Colloquium Series Spring 2004

Chris Harding

Here is the abstract in pdf format.

The modeling (or interpretation) of geological structures from raw data is central to many geoscientific tasks in academia or industry. The advent of immersive displays and other VR technologies in the last years have made it easier for interpreters to interact directly with geoscientific data and ultimately to translate the models developed in their ¡°mind¡¯s eye¡± into a computer model. While geoscience data is now routinely visualized in 3D environments (for example in virtual theaters used by all major oil companies), the addition of touch and sound is still a research topic. The talk focuses on a prototype desktop virtual environment for the investigation of surface meshes - not only via 3D stereo but also with a point-haptic device called the Phantom and with a real-time audio stream. The haptic device is used to digitize line features, such as fault lines, directly on the surface while simultaneously feeling its morphology. The audio stream is a MIDI-interpretation of a (possibly secondary) data attribute such as terrain slope or gravity and useful in augmenting the interaction with the surface data.. The prototype system was primarily used to interpret faults surfaces and other tectonic structures along Mid-ocean ridge in the North Atlantic.

Fei Xie

This talk presents my research on integration of model checking into two emerging software development processes: model-driven development (MDD) and component-based development (CBD). Testing has been the dominant method for validation of software systems. As these systems become complex, conventional testing methods have become inadequate. Model checking is a powerful formal verification method. It supports systematic exploration of all states or execution paths of the system being verified. However, there are two major challenges in practical application of model checking to software systems: (1) applicability - the representations usable by model checkers and the widely used software representations are significantly different, and

(2) computational complexity of model checking - the number of possible states and execution paths in a real-world software system can be extremely large. My research successfully addressed these challenges in integration of model checking into MDD and CBD. The methods and tools developed in this research have enabled verification of real-world software systems of significant size and have led to significant reductions in time and memory usages for model checking these systems. This talk will sketch integration of model checking into MDD, but will concentrate on integration of model checking into CBD.

Qun Li

A sensor network consists of a collection of sensors distributed over an area that form an ad-hoc network. Each sensor is equipped with some limited memory and processing capabilities, multiple sensing modalities, and communication capabilities. Mobile sensor networks are sensor networks in which nodes can move under their own control or under the control of the environment. A key difference between a mobile sensor network and a static sensor network is how information is distributed over the network. Under static nodes, a new task or data can be flooded across the network in a very predictable way. Under mobility this kind of flooding is more complex.

In this talk, I will summarize our effort in developing distributed algorithms that explore the synergy between mobility and communication. Communication can be enhanced with mobility because mobility allows nodes to carry and relay information. Mobility can be enhanced by communication because it is cheaper to transmit sensed data from a remote location than to actuate the sensor to that location. I will briefly describe an infrastructure for mobile sensor networks that relies on new algorithms for global clock synchronization and power-aware communication. Then I will briefly describe algorithms that show how mobility can enhance communication. Finally, I will focus on a navigation application that demonstrates how communication can assist mobility. In the navigation problem a sensor network assists with the creation of adaptive maps and safe paths that can be communicated incrementally to a moving node, to guide the node to a desired spot. I will describe the protocols, their analysis and the performance results from a physical implementation on a 50 node Mote sensor network.

Christian Luer

Based on a vision that, in the future, applications will be flexibly built out of small-grained components, we argue that current technologies do not adequately support component deployment in such a setting. Specifically, current technologies realize deployment processes where most decisions are made by the application manufacturer. When using small-grained components, however, the component user needs to have more control over the deployment process; user- centric deployment is needed. In this talk, we describe our efforts at providing user-centric deployment. We present JPloy, a prototypical tool that gives users more control about the configuration of installed Java components. JPloy extends the Java class loader with support for connecting and adapting components. It allows users to specify modular configurations, and can enforce connections and adaptations over existing components. Name space or versioning conflicts among components can be elegantly resolved, and the tool assists in the generation of glue code.

