Re: research in multimedia communication

Stephen Casner <CASNER@isi.edu> Tue, 02 June 1992 01:06 UTC

Received: from nri.nri.reston.va.us by ietf.NRI.Reston.VA.US id aa04898; 1 Jun 92 21:06 EDT
Received: from nri.reston.va.us by NRI.Reston.VA.US id aa09344; 1 Jun 92 21:13 EDT
Received: from WILMA.BBN.COM by NRI.Reston.VA.US id aa09337; 1 Jun 92 21:13 EDT
Received: from wilma by WILMA.BBN.COM id aa24203; 1 Jun 92 21:01 EDT
Received: from BBN.COM by WILMA.BBN.COM id aa24199; 1 Jun 92 21:00 EDT
Received: from venera.isi.edu by BBN.COM id aa01131; 1 Jun 92 21:01 EDT
Received: by venera.isi.edu (5.65c/5.65+local-6) id <AA20807>; Mon, 1 Jun 1992 18:01:39 -0700
Date: Mon, 01 Jun 1992 18:01:37 -0000
From: Stephen Casner <CASNER@isi.edu>
Subject: Re: research in multimedia communication
To: browne@cs.hope.edu, cip@bbn.com
Cc: raj@ms.uky.edu
Message-Id: <707446897.0.CASNER@ISI.EDU>
In-Reply-To: <9205192019.AA06933@cs.hope.edu>
Mail-System-Version: <VAX-MM(219)+TOPSLIB(128)@ISI.EDU>

The COIP working group is mostly inactive at this point.  The ST-II
specification was completed and released as RFC 1190.  The cip list is
maintained at BBN but I'm not sure by whom.

There are two new working groups related to multimedia conferencing
that may be of interest.  They are the Audio/Video Transport WG that I
chair, and the Remote Conferencing Architecture BOF (to become a WG)
that Ari Ollikainen and Jack Drescher chair.  AVT is addressing the
real-time "transport" level protocols to go above UDP or ST.  One of
the architecture issues to be discussed in the Rem-Conf group is
connection management.  Both groups use the email list
rem-conf@es.net, which you may join by sending a message to
rem-conf-request@es.net.

Experiments are underway in a number of places.  Our work at ISI,
along with several other groups, is based on the DARTnet experimental
testbed network sponsored by DARPA.  We demonstrated packet audio and
IP multicast software from DARTnet at the last IETF by "audiocasting"
live audio from portions of the IETF meeting.  At this point, some of
the protocols are ad-hoc; that's what the IETF WG's are about.  That
software runs on SPARCstations:

  - VT version 1.41.  This implements NVP-II over the original Stream
    Protocol (ST), all in user space.  Available by anonymous FTP from
    venera.isi.edu in the file pub/vt.141.tar.Z.

  - VT version 2.20.  This version does not use the control protocol
    part of NVP-II because that is to be redesigned, but it does use
    the simple data protocol part, either over UDP and IP Multicast,
    or over ST-II, the new version of the Stream Protocol (RFC 1190).
    Available from venera.isi.edu in the file pub/vt.220.tar.Z.  To
    use either IP Multicast or ST-II, these functions must be added to
    the SunOS 4.1.1 kernel, but this can be done without kernel source
    files.  For IP multicast, from gregorio.stanford.edu pick up the
    file vmtp-ip/ipmulti-sunos411.tar.Z.  You can get an ST-II package
    from clynn.bbn.com in the file pub/STII_SOURCE.Z.

  - NEVOT, from University of Massachusetts.  Interoperates with VT in
    either the UDP/IP Multicast or ST-II mode, but provides a much
    nicer X-Windows user interface.  Available by anonymous FTP from
    gaia.cs.umass.edu in pub/nevot-0.9.tar.Z (there may be a newer
    version).

   - vat, from Lawrence Berkeley Laboratory.  Also interoperates with
    VT, but only in UDP/IP Multicast mode.  Also X-based.  No source
    available yet.  Binary available from ftp.ee.lbl.gov in the file
    vat.tar.Z.

Much of the work on DARTnet is on traffic control (resource
management) algorithms.  Experiments are still underway, and there are
not many results published yet.

To collaborate on ST-II in particular, I suggest you contact Karen Seo
(kseo@bbn.com) or Charles Lynn (clynn@bbn.com); they may already have
seen your message on the cip list.
						-- Steve Casner
-------