Colin Perkins
                                                             USC/ISI

                   RTP Interoperability Statement

              draft-ietf-avt-rtp-interop-06.txt

                 draft-ietf-avt-rtp-interop-07.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.

Distribution of this document is unlimited.

Comments are solicited and should be addressed to the author and/or the
IETF Audio/Video Transport working group's mailing list at rem-conf@es.net.

                              Abstract

     It is required to demonstrate interoperability of RTP implementations
     in order to move the RTP specification to draft standard.  This memo
     outlines those features to be tested, as the first stage of an
     interoperability statement.

1   Introduction

The Internet standards process [1] places a number of requirements
on a standards track protocol specification.   In particular, when
advancing a protocol from proposed standard to draft standard it

is necessary to demonstrate at least two independent and interoperable
implementations, from different code bases, of all options and features
of that protocol.   Further, in cases where one or more options or
features have not been demonstrated in at least two interoperable
implementations, the specification may advance to the draft standard
level only if those options or features are removed.   The Real-time
Transport Protocol, RTP, was originally specified in RFC1889 as a
proposed standard [2].   The revision of this specification for draft
standard status is now well underway, so it has become necessary
to conduct such an interoperability demonstration.

This memo describes the set of features and options of the RTP specification
which need to be tested as a basis for this demonstration.   Due to the
nature of RTP there are necessarily two types of test described: those
which directly affect the interoperability of implementations at a ``bits
on the wire level'' and those which affect scalability and safety of the
protocol but do not directly affect interoperability.  A related memo [4]
describes a testing framework which may aid with interoperability testing.

This memo is for information only and does not specify a standard
of any kind.

2   Features and options required to demonstrate interoperability

In order to demonstrate interoperability it is required to produce
a statement of interoperability for each feature noted below.   Such
a statement should note the pair of implementations tested, including
version numbers, and a pass/fail statement for each feature.   It
is not expected that every implementation will implement every feature,
but each feature needs to be demonstrated by some pair of applications.

Note that some of these tests depend on the particular profile used,
or upon options in that profile.   For example, it will be necessary
to test audio and video applications operating under [3] separately.

  1. Interoperable exchange of data packets using the basic RTP header
     with no header extension, padding or CSRC list.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

  2. Interoperable exchange of data packets which use padding.

       o  PASS: IP/TV vs rat
  3. Interoperable exchange of data packets which use a header extension.
     There are three possibilities here:   a) if both implementations
     use a header extension in the same manner, it should be verified
     that the receiver correctly receives the information contained
     in the extension header; b) If the sender uses a header extension
     and the receiver does not, it should be verified that the receiver
     ignores the extension; c) If neither implementation implements
     an extended header, this test is considered a failure.

       o  PASS: jrtplib-2.4 vs UCL RTP library v1.2.2

  4. Interoperable exchange of data packets using the marker bit as
     specified in the profile.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vic

  5. Interoperable exchange of data packets using the payload type
     field to differentiate multiple payload formats according to
     a profile definition.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

  6. Interoperable exchange of data packets containing a CSRC list.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

  7. Interoperable exchange of RTCP packets, which must be compound
     packets containing at least an initial SR or RR packet and an
     SDES CNAME packet.   Other RTCP packet types may be included,
     but this is not required for this test.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

  8. Interoperable exchange of sender report packets when the receiver
     of the sender reports is not also a sender (ie:   sender reports
     which only contain sender info, with no report blocks).

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

  9. Interoperable exchange of sender report packets when the receiver
     of the sender reports is also a sender (ie:   sender reports
     which contain one or more report blocks).

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic (IP/TV never sends SR with report
          blocks, but does successfully receive them from vic/vat).

 10. Interoperable exchange of receiver report packets.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 11. Interoperable exchange of receiver report packets when not receiving
     data (ie:   the empty receiver report which has to be sent first
     in each compound RTCP packet when no-participants are transmitting
     data).

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 12. Interoperable and correct choice of CNAME, according to the rules
     in the RTP specification and profile (applications using the
     audio/video profile [3] under IPv4 should typically generate
     a CNAME of the form `example@10.0.0.1', or `10.0.0.1' if they
     are on a machine which no concept of usernames).

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 13. Interoperable exchange of source description packets containing
     a CNAME item.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 14. Interoperable exchange of source description packets containing
     a NAME item.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 15. Interoperable exchange of source description packets containing
     an EMAIL item.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 16. Interoperable exchange of source description packets containing
     a PHONE item.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic
 17. Interoperable exchange of source description packets containing
     a LOC item.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 18. Interoperable exchange of source description packets containing
     a TOOL item.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 19. Interoperable exchange of source description packets containing
     a NOTE item.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 20. Interoperable exchange of source description packets containing
     a PRIV item.

       o  FAIL: need to test rtplib against rtpdump?

       o  FAIL: Ericsson have an implementation, Magnus Westerlund
          will test against rat.

