Computer networks have long been heralded for their powerful communications capabilities. But this has been in the form of simple text messages, static images and pre-packaged audio and video. Computer technology has finally matured enough so the real-time audio and video benefits of video conferencing can be provided to everyone through their desktop computers.
Why do we want video conferencing? Visions of near-future societies are replete with video-based communications. But except for special circumstances, this would seem to be an unwelcome feature. Who really wants to see me playing Marathon or watching the All-Star game while I talk with them?
The real excitement of video conferencing comes with its potential to allow large numbers of people to participate in remote events. Seminars need no longer be limited by room size. Instead, you could hear Stephen Jay Gould's wisdom from the comfort of your dorm room. Professors would not have to be away from their classes to hear the latest exciting results from distant conferences.
The integration of audio and video broadcasts with computer technology might seem redundant. But it serves those who might want to monitor a broadcast while they work or those who have a computer but no television. It also allows for two-way interaction, something television and radio have yet to offer.
Most methods of video conferencing serve a limited audience at great expense. Participants at each site gather in a specially equipped room with limited seating. The communication is done over an expensive link, with the cost increasing as the distance between the sites increases. While such methods are appropriate for highly interactive tasks for which people have paid money, they lack the ability to easily serve large numbers of people at distant sites.
An experiment called Multicast Backbone (MBONE) on the Internet has for the past seven years been laying the technological foundation for future video conferencing efforts. MBONE utilizes current Internet infrastructure, but contributes little to the problem of traffic congestion due to its efficient distribution method and compression algorithms.
Most network connections use the unicast model in which there is one stream of data between each sender and recipient. For example, imagine a house with two phones where a new copy of a conversation has to be sent across the phone network in order for the second phone to be used. That additional network traffic is wasteful.
MBONE uses a multicast distribution model. Under this scheme, many people can participate using a single copy of the conversation. This is what actually happens when you pick up the second phone. No new connections need to be made, thus there is no additional network burden.
MBONE has been used to broadcast presidential speeches, shuttle missions, music and academic conferences. Until now, this exciting application has been limited to Unix workstations. Now that personal computers also have multicast capabilities, all of us can participate in the excitement of many world-wide events.