Submitted: February 2016
Abstract
This thesis presents an extension of Scenario-Aware Dataflow (SADE), a modeling formalism based on SDF, which is commonly used for signal processing and multimedia applications. The presented variant extends the set of possible processing times and introduces a feature to measure the energy consumption of systems. We show the model’s formal definition and its semantics based on Stochastic Timed Automata. Additionally, we discuss the differences to other variants of the formalism.
As part of this thesis, an implementation of the formalism was developed. It is part of the Modest toolset, a model checking tool to analyze various kinds of automata networks. We discuss the implementation in detail and analyze its performance based on various case studies.