 21. Interoperable exchange of BYE packets containing a single SSRC.

       o  PASS: rat vs vat

       o  PASS: IP/TV vs vat/vic

 22. Interoperable exchange of BYE packets containing multiple SSRCs.

       o  FAIL: rat can send these, but vat only accepts the first
          SSRC

       o  FAIL: IP/TV sends only one SSRC in BYE, but should accept
          multiple

       o  FAIL: need to test rat-3.0.x against rtplib

 23. Interoperable exchange of BYE packets containing the optional
     reason for leaving text.

       o  PASS: tested IP/TV sending to vat.   Also rtplib generates
          and displays them.

 24. Interoperable exchange of BYE packets containing the optional
    reason for leaving text and multiple SSRCs.

     o  FAIL: does anyone implement both?

25. Interoperable exchange of application defined RTCP packets.   As
     with the RTP header extension this test takes two forms:   if
     both implementations implement the same application defined packet
     it should be verified that those packets can be interoperably
     exchanged.   If only one implementation uses application defined
     packets, it should be verified that the other implementation
     can receive compound RTCP packets containing an APP packet whilst
     ignoring the APP packet.   If neither implementation implements
     APP packets this test is considered a failure.

       o  PASS: jrtplib-2.4 vs UCL RTP library v1.2.2

26.

 25. Interoperable exchange of encrypted RTP packets using DES encryption
     in CBC mode.

       o  PASS: rat vs vat

27.

 26. Interoperable exchange of encrypted RTCP packets using DES encryption
     in CBC mode.

       o  PASS (sort of):   rat vs vat (vat gets the padding wrong
          in some cases, but mostly it works).

28. Interoperable exchange of encrypted RTCP packets using DES encryption
    in CBC mode, when those compound RTCP packets have been split
    into an encrypted packet and an unencrypted packet.

     o  FAIL: not tested (rtplib supports this?)

3   Features and options relating to scalability

In addition to the basic interoperability tests, RTP includes a number of
features relating to scaling of the protocol to large groups.  Since these
features are those which have undergone the greatest change in the update
of the RTP specification, it is considered important to demonstrate their
correct implementation.   However, since these changes do not affect the
bits-on-the-wire behaviour of the protocol, it is not possible to perform a
traditional interoperability test.  As an alternative to such testing we
require that multiple independent implementations complete the following
demonstrations.

  1. Demonstrate correct implementation of basic RTCP transmission
     rules:   periodic transmission of RTCP packets at the minimum
     (5 second) interval and randomisation of the transmission interval.

       o  PASS: rat, IP/TV

  2. Demonstrate correct implementation of the RTCP step join backoff
     algorithm as a receiver.

       o  PASS: rat, rtplib
  3. Demonstrate correct implementation of the RTCP step join backoff
     algorithm as a sender.

       o  PASS: rat, rtplib

  4. Demonstrate correct steady state scaling of the RTCP interval
     acording to the group size.

       o  PASS: rat, IP/TV

  5. Demonstrate correct steady state scaling of the RTCP interval
     acording to the group size with compensation for the number of
     senders.

       o  PASS: rat, IP/TV

  6. Demonstrate correct implementation of the RTCP reverse reconsideration
     algorithm.

       o  FAIL: rat is correct, need another implementation

       o  FAIL: Ericsson have an implementation:   Magnus Westerlund
          is testing...

  7. Demonstrate correct implementation of the RTCP BYE reconsideration
     algorithm.

       o  FAIL  FAIL: Ericsson have an implementation:   Magnus Westerlund
          is testing...

  8. Demonstrate correct implementation of the RTCP member timeout
     algorithm.

       o  FAIL  PASS: IP/TV, vat, rat

       o  FAIL: Ericsson have an implementation:   Magnus Westerlund
          is testing...

  9. Demonstrate random choice of SSRC.

       o  PASS: rat, IP/TV, LiveCaster

 10. Demonstrate random selection of initial RTP sequence number.

       o  PASS: rat, LiveCaster

 11. Demonstrate random selection of initial RTP timestamp.

       o  PASS: rat, LiveCaster

 12. Demonstrate correct implementation of the SSRC collision/loop
     detection algorithm.

       o  PASS: IP/TV, vat
 13. Demonstrate correct generation of reception report statistics
     in SR/RR packets.

       o  PASS: rat, IP/TV

 14. Demonstrate correct generation of the sender info block in SR
     packets.

       o  PASS: rat, IP/TV

4   Author's Address

Colin Perkins
USC Information Sciences Institute
4350 North Fairfax Drive
Suite 620
Arlington, VA 22203
USA

Email:   csp@isi.edu

5 Acknowledgments

Thanks to Steve Casner, Jonathan Rosenberg and Bill Fenner for their
helpful feedback.

6   References

[1] S. Bradner, ``The Internet Standards Process -- Revision 3'',
    RFC2026, Internet Engineering Task Force, October 1996.

[2] H. Schulzrinne, S. Casner, R. Frederick and V. Jacobson, ``RTP: A
    Transport Protocol to Real-Time Applications'', RFC1889, Internet
    Engineering Task Force, January 1996.

[3] H. Schulzrinne, ``RTP Profile for Audio and Video Conferences with
    Minimal Control'', draft-ietf-avt-profile-new-08.txt, January 2000.

[4] C. S. Perkins, J. Rosenberg and H. Schulzrinne, ``RTP Testing
    Strategies'', draft-ietf-avt-rtptest-04.txt, November 2000.