Improved resource allocation algorithms for practical image encoding in a ubiquitous computing environment

As a case study of the ubiquitous computing system, we have implemented a prototype for the JPEG encoding application. In order to achieve this eventual development in the real world, we studied resource allocation policies that can improve the overall performance of the system. In this paper, we consider those static and dynamic allocation approaches and then propose four different allocation algorithms. In particular, we extensively studied the dynamic allocation algorithms by exploring various cache policies which include disabled cache, unrestricted cache and restricted cache. Performance of these algorithms in large scale application scenario is also evaluated based on both the improved prototype and a simulation environment. The experimental results show a significant performance improvement achieved by the new proposed algorithms in terms of load balance, execution time, waiting time and execution efficiency.