Guang Song

In this work, we investigate a novel approach for studying protein folding that has evolved from robotics motion planning techniques called probabilistic roadmap methods (PRMs). Our focus is to study issues related to the folding process, such as the formation of secondary and tertiary structure, assuming we know the native fold. A feature of our PRM-based framework is that the large sets of folding pathways in the roadmaps it produces, in a few hours on a desktop PC, provide global information about the protein's energy landscape. This is an advantage over other simulation methods such as molecular dynamics or Monte Carlo methods which require more computation and produce only a single trajectory in each run. We present validation results for several small to moderate sized proteins (60-100 residues) comparing the secondary structure formation order on paths extracted from our roadmaps with known hydrogen exchange experimental results. In addition, we present a case study showing our technique captures known folding differences between the structurally similar proteins G and L. Our work has generated significant interest - it has been reported in Genome Technology (Nov 2001), featured in the NSF CISE quarterly newsletter (3rd Quarter 2002), and one of our protein folding animations appeared on the BBC World News Website (Aug 2002).

Portions of this work are joint with my advisor Nancy Amato, Dr. Marty Scholtz (Biochemistry, Texas A&M), Dr. Ken Dill (Pharmaceutical Chemistry, UCSF), and Shawna Thomas, a graduate student in our group. More information regarding our work, including movies, can be found at http://parasol.tamu.edu/.

Armin R. Mikler

The domains of Computational Biology and Bio-Informatics have harnessed the computational power of high-performance computing to tackle the computational complexity of tasks associated with genomics research. This effort undoubtedly deserves its own field of research; however, there are other domains in which scientists can take advantage of recent advances in Computer Science and Scientific Computing. One such domain is Epidemiology with its multiplicity of sub-domains, including Field Epidemiology, Epidemiological Genomics, Infectious Disease Epidemiology, Social and Behavioral Epidemiology, and Surveillance.

Epidemiologists are often faced with the challenge of dealing with data that are sparse, widely distributed, and incomplete (often due to confidentiality and other constraints). This may result in conflicting information that confound or disguise the evidence leading to wrong conclusions. Today, the role of epidemiologists has become even more pronounced as the significance of Public Health has been recognized. To meet the increasing demands, the field of Epidemiology is in need of specific computational tools that would enable the professionals to respond promptly and accurately in their efforts to control and contain disease outbreaks. Increased globalization, highly mobile populations, and possible exposure to infectious diseases pose new public health threats. It is vital to develop new tools that take advantage of today¡¯s communication and computing infrastructures. Computational models for the simulation of global disease dynamics are required to facilitate adequate what-if analyses. This necessitates adapting fundamental Computer Science concepts to the specific problems in Epidemiology.

This talk focuses on the design of a simulation infrastructure that facilitates the study of communicable diseases in different spatial domains. After motivating the need for the field of computational epidemiology, this presentation will highlight various issues that must be considered when developing an experimental simulation environment. Depending on the level of spatial resolution of the model, different computing paradigms may be applicable. Two of these computational paradigms are: Multi-Agent Systems and Stochastic Cellular Automata. Their role in the design of a disease outbreak simulator will be discussed in detail. Recent advances in high-performance computing facilitate the execution of complex simulation models. The availability of data from geographic information systems (GIS), new visualization techniques (i.e. virtual reality) and high-performance computing paradigms, such as cluster and grid computing will greatly contribute to the development of tools that facilitate the work of today¡¯s epidemiologists. The talk concludes with the discussion on a proposed design of a simulation environment that can take advantage of cluster and grid computing infrastructures.

Erik Hilsdale

AspectJ is a seamless aspect-oriented extension to Java[tm]. It can be used to cleanly modularize the crosscutting structure of concerns such as exception handling, multi-object protocols, synchronization, performance optimizations, and resource sharing. When implemented in a non-aspect-oriented fashion, the code for these concerns typically becomes spread out across entire programs. AspectJ controls such code-tangling and makes the underlying concerns more apparent, making programs easier to develop and maintain. This talk will introduce Aspect-oriented programming and show how to use AspectJ to implement crosscutting concerns in a concise, modular way. AspectJ is freely available at http://eclipse.org/aspectj

