Tohru Hoshi

B-ISDN multimedia communication and collaboration platform using advanced video workstations to support cooperative work

A distributed real-time computer-supported cooperative work (CSCW) platform called the B-ISDN group teleworking system is described. The system combines the high-speed multimedia communication capability with advanced video integrated workstations. The aim of this system is to support ad hoc meetings and real-time collaborative work between small groups of people who are geographically distributed. The main features of the system are face-to-face communication and the capability for sharing and comanipulating shared multimedia information including video. The concept and architecture of the system are described. A prototype group teleworking system that uses B-ISDN multimedia workstations is also described. A mock remote medical consultation application for the system is presented. >

Proposal of a method of voice stream multiplexing for IP telephony systems

IP telephony systems are expected to be deployed worldwide in the near future because of their potential for integrating the multimedia communication infrastructure over IP networks. Phone-to-phone connection over an IP network via IP telephony gateways (IP-GWs) is a key feature of the system. In an IP telephony system, a low-bit-rate voice codec is used to improve bandwidth efficiency. However, due to the packet transfer method over the IP network, it is necessary to add packet headers, including IP, UDP, and RTP headers, which increases the header overhead and thus decreases transfer efficiency. Moreover, because there will be large numbers of short voice packets flowing into the IP network, the load on the Internet will increase. We propose voice stream multiplexing between IP-GWs to solve these problems. In this scheme, multiple voice streams are connected between a pair of IPCWs, enabling multiplexed voice stream transfer. The voice stream multiplexing mechanism can reduce the header overhead as well as decrease the number of voice packets. The voice stream multiplexing we propose is to concatenate RTP packets destined for the same IP-GW at a multiplexing interval period into a single UDP packet. The advantage of this method is that no new additional header is required and the current well-defined H.323 and RTP standards can be applied with minimum changes. We implemented and tested the system.