ISS Seminar Series - Spring 2009

Thursday April 9th, 2009 11:30 AM 
Dr. Randolph Moses 
Topic: Statistical signal processing approached for information extraction in sensor systems. 

Dr. Moses is a professor in the OSU Department of Electrical and Computer Engineering. His research interests include: Stochastic digital signal processing; spectral estimation; time series analysis; parameter estimation; statistical properties of algorithms; array signal processing. Applications to automatic target recognition and sensor networks. Most of this research is conducted with colleagues and students in the Information Processing Systems (IPS) Laboratory.

Tuesday April 14th, 2009 11:30 AM 
Junshui Ma (Merck) Host: Ashok Krishnamurthy (Electrical and Computer Engineering) 
Topic: Brain Signals for Drug R&D. 
This presentation mainly covers the following three topics

1. A brief introduction to Pharmaceutical industry and its drug Research & Development (R&D)
2. An overview of the application of brain signals, especially EEG, to drug development.
3. A case study : applying a new independence criterion to explore the relationship between different EEG related measurements.

Tuesday April 21st, 2009 11:30 AM 
Dr. Steven Bibyk  (Slides) (Lecture and Slides)

Topic: Wireless sensor network research and curriculum change. 
Electronics research for wireless sensor networks, national contest sponsored by TI that has been part of senior projects that uses their wireless sensor platforms. 

Dr. Bibyk is an associate professor in the OSU Department of Electrical and Computer Engineering. His research interests include: Electronics, communication systems, parallel computation for signal processing, solid-state device theory and measurements.

Friday May 1st, 2009 11:30 AM 
Sanjay Shakkottai (U. of Texas at Austin, Electrical and Computer Engineering) 
Host: Ness Shroff (Electrical and Computer Engineering) 
Topic: Scheduling over wireless networks with delayed network state information. 
We study the problem of scheduling/routing over a mobile ad hoc wireless network (MANET). Typically, scheduling/routing algorithms for MANETs that are based on the well-known back-pressure algorithm by Tassiulas and Ephremides assume instantaneous channel, topology and queue-length knowledge (together referred to as Network State Information - NSI) in order to make decisions. Often however, only partial and/or delayed NSI is available and this information is not consistent across nodes (i.e., different nodes in the network have differing information). In this talk, we present recent results on characterizing the throughput-region with partial/delayed NSI and describe associated distributed algorithms that are shown to be throughput-optimal. (Based on joint work with Lei Ying, ISU) 

Dr. Shakkottai received his Ph.D. from the ECE Department at the University of Illinois at Urbana-Champaign in 2002. He is with The University of Texas at Austin, where he is currently an Associate Professor and the Engineering Foundation Endowed Faculty Fellow in the Department of Electrical and Computer Engineering. He received the NSF CAREER award in 2004. His research interests include wireless and sensor networks, stochastic processes and queuing theory.

Tuesday May 5th, 2009 11:30 AM 
Dr. Lee Potter  (Slides)
Topic: Inverse problems for sensor signal processing.
The presentation aims to describe goals, challenges, and techniques in sensor signal processing, with applications to radar imaging and electron paramagnetic magnetic resonance.  For the ISS forum, the brief overview will conclude with a discussion of conjectured directions. 

Dr. Potter is an Associate Professor in the OSU Department of Electrical and Computer Engineering. He recieved his Ph.D. in 1990 at the University of Illinois at Urbana-Champaign. His research interests include: Statistical signal processing; inverse problems; detection and estimation; applications to radar imaging and ultra wide-band systems.