Benjamin Wah

The solution of mixed-integer constrained optimization problems with highly nonlinear objectives and/or constraints has long been considered challenging. These problems are abundant in many engineering applications, including IC design automation, signal processing, planning and scheduling, computer aided manufacturing, robotics, data mining, and neural-network learning. In this talk, we present a new theory based on the extended saddle point condition that is proved to be necessary and sufficient for characterizing constrained local minima in these problems. An important property of the condition is that it can be partitioned into multiple necessary conditions that collectively are necessary and sufficient. This allows complex problems to be partitioned into simpler subproblems that can be solved independently, before resolving violated global constraints that relate the subproblems. We illustrate the application of the theory in temporal planning and the design of a new planner for space applications. Partitioning is attractive in temporal planning problems because many of their constraints and objectives are related to activities with temporal locality. We also discuss other related applications in machine learning, signal processing, and nonlinear optimization.

Arnab Ray

The last two decades has seen an increased reliance on distributed computer-controlled systems in virtually all spheres of our daily lives. That is why there is an urgent need for computer-based systems to be designed in a bug-free fashion. While there have been rapid advancements and intense research activity in the formal methods community to discover tools and techniques for mathematically certifying designs, their adoption in the software engineering world has been slow and guarded. Developers complain about the unintuitiveness and steep learning curve for available tools, the enormous memory-space requirements for even moderately complex models (due to the infamous "state-space" explosion problem), and the poor returns obtained on the time spent in creating detailed models. My research handles each of these problems by proposing intuitive design notations, providing methodologies for "state-space-efficient" verification and supplying compositional semantics (which enables mathematically precise component-wise reasoning) to traditionally non-compositional formalisms. In my talk, I shall broadly outline my research goals pursued in the course of my PhD work and concentrate on Architectural Interaction Diagrams (AID): a hierarchical compositional architecture description language that is unique in its ability to provide a parameterized notion of communication.

Benyuan Liu

Mobile and wireless networks have undergone tremendous growth in the past few years. Many new applications and services have started to emerge. The inherent characteristics of mobile and wireless networks, such as limited bandwidth and battery power, dynamic topology, error-prone channels, pose great challenges to the performance of wireless networks. In this talk, I will present my work on the design and performance modeling of wireless networks. First, I will present the performance modeling for the coverage of large-scale sensor networks. Coverage of a sensor network represents the quality of service (surveillance) that it can provide, for example, how well a region of interest is monitored by sensors, how effective a sensor network is in detecting intruding objects. I will define several fundamental coverage measures, present the characterizations of these coverage measures for different network scenarios, and discuss the implications to network design. The second part of my talk addresses the data transport capacity of wireless ad hoc networks, which is a fundamental measure of the data transmission rate that a network can sustain. At the end of the talk, I will summarize my current work and outline the directions for my future research.

Beichuan Zhang

A ubiquitous and efficient multicast mechanism is essential to the success of large-scale group communication applications. Traditional IP Multicast is implemented at IP layer on routers, which achieves great transmission efficiency but also poses significant challenges to deployment. End-host Multicast moves the functionality to transport or application layer on end hosts in order to have minimal deployment barrier, but incurs performance penalty due to duplicate packet transmission and inefficient routing. Between them, there are a spectrum of other schemes, such as Source-Specific Multicast and multicast server overlay, trying to make better trade-off between system performance and deployability. Nevertheless, the lack of ubiquitous multicast support on the current Internet hinders the development of multicast applications, which in turn reduces incentives for multicast deployment.

In this dissertation research, we designed and implemented the Universal Multicast (UM) framework to provide ubiquitous multicast delivery service on the Internet. It is conceivable that the deployment of multicast will take several stages. Unlike previous works, we do not implement multicast functionality at a fixed place in the network for a particular deployment stage; instead, UM is a general framework within which applications can be provided with IP Multicast delivery immediately, and native multicast support can gradually spread from the network edges into the network core. Therefore the trade-off between system performance and deployability is not fixed by the design, but can evolve from time to time driven by user demand for the service.

