 |
Architecture and Real-Time Systems (ARTS) Laboratory
|
Researchers in ARTS seek new architectures for high performance parallel and distributed computing systems with a particular emphasis on embedded real-time systems and fault-tolerance in parallel systems.
Currently we are developing a unified approach to tolerate system and network failures in distributed systems. Our techniques will be characterized by quick fault detection and efficient fault recovery without the use of expensive hardware. These techniques will be tested and validated on the RAPIDS framework. Please contact Vijay Lakamraju to obtain the latest version of the RAPIDS toolset.
For a brief overview of fault-tolerance related research in the ARTS Lab, check out these presentations: NetFT , ALFT.
In addition, extensive research is being carried out in system-level power aware computing.
In the past, the ARTS lab has developed new measures for system and network reliability and defining methods and algorithms by which these measures may be calculated. The TRIDENT software tool is being developed to allow system designers to make use of these new measures quickly and easily in order to better evaluate real-time systems.
Work has also been done in developing algorithms for dynamic recovery (from failure), adaptive resource allocation, scheduling and rescheduling, and load balancing. The RAPIDS simulator is a tool for verifying these and other real-time system algorithms.
We would like to thank our Sponsors:
- National Science Foundation
- Jet Propulsion Laboratory - NASA
- Defense Advanced Research Projects Agency
Professors:
Past Post-Doctoral Researcher:
Current Students:
Past Students:
- Yongkui Han
- Yizheng Zhou
- Osman S. Unsal
- Akash Goel
- Hemant Kumar
- Jayakrishnan Nair
- Meghna Misra
- Diganta Roychowdhury
- Janhavi Rajagopal
- Eric C. Ciocca
- Steve Morin
- Sivakumar Murgesan
- Joshua W. Haines
- Luis R. V. Rivera
- Miriam Allalouf
- Jing Chang
- Gopinath Durairaj
- Kiran Toutireddy
- Kai Yu