Abstract
In this paper, we present an algorithm for minimizing the processing time of a video sequence on a handheld-based distributed computing system. The algorithm considers the energy use and the remaining energy of the nodes, and steers a Monte-Carlo-based allocation according to the following rationale: images should be mapped more likely onto those nodes that communicate and process faster, consume less energy, and possess more energy reservoirs. Simulation results show that these simple ideas allow the allocation algorithm to deal efficiently with the fundamental tradeoff between turnaround time and energy.