Our approach is to build an end-host overlay to connect IP-Multicast "islands" by adaptive unicast tunnels, and use native IP Multicast within each island. The design consists of three major components: Host Multicast Tree Protocol (HMTP) for inter-island routing, Host Group Management Protocol (HGMP) for intra-island management, and a user Agent implementing the functionalities on a host. UM takes full advantage of network support where available and utilizes End-host Multicast where needed. It is fully distributed, self-organized, and independent from underlying routing protocols. We evaluated the system performance through simulation, implemented a prototype and tested with existing multicast applications.

Yanxin Wang

Anomaly intrusion detection normally has high false alarm rates, and a high volume of false alarms will prevent system administrators from identifying the real attacks. Machine learning methods provide an effective way to decrease the false alarm rate and improve the detection rate of anomaly intrusion detection. In this research, we propose a novel approach using kernel methods and Support Vector Machine (SVM) for improving anomaly intrusion detectors' accuracy. We also propose an anomaly detection approach, using STIDE kernel and Markov Chain kernel based one class SVM, which does not need labeled training data.

To further increase the detection rate and lower the false alarm rate, an approach of integrating specification based intrusion detection with anomaly intrusion detection is also proposed. Using specification based method as a filter and feature selection tool, the accuracy rate of anomaly detection intrusion can be further improved.

This research also establish a platform which generates automatically both misuse and anomaly intrusion detection software agents. In our method, a SFT representing an intrusion is automatically converted to a Colored Petri Net (CPNs) representing an intrusion detection template, subsequently, the CPN is compiled into code for misuse intrusion detection software agents using a compiler and dynamically loaded and launched for misuse intrusion detection. On the other hand, a model representing a normal profile is automatically generated from training data, subsequently, an anomaly intrusion detection agent which carries this model is generated and launched for anomaly intrusion detection. By engaging both misuse and anomaly intrusion detection agents, our system can detect known attacks as well as novel unknown attacks.

Doina Caragea

Recent advances in computing, communications, and digital storage technologies, together with development of high throughput data acquisition technologies have made it possible to gather and store large volumes of data in digital form. These developments have resulted in unprecedented opportunities for large-scale data-driven knowledge acquisition with the potential for fundamental gains in scientific understanding (e.g., characterization of macromolecular structure-function relationships in biology) in many data-rich domains. In such applications, the data sources of interest are typically physically distributed, semantically heterogeneous and autonomously owned and operated, which makes it impossible to use traditional machine learning algorithms for knowledge acquisition.

However, we observe that most of the learning algorithms use only certain statistics computed from data in the process of generating the hypothesis that they output and we use this observation to design a general strategy for transforming traditional algorithms for learning from data into algorithms for learning from distributed data. The resulting algorithms are provably exact in that the classifiers produced by them are identical to those obtained by the corresponding algorithms in the centralized setting (i.e., when all of the data is available in a central location) and they compare favorably to their centralized counterparts in terms of time and communication complexity.

To deal with the semantical heterogeneity problem, we introduce ontology-extended data sources and define a user perspective consisting of an ontology and a set of interoperation constraints between data source ontologies and the user ontology. We show how these constraints can be used to define mappings and conversion functions needed to answer statistical queries from semantically heterogeneous data viewed from a certain user perspective. That is further used to extend our approach for learning from distributed data into a theoretically sound approach to learning from semantically heterogeneous data.

The work described above contributed to the design and implementation of AirlDM, a collection of data source independent machine learning algorithms through the means of sufficient statistics and data source wrappers, and to the design of INDUS, a federated, query-centric system for knowledge acquisition from distributed, semantically heterogeneous, autonomous data sources.

Chris Harding received a Masters Degree in 1993 in Geology from the Free University in Berlin, Germany, where he specialized in mathematical geology such as geostatistics and Geographic Information Systems (GIS). He worked as a programmer for mine planning and geotechnical applications and received a Ph.D. in Geology from the University of Houston in 2001 where he worked on multi- sensory Virtual Environments for petroleum exploration data. He worked for ExxonMobil and Shell on the application of audio and force-feedback (haptics) to improve seismic interpretation. He is currently employed as Assistant professor at the geoscience department at Iowa State University and is also part of its Human-Computer-interaction (HCI) program which is centered at the Virtual Reality Applications Center (VRAC).

