Optimization methods for the performance enhancement of thin client computing over bandwidth-limited and slow links

Abstract

Thin client computing is gaining more and more popularity in a wide range of industries. Nevertheless, to deliver a satisfactory user experience, especially over the bandwidth-limited network links, performance enhancement of existing thin client systems is still essential. Analysis shows that there is still a large amount redundancy in the screen updates generated by existing thin client systems, and improving the interactive performance of thin client computing by reducing the redundancy is still practicable. Towards this direction, we proposed a few optimization techniques for thin-client systems. These techniques include a static object cache technique, a long-distance redundancy reduction scheme, and a data spike reduction scheme. Our experimental results show that these techniques can considerably reduce network traffic and the latency of user operations over bandwidth-limited links with reasonable cache memory and slightly more computing time.