All things related to multicast

["Richard Mavrogeanes" <>]

Folks,

I need your help. But first some background:

On or about July 25th, we will begin broadcasting live MPEG-2 and MPEG-4 
audio/video from a research ship in the Black Sea.  This is part of a $5M 
project to explore the fantastic underwater treasures of human history that 
have been preserved under hundreds of meters of oxgen-free water.  The 
research is being conducted by Dr. Bob Ballard (who discovered the Titantic, 
PT-109, etc), and is sponsored by the National Geographic among others.  You 
can learn more about the expedition at www.expedition2003.org.

The significant part of this project is that viewers will be able to witness 
the discoveries live, as they occur.  The video is sourced from the ship 
deck, but more importantly from the Remote Vechicle's cameras as it makes 
discoveries.  This promises to be truly compelling, although I would 
anticipate long periods of "not much".

This event has made national press, and I suspect will be on the evening news 
at some point during the project (July25 - August 25).  Dr. Ballard promotes 
the Internet-2 in almost every talk he gives. As a result, the multicast 
traffic on Abilene may reach record levels (or not).  

The video signals originate at the ship, are sent via satellite to an earth 
station in Portland Maine, travel via DSL to Plano TX (EDS), then travel via 
I2 to the University of Rhode Island where they are displayed on a specially 
built console.  The video signals are then encoded into MPEG-2 at 6Mbps and 
MPEG-4 at 200 Kbps by a VBrick and presented to the URI campus network and 
Abiliene.

Several institutions have built large venue theaters to exhibit the live 
video.  Others are using StreamPlayer to view the live videos at their 
desktops.  In theory, virtually every I2 participant should be able to view 
the events live, but we know that the penatration of multicast in the campus 
networks is not what it might be.

The videos are sent via multicast, and I can see it listed at several of the 
SDR monitors . As you would imagine, this is not without issues since the 
source campus network was not configured for IGMP (CGMP in this case).  I do 
not have I2 access in my office, so I cannot monitor the video in real time.

As a matter of interest, the MPEG-4 video is also sent from the VBrick via 
RTSP unicast on the public Internet, and will appear on the project web site 
during the mission.  

For the last few weeks, we've been multicasting on  233.56.12.14 and 
233.56.12.114, and will may bring up another stream next week.  You can see 
the video as "Black Sea Project" in both MPEG-2 and MPEG-4, and you can get a 
free player (if you don't already have it) from the VBrick web site. 
Currently, the video is just a very cheap camera pointed out the window, but 
it will switch to the Black Sea audio/video on July 25th. As of this message, 
we are not sending but it should be back up on Monday although it will up and 
down as we make changes during the next few days.

Now for my request for help. I'm looking for someone to work with with good 
access that I can call with a regular "can you see it now?".  On Monday, for 
example, the the good folks at URI (who are unfamilar with multicast) wish to 
prove to themselves that the bandwidth requirement from their campus stays 
the same regardless of the number of viewers (as I said, unfamilar with 
multicast).  Beyond this, it would be most useful if I could find someone 
with the time to assist others troubleshooting.  I am handicapped by not 
having local I2 access, and further handicapped by marginal backbone 
knowledge and familarity with the tools.

Thanks in advance,

Rich Mavrogeanes

VBrick Systems

1-203-303-0200