BACK

Fei Xie

Fei Xie is a Ph.D Candidate at the University of Texas at Austin, working with Prof. James C. Browne. His research interests are primarily in software engineering, especially software safety, security, and reliability. He is particularly interested in developing techniques and tools that are based on formal methods and enable design and development of safe, secure, and reliable software systems. He is also interested in hardware/software co-design and co- verification.

Visit Fei Xie's homepage here.

BACK

Qun Li

Qun Li is currently a PhD student in the Computer Science Department at Dartmouth College. He received a B.S. degree from Changsha Institute of Technology and a master degree from Southeast University, both in computer science. His research interests include sensor networks, mobile networks, and wireless networks. He has been working on mobility, power conservation, clock synchronization, information diffusion, and security. His work has involved a wide range of categories, including theoretical algorithm design and analysis, simulation, real experimentation, and measurement.

Visit Qun Li's homepage here.

BACK

Christian Luer

Chris L¨¹er is a Ph. D. candidate in the School of Information and Computer Science at the University of California, Irvine. He received a joint B.S. and M.S. degree in Computer Science from the University of Dortmund in 1999. Chris's research is centered around component-based software development.

Visit Christian Luer's homepage here.

BACK

Guang Song

Guang Song received his Ph.D. degree in Computer Science from Texas A&M University in 2003. He is currently a post doctorate research associate at L.H. Baker Center for Bioinformatics and Biological Statistics at Iowa State University. His research interests are: computational biology, motion planning, robotics, virtual reality, molecular docking, and quantum computing. More information related to his research is available at http://parasol.tamu.edu/~gsong.

Visit Guang Song's homepage here.

BACK

Armin R. Mikler

Armin R. Mikler received his Diploma in Informatics from Fachhochschule Darmstadt, Germany in 1988. After spending one year as a Fulbright exchange student at Iowa State University (ISU), he joined ISU as a graduate student and received his MS and Ph.D. in Computer Science in 1990 and 1995 respectively. From 1995 to 1997, he worked as a postdoctoral research associate in the Scalable Computing Laboratory at Ames Laboratory, USDOE. In 1997, Dr. Mikler joined the faculty in Computer Science at the University of North Texas (UNT) where he holds the rank of associate professor in Computer Science with joint appointment in the Department of Biological Sciences. His research interests include: Intelligent Network Management, Distributed Coordination of Intelligent Mobile Agents, Distributed Decision Making, Multi-Agent Simulation and Stochastic Cellular Automate applied to Computational Epidemiology. Dr. Mikler has established and is the director of the Network Research Laboratory (NRL) which provides the necessary computational infrastructure to conduct large scale simulations. As a member of the Institute of Applied Science at UNT, he has been conducting collaborative and interdisciplinary research in computational science, specifically in the areas of quantitative analysis of ecological processes and Biocomplexity. Dr. Mikler is an associate editor of the Telecommunication Systems Journal and a member of the ACM and the IEEE Computer Society.

Visit Armin R. Mikler's homepage here.

BACK

Erik Hilsdale

Erik Hilsdale is a researcher at the Palo Alto Research Center and a PhD student at Indiana University. As a PhD student he works in the area of programming languages and compilers. As a member of the AspectJ team he focuses on language design, pedagogy and compiler implemetation. He is an experienced and energetic presenter with a long background with AspectJ.

Visit Erik Hilsdale's homepage here.

BACK

Benjamin Wah

Benjamin W. Wah is currently the Franklin W. Woeltge Professor of Engineering and a Professor in the Department of Electrical and Computer Engineering and the Coordinated Science Laboratory of the University of Illinois at Urbana-Champaign, Urbana, IL. He received his Ph.D. degree in computer science from the University of California, Berkeley, CA, in 1979. Previously, he had served on the faculty of Purdue University (1979-85), as a Program Director at the National Science Foundation (1988-89), as Fujitsu Visiting Chair Professor of Intelligence Engineering, University of Tokyo (1992), and McKay Visiting Professor of Electrical Engineering and Computer Science, University of California, Berkeley (1994). In 1989, he was awarded a University Scholar of the University of Illinois; in 1998, he received the IEEE Computer Society Technical Achievement Award; in 2000, the IEEE Millennium Medal; and in 2003, the Raymond T. Yeh Lifetime Achievement Award from the Society for Design and Process Science.

