Semantic Domain Integration for Embedded and Hybrid SystemsΒΆ

Society is increasingly dependent on complex mission-critical engineered systems such as supervisory control and data access systems for power grid management, industrial control systems for automated manufacturing, and medical device systems for patient monitoring and treatment. The potential failure of these systems puts safety, health, and economic concerns of vital national interest in jeopardy. To protect these vital interests, it is crucial that these engineered systems maintain rigorous control over physical properties such as power flows, drug release rates, and spatial positioning. Furthermore, controlling these physical properties requires precise control over systemic properties such as communication and computation latencies, sensor sampling rates, and actuation response times. The system software that manages these engineered systems must monitor, evaluate, and respond to changes in the engineered system, while also coordinating computation, communication, sensing and actuation resources across heterogeneous and time-varying application requirements.

The current lack of integration among the following semantic domains limits the ability of system developers to exert precise control over physical and systemic properties of engineered systems: (1) application: application-specific quality of service (QoS) semantics required to ensure that high-fidelity control over the engineered system can be maintained; (2) system software: the QoS semantics of the system software components used to implement the application; (3) resource management: rigorous run-time resource management to ensure application-level QoS requirements can be met within the context of the system software QoS semantics; and (4) behavioral information: information about the observed run-time behavior of the system software and the engineered system itself. The problem that this research addresses is the disjointed manner in which these highly inter-dependent semantic domains are handled in the current state of the art, which limits the system developers’ ability to address key current challenges, such as preventing (or at least mitigating) cascading power grid system failures.

Our research group at the University of Kansas, in collaboration with researchers at Washington University in St. Louis and the University of Missouri-Rolla, is developing a revolutionary approach to system software for complex systems in which application QoS requirements, system software QoS semantics, resource management, and behavioral information are integrated through (1) mutually consistent formal and verifiable models of each semantic domain; (2) novel policies and mechanisms for exerting precise run-time control across semantic domains; and (3) detailed, efficient, and timely collection and dissemination of behavioral information to improve run-time control fidelity. Through rigorous integration of these semantic domains we aim to achieve a much greater correspondence among their respective semantics, and establish a foundation for revolutionary improvements in the state of the art, particularly for increases in system accuracy and reliability in producing desired behaviors (and in preventing undesired behaviors) in complex mission-critical engineered systems.

The main page for this project is at:

http://www.ittc.ku.edu/kusp/semantic_domains/index.html

NSF Support

This research is supported by the National Science Foundation, grant CCF-0615035.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Previous topic

Non-bypassable Kernel Services for Execution Security

Next topic

Kernel Testing

This Page