Tuesday May 12th, 2009 11:30 AM 
Dr. Prasun Sinha 
Topic: Fair Data Collection in Rechargeable Sensor Networks. 
Various forms of energy, such  as  solar,  wind,  vibration,  and thermal,  can  be  harvested  from  the environment to extend the lifetime of sensor networks. A network based on energy harvesting can  operate perpetually without running out of energy, using the available energy judiciously. Several applications based on  data collection  require high and fair data rate from all nodes in the network. However, dissimilar recharging rates of nodes and  their time-varying  nature  makes  it  a  challenging  problem. We have explored  notions  of  both  maxmin  fairness  and   proportional fairness  in  this  context.   I  will present our recent work on designing distributed solutions for  both  notions  of  fairness. Our  solutions have been implemented on TinyOS and evaluated on a sensor network testbed using emulated recharging rates  based  on the  solar  radiation  measurements  collected  by  the  National Renewable Energy Laboratory.

Dr. Sinha is an Assistant Professor in the OSU Department of Computer Science and Engineering. Prasun Sinha  received  his  PhD  from  University  of  Illinois, Urbana-Champaign  in  2001,  MS from Michigan State University in 1997, and B. Tech. from IIT Delhi in 1995.   He  worked  at  Bell Labs,  Lucent Technologies from 2001 to 2003. Since 2003 he is an Assistant Professor in the Department  of  Computer  Science  and Engineering at Ohio State University. His research focuses on the design of network protocols  for  sensor  networks  and  wireless networks.   He   has  won  several  awards  including  Ray  Ozzie Fellowship (UIUC, 2000), Mavis Memorial Scholarship (UIUC, 1999), and  Distinguished  Academic  Achievement  Award  (MSU, 1997). He received the prestigious NSF CAREER award  in  2006.

Tuesday May 19th, 2009 11:30 AM 
Dr. Gagan Agrawal  (Slides)
Topic: Data management issues on (sensor-based) scientific data and sensing-event-driven real-time distributed processing issues.
This talk will describe two computer science projects that have been driven by sensing scenarios.  The goal of the first project is to provide effective utilization of massive datasets being collected  by sensors.  We have developed  an autonomous scientific workflow system that enables high-level, natural language based, queries over low-level data sets. Our technique involves a combination of natural language processing, metadata indexing, and a semantically-aware workflow composition engine which dynamically constructs workflows for answering queries based on service and data availability. This system has been evaluated with many query scenarios and datasets from the Mapping  and GIS Lab at Ohio State.

The motivation for the second project is as follows. In sensing scenarios, a timely response to an important event may be  needed. Often such response can require significant computation and possibly communication, and it can be very challenging to complete it within the time-frame the response is needed.  At the same time,  there could be application-specific flexibility in the computation that may be desired. We have  designed, and  implemented   a middleware that can support such applications. The middleware enables  the time-critical event handling to achieve the maximum benefit, as per the user-defined  benefit function, while satisfying the time constraint. 

(Joint work with David Chiu, Qian Zhu, Ron Li, Keith Bedford).

Dr. Agrawal is a Professor in the OSU Department of Computer Science and Engineering. His research interests include: Parallel and Distributed Systems,  Data Mining, Compiler and Middleware Systems, and Grid/Cloud Computing. He has extensive published in these areas and his work has been supported by numerous NSF grants.  He obtained his B. Tech degree from IIT Kanpur, and MS and PhD degrees from University of  Maryland, College Park. He received an NSF CAREER award in 1998.

Friday May 29th, 2009 9:30 AM 
Dr. John Makhoul Host: Eric Fosler-Lussier (Computer Science and Engineering) 
Topic: A Model-Based Approach to Speech and Language Processing.

The last two or three decades have witnessed significant advances in speech and language technologies, thanks to a mathematically rigorous, model-based approach.  In this approach, statistical models, specified by experts, are trained automatically from annotated data where only domain knowledge is needed to perform the annotations and can usually be supplied by non-expert labor.  The resulting systems have achieved state-of-the-art performance, are highly robust in the face of degraded input, and have proven to be language-independent, requiring only annotated training data from a new language.  Best results are obtained when the model designs are properly informed by structural and linguistic knowledge of the problem at hand.

This talk will present model-based approaches to three technologies: speech recognition, optical character recognition, and machine translation.  In each case, we show how the model-approach has been used to advance the state of the art.  The presentation will end with a demonstration of a real-time broadcast monitoring system where live broadcasts in Arabic and Mandarin Chinese are transcribed automatically and translated into English using off-the-shelf systems.

