Tomoharu Imai

An adaptive desktop transfer protocol for mobile Thin Client

On a Thin Client system, the user manipulates the server system and sees remote desktop images which are transferred over a network. However, it is difficult to use Thin Client systems on a mobile network, because the network bandwidth is insufficient to transfer the remote display image with sufficient frequency. In this study, we implement and evaluate an improved remote desktop system for mobile users that can detect a high motion zone (on the desktop), and automatically encode the detected zone into a video stream. Also we propose the scene detection method, which can optimize reducing the network bandwidth consumption depending on the situation. As a result, we show reducing the bandwidth consumption for the user can use Thin Client system on a 3G(HSPA)-like network offered narrow bandwidth comfortably.

Network Data Buffering for Availability Improvement of Mobile Web Applications

In recent years, we have seen an explosion in the use of smart devices. With several different competing OS platforms on these smart devices, developers have turned to web applications as an effective way to provide cross-platform services. However, because many web applications are designed to handle UI interactions locally on a user’s device and to process data remotely in the cloud, it is difficult for them to continue running while offline. In this paper, we propose a data synchronization technology that buffers and minimize network communications to address problems associated with dropped network connections as well as low bandwidth and/or high latency environments. To test the technology’s effectiveness, we applied it to some typical web applications and compared their performance in environments with dropped connections.