CSU EAST BAY
DEPARTMENT OF MATHEMATICS AND
COMPUTER SCIENCE
THESIS PRESENTATION
Tuesday, June 7, 2005; 12:00 Noon Sc S105C
Speaker: Yipkei Kwok, Candidate for the M.S. Degree in Computer Science
The popularity of embedded systems is growing rapidly. Due to the ^M increase in application complexity that causes extensive CPU and memory usage, system devices are more power-consuming than ever. Since most of the embedded systems are operated under limited power supply, energy optimization has become a popular research area.
In this work, we introduce the Smart Caching approach that cooperates ^M with the Direct Memory Access (DMA) technique to taper process memory footprints. In addition, we propose the Process Rearrangement app roach that lowers the number of node activations by optimizing process execution patterns. Consequently, more memory nodes can be put i nto a power-saving mode. As a result, the battery-powered supply of an embedded device can last longer.
Experiments were conducted on a simulated operating system. Analysis and comparisons were made to demonstrate the effectiveness of our technique.