A method and apparatus for cost reduction of telephony conference using VoIP
Original Publication Date: 2002-Jul-01
Included in the Prior Art Database: 2003-Jun-19
The method telephony conferencing works today is based on a telephony conferencing bridge. Every participant in the conference, calls the bridge, and when connected he is prompted to dial the identification of the conference session he wants to participate. In principal the conference bridge gets the audio streams of all participants in the same conferencing session, mixes the streams together, and sends them back to each of the participant. By doing so each participant hears what everybody else in the conference is saying, and when number of participants talk at the same time, each participant hears them simultaneously too, since their voice streams are being mixed. In many conferences the participants split themselves naturally into two classes, those that actively participate in the discussion, and those which are passive listeners. From time to time depending on the content of the discussion, participants move from between the two classes. Participants belonging to the listeners class, will normally put their phone on mute since they want only to listen, and they don’t want their normal office noises to interfere with the conference. Thus in the situation today all of the conference participants have to pay for a phone conversation to the conferencing bridge (which may be an overseas call) even if they want just to listen. In this invention we propose to reduce the cost of a conference by allowing a listening only participants to participate in the conference using Voice over IP (VoIP). Only when such participant will want actively to participate in the conference he will switch from IP line to a PSTN line. The advantage of this method is that while a participant is in a listening mode only, he can get a good quality of voice, without having to pay for a phone call. A good quality may be reached because the participant is not engaged in an interactive call. In this situation the audio of the other participants can be stream to him with somewhat higher delay to compensate for communication lines bandwidth variation, and jitter. When a passive participant detects that he want to become an active participant, he will dial to the conferencing bridge, and patch in to the conference on a PSTN line. When a participant will want to become a passive participant again he will disconnect from the PSTN line which will switch him back to an IP line. For ISDN lines the dial and connect process is in the order of half a second. For regular PSTN lines this may take a couple of seconds. This means that for non PSTN lines a passive participant has to plan a head of time (in the order of seconds) his transition to an active participant. As already said above to achieve good voice quality, the audio streamed on IP communication lines is with some delay (2-3secs) with regards to the audio streamed on PSTN lines. Thus in a simplest implementation, when switching to PSTN, these couple of seconds of conversation will be lost. More sophisticated algorithms nay be deployed to artificially speed the audio streaming at the beginning if the switching to the PSTN to compensate for this lose.