Development and evaluation of processor farm algorithms/softwarefor robot sensing applications

摘要

Using a parallel multi-transputer pipeline hardware architecture,this paper analyzes the processor farm paradigm for robot sensingapplications. Algorithms for parallel search in a database, parallelsorting and parallel image processing were developed. We show how theprocessor farm is a dynamic load balancing software depending on factorswhich are automatically determined by the system. Considering softwaremodifications in the farmer, like the activity of the sender or the useof buffers, we have evaluated the performance of several variants offarm algorithms. The results, obtained in real situations, give usindications for the best choice of the farmer software and the optimumnumber of worker processors