Wah's current research interests are in the areas of nonlinear search and optimization, multimedia signal processing, and computer networks.

Wah was the Editor-in-Chief of the IEEE Transactions on Knowledge and Data Engineering between 1993 and 1996, and is the Honorary Editor-in-Chief of Knowledge and Information Systems. He currently serves on the editorial boards of Information Sciences, International Journal on Artificial Intelligence Tools, Journal of VLSI Signal Processing, World Wide Web, and Neural Processing Letters. He had chaired a number of international conferences and was the International Program Committee Chair of the IFIP World Congress in 2000. He has served the IEEE Computer Society in various capacities; among them include Vice President for Publications (1998 and 1999) and President (2001). He is a Fellow of the IEEE and the Society for Design and Process Science.

Visit Benjamin Wah's homepage here.

BACK

Arnab Ray

Arnab Ray is a PhD candidate in the Department of Computer Science at the State University of New York at Stony Brook advised by Professor Rance Cleaveland. Arnab received his MS degree from the State University of New York at Stony Brook in 2001 and his BE degree in Computer Science and Engineering from Jadavpur University, India in 1999. His research interests include software engineering (specifically software architecture), formal methods, specification and verification methodologies and computer security.

BACK

Benyuan Liu

Benyuan Liu received the Ph.D. degree in computer science from the University of Massachusetts at Amherst in 2003, and he is currently an assistant professor in the Computer Science Department at the City College of New York. He also received a B.S. degree from the University of Science and Technology of China and M.S. degree from Yale University, both in physics. His research interests are in the area of mobile and wireless networking, sensor networks, Internet technologies and applications.

BACK

Beichuan Zhang

Beichuan Zhang received his PhD in computer science from UCLA in 2003. He got his M.S. from UCLA and B.S. from Beijing University, China. His research interests include Internet routing dynamics, multicast, application-level overlays, and network measurement and performance. He is now with USC/ISI as a Postdoc researcher.

BACK

Yanxin Wang

Yanxin Wang is a Ph.D. candidate in Computer Science at Iowa State University (ISU). Currently, she is working at Microsoft as a software engineer in the Distributed Systems Group. Yanxin received her B.S. degree from Beijing University of Aeronautics and Astronautics in 1993, M.S. degree in Computer Science at Tsinghua University in 1996 and also M.S. in Computer Science at ISU in 2002. Her research interests include computer intrusion detection & countermeasures, security in operating systems and networks, formal methods and machine learning methods for intrusion detection.

Visit Yanxin Wang's homepage here.

BACK

Doina Caragea

Doina Caragea is a Ph.D. candidate in Computer Science Department at Iowa State University. Doina received her B.S. degree in 1996 and her M.S. degree in 1997 from University of Bucharest, Romania. During her graduate studies at ISU, Doina was awarded an IBM fellowship two years in a row. Her research interests include machine learning and data mining, information integration, visualization and computational biology.

Visit Doina Caragea's homepage here.

BACK

---

Robert Stewart Distinguished Lecture

The Robert Stewart Distinguished Lecture in Computer Science is made possible by the generous contribution of Dr. Long Nguyen, who got his doctorate from the Computer Science Department at Iowa State University in 1975. It is in honor of his mentor, Dr. Robert Stewart, Professor Emeritus and the first Department Chair of Computer Science at Iowa State University.

---

Archives

• Fall 2003
• Spring 2003
• Fall 2002
• Spring 2002
• Fall 2001
• Spring 2001
• Fall 2000
• Spring 2000
• Fall 1999
• Spring 1999
• Fall 1998
• Summer 1998
• Spring 1998
• Fall 1997
• Spring 1997

(Top)

---

Contacts

Thank you for visiting this page. Please send your suggestions and comments to one of us in the Computer Science colloquium committee.

• Soma Chaudhuri . chaudhur@cs.iastate.edu
• Dalei Li. daleili@cs.iastate.edu

(Top)

---