The configuration of campus networks is drastically changing in such a way that the existing FDDI is giving way to the ATM-LAN. Although ATM enables transfer speeds beyond 100Mbps, handles not merely data but voice and moving images as well and is believed to best suit the needs of today's multimedia communications, it is still partly incomplete as a technology. Technology to control traffic, especially burst traffic, and to control quality of service (QoS) of applications in an ATM environment needs to be developed quickly.

In April 1993, the National Center for Science Information Systems (NACSIS) and NTT Multimedia Networks Laboratories jointly started a five-year research and development project aiming for the high-speed multimedia communications network. NACSIS is a national institute which belongs to the Ministry of Education, Science and Culture providing the computer network backbone to national universities and laboratories. The main purpose of the project is to create a high-performance academic network which will serve researchers requiring advanced research environment by conducting research:

• to develop the future operation both of the high-performance LAN for campus networks,
• to demonstrate applications on high-performance communication networks, and
• to conduct experimental studies to evaluate those feasibility.

The project started with 4 core institutes (NACSIS, Waseda university, Institute of Industrial Science belonging to University of Tokyo and NTT) in April, 1993, and established a testbed connecting these 4 institutes in September, 1994. The testbed uses the high speed backbone network provided by NTT and each institute is connected to the testbed by one or several STM-1 circuits. Virtual paths connecting two institutes are assigned with different bandwidths according to applications' requirements.

The contributions of the research project are concerned primarily with the technological foundations of ATM networking and with some high-speed applications as follows:

• network architecture for multimedia services,
• element technologies for multimedia services, such as high-speed routing algorithms, high-speed protocols, resource control for multicast communications, etc.,
• operation and management of integrated ATM networks,
• QoS control schemes of multimedia applications,
• distributed image database system with hierarchical definition control,
• massive database system for satellite image,
• electronic library system, and
• multimedia telemedicine and telesurgery.

This paper outlines the activities of the project focusing on the new approach to guarantee end-to-end application QoS on ATM network.