Dr. Makhoul is one of the world's leading experts on speech and signal processing. He has been with BBN Technologies since 1970, working on various aspects of speech and language processing, including speech coding, speech synthesis, speech recognition, speaker identification and verification, artificial neural networks, digital signal processing, optical character recognition, language understanding, speech-to-speech translation, and human-machine interaction using voice. In addition, Dr. Makhoul is Adjunct Professor at Northeastern University where he supervises students doing their Ph.D. work at BBN in the area of speech and language processing.

Dr. Makhoul is a Fellow of the IEEE and a Fellow of the Acoustical Society of America. His 1975 IEEE Proceedings paper on linear prediction was named a "Citation Classic" by the Institute for Scientific Information. His honors include the 1978 Senior
Award, the 1982 Technical Achievement Award, the 1988 Society Award of the IEEE Signal Processing Society, the IEEE Third Millennium Medal, and most recently, the recipient of the IEEE 2009 James L. Flanagan Speech & Audio Processing award "for pioneering contributions to speech modeling."

He is also an alumnus of The Ohio State University, having graduated with an MS from Ohio State in the Electrical Engineering department, and subsequently he received his Ph.D. from the Massachusetts Institute of Technology.

Friday May 29th, 2009 2:30 PM 
Dr. Yannis Paschalidis Host: Ness Shroff (Electrical and Computer Engineering) (Slides)
Topic: A New Statistical Localization Framework for Wireless Sensor Networks

We consider a class of so called "localization" problems generally dealing with sensor networks attempting to determine the physical location of their nodes. A localization functionality can be extremely useful especially indoors where other alternatives like the Global Positioning System are not available. An important application is asset and personnel tracking in large buildings, warehouses, hospitals, and other settings where one needs to track movable valuable assets. We develop a mesurement-based statistical localization approach motivated by the lack of accuracy and robustness exhibited by triangulation approaches. To that end, we discretize the problem by partitioning the coverage area of the sensor network into a set of regions. We seek to identify the region of a sensor based on observations by stationary clusterheads. Observations (e.g., signal strength) are assumed random to account for the complexity and dynamic character of an indoor environment. We pose the localization problem as a composite multi-hypothesis testing problem, develop the requisite theory using tools from information theory, and address the problem of optimally placing clusterheads. We show that localization decisions can be distributed by appropriate in-network processing. We also tackle related problems, including detecting whether a node has moved from its most recent position and formation detection where we seek to identify the formation of a set of nodes (out of discrete set). The latter problem has applications in robot formation detection and body area networks. Time permitting, we will also briefly discuss issues related to efficient data collection, resulting in a throughput maximizing transmission scheduling problem. Our localization approach has been validated in a Boston University testbed yielding promising results and has been implemented in a large warehouse to track a fleet of forklifts.

Yannis Paschalidis is an Associate Professor of Electrical & Computer and of Systems Engineering at  Boston University, a Co-Director of the  Center for Information and Systems Engineering (CISE), and the Academic Director of the  Sensor Network Consortium (SNC) - a industry consortium he spearheaded which currently consists of 14 companies focusing in sensor networks. He completed his graduate education at the Massachusetts Institute of Technology (MIT) receiving an M.S. (1993) and a Ph.D. (1996) degree, both in Electrical Engineering and Computer Science. In September 1996 he joined Boston University where he has been ever since. 

His current research interests lie in the fields systems and control, networking, applied probability, optimization, operations research, computational biology, and bioinformatics.

Tuesday June 2nd, 2009 CANCELED
Dr. Emre Ertin 
Topic: This presentation has been canceled.

Dr. Ertin is an Assistant Research Professor at OSU for ISS and the Dapartment of Electrical and Computer Engineering. His research interests include: Statistical Signal Processing, game theory and machine learning with applications to Synthetic aperture radar, Sensor networks, Software defined radar and Smart biomedical sensors.