In this paper, we introduce LargeSpace, the world's largest immersive display, and discuss the principles of its design. To clarify the design of large-scale projection-based immersive displays, we address the optimum screen shape, projection approach, and arrangement of projectors and tracking cameras. In addition, a novel distortion correction method for panoramic stereo rendering is described. The method can be applied to any projection-based immersive display with any screen shape, and can generate real-time panoramic-stereoscopic views from the viewpoints of tracked participants. To validate the design principles and the rendering algorithm, we implement the LargeSpace and confirm that the method can generate the correct perspective from any position inside the screen viewing area. We implement several applications and show that large-scale immersive displays can be used in the fields of art and experimental psychology.

The development of a mobile network to provide intelligent transport system (ITS) services is underway as a national project in Japan called the "Smart Gateway", responding to nationwide needs for such a network. System issues for implementation include the necessity of measures for phased system expansion and efficient system maintenance. In addition, a data transmission method must be established to provide uninterrupted communication across microcells following radio signal environment changes to deliver information to fast-moving vehicles. To cope with these issues, this system uses an autonomous decentralized system architecture and supports online extendibility, online maintainability and fault tolerance that are critical to set up a system that operates non-stop. Thus, the authors developed microcell mobility management to enable uninterrupted communication across microcells. The authors developed a prototype incorporating the above methods and made an evaluation.

We propose an autonomous group-management model where each node autonomously judges whether to work with other nodes and dynamically forms a group to complete a requested job. This model was developed for mobile systems that consist of mobile terminals and base stations (BSs). The BSs are connected by a network and each BS judges if it should form a group with other BSs based on their physical locations and the type of job without relying on direction from the other BSs. Therefore, each BS is "autonomous". The BSs dynamically form local groups without any centralized manager share information, and complete a job, so this system is well suited to highly distributed systems where high performance is necessary. An intelligent transport system (ITS) is a typical application that matches this model. Various information is provided front BSs to moving cars in an ITS. Such a system will cover a wide area (all of Japan) and will include a huge number of BSs. High performance is required because many cars move at high-speed. We applied our proposed model to a smart gateway (SG) system - one of the ITSs promoted by the Telecommunication Advancement Organization of Japan. We developed a new type of BS according to this model and installed a series of such BSs along roads for our evaluation of the system.