On the problem of specifying the number of floors for a voice-only conference on packet networks

Voice conferencing is an essential block of any multimedia system used for collaborative work. In a collaborative environment floor control is an important issue that is dealt by many; yet fixing the number of floors is an open problem. In an audio conference, mixing streams from too many concurrent speakers degrades the voice quality. Therefore setting an upper bound for the number of streams (floors) that may be mixed is sine qua non for providing quality conferencing service. We address the problem of setting the upper bound on number of floors for a system meant to support concurrent multiparty audio sessions on top of IP multicasting. A measure called "Loudness Number" that is used to manage the number of floors is briefly outlined to the extent of making this paper self-contained. Our implementation at a functional level on a campus-wide network of Windows/sup /spl copy// systems has yielded satisfactory performance.