Design Space Exploration Methodology Based on Decision Trees for Cyber-Physical Systems

As cyber-physical systems become increasingly complex, effective design space exploration becomes more and more challenging. Exploration parameters vary a lot between different architectures, while the design spaces are often large and complex. This work presents a systematic way for defining design spaces and it is based on a decision tree representation. The proposed methodology assists the effective exploration and the development of tools that support it. It is demonstrated in two different use cases in the context of cyber-physical systems.