A secure audio teleconference system

Abstract

Users of large communication networks often require a multi-party teleconferencing facility. The most common technique for providing secure audio teleconferencing requires the speech of each participant to be returned to clear form in a bridge circuit where it is combined with the speech of the other participants. The combined signal is then re-encrypted for distribution to the conferees. This introduces a security weakness as the bridge works with the clear speech and the cipher keys for all of the participants. In this paper we describe secure conferencing systems in which some of the bridge functions are distributed among the users so that no additional security weakness is introduced. The network conference bridge knows the addresses of the participants and receives and distributes the encrypted speech without modification. The conference system can be used with a number of encryption algorithms and the system is suitable for deployment on digital networks such as ISDN. High quality and robust secure voice communications can be provided with this technique.