CSR: Small: Network-Level Design of Cyber-Physical Systems

点击翻译按钮获取中文摘要

Cyber-physical systems (CPS) are engineered systems that are built from, and depend upon, the seamless integration of computational algorithms and physical components. CPS promise to revolutionize how we approach global societal-scale problems, through application of machine intelligence and autonomy in areas such as energy, healthcare, and transportation. A key element of the CPS promise depends on achieving effective networking of hitherto disconnected computer systems. However, the networked aspect of CPS also poses new design challenges. Machine-to-machine communication comes with inherent uncertainties, such as data losses, yet CPS have to perform under tight real-time, reliability, safety, efficiency and performance guarantees. When combined with rapidly growing increases in complexity, traditional approaches in which computation and communication aspects of such networked systems are designed separately, and in an ad-hoc manner, hinder achieving system-wide correctness guarantees and leave a large optimization potential untapped. The long-term goal of this project is to close this gap by investigating novel, systematic, and ultimately automated design approaches that enable network-wide optimization and validation of CPS across computation and communication boundaries. This is expected to advance how future intelligent and autonomous cyber-physical systems are designed, enabling whole new classes of smart applications to emerge.The foundation for any rigorous design process is a sound formalization with well-defined abstractions and models that allow users or automated tools to reason about intended behavior before a system is built. While the integration of cyber and physical models has been studied, comparable theories for a unification of computation and communication are lacking. A key question is how application-specific notions of real-time behavior and quality-of-service can be provided for overall optimized implementations in the presence of inherent uncertainty, adaptivity and reactivity in networked and distributed CPS. This project aims to answer such questions by exploring formal foundations for a network-level computation and communication co-design science for cyber-physical systems. Specific innovations include: formal models of computation and communication for system specification, including associated analysis and synthesis algorithms; fast yet accurate performance models for rapid, early validation, including associated model generation, synthesis and design space exploration capabilities. These innovations are evaluated in the context of specific cyber-physical system case studies, including virtual and physical testbeds. Research is combined with an educational agenda aimed at establishing a holistic view of networking and embedded computing, including public release of models and tools as they are developed, as well as development of online teaching materials for a broad audience.