Guest editorial: Future voice technologies

With recent technological advances and accompanying progress in standardization of telecommunication interfaces, the promise of widespread availability of goodquality packetized voice service should be fulfilled soon. The papers appearing in this issue cover a wide range of aspects of packet voice technologies including new ideas on performance enhancements, quality improvements, network dimensioning, speech recognition, and telephony architecture. The international Telecommunications Union (ITU) has already standardized the transport of constant bit rate (CBR) voice over an ATM packet network via an adaptation protocol called AAL1 (ATM adaptation layer type 1). The current standardization effort concentrates on the development of protocols, AAL2, that allow the multiplexing of packet streams from variable bit rate (VBR) voice and data sources into a single ATM virtual circuit. The basic aspects of AAL2 are defined in Recommendation I.363.2, and the voice aspects are defined in draft Recommendation I.366.2. In particular, the latter standard provides an implicit time stamp, by sequence numbering the packets of each voice stream, which can be used by the receiver to mitigate the reconstruction anomalies resulting from lost and excessively jittered packets.