Active System Characterization and Control
It is very difficult to understand how systems actually behave based on design documents and models. In addition, systems are generally composed of legacy systems, subsystems and components which interact in complex and often unforeseen ways, sometimes using many "undocumented features": The complexity of these systems makes it exceedingly difficult to predict which and when these interactions are critical and significantly impact realized behavior. Transient behaviors, which are generally difficult to understand, tend to dominate the interesting behavior of the system. Finally, with the introduction of the human element, systems may be used in ways which were not envisioned by the developers. However, for new capabilities and features to be added to an existing system, it is often critical to understand its current behavior. While model-based engineering can be made to be quite effective, it is only as good as the models that it uses. While this may not be a major issue in the development of self contained greenfield systems with few human interactions and accurate existing models, model fidelity is much more problematic in the development of new features and/or capabilities for systems which depend on existing infrastructure or systems that have a high degree of human interactivity or legacy. In fact, it might well be that most system developments have to contend with legacy system modeling issues. As a result, to get the full benefits of model-centric engineering, the existing, active system needs to be understood and characterized.
This project involves research in the following potential areas:
- Documenting the potential uses and benefits of these techniques based on a literature search of applicable case studies.
- Analyzing an existing system and determining how system monitoring and feedback can be used to facilitate future system improvements.
- Analyzing challenges in Model Based Systems Engineering (MBSE) and determining where active system monitoring and characterization can most effectively be deployed.
- Wade, J. et al.(20 1 0), "Development of 3-Year Roadmap to Transform the Discipline of Systems Engineering." Final Technical Report -SERC-2009-TR-006, March 2010.
Karacali, B.; Rao, B. (2007), "Network Instrumentation for End-to-End Measurements," Integrated Network Management, 2007. 1M '07. 10th IFIP/IEEE International Symposium on, vol., no., pp.314-322, May 21 2007-Yearly 252007.
- Experienced Systems Engineers with an interest in active system measurement and control.