The development of distributed real-time embedded systems requires adaptive control technologies that are capable of assuring stable and robust operation in the presence of dynamic and unpredictable changes in the workloads and resource capabilities. Such changes can be smooth variations caused by the dynamic behavior of existing tasks or abrupt changes caused by the occurrence of aperiodic events. An instance of such changes occurs, for example, when the workload causes the bottleneck resource of a distributed system to switch between CPU and network bandwidth. The objective of the control framework is to manage the system resources adaptively so that the system can robustly react to such changes. Control theory provides the state of the art in the design of adaptive controllers and can be applied to systems that exhibit smooth variations and abrupt changes.

Controller design requires the development of analytical models using differential/difference equations that describe resource utilization and contention. We use simulation, analytical, and experimental techniques to validate our models. We also design software components that realize the control capabilities and integrate them in open middleware architectures. The development of control technologies allows the design of provably correct distributed adaptive resource management methods and will push the performance of embedded systems. Our approach also contributes to the evolution of such systems by providing tools for evaluating the system design using analytical models of aggregate behavior.

Personnel

• Radhika Tekumalla, MS in Computer Science, December 2005

Publications

• X. Wang, Y. Chen, C. Lu, and X. Koutsoukos, FC-ORB: A Robust Distributed Real-time Embedded Middleware with End-to-End Utilization Control, Elsevier Journal of Systems and Software, Special Issue on Dynamic Resource Management in Distributed Real-Time Systems. To appear. (This is the extended version of the RTSS'05 paper on FC-ORB).
  
  
• Chenyang Lu, Xiaorui Wang, and Xenofon Koutsoukos, "Feedback Utilization Control in Distributed Real-Time Systems with End-to-End Tasks". IEEE Transactions of Parallel and Distributed Systems, 16(6), 550-561, June 2005. (This is the extended version of the ICDCS'04 paper on EUCON.)
  
  
• Nishanth Shankaran, Xenofon Koutsoukos, Douglas Schmidt, Yuan Xue, and Chenyang Lu, "Hierarchical Control of Multiple Resources in Distributed Real-Time and Embedded Systems", 18th Euromicro Conference on Real-Time Systems (ECRTS 06). July, 2006.
  
  
• Xiaorui Wang, Dong Jia, Chenyang Lu, and Xenofon Koutsoukos "Decentralized Utilization Control in Distributed Real-Time Systems". 26th IEEE Real-Time Systems Symposium (RTSS) . December 2005.
  
  
• Xiaorui Wang, Chenyang Lu, and Xenofon Koutsoukos "Enhancing the Robustness of Distributed Real-Time Middleware via End-to-End Utilization Control". 26th IEEE Real-Time Systems Symposium (RTSS) . December 2005.
  
  
• Nishanth Shankaran, Xenofon Koutsoukos, Douglas C. Schmidt, and Aniruddha Gokhale, "Evaluating Adaptive Resource Management for Distributed Real-Time Embedded Systems". The 4th Workshop on Adaptive and Reflective Middleware (ARMS2005). November 2005.
  
  
• X. Koutsoukos, R. Tekumalla, B. Natarajan, and C. Lu, "Hybrid Supervisory Control of Real-Time Systems", 11th Real-Time and Embedded Technology and Applications Symposium - RTAS 2005, March 2005, San Francisco, CA.
  
  
• Chenyang Lu, Xiaorui Wang, and Xenofon Koutsoukos, "End-to-End Utilization Control in Distributed Real-Time Systems", International Conference on Distributed Computing Systems (ICDCS 2004), Tokyo, Japan, March 2004.

Updated list of publications and preprints can be found at http://www.vuse.vanderbilt.edu/~koutsoxd/www/publications.html

For additional information or if you are interested for graduate student research positions, contact me at: Xenofon.Koutsoukos@vanderbilt.edu