Overview
Yanyong's research focuses on bridging the gap between large-scale parallel and distributed
systems and emerging applications that run on these systems. In addition to the performance
requirements of the applications, Yanyong also emphasizes other requirements such as
reliability, robustness, trustworthiness, and availability. Yanyong tries to address this
challenge at the OS/middleware layer.
Specifically, she is examining
two broad types of platforms: parallel systems and networked sensor systems.
Over the past 5 years, Yanyong has built a research program, called PROSE (Providing Robustness to Large-Scale Networked Systems).
Large-Scale Networked Sensor Systems
For many sensor applications, the most important service a sensor network provides
is the continuous monitoring
of the sensor field, which involves continuously collecting data from physical environment and continuously
delivering data to the application. At the same time,
sensor systems face many new challenges: sensor nodes have severe energy constraints;
network operations can be disrupted by natural reasons such as node failures or
congestion as well as malicicous attacks such as radio interference attacks.
The objective of this work, therefore, is to deliver the required service under varying
network conditions.
Large-Scale Parallel Systems and Clusters
High-performance systems mainly serve large-scale, long-running
scientific and engineering applications, which require maximized
system utilization and minimized job response times. As both the applicattions
and the underlying platforms scale to thousands of processors,
failures become a norm rather than an exception. As a result,
instead of preventing failure occurrences, we seek to
to provide "graceful" operations under faulty
conditions by mitigating the adverse impact of failures.
Funding:
- "TRIESTE: A Trusted Radio Infrastructure for Enforcing Spectrum Etiquettes," NSF, 09/07-08/08
- "A Roll-Call System for Asset Tracking," NSF, 06-07
- "Pro-active Runtime Health Enhancement of Large-scale Parallel Systems Using PROGNOSIS," NSF, 09/06 - 08/09
- "CAREER:PROSE: Providing Robustness in Systems of Embedded Sensors," NSF, 07/06 - 06/11
- "PROGNOSIS to Enhance the Runtime Health of Large Scale Parallel Systems," NSF, 08/05 - 07/06
- "PARIS: A Framework for Privacy Augmented Relaying of Information from Sensors," NSF, 09/04 - 09/07
- "ORBIT: Open-Access Research Testbed for Next-Generation Wireless Networks," NSF, 09/03 - 09/07
Currently Supervising the following wonderful Students:
Yinglung Liang (expected graduation: 09/07)
Shengchao Yu (expected graduation: 12/2007)
Pandurang Kamat (expected graduation: 09/07, co-advise with Prof. Trappe)
Gautam Bhanage
Lijun Dong
Tingting Sun
Benard Firner
Students who have graduated:
Ke Ma
Wenyuan Xu (Currently an Assistant Professor at the University of South Carolina, co-advise with Prof. Trappe)
Jaewon Kang (Currently employed at Telcordia Research, co-advise with Prof. Nath)
Celal Ozturk
Back
to Home