DAISy Dynamic Ambient Intelligent Systems Unit

This paper deals with the effective representation of flow graph (DFG) models for distributed real- time applications. To this end, we first present the real- time data flow graph (RDFG) model, which is a modification of the traditional DFG model that will be using to cope with the special requirements of distributed real- time tasks. Then we present a new technique for RDFG compression, which can also be applied to DFGs. Using it, CASE tools based on the DFG model can benefit both in storage and speed by transforming the graph to a more compact form. The algorithm consists of three phases applied consecutively to a DFG that has been previously transformed to a task tree. The whole process has been developed in the context of SEPDS project, which aims at producing an experimental CASE tool based on the DFG model. Finally, we present the conclusions draw from the analysis of the computational complexity of this technique, together with future directions of our research on the subject.