Partitioning metrics for improved performance and economy of distributed embedded systems

This paper describes some metrics which can be applied to optimize the constrained partitioning of very large distributed software/hardware systems. These metrics are tailored for software component models for UML specifications. The use of clear defined metrics allows us to capture more aspects of the design (abstraction level, aggregation and causality relations between the components). Besides, the appropriate formalization of these metrics has a great impact on the results that can be obtained from the partitioning regarding both the performance and the economy of the system under design.