Requirements of a Model-Driven Methodology and Tool-Chain for the Design and Verification of Hierarchical Controllers of an Autonomic Network

Model-driven approaches to design, simulation, verification, validation and testing of complex systems are increasingly required in every domain of systems and software engineering where the systems are complex. The domain of autonomic network engineering is one such new area calling for such model-driven approaches. In this paper we provide the requirements of a model-driven methodology and Tool-Chain for the design and verification of hierarchical controllers of an autonomic network. The requirements are classified as behavioral, structural and stability requirements. Based on these requirements, we propose a hybrid approach that combines design techniques from modern control theory, conventional protocol design and model-driven systems engineering for the design of hierarchical controllers that meet the stability, structural and behavioral requirements of their autonomic network architectures. The methodology addresses the problems of behavior conflicts, stability, response time problems and operating regions conflicts of hierarchical controllers. A Tool-Chain to realize the proposed methodology is also presented.