current media policy description. Operations for selecting specific streams from a well-known group will also be defined. The appendix will list the operation specifying the input and output of each operation and will explain the semantics of the operation. 8.1 Stream selection operations Loudest(N) - This operation will receive as input a set of audio streams and will output N streams selected according to the volume levels starting from the highest LastLoudest - This operation will receive as input a set of audio streams and will select the audio stream of the participant that was the loudest speaker in the previous mix (This is the audio mix set before the current one) Not(Qualifier) - This operation will receive as input a set of audio streams and will select a stream that does not include a stream based on the Qualifier. The qualifier can be a stream name or alias or an operation. Example Not(Loudest) will select the stream that does not have the Loudest stream as one of its contributing sources. SelectMain- the operation will select from a set of input stream the Even, et al. Expires August 24, 2003 [Page 12] Internet-Draft media policy February 2003 stream that represent the main mix that include all contributing sources. The operation may be replaced by defining a well-known stream definition call MainAudio. In(transformation) - The operation will perform the operation on the incoming stream or set of streams. An example is In(0.4) will attenuate the incoming stream by 0.6 VofLoudest(N) - The operation will get as input a set of video streams and will select the N video streams of the participants according to their audio volume level. AssignToWID(N) - This operation will assign the incoming video stream to the subwindow N in a tiled video display. 8.2 Mixing Operation MixMinus(N) - The operation will mix the N incoming streams and will create N+1 output streams. The output streams will include one stream that is a mix of all input streams and N different streams that are a mix of all input streams without one of them. This is very useful for audio conferences where participants do not want to have their voice in the mix they receive from the conferencing bridge. Mix - The operation creates a single mix of all input sources. It has one output stream. Add - This operation gets two input sets of streams and add them to create the added mix. For example if one set has 2 stream2 and the second set has 3 streams the result will be 6 output streams. Each of the combined streams will be a mix of a stream from the first set with a stream from the second set. VideoMix(Layout) - the operation will take the incoming video streams and assign them according to their sub-window IDs to the layout. The layout may be predefined by a well known name or custom defined. The output stream will be the tiled video mix. 8.3 Well known streams MainAudioIn- this is the set of all contributing sources to the main audio conference. LoudestSpeaker(N) - This is the stream of the Nth loudest speaker. N=1 is the loudest. VideoLayout(N) - This is a tiled video layout. N specifies the table Even, et al. Expires August 24, 2003 [Page 13] Internet-Draft media policy February 2003 entry number that describes the specific layout. A table of all pre-defined video layout will be defined. References [1] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, A., Peterson, J., Sparks, R., Handley, M. and E. Schooler, "SIP: Session Initiation Protocol", RFC 3261, June 2002. [2] Rosenberg, J., "A Framework for Conferencing with the Session Initiation Protocol", draft- rosenberg-sipping-conferencing-framework-01 (work in progress), February 2003. [3] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model with Session Description Protocol (SDP)", RFC 3264, June 2002. [4] Rosenberg, J. and H. Schulzrinne, "A session initiation protocol (SIP) event package for conference state", , June 2002. Authors' Addresses Roni Even Polycom 94 Derech Em Hamoshavot Petach Tikva 49130 Israel EMail: roni.even@polycom.co.il Orit Levin RADVISION 266 Harristown Road Glen Rock NJ USA EMail: orit@radvision.com Nermeen Ismail Cisco Systems, Inc. 170 West Tasman Drive San Jose 95134 CA USA EMail: nismail@cisco.com Even, et al. Expires August 24, 2003 [Page 14] Internet-Draft media policy February 2003 Intellectual Property Statement The IETF takes no position regarding the validity or scope of any intellectual property or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; neither does it represent that it has made any effort to identify any such rights. Information on the IETF's procedures with respect to rights in standards-track and standards-related documentation can be found in BCP-11. Copies of claims of rights made available for publication and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementors or users of this specification can be obtained from the IETF Secretariat. The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights which may cover technology that may be required to practice this standard. Please address the information to the IETF Executive Director. Full Copyright Statement Copyright (C) The Internet Society (2003). All Rights Reserved. This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English. The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assignees. This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION Even, et al. Expires August 24, 2003 [Page 15] Internet-Draft media policy February 2003 HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Acknowledgement Funding for the RFC Editor function is currently provided by the Internet Society. Even, et al. Expires August 24, 2003 [Page 16]