A partition methodology to develop data flow dominated embedded systems

Abstract

This paper proposes an automatic partition methodology oriented to develop data flow dominated embedded systems. The target architecture is CPU-based with reconfigurable devices on attached board(s), which closely matches the PSM meta-model applied to system modelling. A PSM flow graph was developed to represent the system during the partitioning process. The partitioning task applies known optimization algorithms - tabu search and cluster growth algorithms - which were enriched with new elements to reduce computation time and to achieve higher quality partition solutions. These include the closeness function that guides cluster growth algorithm, which dynamically adapts to the type of object and partition under analysis. The methodology was applied to two case studies, and some evaluation results are presented.